Effect of Fire Exposed on the Behavior of Reactive Powder Concrete Columns under Concentric Compression Loading

Abstract

With growing of the structural design requirements for higher strength materials in applications where member size is governing factor in the design especially in construction of columns in tall buildings, reactive powder concrete (RPC) as a king of ultra-high performance fiber reinforced concrete is catching more attention nowadays due to its extraordinary strength and durability properties. In addition to the developments in concrete technology, the fire engineering design of buildings made from this concrete has become essential. Accordingly, the major component of this Thesis aims to contribute new knowledge with respect to the understanding of the fire performance of concentrically loaded reinforced (RPC) columns.

The first part of this research is the experimental investigation of the behavior of reactive powder concrete (RPC) and normal strength concrete (NSC) columns exposed to fire at 2-adjacent sides. Forty-eight reduced scale models of RPC and NSC column specimens had dimensions of (100x100x900 mm) with longitudinal reinforcement of (406mm) were concentrically loaded up to failure at age of 56 days prior to and later on fire heating at (200, 400 and 600°C) temperature levels and for (1-hour), and compared with reference columns (at 25°C) to analyse the influence of some determinant parameters in improving fire resistance of these columns including the effect of using transverse reinforcement of (05mm) with different tie spacing (50, 100, 200 mm, and No ties) for RPC columns and (50 and 200 mm) for NSC columns, and the effect of increasing concrete cover (from 15 to 30 mm). The second part is a three-dimensional finite element (FE) modeling of the reinforced RPC columns in ABAQUS using a four-nodded tetrahedral element capable of coupled thermal displacement analysis of the fire exposure which simulated using a heat flux. Fire-induced spalling is also presented based on pore pressure calculations in concrete. After the validation of the developed (FE) model with the results obtained from experimental work, the validated model was used to conduct a detailed parametric study on the behavior of RPC columns with different fire scenarios including directions of fire exposure and fire durations.

The comparison of the results showed that there is a good agreement between the simulation and test results of the RPC columns in which all the theoretical values of ultimate load of the reference RPC column specimens and those exposed to fire exceed the experimental values by a margin ranging between (4.9-8.3%) with a standard deviation in the range (0.0112­0.0135). The numerical predicted ultimate axial deformations are found to be lower compared to the experimental values with an average experimental to the numerical ratio of (1.0833). The obtained results of this study showed that the load carrying capacity of RPC columns is slightly increased after heating at 200°c. While it is decreased significantly after exposure to fire temperatures of (400 and 600°C). RPC columns lost about (39 to 45%) of their load carrying capacity after heating at 600°c. On the other hand, NSC columns provide a better fire resistance compared with RPC columns in which the reduction was only about (15 to 24%) after heating at the same temperature level. The effect of reducing tie spacing for improving the fire resistance of RPC columns had a slight influence at 400°c fire temperature. While at 600°c, the effect was insignificant. In addition, the effect of increasing concrete cover was nearly negligible and the load carrying capacity of RPC columns with different concrete covers became close to each other at higher fire temperature levels. It is found from the parametric study that the deterioration in the load carrying capacity increased significantly with increasing number of sides of the exposure and fire duration.

Acknowledgements

In The Name Of Allah, the Most Gracious, the

Most Merciful

he path towards this thesis has been circuitous. It is completion is thanks in large part to the special people who challenged, supported, and stuck with me along the way.

Foremost, I want to offer this endeavor to our GOD Almighty who helped and enabled me to complete this research and for the wisdom he bestowed upon.

I would like to express my appreciation and deepest gratitude to thank my thesis advisor Dr. Mohammed Mansour Kadhum of the College of engineering at University of Babylon. The door to Prof. Kadhum office was always open whenever I ran into a trouble spot or had a question about my research or writing. He consistently allowed this thesis to be my own work, but steered me in the right direction whenever he thought I needed it. I have the full luck and honor of being under his supervision, for his continuous encouragement and invaluable guidance throughout this thesis.

Special thanks and gratitude are due to my family for their care, patience and encouragement throughout the research period.

ไร,

Mustafa Siham Abdulraheem (2017)

List of Contents

Abstract

List of Contents

List of Tables

List of Figures

List of Plates

List of Abbreviation

1 INTRODUCTION
1.1 General Overview
1.2 Concrete Structures in Fire
1.3 Column Behavior at Normal Temperature
1.3.1 Slenderness Ratio and Column Failure Mode
1.3.2 Concrete Protective Cover
1.4 Column Behavior under Fire Exposure
1.5 Problem Statement of Reinforced Concrete Columns Exposed to Fire
1.6 Significance of the Present Study
1.7 Obj ectives of the Present Study
1.8 Outline of the Thesis

2 REVIEW OF LITERATURE
2.1 Introduction
2.2 Reactive Powder c oncrete
2.3 Composition of Reactive Powder Concrete
2.4 Column Behavior at Ambient Temperature
2.4.1 Nominal Strength of Reinforced Concrete Column
2.4.2 Confinement of Reinforced Concrete Column
2.4.3 Ductility of Reinforced Concrete Column
2.4.4 Stiffness of Reinforced Concrete Column
2.5 Column Behavior under Fire Exposure
2.5.1 Effect of Fire Exposure on Concrete
2.5.2 Spalling of Concrete under Fire Exposure
2.5.2.1 Mechanisms of Spalling
2.5.3 Physiochemical Transformations of Concrete under Fire Exposure...
2.5.4 Effect of Fire Exposure on Steel Reinforcement
2.5.5 Latent Heating
2.5.6 Plastic Centroid Movement
2.5.7 Increasing of Slenderness Ratio
2.6 Literature Review of the Experimental Studies on the Behavior of Concrete Columns under Fire Exposure
2.6.1 Lie, 1989
2.6.2 Ali et al., 2010
2.6.3 Bikhiet et al, 2014
2.6.4 Seręga, 2015
2.6.5 Bodalal etai, 2017
2.7 Literature Review of the Finite Element Analysis of Fire Exposed Concrete Column Models
2.7.1 Allen and Lie, 1974. Model
2.7.2 Lie and Irwin, 1993. Model
2.7.3 Ali et al, 2010. Model
2.7.4 Raut and Kodur, 2011. Model
2.7.5 Kadhum, 2013. Model
2.7.6 Emberley, 2013. Model
2.7.7 Seręga, 2015. Model
2.7.8 Baloji et al, 2016. Model
2.8 Case Study (Full-Scale Fire Resistance Test)
2.9 Conclusion Remarks

3 THE EXPERIMENTAL PROGRAM
3.1 Introduction
3.2 Flow Chart of the Research
3.3 Materials Selection and Its Basic Properties
3.3.1 Portland Cement
3.3.2 Fine Aggregate
3.3.3 Coarse Aggregate (Gravel)
3.3.4 Micro Silica Fume (SF)
3.3.5 High Range Water Reducing Admixture (HRWRA)
3.3.6 Micro Steel Fibers
3.3.7 Water
3.3.8 Steel Reinforcement
3.4 Determination of the Workability of RPC Mixtures
3.5 Strength Activity Index for Mineral Admixture
3.6 Concrete Trial Mixes
3.7 Description of the Tested Column Specimens
3.7.1 Column Specimens Identification
3.8 Mixing Procedures
3.9 Casting and Curing Procedures of the Specimens
3.10 heating Process
3.10.1 Brick Furnace
3.10.2 Thermocouple
3.10.3 Electrical Gas Regulator and Ignition-Heating System
3.10.4 Hea ting Procedure of the Specimens
3.11 Cooling Procedure
3.12 Testing Procedure of Reinforced NSC and RPC Columns
3.13 Testing of Control Samples
3.13.1 Compressive Strength Test
3.13.2 Splitting Tensile Strength
3.13.3 Modulus of Rupture
3.13.4 Static Modulus of Elasticity

4 RESULTS AND DISCUSSION
4.1 Introduction
4.2 Mechanical Properties of Hardened Concrete
4.2.1 Compressive Strength
4.2.2 Splitting Tensile Strength
4.2.3 Flexural Strength (Modulus of Rupture)
4.2.4 Static Modulus of Elasticity
4.3 Description of Experimental Program of Column Specimens
4.4 Load Carrying Capacity of RPC and NSC Column Specimens
4.4.1 Behavior of RPC and NSC Columns at Normal Temperature (Group One)
4.4.2 Fire Exposed RPC and NSC Column Specimens
4.4.2.1 Group Two (Heatingat 200°c,Fire Temperature)
4.4.2.2 Group Three (Heatingat 400°c,Fire Temperature)
4.4.2.3 Group Four (Heating at 600°c Fire Temperature)
4.5 Load-Displacement Relationships of RPC and NSC Column Specimens..
4.6 Ductility of RPC and NSC Column Specimens
4.7 Stiffness of RPC and NSC Column Specimens
4.8 Energy Absorption of RPC and NSC Column Specimens
4.9 Effect of Fire Exposure on Steel Reinforcement
4.10 Appearance and Color Change of the column specimens
4.10.1 RPC Column Specimens
4.10.1 NSC Column Specimens
4.11 Discussions of Failure Process of RPC and NSC Columns

5 FINITE ELEMENT ANALYSIS
5.1 Introduction
5.2 FE mesh and boundary conditions
5.3 Modelling and Analysis of RPC Column Specimens
5.3.1 Parts and Assembly
5.3.2 Static analysis model.
5.3.3 Coupled temperature-displacement analysis model
5.4 Finite Element Analysis Results and Discussion
5.4.1 Analytical Results of Group One (Reference RPC Columns)
5.4.2 Analytical Results of Groups Two and Three (Heating at 400 and 600°c Fire Temperature)
5.5 Parametric Study
5.5.1 Effect of Fire Exposure Scenarios on Load Carrying Capacity
5.5.2 Effect of Fire Exposure Scenarios on Stiffhess and Ductility
5.5.3 Spalling Implementation
5.5.4 Effect of slenderness ratio on the fire behavior of RPC columns

6 CONCLUSIONS AND RECOMMENDATIONS
6.1 Introduction
6.2 Key Findings
6.2.1 Mixture Preparation and Mechanical Properties of RPC
6.2.2 Experimental Residís of RPC and NSC Columns
6.2.3 Conclusions Based on the Numerical Model
6.3 Recommendations for Further Research

REFERENCES

MATERIAL SAFETY DATASHEETS APPENDIX A

(Appendix A: Removed for copyright reasons)

MODELLING OF MATERIAL PROPERTIES APPENDIX в

GOVERNING EQUATIONS AND VALIDATION STUDIES APPENDIX c

List of Tables

3-1 Main components and chemical composition of AL-Mass cement

3-2 Physical properties of AL-Mass cement

3-3 The original fine sand grading compared with the requirements of (IQS no.45/1984)

3-4 The physical and chemical properties of fine aggregate

3-5 The separated fine sand grading compared with the requirements of (IQS N0.45/1984)

3-6 Grading of natural coarse aggregate

3-7 Chemical analysis of silica fume used

3-8 Physical properties of silica fume used

3-9 Technical description of (Hyperplast PC200)

3-10 Chemical composition of the micro steel fiber

3-11 Micro steel fiber properties

3-12 Specifications and test results of steel reinforcing bars

3-13 Mix proportions of the reference and test samples

3-14 Pozzolanic activity index and w/c or w/cm ratios for test mortars

3-15 Details of RPC trial mixes (kg/m[3])

3-16 Details of the mixes used in this research

3-17 Summary of RPC and NSC column test specimens

4-1 Test values of cube compressive strengths of RPC samples before and after exposure to fire

4-2 Test values of cylinders splitting tensile strength for RPC samples before and after heating

4-3 Test values of prisms flexural strength of RPC samples before and after heating

4-4 Test values of modulus of elasticity of RPC specimens before and after heating

4-5 Experimental test results of RPC and NSC column specimens

4-6 Ductility index test results of RPC and NSC column specimens

4-7 Secant and initial stiffness test results of RPC column specimens

4-8 The energy absorption capacity test results of RPC and NSC columns

4-9 The effect of fire exposure on yield stress of steel reinforcement

5-1 The selected RPC columns for the theoretical study

5-2 The experimental and theoretical results for the reference RPC columns ...

5-3 The experimental and theoretical results for the exposed RPC columns at 400 and 600°c fire temperature

5-4 The theoretical results of the exposed RPC columns at 600°c Fire Temperature and with different fire configurations

5-5 Percentage of reduction in the load carrying capacity (Pu) of RPC columns after fire exposure at 600°c and with different fire scenarios

5-6 Initial stiffness and percentage of reduction in initial stiffness of RPC columns after fire exposure at 600°c and with different fire scenarios

5-7 Energy absorption capacity and percentage of reduction in energy absorption capacity of RPC columns after fire exposure at 600°c and with different fire scenarios

List of F igures

1-1 Types of reinforcement in columns (Assakkaf, 2004)

1-2 Failure modes of concrete columns for different (Le/d) ratios reproduced from (Module, 2010)

1-3 Column failure patterns for: a) Short column b) Long column (Module, 2010)

1-4 Concrete protective cover (The Constructor, 2016)

2-1 Uniaxial stress-strain curves for steel and concrete reproduced from (Caprani, 2006)

2-2 Calculation of beam stiffness (Ugural, 2003)

2-3 Illustration of occurrence of spalling (Harmathy, 1993 and Anderberg, 1997)

2-4 Moisture clog model for spalling of concrete according to (Harmathy, 1993)

2-5 Illustration of thermal dilation mechanism for fire induced spalling (Bažant, 1997)

2-6 Explosive spalling caused by combined thermal stresses and pore pressure by (Khoury, 2008) based on (Zhukov, 1975)

2-7 Concrete in fire, physiochemical process for one-hour duration reproduced from (Khoury, 2000)

2-8 Cardington test - column deformation and area of heating (Bailey, 2002)

3-1 Flow chart of the research plan

3-2 Grading curves for fine aggregate compared with requirements of (IQS N0.45/1984, Zone 2)

3-3 Grading Curves for Separated Fine Aggregate Compared with Requirements of (IQS N0.45/1984, Zone 4)

3-4 Materials proportion used in RPC mixture (% of total mix weight)

3-5 Materials proportion employed in NSC mixture (% of total mix weight)

3-6 Reinforcement details of the reinforced column specimens

3-7 RPC mixing time process

3-8 Illustrates the turnace and equipment

3-9 Top view and section view of the furnace and equipment

3-10 Electro-hydraulic testing machine used for testing column specimens

4-1 The effect of fire exposure on the compressive strength of RPC

4-2 Percentage of reduction in compressive strength at (400 and 600°C)

4-3 A comparison of compressive strength values at 28 days for different researchers

4-4 The effect of fire exposure on the splitting tensile strength of RPC

4-5 Comparison of residual splitting tensile strength for the present study and other researchers

4-6 The effect of fire exposure on the flexural strength of RPC

4-7 Comparison of residual flexural strength for the present study and other researchers

4-8 The effect of fire exposure on the modulus of elasticity of RPC

4-9 Comparison of the residual modulus of elasticity for the present study and other researchers

4-10 Effect of the transverse steel reinforcement on the load carrying capacity of reference RPC columns

4-11 Effect of the transverse steel reinforcement on the load carrying capacity of reference NSC columns

4-12 Effect of the transverse reinforcement on the percentage of increase in load carrying capacity of reference RPC and NSC columns

4-13 Effect of concrete cover on the load carrying capacity of reference RPC and NSC columns with different transverse reinforcement

4-14 The percentage of change in load carrying capacity of the exposed RPC and NSC column specimens at 200°c fire temperature

4-15 The percentage of reduction in load carrying capacity of the RPC and NSC columns after heating at 400°c fire temperature

4-16 The percentage of reduction in load carrying capacity of the RPC and NSC columns after heating at 600°c fire temperature

4-17 The residual load carrying capacity of RPC and NSC columns prior to and later on heating at different temperature levels

4-18 Axial load versus axial displacement of the reference RPC and NSC column specimens

4-19 Axial load versus axial displacement of RPC 10 and RPC20 at different temperature levels

4-20 Axial load versus axial displacement of RPC 11 and RPC21 at different temperature levels

4-21 Axial load versus axial displacement of RPC12 and RPC22 at different temperature levels

4-22 Axial load versus axial displacement of RPC 13 and RPC23 at different temperature levels

4-23 Axial load versus axial displacement of NSC11 and NSC21 at different temperature levels

4-24 Axial load versus axial displacement of NSC 13 and NSC23 at different temperature levels

4-25 Axial load versus mid-height lateral displacement of the RPC and NSC columns prior to and later on heating at 600°c

4-26 Determination procedures of column displacement ductility reproduced from: (a) (Pessiki and Pieroni, 1997), (b) (Azizinamini et al., 1999)

4-27 Example of diagram for determining the ductility index of RPC columns prior to and later on heating by (Pessiki and Pieroni, 1997) approach

4-28 Example of diagram for determining the ductility index of RPC columns prior to and later on heating by (Azizinamini et ak, 1999) approach

4-29 Effect of transverse reinforcement on the displacement ductility index of RPC and NSC columns at prior to and later on fire exposure determined by (Pessiki and Pieroni, 1997) approach

4-30 Effect of transverse reinforcement on the displacement ductility index of RPC and NSC columns at prior to and later on fire exposure determined by (Azizinamini et ah, 1999) approach

4-31 Illustration of secant and initial stiffness of the displacement-based design (DBD) approaches reproduced from (Sullivan et ak, 2004)

4-32 Example of diagram for calculations of secant and initial stiffness of RPC and RPC 13 columns prior to and later on heating

4-33 Percentage secant stiffness of RPC and NSC columns after heating at 400°c fire temperature levels

4-34 Percentage secant stiffness of RPC and NSC columns after heating at 600°c fire temperature levels

4-35 Example of diagram for determinations of area under the load-displacement curves of RPC 10 and RPC 13 columns prior to and later on heating

4-36 Effect of concrete cover in protecting the steel reinforcement after fire exposure at 600°c

5-1 The convergence study

5-2 The considered meshes for the RPC column specimens

5-3 The assembled parts of RPC column specimens

5-4 The upper and lower ends boundary conditions of columns

5-5 Fire temperature-time curve used for the exposed RPC columns at 600°c and for 1-hour fire duration

5-6 Experimental and theoretical load-axial displacement relationship of RPC column specimens prior to heating

5-7 Variation of vertical displacement, stress and plastic strain (PEMAG) for RPC columns prior to heating

5-8 Experimental and theoretical load-axial displacement relationship of the exposed RPC column specimens at 400°c

5-9 Experimental and theoretical load-axial displacement relationship of the exposed RPC column specimens at 600°c

5-10 The temperature distribution in the RPC column specimen (RPC 13) after 60- minutes of heating at 400°c

5-11 Heat transfer analysis with time for RPC23 column specimen after 60-minutes of heating at 600°c

5-12 Fire exposure of RC columns on different sides according to its position...

5-13 Fire scenarios included in the parametric study

5-14 Comparison between analytical values of different fire configurations for (a- RPC 10, b-RPC13, C-RPC20 and d-RPC23) column specimens at 600°c temperature level

5-15 Heat transfer analysis with time of different fire configurations for (a-RPC 10, b-RPC13, C-RPC20 and d-RPC23) columns after 120-mintues of heating at 600°c

5-16 Heat transfer analysis after 120-minutes of heating at 600°c and with (2- opposite and 2-adjcent sides) fire exposure

5-17 Spalling analysis with time for: (a) 2-adjacent sides fire exposure, (b) 4-sides fire exposure

5-18 Slenderness ratios included in the parametric study

5-19 load-vertical displacement of RPC13 with slenderness ratio of: 5(a) 9 and (b) 20

5-20 load-lateral displacement of RPC 13 with different slenderness ratios

List of Plates

2-1 Surface cracking after subjected to high temperatures (Balázs et al, 2010)...

2-2 Structural failure (Balázs et al, 2010)

2-3 Relative spalling in NSC and HSC columns under fire conditions (Kodur, 2000)

2-4 Spalling of flat slab with reinforcing bars draped towards floor of test room (Bailey, 2002)

3-1 Micro silica employed in this investigation

3-2 Sample of micro steel fiber used, after and before magnification

3-3 Steel reinforcement used in column specimens

3-4 Process of the flow table test

3-5 Mixers used in this study

3-6 Steel molds used for casting column specimens

3-7 The molded specimens and samples after covering

3-8 Column specimens steel molds and reinforcement bars

3-9 Steel reinforcement spacer to provide the required cover

3-10 Pouring concrete into the steel mold

3-11 Curing procedures for column specimens and samples

3-12 Details of the brick furnace

3-13 Thermocouple details

3-14 Digital gage details

3-15 Electrical gas regulator

3-16 Infrared ray thermometer used in this investigation

3-17 The full details of the heating process and furnace with the connections

3-18 Extinguishing of column specimens by foam spray fire extinguisher

3-19 Checking the verticality using a spirit level at both sides of the column

3-20 Compressive strength test machine

3-21 Splitting tensile strength test

3-22 Flexural strength test and machine

3-23 Modulus of elasticity test device

4-1 The appearance of cube samples after exposure to fire at different temperature levels

4-2 The appearance of cylinder samples after exposure to fire at different temperature levels

4-3 Moisture migration observed during: (a) heating process, (b) Testing process

4-4 Color change of RPC column specimens after exposure to fire with different temperature levels

4-5 Hairline cracks distributed along the exposed surfaces of NSC columns 140 4-6 Examples of the failure modes observed in this investigation

List of Abbreviations

illustration not visible in this excerpt

CHAPTER ONE

Introduction

1.1 General Overview

The development of concrete technology is becoming more advanced as engineers started to use other than just normal concrete in their designs. The upper limit of compressive strength for concrete that can be used in structures continues to be pushed higher and higher. In 1990s, the reactive powder concrete (RPC) has been developed which has a compressive strength greater than 140 MPa. These materials enhanced mechanical properties of concrete and give excellent flexural strength and extremely high ductility, more than 250 times greater than that of conventional concrete (Mehta, 1999).

Reactive Powder Concrete (RPC) is defined as a mixture of fiber- reinforced, superplasticized, with high cement and silica fume contents and low water/binder ratio and the presence of very fine quartz sand (0.15-0.60 mm) instead of regular aggregate. Coarse aggregate is eliminated for the enhancement of microstructure and performance of the RPC in order to reduce heterogeneity between the cement matrix and the aggregate. However, due to the elimination of coarse aggregate and the use of very fine sand, the cementitious materials used in reactive powder concrete is as high as (900-1200 kg/m[3]) (Collepardi, et al., 1997).

The excellent properties of (RPC), make it very attractive for structural applications. Columns transfer loads to the foundation so using RPC resulting in a significant dead load reduction of columns and reducing settlement of foundation. This can lead to an increase in the usable floor space within high-rise building (Prabha, et al., 2010).

1.2 Concrete Structures in Fire

In addition to the developments in concrete technology, the fire engineering design of buildings has become commonplace. Construction professionals pay particular interest to the behavior of building structures in fire conditions because it is important to create buildings and structures that minimize risk to both people and property as effectively and as efficiently as possible (Kodur and Mcgrath, 2003). (Al-Owaisy, et al., 2007) summarized that the fire resistance and post heat exposure behavior of structural members depend on thermal and mechanical properties of the materials composing these members. One of the benefits of concrete is it is considered a fire resistance material due to the good mechanical and thermal properties of its constituents, but concrete structures must still endurance fire by using proper design criteria. Structural components still must be able to withstand dead and live loads without collapse during the fire, which causes a decrease in strength and modulus of elasticity for concrete and steel reinforcement (Chen, et al., 2009). The mechanical properties of (RPC) exposed to high temperatures have gotten significant consideration in recent years (Sideris et al., 2009; El-Dieb, 2009; Al-Jabiri, 2015; Kadhum, 2015). These studies reported that RPC is vulnerable to explosive spalling when exposed to fire, which has raised great safety concerns. Thus, it is of basic criticalness to understand the assumptions that induce material degradation and explosive spalling to prevent RPC structures from failure.

For many years, the most common method of ensuring compliance with the requirements of the building regulations regarding the fire safety of concrete structures has been to rely on tabulated values for minimum dimensions and minimum cover to reinforcement (Lennon, 2004). The uncertainties in using tabulated values (usually derived from a regression of test data) for fire design have raised the urgent requirement for better fire design method. Recent years have seen a gradual transition from the prescriptive approach to the performance-based approach in the fire safety design since the latter provides a more cost-effective, flexible and rational tool and allows designers to use multiple routes to achieve the required fire safety (Beyler, et al., 2007; Engineers and Staff, 2007). The performance- based fire safety design approach requires tools for the accurate fire resistance analysis of reinforced concrete (RC) members, which has motivated the development of numerical simulation tools with the desired capability. Such a numerical simulation tool is capable of a three-step analysis: (a) fire scenario analysis, (b) heat transfer analysis, and (c) mechanical response analysis (Bailey, 2009).

Given the many benefits of RPC and their increased use in structural applications such as columns, it is essential that the fundamental behavior of them at elevated temperatures be understood to ensure that structural fire design involving RPC will be safe.

1.3 Column Behavior at Normal Temperature

According to (ACI Code 318, 2014) a structural element with a ratio of height-to-least lateral dimension exceeding three used primarily to support compressive loads is defined as a column. Columns support vertical loads from the floor and roof slabs and transfer these loads to the footings. Columns usually support compressive loads with or without bending.

Despite the tensile forces or stresses that may be produced, columns are generally referred to as (compression members) because the compression forces or stresses dominate their behavior.

illustration not visible in this excerpt

According to function, there are two types of reinforcement of column (Assakkaf, 2004), as shown in Figure (1-1):

1. Longitudinal Reinforcement: To take care of the moments and axial forces in columns reinforcing bars are provided parallel to the longitudinal axis of columns.
2. Lateral Reinforcement: It is provided to restrain local buckling, provides shear resistance, holds longitudinal steel and confines concrete.

illustration not visible in this excerpt

Figure (1-1): Types of reinforcement in columns (Assakkaf, 2004).

