Faculty of Engineering Cairo University MSc Degree, Civil Engineering Department, Faculty of Engineering, Cairo University, Egypt, 2018 REPAIR FOR SLABS SUBJECTED TO FIRE & CASE STUDIES Submitted By: 1- Mohamed Ashraf Mokhtar 2- Ahmed Yousry Tawfiq 3- Mohamed Sayed Mourad 4- Mansour Mohamed Mansour 5- Mohamed Ahmed Kamal 6- Nada Abd El-Kader Ragab 7- Maggi Mamdouh Ahmed 8- Mohamed Ayman Ibarheem 9- Mohamed Refaat Mahmoud Submitted To: Dr/ Hatem Hassan Ali Date: April 15, 2018 1 Table of Content Page Introduction ………………………………………………………………………………….... 3 1. Main Considerations in Fire Protection Structural Design Approach ……………………… 3 2. Comprehensive Concrete Fire Resistance …………………………………………………. 4 2.1. Concrete Provides Comprehensive Fire Protection ………………………………….. 4 2.2. Concrete’s Thermal Performance and Physical Process in Fire ……………………… 6 2.3. Concrete provides effective compartmentation ………………………………………. 8 2.4. On-Site Investigation ………………………………………………………………… 8 2.5. Fire Endurance of Structures ………………………………………………………… 9 2.6. Advanced Analytical Method ………………………………………………………... 9 2.7. Concrete is easier to repair after a fire ……………………………………………….. 10 2.8. Design for Fire Safety with Concrete ………………………………………………… 12 Case Study 1: The Windsor Tower, Madrid, Spain (2005) ………………………………. 14 3. Current design methods and perception …………………………………………………… 15 Case Study 2: Case Study in Lab ……………………………………………………………... 17 A. Description of Concrete Building ……………………………………………………… 17 B. Test results and observations ……………………………………………………………20 C. Damage to the structure …………………………………………………………………22 Case Study 3: Rehabilitation Case Study ……………………………………………………… 24 A. Visual Inspection of Damaged Structural Components ……………………………….. 25 B. Material Tests ………………………………………………………………………….. 28 C. Numerical Analysis ……………………………………………………………………. 31 D. Rehabilitation Plan …………………………………………………………………….. 31 Case Study 4 : Distress Assessment & Rehabilitation of a Fire Damaged Building in Delhi … 33 A. Overview ……………………………………………………………………………….. 33 B. Field Investigation ……………………………………………………………………… 33 C. Lab Investigation ………………………………………………………………………. 34 D. Discussion of Results …………………………………………………………………... 34 E. Recommended Repair and Rehabilitation Measures …………………………………… 35 Case Study 5: Repair Case Study ……………………………………………………………… 35 A. General information about the fire and building ……………………………………….. 35 B. Quality of Concrete …………………………………………………………………….. 37 C. Visual Inspection ……………………………………………………….………………. 38 D. Assessment of the Structure ……………………………………………………………. 40 E. Repair and Strengthening of the Structure ……………………………………………… 40 4. Additional Concrete Applications according to Fire Resistance in Slabs …………………... 42 A. Improving Fire Safety in Road Tunnels ………………………………………………... 42 B. Fire Safety in the residential Buildings …………………………………………………. 43 C. Independent Fire Damage Assessment …………………………………………………. 44 D. Concrete prevents fire spread following earthquakes ………………………………….. 45 E. Lower insurance premiums with concrete ……………………………………………… 45 Conclusion and Recommendations…………………………………………………………….. 47 2 Introduction There are two basic approaches to providing design methods for concrete structures. The first is the traditional, prescriptive code-based approach that addresses the various aspects of life safety independently of one another, using generic rules and formulas. As designs for new concrete structures have increased in complexity, so the prescriptive codes are falling outside their applicable realms. The alternative is to adopt a performance-based, fire engineering design approach that is described in an SFPE guide (SOCIETY OF FIRE PROTECTION ENGINEERS). The performance-based approach takes into account the complexity of modern structures and the interrelationship between the various fire safety measures and systems. Provided any deviations from the Standard can be proven by engineering analysis (often supported by a qualitative or quantitative risk assessment) to be equivalent or superior to the required Standard, then a performance based approach is considered acceptable. Engineering and risk analyses are based on fundamental research into issues such as fire and smoke spread, the behavior of people and response of facilities to fire, to establish a performance based solution. Everyday examples and international statistics provide ample evidence of concrete’s fire protecting properties, and so building owners, insurers and regulators are making concrete the material of choice, increasingly requiring its use over other construction materials. By specifying concrete, you can be sure you have made the right choice because it does not add to the fire load, provides fire-shielded means of escape, stops fire spreading between compartments and delays any structural collapse, in most cases preventing total collapse. In comparison with other common construction materials, concrete offers superior performance on all relevant fire safety criteria, easily and economically. A. Main Considerations in Fire Protection Structural Design Approach Predicting the temperature distribution in the fire compartment Predicting the worst-case temperature distribution involves calculating the fire severity curve according to the different characteristics of the fire compartment. Depending on the complexity of the building geometry, different types of analysis can be adopted. Predicting the temperature of structural members This stage depends on the location of the structural member within the fire compartment, the cross-section factor (exposed perimeter to the cross-sectional area) and the applied fire protection to the structural member. Structural fire analysis This involves carrying out the necessary structural fire analysis to predict the stability of the whole structure, based on the results of the previous stages, using a computer model (see Figure 1.1). It is important to note that the stability of the member and overall structure depends on the following factors: 3 • • • • • • • Maximum temperature and temperature distribution (including penetration) Material properties The applied load The level of