1.3.1 Slenderness Ratio and Column Failure Mode

Columns are classified as short or long depending on their slenderness ratios. The slenderness ratio given by kL/r where kL is the column’s effective length and r is the column’s radius of gyration (ACI Code 318, 2014). According to the (BIS 456, 2000), if the length of the column is more than 12 times the least dimension of its cross-section, then the column will be categorized as a long column (Slenderness ratio of the long column is > 22). While, if the length of the column is < 12 the least dimension of its cross-section, then the column will be a short column (Slenderness ratio of the short column is < 22).

Based on the height-to-least lateral dimension ratio of the column, there are three modes of failure of reinforced concrete columns (Module, 2010) as shown in Figure (1-2):

Mode -1: Compression failure of the concrete or steel reinforcement.
Mode -2: Combination of buckling and compression failure.
Mode -3: Buckling failure.

illustration not visible in this excerpt

Figure (1-2): Failure modes of concrete columns for different (Le/d) ratios reproduced from (Module, 2010).

Compression failure is likely to occur with columns which are short and stocky in result of this, column material fails and gets crushed, while Buckling is probable with columns which are long and slender, bending stress produces in the result of buckling which results in column failure as shown in Figure (1-3). A long column of the same material and the same cross section will carry less load as compared to the shorter column.

illustration not visible in this excerpt

a) Compression failure b) Buckling failure (short column) (longcolumn)

Figure (1-3): Column failure patterns for a) Short column,
b) Long column (Module, 2010).

1.3.2 Concrete Protective Cover

Concrete cover in reinforced concrete is the least distance between the surface of embedded reinforcement and the outer surface of the concrete (ACI Code 318, 2014) as shown in Figure (1-4) (The Constructor, 2014).

The concrete cover must have a minimum thickness for three main reasons:

1) To protect the steel reinforcement bars from environmental effects to prevent their corrosion.
2) To provide thermal insulation, which protects the reinforcement bars from the fire.
3) To give reinforcing bars sufficient embedding to enable them to be stressed without slipping.

illustration not visible in this excerpt

Figure: (1-4): Concrete protective cover (The Constructor, 2014).

1.4 Column Behavior under Fire Exposure

Generally, codes for fire design require columns in high rise buildings to have a fire resistance of 1 -4 hours based on the use and occupancy of the building, in order to provide a sufficient time to evacuate the building safely and before collapse (Raut and Kodur, 2011).

The influence of transverse reinforcement on the load carrying capacity of reinforced concrete columns at ambient temperature was the subject of many research projects (Saatcioglu and Razvi, 1998; Assa et al., 2001). These researchers reported that due to the confinement effect there is a noticeable increase in load carrying capacity for axially loaded columns with closer spacing of transverse reinforcement and more ductile behavior of a column was observed. The positive influence of transverse reinforcement on the load carrying capacity of reinforced concrete columns at ambient temperature raises a question whether a similar phenomenon would be seen in fire conditions.

Another important parameter that should be considered in the fire design of reinforced concrete columns is the minimum thickness of concrete cover to the reinforcement. Nearly all collapses in fire-damaged concrete structures are caused either by poor detailing or, in severe cases, by the failure of the steel reinforcement (Shi et al., 2004; Kigha et al., 2015). The reason for that is because the reinforcement is usually placed near the concrete member surface. Hence, the reinforcement is subjected to a greater temperature increase, and its strength is first affected in comparison to the main body of concrete (Kigha et al., 2015).

Depending on the architecture and structural layout of a building, columns can be exposed to a fire in some different scenarios (Raut and Kodur, 2011). For example, a wall could work as a barrier to column exposing only one, two or three faces of the column to fire. On the other hand, a column could be located in the middle of a room thereby exposing all four sides of the column to fire. In the case of the comer and peripheral columns, the columns may undergo bending due to the development of thermal gradients or due to the occurrence of non-uniform spalling, which can result in uniaxial (1-side fire exposure) or biaxial (2 adjacent sides fire exposure) bending of the column.

1.5 Problem Statement of Reinforced RPC Columns Exposed to Fire

Fire is a catastrophic occasion to which can attack any building during its lifetime. Not only does it represent an immediate risk to the occupants through the emission of harmful gases and heat, but this heat also considered a serious risk to the structural integrity of the entire building (Shihada and Nassar, 2012).

As mentioned before the evacuation process in high rise buildings must be carefully considered and the structural safety of the building should be assured for (1-4) hours. Previous researchers reported that columns made of normal strength concrete (NSC) can provide the required fire resistance without any external fire proofing. However, RPC columns may not exhibit the same level of performance and provide fire resistance for (1-4) hours as that of NSC columns (Raut and Kodur, 2011).

1.6 Significance of the Research

One of the major uses of (RPC) in buildings is for structural framing consisting of beams and columns, which are the primary load-bearing components, and hence the provision of appropriate fire safety measures for these columns is one of the major safety requirements in building design. The reason for this requirement can be referred to the fact that, when other measures for prohibiting the fire fail, the fire resistance of these members for adequate time is the last of defense (Kodur and Sultan, 1998). In addition, Mechanical behavior of (RPC) structures under high temperatures has seldom been studied. Therefore, Results of this study provide a valuable reference for future construction applications and design.

1.7 Objectives of the Present Study

The main objective of this research is to study the behavior of RPC columns exposed to fire at two adjacent sides and subjected to concentric loading, with the effect of various parameters such as fire temperature level, cover thickness, lateral tie spacing, and type of concrete.

From the above discussion, it is clear that there is a lack of knowledge on the fire response of RPC columns under realistic fire, to address this knowledge gap the following research objectives are set-out as a part of this research:

1) Searching the effect of fire exposure with different temperature levels (200, 400 and 600°C) on some mechanical properties of RPC samples, such as compressive strength, splitting tensile strength, flexural strength, and modulus of elasticity and comparing the results with control samples (at normal temperature).
2) Studying the role of concrete type, concrete cover, volumetric ratio of transverse reinforcement and fire temperature levels on the structural behavior of reactive powder concrete (RPC) and normal strength concrete (NSC) columns exposed to fire at two adjacent sides (mid-height lateral displacement, axial displacement characteristics, load versus displacements and the maximum load carrying capacity).
3) Investigating the effect of fire exposure on the ductility, stiffness and energy absorption of reactive powder concrete (RPC) and normal strength concrete (NSC) columns.
4) Experimentally examining the axial behavior of square reactive powder concrete (RRC) columns with different concrete covers and volumetric ratios of transverse reinforcement and comparing it to normal strength concrete (NSC) column specimens.
5) Examining the steel reinforcement bars after exposure to elevated temperatures through the extraction of steel bars from column specimens, then testing its yield strength.
6) Developing a numerical finite element (FE) model for
simulating the effect of fire heating on the behavior of RPC columns and simulating the fire-induced spalling based on pore pressure calculations in concrete.
7) Validating the above developed finite element (FE) model using data extracted from the experimental work.
8) Carrying out parametric studies to quantify the influence of different fire scenarios on the behavior of reactive powder concrete (RPC) columns including fire exposure at different sides (1, 2-opposite, 3 and 4 sides) with temperature level of 600°c, different fire durations (1 and 2 hours) and different slenderness ratios.

1.8 Outline of the Thesis

This thesis consists of six chapters as follows:

Chapter 1 (Introduction): This chapter gives some background on fire effect on concrete structures, especially reinforced concrete columns. Also, it gives a description of the research importance, problem, and objectives, in addition to the research outline.

Chapter 2 (Literature Review): This chapter reviews a number of studies and scientific researches which have been published on the impact of fire on reinforced concrete columns by accredited scholars and researchers.

Chapter 3 (Experimental Program): This chapter determines the basic properties of materials used, preparation of concrete, columns samples, heating process, devices and test of samples after heating.

Chapter 4 (Results and Discussion): This chapter discusses the results of the tests that are performed on (RPC) samples, reinforced concrete column specimens and reinforcement steel bars samples.

Chapter 5 (Numerical Simulation): This chapter offers a numerical simulation by developing a nonlinear finite element model using a powerful nonlinear finite element method package ABAQUS/Standard 2016 to simulate fire effect on the RPC columns and for better prediction of the response of RPC columns to a real fire in some parametric studies. Chapter 6 (Conclusions and Recommendations): This chapter includes main conclusions and recommendations drawn from the research work.

CHAPTER TWO

Review of Literature

2.1 Introduction

This chapter presents the literature relevant to this research project. This literature review covers the background of reactive powder concrete (RPC), effect of fire on concrete structures, properties of concrete and steel at elevated temperatures, behavior of concrete columns before and after fire exposure and previous experimental and numerical investigations on the fire performance of concrete columns.

2.2 Reactive Powder Concrete

Reactive Powder Concrete (RPC) is an ultra-high-strength and high ductility cementitious composite with advanced mechanical and physical properties. It consists of a unique concrete where the microstructure is optimized by precise gradation of all particles in the mix to yield maximum density. It uses extensively the pozzolanic properties of highly refined silica fume and optimization of the Portland cement chemistry to produce the highest strength hydrates. It is a special type of concrete that is rather a mortar than an actual concrete mixture because traditional coarse and fine aggregate are replaced by very fine sand with a particle size in the range of (150-600 μιη). Reactive powder concrete is composed of very fine particles (cement, sand, quartz powder and silica fume), steel fibers and super plasticizer. The super plasticizer, used at its optimal dosage, decreases the water to cement ratio (W/C) while improving the workability of the concrete.

These reactive powder concretes have compressive strengths ranging from 140 MPa to 800 MPa (Richard and Cheyrezy, 1995).

2.3 Composition of Reactive Powder Concrete

Griffith theory implies that the inherent defects (micro-cracks) in brittle materials leading to stress concentration and gives lower fracture strength of the materials (Lajtai, 1971). According to this theory, reactive powder concrete due to its compacted microstructure is able to achieve a greater strength and obtains enhanced durability properties. RPC is able to achieve its improved properties by using a very dense mix, consisting of fine particles and fibers.

Richard and Cheyrezy (1995) indicate the following principles for developing RPC in which the heterogeneity problems are substantially reduced:

1) Improving homogeneity by using fine sand with a particle size less than 600μιη and elimination of coarse aggregate.
2) Increasing matrix properties by using superplasticizer to reduce water- to-binder ratios.
3) Utilization of the pozzolanic properties of silica fume in the mix.
4) Increasing the compacted density, by using a fine granular mixture.
5) Improving compaction of the mixture by Application of pressure (before and during setting).
6) Heat-treatment of the mixture after hardening for the enhancement of the microstructure.

Application of these principles without steel fibers produces a matrix with very high compressive strength, but with ductility no better than that of conventional mortar. The inclusion of fibers improves tensile strength and makes it possible to obtain the required level of ductility.

2.4 Column Behavior at Ambient Temperature

2.4.1 Nominal Strength of Reinforced Concrete Column

The column structural member is vital mainly transmitting axial compressive load and their failure can lead to extensive damages, which can cause progressive collapse. As mentioned in chapter one, the most important factor affecting nominal strength and failure mode of the reinforced concrete column is the slenderness ratio. For long columns (KL/r > 22), bending stress produces in the result of buckling which results in column failure, a long column of the same material and the same cross section will carry less load as compared to the shorter column. While, for short columns (KL/r < 22), the key parameters affecting the strength of the column are the area of the cross-section of the column and the yield strength of both the compression and tension of concrete and steel (Emberley, 2013). When axial compressive loads are applied through the centroid of the cross section of a short column, concrete and steel reinforcement are shortened by the same amount due to their composite action. The ultimate load is attained when the reinforcement reaches its yield stress and the concrete reaches its 28-day compressive strength simultaneously (Caprani, 2006), as shown in Figure (2-1).

illustration not visible in this excerpt

Where:

pn0: nominal axial capacity of section at zero eccentricity

pnc: nominal axial load carried by concrete

Pns: nominal axial load carried by steel reinforcement

Ag: gross sectional area of column

As: cross sectional area of reinforcement

fc': concrete compressive strength at 28-days

Rationally, using RPC in columns achieve economic and technical advantages, for instance, reducing the total columns cost price. In addition increasing the usable floor space and allowing additional stories would be the essential benefits of utilizing RPC in columns. However, the effect of using RPC in reinforced columns, in improving the nominal strength of the column may differ from the calculated accordance with design criteria, from equation (2-2).

Kadhum and Mankhi (2015) carried out an experimental work to investigate the behavior of reactive powder concrete (RPC) columns with or without steel ties. The main objective is to investigate experimentally the behavior of RPC columns, to search the effect of the experimental variables, type of concrete (RPC and NSC) and spacing between steel ties. Twelve RPC columns were cast and tested under concentric axial compression load up to failure. The experimental results showed that RPC column specimens failed in a controlled manner without observing spalling of concrete cover or buckling of the longitudinal reinforcement to well beyond the peak load due to the inclusion of steel fibers in RPC. In addition, the space and amount of steel ties affect the load carrying capacity of columns by increasing the load carrying capacity with decreasing spacing of lateral ties. In comparison with normal concrete column specimens (with the same steel reinforcement), the results of using 150mm tie spacing showed an increase in the load carrying capacity by 280.4% for RPC columns with 2% steel fibers. While for 100mm tie spacing, the increase was 273.9% for RPC columns contained 2% steel fibers. At 50mm tie spacing, the increase was 272.8% for RPC columns with 2% steel fibers.

2.4.2 Confinement of Reinforced Concrete Column

It was not until very recently that design specifications and codes of practice started realizing the importance of introducing extreme event load cases that necessitate accounting for advanced behavioral aspects like confinement. Confinement adds another dimension to columns analysis as it increases the column’s capacity and ductility.

King (1946) presented the first comprehensive investigation into this subject. He demonstrated the provision of links in reinforced concrete columns may increase the capacity of a column in two ways, firstly by preventing buckling of the main longitudinal reinforcement and, secondly, if sufficient lateral reinforcement is provided by restraining the central core of concrete which, being in tri-axial compression, can sustain a higher load. In a column, these links are provided in the form of rectangular, circular, or spiral rings from top to bottom, which confines the concrete.

Ahmed and Shah (1982) summarized the benefits of confinement as the following:

1) Confinement compensates the strength loss, which results due to spalling of concrete cover.
2) Confinement increases the capacity of concrete to carry on large deformations without considerable strength loss.

Since the past few decades, thousands of concrete-framed high-rise buildings were added to the skyline, as a high-rise residential boom takes place across the world. The development in concrete technology over the twentieth century made it possible to design high-rise buildings exceeding 400m (Rizk, 2010). The use of higher strength concrete in columns is becoming increasingly frequent which are particularly prominent for the possibilities of increased load-carrying capacities and stiffness (Barrera, 2012). Currently ultra-high strength concrete (UHSC) with a compressive strength 200 MPa has been developed (Kimura et al. 2007). However, UHSC columns exhibit sudden cover spalling and exhibit an extremely brittle failure mode unless adequate confinement is provided. A considerable amount of work has been carried out to find out the positive effect of the lateral confinement on (UHSC) columns, which indicate that the presence of confinement on (UHSC) columns would affect the actual stress-strain curve of concrete. This effort gives a more accurate prediction on the compressive force of concrete in a column thus, resulting in further more efficient column cross section. The positive influence of transverse reinforcement on the load carrying capacity of reinforced concrete columns at ambient temperature raises a question whether a similar phenomenon would be observed in fire conditions. However, an answer to this question is not obvious and need for further intensive investigations.

2.4.3 Ductility of Reinforced Concrete Column

The word ductility used to describe the ability of any material to sustain inelastic deformation before fracture. Concrete is a brittle material, but it is generally accepted that conventionally reinforced concrete (RC) members can attain suitable ductile behavior by proper design and detail of steel reinforcement. The yield point of steel is thus treated as an important datum beyond which inelastic deformation of the RC member takes place, thus enabling the full stress and strain capacity of concrete to be developed before ultimate failure (Ahmed and Shah, 1982). Knowledge of behavior of confined concrete helps in calculating the most suitable quantity of confining steel.

The ductility of a column subjected to an axial load is a measure of how much loss of load carrying capacity occurs under increasing axial deformation once the cover region begins to fail. Different reinforced concrete elements have varying degrees of ductility, and this is recognized by design codes with the inclusion of strength reduction factors so as to prevent sudden failure of elements and jeopardize the lives of occupants. Confinement of concrete increases both axial strength and ductility of RC columns. At peak loads, after spalling of concrete cover, the strength and ductility of the member will depend upon the confinement of concrete core (Ho, 2011).

Maha et al. (2013) studied the ductility of reactive powder concrete (RPC) columns, reinforced reactive powder concrete (RRPC) columns, and compared the results with reinforced high strength concrete (RHSC) columns. The comparisons indicated that the ductility for both RPC and RRPC columns is greater than RHSC columns by about (100-130%). wili eh was due to the inclusion of steel fibers in RPC that enhances the ductility of columns with and without reinforcement.

Agha and Kadhum (2015) experimentally investigated the ductility of RPC columns with and without steel reinforcement, RPC cast with (1, 1.5, 2%) steel fiber. The results were compared with the similar NSC columns and with the same steel reinforcement. The results indicated that using RPC instead of NSC increased the ductility of columns by about (200-330%).

2.4.4 Stiffness of Reinforced Concrete Column

The general definition of Stiffness is the rigidity of an object and the extent to which it resists deformation in response to an applied force. The complementary concept is flexibility or pliability, in which the more flexible an object is, the less stiff it is (Baumgart and Linder, 2000). The stiffness (k) of a body is a measure of the resistance offered by an elastic body to deformation (Ugural and Fenster, 2003). For an elastic body, the general formula to calculate its stiffness (Hooke's law) as shown in Figure (2-2) is:

illustration not visible in this excerpt

Figure (2-2): Calculation of beam stiffness (Ugural and Fenster, 2003).

The most important factor affecting the stiffness of concrete structures is Young’s modulus of elasticity. In the elastic deformation region, the ratio between stress and strain (Young’s modulus of elasticity) is the mathematical description of the stiffness of a material. A stiff material has a strong supporting structure and does not deform much when a stress is applied. Therefore, stiff materials, by definition, have a high modulus of elasticity (Lei et. al, 2013). Therefore, the modulus of elasticity is often one of the primary properties considered when selecting a material, in which a high modulus of elasticity is sought when deflection is undesirable. For an element in tension or compression, the axial stiffness is:

illustration not visible in this excerpt

Where:

A: is the cross-sectional area,

E: is the (tensile) elastic modulus (or Young's modulus),

L: is the length of the element.

Many extensive studies have been carried out to study the ductility and stiffness of UHSC columns. It is well recognized that ductility and stiffness are important structural properties, which provides a noticeable warning at the beginning of the failure of the structure to users for evacuation. According to that, the effect of fire on stiffness and ductility of these columns must be understood and need further investigations, especially that these properties affected mainly on the strength and stress-strain behavior of the concrete and steel reinforcement, which are highly deteriorating during the fire.

2.5 Column Behavior under Fire Exposure

The main objectives of fire safety engineering are to prevent the loss of life or injury and property loss during the fire (Lin et al., 1992). Obviously, these objectives can be best achieved by preventing ignition, which is the start of the fire by using active fire protection systems, such as sprinkler systems, minimize the spread of fire and smoke generation before the fire fully develops. If active fire protection measures are unable to contain the fire, and the fire does become fully developed. The fire safety objectives are then addressed by ensuring the structural members have adequate fire resistance to prevent both the spread of fire and collapse of the structure for the adequate period for the safe evacuation of occupants and for the safety of firefighters.

In this section, the main parameters that deteriorated and the phenomenon's that affect the behavior of concrete columns during the fire which reported by previous researchers are summarized below.

2.5.1 Effect of Fire Exposure on Concrete

Concrete is a composite material that consists mainly of mineral aggregates bound by a matrix of hydrated cement paste. The matrix is highly porous and contains a relatively large amount of free water unless artificially dried. When exposing it to high temperatures, concrete undergoes changes in its chemical composition, physical structure, and water content. These changes occur primarily in the hardened cement paste in unsealed conditions. Such changes are reflected by changes in the physical and the mechanical properties of concrete that are associated with temperature increase. Deterioration of concrete at high temperatures may appear in two forms (Lublóy, 2010):

1) Local damage (Cracks) in the material itself, Plate (2-6).
2) Global damage resulting in the failure of the elements, Plate (2-7).

illustration not visible in this excerpt

Over the last three decades, there has been significant research and development activity in improving the properties of concrete. This has led to new types of concrete which are often referred to as high strength concrete (HSC), fiber reinforced concrete (FRC), high-performance concrete (HPC) and ultra-high performance fiber reinforced concrete (UHPFRC) including reactive powder concrete (RPC) (Mehta, 1999). In addition to the developments in concrete technology, the fire engineering design of buildings has become essential. The construction professionals pay a particular interest in the behavior of concrete structures in fire conditions because it is important to create buildings and structures that minimize risk to both people and property as effectively and as efficiently as possible. Generally, concrete structural members (traditionally used to be made of NSC) exhibit good performance under fire situations because concrete is non-combustible and has a relatively low thermal conductivity. In addition, if the concrete cover remains in place during heating (there is no cover spalling), heat flow to the interior of a reinforced concrete element in fire occurs slowly. Concrete structures are therefore widely assumed to possess inherent fire resistance, which is typically ensured in design by prescribing minimum overall member dimensions and minimum concrete cover to the steel reinforcement (Kodur and McGrath, 2003). However, the behavior of RPC, when exposed to fire scenarios, is different from that of NSC and may not exhibit the same level of performance as that of NSC. Further, the spalling of concrete under fire conditions is one of the major concerns due to the low porosity (permeability) in RPC (Tai, et al., 2011).

The mechanical properties that are of primary interest in fire resistance design are the compressive strength, tensile strength, elastic modulus and stress strain response in compression. Mechanical properties of concrete at elevated temperatures have been studied extensively (Kodur and Harmathy, 2008; Harmathy, 1993; Tang and Lo, 2009). It is helpful to consider recent literature on the effects of fire on concrete elements as this is relevant to understanding the potential effects of fire on modern concretes. Recent testing has been fundamentally concerned with considering the strength and spalling behavior of HSC and RPC structures under various fire exposure conditions (Lennon et al., 2002).

Tai et al. (2011); Bashandy (2013); Al-Jabiri (2015); Kadhum (2015) studied the effect of heating by fire flame on the mechanical properties of reactive powder concrete. These studies summarized that at fire temperature below 300°c, a slight increase in compressive strength was indicated when comparing with the results at ambient temperature. Splitting tensile, flexural strength and modulus of elasticity were also increased at this temperature. While, above 300°c fire temperature, a rapid deterioration in all mechanical properties of the RPC has taken place.

2.5.2 Spalling of Concrete under Fire Exposure

Spalling is the separation of concrete from the surface in concrete structures when they are exposed to high temperatures. Fire induced spalling in concrete members can broadly be classified into three stages, namely: early spalling, intermediate spalling, and late spalling (Dwaikat and Kodur, 2009). In typical RC members (beams and columns), early spalling can start immediately after fire exposure (5 to 10 minutes) and can continue up to 30 to 45 minutes, This type of spalling is generally of explosive nature and often results due to the development of high thermal gradients within the parts (surface close to fire exposure) of concrete members. Early spalling results in the break-up of large chunks of concrete and thus may have a detrimental effect on the fire resistance of RC columns, particularly when spalling reduces the concrete cover thickness to the main reinforcement (Dwaikat and Kodur, 2009; Kodur and Phan, 2007). Intermediate spalling occurs midway into fire exposure (about 30 to 45 minutes), results in the form of surface scaling (breaking off small pieces of concrete), and this can continue through final stages of fire exposure. Such type of spalling has less effect on the fire resistance of the members since it generally results in breakup of thin layers of concrete (Kodur and Phan, 2007). Late spalling occurs just prior to failure of RC columns and primarily results from significant loss of strength and stiffness of these columns due to prolonged fire exposure. This late spalling has a minor influence on the overall fire response of RC columns. Such type of spalling occurs even in room temperature conditions prior to failure of RC columns (Dwaikat and Kodur, 2009).

A review of the literature indicates that the main factors influencing fire induced spalling in concrete are moisture content, concrete permeability, concrete strength, fire scenario (rate of heating) and stress levels (Phan 1996; Phan and Carino, 2000; Kodur and Phan, 2007).

2.5.2.1 Mechanisms of Spalling

Dwaikat (2009) stated that the review of literature presented a conflicting picture on the occurrence of spalling in RC members and on the exact mechanism for spalling. While some researchers, based on fire test data, reported explosive spalling in concrete members, a number of other studies reported little or no significant spalling. One possible explanation for this confusing trend of observations is a large number of factors that influence the spalling and the interdependency of some of these factors. However, most researchers agree that major cause's for spalling in concrete are due to the low permeability of concrete and moisture migration at elevated temperatures. There are three broad theories of which the spalling phenomenon can be explained (Kodur, 2000):

- Pressure build-up: Spalling is believed to be caused by the build-up of pore pressure during heating. RPC is believed to be more susceptible to this pressure build-up than NSC because of its low permeability. The extremely high water vapor pressure, generated during exposure to fire, cannot escape due to the high density (and low permeability) of RPC. When the effective pore pressure exceeds the tensile strength of concrete, chunks of concrete fall off from the structural member as shown in Figure (2-3). This falling off can often be explosive depending on the fire and concrete characteristics (Harmathy, 1993; Anderberg, 1997). Plate (2-8) shows photographs of the column specimens with NSC and HSC after the fire resistance test is applied (Kodur, 2000).

illustration not visible in this excerpt

Figure (2-3): Illustration of occurrence of spalling (Harmathy, 1993;

illustration not visible in this excerpt

Plate (2-3): Relative spalling in NSC and HSC columns under fire conditions (Kodur, 2000).