composite action with slab The restraints provided by the surrounding structure The continuity/interaction with other cold/heated members Engineering factors. Figure 1.1 Finite element analysis of the deflection profile for a composite concrete floor slab at elevated temperatures B. Comprehensive Concrete Fire Resistance 2.1. Concrete Provides Comprehensive Fire Protection Using concrete in buildings and structures offers exceptional levels of protection and safety in fire: • • • • • • 4 Concrete does not burn, and does not add to the fire load. Concrete has high resistance to fire, and stops fire spreading. Concrete is an effective fire shield, providing safe means of escape for occupants and protection for firefighters. Concrete does not produce any smoke or toxic gases, so helps reduce the risk to occupants. Concrete does not drip molten particles, which can spread the fire. Concrete restricts a fire, and so reduces the risk of environmental pollution. • • • • • • Concrete provides built-in fire protection – there is normally no need for additional measures. Concrete can resist extreme fire conditions, making it ideal for storage premises with a high fire load. Concrete’s robustness in fire facilitates firefighting and reduces the risk of structural collapse. Concrete is easy to repair after a fire, and so helps businesses recover sooner. Concrete is not affected by the water used to quench a fire. Concrete pavements stand up to the extreme fire conditions encountered in tunnels. We have to be prepared for the possible outbreak of fire in most buildings, and its effects. The aim is to ensure that buildings and structures are capable of protecting both people and property against the hazards of fires. Although fire safety codes are written with both these aims in mind, understandably it is the safety of people that often assumes the greater importance. But private owners, insurance companies and national authorities may also have interests in fire safety for other reasons, such as economic survival, data storage, environmental protection and upkeep of critical infrastructure. All of these factors are taken into account in European and national legislation on fire safety, see Figure 2.1.1. Structural fire protection measures must fulfil three aims: • Personal protection to preserve life and health. • Protection of property to preserve goods and other belongings. • Environmental protection to minimize the adverse effects on the environment through smoke and toxic gases as well as from contaminated water used for extinguishing fires. Figure 2.1.1: The Comprehensive Approach to Fire Safety (Courtesy Neck.2002) With concrete construction all three aims can be achieved. Its non-combustibility and high fire resistance mean that concrete provides comprehensive fire protection for people, property and the environment. Concrete’s natural fire resistance properties are compared 5 with other building materials in Table 2.1.1, which shows how concrete scores against a range of key properties. Table 2.1.1: Summary of unprotected Construction Materials Performance in Fire 2.2. Concrete’s Thermal Performance and Physical Process in Fire Due to the low rate of increase of temperature through the cross section of a concrete element, internal zones do not reach the same high temperatures as a surface exposed to flames. The standard ISO 834 fire test on 160 mm wide x 300 mm deep concrete beams exposed three sides to fire for one hour. While a temperature of 600°C was reached at 16 mm from the surface, this was halved to just 300°C at 42 mm from the surface – a temperature gradient of 300°C in just 26 mm of concrete! (Kordina, Meyer-Ottens, 1981). this shows clearly how concrete’s relatively low rate of increase of temperature ensures that its internal zones remain well protected. Even after a prolonged period, the internal temperature of concrete remains relatively low; this enables it to retain structural capacity and fire shielding properties as a separating element. When concrete is exposed to the high temperatures of a fire, a number of physical and chemical changes can take place. These changes are shown in Figure 2.2.1, which relates temperature levels within the concrete (not the flame temperatures) to changes in its properties. 6 Figure 2.2.1 Concrete in fire: physical processes. (Khoury 2000) Heating can change the color of concrete and this may indicate the temperature attained. At above 300ºC a red discoloration is important as it coincides approximately with the onset of significant strength loss. Consequently, any pink/red discolored concrete should be regarded as being potentially weakened. Actual concrete colors observed depend largely on the types of aggregate present in the concrete. Color changes are most pronounced for siliceous aggregates and less so for limestone, granite and sintered pulverized fuel ash. Perhaps the most striking colors are produced by flint (see Figure 2.2.2). The red color change is a function of (ox-disable) iron content and it should be noted that not all aggregates undergo color changes on heating. Also, due consideration must always be given to the possibility that the pink/red color may be a natural feature of the aggregate rather than heat-induced. A number of complementary non-destructive techniques can be used to assess material strength in-situ. These include Schmidt (rebound) hammer and ultrasonic pulse velocity (UPV). Samples of damaged material (and undamaged references) may be removed for laboratory investigation, often by diamond drilling of cores. If it is suspected that the reinforcement bars have been heat damaged it will be necessary to obtain samples and test the tensile strength in the laboratory. Figure 2.2.2 View of a fire-damaged concrete column showing red discoloration of flint fine aggregate particles 7 2.3. Concrete provides effective compartmentation Concrete protects against all the harmful effects of a fire, so that, it is commonly used to provide stable compartmentation in large industrial and multi-story buildings. By dividing these large buildings into compartments, the risk of total loss in the event of a fire is virtually removed – the concrete floors and walls reduce the fire area both horizontally (through walls) and vertically (through floors). Concrete thus provides the opportunity to install safe separating structures in an easy and economic manner; its fire shielding properties are inherent and do not require any additional fire stopping materials or maintenance. 2.4. On-Site Investigation The primary on-site investigation technique is visual inspection, which is used to classify the degree of damage for each structural concrete member. Visually apparent damage induced by heating includes collapse, deflection, spalling, cracking, surface crazing, color changes and smoke damage. Visual survey of reinforced concrete structures is performed using a classification scheme from Concrete Society (see Table 2.4.1). This uses visual indications of damage to assign each structural member a class of damage from 1 to 5. Each damage classification number has a corresponding category of repair, ranging from decoration to major repair. Spalling of the surface layers of reinforced concrete is a common effect of fires. Explosive spalling (see Figure 2.4.1) is erratic and generally occurs in the first 30 minutes of the fire. A slower spalling (‘sloughing off’) occurs as cracks form parallel to the fire-affected surfaces. Forms of cracking include those caused by differential thermal expansion, and thermal shock from quenching by fire-fighting water. Also, differential incompatibility between aggregates and cement paste may cause surface crazing. Table 2.4.1: Simplified visual concrete fire damage classification (after Concrete Society, 1990 and Smart, 1999) 8 Figure 2.4.1 View of the interior of a firedamaged building showing spalling of a reinforced concrete slab soffit. 2.5. Fire Endurance of Structures Figure 2.5 shows the effect of fire on the resistance of a simply supported reinforced concrete slab. If the bottom side of the slab is subjected to fire, the strength of the concrete and the reinforcing steel will decrease as the temperature increase. However, it can take up to three hours for the heat to penetrate through the concrete cover to the steel reinforcement. As the strength of the steel reinforcement decreases, the moment capacity of the slab decreases. When the moment capacity of the slab is reduced to the magnitude of the moment caused by the applied load, flexural collapse will occur. It is important to point out that duration of fire until the reinforcing steel reaches the critical strength depends on the protection to the reinforcement provided by the concrete cover. Figure 2.5 Effect of fire on the resistance of a simply supported reinforced concrete slab 2.6. Advanced Analytical Method Recently some engineers have suggested using 3D finite element software to calculate the change in spatial temperatures over time in structural components using as input the time, temperature, and pore pressure data from the fire analysis described in previous sections. The software has to be able to model the non-linear non-isotropic behavior of reinforcement steel and concrete including crack development and crushing of the concrete. In addition to the external service loads, the model has to be able to include the following: 9 1. Internal forces due to restraints that prevent free expansion. 2. Internal forces due to pore pressure changes. 3. Internal forces due to redistribution due to degradation of the mechanical properties of the steel reinforcement and concrete. 4. Internal forces due to second order effects from the interaction of external loads and the deformations due to the three types of internal forces mentioned above. 2.7. Concrete is easier to repair after a fire The majority of concrete structures is also it can usually be easily repaired afterwards, thereby minimizing any inconvenience and cost. The modest floor loads and relatively low temperatures experienced in most building fires mean that the loadbearing capacity of concrete is largely retained both during and after a fire. Speed of repair and rehabilitation is an important factor in minimizing any loss of business after a major fire; it is obviously preferable to demolition and reinstatement. Up to 300ºC the residual strength of structural quality concrete is not severely reduced. Between 300ºC and 500ºC the compressive strength reduces rapidly and concrete heated in excess of 600ºC is of no use structurally. 300ºC centigrade is normally taken to be the critical temperature above which, concrete is deemed to have been significantly damaged and normally this is replaced if possible. Otherwise the dimensions are increased, depending upon the design loads. At 200-400ºC prestressed steel shows considerable loss of strength, at >450ºC cold-worked steel loses residual strength and at >600ºC hot-rolled steel loses residual strength. Reinforcing steel is much more sensitive to high temperatures than concrete. Figure 2.7.1 shows the effect of high temperature on the yield strength of steel. Figure 4 shows the effect on the modulus of elasticity. As indicated in the figures, hot-rolled steels (reinforcing bars) retain much of their yield strength up to about 800 °F, while cold-drawn steels (prestressing strands) begin to lose strength at about 500 °F. Fire resistance ratings therefore vary between prestressed and non-prestressed elements, as well as for different types of concrete. 10 Figure 2.7.1 Effect of high temperature on the yield strength of steel Concrete element repair will usually include three main processes, the first being removal of damaged concrete by using either power breakers or water jetting (see Figure 2.7.2). After a severe fire it is likely that the second process will comprise removal of weakened reinforcement and connection of new reinforcement. The final part of the repair stage will comprise reinstatement of concrete to provide adequate structural capacity, the necessary durability and fire resistance, and an acceptable appearance (see Figure 2.7.3). An alternative to providing additional steel reinforcement is the use of fiber composite materials (FRPs), bonded to the surface using an epoxy adhesive. Figure 2.7.2 far right: Removal of fire-damaged concrete by water jetting. Figure 2.7.3 Application of sprayed concrete to repair a fire-damaged slab soffit following removal of fire-damaged concrete. 