Moisture is clearly an important factor, One classical approach to the role of moisture was formulated by (Harmathy, 1993) (the Moisture Clog Theory), When the concrete specimen is heated, the steam pressure in the pores rises close to the surface. The pressure gradient then drives moisture not only out of the specimen but also towards thinner colder regions. When the steam meets a neighboring colder layer, it will condense. This process will continue, moving further into the cross-section, until a fully saturated region of considerable thickness will be created, this region is the so called (moisture clog). When the moisture clog is created, further movement of steam inwards towards the colder regions is restricted which will lead to the situation shown in Figure (2-4). The main point is that the highest pressure will be developed at the boundary between the moisture clog and the steam moving inwards and when (or if) this pressure exceeds the tensile strength of the concrete a piece will spall away.

- Restrained thermal dilatation: This hypothesis considers that spalling results from restrained thermal dilatation close to the heated surface, which leads to the development of compressive stresses parallel to the heated surface as shown in Figure (2-5), these compressive stresses are released by brittle fractures of concrete (spalling). The pore pressure can play a significant role in the onset of instability in the form of explosive thermal spalling (Bažant et al., 1997).

(a) Restrained Concrete (b) Developed Stresses

illustration not visible in this excerpt

Figure (2-5): Illustration of thermal dilation mechanism for fire induced

spalling (Bazantet al., 1997).

- Combined thermal dilatation and pore pressure induced explosive spalling: The experience and analysis from several tests carried out over the last few decades on explosive spalling showed that it is almost impossible to assign the different observed types of failure to just one of the two mechanisms of spalling. Either spalling due to high pore pressure or spalling due to thermal stresses.

Zhukov (1975) first proposed starting in the middle of 1970’s, the idea of a combined thermal stress and pore pressure-induced explosive spalling. (Connolly, 1995) developed new ideas on combined pore pressure and thermal stress spalling. The main results from these findings can be summarized as follows:

- Connolly considered uniform stresses due to thermal expansion and an applied load causing stresses and cracks in the longitudinal direction.
- Additional stresses due to a high pore pressure perpendicular to the heated surface will cause concrete parts to spall.

Based on Zhukov’s ideas, (Khoury, 2008) presented a general sketch of combined thermal stress and pore pressure-induced explosive spalling as shown in Figure (2-6).

illustration not visible in this excerpt

Figure (2-6): Explosive spalling caused by combined thermal stresses and
pore pressure by (Khoury, 2008) based on (Zhukov, 1975)

In the case of RPC, thermal spalling has seldom been studied. However, it is probable that the lower permeability of RPC prevents water vapor from escaping, causing considerable internal vapor pressure that often results in significant spalling. Therefore, a decrease in strength and an increase in the risk of spalling of RPC at elevated temperatures should be investigated.

2.5.3 Physiochemical Transformations of Concrete under Fire Exposure Khoury (2000) summarized the behavior of reinforced concrete during the fire and the degree of effect as shown in Figure (2-7). At each degree of heating after one hour of exposure, where the increase of temperature degree increases the deterioration of concrete member harm which is caused by a combination of the effects of gases, which are emitted from heating materials, and the effects of flames and high air temperatures (Khoury, 2000).

Figure (2-7): Concrete in the fire, the physiochemical process for the one-
hour duration reproduced from (Khoury, 2000).

Bilow et al. (2008) summarized the following physiochemical transformations can be observed by the increase of temperature: At around 100°c, the weight loss indicates water evaporation from the micro pores. Dehydration of ettringite (3Ca0A1203-3CaS0431־H20) occurs between (50 and 110°C). In the temperature range (100-300°C), free and bound water from C-S-H gel is evaporated. Above 300°c a reduction in strength in the range of (15-40%) occurs. At 550°c, the reduction of strength in the range (55-70%) and dihydroxylation of Ca(OH)2 takes place. The dehydration of calcium silicate hydrated and the thermal expansion increase internal stresses and micro cracks, which are induced by the cementing material. Fire is generally extinguished by water and CaO turns into Ca(OH)2 causing cracking and crumbling of concrete. Dehydration of calcium silicate hydrates was found at the temperature of 700°c.

The alterations produced by high temperatures are more evident when the temperature surpasses 500°c. Most changes experienced by concrete at this temperature level are considered irreversible. C-S-H gel, which is the strength-giving compound of cement paste, decomposes further above 600°c. At 800°c, concrete is usually crumbled and this stage of concrete is characterized by the collapse of its structural integrity, revealing residual compressive strength (Bilow et al., 2008).

2.5.4 Effect of Fire Exposure on Steel Reinforcement

The performance of steel during a fire is understood to a higher degree than the performance of concrete, and the strength of steel at a given temperature can be predicted with reasonable confidence (Fletcher et al., 2007). It is generally held that steel reinforcement bars need to be protected from exposure to temperatures in excess of 250-300°C; this is because steels with low carbon contents are known to exhibit blue brittleness between 200 and 300°c. Concrete and steel exhibit similar thermal expansion at temperatures up to 400°c. However, higher temperatures will result in insignificant expansion of the steel compared to the concrete and, if temperatures of the order of 700°c are attained, the load-bearing capacity of the steel reinforcement will be reduced to about 20% of its design value (Fletcher et al., 2007).

Topçu and Karakurt (2008) investigated the mechanical properties of steel reinforcement bars after the exposure to high temperatures. Plain steel, reinforcing steel bars embedded into a mortar and plain mortar specimens were prepared and exposed to (20, 100, 200, 300, 500, 800, and 950°C) fire temperatures for 3 hours individually. Concrete cover of 25 mm provides protection against high temperatures up to 400°c. The high temperature exposed plain steel and the steel with 25mm cover has the same characteristics when the reinforcing steel with the 25mm cover was exposed to a temperature 250°c above the exposure temperature of plain steel.

Mamillapalli (2009) investigated the impact of the elevated temperatures on reinforcement steel bars by heating the bars to (100, 300, 600 and 900°C). The heated samples were rapidly cooled by quenching in water and normally by air-cooling. The changes in the mechanical properties were studied using the universal testing machine (UTM). The impact of elevated temperature above 900°c on the reinforcement bars was observed. There was a significant reduction in ductility when rapidly cooled by quenching. In the same case, when cooled under normal atmospheric conditions the impact of temperature on ductility was not high.

Topçu and Isikdag (2008) investigated the mechanical properties of reinforcement steel bars after exposure to elevated temperatures. The в 16 mm-ribbed steel bars were used to prepare (56 X 56 X 290 mm), (76 X 76 X 310 mm), (96 X 96 X 330 mm) and (116 Xİ16 X 350 mm) specimens with concrete covers of (20, 30, 40, and 50 mm) against elevated temperatures up to 800°c. The reinforcement steel bars were embedded in mortars, and then specimens were exposed to (20, 100, 200, 300, 500, and 800°C) temperatures for 3 hours, individually. After the cooling process, the specimens were cured for 28 days. The mechanical tests were conducted on cooled specimens, and the ultimate tensile strength, yield strength, and elongation of mortar specimens at various temperatures were determined at the end of the experiments. The results indicated that a cover of (20, 30, 40, and 50mm) thickness provided a protection to rebar in the exposure of high temperatures. The cover reduced the losses in yield and tensile strengths of rebar and provide a 15% higher strength compared to rebar without cover. For temperatures up to 300°c, a rebar with cover had the same yield and tensile strengths with that of the rebar without cover in exposure to elevated temperatures. However, when the temperature increased up to 800°c, the rebar without cover lost an average of 80% of its strength capacities compared with a 20% loss for the rebars with cover. It was observed that (20, 30, 40 and 50mm) cover thickness was not sufficient to protect the mechanical properties of rebars in the exposure of temperatures above 500°c.

2.5.5 Latent Heating

The general definition of Latent heating is referred to anything absorbing large amounts of heat through physical and/or chemical processes. Materials containing large amounts of chemically combined water in their structure, like concrete, can form heat sinks. They absorb significant amounts of heat due to the energy consumed in the water-evaporation process (Al-fawakhiri et al., 2002).

Latent heating which is a significant phenomenon that occurs in concrete when it is subject to fire, due to its low thermal conductivity and low thermal diffusivity, concrete acts like a heat sink (Buchanan and Abu, 2002). It takes much longer for a concrete member to rise in temperature than a conductive material such as steel. On the other hand, this also means that heat is not dissipated as quickly in concrete. In the cooling phase, even though the temperature is decreasing, the interior of the concrete member could still be increasing in temperature. Due to the still high temperatures in the cross-section, the outer portion of the column cools first while the interior is still increasing in temperature. This can be a significant problem because concrete does not fully recover its initial strength even when cooled back to ambient conditions. With latent heating, the cross could still be losing unrecoverable strength due to developing of the thermal gradient between the inner and outer of the column when externally everything appears safe (Salah et al., 2011).

Khoury (2003) reported that there is a post-cooling spalling occurs after the fire is over, after cooling down or maybe even during extinguishing. Which assumed to be due to the developing of the thermal gradient between the inner and outer of the column. In addition to the rehydration of CaO to Ca(OH)2 after cooling, with an expansion of over 40%.

2.5.6 Plastic Centroid Movement

Raut and Kodur (2011) explain the occurrence of this phenomenon in the cases of asymmetric fire exposure of concrete columns. The plastic centroid is the location of the centroid of the resistance forces, tension and compression steel and concrete section, in a column (Park and Paulay, 1975). Under normal conditions, the plastic centroid for a symmetrically reinforced column is located directly in the center of the column. However, in fire conditions, the plastic centroid shifts depending on the exposure. For one, three, and two adjacent sided fire exposures, the column strength will deteriorate asymmetrically leading to a shift to either the right or the left of the original plastic centroid location. The shifting of the centroid will decrease the load carrying capacity of the column due to the failure of deteriorated sides of the column.

2.5.7 Increasing of Slenderness Ratio

As mentioned before, the most important factor affecting nominal strength and failure mode of the reinforced concrete column is the slenderness ratio. Columns with same characteristics and different slenderness ratios will exhibit different nominal strengths. Column with larger slenderness ratio is more susceptible to flexural buckling failure leading to lower nominal strength. According to that, (Emberley, 2013) assumed that, as long as the radius of gyration is based on the width and depth of a column’s cross-section, the spalling of concrete cover will increase the slenderness ratio throughout a fire exposure. This means that a column will exhibit a lower nominal strength later in a fire exposure than at ambient conditions.

2.6 Literature Review of the Experimental Studies on the Behavior of Concrete Columns under Fire Exposure

2.6.1 Lie (1989)

Experimental studies were carried out by (Lie, 1989) to develop general methods for the prediction of the fire resistance of reinforced concrete columns, 41 columns were tested. The parameters studied included the amount of steel reinforcement, concrete strength, and moisture content. It is concluded from the results that increasing of the amount of reinforcing steel of the columns produce relatively small increases in fire resistance. While, The influence of concrete strength, concrete moisture content is insignificant.

2.6.2 Ali et al. (2010)

This researcher examined the behavior of high strength concrete columns in fire. (30) High strength concrete columns were exposed to fire to clarify the effect of the parametric study investigated. The high strength concrete columns subjected to two heating rates, with special attention directed towards explosive spalling. In all the tests, explosive spalling happened in a period of the first 50 minutes of heating. In all cases, minor spalling took place first, followed by major and severe spalling. Explosive spalling never happened at the late heating time.

2.6.3 Bikhiet et al. (2014)

The author in this paper examined the behavior of fifteen column specimens (15x15x100 cm) exposed except one specimen to (600°C) fire exposure and subjected to an axial load to evaluate the reduction in column compressive capacity after the fire. The main studied parameters were the concrete strength, fire duration, the percentage of longitudinal reinforcement and bar diameters. Experimental results indicated that for columns not exposed to fire, the first crack appeared at 80% of column failure load while the first crack occurred at 50% of column failure load for columns exposed to fire. Columns with the same reinforcement percentage but with smaller bar diameters gained less lateral strain and smaller vertical displacement than columns with bigger bar diameters. Using high-grade steel as main reinforcement showed failure load higher by 55% than that of column reinforced by mild steel.

2.6.4 Seręga (2015)

This research presented an experimental study of the influence of transverse reinforcement spacing on the fire resistance of axially loaded, HSC columns with a circular cross-section. The results of full-scale tests indicated that columns with a spacing of ties recommended by the code provisions for design of concrete structures could suffer a premature failure because of inelastic buckling of main reinforcing bars between adjacent ties. In addition, the confinement effect was not observed for the columns with a congested spacing of transverse reinforcement.

2.6.5 Bodalal et al. (2017)

This research proposed an experimental work to study the effects of high temperature generated from fire on the high strength concrete (HSC) columns performance. In this study, first, the spalling failure mode of HSC columns was reviewed and the spalling temperature and methods to improve it was discussed. Then the fire resistance criteria based on time was highlighted. The previous deals with total failure. It was indicated from the results that fire more adversely affects high strength concrete (HSC) columns than normal strength concrete (NSC) columns. With increasing temperature, the tensile strength of concrete decreases and thus the risk of spalling increase. The faster degradation of compressive strength with temperature, combined with the occurrence of spalling, leads to lower fire resistance in HSC members.

2.7 Literature Review of the Finite Element Analysis of Fire Exposed Concrete Column Models

The following section is a summary of the important papers on the subject of both reinforced concrete column fire performance and numerical modeling of heat transfer differential equations. The works listed are those that contributed significantly and have a close relationship to the present work.

2.7.1 Allen and Lie (1974)

This paper outlined the process of developing a numerical procedure to calculate the fire performance of reinforced-concrete columns. The authors describe material property changes due to an elevated temperature well. One of the most important details about this article is that the authors developed their own model and compared it against actual column furnace tests. The furnace tests used both standard and natural fire curves.

2.7.2 Lie and Irwin (1993)

The authors of this paper developed a model to analyze the fire performance of reinforced concrete columns. One of the unique difference between this paper and the work of (Allen and Lie, 1974) is that this paper incorporated a heat transfer term to account for moisture in the concrete.

2.7.3 Ali et al. (2010)

Those authors developed a three-dimensional finite element (FE) model to examine the behavior of high strength concrete columns in the fire, taking into account exposure to high temperatures. The concrete columns were modeled taking into account the embedded reinforcement and crack formation and propagation using the smeared cracks model, which allowed a nonlinear transient structural analysis to be conducted. The comparison of the results of the FE analysis and the tests performed showed a reasonable agreement and a divergence in some cases due to concrete spalling. An assessment of stresses generated in the high strength concrete columns under fire using the FE model was also presented. The following conclusions were drawn from this model:

1) FEM calculations have shown that compressive and tensile stresses can be created in the same section, depending on the temperature profile. The diversity in stress profiles can be one of the main parameters to cause spalling as the calculated tensile stresses in hot surfaces reached 8.69 MPa higher than the tensile strength of the concrete.
2) It is recommended (where it is feasible) to use experimentally obtained mechanical and thermal proprieties as the input parameters to numerical models to obtain more accurate analysis results.

2.7.4 Raut and Kodur (2011)

This research proposed a numerical model, in the form of a computer program, was developed for tracing the fire response of RC column. The model, based on macroscopic finite element approach. The model accounts for fire induced spalling, various strain components, and high-temperature material properties. The proposed macroscopic finite element model was capable of predicting the fire response of RC columns, in the entire range: from pre-fire stage to collapse stage, with a good amount of accuracy. This model accounts for critical parameters such as different fire scenarios, high- temperature material properties, various strain components, fire induced spalling in RC columns arising from 1-, 2-, 3-side exposure.

2.7.5 Kadhum (2013)

This author conducted an investigation to study the behavior and load carrying capacity of reinforced concrete columns exposed to fire. Finite element method was used to idealize the effect of heating by fire exposed to reinforced concrete columns. The specimens that were subjected to fire exposure at temperature levels of (400, 600, and 750°C) for 1.5-hour period of exposure at age of 60 days were determined. A three­dimensional nonlinear finite element model was adopted to investigate the structural behavior of reinforced concrete column specimens with and without exposure to heating. The comparison indicated a good agreement between the adopted finite element model results and experimental results. The experimental to theoretical ultimate load ratio ranged from (1.01-1.07) before heating and (1.03-1.23) after heating for the analyzed column specimens. The adopted finite element analysis showed, also, good agreement with experimental results throughout the load-deflection behavior before and after heating.

2.7.6 Emberley (2013)

This research proposed a finite element model using ANSYS and several published column furnace tests were used to benchmark the heat transfer and structural analysis portions of the model. One, three and four­sided fire exposures along with the ASTM El 19 fire curve and a natural fire curve were used to study latent heating effects on failure modes. Assessments of column structural capacity were performed in accordance with the provisions of ACI 318. The completed model served as an effective tool for the thesis and is available to help aid students and engineers investigate the design of reinforced concrete columns under fire conditions through integration the heat transfer analyses and the structural evaluations.

2.7.7. Seręga (2015)

This paper proposed a numerical analysis of tested columns. The columns were modeled with embedded reinforcement. The applied material model took into account the influence of transient temperature on mechanical properties of concrete and steel. The effect of cracking, development of transient creep strains and plastic strains for concrete were also included in the analysis. The inelastic buckling of main reinforcement was modeled using average stress-strain relationships for steel in compression. The comparison of numerical simulations and experiments showed reasonable agreement. The results of calculations indicated that during the completely heating period high thermal gradients generated tensile stresses in the plane of the cross-section of the columns. Due to this fact, the confinement effect was not observed for the columns with a congested spacing of transverse reinforcement.

2.7.8 Balaji et al. (2016)

This article investigated the axial capacity of reinforced columns exposed to fire. Finite element software ANSYS was used to perform the thermal analysis. A set of numerical studies were carried out to quantify the effect of various parameters on short columns subjected to fire. The study was performed on columns of different cross-sections to investigate the effect of these parameters, the thermal boundary conditions, grades of concrete and steel and concrete cover. The fire ratings based on various failure criteria were determined. The following conclusions were drawn from this model:

1) The axial capacity and fire resistance decreased directly with thermal boundary conditions, having maximum effect on columns with four-side exposure. For a particular time of exposure, the four-sided exposure caused a 65% reduction in axial capacity compared to single side exposure.
2) The grade of concrete and steel had less effect in thermal criteria of failure but a significant effect on axial capacity and fire rating based on strength criteria. Contradictory to normal strength design at ambient temperature, increasing the grade of concrete and steel had an adverse effect on fire rating based on strength criteria.

2.8 Case Study (Full-Scale Fire Resistance Test)

While it is important to understand the performance of individual concrete members during the fire, the behavior of the same structural elements within the context of a complete structure can vary widely from their independent responses. It must be recognized that in situations where an individual concrete member might have failed, the overall structure may not fail, due to the inherent redundancies and load redistribution. This is a common phenomenon in composite structures (Mostafaei, 2014).

There have been very few full-scale tests on concrete structures. The only one reported is that conducted at Cardington in the large test enclosure. This test is reported by (Bailey, 2002). The tested structure was a six-story reinforced concrete building with high strength concrete columns and reinforced concrete flat slabs. The concrete used for the slabs had 28-day cube strength of 61 MPa. While concrete used in the columns had a compressive strength of 103 MPa at 28 days. The building was constructed within the large test enclosure at Cardington. The columns incorporated 2.7 kg/m[3] of polypropylene fibers within the mix. No fibers were added to the concrete mix used for the floors, as this was considered not to be a threat with respect to spalling. Part of the ground floor of the building was compartmentalized to form a fire enclosure measuring (15/15m); details are given in Figure (2-8). The resulting fire induced very significant spalling of the floor such that the bottom reinforcement was exposed with some of it being draped towards the floor; Plate (2-9). The temperatures within the test enclosure were unable to be measured beyond a certain point due to the loss of instrumentation. Despite the loss of concrete from the floor due to spalling, the floor continued to support the imposed load. The floor expanded considerably with one of the columns having an outwards lateral residual displacement of 57 mm at the floor location.

Bailey (2002) suggested that the spalling of the floor slab was increased by the development of compressive stresses within the slab due to the expansion of the floor on heating.

illustration not visible in this excerpt

Figure (2-8): Cardington test - column deformation and area of heating

illustration not visible in this excerpt

2.9 Concluding Remarks

From the previous review of literature in the field of effect of fire on concrete structures, it can be noted that there is a large number of research, which has been conducted in studying the behavior of concrete members made of NSC under fire exposure, but the behavior of RPC members under fire exposure has seldom been studied. Only few studies were reported on the behavior of RPC members exposed to fire. (Mohammed and Kadhum, 2016) investigated the effect of fire exposure on reactive powder concrete beams and (Alwash, 2015) studied the behavior of reactive powder concrete slabs exposed to fire. Those researchers illustrated that RPC specimens were more pronounced to fire-induced spalling when compared with the NSC specimens. On the other hand, there is no data available on the behavior of RPC columns during fire exposure. Therefore, this investigation provides a vulnerable information and better understanding of the conduct of RPC columns under real fire, especially that RPC is majorly utilized in columns, which are the primary load-bearing component, and hence the provision of appropriate fire safety measures for these columns is one of the major safety requirements in building design.

The following conclusions can be summarized from the review of the literature:

- Previous researchers reported that concrete structures made from NSC exhibit an excellent performance under fire situation and usually provide a sufficient time to evacuate the building safely, but the behavior of RPC structures may differ and need further investigations.
- The positive influence of transverse reinforcement on the load carrying capacity of reinforced concrete columns at ambient temperature raises a question whether a similar phenomenon would be observed in fire conditions and need further intensive investigations.
- The permeability of concrete has a significant influence on the spalling of concrete at elevated temperature, therefore, RPC exhibit higher spalling than normal concrete at the same temperature due to its compacted microstructure and low permeability.
- The use of fibers greatly influences the occurrence of spalling in RPC under high temperatures, in general mixes with higher steel fiber content show excessive cracking and pulverized spalling.
- The reduction in the load carrying capacity of a reinforced concrete columns exposed to fire is believed to be due to: (1) Reduction in the characteristics of both concrete and steel. (2) Spalling of concrete cover. (3) Latent heating due to low thermal conductivity of concrete. (4) Plastic centroid movement due to non-uniform spalling. (5) Increasing of slenderness ratio due to spalling of concrete cover.

CHAPTER THREE

Experimental Program

3.1 Introduction

In this chapter, the details of the experimental program are presented. It elaborates the research methodology adopted in achieving the objectives mentioned in chapter one, details of the materials used, mix proportions, preparation and curing of column specimens, heating process of column specimens and testing program.

Limited previous experimental work was carried out to study the effect of fire exposure on the structural behavior of reinforced RPC column specimens, The aim of the experimental investigations was undertaken in order to explore the effect of fire heating at two adjacent sides on the structural behavior of reinforced column specimens cast with two types of concrete mixes, normal strength concrete (NSC) with compressive strength at the age of 28 days of about 40 MPa and reactive powder concrete (RPC) with compressive strength at age of 28 days of about 140 MPa. In addition, the following mechanical properties were also investigated in this research:

1) Compressive strength;
2) Splitting tensile strength;
3) Modulus of rupture;
4) Modulus of elasticity;
5) Axial deformations;
6) Load mid-height lateral deflection relationship; and
7) Ultimate load carrying capacity of the columns.

The experimental variables investigated for the column specimens

were:

1) Type of concrete: (RPC and NSC);
2) Transverse reinforcement spacing: (50, 100, 200 mm and No ties);
3) Thickness of concrete cover to the reinforcement: (15 and 30 mm); and
4) Fire temperature levels (200, 400 and 600°C).

3.2 Flow Chart of the Research

The experimental program of this research comprised of two stages. In the first stage, the selection, preparation of raw materials and evaluation the physical and chemical properties used in this study. Subsequently, this stage also performs trail mixes to adopt the optimum mix weight proportions and to select the proper mineral and chemical additives type and dosage. After that, the selected materials are then mixed with the adopted optimum mix proportions and with a suitable mixing procedure. Finally, the casted samples and column specimens cured for the required ages (3, 7, 28 and 56 days).

While the second stage consists of the heating of the column specimens and samples with different temperature levels after reaching the required age, preparing and testing of the exposed and unexposed concrete samples and column specimens after the heating process is finished. The overall experimental investigation is shown in the flowchart given in Figure (3-1). This chapter describes the testing procedure only, while the obtained results from the mentioned tests will be discussed in Chapter Four.

illustration not visible in this excerpt

3.3 Materials Selection and Its Basic Properties

It is imperative to know the properties and characteristics of constituent materials of concrete, as it is aware, concrete is a composite material made up of several different materials such as gravel, sand, water, cement and admixtures. These materials have properties and different characteristics such as (Unit weight, Specific gravity, gradation and water content).

In order to ensure adequate production of concrete, stringent procedures adopted in materials selection, controlling and proportion the whole ingredient. Under this experimental investigation, the sources of materials, chemical compositions and physical properties of the materials that were used are described in details; the necessary tests are conducted at Babylon University, College of Engineering, Civil Department laboratories. The main properties of these materials are as follows:

3.3.1 Portland Cement

The type of cement used in this study was ordinary Portland cement (type I) manufactured in Iraq by (AL-Mass Company) and taken from local markets. It was kept in a dry place to avoid exposure to different atmospheric conditions. The chemical analysis and physical test results of the used cement are given in Tables (3-1) and (3-2) respectively. This cement complied with the Iraqi Standard Specification (IQS N0.5/1984).