11 Example 1: Fire in a high-rise building in Frankfurt, Germany (1973) In 22 August 1973 a severe fire broke out on the 40th floor of the first high-rise building in Frankfurt. The fire rapidly spread to the 38th and 41st floor, the top floor of this twin block, 140m high office building. The entire vertical and horizontal load-bearing structure of this building was made of reinforced concrete with a double-T shaped flooring system. Because the riser pipes had not been correctly connected, the firefighting could only begin two hours after the fire had started. Three hours later the fire was under control. In all it took about eight hours for the fire to be extinguished (Beese, Kürkchübasche, 1975). All the structural elements withstood the fire although they were exposed to the flames for some four hours. In many places the concrete spalled and in several cases the reinforcement was fully exposed. Fortunately the structure did not fail during the fire and afterwards it was not necessary to demolish entire stories. It was possible to repair most of the elements on site by reusing and strengthening the reinforcement and by concrete guniting . Figure EX1.1: Frankfurt building fire (Courtesy DBV, Germany) Figure EX1.2 Repairing elements with guniting (sprayed concrete) (Courtesy DBV, Germany) 2.8. Design for Fire Safety with Concrete There are five principal objectives that have to be fulfilled when designing a building to be fire safe. Concrete can satisfy all the objectives of fire safety with ease, economy and with a high degree of reliability. The main requirements are shown in Tables 2.8.1 and 2.8.2 shows some examples of how the requirements can be met using concrete construction and demonstrates the comprehensive protective functions of concrete structures. 12 Table 2.8.1: Requirements for fire safety and their relation to concrete Table 2.8.2: The three main fire protection criteria, adapted from Eurocode 2, Part 1–2 A number of different design fire scenarios were considered. The first scenario predicted low peak temperatures (840°C) with a long duration (in excess of 100 minutes). The second scenario predicted high maximum temperatures (in excess of 1000°C) but had a short duration of approximately 30 minutes. It is a matter of some debate as to which is the most severe scenario for this form of construction. For example, a high temperature, short duration fire, may induce concrete spalling due to the thermal shock, whereas a lower temperature but longer duration will result in a greater average temperature in the concrete members. 13 Case Study 1: The Windsor Tower, Madrid, Spain (2005) The Windsor tower was built between 1974 and 1978, it consisted of 32 office stories, five basement levels and two ‘technical floors’ above the 3rd and 16th floors. The structural frame used normal strength concrete in its central core, steel external columns and waffle slab floors, but the most important feature of the tower was to be its two concrete ‘technical floors’. These two ‘technical’ each with eight super-deep concrete beams (measuring 3.75 m in depth; the floor to ceiling height elsewhere), were designed to act as massive transfer beams, preventing progressive collapse caused by structural elements falling from above. The fire broke out, the building was unoccupied. It started on the 21st floor and spread quickly; fire spread upwards through openings made during the refurbishment and via the façade (between perimeter columns and the steel/glass façade), and downwards via burning façade debris entering windows below. The main factors leading to the rapid fire growth and the fire spread to almost all floors included: • The lack of effective fire-fighting measures, such as automotive sprinklers. • The “open plan” floors with a floor area of 1000m2. • The failure of vertical compartmentation measures, in the façade system and the floor openings. The height, extent and intensity of the blaze meant firefighters could only try to contain it and protect adjacent properties, so the fire raged for 19 hours, engulfing almost all the floors (see Figures EX2.1 and EX2.2). When the fire was finally extinguished, the building was burnt out completely above the 5th floor, much of the façade was destroyed and there were fears that it would collapse. However, throughout the fire and until eventual demolition, the structure remained standing; only the façade and floors above the upper concrete ‘technical floor’ suffered collapse. The passive resistance of the concrete columns and core had helped prevent total collapse, but the role of the two concrete ‘technical floors’ was critical, particular the one above the 16th stories, which contained the fire for more than seven hours. It was only then, after a major collapse, that falling debris caused fire to spread to the floors below this, which burned, but again damage was limited to the stories above the lower ‘technical floor’ at the 3rd level. 14 Figure EX2.1 Above The fire rages in the Windsor Tower, Madrid. (Courtesy IECA, Spain) Figure EX2.2 Top left The façade above the technical floor at level 16 was totally destroyed. (Courtesy IECA, Spain) Figure EX2.3 Buckling of unprotected steel perimeter columns at the 9th floor (Photo: Colin Bailey) The Spanish research Centre Institute de Ciencias de la Construcción Eduardo Torroja (IETcc) in collaboration with the Spanish Institute of Cement and its Applications (IECA), investigated the reinforced concrete structural elements of the Windsor Tower. It was observed that the temperature reached inside the concrete was 500 ºC at a distance of 5 cm from the surface subjected to fire. C. Current design methods and perception It is thought that a common cause of collapse is due to the structure as a whole being unable to accommodate or resist the large horizontal displacements induced by the thermal expansion of heated floor slabs. Also the removal of cover to reinforcing bars by spalling could lead to premature failure of members or entire structures, and the cause has generally been attributed to poor continuity of reinforcement or poor workmanship during construction. The underlying assumption is that the fire resistance of the complete structure will be at least equal to the fire resistance of its individual members. However, the behavior of the entire structure when considered as a collection of connected members is significantly different from the behavior of individual isolated members, with both beneficial and detrimental effects occurring. 15 An obvious difference between whole building and member behavior is that the structure will utilize actual load path mechanisms that cannot be identified from member testing. For example, restraint to thermal expansion of a floor slab, caused by a cooler part of the structure surrounding a heated compartment, can induce high compressive stresses into the heated slab, which can be beneficial by inducing compressive membrane action to support the applied load. These high compressive forces can also, however, increase the slab’s susceptibility to spalling. Another form of whole building behavior that can lead to premature collapse results from the inability of vertical members to resist or accommodate the large lateral movements caused by the thermal expansion of the heated floor slab. The spalling behavior is difficult, if not impossible, to predict and this is probably the reason why design codes provide no definitive guidance. The behavior of spalling is typically classified into three types; aggregate spalling, explosive spalling and corner spalling. Aggregate spalling occurs in the early stages of the fire and involves small pieces flying off the surface. As the name suggests aggregate spalling is mainly caused by the break-up of the aggregate at elevated temperatures, but typically results in only superficial damage. Flint and Thames gravel are particularly susceptible to this type of spalling. Explosive spalling, which is extremely violent, also occurs early in the fire, and is the most serious in terms of causing premature collapse of the structure. The consequence of explosive spalling has resulted in it being the most researched, with the main causes being attributed to heating rate, moisture content, and permeability of the concrete and mechanical stress levels. Corner spalling is non-violent and occurs later in the fire and results in large sections of concrete at the corners of columns or beams breaking off. In real fires corner spalling will not significantly affect the exposed steel reinforcement, since the fire temperatures are typically low towards the end of the fire. However, in fire resistance tests on members where the temperatures continue to rise for the duration of the test, corner spalling can be significant, resulting in failure of the member. In the current UK Code BS8110-Part 1 of BS8110 assume that severe spalling of the concrete will not occur if the cover is less than 40mm. In cases where the cover is greater than 40mm the designer is referred to Part 2 of BS8110, where the use of steel fabric is specified as additional reinforcement. Although a reasonable solution, practically placing steel fabric between the main bars and face of the member is, at best, difficult. However, the Code suggests that spalling of concrete is only a particular problem for reinforcement covers exceeding 40 to 50mm. The limited guidance given in the Code ignores the severe spalling that has been experienced in some fire tests, and observations from real fires in concrete structures, where the cover to the reinforcement was significantly less than 40mm. The latest draft of the EN version of the Eurocode has provided some design guidance on spalling of concrete, which states that it is unlikely when the moisture content of concrete is less than 3% by weight. The Code also states that this moisture content will not be exceeded if the structure is subjected to an internal exposure. The draft EN Code also provides the guidance that if tabulated data are used, which are similar to the prescriptive rules in BS8110, then the effects of spalling can be ignored. 16 Where spalling is considered possible, the Eurocode states that the member should be designed assuming local loss of cover to the reinforcement. However, the Code also recommends that the reduction in design strength due to spalling is not necessary for solid slabs with evenly distributed bars or beams with a width larger than 400mm and containing more than 8 bars in the tensile area. Outside the UK, the use of high-strength concrete is starting to become popular since it enables thinner cross-sections to be specified. However, it has been shown that this type of concrete has a high propensity for explosive spalling in fire. The latest draft of the EN version of the Eurocode has identified this problem and provides some design guidance. This guidance consists of using a supplementary steel mesh or including in the concrete mix a minimum of 2.0kg/m3 polypropylene fibers, which melt at approximately 160°C and have been shown to reduce the risk of spalling. Case Study 2: Case Study in Lab A. Description of Concrete Building The full-scale seven-story insitu-concrete building constructed at the BRE Laboratories in Cardington, was designed to Eurocode 2 and BS8110 and represented a commercial office building situated in the center of Bedford. The building was constructed between January and April 1998 and incorporated different concrete mixes, construction techniques and rationalization of reinforcement layouts. As a consequence, the finished building had a mixture of concrete strengths and different reinforcement layouts for each floor. The completed building comprises 3 bays by 4 bays each 7.5m, with two core areas (Figure CS2.1), which include steel cross-bracing to resist lateral loads. Each floor slab is nominally 250mm thick, grade C37 normal weight concrete and designed as a flat slab supported by internal columns 400mm square and external columns 400mm by 250mm. The reinforcement in the first and second floor slabs is traditional loose bar, with hookand-bob links for shear resistance around the columns. Above the first and second floors the reinforcement layouts change to allow different design methods and reinforcement rationalization to be investigated. 