Table (3-1): Main components and chemical composition of AL-Mass cement.

illustration not visible in this excerpt

Table (3-2): Physical properties of AL-Mass cement.

illustration not visible in this excerpt

3.3.2 Fine Aggregate

Local natural sand from (Al-Ekhaider) region was used as fine aggregate. The grading of original fine aggregate is shown in Table (3-3). The sand was speared out and left in the air to dry before use. Results indicate that fine aggregate grading is within the requirements of the Iraqi Specification (IQS N0.45/1984). Physical and chemical properties of the fine aggregate are presented Table (3-4). The sulphate content, fine materials, specific gravity, fineness modulus and absorption of the fine aggregate were carried out in accordance with (IQS N0.45/1984). Figure (3-2) shows the grading curve of the natural sand.

The large particles of aggregate are undesirable for producing RPC. In RPC, the nominal size ranges from 150 to 600μ1η for quartz sand (fine aggregate) which are prepared by sieving to satisfy the grading according to the requirements. The separated sand was also compared with the Iraqi specification (IQS N0.45/1984), as presented in Table (3-5) and Figure (3-3). In addition, it is important to ensure that the aggregates are clean since a layer of silt or clay will reduce the cement aggregate bond strength, in addition to increasing the water demand. It must be noted that the natural sand presented in Table (3-3) was also used for casting NSC columns.

Table (3-3): The original fine sand grading compared with the requirements of (IQS no.45/1984)'

illustration not visible in this excerpt

Table (3-4): The physical and chemical properties of fine aggregate.

illustration not visible in this excerpt

Table (3-5): The separated fine sand grading compared with the requirements

illustration not visible in this excerpt

Figure (3-2): Grading curves for fine aggregate compared with requirements

illustration not visible in this excerpt

Figure (3-3): Grading curves for separated fine aggregate compared with
requirements of (IQS N0.45/1984, Zone 4).

3.3.3 Coarse Aggregate (Gravel)

For NSC columns, the coarse aggregate used was natural rounded gravel with a maximum size of 10mm, which was also obtained from the local sources in Iraq from (AL-Nibaai) region. Results revealed that coarse aggregate used in this research is complying with (ASTM C33, 2006) specifications. The gravel washed, stored in the air to dry the surface and then stored in containers in a saturated dry surface condition before use. The grading, physical and chemical properties of coarse aggregate are reported in

Table (3-6): Grading of natural coarse aggregate.

illustration not visible in this excerpt

3.3.4 Micro Silica Fume (SF)

Silica fume (SF) is a byproduct of the smelting process in the silicon and ferrosilicon industry. The reduction of high-purity quartz to silicon at temperatures up to 2000°c produces Sİ02 vapors, which oxidizes and condense in the low-temperature zone to tiny particles consisting of non­crystalline silica. The condensed silica fume then processed to remove impurities and to control particle size.

Silica fume is also known as micro silica, condensed silica fume, volatilized silica or silica dust. The American Concrete Institute (ACI) defines silica fume as a very fine non-crystalline silica produced in electric arc furnaces as a byproduct of the production of elemental silicon or alloys containing silicon. It is usually a grey colored powder, somewhat similar to Portland cement or some fly ashes. It can exhibit both pozzolanic and cementitious properties.

Plate (3-1) shows the silica fume used in this research; its particle size ranges from (0.1 to 1 pin) about 100 times lesser than the average cement particle. In concrete construction, silica fume was an active pozzolanic agent due to its large exterior area, high content of Sİ02 and its small particles. The silica fume (SF) used in this work conforms to the physical and chemical requirements of (ASTM C1240, 2015).

Densified micro silica from (FINNFJORD) Company localized in Norway was used throughout this research. Table (3-7) and (3-8) illustrated its chemical analysis and physical properties respectively of the used micro silica fume and compared with the requirements of (ASTM C1240, 2015).

illustration not visible in this excerpt

Table (3-7): Chemical analysis of silica fume used* ■

illustration not visible in this excerpt

Table (3-8): Physical properties of silica fume used*.

illustration not visible in this excerpt

3.3.5 High Range Water Reducing Admixture (HRWRA)

The high range water reducing admixture used in this study is a third generation super plasticizer for concrete and mortar, which is known commercially as (Hyperplast PC200), It is a High-Performance Super plasticizer Concrete Admixture which is imported from (DCP) company, it is aqueous solution of modified polycarboxylic polymers with long chains, free from chlorides and complies with (ASTM C494 / C494M, 2017), which is specially designed to enable the water content of the concrete to perform more efficiently by significantly improves cement dispersion, at the beginning of the mixing process the polymer chains increase the negative charge on the surface of the cement particle and dispersion of the cement occurs by electrostatic repulsion.

Production of RPC required to decrease the w/cm ratio to below 0.25 (about 0.16 to 0.2) which is only possible where using an admixture (super plasticizer), because of the fluidizing power of high-quality third generation super-plasticizing agents.

Table (3-9) presents the technical description of (Hyperplast PC200). It has many benefits such as:

Table (3-9): Technical description of (Hyperplast PC200)*.

illustration not visible in this excerpt

3.3.6 Micro Steel Fibers

It is known that plain cement concrete does not have good tensile properties to resist flexure in structural members. Therefore, the fibers are used to improve hardened concrete properties and to improve the ductility of the RPC because of the delayed initiation and slower propagation of cracks.

The micro steel fibers employed in this investigation are 13 mm length and 0.2 mm diameter, clean of rust or oil. This type of micro steel fibers was manufactured by the (Ganzimi Daye Metallic Fibers Co., Ltd, China).

The steel fibers were imported in 25 Kg bags, sample of steel fibers before and after magnification with macro lens (25-58) mm shown in Plate (3­ 2) . Straight and brass-coated micro steel fibers were used and the properties of the used steel fibers are presented in Tables (3-10) and (3-11).

Table (3-10): Chemical composition of micro steel fiber*.

illustration not visible in this excerpt

Table (3-11): Properties of Micro steel fiber.

illustration not visible in this excerpt

3.3.7 Water

Drinkable tap water available on the college campus was used in all concrete mixtures and in the curing of specimens, water was free from organic content, turbidity and salts, the temperature of the mixing water was maintained at (25±2°C).

3.3.8 Steel Reinforcement

Deformed Steel bars with (05 mm) diameter used as lateral reinforcement (ties) and (06 mm) bars used as longitudinal reinforcement as shown in Plate (3-3). The mechanical property of the reinforcement bars obtained by a digital computer complementary with the testing machine as presented in Table (3­12).

illustration not visible in this excerpt

Table (3-12): Specifications and test results of steel reinforcing bars*.

illustration not visible in this excerpt

3.4 Determination of the Workability of RPC Mixtures

The workability of RPC mixtures measured using the mortar flow table test according to (ASTM C1437, 2015), the flow table device and workability test is shown in Plate (3-4). The procedure of the test is summarized as follows:

1) The mold is filled with concrete in 25 mm layers, each layer rodded with 20 strokes with a tamping rod, in a uniform manner over the cross section of the mold, Plate (3-4-a) shows the RPC sample before the removal of the brass cone.
2) The mold is then removed from the concrete by a steady upward pull.
3) The table is then immediately dropped 25 times in 15 seconds.
4) The diameter of the spreading concrete is then measured, which is the average of four symmetrically distributed measurements read to the nearest 5 mm, as shown in Plate (3-4-b).

Results of the flow table revealed that the diameter of the spreading concrete between 205 and 215 mm is adequate in providing the required flow characteristics for RPC (110±5). The flow expressed as a percentage of the increased base diameter of the mortar mass to the original base diameter (100 mm):

illustration not visible in this excerpt

Where:

D: Average diameter of the spreading concrete measured in four directions, (mm)

3.5 Strength Activity Index for Mineral Admixture

The sustainability movement has spurred renewed interest in reducing the cement content of concrete mixtures by replacing an ever-increasing portion of the Portland cement with supplementary cementitious materials such as silica fume (Mehta, 2009), one essential characteristic of silica fume is their capability to undergo a pozzolanic reaction with the calcium hydroxide produced during conventional cement hydration to produce an increased quantity of calcium silicate hydrate gel (C-S-H), often leading to long term benefits, including increases in compressive strength.

Test methods for detection of pozzolan activity are divided into two broad categories: direct methods measuring consumption of Ca(OH)2, and indirect methods measuring the relative compressive strength of test specimens, strength activity index (SAI).

Indirect method (SAI) was used in this research. SAI is in principle determined by molding two sets of test specimens according to a standard recipe: one with 100 % Portland cement (reference), and another where a standardized part of the Portland cement is replaced by a corresponding mass of the pozzolan to be tested. The strength activity index of the used micro silica fume utilized in this research was tested according to (ASTM C1240, 2015) and (ASTM C311, 2017), in which The specimens were prepared for strength activity index (Pozzolanic activity) consisted of one part of cement and 2.75 part of standard graded sand by weight. The mix proportions of the reference specimens and the test specimens have been described in Table (3­13).

Table (3-13): Mix proportions of the reference and test samples.

illustration not visible in this excerpt

In this test six cubes (50x50 mm) molds were used, three for the referee mixture and three for test mixture. The cubes demolded after 24 hours of initial curing in the wet room with a temperature of (23 ± 2°C) and relative humidity of not less than 95%, then the cubes placed in airtight glass containers and stored for six days at ( 65± 2°C). The specimens tested as specified in testing method C109/C 109M to determine the compressive strength at 7 days after molding. The result of the tested samples used according to (ASTM c 311, 2017) equation to calculate the Pozzolanic Activity Index (P.A.I). Table (3-14) shows w/c or w/cm ratio and pozzolanic activity index.

illustration not visible in this excerpt

A: Average compressive strength of standard mortar of cement, sand and pozzolanic material at given age, (MPa).

B: the Average compressive strength of control mortar at the same age as that of age A, (MPa).

Table (3-14): Pozzolanic activity index and w/c or w/cin ratios for test mortars.

illustration not visible in this excerpt

3.6 Concrete Trial Mixes

The trial mixture approach is best for selecting proportions for RPC. In order to obtain high strength, it is necessary to use a low water to cementing materials ratio and a high Portland cement content. The unit strength received for each unit of cement used in a cubic meter of concrete can be plotted as strength efficiency to assist with mix design.

A number of mixing proportions from the previous literature were taken into consideration and playing around with these mix proportions, numerous mix proportions were established.

In this kind of concrete, homogeneity is improved by using a powder concrete in which aggregates and traditional sand are replaced by ground quartz less than 600 microns in size. The high cement content of ordinary Portland cement, with a different amount, is used to achieve greater strength. Another significant characteristic is applying low water-cement ratio in the range of 0.11-0.20. Different amount of chemical admixtures are used to improve the workability of the fresh concrete in so low water-cement ratio. Moreover, high volume silica fume (SF) more than 20 percent, is used to work as the filler, increase the density, in addition, some chemical reactions in the particular curing condition occur. These limitations were used to design the trials mixes proportion of the RPC to be employed in this research in order to achieve the maximum compressive strength and flow (110%±5) according to (ASTM C109, 2016) and (ASTM C1437, 2015) respectively. Fifteen RPC

Table (3-15): Continued.

illustration not visible in this excerpt

Table (3-16) and Figures (3-4) and (3-5) show the optimum mixes used in this study with weight proportions.

Table (3-16): Details of the mixes used in this research.

illustration not visible in this excerpt

Figure (3-4): Materials proportion used in RPC mixture (% of total mix weight).

illustration not visible in this excerpt

Figure (3-5): Materials proportion employed in NSC mixture (% of total mix weight).

3.7 Description of the Tested Column Specimens

The experimental program consisted of testing (48) column specimens of RPC and NSC. (32) Reinforced RPC column specimens cast in the laboratory were tested under (concentric) loading and compared with the reference (16) reinforced NSC column specimens. The selected dimensions were fixed for all column specimen at (900mm), and a square cross section of (100x100 mm), each column was reinforced with 406mm as longitudinal reinforcement and 05mm bars were used for transverse reinforcement (ties) with different spacing as shown in Figure (3-6). Column specimens were divided into four groups depending on fire temperature levels, each group consisted of (8) RPC columns and (4) NSC columns depending on concrete cover and transverse reinforcement (ties) spacing as shown in Table (3-17).

3.7.1 Column Specimens Identification

Due to a large number of column specimens investigated in this research and in order to facilitate the comparison between these columns, each RPC column is identified by three symbols. The first symbol is letters that refer to the concrete type. The letters (RPC) and (NSC) refer to reactive powder concrete and normal strength concrete respectively. The second symbol is a number refers to the concrete cover (C.C); (1) and (2) for (15 and 30 mm) concrete cover respectively. While, the third symbol is a number that refers to the tie spacing; (0) for no ties, (1), (2) and (3) for (200, 100 and 50 mm) tie spacing respectively.

Table (3-17): Summary of RPC and NSC column test specimens.

illustration not visible in this excerpt

Figure (3-6): Reinforcement details of the reinforced column specimens.

illustration not visible in this excerpt

3.8 Mixing Procedure

An important factor for studying new cementitious materials is the mixing procedure, due to the nature of the raw materials comprising RPC and the high mix quality required, the conventional mixing procedures are not entirely sufficient. The extended mixing time is necessary both to fully disperse the silica fume, breaking up any agglomerated particles and to allow the super-plasticizing agent to develop its full potential. All trial mixes were performed in a small rotary mixer of (0.01m[3]) capacity, while the specimens' mixes were performed in a horizontal rotary mixer with a capacity of 0.09 m[3] as shown in Plate (3-5). Any residual particles of concrete from prior batch must be cleaned off before using the concrete mixer. A moist cloth is used to clean the blades and pan of the concrete mixer.

illustration not visible in this excerpt

The adopted method in this research, presented in Figure (3-7), depended on previous researches and approved during trail mixes to produce satisfactory workability and good dispersion of fiber without evidence of fiber balling. After finishing mixing the workability of fresh concrete was performed by using flow table test before casting of samples and column specimens.

The following sequence in mixing RPC proposed by (Wille et al., 2011) was used:

1) Dry mixing of fine sand and silica fume for 3 minutes.
2) Adding cement and dry mixing of the components for about 2 minutes to break the agglomerates.
3) Addition of water containing of superplasticizers gradually during mixing for about 4 minutes.
4) When the flowable consistency achieved the steel fiber added slowly during the operation of the mixer to insure the uniformly dispersion which takes about 5 minutes.
5) The whole mixing process takes about 14 minutes.

Mixing the Fine Sand and Micro Silica

Adding the Cement and Mixing the Dry
Components

Addition of Water Containing of
Superplasticizers Gradually During Mixing

Adding the Steel Fiber Slowly During the
Operation of the Mixer

End of Mixing and Ready of Pouring

illustration not visible in this excerpt

Figure (3-7): RPC mixing time process.

3.9 Casting and Curing Procedures of the Specimens

Before casting, the selection materials was prepared and weighed according to the volume of the mix, All columns test specimens used in this research were cast in steel molds, the specimens cast in a short period to mitigate the variation in exposure condition, the steel molds have a cross section of (100x100 mm) and a clear height of 900mm as shown in Plate (3­

6) . Each batch from the mixer was sufficient to cast three columns as shown in Plate (3-7).

The process of casting is described by the following stages:

1) The steel molds were prepared by proper cleaning and lubricating the internal faces lightly by oil to prevent the adhesion with hardened concrete before each casting and placed on a horizontal ground.
2) The longitudinal reinforcement was positioned carefully inside the steel molds with the required bottom and sides cover being accurately maintained as shown in Plate (3-8). The concrete cover was provided by using steel reinforcement bars at the ends of column reinforcement instead of plastic spacer to prevent melting during fire exposure as shown in Plate (3-9).
3) After the mixing was finished, the fresh concrete mixture poured into the molds by placing the specific concrete directly from the mixer in three layers with each layer being vibrated using a vibrating table and full compaction was made sure by observing the air bubbles on the surface as presented in Plate (3-10).
4) Soon after casting and finishing the top surface of all specimens, the molded specimens were covered (to prevent loss of moisture) and left in the casting room at (25±2°C) for 24 hours until de-molding as shown in Plate (3-7).

With each mix, samples were prepared to determine the mechanical properties of concrete involve 3 cubes (50x50x50 mm) for compressive strength, 3 cylinders (100x200 mm) for splitting tensile strength, 3 cylinders (100x200 mm) for modulus of elasticity and 3 prisms (50x50x300 mm) for flexural strength.

One of the principles for developing RPC as mentioned before (in chapter two) is the heat treatment curing of RPC, but this method was not adopted in this study in order to gain an advantage of producing RPC of exceptional mechanical properties (compressive strength more than 140 MPa) using conventional curing method without any additional provisions and also to simulate the practical site conditions. Thereafter, column specimens were cured with saturated wet coverings by using burlap, while samples were cured in a water tank with a temperature of (23±2 °C) as shown in Plate (3-11).

illustration not visible in this excerpt

3.10 Heating Process

The main purpose of the lire exposure furnace is to raise the temperature levels of concrete models to the target temperature and holds the temperature constant for a required duration. In order to control the heating procedure, heating process consists of the following equipment:

1) Brick furnace.
2) Network .dZlclrbO 30๙( .Cİ+ /, g burners.
3) Thermocouple.
4) Electrical gas regulator control.
5) Digital gage.
6) Electrical network.
7) Gas bottle.
8) Gas connections and pipelines.
9) furnace steel cover.
10) Ignition tool.

The full details of the furnace and equipment are shown in Figure (3-8).

illustration not visible in this excerpt

The furnace is manufactured with dimensions of (1500x1500x1250 mm) (lengthX widthX height respectively) and the wall thickness in all sides are 250 mm, the main structure is made of Perforated brick and mortar with small opening to provide the necessary fresh oxygen for the burners, the furnace v( ) c* is made of insulator plate with 8 mm thickness to keep the temperature constant as shown in Plate (3-12). The network of burners consists of eight liquefied petroleum gas burners arranged in two lines at each side of exposure (4 burner at each side) and distributed along the length of the column,, 11 liquefied petroleum gas burners connected together in one pipeline to control /, g0<įyf, */ c and connected to the gas bottle. The bars of flame were intended e( gcE 1., ec the heating condition in an actual fire.

illustration not visible in this excerpt

Figure (3-8): Illustrates the !limace and equipment.

illustration not visible in this excerpt

Plate (3-12): Details of the brick furnace.

3.10.2 Thermocouple

Thermocouple is a temperature sensing element consists of two wire legs made from different metals inside a metal tube (thermocouple cover). The wires legs are welded together at one end, creating a junction. This junction is where the temperature is measured. When the junction experiences a change in temperature, a voltage is created, and this voltage can be interpreted to measure temperature. Thermocouple usually connected to a thermometer or any other thermocouple-capable device (digital gage used in this investigation) by a thermal-insulated electric wire to resist high temperatures during heating. Plates (3-13) and (3-14) show the details of thermocouple and the digital gage.

illustration not visible in this excerpt

Plate (3-13): Thermocouple details.

illustration not visible in this excerpt

Plate (3-14): Digital gage details.

3.10.3 Electrical Gas Regulator and Ignition-Heating System

Gas regulator device is an electronic valve designed to provide very fast and accurate gas flow for the furnace burners. The electronic circuit compares the gas injection pressure to its set point of temperature and regulates the gas flow in order to maintain the temperature at its set point by closing or opening the gate. The electrical gas regulator is connected to digital gage to provide gas flow according to the set temperature requires, these connections permit to remain fire in burners as an ignition system during the firing process. Ignition equipment usually remains ignited during gas locking intervals to be ready for reigniting process again when temperature degree reduces below the level required. Plate (3-15) shows the electrical gas regulator and ignition­heating system.

illustration not visible in this excerpt

Plate (3-15): Electrical gas regulator.

3.10.4 Heating Procedure of the Specimens

After curing was finished the heating procedure was applied. The specimens were heated with direct fire exposure by a net of burners inside a brick furnace to the target temperature. The heating procedure was done by tf c following steps:

1) After preparing the furnace with all connections, the concrete specimens were held carefully by hand, weighed and placed inside the furnace.
2) The concrete specimens were positioned fair-faced toward the 41*-c*g with the same distance from each line of burners.
3) The gas valve was opened and the burners was ignited by an ignition tool, the measured temperature was increased gradually.
4) When the target temperature reached, the electrical gas regulator started to work by regulating the gas flow so the target temperature stayed constant, the target temperature was measured by thermocouple placed in direct contact with fire.
5) The target temperature and the temperature of concrete were also measured by applying infrared ray by infrared ray thermometer from about approximately one meter from the exposed concrete surface as shown in plate (3-16).
6) After the duration of heating was finished (1 hour of fire exposure) the gas valve was closed and the specimens were cooled immediately.

The full details of the heating process and heating furnace with all connections are shown in Plate (3-17) and Figure (3-9).

illustration not visible in this excerpt

Plate (3-17): The full details of the heating process and furnace with the connections.

illustration not visible in this excerpt

Figure (3-9): Top view and section view of the furnace and equipment.

3.11 Cooling Procedure

After the duration of heating was finished, the gas valve was switched off, the concrete specimens were immediately extinguished by foam spray fire extinguisher and then lifted from furnace by using thermal gloves as shown in Plate (3-18). This process of cooling adopted in this investigation in order to simulate this problem to practical site conditions.

illustration not visible in this excerpt

Plate (3-18): Extinguishing of column by foam spray fire extinguisher.

3.12 Testing Procedure of Reinforced NSC and RPC Columns

One day before testing, the column specimen surface was cleaned then, the column specimen was placed in its position for testing, and making sure that it is in a vertical position, then the dial gauges were fixed in their position. During testing, the main characteristics of the structural behavior of the specimens were detected at every stage of loading. At each test, axial deformation, mid-height lateral displacements, and ultimate load were recorded.

All of the column specimens were tested at the age of 56 days under increasing concentric load up to failure by using a calibrated electro-hydraulic testing machine with a maximum range capacity of 2500 KN as shown in Figure (3-10).

Two dial gauges having accuracy of 0.001 mm per deviation and capacity of 20 mm and a maximum sensor length of 50 mm were used to measure the lateral deflection at the mid-height and axial deformation. These dial gauges were mounted at the mid-height along the vertical center line of the column specimens and on the piston of electro-hydraulic testing machine as shown in Figure (3-10). For each column the verticality was checked accurately at both sides using a spirit level, before applying the load as shown in Plate (3-19). The load was applied in small increments and the reading of displacements versus load were recorded simultaneously for each load increment by connecting multiple HD video Cameras to a computer which recording these parameters until failure occurred.

3.13 Testing of Control Samples

After selecting the mixes for NSC and RPC of column specimens, cubes, cylinders and prisms were casted in order to test it at ages of (3,7,28 and 56 days) at room temperature and after exposure to fire of 200, 400, 600)°c and with the same procedures and conditions of heating of column specimens. These testing consisted of cube compressive strength (/cu), modulus of elasticity (Ec), splitting tensile strength ifsp) and modulus of rapture ifr). All the tests were carried out at Babylon University, College of Engineering, Civil Department, Construction Materials laboratory.

3.13.1 Compressive Strength Test

The compressive strength was an established measure that represent one of the important engineering properties of concrete, which could provide an overall image of the quality of concrete. Three cubes of (50x50x50 mm) for each mix were cast and burned in a similar way as for column specimens and with different temperature levels to determine the compressive strength, and an average value of these cubes was obtained according to (ASTM 009/ C109M, 2016).

The compressive strength test was determined by crushing three cubes at ages of (3, 7, 28 and 56 days) using a digital testing machine with a capacity of (1900 kN), the loading rate was applied at (0.3 MPa/sec). Plate (3-20) shows the compressive strength test machine.

illustration not visible in this excerpt

Plate (3-20): Compressive strength testing machine.

3.13.2 Splitting Tensile Strength

Splitting tensile strength of concrete has been standardized by (ASTM C496, 2017) specification. Three (100x200 mm) cylindrical concrete specimens performed for each mix and each ages (3, 7, 28 and 56 days), cast and burned in a similar way as for column specimens and with different temperature levels. Two bearing strips of 3 mm thick of plywood, 25 mm wide and 200 mm length were placed below and above the specimen which was placed between the bearing blocks of the testing machine, an electrical testing machine with a capacity of 1900 kN as shown in Plate (3-21).

The splitting tensile strength ifsp) of specimen is calculated by the following Equation:

illustration not visible in this excerpt

fsp : Splitting tensile strength (MPa), p : Maximum applied load in splitting test (N).

D : Diameter of cylinder (mm).

L : Length of cylinder (mm).

illustration not visible in this excerpt

Plate (3-21): Splitting tensile strength test.

3.13.3 Modulus of Rupture

The concrete modulus of rupture was determined by testing (50x50x300 mm) prisms samples in conformity with (ASTM C78, 2016). The prisms were cast and burned in a similar manner as the column specimens, Each prism was simply supported and subjected to a two point loading using testing machine with a capacity of ( 1900 kN) as shown in Plate (3-22), and the average modulus of rupture (often referred to as flexural tensile strength) of three prisms for each age (3, 7, 28 and 56 days). The ultimate modulus of rupture {fr) was calculated by using the following Equation.

illustration not visible in this excerpt

fr: Modulus of rupture, (MPa). p: Failure load, (N).

L: distance between the supports, (mm). b: Width of prisms, (mm). h\ Depth of prisms, (mm)

3.13.4 Static Modulus of Elasticity

The static modulus of elasticity of concrete was performed on (100x200 mm) cylindrical specimens and according to (ASTM C469, 2014). The top surface of cylinder was well finished and smoothed by using electric grinding machine to prevent any loss of strength. The specimens were tested after heating in a similar manner as the column specimens by using a hydraulic machine with a capacity of (1900 kN). The 40% of ultimate compressive strength of concrete specimen was applied on the concrete cylinders to accomplish the elastic modulus test as stated by as illustrated in Plate (3-23). The specimens were tested at ages (3, 7, 28 and 56 days) and the average of three specimens were taken. Static modulus of elasticity was computed by the following equation:

illustration not visible in this excerpt

Plate (3-23): Modulus of elasticity test device.

CHAPTER FOUR

Results and Discussion

4.1 Introduction

The main objective of this study is to investigate experimentally the behaviour of reinforced reactive powder concrete (RRPC) columns exposed to fire at two adjacent sides and subjected to axial load.

In this chapter, the results of the experimental program illustrated in chapter three are presented and discussed. The results of testing forty-eight concrete column specimens are discussed in this chapter. The specimens included thirty-two RPC columns along with sixteen normal strength concrete (NSC) columns. The experimental results compared with each other to investigate the significance of the considered experimental variables. The evaluated experimental variables test results were the ultimate load carrying capacity, failure behaviour and the load displacement relationships for each loading increments. Based on the estimated load-displacement relationships, ductility, stiffness and energy absorption were calculated and compared for the exposed and non-exposed RPC and NSC columns to assess experimentally the residual ductility and stiffness of concrete columns after exposed to real fire. The considered experimental variables were the type of concrete (RPC and NSC), transverse reinforcement spacing, the thickness of the concrete cover and temperature levels of fire exposure.

This chapter also includes the results of the tests conducted on the associated control samples to specify their mechanical properties. Additionally, some physical changes have been recorded in order to study and follow the behaviour of the reactive powder concrete under real fire exposure conditions. These changes included the appearance and colour vf, -/ c of RPC.

4.2 Mechanical Properties of Hardened Concrete

Studying the mechanical properties of concrete considered an imperative process to comprehend the behaviour of the concrete column specimens. The mechanical properties of hardened concrete conducted in this study consist of compressive strength, splitting tensile strength, modulus of rupture and modulus of elasticity. These properties are investigated in order to estimate the resistance and capability of RPC under fire exposure. The heating process of the RPC samples was similar to the heating process of column specimens and with the same duration (1 hour).

4.2.1 Compressive Strength

Compressive strength test is the most common among all tests on hardened concrete. It is used to estimate the potential strength of concrete. The characteristic of concrete compressive strength is used to classify the concrete in the national and international codes. The ultra-high compressive strength is the most significant property of RPC. The test of compressive strength (feu) was carried out using (50x50x50) mm cubes for RPC; these cubes were cast using RPC with 2% micro steel fibre content. The cubes were exposed to fire at different temperature levels (200, 400 and 600°C) and tested to determine the compressive strength. The results were compared with RPC cubes tested at laboratory temperature (25°C) at ages of (3, 7, 28 and 56) days.

Table (4-1) shows the compressive strength of reactive powder concrete before and after heating. Each value in this Table represents the average value of (3) cubes in order to minimize the expected error in any measured result.

It is cleared from the results of Table (4-1), that RPC gains strength rapidly during the early age (RPC at 25 °c gained 55.2% of the compressive strength in 3 days and continues until about 69.3% of the compressive strength achieved at 7 days). However, the strength gain continued until the end of the curing but in a slower rate of strength development. These results also observed by other researchers (Kadhum, 2015) and (Al-Jabiri, 2015).

The results of compressive strength tests for RPC, which illustrated in Table (4-1) and Figure (4-1) show that temperatures can be divided into two ranges in terms of the effect on RPC’s strength, namely (25-200°C) and (200-600°C).

At 200°c the compressive strengths /си of all RPC mixes are slightly higher than the corresponding compressive strength for 25°c. The increasing ratios for all ages are (3.1, 4, 4.3 and 4.8%), for (3, 7, 28 and 56 days) respectively. The increase in compressive strength in the range of (25- 200°C) might be due to the internal autoclaving for RPC since RPC was cured normally, which leads to further hydration of cement with silica fume as a pozzolanic reaction. This reaction was resulted in increasing the hydration products and reduced the pore size (Kadhum, 2015). In addition, silica fume can react with cement hydrates to perform C-S-H. The structure of C-S-H was seemed to be complete and closely knit. These chemical changes were found to be well affected in the activate changes of the hydrates microstructure (Tai, et al., 2011), chemical changes lead to increase in the compressive strength at this range of temperature.

At both 400°c and 600°c, the compressive strength decreases. The decreasing ratios at 400°c were (35.1, 33.3, 30.9 and 29.6%), for (3, 7, 28 and 56 days) respectively. The reason for this is the loss of water from the hydrated cement paste and possibly internal collapse (Tai, et al., 2011). While at 600°c the decreasing ratios are (69.4, 65.3, 62 and 58.7%), for (3, 7, 28 and 56 days) respectively. This may be attributed to the decomposition of calcium hydroxide (Neville, 1995).

On the other hand, the test results of the residual compressive strength after heating show that the reduction at 3 days-age was more than the reduction at the ages (7, 28 and 56 days). This may be attributed to the fact that the hydration of cement paste is more finished at later ages as shown in Figure (4-2).

The results of residual compressive strength after exposure to fire compared to the results of the studies carried out by other researchers are presented in Figure (4-3). The recorded results of the residual compressive strength ratios in the present study indicated a close behaviour with (Mohammed, 2016), and semi identical behaviour with (Kadhum, 2015) and (Al-Jabiri, 2015).

The appearance of the cube samples after exposure to fire at different temperature levels are illustrated in Plate (4-1), the images of the exposed samples showing different appearances of spalling at various temperature levels captured by HD digital camera.

Table (4-1): Test values of cube compressive strengths of RPC samples before and after exposure to fire.

illustration not visible in this excerpt

Figure (4-3): A comparison of compressive strength values at 28 days for

illustration not visible in this excerpt

Plate (4-1): The appearance of cube samples after exposure to fire at different temperature levels.

4.2.2 Splitting Tensile Strength

The values of splitting tensile strength ifsp) of the specimens considered in the present investigation are summarized in Table (4-2). The tensile strength governs the cracking behaviour and it affects other properties such as stiffness and durability of RPC. Splitting tensile strength was performed by using a cylinder with dimensions (100x200) mm at ages (3, 7, 28 and 56 days), these cylinders were cast using RPC with 2% micro steel fibre content. By applying the same procedure used in the calculation of compressive strength, each test was conducted by three cylinders.

It is found from the results of Table (4-2) that the splitting tensile strengths of RPC generally increases with age. It is also noted that the rate of strength development of the RPC is higher during the early ages (RPC at 25 °c gained 63.5% of the splitting tensile strength in 3 days and continues until about 89.3% of the splitting tensile strength achieved at 7 days), which is similar to the observation of the compressive strength in this study. However, the tensile strength of concrete is much less, than the compressive strength and this attributed to the ease propagation of cracks under tensile loads (Agha, 2015). Increasing of strength with time is the result of the progress of the hydration process of the matrix through the completion of its inner microstructure leading to additional strength added to the composite material, along with the increase in the bond strength between the fibres and the matrix (Agha, 2015). As in compressive strength, the effect of the fire exposure at age of 3 days was more than at (7, 28 and 56 days). This is due to further completion of hydration of the cement at later ages.

The results of splitting tensile strength tests for RPC, which illustrated in Table (4-2) and Figure (4-4), show that at 200°c the splitting tensile strengths /sp for all ages are slightly higher than the corresponding splitting tensile strength for 25°c. The increasing ratios for all ages are (3, 3.5, 4.6 and 5%), for (3, 7, 28 and 56 days) respectively. This increasing might be due to dry hardening and internal autoclaving of the un-hydrated cement and silica fume particles since RPC was cured normally (Kadhum, 2015).

At both 400°c and 600°c, the splitting tensile strength decreases. The decreasing ratio at 400°c is (37.2, 36.6, 36.1 and 34.6%), for (3, 7, 28 and 56 days) respectively. While at 600°c the decreasing ratios are (64.4, 63.6, 62.9 and 61.6%), for (3, 7, 28 and 56 days) respectively. The reason for this is the same reason of the compressive strength decreasing which is the loss of water from the hydrated cement paste and possibly internal collapse (Tai, et al., 2011), or due to the decomposition of calcium hydroxide (Neville and Brooks, 2010). It is also possible that the splitting tensile strength decreasing is may be due to the quick effects of fire on fiber rather than on concrete, so the fiber will be heated faster than the other components, so that the connection area between the fiber and the concrete will be affected and being weak (Al-Jabiri, 2015).

Researchers also study the effect of heating on the splitting tensile strength of RPC. Figure (4-5) shows their results and it is different with the current research.

The appearance of cylinder samples after exposure to fire at different temperature levels are illustrated in Plate (4-2), different appearances of spalling can be seen from the images of the exposed samples at various temperature levels captured by HD digital camera.

Table (4-2): Test values of cylinders splitting tensile strength for RPC samples before and after heating■

illustration not visible in this excerpt

Figure (4-4): The effect of fire exposure on the splitting tensile strength of RPC.

illustration not visible in this excerpt

Figure (4-5): Comparison of residual splitting tensile strength for the present study and other researchers.

illustration not visible in this excerpt

Plate (4-2): The appearance of cylinder samples after exposure to fire at different temperature levels.

4.2.3 Flexural Strength (Modulus of Rupture)

Modulus of rupture test was used to express the flexural strength. Each prism was tested as simply supported and subjected to two-point loading. Table (4-3) presents the tested results in which each value in this table represents the average value of (3) prismatic samples in order to minimize the expected error in any measured result.

In the same manner of the observation in the compressive strength and the splitting tensile strength in this study. It is found that the rate of flexural strength development of the RPC is higher during the early ages until about 74.8% of the flexural strength achieved at 7 days, then the strength gain continued until the end of the curing but in a slower rate of strength development. It is clear from the test results illustrated in Table (4-3) and Figure (4-6) that the effect of the fire at age of 3 days was more than at (7, 28 and 56 days). As the compressive strength, this was due to further completion of hydration of the cement at later ages.

After fire exposure at 200°c, the flexural strengths iff) of all RPC mixes are slightly higher than the corresponding flexural strength for 25°c. It is observed from the results that the increasing ratios are (3.6, 4.5, 5.9 and 7%), for (3, 7, 28 and 56) days respectively. This increasing might be due to dry hardening and the internal autoclaving for RPC since RPC was cured normally. At both 400°c and 600°c fire temperature, the flexural strength decreases. The decreasing ratio at 400°c is (39.9, 38.9, 36.8 and 34.5%) for (3, 7, 28 and 56 days) respectively. While at 600°c, the decreasing ratios are (71.1, 69.4, 66.8 and 63.9%) for (3, 7, 28 and 56 days) respectively. The reason for this is the same reason of the compressive strength decreasing which is the loss of water from the hydrated cement paste and possibly collapsed by internal stresses and cracks.

The results of residual flexural strength after exposure to fire compared to the results of the studies carried out by other researchers (Mohammed, 2016), (Al-Jabiri, 2015) and (Kadhum, 2015) are presented in Figure (4-7).

Table (4-3): Test values of prisms flexural strength of RPC samples before and after heating.

illustration not visible in this excerpt

Figure (4-6): The effect of fire Ftuvwx*F on the flexural strength of RPC.

illustration not visible in this excerpt

Figure (4-7): Comparison of residual flexural strength for the present study and other researchers.

illustration not visible in this excerpt

4.2.4 Static Modulus of Elasticity

The static modulus of elasticity (Ec) is a material property that describes its stiffness and is, therefore, one of the most important properties of solid materials.

The modulus of elasticity is strongly influenced by the concrete materials and their proportions. An increase in the modulus of elasticity is expected with an increase in compressive strength since the slope of the ascending branch of the stress-strain diagram becomes steeper (Ramezanianpour, 2009).

Figure (4-8) depicts the relationship between modulus of elasticity and fire temperature levels for RPC samples. From the results of Table (4­4), it can be concluded that the reduction in values of modulus of elasticity is more significant than in compressive strength at identical fire exposure temperatures. The reduction of the modulus of elasticity is due to the breakage of bonds in the microstructure of the cement paste and the differential movement between the cement paste and the aggregate, which take place when the RPC subjected to high temperatures of firing process (Kadhum, 2015).

At heating temperature of 200°c, the residual modulus of elasticity of all RPC mixes are slightly higher than the corresponding modulus of elasticity for 25°c. The increasing ratios for all ages are (2.8, 3.5, 5.8 and 6.8%), for (3, 7, 28 and 56 days) respectively. This was due to improving in strength compared with the original as a result from the more hydration of cement with silica fume as a pozzolanic reaction. This reaction was resulted in increasing the hydration products and reduced the pore size. The reductions in modulus of elasticity after exposure to fire temperature 400°c were (49.1, 45.7, 42.5 and 39.8%), for (3, 7, 28 and 56 days) respectively.

While at 600°c were (79.8, 78, 76.4 and 75%) for (3, 7, 28 and 56 days) respectively.

Experimental studies carried out by other researchers compared with the results of the residual modulus of elasticity in the present study are given in Figure (4-9). The recorded results of the residual modulus of elasticity ratios in the present study indicated a close behaviour with (Mohammed, 2016), and semi identical behaviour with (Kadhum, 2015) and (Tai et al., 2011).

Table (4-4): Test values of modulus of elasticity of RPC specimens before and after heating.

illustration not visible in this excerpt

Figure (4-8): The effect of fire exposure on the modulus of elasticity of RPC.

illustration not visible in this excerpt

4.3 Description of Experimental Program of Column Specimens

In order to evaluate the structural behaviour of the tested column specimens the main structural characteristics observed and recorded during the test of each column specimens and at each stage of loading. Electrohydraulic jack compression machine with 2500kN capacity used to apply the load on the concrete column specimens. The observed characteristics are the axial displacement, mid-height lateral displacement and the ultimate failure load. The details of tests results of RPC and NSC columns that include the Load carrying capacity of columns, the change ratios of the load carrying capacity after fire exposure, the percentages of residual load carrying capacity of columns after fire exposure, the axial displacements measured over the full height of the specimen and corresponding lateral displacements at the mid-height of the specimen are given in Tables (4-5).

Table (4-5): Experimental test results of RPC and NSC column specimens.

illustration not visible in this excerpt

4.4 Load Carrying Capacity of RPC and NSC Column Specimens

The load carrying capacity reflected the ultimate applied load that can be subjected to the tested column specimens, after that a drop in machine reading appeared with a rapid deformation on the column, which termed as a failure. In order to investigate the effect of fire exposure on the load carrying capacity of column specimens, the behaviour of RPC and NSC columns (group one) prior to heating must be studied and summarized as the following:

4.4.1 Behaviour of RPC and NSC Columns at Normal Temperature (Group One)

As mentioned previously the tested column specimens were twelve reference columns divided into eight RPC columns and four NSC columns. Figures (4-10), (4-11) and (4-12) reveal the effect of the spacing and the amount of transverse reinforcement (ties) on load carrying capacity of RPC and NSC column specimens. The basic philosophy of the ACI 318-14 Building Code for lateral confinement is that the increase in the strength of the core of a column due to confinement must offset the loss of strength due to spalling of the unconfined cover. It can be seen from these Figures that the reducing the tie spacing from (200 to 50 mm) increased load carrying capacity by (19 and 16.5%) for RPC columns with concrete cover (C.C) of (15 and 30 mm) respectively, and by (28 and 21%) for NSC columns with (15 and 30 mm) concrete cover (C.C) respectively. It can be concluded from these results that the transverse reinforcement is slightly less effective for RPC column specimens than NSC column specimens. The reason for this is believed to be due to that the stress in the transverse reinforcement at peak load for higher strength concrete columns is often significantly less than the yield strength of the transverse steel reinforcement (Ahmad and Shah, 1979).

It is also indicated from Figures (4-10), (4-11) and (4-12), that the value of load carrying capacity of RPC and NSC columns before heating increases with reducing the concrete cover. The reducing of the concrete cover (C.C) from (30 to 15 mm) for RPC columns with different tie spacing (No Ties, 200, 100 and 50 mm) gave an increase in load carrying capacity of (8, 12, 13 and 14%) respectively. While for NSC columns with tie spacing of (200 and 50mm) the increase in load carrying capacity was about (15 and 22%) respectively, as shown in Figure (4-13). The reason for this is believed to be due to that reducing the concrete cover (C.C) for the same cross-section of the column specimen will result in a larger confined concrete area and reflects the higher confinement efficiency especially in concrete columns with closer spacing of transverse reinforcement.

Figure (4-10): Effect of the transverse steel reinforcement on the load carrying capacity of reference RPC columns.

illustration not visible in this excerpt

Figure (4-11): Effect of the transverse steel reinforcement on the load carrying capacity of reference NSC columns.

illustration not visible in this excerpt

Figure (4-12): Effect of the transverse reinforcement on the percentage of increase in load carrying capacity of reference RPC and NSC columns.

illustration not visible in this excerpt

Axial Spacing of Transverse Rinforcement (mm)

Figure (4-13): Effect of concrete cover on the load carrying capacity of reference RPC and NSC columns with different transverse reinforcement.

4.4.2 Fire Exposed RPC and NSC Column Specimens

In this section, twelve column specimens similar to those investigated in group one (reference columns) and with the same parameters are exposed to fire in order to investigate the effect of variables study mentioned in chapter three on the structural behaviour of RPC and NSC columns after fire heating. The experimental program consisted of three groups of the exposed column specimens for each temperature level (200, 400 and 600°c).

4.4.2.1 Group Two (Heating at 200°c Fire Temperature)

From the test results of Table (4-5), it can be seen that the value of load carrying capacity of the exposed RPC columns at 200°c was slightly higher than those of the reference RPC columns. The increasing ratios of load carrying capacity were almost identical, with slightly higher increasing ratios for RPC columns with 30 mm concrete cover, see Figure (4-14).

On the other hand, the test results stated that the load carrying capacity values of the exposed NSC columns at 200°c were decreased. It can be по observed from Figure (4-14) that the decreasing ratios of load carrying capacity of NSC columns with 30 mm concrete cover were slightly lower than these of NSC columns with 15mm concrete cover, which gives an impression of the positive effect of the concrete cover for protecting the concrete column and its reinforcement.

As well as, the same observation was noticed with reducing the tie spacing of the exposed NSC columns, in which the decreasing ratios of load carrying capacity after burring at 200°c were lower.

illustration not visible in this excerpt

Axial Spacing of Transverse Reinforcement (mm)

Figure (4-14): The percentage of change in load carrying capacity of the
exposed RPC and NSC column specimens at 200°c fire temperature.

4.4.2.2 Group Three (Heating at 400°c Fire Temperature)

The test results presented in Table (4-5) showed that the load carrying capacity of exposed RPC columns at 400°c was decreased significantly. The decreasing ratios for RPC columns with 15 mm concrete cover (C.C) and with tie spacing of (No Ties, 200, 100 and 50 mm) were (26.0, 24.5, 22.2 and 21.8%) respectively. While, For RPC columns with 30 mm concrete cover the decreasing ratios were (25.1, 23.1, 20.6 and 19.3%) respectively, as shown in Figure (4-15).

On the other hand, the test results indicated that the decreasing ratios of load carrying capacity for NSC columns after heating at the same temperature level were lower. NSC columns with 15 mm concrete cover and tie spacing of (200 and 50 mm) lost only about (13.1 and 10.2%) respectively. While NSC columns with 30 mm concrete cover and tie spacing of (200 and 50 mm) lost about (9.8 and 6.2%) respectively. In addition, Figure (4-15) accentuates that the contribution of increasing concrete cover (C.C) in improving the fire resistance of RPC columns after heating at 400°c was insignificant. While for NSC columns the increase in concrete cover (C.C) to the reinforcement has a slight positive impact on the reduction in load carrying capacity.

The reason may attribute to that, RPC due to its low water-cement ratio and its low permeability compared to that of NSC, will experiences significant spalling which is attributed to the build-up of pore pressure during heating which cannot escape because of the high density (and low permeability) of RPC, spalling results in the rapid loss of the surface layers of the concrete and exposes the core concrete to fire temperatures, thereby increasing the rate of transmission of heat to the core concrete and the reinforcement.

The effect of spacing and amount of ties on the load carrying capacity are also clarified from this Figure. It can be indicated that the effect of transverse reinforcement in improving the fire resistance of RPC columns has a slightly better influence on the reduction in load carrying capacity than the effect of increasing concrete cover. The reason may also attribute to the serious spalling of the concrete cover for RPC columns which takes place immediately after the fire temperature exceeds 300°c which results in the rapid loss of the concrete cover thus the fire endurance of the core concrete depends mainly on the efficiency of the confinement.

illustration not visible in this excerpt

Figure (4-15): The percentage of reduction in load carrying capacity of the

RPC and NSC columns after heating at 400°c fire temperature.

4.4.2.3 Group Four (Heating at 600°c Fire Temperature)

At this temperature, the reduction in the load carrying of RPC columns became more severe. Figure (4-16) shows that the reduction in the load carrying capacity of RPC columns after heating at 600°c was in the range of (44.7 to 39.4%) for RPC columns with different concrete cover and tie spacing. While for NSC columns the reduction was in the range of (24.3 to 15.3%). The same observations of (Group Three) were seen in regard to the case of the effect of tie spacing and concrete cover on the load carrying capacity after fire exposure at 600°c. The contribution of increasing concrete cover (C.C) and reducing tie spacing in improving the fire resistance of RPC columns at this temperature was nearly negligible. And it was found that the load carrying capacity of RPC columns with different concrete cover became close to each other at higher fire temperature levels. It was also noticed that the increase of concrete cover (C.C) for the exposed NSC columns at 600°c has a slightly better effect in improving the fire resistance than in the case of heating at 400°C.T11US, it can be concluded from the results of (Group Three and Four) that the effect of concrete type had a major impact in performance of concrete column specimens during fire exposure. NSC provides excellent fire resistance compared with RPC, in which the percentages of the residual load carrying capacity of the exposed NSC columns are much higher than these of the exposed RPC columns both at (400 and 600°C) fire temperature level, see Figure (4-17). Figures (4-16) and (4-17) showed that nearly 42% of the load carrying capacity for RPC columns were gone after heating at 600°c. While NSC columns lost only about 20% after heating at the same temperature level. The reason for this excellent fire resistance of NSC is believed to be due to that NSC is non-combustible and has a relatively low thermal conductivity, high heat capacity and slower strength degradation with temperature, this slow rate of heat transfer and strength loss that enables NSC to act as an effective fire shield (Kodur and McGrath, 2003). In addition, fire induced spalling which was occurs significantly in RPC columns, was not observed during heating process of NSC columns. So if the concrete cover remains in place during heating (there is no cover

illustration not visible in this excerpt

Figure (4-16): The percentage of reduction in load carrying capacity of theRPC and NSC columns after heating at 600°c fire temperature.

illustration not visible in this excerpt

energy dissipation of concrete columns. The ultimate displacements of the reference columns and exposed columns at different temperature levels are outlined in Table (4-5). From the test results of the reference columns, it can be observed that concrete columns with higher transverse reinforcement amount exhibit higher axial displacement than concrete columns without transverse reinforcement, the same behaviour observed concerning reducing the concrete cover (C.C), load versus axial displacement relationships of each pair of the reference columns having the same transverse reinforcement amount and different concrete cover (C.C) presented in Figure (4-18). This Figure exhibited that the ascending portion of the curves for all RPC columns are usually linear however, RPC columns with less transverse reinforcement were steeper. This can be attributed to the fact that the presence of higher transverse reinforcement modified the failure mode of RPC columns to become more ductile rather than brittle, thus the sudden drop in capacity is almost missing. It is also illustrated that the ascending portion of the curve for the NSC specimens are liner while the descending curve drop rapidly in a short range of deformation.

illustration not visible in this excerpt

Figure (4-18): Axial load versus axial displacement of the reference RPC and NSC column specimens.

After fire exposure of RPC and NSC columns with different temperature levels, the load-displacement relationships changed, at temperature of 200°c the load-displacement relationships of the RPC columns were nearly semi-identical with those of the reference RPC columns. While, for heating temperatures of (400 and 600°C), the results elucidate that the axial deformation of the RPC columns increased with respect to the level of the applied load compared with those of the reference RPC columns and the load-displacement relationships of the RPC columns indicated softer behaviour. This could be attributed to the relative reduction in stiffness of RPC columns exposed to fire, which was essentially due to the reduction in the modulus of elasticity of concrete and the formation of hairline cracks. Figures (4-19) to (4-22) reveal the load versus axial displacement relationships of each pair of RPC columns having the same transverse reinforcement amount and different concrete cover (C.C).

On the other hand, for NSC columns at heating temperatures of (200, 400 and 600°C), the behaviour of load-displacement relationships were less affected than those of the RPC columns, the ascending part of the load­displacement curves was little flattening and became more linear, which indicated increasing of the columns axial deformation with respect to the level of the applied load compared with the reference NSC columns. Figures (4-23) and (4-24) reveal the load versus axial displacement relationships of each pair of NSC columns having the same transverse reinforcement amount and different concrete cover (C.C).

illustration not visible in this excerpt

Figure (4-19): Axial load versus axial displacement of RPCIO and RPC20 at different temperature levels.

illustration not visible in this excerpt

Figure (4-21): Axial load versus axial displacement of RPC 12 and RPC22 at different temperature levels.

illustration not visible in this excerpt

Figure (4-22): Axial load versus axial displacement of RPC 13 and RPC23 at different temperature levels.

illustration not visible in this excerpt

Figure (4-23): Axial load versus axial displacement of NSC 11 and NSC21 at different temperature levels.