17 Figure CS2.1 Plan of building showing location of fire compartment The required funding for a large-scale compartment fire test on the concrete building was secured on the 5th July 2001. Due to the BRE corporate decision to ‘mothball’ the Cardington Laboratories on the 28th September 2001, the design of the test, construction of the compartment, installation of the instrumentation and execution of the test had to be carried out within a twelve-week period. This limited time-scale restricted the amount of instrumentation used and forced the test to be carried out between the ground and first floor, which saved time by removing the need to transport personnel and materials up the building. A fire compartment, with a floor area of 225m2, was constructed between the ground and first floors (Figures CS2.1 and CS2.2). Due to the height of the footings, the overall height of the compartment from the laboratory floor to the underside of the first floor was 4.25m (Figure CS2.2). The compartment walls were constructed using 140mm thick Top-Crete blocks and lined, inside the compartment, with one skin of plasterboard. The laboratory floor was also lined with plasterboard. The blockwork was constructed such that a gap of 325mm was left between the top of the wall and underside of the slab. This gap was filled with a ceramic blanket which was fixed to the wall and shot fired to the underside of the slab (Figure CS2.3). It was assumed that the design and construction of the compartment wall would contain the fire, whilst allowing vertical movement of the tested structure, without providing any support. 18 The tested first floor slab was nominally 250mm thick and designed as a flat slab. The reinforcement comprised 12mm and 16mm loose bars at various centers. 12mm diameter hook-and-bob links were provided around the columns consisting of 72 links around the internal columns, 60 links around an external column and 41 links around a corner column. The floor was constructed using a C37 normal-weight concrete, with a cube strength at 28 days of 61N/mm2. Tests on the wet properties of the concrete using a Rapid Analysis Machine (RAM) showed that the average cement content was 407 kg/m3 and the average water content was 205 kg/m3. The aggregate source makes the concrete susceptible to aggregate spalling. The average moisture content, measured seven days before the fire test, was 3.8% by weight. Figure CS2.2 Cross-section through the building showing location of fire test Figure CS2.3 Detail showing fixing of ceramic blanket to first floor soffit and compartment wall The design specification for the building stated that the nominal cover to all reinforcement should be 20mm. To check if this cover was achieved for the completed structure, a cover survey was carried out prior to the test, with the results shown in Table 1. This survey showed that the cover to the reinforcement was more than sufficient to achieve at least 60 minutes fire resistance, as specified in current codes. The codes also state a minimum size for members to achieve a required fire resistance. For 60 minutes BS8110 Part 1 gives a minimum slab thickness of 95mm and a minimum column width of 200mm, for columns fully exposed to the fire. The section sizes on the Cardington building where significantly greater than these values and therefore had a greater fire resistance than the minimum specified. 19 Table CS2.1: Measured cover to main reinforcement The design load for the floors is shown in Table 2. The behavior of structures subjected to fire can be treated as an accidental limit state, with appropriate load factors. These factors are shown in Table 2 and are based on the values given in ENV 1991-1(20) and the proposed amendment to BS5950 Part 8. Table CS2.2 Design load at the fire limit state To represent the design imposed, partition, raised floor, ceiling and services load, sand bags each weighing 10.79kN, were placed on the first floor to achieve a vertical load of 3.25kN/m2 over the fire compartment area. Additional sandbags were placed on the floors above the tested first floor to achieve a total design axial load, excluding self-weight, of 925kN in column C3 (central column within the fire compartment area) and 463kN in column D3. This design load assumes a 40% reduction in the partition and imposed load as stated in BS6399 Part 1. A protected steel safety cage, comprising four steel columns, which were tied together, was placed around column C3 within the fire compartment. The cage was designed to support the first floor slab should the tested 400mm × 400mm high-strength concrete column (C3) fail during the fire. The floors above the first floor were shown, by simple calculation, to be adequate to bridge column C3 by tensile membrane action. The constructed cage is shown in Figure 4. The cage was not required during the test since the concrete column performed adequately, as discussed in Section 6.0. B. Test results and observations Figure CS2.4 shows the recorded maximum atmosphere temperature and the average atmosphere temperature throughout the fire compartment. Considering Figure CS2.4, it can be seen that the temperatures reduced considerably between 12 and 13 minutes after the start of logging (6 minutes after ignition). A video and time-lapse photos of the test show clearly that explosive spalling of the soffit of the floor slab began at this time. The spalling, which was extensive, reduced the severity of the fire throughout the test, which is an effect that has also been shown from other fire tests. 