It’s also illustrated from the test results, that the discrepancy of the ultimate mid-height lateral displacement of the reference columns were not tangible and there is a slightly higher displacement with increasing transverse reinforcement. Also the load versus mid-height lateral displacement relationships of RPC and NSC after heating at fire temperatures of (400 and 600°C) were also exhibited the same behaviour but with higher displacement with respect to the level of the applied load compared with the reference columns as shown in Figure (4-25). The deflection readings at each load stage were plotted along the length of the column as depicted in this Figure.

illustration not visible in this excerpt

Figure (4-25): Axial load versus mid-height lateral displacement of the RPC
and NSC columns prior to and later on heating at 600°c.

4.6 Ductility of RPC and NSC Column Specimens

The parametric study of ductility of the concrete columns at ambient temperatures have been investigated by many researchers in the past, but there is no information about the ductility of concrete columns after exposed to real fire especially that the ductility of the structure during and after fire is very important, in order to indicate the time of the beginning of the failure which represent a real threat to fire fighters and residents.

The ductility of the tested columns were computed using two approaches developed by considering a displacement ductility ratio as an index, these approaches proposed by (Pessiki and Pieroni, 1997) and (Azizinamini et al., 1999). (Pessiki and Pieroni, 1997) defined column "displacement ductility" as the ratio of the axial displacement of the column at an axial load corresponding to 85% of the maximum axial load on the descending branch of the axial load-displacement curve (Δ85), to the displacement at the limit of elastic behaviour (Ay), Figure (4-26-a) illustrate how the limit of elastic behaviour, (Ay) is determined. This mean the ductility is calculated using the equation below:

The displacement ductility index (μ85) = (4-6)

While (Azizinamini et al., 1999) defined the displacement ductility as the ratio between the displacement at peak load (Au) and the notional yield displacement (Ay), as presented in Figure (4-26-b), the notional yield displacement (Ay) is defined as the intersection of the two straight lines associated with the load-displacement curves at the elastic and post-elastic stages respectively.

The displacement ductility index (μΔ) = ^ (4-7)

The results of ductility index of RPC and NSC columns were tabulated in Table (4-6). As mentioned before the load-displacement relationships of the RPC columns at temperature of 200°c were nearly semi­identical with those of the reference RPC columns, thus the ductility of RPC columns at 200°c were not taken into account.

illustration not visible in this excerpt

Figure (4-26): Determination procedures of column displacement ductilitydeveloped by: (a) (Pessiki and Pieroni, 1997), (b) (Azizinamini et al., 1999).

illustration not visible in this excerpt

Table: (4-6): Ductility index test results of RPC and NSC column specimens.

illustration not visible in this excerpt

In order to clarify the calculation procedures of the ductility results presented in Table (4-6), the determinations of the displacement ductility by the two approaches adopted for RPC 10 and RPC 13 columns prior to and later on fire exposure at 600°c were depicted in Figures (4-27) and (4-28).

Due to the difficulty of identifying the limit of elastic behaviour (Ay) for (Pessiki and Pieroni, 1997) approach, a best-fit line to the ascending part of the load-displacement curve for each column was obtained by linear regression analysis. This line was then extended to intersect with the maximum load of the column.

illustration not visible in this excerpt

Figure (4-27): Example of diagram for determining the ductility index of RPC columns prior to and later on heating by (Pessiki and Pieroni, 1997) approach.

illustration not visible in this excerpt

Figure (4-28): Example of diagram for determining the ductility index of RPC columns prior to and later on heating by (Azizinamini et al., 1999) approach.

It can be concluded from the results of Table (4-6) that the amount of transverse reinforcement is the most crucial parameter which affects the behaviour of columns with respect to ductility. An increase of the amount of transverse reinforcement leads to an increase in lateral confining pressure, resulting in improvement of both the strength and ductility of confined concrete. Figures (4-29) and (4-30) illustrate the effect of transverse reinforcement on the ductility of RPC and NSC columns prior to and later on fire exposure at 600°c. It can be seen from these Figures that the closer the spacing of ties, the more ductile is the behaviour of columns. A more sudden drop of load resistance after peak load occurred in the columns with larger spacing of transverse steel.

It is noted from the results of Table (4-6) that the displacement ductility index for RPC columns was not affected after fire exposure and even it got a slight increase for the exposed RPC columns. On the other hand, it was slightly decreased for the exposed NSC columns. This may attribute to that the calculation of displacement ductility index depends only on displacement of the columns, NSC columns after exposed to fire exhibit a steeper and more linear behaviour in the ascending part of the load­displacement curve and then a quick descending portion prior to failure, while in the case of the exposed RPC columns and due to the presence of steel fibres, the load-displacement curve response as shown before, is more ductile which exhibits a higher value of (Δ85%) despite the aggressive reduction of the load carrying capacity. Since displacement ductility index depends only on the displacement values and does not differentiate the ductility between the columns with a small residual load resistance and those with a large load resistance, therefore this method is not adequate in determining the ductility of the exposed columns. Accordingly, a better approach has been proposed to measure the ductility of those columns may be by using the energy absorption capacity which is based on estimating the area enclosed by the load-displacement curve for each column, to take into account the reduction of load carrying capacity for these columns after fire.

Figure (4-29): Effect of transverse reinforcement on the displacement ductility index of RPC and NSC columns at prior to and later on fire exposure determined by (Pessiki and Pieroni, 1997) approach.

illustration not visible in this excerpt

Figure (4-30): Effect of transverse reinforcement on the displacement ductility index of RPC and NSC columns at prior to and later on fire exposure determined by (Azizinamini et al., 1999) approach.

4.7 Stiffness of RPC and NSC Column Specimens

In the last decade, various displacement-based design (DBD) approaches have been proposed to better control the displacements of structures in earthquakes and thereby enable performance based seismic design.

Sullivan et al. (2004) proposed two procedures as illustrated in Figure (4-31) that are commonly utilized to determine the stiffness of reinforced concrete (RC) columns, secant stiffness (Ks) and initial stiffness (Kin). Secant stiffness of RC columns, which is also called effective stiffness is defined as the ratio of the maximum applied load on the specimen (Pu), to the maximum displacement (Au). While Initial stiffness determined by a simple approach as shown in Figure (4-31), in which a secant passing through a point on the load-displacement envelope corresponding to 70% of the maximum applied load on the specimen (0.7PU) is extended to intersect with the horizontal line at (Pu). The results of the secant and initial stiffness of RPC and NSC columns were tabulated in Table (4-7).

The same RPC columns (RPC 10 and RPC 13) presented in determining ductility index, are proposed in Figure (4-32) to show the calculation procedures of the secant and initial stiffness of these columns prior to and later on fire exposure at 600°c. It must be noted that for the same reason mentioned in the ductility of RPC columns, the stiffness of RPC columns at 200°c were not taken into account.

Figure (4-31): Illustration of secant and initial stiffness of the displacement-
based design (DBD) approaches reproduced from (Sullivan et al., 2004).

Table: (4-7): Secant and initial stiffness test results of RPC column

illustration not visible in this excerpt

Figure (4-32): Example of diagram for calculations of secant and initial stiffness of RPC 10 and RPC 13 columns prior to and later on heating.

As indicated in previous Table, that the secant and initial stiffness after heating at different temperature levels deteriorated significantly with the increase in fire temperature level which can be observed also from Figures (4-33) and (4-34), and the reduction in stiffness is accompanied with a reduction in the load carrying capacity. The stiffness of RPC columns decreased from 100% at ambient temperature to almost about (75 and 55%) after eat$1& at (400 and 600°C), respectively. While for NSC columns the stiffness decreased from 100% at ambient temperature to about (79 and 60%) at (400 and 600°C), respectively. The reason for this is believed to be due to the deterioration of the mechanical properties of both steel reinforcement and concrete including (stress-strain relationships and Young's modulus) as well as chemical change actions which increasing the concrete permeability and porosity at high temperature and principally producing micro-cracks by changes of material inner structure, as well as by crack opening due to high gas pressure values, these micro-cracks will seriously

illustration not visible in this excerpt

Figure (4-33): Percentage secant stiffness of RPC and NSC columns after
heating at 400°c fire temperature levels.

illustration not visible in this excerpt

Figure (4-34): Percentage secant stiffness of RPC and NSC columns after
heating at 600°c fire temperature levels.

4.8 Energy Absorption of RPC and NSC Column Specimens

The energy absorption capacity of the concrete column defined as the area enclosed by the load-displacement curve until the maximum load was reached, which represents the energy absorption of the concrete column that could sustain before displaying a significant drop in load carrying capacity (Barros and Sena-Cruz, 1999); (Yu et al., 2016); (Kumar et al., 2017). Previous research reported that energy absorption capacity is the most suitable index of concrete structures not only for its structural response against earthquake motion but also to be seriously considered for concrete structures that have to resist fires and impact load caused by incidents or terrorist attacks (Yu et al., 2016). Results of the energy absorption of RPC and NSC columns before and after exposure were tabulated in Table (4-8). It's obvious that the spacing and amount of ties affect the energy absorption capacity of RPC and NSC column specimens by increasing the energy absorption capacity with decreasing spacing of lateral ties. For the reference columns, reducing tie spacing from (200 to 50 mm) gave an increasing in the energy absorption capacity of (58 and 45%) for RPC columns with concrete cover (C.C) of (15 and 30 mm) respectively, and an increasing of (70 and 54%) for NSC columns with concrete cover (C.C) of (15 and 30 mm) respectively.

From these results, it is clear that the values of energy absorption capacity of the reference and exposed column specimens decrease with increasing of concrete cover for the two types of concrete (RPC and NSC). It is also illustrated that RPC columns exhibit a much higher energy absorption capacity than NSC columns at ambient temperature, in which the energy absorption capacity of RPC columns with concrete cover of (15 and 30 mm) were higher than that of similar NSC columns with concrete cover of (15 and 30 mm) by (673 and 733)% respectively. The reason for this is believed to be due to the addition of fibres in concrete can significantly improve ductility, and it provides further reinforcement benefits for structures in areas that are prone to strong earthquakes. The combination of steel fibre with superior strength concrete reduces the need for mounting confining reinforcements significantly for the columns (Massicotte et al., 1999).

Table (4-8): The energy absorption capacity test results of RPC and NSC columns.

illustration not visible in this excerpt

The calculations of the area under the load-displacement curves until the maximum load for RPC 10 and RPC 13 columns prior to and later on fire heating at 600°c were represented in Figure (4-35).

At eating temperature of 400°c, for RPC and NSC columns with concrete cover of 15 mm, the residual energy absorption capacity were in the range of (73.1 to 77.8%) and (86.9 to 90.2%) respectively. While for RPC and NSC columns with concrete cover of 30 mm, the residual energy absorption capacity were in the range of (77.6 to 82.3%) and (84.8 to 96.7%) respectively.

At heating temperature of 600°c, for RPC and NSC columns with concrete cover of 15 mm, the residual energy absorption capacity were in the range of (51.8 to 57.9%) and (75.2 to 82.9%) respectively. While for RPC and NSC columns with concrete cover of 30 mm, the residual energy absorption capacity were in the range of (77.6 to 82.3%) and (79.2 to 79.5%) respectively.

It can be seen clearly that the energy absorption capacity provides a better approach for investigation the ductility of the concrete structures during fire. This is due to that the calculation of energy absorption capacity takes into account both the load resistance of the column and its displacement. This is contrast with displacement ductility index which did not differentiate the ductility between the columns with a small residual load resistance and those with a large load resistance.

illustration not visible in this excerpt

Figure (4-35): Example of diagram for determinations of area under the load-displacement curves of RPC 10 and RPC 13 columns prior to and later on heating.

4.9 Effect of Fire Exposure on Steel Reinforcement

Steel reinforcement if protected by the minimum cover specified by the code it is expected that the effect of high temperature on the reinforcement bars will be negligible. However, deformation due to thermal expansion and loss of bond between concrete and steel might cause the structure instability and affect the structural integrity. It is generally held that steel reinforcement bars need to be protected from exposure to temperatures in excess of 250-300°C, this is due to the fact that steels with low carbon contents are known to exhibit blue brittleness between (200 and 300°C) (ünlüoölu et al., 2007).

For the investigation of fire effect on steel reinforcement bars, three groups of steel reinforcement bars 06 mm in diameter were used. In the first group, steel reinforcement samples were embedded in RPC and NSC columns at 30 mm concrete cover (C.C). The samples of second group were embedded in RPC and NSC columns at 15 mm concrete cover, and the third group samples were exposed to fire directly without concrete cover. The three groups of samples were burned at temperature of 600°c and for 1 hour fire duration, then the steel reinforcement were extracted from concrete and a yield stress test was conducted. Table (4-9) and Figure (4-36), show the comparison of the results of yield stress test for the reinforcement steel bars after exposure to 600°c and these of the reference sample.

Table (4-9) and Figure (4-36) demonstrate that the increase in concrete cover to the reinforcement steel bars has a positive impact on the yield stress where the reduction in the yield stress for columns with concrete cover of 30 mm were (3.4 and 10.5%) for RPC and NSC respectively. And for columns with 15 mm concrete cover the reduction were (6.8 and 14.8%) for RPC and NSC respectively. While for samples which were exposed directly to fire the reduction was 31.9%. It is also observed that NSC provides excellent protection for the reinforcement bars when compared with RPC with the same cover, in which the reduction in the yield stress for the NSC columns were (7.1 and 8%) lower than RPC columns with concrete cover (C.C) of (30 and 15 mm) respectively.

Table: (4-9): The effect of fire exposure on yield stress of steel reinforcement.

illustration not visible in this excerpt

Figure (4-36): Effect of concrete cover in protecting the steel reinforcement after fire exposure at 600°c.

4.10 Appearance and Colour Change of the column specimens

After the instillation of the column vertically inside the furnace chamber, the heating process was started and some observations were recorded and compared between RPC and NSC column specimens.

4.10.1 RPC Column Specimens

- The most visible phenomenon during the heating process of RPC columns is the spalling of concrete cover which was observed from the beginning of the heating after the fire temperature exceeded 300°c, a crackle sound started to be heard followed by breaking of small chunks of concrete cover in explosive way which was indicated clearly by the sound of the hitting of these chunks at the steel cover of the furnace.
- For all RPC columns the explosive spalling continuous for almost about 30 minutes, then the crackle sound begins to fade and spalling appeared to be gradually reduced but continued until the end of the heating. -
- In few cases, spalling was also observed immediately after cooling of RPC columns.
- After completing 60 minutes and removing the fire source, RPC columns were found with different shapes and the comer of the columns were significantly damaged specially in the case of heating at 600°c fire temperature.

illustration not visible in this excerpt

Plate (4-3): Moisture migration observed during: (a) Heating process, (b) Testing process.

Plate (4-4) illustrates an analysis of colour change in RPC under the influence of fire with temperature levels of up to 600°c, The colour change observed in concrete is primarily a result of the gradual dehydration of the cement paste, but also of transformations occurring within the aggregate (Hager et al., 2013). Additionally, steel fibres at the surface of specimen underwent oxidation after exposed to high temperatures, subsequently producing black carbon particles that were absorbed by the concrete. Concrete with steel fibres thus had a relatively darkish colour (Tai, et al. 2011). The images obtained from the exposed the RPC columns surface prior to and later on heating at different temperature levels, were taken after magnification with macro lens (25-58) mm. At 200°c, the colour of RPC specimens are slightly the same as those at room temperature and there is no spalling occurred at the exposed surface. At 400°c, small cracks began to appear and the specimens turned to a greyish brown colour. At 600°c, cracks at the specimen surface expanded and the specimen turned to a darkish grey to buff colour.

illustration not visible in this excerpt

Plate (4-4): Colour change of RPC column specimens after exposure to fire with different temperature levels.

4.10.2 NSC Column Specimens

- Spalling was not observed during the heating process of NSC columns until the 1 hour fire duration was finished.^[2]

illustration not visible in this excerpt

Plate (4-5): Hairline cracks distributed along the exposed surfaces of NSC columns.

4.11 Discussions of Failure Process of RPC and NSC Columns

Columns are very important members of concrete structural systems. They are subjected to axial compressive force associated with bending in most practical cases . As compression elements, failure of one column in a critical location can cause the progressive collapse of the entire structure to which it is connected. Therefore, extreme care must be taken in column analysis and design, because failure of columns can provide little visual warning and cause a significant economic and human loss.

During the testing process, the main difference observation indicated between the failure mode of NSC and RPC columns is that the failure mode of all NSC columns was a gradual failure started by falling of a large pieces of the concrete and steel reinforcement started to appear until a localized failure occurs. For NSC columns, the failure surface was similar to a pyramid as shown in Plate (4-6, a and b) and usually accompanied with spalling of the cover and the steel reinforcing bars buckle and the confinement steel finally fractures. After the failure takes place a sudden drop of the load was observed for all NSC columns.

While RPC column specimens express better core integrity and there is no evidence of any cover spalling or indication of the failure until reached approximately 90% of the failure load, diagonal cracks occurred on the column sides at the ends. The beginning of failure can be noticed for all RPC columns by listening to sound of the popping and pulling out of steel fibres then an explosive noisy failure takes place and the applied load was gradually dropped down, Plate (4-6, c and d).

In some cases and for the exposed RPC columns specially after heating at 600°c, the column surfaces were highly deteriorated and the location of the failure was not visible at the deteriorated surfaces, Plate (4-6, e, f, g and 11). It is also indicated that for the exposed RPC columns the failure was usually occurred at the most deteriorated part of the column, Plate (4-6, i and j).

In addition, it was also observed in the exposed RPC columns at 600°c and with 15 mm concrete cover, buckling of the longitudinal bars was visible at the deteriorated faces of the column as shown in Plate (4-6, k) which is due to sloughing of concrete cover after fire exposure.

Plate (4-6): Examples of the failure modes observed in this investigation.

illustration not visible in this excerpt

CHAPTER FIVE

Finite Element Analysis

5.1 Introduction

n this chapter, a comparison between the experimental and theoretical studies is presented to ensure that modelling of the column is adequate containing: elements type, material properties, real constant and convergence study. A nonlinear finite element analysis has been carried out to analyze the RPC columns tested in the current study. The tested column specimens with and without exposure to fire were analyzed.

The goal of this chapter is to develop a nonlinear finite element model to simulate the effect of fire exposure on the RPC columns and for better prediction of the response of RPC columns to a real fire. In addition, this chapter also includes the analysis of the tested columns with some important parametric studies by using a powerful nonlinear finite element method package ABAQUS/Standard 2016. From the finite element analysis of each of these tested columns, the predicted ultimate load carrying capacity versus axial deformation response were obtained and compared with the corresponding experimental results.

In this chapter, twelve RPC column specimens with and without lateral ties and with different concrete covers were selected to be modeled. The description of the selected RPC columns characteristics along with the reference as illustrated in chapter four, were summarized in Table (5-1).

Table (5-1): T

illustration not visible in this excerpt

5.2 FE mesh and boundary conditions

An important step in the finite element modelling is the selection of the mesh density. Before the numerical analysis is adopted, a sufficient pre­analysis of different densities of mesh were carried out to select the best mesh density that gives the required accuracy depending on the level of complexity for the analysis with the time spent in the processing. Convergence of results is obtained when the structure is divided into an adequate number of elements. This is practically achieved when an increase in the mesh density has a negligible effect on the results. Therefore, in the present finite element modelling, a convergence study on the model of the column was carried out to determine an appropriate mesh density. The convergence study was made by increasing the number of elements (mesh) in each direction X, Y and z. The column of same material properties was modelled with a decrease in the element size (25, 20, 15, 10, and 5 mm) (the number of elements are 576, 1125, 2667, 9000 and 72000 respectively). The axial displacement was observed for the same applied load level. It can be concluded from convergence study shown in Figure (5-1), that the change in axial displacement can be neglected when the mesh size decreased from (15) mm to (5) mm, also the value of the axial displacement become more accurate with experimental results. Therefore, 10 mm element size were selected for the mesh density, which guarantees a good adjustment between the size of the elements and the stability of the numerical solution. Figure (5­2) indicates the varied mesh of columns.

illustration not visible in this excerpt

< 3.25

Mesh 25 Mesh 20 Mesh 15 Mesh 10 Mesh 5 Figure (5-2): The considered meshes for the RPC column specimens.

5.3 Modelling and Analysis of RPC Column Specimens

A nonlinear finite element model for RPC columns was developed using ABAQUS finite element code. Two analysis models were illustrated below for both the reference and exposed RPC columns.

5.3.1 Parts and Assembly

The assembly of parts, which were used in modelling these columns, is shown in Figure (5-3). In addition, this Figure depicts the details about the orientation of steel reinforcement in reinforced RPC columns with tie spacing (transverse steel ratio). It must be noted that the modelling of material properties for RPC and steel reinforcement are included in (Appendix B).

illustration not visible in this excerpt

Abaqus/Standard 3DEXPERIENCE R2016X Sat Aug 05 15:16:20: Arabic standard Time 2017 Figure (5-3): The assembled parts of RPC column specimens.

5.3.2 Static analysis model

For the reference (non-exposed) RPC columns, the structural analysis was done by using only one-step (static general analysis step). Reference RPC columns were modelled using three-dimensional solid elements with reduced integration in the modelling of the concrete with an 8-node brick element (C3D8R). While for reinforcement, a three-dimensional two-node truss element (T3D2) was used. For each RPC column, a uniform pressure was applied at the top of the column in which the bottom of the column is fixed (displacement is restrained in all directions X, Y and Z-axis). While, the top of the column is restrained to displacement in X and Z-axis and free displacement is assumed in the direction of the longitudinal axis of the column (Y-axis) and a uniform loading was applied at the top of the column, as shown in Figure (5-4).

5.3.3 Coupled temperature-displacement analysis model

The numerical model, proposed here, uses two steps of coupled temperature-displacement analysis. The first step was used for fire simulation. While the second step was used to evaluate the structural response of RPC columns after exposure to fire, with the same boundary and loading conditions used in the reference RPC columns which illustrated in static analysis model, Figure (5-4). The exposed RPC columns were modelled using three-dimensional solid elements in the modelling of the concrete with an 8-node thermally coupled brick, trilinear displacement and temperature (C3D8T), and for reinforcement, a three-dimensional thermally coupled two-node truss element (T3D2T). So as to study the effect of the fire heating at (400 and 600°C) and on two adjacent sides on the structural behavior of RPC column specimens. The fire effect simulation is carried out by using a heat flux, in such a way that the temperature was changing by using an amplitude. The amplitude was used to input the fire temperature-time curve in order to simulate the real fire used in the experimental program and to cool the column specimens to room temperature before starting the second step, Figure (5-5). At each time interval, the analysis is performed through two main procedures, namely: 1) Establishing fire temperature due to fire exposure, 2) Carrying out coupled hydro-thermal analysis to evaluate the fire temperatures distribution history inside the column and for generalization of the spalling along the element, fire-induced spalling. It must be mentioned that the governing equations for the hydrothermal analysis employed in this model were proposed by (Dwaikat and Kodur, 2009), (Ali et al., 2010) and (Raut and Kodur, 2011) which are summarized in (Appendix C).

illustration not visible in this excerpt

Figure (5-4): The upper and lower ends boundary conditions of columns.

illustration not visible in this excerpt

Figure (5-5): Fire temperature-time curve used for the exposed RPC columns at 600°c and for 1-hour fire duration.

5.4 Finite Element Analysis Results and Discussion

5.4.1 Analytical Results of Group One (Reference RPC Columns)

A comparison between the ultimate loads (load at failure) and axial displacement of the tested RPC column specimens with theoretical values from finite element analysis is listed in Table (5-2). In Figure (5-6) finite element results are compared against experimental data. The good agreement between the test and the analysis indicates that the proposed model is consistent and can be used with confidence. It can be noted also that all the theoretical values of ultimate load of the reference RPC column specimens and those exposed to fire exceed the experimental values by a margin ranging between (4.9-8.3%) with a standard deviation in the range (0.0112­0.0135). The numerical predicted ultimate axial deformations are found to be lower compared to the experimental values with an average experimental to the numerical ratio of (1.0833). In addition, it can be noticed from the axial load-displacement relationships that all the theoretical models show a stiffer behavior when compared with the experimental axial load-displacement relationships. On the other hand, the maximum vertical displacement, stresses and strains distribution along the longitudinal у-axis are found to be very close in RPC columns with no ties, see Figure (5-7). While, in graphs shown in this Figure, it can be deduced that there is a significant change in these parameters along of the RPC columns with 50 mm tie spacing.

Table (5-2): The experimental and theoretical results for the reference RPC columns.

illustration not visible in this excerpt

Figure (5-6): Experimental and theoretical load-axial displacement relationship of RPC column specimens prior to heating.

illustration not visible in this excerpt

Figure (5-7): Continued.

5.4.2 Analytical Results of Groups Two and Three (Heating at 400 and 600°c Fire Temperature)

In the analysis, the RPC columns were exposed to fire from two sides. The column specimens were burnt at temperature levels (400 and 600°C) for a 60-minutes period of exposure. In Figures (5-8) and (5-9) the finite element results are compared against experimental data which showed that the axial deformations increased rapidly with a lower level of the applied load compared with those of the reference columns. A control plot of the temperature distribution in the RPC column specimen (RPC 13) after 60- minutes of heating is shown in Figure (5-10), while The computed relations between fire temperatures at different depths from the exposed faces of RPC23 column with time are shown in Figure (5-11). From these Figures, it is clear that with the increase in time of exposure to fire the temperature increases towards the core of the RPC column specimens. These Figures depict that the predicted temperatures are in good agreement with experimental values, which confirm the validity and accuracy of the thermal analysis model, and the program used. The results of load carrying capacity and axial deformation of RPC column specimens exposed to fire are given in Table (5-3). From this Table, it can be seen that the values of ultimate load decrease when the column specimen is exposed to fire.