20 The vertical displacements of the slab, at the locations shown in Figure CS2.1, are shown in Figure CS2.5, up to the time at which the instruments malfunctioned. It can be seen that the displacements near the center of the building, at locations V4, V5, V8, V11 and V12 show a distinct plateau of constant displacement after 15 minutes. The displacements at locations V6, V9 and V13, towards the edge of the building, are much larger than the displacements near the center and show no signs of a plateau. The failure of the blanket around the top of the compartment increased the ventilation to the fire. Numerical models used to predict the time-temperature response suggest that this increase in ventilation had the effect of slightly increasing the maximum temperature of fire and shortening the duration of the fire. Figure CS2.4 Recorded atmosphere temperatures throughout the compartment 300mm below the soffit of the slab. Figure CS2.5 Measured displacements of the slab (For location of measurements refer Figure CS2.1) 21 C. Damage to the structure The structure showed no signs of collapse during or after the fire. Following the fire, the displacement transducers were re-wired and the residual displacements recorded. The residual vertical and horizontal displacements at first floor level are shown in Figure CS2.6. The extent of the horizontal movement of the columns along gridline D, caused by thermal expansion of the heated slab. The horizontal movement of the floor slab also caused buckling of the steel cross-bracing positioned on gridline B and C (refer Figure CS2.6), as shown in Figure CS2.7. An obvious question is why is there a lateral residual movement of 67mm of column D3, and why did it not recover as the slab cooled and the thermal expansion reduced? The column itself had insufficient strength to hold the slab out and no significant tensile cracks were observed in the floor slab. The only plausible explanation, at present, is that when the slab is heated the aggregate expands and the cement paste contracts. The difference in expansion causes internal cracking between the aggregate and paste resulting in overall irrecoverable thermal expansion. In addition the phase changes of siliceous aggregates when heated produce an increase in volume that is irreversible on cooling. Figure CS2.6 Residual horizontal and vertical displacements 22 Figure CS2.7 Buckling of bracing between columns B4 and B5 (similar buckling of bracing between columns C2 and C1 was also observed). There was significant cracking around column D3, which can be attributed to the hogging moment from the slab and the moment caused by the P-δ effect from the vertical load in the column and its lateral movement due to thermal expansion of the floor slab. The behavior of the structure and recorded crack layout can be investigated more fully by the planned computer modelling of the test. The extent of the spalling, which exposed the bottom reinforcement in the heated floor slab is shown in Figure CS2.8. The pictures shown in Figure CS2.9 also provide an appreciation of the extent of the spalling, and show reinforcement hanging from the soffit of the floor slab. A total of 10 reinforcement bars, 12mm diameter and 2.6m long were found on the compartment floor. Figure CS2.8 Extent of the spalling exposing bottom reinforcement in the slab Figure CS2.9 Picture showing spalling (looking at column B2) Examination of the area of the spalling, and the type of explosive nature of the spalling observed during the test, suggests that high compressive forces were induced in the slab due to restraint to thermal expansion. The behavior of the spalling around the service holes 23 (Figure CS2.10) also suggests that spalling was primarily due to high compressive stresses, since where the compressive stresses are reduced along the edges of the hole no spalling occurred. Although these compressive stresses almost certainly led to the severity of the spalling, they were also probably beneficial to the survival of the floor slab due to compressive membrane action. Provided the slab does not reach vertical displacements greater than approximately 0.5 times its depth, then the effect of compressive membrane action will considerably enhance the load-carrying capacity of the slab above that calculated assuming flexural action. The behavior of compressive membrane action will be considered in more detail during the planned numerical modelling of the test. The debris from the spalling on the floor of the compartment (Figure CS2.11) was 90mm deep at the worst location. This debris clearly reduced the severity of the fire, as observed during the test. Figure CS2.10 Picture showing spalling around service hole Figure CS2.11 Picture showing extent of spalling (debris) on compartment floor (50 pence shown for scale) Case Study 3: Rehabilitation Case Study This example presents a case study to provide a reference for the rehabilitation of firedamaged RC structures. The target structure is the main control building of a thermal power plant. The fire broke out in one of battery rooms in the building for an unknown reason and lasted for three hours before it was extinguished. The fire was contained within the battery room and did not spread to the other part of the building. As the power plant was under construction at the time of fire, the building was not in use and there were no casualties due to the fire other than to the building structure. A team of structural and material researchers/engineers was dispatched to the site to evaluate the condition of the structure and to set up appropriate corrective measures. The location and identification number of each structural member are indicated in Figure CS3.1. 24 Figure CS3.1 Locations of structural elements and batteries in the battery room A. Visual Inspection of Damaged Structural Components The visual inspection revealed that most of the batteries in the room were burned down, especially the Batteries #1, #4, #5 and #6. For Batteries #2 and #3, part of the battery structures still remained. This condition coincides with accounts stating that the fire burned longest in the region of Batteries #4 and #5. As a result, the structural elements located above Batteries #1, #4, and #5 experienced relatively severe damage compared with those located above Batteries #2 and #3. The soffits of the slabs were mostly damaged due to fire as shown in Figure CS3.2. The concrete in the bottom cover and at mid-depth spalled, exposing the reinforcing bars. The damage was more severe in Slabs #3 and #4 than in Slabs #1 and #2. The condition of each slab can be summarized as below: • Slab #1: More than 50% of the concrete spalled and 60% of bottom reinforcing bars were exposed. • Slab #2: About 50% of the concrete spalled and 50% of bottom reinforcing bars were exposed. • Slab #3: More than 90% of the concrete spalled. About 90% of bottom reinforcing bars and 40% of top reinforcing bars were exposed. • Slab #4: More than 90% of the concrete spalled. About 90% of bottom reinforcing bars and 30% of top reinforcing bars were exposed. 