Table (5-3): The experimental and theoretical results for the exposed RPC columns at 400 and 600°c fire temperature.

illustration not visible in this excerpt

Figure (5-8): Experimental and theoretical load-vertical displacement of RPC 10 and RPC 13 after heating at (400 and 600°C).

illustration not visible in this excerpt

Figure (5-9): Experimental and theoretical load-vertical displacement of RPC20 and RPC23 after heating at (400 and 600°C).

illustration not visible in this excerpt

Figure (5-10): The temperature distribution in the RPC column specimen
(RPC 13) after 60-minutes of heating at 400°c.

illustration not visible in this excerpt

5.5 Parametric Study

The main objective of this section is to investigate the effect of several important parameters on the behavior of RPC columns. Depending on the architecture and structural layout of a building, columns can be exposed to a fire in a number of different scenarios. If a structural frame is exposed to fire as shown in Figure (5-12), columns may expose to fire from different sides (1-side, 2-sides, 3-sides or all four sides) depending on the position of columns in the building. According to that, the parametric studies included in this section deals with the effect of fire exposure at different fire scenarios (direction of the fire exposure and fire duration). Same RPC column models used in the finite element analysis in this chapter (RPC10, RPC13, RPC20 and RPC23) will exposed to fire at different sides (1-, 2-opposite, 3- and 4-sides) and with different fire durations (1 and 2 hours) as shown in Figure (5-13). The finite element model generated in this investigation will be used to conduct a detailed parametric study on the behavior of RPC column specimens.

For all RPC columns, following parameters had been kept constant (similar to the parameters of the finite element analysis):

1) Column dimensions.
2) Loading pattern.
3) Analysis steps and boundary conditions.
4) Material properties.
5) Mesh density.

The output parameters that have been extracted from the analysis are the axial load and displacement data, which are directly obtained from the ABAQUS simulation. The summary of the results of the ultimate load carrying capacity (Pu) and axial deformation at ultimate load (Au) are given in Table (5-4). The load versus axial displacement of the RPC column specimens exposed to fire from different sides are then generated from the numerical analysis, is investigated in this study as shown in Figure (5-14 a, b, c and d). In addition, the computed relations between fire temperatures at different depths with time are shown in Figure (5-15).

illustration not visible in this excerpt

Figure (5-12): Fire exposure of RC columns on different sides according to its position.

illustration not visible in this excerpt

Figure (5-13): Fire scenarios included in the parametric study.

illustration not visible in this excerpt

Table (5-4): The theoretical results of the exposed RPC columns at 600°c Fire Temperature and with different fire configurations.

illustration not visible in this excerpt

Figure (5-14): Comparison between analytical values of different fire configurations for (a-RPC10, b-RPC13, C-RPC20 and d-RPC23) column specimens at 600°c temperature level.

illustration not visible in this excerpt

5.5.1 Effect of Fire Exposure Scenarios on Load Carrying Capacity

To evaluate the influence of fire scenarios on the load carrying capacity, the percentage of reduction in load carrying capacity were calculated by comparing the ultimate load carrying capacity (Pu) of the exposed RPC columns with different fire scenarios, which extracted from the finite element analysis with the theoretical results of the reference RPC columns before heating. It can be indicated from the results of Table (5-5) that the effect of fir e heating is more severe with increasing the number of sides of the exposure. After fire exposure at 600°c temperature level and for 2 hours duration, RPC columns lost about (53.2-58.4%) of the load carrying capacity for 1-side fire exposure and about (68.5-72.3%) of the load carrying capacity for all 4-sides fire exposure. The same observation can be seen with increasing the fire duration from (1 to 2 hours), in which the reduction in the load carrying capacity were increased from (39 to 58.4%) for the 1-side fire heating and from (61.5 to 72.3%) for the all 4-sides fire exposure.

In addition, it has indicated from the results that the effect of fire heating was more pronounced in the case of the 2-adjcent sides' fire exposure scenario compared with 2-opposite sides' fire exposure. This may be due to the development asymmetric thermal gradients along both axes when exposed to fire on 2-adjacent sides. While, in the case of the 2-opposite sides, the thermal gradients remain symmetric along both axes as shown in Figure (5-16). In addition, it can be seen from the Figure that in the case of the 2-adjcent fire exposure scenario the fire temperatures distribution were more concentrated at the exposed comer with fire temperature levels above 300°c reached nearly the center of the core concrete, at this temperature RPC begin to loss its strength rapidly. While in the case of the 2-opposite fire exposure the fire temperatures dispersed in a larger area of the column cross­section with temperature levels below 300°c.

Table (5-5): Percentage of reduction in the load carrying capacity (Pu) of RPC columns after fire exposure at 600°c and with different fire scenarios.

illustration not visible in this excerpt

Figure (5-16): Heat transfer analysis after 120-minutes of heating at 600°c and with (2-opposite and 2-adjcent sides) fire exposure.

illustration not visible in this excerpt

5.5.2 Effect of Fire Exposure Scenarios on Stiffness and Ductility

Initial stiffness (Kin) of the displacement-based design (DBD) are calculated for the exposed RPC columns with different fire exposure scenarios by using the same approach illustrated in chapter four proposed by (Sullivan T. J. et al., 2004). Table (5-6) elucidates the results of the initial stiffness of the both reference and exposed columns with different fire scenarios. In addition, the percentage of reduction in the initial stiffness after fire exposure in different scenarios are also calculated and summarized in tfcğ Table.

It can be clarified from the results that the reduction in the stiffness of RPC columns were more severe with increasing the number of sides of the exposure and increasing the duration of the fire exposure from (1 to 2 hours). In addition, it can be seen form the results that RPC columns with (30 mm) concrete cover provide a better performance after 1-hour fire heating. In which the reduction in initial stiffness was slightly lower compared with the similar RPC columns with (15 mm) concrete cover.

Table (5-6): Initial stiffness and percentage of reduction in initial stiffness of RPC columns after fire exposure at 600°c and with different fire scenarios.

illustration not visible in this excerpt

As motioned in chapter four, that calculating the energy absorption capacity provides a better approach for determining the ductility of the exposed concrete columns. According to that, the effect of fire exposure with different scenarios on the ductility of RPC columns determined by calculating and comparing the energy absorption capacity of both reference and exposed RPC columns. Table (5-7) summarizes the results of the energy absorption capacity and percentage of reduction in it after heating with different fire scenarios. It can be clarified from the results that the effect of increasing fire duration on the ductility of RPC columns was more severe than the effect of increasing the number of sides of the exposure. Also the effect of fire exposure on the ductility of RPC columns was lower somehow compared with the effect of the same fire scenario on the stiffness of similar RPC columns.

Table (5-7): Energy absorption capacity and percentage of reduction in
energy absorption capacity of RPC columns after fire exposure at 600°c and with different fire scenarios.

illustration not visible in this excerpt

5.5.3 Spalling Implementation

A 2-D heat transfer analysis of the RPC column cross section is performed using ABAQUS finite element software (ABAQUS 2016). The onset of spalling is triggered when the fire temperature level reaches about 350°c. Spalling resulted of removing all the elements making up the concrete cover. The analysis is continued and the temperature distribution for the reduced cross section is calculated. Two fire exposure scenarios were analyzed as shown in Figure (5-17). It can be indicated from this Figure, that the spalling is initiated firstly at the exposed corners of the RPC column then continues to take place along the concrete cover of the exposed surfaces. In addition, it can be indicated that the spalling increased significantly during the first 30-minutes of the fire exposure in which nearly a half of the mass loss of the exposed column takes place during this period.

Figure (5-17): Spalling analysis with time for: (a) 2-adjacent sides fire
exposure, (b) 4-sides fire exposure.

5.5.4 Effect of slenderness ratio on the fire behavior of RPC columns

As mentioned before, the column may undergo bending (uni-axial or bi-axial bending) due to eccentric loading and uneven exposure to fire. In addition, columns (especially those made of RPC) may experience spalling and may be subjected to fire scenario different from a standard fire based on the fuel and ventilation available in the building. To illustrate the effect of these parameters, a simulation was performed using the same FE model developed in this chapter to check the presence of (fire induced biaxial bending) for RPC 13 column and with different slenderness ratios (different height of the columns: 900, 1500 and 2000 mm) as shown in Figure (5-18).

Consider Figure (5-19) which plots comparison of the load-axial displacement curves after exposure to fire at 2-adjacent sides and at 4-sides and with slenderness ratio (λ) of (9 and 20). From this figure, it can be seen that RPC column exhibited higher axial displacement after exposure to fire at 4-sides compared with 2-adjacent sides' fire exposure. The lower axial displacement obtained after fire exposure at 2-adjacent sides can be contributed to the biaxial bending induced by the uneven spalling of RPC columns, Figure (5-20). Contrary to expectations, it is apparent from Figure (5-19) that RPC column with slenderness ratio of (20) gives slightly lower ultimate load carrying capacity after exposure to fire at 2-adjacent sides compared with 4-sides fire exposure at the same temperature level and duration. An important issue emerging from the results of Figures (5-19) and (5-20), as it give an example of the eccentricity induced by the uneven spalling when exposed to fire at 2-adjacent sides in which the column strength will deteriorate asymmetrically leading to shifting the plastic centroid location. This shifting will increase the applied moment on the column causing an increasing of the lateral displacement of RPC columns.

Figure (5-18): Slenderness ratios included in the parametric study.

illustration not visible in this excerpt

Figure (5-19): load-vertical displacement of RPC 13 with slenderness ratio of (a) 9 and (b) 20.

illustration not visible in this excerpt

Figure (5-20): load-lateral displacement of RPC 13 with different slenderness ratios

CHAPTER SIX

Conclusions and Recommendations

6.1 Introduction

The primary goal of this thesis is to study the behavior of reinforced- reactive powder concrete (RPC) columns under fire conditions and to analyses the various characteristics of a reinforced-concrete column that affect its overall fire performance. A secondary goal of the thesis is to develop a finite element model to help further investigations in the analysis of reinforced RPC structures under fire exposures.

6.2 Key Findings

Based on the investigated variables in this research different conclusion points concerning the structural behavior of concrete column specimens under exposure to real fire were drawn which can be outlined under different titles as follows:

6.2.1 Mixture Preparation and Mechanical Properties of RPC

1) After fire heating at 200°c, there was a slight increases in the mechanical properties of RPC in the range of (2 to 7%) due to the internal autoclaving of RPC since RPC was cured normally, which leads to further hydration of cement with silica fume as a pozzolanic reaction.
2) After fire exposure at (400 and 600°C), the reduction in the mechanical properties of RPC at early ages was higher than the reduction at later ages.
3) After fire exposure at (400 and 600°C), the reduction in values of modulus of elasticity of RPC was more significant than the other mechanical properties.

6.2.2 Experimental Results of RPC and NSC Columns

1) For the reference RPC and NSC columns, the load carrying capacity was increased with increasing the transverse reinforcement amount and reducing tie spacing.
2) The effect of the transverse reinforcement on the load carrying capacity was slightly less effective for RPC column specimens than NSC column specimens. With reducing the tie spacing from (200 to 50 mm) the load carrying capacity was increased by (19 and 16.5%) for RPC columns with concrete cover of (15 and 30 mm) respectively, and by (28 and 21%) for NSC columns with (15 and 30 mm) concrete cover respectively.
3) The same observation indicated in the case of the effect of transvers reinforcement on the load carrying capacity was also indicated with reducing of concrete cover from (30 to 15 mm) for both NSC and RPC columns.
4) It is found that there is a slight increase in the ultimate load carrying capacity of RPC columns at fire temperature of 200°c and the increasing ratios of load carrying capacity were almost identical for all RPC columns. On the other hand, it's found that the load carrying capacity values of the exposed NSC columns at 200°c were slightly decreased and the decreasing ratio was in the range of (2 to 6%).
5) It is found that after heating at fire temperature of (400 and 600°C), NSC provides excellent fire resistance compared with RPC columns in which it is found nearly 42% of the load carrying capacity for RPC columns were gone after heating at 600°c. While NSC columns lost only about 20% after heating at the same temperature level.
6) It is concluded from the test results that the contribution of increasing concrete cover and reducing tie spacing in improving the fire resistance of RPC columns after heating at 600°c was nearly negligible. And it was found that the load carrying capacity of RPC columns with different concrete covers became closer to each other at higher fire temperature levels. It was also noticed that the increase of concrete cover for the exposed NSC columns at 600°c has a slightly better effect in improving the fire resistance than in the case of heating at 400°c.
7) For the reference columns, it is observed that concrete columns with higher transverse reinforcement amount exhibit higher axial displacement than concrete columns without transverse reinforcement, the same behavior was also observed concerning reducing the concrete cover.
8) It is found form the load-displacement relationships that the ascending portion of the curves for all RPC columns was usually linear however, RPC columns with less transverse reinforcement were steeper. While the descending portion was softer. On the other hand, the ascending portion of the curve for the NSC specimens are liner while the descending curve drop rapidly in a short range of deformation.
9) In the present study, it is noticed that RPC columns exhibited a much higher ductile behavior compared with similar NSC columns, also it is found that the amount of transverse reinforcement is the most crucial parameter which affects the behavior of columns with respect to ductility, the closer the spacing of ties, the more ductile is the behavior of columns.
10) It is concluded from the test results that the stiffness of the reference RPC columns was significantly higher compared with the similar NSC columns. While after heating at (400 and 600°C) stiffness values
deteriorated significantly especially for RPC columns. The stiffness of RPC columns decreased from 100% at ambient temperature to almost about (75 and 55%) after heating at (400 and 600°C), respectively. While for NSC columns the stiffness decreased from 100% at ambient temperature to about (79 and 60%) at (400 and 600°C), respectively.
11) It can be deduced from the experimental results that, the energy absorption capacity of RPC and NSC columns clearly decreasing at higher levels of heating.
12) From the calculations of the ductility of RPC and NSC columns, it can be seen clearly that the energy absorption capacity provides a better approach for investigating the ductility of the concrete structures during fire than the displacement ductility index. This is due to that the calculation of energy absorption capacity takes into account both the load resistance of the column and its displacement. This is contrasted with displacement ductility index which did not differentiate the ductility between the columns with a small residual load resistance and those with a large load resistance.
13) The yield stress of steel reinforcement bars deteriorated significantly after fire exposure. The concrete cover has a good contribution in protecting the steel bars after heating at high temperatures especially in NSC columns. Where the loss indie yield stress after heating at 600°c for NSC and RPC columns with 30 mm concrete cover was smaller by (28.5 and 21.4%) respectively, than the reduction in the yield stress for the unprotected steel reinforcement after heating at the same temperature level.

6.2.3 Conclusions Based on the Numerical Model

A nonlinear finite element analysis has been carried out by using a powerful nonlinear finite element method package ABAQUS/Standard 2016 to analyze the RPC columns tested in the current study. The tested column specimens with and without exposure to fire were analyzed. And after a good agreement between the experimental and theoretical results, the analysis of the tested columns with some important parametric studies were also done and the following points were deduced:

1) The developed model provides a good simulation of the behavior of RPC columns after fire exposure in which the numerical analysis of the exposed columns based on both mechanical degradation of concrete and steel, and spalling prediction which is the main concern that effecting the behavior of RPC at fire.
2) There was good agreement between the test results and the analysis results in which all the theoretical values of ultimate load of the reference RPC columns and those exposed to fire exceed the experimental values by a margin ranging between (4.9-8.3%) with a standard deviation in the range (0.0112-0.0135). The numerical predicted ultimate axial deformations are found to be lower compared to the experimental values with an average experimental to numerical ratio of (1.0833). Therefore, the proposed model is consistent and can be used with confidence.
3) It is found from the axial load-displacement relationships that all the theoretical models show a stiffer behavior when compared with the experimental axial load-displacement relationships.
4) There is a lack in the information about the performance of the concrete columns during fire when the exposure of the fire is changing according to the position of the column in the building.
5) It à concluded that the effect of fire heating is more severe with increasing the number of sides of the exposure. After fire exposure at 600°c and for 2 hours, RPC columns lost about (53.2-58.4%) of the load carrying capacity for 1-side fire exposure and about (68.5-72.3%) of the load carrying capacity for all 4-sides fire exposure.
6) The same conclusion was found with increasing the fire duration from (1 to 2 hours), in which the reduction in the load carrying capacity were increased from (39 to 58.4%) for the 1-side fire exposure and from (61.5 to 72.3%) for the all 4-sides fire exposure.
7) It's deduced from the results that the effect of fire exposure was more pronounced in the case of the 2-adjcent sides fire exposure scenario compared with 2-opposite sides fire exposure.
8) Its concluded from the results that the reduction in the stiffness of RPC columns were more severe with increasing the number of sides of the exposure and increasing the duration of the fire exposure from (1 to 2 hours).
9) It's found from the results that the effect of increasing fire duration on the ductility of RPC columns was more severe than the effect of increasing the number of sides of the exposure. Also the effect of fire exposure on the ductility of RPC columns was lower somehow compared with the effect of the same fire scenario on the stiffness of similar RPC columns.

6.3 Recommendations for Further Research

For the purpose of involving the study of the structural behavior of RPC columns exposed to fire, the following recommendations for future studies may be taken into consideration:

1) For better simulation of the real fire exposure inside concrete structure further study is required on the effect of different fire scenarios according to the position of the column with eccentric loading, for example in the case of the corner column the analysis must including the effect of biaxial bending with 2 adjacent sides fire exposure. While for the peripheral column uniaxial bending with 1 side fire exposure is the best simulation of the problem.
2) Further experimental and theoretical study is required to investigate the effect of different fire exposure scenarios on slender RPC columns, in which a critical situation may arise due to the eccentricity induced by the uneven spalling.
3) Further study is required to investigate the effect of increasing the dimensions of RPC column specimens on its structural behavior under the influence of fire exposure.
4) Studying the structural behavior of RPC columns containing polypropylene fiber or hybrid fibers (steel fiber + polypropylene fiber) which is believed to exhibit a better performance against fire compared with RPC columns containing steel fiber.
5) Investigating the effect of changing the longitudinal reinforcement ratio (p) on the structural behavior of RPC columns exposed to fire.

REFERENCES

References

ACI Committee. (2014). Building Code Requirements for Structure Concrete and Commentary AC I 318M. American Concrete Institute, Michigan.

Agha A. Ahmed & Kadhum. M. M. (2015). Behavior of Reactive Powder Concrete Columns under Axial Compression.

Ahmad, s. H., & Shah, s. p. (1979). Complete stress-strain curve of concrete and nonlinear design. Progress Report Illinois Univ., Chicago. Dept, of Materials Engineering.

Ahmad, s., & Shah, s. p. (1982). Stress-strain curves of concrete confined by spiral reinforcement. Paper presented at the Journal Proceedings.

Alfawakhiri, F., Hewitt, c., & Solomon, R. (2002). Structural fire protection. Modem Steel Construction, 42( 12), 45-50.

Ali, F., Nadjai, A., & Choi, s. (2010). Numerical and experimental investigation of the behavior of high strength concrete columns in fire. Engineering structures, 52(5), 1236-1243.

Al-Jabiri, Z. s. (2015). Effect of Burning By Fire Flame on Mechanical Properties of Reactive Powder Concrete. (Degree of Master in Civil Engineering - Construction Materials), University of Babylon -College of Engineering.

Allen, D. E., & Lie, T. T. (1974). Further studies of the fire resistance of reinforced concrete columns׳. National Research Council Canada, Division of Building Research.

Al-Owaisy, L. s. R., & Shallal, M. A. (2007). Strength and elasticity of steel fiber reinforced concrete at high temperatures. Journal of Engineering and Development, 11(2).

Anderberg, Y. (1997). Spalling phenomena ofHPC and oc. Paper presented at the NIST Workshop on Fire Performance of High Strength Concrete in Gaithersburg.

ASTM. (2006). Standard Specification for Concrete Aggregates ASTM C33. ASTM International, West Conshohocken, PA.

ASTM. (2014) Standard Test Method for Static Modulus of Elasticity and Poisson’s Ratio of Concrete in Compression ASTM C469 C469M.

ASTM International, West Conshohocken, PA.

ASTM. (2015). Standard Specification for Portland Cement ASTM Cl240. ASTM International, West Conshohocken, PA.

ASTM. (2015). Standard test method for Flow of Hydraulic Cement Mortar ASTM Cl427. ASTM International, West Conshohocken, PA.

ASTM. (2016). Standard Test Method for Compressive Strength of Hydraulic Cement Mortars (Using 2-in. or [50-mm] Cube Specimens) ASTM Cl 09 C109M. ASTM International, West Conshohocken, PA.

ASTM. (2016). Standard Test Method for Flexural Strength of Concrete (Using Simple Beam with Third-Point Foading) ASTM C78 C78M. ASTM International, West Conshohocken, PA.

ASTM. (2017). Standard Specification for Chemical Admixtures for Concrete ASTM C494 C494M. ASTM International, West Conshohocken, PA.

ASTM. (2017). Standard Specification for Portland Cement ASTM C150. ASTM International, West Conshohocken, PA.

ASTM. (2017). Standard Test Methods for Sampling and Testing Fly Ash or Natural Pozzolans for Use in Portland-Cement Concrete ASTM C311 C311M. ASTM International, West Conshohocken, PA.

ASTM. (2017). Standard Test Method for Splitting Tensile Strength of Cylindrical Concrete Specimens ASTM C496 C496M. ASTM

International, West Conshohocken, PA.

Assa, В., Nishiyama, M., & Watanabe, F. (2001). New approach for modeling confined concrete. I: Circular columns. Journal of Structural Engineering, 127(7), 743-750.

Assakkaf, I. (2004). ENCE 454 Fecture Notes-Design of concrete structures. Department of Civil and Environmental Engineering, University of Maryland, College Park.

Azizinamini, A., Darwin, D., Eligehausen, R., Pavel, R., & Ghosh, s. K. (1999). Proposed modifications to ACI 318-95 tension development and lap splice for high-strength concrete. Structural Journal, 96(6), 922-926.

Bailey, c. (2002). Holistic behaviour of concrete buildings in fire. Proceedings of the Institution of Civil Engineers-Structures and Buildings, 752(3), 199-212.

Bailey, c. (2009). Science and technology developments in structural fire engineering. Structural Engineering International, 79(2), 155-164.

Balaji, A., Nagarajan, P., & Pillai, T. M. (2016). Studies on the behavior of Reinforced Concrete Short Column subjected to fire. Alexandria Engineering Journal, 55(1), 475-486.

Barros, J. A., & Sena-Cruz, J. (1999). Energy absorption capacity of steel fibre reinforced concrete. IRF'99: Integrity, Reliability and Failure, 200.

Bashandy, A. A. (2013). Influence of elevated temperatures on the behavior of economical reactive powder concrete. Journal of Civil Engineering Research, 3(3), 89-97.

Baumgart, c., Cave, s., & Linder, K. (2000). Imaging System for the Automated Determination of Microscopical Properties in Hardened Portland Concrete: Kansas City Plant, Kansas City, MO (US).

Bažant, Z. p., Hauggaard, A. B., Baweja, s., & Ulm, F.-J. (1997). Microprestress-solidification theory for concrete creep. I: Aging and drying effects. Journal of engineering mechanics, 723(11), 1188­1194.

Barrera, A. c., Bonet, J. L., Romero, M. L., & Fernández, M. A. (2012). Ductility of slender reinforced concrete columns under monotonie flexure and constant axial load. Engineering structures, -/¿?(Supplement C), 398-412.

Beyler, c., Beitel, J., Iwankiw, N., & Lattimer, B. (2007). Fire Resistance Testing For Perfomance Based Fire Design of Buildings: final report. Ouincy, MA: The Fire Protection Research Foundation, Batterymarch Park.

Biklliet, M. M., El-S11afey, N. F., & Hl-Hash im y. H. M. (2014). Behavior of reinforced concrete short columns exposed to fire. Alexandria Engineering Journal, 55(3), 643-653.

Bilow N. David , p. E., Kamara E.Mahmoud , Ph.D. (2008). Fire and Concrete Structures Structures Congress 2008.

BIS, I. (2000). 456: 2000, Plain and reinforced concrete code of practicei Bureau of Indian Standards. Fourth revision.

Bodalal, A. s., Shuaeib, F. M., & Karakale, V. (2017). Thermal Simulation and Performance Prediction of High Strength Concrete (HSC) Columns Subjected to Fire.

Buchanan, A. H., & Abu, A. K. (2002). Structural design for fire safety. John Wiley & Sons.

Caprani, c. (2006). Preliminary Design of Building Structures 3rd Year Structural Engineering.

Chen, J., & Young, B. (2006). Stress-strain curves for stainless steel at elevated temperatures. Engineering structures, 28(2), 229-239.

Chen, Y.-H., Chang, Y.-F., Yao, G. c., & Sheu, M.-S. (2009). Experimental research on post-fire behaviour of reinforced concrete columns. Fire Safety Journal, 44(5), 741-748.

Collepardi, s., Coppola, F., Troli, R., & Collepardi, M. (1997). Mechanical properties of modified reactive powder concrete. ACI Special Publications, 172, 1-22.

Connolly, R. J. (1995). The spalling of concrete in fires. Aston University.

Dwaikat, M. B. (2009). Flexural response of reinforced concrete beams exposed to fire׳. Michigan State University.

Dwaikat, M. B., & Kodur, V. (2009). Hydrothermal model for predicting fire-induced spalling in concrete structural systems. Fire Safety Journal, 44(3), 425-434.

El-Dieb, A. s. (2009). Mechanical, durability and microstructural characteristics of ultra-high-strength self-compacting concrete incorporating steel fibers. Materials & Design, 30(10), 4286-4292.

Emberley, R. L. (2013). A Study into the Behavior of Reinforced-Concrete Columns under Fire Exposures using a Spreadsheet-Based Numerical Model. Worcester Polytechnic Institute.

Engineers, I. o. s., & Staff, I. o. s. E. (2007). Guide to the Advanced Fire Safety Engineering of Structures: Institution of Structural Engineers.

Fletcher, I. A., Welch, s., Alvear, D., Lazaro, M., & Capote, J. A. (2007). Model-based analysis of a concrete building subjected to fire.

Fu, Y., & Li, L. (2011). Study on mechanism of thermal spalling in concrete exposed to elevated temperatures. Materials and structures, 44( 1), 361-376.

Hager, L, Zdeb, T., & Krzemień, к. (2013). The impact of the amount of polypropylene fibres on spalling behaviour and residual mechanical properties of Reactive Powder Concretes. MATEC Web of Conferences, 6, 02003.

Harmathy, T. z. (1993). Fire Safety Design and Concrete: Longman Scientific & Technical.

Ho, J. C. M. (2011). Limited ductility design of reinforced concrete columns for tall buildings in low to moderate seismicity regions. The Structural Design of Tail and Special Buildings, 20( 1), 102-120.

IQS. (1984). Iraqi Specifications No. (5), 1984 for Portland Cement.

IQS. (1984). Iraqi Specifications No. (45), 1984 for Aggregates of Natural Resources used for Concrete and Construction.

Kadhum, M. M. (2013). Fire Flame Exposed Behavior Analysis of Reinforced Concrete Columns Eccentrically Under Axial Loads Using Finite Element Method.

Kadhum, M. M., & Manklii, B. s. (2015). Behavior of Reactive Powder Concrete Columns with or without Steel Ties.

Kadhum, M. M. (2015). Prediction of Mechanical Properties of Reactive Powder Concrete by Using Artificial Neural Network and Regression Technique after the Exposure to Fire Flame. 9, 381-399. doi: 10.14525/jjce.9.3.3079

Klioury, G. (2000). Effect of fire on concrete and concrete structures. Progress in Structural Engineering and Materials, 2(4), 429-447.

Khoury, G. (2003). Stress-strain of concrete. Course on effect of heat on concrete, 913.

Khoury, G. (2008). Concrete spalling assessment methodologies and polypropylene fibre toxicity analysis in tunnel fires. Structural concrete, 9(1), 11-18.

Kigha, F., Sadeeq, J., & Abejide, o. (2015). Effects of Temperature Levels and Concrete Cover Thickness on Residual Strength Characteristics of Fire Exposed Reinforced Concrete Beams. Nigerian Journal of Technology, 34(3), 429-437.

Kimura, H., Ishikawa, Y., Kambayashi, A., & Takatsu, H. (2007). Seismic behavior of 200 MPa ultra-high-strength steel-fiber reinforced concrete columns under varying axial load. Journal of Advanced Concrete Technology, 5(2), 193-200.

King, J. (1946). The effect of lateral reinforcement in reinforced concrete columns. Struct. Eng, 24(1), 355-388.

Kodur, V. K. R., & Sultan, M. A. (1998). Structural behaviour of high strength concrete columns exposed to fire.

Kodur, V. (2000). Spalling in high strength concrete exposed to fire: concerns, causes, critical parameters and cures Advanced Technology in Structural Engineering (pp. 1-9).

Kodur, V., & Mcgrath, R. (2003). Fire endurance of high strength concrete columns. Fire technology, 39( 1), 73-87.

Kodur, V., & Phan, L. (2007). Critical factors governing the fire performance of high strength concrete systems. Fire Safety Journal, 42(6), 482-488.

Kodur, V., & Harmathy, T. (2008). Properties of Building Materials, SFPE Handbook of Fire Protection Engineering, PJ DiNenno, Ed. National Fire Protection Association, Ouincy, Mass, USA.

Kumar, V., Iqbal, M. A., & Mittal, A. к. (2017). Energy absorption capacity of prestressed and reinforced concrete slabs subjected to multiple impacts. Procedia Structural Integrity, 6, 11-18. doi: https:// doi.org/10.1016/j .prostr .2017.11.003

Lajtai, E. (1971). A theoretical and experimental evaluation of the Griffith theory of brittle fracture. Tectonophysics, 11(2), 129-156.

Lei, G., Chen, Q., Liu, Y., & Jiang, J. (2013). An inverse method to reconstruct complete stiffness information of rubber bushing. Advances in Materials Science and Engineering, 2013.

Lennon, T., Bailey, c., & Clayton, N. (2002). The performance of high grade concrete columns in fire. Paper presented at the Symposium on Utilization of HPC. Ed.: König, G., Dehn, p., Laust, T., Leipzig.

Lennon, T. (2004). Lire safety of concrete structures. Structural Engineer, 82(6), 22-24.

Lie, T. T. (1989). Lire Resistance of Reinforced Concrete Columns: a Parametric Study. Journal of Fire Protection Engineering, 7(4), 121­129. doi: 10.1177/104239158900100402

Lie, T., & Erwin, R. (1993). Method to calculate the fire resistance of reinforced concrete columns with rectangular cross section. ACI Structural Journal, 90( 1), 52-60.

Lin, T., Zwiers, R., Burg, R., Lie, T., & McGrath, R. (1992). Lire resistance of reinforced concrete columns.

Lublóy, G. L. B. E. (2010). Reinforced Concrete Structures in and After Lire. Concrete Structures, 72.

Maha, M. s. R., Taghreed, к. M. A., & Zeena, w. A. (2013). Behavior of Axially Loaded Reactive Powder Concrete Columns Journal of Engineering and Sustainable Development, 77(2), 193-209.

Mamillapalli, R. s. (2009). Impact of Fire on Steel Reinforcement of RCC Structures.

Massicotte, B., Mossor, в., Piliatrault, А., & Tremblay, s. (1999). Compressive Strength and Ductility of Steel Liber Reinforced Concrete. Special Publication, 182, 163-180.

Melita, P. к. (1999). Advancements in concrete technology. Concrete International, 27(6), 69-76.

Moehle, J. p., & bavanagh, T. (1985). Confinement effectiveness of crossties in RC. Journal of Structural Engineering, 777(10), 2105-2120.

Mohammed, A. z., & Kadhum, M. M. (2016). Structural Behavior of Reactive Powder Concrete Beams Exposed to Fire Flame. (Degree of Master in Civil Engineering - Construction Materials), University of Babylon -College of Engineering.

Module 10 Compression Members (2010). Short Compression Members under Axial Load with Uniaxial Bending. Lesson 23, Version 2 CE IIT, Kharagpur.

Mostafaei, H. (2014). State-of-the-Art on Fire Resistance of Concrete Structure Structure-Fire Model Validation.

Neville, A. M. (1995). Properties of concrete (Voi. 4): Longman London.

Neville, A. M., & Brooks, J. J. (2010). Concrete Technology. Prentice Hall.

Park, R., & Paulay, T. (1975). Reinforced concrete structures׳. John Wiley & Sons.

Pessiki, s., & Pieroni, A. (1997). Axial load behavior of large-scale spirally- reinforced high-strength concrete columns. ACl Structural Journal, 94(3), 304-314.

Pham, Т.-P., & Li, B. (2013). Splitting failure of reinforced concrete columns. Journal of Structural Engineering, 140(3), 04013081.

Phan, L. T. (1996). Fire performance of high-strength concrete: A report of the state-of-the art׳. US Department of Commerce, Technology Administration, National Institute of Standards and Technology, Office of Applied Economics, Building and Fire Research Laboratory.

Phan, L. T., & Carino, N. J. (2000). Fire performance of high strength concrete: research needs Advanced Technology in Structural Engineering (pp. 1-8).

Prablia, s. L., Dattatreya, J., Neelamegam, M., & Seshagirirao, M. (2010). Study on stress-strain properties of reactive powder concrete under uniaxial compression. International Journal of Engineering Science and Technology, 2(11), 6408-6416.

Ramezanianpour, A. A., Ghiasvand, E., Nickseresht, I., Mahdikhani, M., & Moodi, F. (2009). Influence of various amounts of limestone powder on performance of Portland limestone cement concretes. Cement and Concrete Composites, 57(10), 715-720.

Raut, N., & Kodur, V. (2011). Response of reinforced concrete columns under fire-induced biaxial bending. ACI Structural Journal, 108(5), 610. doi: 10.1016/j.engstruct.2012.03.054

Richard, p., & Cheyrezy, M. H. (1994). Reactive powder concretes with high ductility and 200-800 MPa compressive strength. Special Publication, 144, 507-518.

Richard, p., & Cheyrezy, M. (1995). Composition of reactive powder concretes. Cement and concrete research, 25(7), 1501-1511.

Rizk, A. S. s. (2010). Structural Design of Reinforced Concrete Tall Buildings. CTBUH Journal (1), 34-41.

Saatcioglu, M., & Razvi, s. R. (1998). High-strength concrete columns with square sections under concentric compression. Journal of Structural Engineering, 124(12), 1438-1447.

Salah Dimia, M., Guenfoud, M., Gemay, T., & Franssen, J.-M. (2011). Collapse of concrete columns during and after the cooling phase of a fire. Journal of Fire Protection Engineering, 27(4), 245-263.

Seręga, s. (2015). Effect of transverse reinforcement spacing on fire resistance of high strength concrete columns. Fire Safety Journal, 71, 150-16F

Shi, X., Tan, T.-H., Tan, K.-H., & Guo, z. (2004). Influence of concrete cover on fire resistance of reinforced concrete flexural members. Journal of Structural Engineering, 730(8), 1225-1232.

Shihada, s., & Nassar, K. (2012). Improving Fire Resistance of Reinforced Concrete Columns.

Shin, H., Yoon, Y., Lee, s., Cook, w., & Mitchell, D. (2014). Effect of Steel fibers on the performance of ultrahigh-strength concrete columns. Journal of materials in civil engineering, 27(4), 04014142.

Sideris, K., Manita, p., & Chaniotakis, E. (2009). Performance of thermally damaged fibre reinforced concretes. Construction and Building Materials, 23(3), 1232-1239.

Sullivan, T., Calvi, G., & Priestley, M. (2004). Initial stiffness versus secant stiffness in displacement based design. Paper presented at the 13th World Conference of Earthquake Engineering (WCEE).

Tang, w., & Lo, T. (2009). Mechanical and fracture properties of normal­and high-strength concretes with fly ash after exposure to high temperatures. Magazine of Concrete Research, 67(5), 323-330.

Tai, Y.-S., Pan, H.-H., & Kung, Y.-N. (2011). Mechanical properties of steel fiber reinforced reactive powder concrete following exposure to high temperature reaching 800°c. Nuclear Engineering and Design, 241(2), 2416-2424.

The Constructor. (2014). Different Failure Modes of Concrete Columns - Compression Members. URL: https://theconstmctor.org/stmctural- engg/failure-modes-of-concrete-columns/8933.

Topçu, B. í., & Karakurt, c. (2008). Properties of Reinforced Concrete Steel Rebars Exposed to High Temperatures. Research Letters in Materials Science, 2008,4. doi: 10.1155/2008/814137

Topçu, B. í., & Işlkdağ, в. (2008). The effect of cover thickness on rebars exposed to elevated temperatures. Construction and Building Materials, 22(10), 2053-2058.

Ugural, A. c., & Fenster, s. K. (2003). Advanced strength and applied elasticity. Pearson Education.

ünlüoğlu, E., Topçu, i. B., & Yalaman, в. (2007). Concrete cover effect on reinforced concrete bars exposed to high temperatures. Construction and Building Materials, 27(6), 1155-1160.

Willam, K., Xi, Y., Lee, K., & Kim, B. (2009). Thermal response of reinforced concrete structures in nuclear power plants. A report submitted at College of Engineering and Applied Science.

Wille, K., Naaman, A. E., & Parra-Montesinos, G. J. (2011). Ultra-High Performance Concrete with Compressive Strength Exceeding 150 MPa (22 ksi): A Simpler Way. ACIMaterials Journal, 10S( \ ).

Yu, R., Spiesz, p., & Brouwers, H. J. H. (2016). Energy absorption capacity of a sustainable Ultra-High Performance Fibre Reinforced Concrete (UHPFRC) in quasi-static mode and under high velocity projectile impact. Cement and Concrete Composites, ^(Supplement C), 109­122. doi: https://doi.Org/10.1016/j.cemconcomp.2016.02.012

Zhukov, V. (1975). Explosive failure of concrete during a fire. Translation No. DT, 2124.

APPENDIX-A

A

< I

Material Safety Datasheets

A.l Data sheet of steel fibers provided by the manufacturer

(Removed for copyright reasons)

A.2 Data sheet of silica fume provided by the manufacturer

A.3 Data sheet of superplasticizer provided by the manufacturer

APPENDIX-B

Modelling of Material Properties

B.l Introduction

ABAQUS is a complex finite element (FE) package widely used in civil engineering practice. In particular, it is used for modelling of reinforced concrete structures. One of the concrete models incorporated in ABAQUS is the Concrete damaged plasticity model, which considers both the tensile cracking and compressive crushing of concrete as possible failure modes. A short description of concrete and steel reinforcement modelling is given and all the modelling assumptions and element property definition are described below.

B.2 Material Model Properties

B.2.1 Concrete Material

Concrete damaged plasticity is capable of modeling all structural types of reinforced or unreinforced concrete or other quasi-brittle materials subjected to monotonie, cyclic, or dynamic loads. This model is based on a coupled damage plasticity theory and the multi-axial behavior of concrete in damaged plasticity model governs by a yield surface, which proposed by (Lubliner, 1989), as shown in Figure (В-l). Tensile cracking and compressive crushing of concrete are two assumed main failure mechanisms in this model. Furthermore, the degradation of material for both tension and compression behavior have been considered in this model.

illustration not visible in this excerpt

Figure (В-l): Yield surface in plane stress reproduced from (Lubliner, 1989).

Different input data, which should be defined in concrete damaged plasticity, are:

- ψ: is the dilation angle, measured in p-q plane, and should be defined to calculate the inclination of the plastic flow potential in high confining pressures Figure (B-2). In higher level of confinement stress and plastic strain, dilation angle is decreased. Maximum value of it equal 56.3° and minimum value is close to zero. Upper values represent a more ductile behavior and lower values show a more brittle behavior. According to (Malm, 2006) the effect of the dilation angle in values between 30° <ψ< 40°. In some cases can be neglected and for normal concrete 30° is acceptable.

- e: is the flow potential eccentricity. It is a small positive number, which defines the range that the plastic potential function closes to the asymptote as shown in Figure (B-3). The default value in ABAQUS is 0.1 and indicates that the dilation angle is almost constant in a wide range of confining pressure. In higher value of e, with reduction of confining pressure, the dilation angle increases more rapidly. Very small values of in comparison with the default value may cause convergence problems in cases with low confining pressure, due to very tight flow-potential curvature at the point of intersection with the p-axis (Malm, 2006).

- FbO/fcO: is the proportion of initial equibiaxial compressive yield stress and initial uniaxial compressive yield stress. The default value in ABAQUS is 1.16.

- Kc: is the ratio of the second stress invariant in the tensile meridian to compressive meridian for any defined value of the pressure invariant at initial yield. It is used to define the multi-axial behavior of concrete and is 0.5<Kc<l. The default value in ABAQUS is 2/3.

- μ: is the viscosity parameter. According to (Malm, 2006) p=10[7]־is recommended because in comparison with characteristic time increment it should be small.

illustration not visible in this excerpt

Figure (B-2): Uniaxial compressive behavior of concrete reproduced from (Malm, 2006).

illustration not visible in this excerpt

Figure (B-3): Hyperbolic plastic flow rule reproduced from (Malm, 2006).

The properties of concrete are summarized in Tables (В-l), (B-2), (B- 3), (B-4), (B-5) and (B-6).

Table (B-l): General properties of concrete.

illustration not visible in this excerpt

Table (B-2): Elastic properties of concrete.

illustration not visible in this excerpt

Table (B-3): Plastic properties of concrete.

illustration not visible in this excerpt

The stress-strain curves of RPC with 2% steel fiber exposure to high temperatures summarized by (Tai et al., 2011) were used in concrete modelling after converting the nominal strain to inelastic strain as shown in Table (B-4).

Table (B-4): Concrete compression behavior.

illustration not visible in this excerpt

While, stress-strain relationship of RPC with strain rate effects in tension as shown in Figure (B-4), calculated using the tensile stress-crack opening relationship of RPC as shown in Table (B-5), which proposed by

(Fujikake et al., 2006) as follows:

illustration not visible in this excerpt

Cw = crack opening, 02 = 4.8 MPa; W1= 0.4 mm, W2 = 2.0 mm ‘W3 = 4.4 mm. The dynamic tensile strength (Ztf5 d) was given as the function of strain rate ε :

fis, s = static tensile strength =10.8 MPa, and 8st= strain rate corresponding to static loading = 1.0xl0[6]־/s.

Table (B-5): Concrete tensile behavior.

illustration not visible in this excerpt

Table (B-6): Thermal properties of concrete.

illustration not visible in this excerpt

B.2.2 Steel Material

The properties of steel reinforcement are shown in Tables (B-7), (B-8) and (B-9). (Chen and Young, 2006) summarized the elastic and plastic properties of steel reinforcement as shown in Tables (B-7) and (B-8).

Table (B-7): Elastic properties of steel reinforcement.

illustration not visible in this excerpt

Table (B-8): Plastic properties of steel reinforcement.

illustration not visible in this excerpt

Table (B-9): Thermal properties of steel reinforcement.

illustration not visible in this excerpt

REFERENCES

Fujikake, K., Senga, T., Ueda, N., Olmo, T., & Katagiri, M. (2006). Effects of strain rate on tensile behavior of reactive powder concrete. Journal of Advanced Concrete Technology, -/(1), 79-84.

Chen, J., & Young, B. (2006). Stress-strain curves for stainless steel at elevated temperatures. Engineering structures, 28(2), 229-239.

Lubliner, J., Oliver, J., oiler, s., & Oñate, E. (1989). A plastic-damage model for concrete. International Journal of Solids and Structures, 25(3), 299-326. doi: https://doi.org/10.1016/0020-7683(89)90050-4

Malm, R. (2006). Shear cracks in concrete structures subjected to in-plane stresses. KTH.

Tai, Y.-S., Pan, H.-H., & Kung, Y.-N. (2011). Mechanical properties of steel fiber reinforced reactive powder concrete following exposure to high temperature reaching 800 c. Nuclear Engineering and Design, 241(1), 2416-2424.

APPENDIX-C

Governing Equations and Validation Studies

c.l General Approach

The numerical model, proposed here, uses coupled temperature- displacement analysis. RPC column is divided into a number of segments along its length and the midsection of the segment is assumed to represent the behavior of the whole segment. The fire effect simulation is carried out by incrementing time in steps. At each time interval, the analysis is performed through two main steps, namely:

1) Establishing fire temperature due to fire exposure,
2) Carrying out coupled hydro-thermal analysis in each segment to predict cross-sectional temperature and fire-induced spalling.

Dwaikat and Kodur (2009); Ali et al. (2010); Raut and Kodur (2011) developed a governing equations for both of the thermal analysis and pore pressure calculations in order to simulate the effect of fire induced spalling in concrete.

C.2 Thermal Analysis

The temperature is assumed to be uniform along the length of the segment and thus the calculations are performed for a unit length of each segment, the familiar equation defining heat generation and transfer is based on the conservation of energy in (i-directions):

illustration not visible in this excerpt

Applying Fourier’s law, which relates the heat flux to the temperature gradients as:

illustration not visible in this excerpt

Equation (C-l) can be written as:

illustration not visible in this excerpt

Thus, the governing heat transfer equation within a square column cross-section can be written as:

illustration not visible in this excerpt

Where: к = thermal conductivity, pc = heat capacity, T = temperature, t = time, О = heat source, and q = heat flux.

At the fire-column interface, the mechanism for heat transfer is through radiation and convection. The heat flux on the boundary due to convection and radiation can be given by the following two equations, which relates the heat flux due to convection and radiation to the temperature of fire (or ambient temperature for faces not exposed to fire):

illustration not visible in this excerpt

Where:

qrad and qcon = radiative and convective heat fluxes,

hrad and hcon = radiative and convective heat transfer coefficient,

illustration not visible in this excerpt

Те = temperature of the environment surrounding the boundary, σ = Stevan-Boltzman constant = 5.67xl0[8]־ (W/m[2]. °K[4]), and 8 = emissivity.

Hence, the total heat flux on the column boundaries (qb) can be given by the following equation:

illustration not visible in this excerpt

Using Fourier’s Law, the governing heat transfer equation on the boundary of the column can be written as:

illustration not visible in this excerpt

Where:

nx, Пу and nz = components of the vector normal to the boundary, and h = hcon + hrad.

Since the column may be exposed to fire from 1-, 2-, 3-, or 4- sides, two types of boundary equations are to be considered for thermal analysis, namely:

Fire exposed boundaries where the heat flux is governed by the following equation:

illustration not visible in this excerpt

Unexposed boundary where the heat flux equation is given by:

illustration not visible in this excerpt

Where:

hf and he = heat transfer coefficient of the fire side and the cold side, respectively, and Tf and To = fire and cold side temperature, respectively.

C.3 Pore Pressure Calculations

To evaluate fire induced spalling, pore pressure calculations are carried out for each segment at each time step as proposed by (Dwaikat and Kodur, 2009) and (Raut and Kodur, 2011).

Assumptions

The following assumptions are made in the development of spalling sub-model:

- Water vapor is an ideal gas, which is valid for most engineering applications (Harmathy, 1969) and (Huang et al., 1979).
- Mobility of liquid water is ignored. This assumption is valid because Darcy’s coefficient (permeability) for liquid water in concrete is much smaller than that for water vapor (Harmathy, 1969).
- The effect of air is ignored in the analysis. This assumption is considered to be valid because the mass of air in concrete is much smaller than the mass of water.

All the governing equations for the calculation of vapour pressure distribution inside the concrete are derived and presented in (Dwaikat and Kodur, 2009) research and the general form of the equation is as the following:

illustration not visible in this excerpt

Pv= pore pressure, mv= the mass of water vapor, Vv= the volume of water vapor, pL= the density of liquid water, mL= the mass of liquid water, R= the gas constant equal to 8.31446 J/mol/K, M= the molar mass of water, T= temperature, кт= the permeability of concrete at temperature T, μν= the dynamic viscosity of water vapor, mD= the mass of liquid water that formed due to dehydration and t represents time.

C.4 Model Validation

The validation is done to quantitatively measure the agreement between the predictions provided by the model and the real world represented by observations in experiments.

In this section, the computational model developed in this research is compared with the tested columns in two investigations (Ali et al., 2010) and (Raut and Kodur, 2011), in order to ensure that the governing equations of the thermal analysis proposed by those researchers are accurately representing the underlying model and its solution.

The above-described model is validated by comparing temperature variation with time predicted from the model with the measured experimentally in two tested columns by those researchers. The first column was tested by (Ali et al., 2010) and the second column was tested by (Raut and Kodur, 2011). Details of columns configuration are shown in Table (C-

1). It must be illustrated that the validation of the structural response analysis of this model depended only on the comparison with the results of the experimental work presented in this research due to the lack of the data about material characteristics at each temperature level in those researches including (stress-inelastic strain) relationships of concrete and steel, and plastic parameters of concrete which required for the ABAQUS F.E analysis. Also, it should be noted that there is no data available on the behaviour of RPC columns exposed to fire from different sides.

Figures (C-l) and (C-2) show the experimental and numerical predicted temperatures at different positions (column surface, steel reinforcement and column centre) as a function of time for the two columns. These figures indicated that the numerical model results for these columns agree reasonably well with the experimental results.

Table (C-l): Properties of tested column specimens.

illustration not visible in this excerpt

Figure (C-l): Comparison of predicted and measured temperatures in column I, (Ali et al., 2010).

illustration not visible in this excerpt

Figure (C-2): Comparison of predicted and measured temperatures in column II, (Raut and Kodur, 2011).

REFERENCES

Ali, F., Nadjai, A., & Choi, s. (2010). Numerical and experimental investigation of the behavior of high strength concrete columns in fire. Engineering structures, 52(5), 1236-1243.

Dwaikat, M. B., & Kodur, V. (2009). Hydrothermal model for predicting fire-induced spalling in concrete structural systems. Fire Safety Journal, 44(3), 425-434.

Harmathy, T. z. (1969). Simultaneous moisture and heat transfer in porous systems with particular reference to drying. Industrial & Engineering Chemistry Fundamentals, 8(1), 92-103.

Huang, C. L. D., Siang, H. H., & Best, c. H. (1979). Heat and moisture transfer in concrete slabs. International Journal of Heat and Mass Transfer, 22(2), 257-266.

Raut, N., & Kodur, V. (2011). Response of reinforced concrete columns under fire-induced biaxial bending. ACI Structural Journal, 108(5), 610. doi: 10.1016/j.engstruct.2012.03.054

^[1] In few cases, bleeding of the water and vapour from the column surface was observed during the heating process as shown in Plate (4-3, a). The reason for this is may be due to the presence of the moisture within the pore system of the concrete and after the temperature exceeded 100°c, the water started to vaporize, increasing the pore pressure inside RPC and leading to spalling. A part of this vapor is liberated through the debonded interfaces between steel fibre and concrete near the exposed surfaces or from the micro cracks developed at the exposed surfaces. In addition, moisture was also observed in few cases even after testing of the RPC column as shown in plate (4-3, b).

^[2] After the fire duration was finished and fire source was removed, the exposed NSC columns maintained its shape after heating and no evidence of spalling was found, only hairline cracks were distributed along the exposed surfaces as shown in Plate (4-5).