25 Figure CS3.2 Damage in each span of slab on its soffit, observed from the north. (a) Slab #1, (b) Slab #2,n(c) Slab #3 and (d) Slab #4 Figure CS3.3a shows the floor conditions of the HVAC (Heating, Ventilation and Air Conditioning) room, which is located right above the battery room where the fire occurred. It indicates that, in the area where the concrete has spalled, a network of cracks has developed on the top surface of the floor slab. At the time of the visual inspection, the damage to the floor slab of the battery room was not identifiable, since the debris on the floor had not been removed (see Figure CS3.3b). Inferring from the fact that the steel structures supporting the batteries remain without excessive deformation, it is assumed that there was little or no structural damage to the floor slab. At the time of the visual inspection, the damage to the floor slab of the battery room was not identifiable, since the debris on the floor had not been removed (see Figure CS3.3b). Inferring from the fact that the steel structures supporting the batteries remain without excessive deformation, it is assumed that there was little or no structural damage to the floor. 26 Figure CS3.3 (a) Damage to top surface of Slab #3 and (b) pileup of debris under Slab #3 and #4. Having lost more than half of the concrete volume, all the spans from Slabs #1 to #4 are judged to have degradation in their structural capacity. The main role of the slab is to transfer the transverse load on the slab to adjacent girders and beams. This role must be restored by either replacing the damaged span of slab with new ones or by providing additional load carrying capacity to the slab. The exposed reinforcing bars might have experienced a strength reduction due to the elevated temperature during the fire. To identify the degree of the strength reduction, tests on the rebar specimens from the respective spans of slab were necessary. Although not identifiable at the time of inspection, the floor slab of the room was assumed to have retained its structural capacity and needed no retrofitting. The degree of damage in the girders varied by component. At some points, the concrete covers at the bottom corner and side of girders and beams spalled, although it is not dominant, the longitudinal bars and stirrups were exposed. As shown in Figures CS3.4, CS3.5 the condition of the girders can be summarized as follows: • • • Girder #1: One longitudinal rebar at the bottom corner was exposed by 1–2 m, and 7 stirrups were exposed due to the spalling of concrete both at side cover and inside the stirrups. Girder #2: No marked exposure of the longitudinal rebar was observed. Concrete spalled on the side of Slab #3 at a slight depth. Two stirrups were partly exposed toward the end of the span at the location of Column #3. Girder #3 to #6: These girders comprise the longitudinal edge of the room. No marked damage observed on these girders except that a longitudinal bar at the corner of Girder #6 was exposed by 3 m. Longitudinal rebar was exposed at the corners, and stirrups were slightly visible at some areas. In spite of the spalling of the concrete cover and the exposure of reinforcing bars, the girders and beams were deemed to still retain their designed structural capacity, judging from the fact that neither flexural cracks at mid‐ span nor diagonal cracks toward the span ends were found, and deflection at the center of the element was not detected with the naked eye. However, a possible degradation in the load‐ carrying capacity must be retrofitted by testing the strength of the damaged concrete and the exposed reinforcing bars, and then by providing the required amount of reinforcement to compensate for the degraded capacity. Figure CS3.4 Damage to Girder #1. (a) Exposed stirrups in the north region. 27 (b) West side of girder seen from the north Figure CS3.5 Damage to Girder #2. (a) East side of girder seen from the north. (b) West side of girder seen from the north B. Material Tests from the north In order to supplement the results of the visual inspection discussed in the previous section, both an indicator test and a strength test were performed. This section summarizes and discusses the results of these tests. 1. Neutralization Test of Concrete To determine the degree of neutralization of the concrete by its pH level, the phenolphthalein (C20H14O4) indicator method was adopted. The key indicator of this test is that a solution of Phenolphthalein remains colorless in acids or neutral liquids, but turns bright reddish/pink in alkalis. Therefore, it is possible to infer the temperature history and neutralization depth through the change of color in the test surface of concrete, which permits the soundness and the reusability of the concrete exposed to fire to be judged. For the normal condition of concrete, calcium hydroxide, Ca(OH)2, is created in the hydration process of cement, which is described by CaO + H2O = Ca(OH)2 (1) Calcium hydroxide is a strongly alkaline product with a pH value ranging from 12 to 13. One of the particular characteristic of calcium hydroxide is that it decomposes into calcium oxide (CaO) and water by chemical decomposition at elevated temperatures (above 500 _C), such as those in a fire environment. As a result, the concrete gets neutralized, which is described by Ca(OH)2 ]above 500 _C[= CaO + H2O (2) During the test on the girder, 1% phenolphthalein solution was first applied to its side surfaces that was affected by the fire without treating the surfaces. As shown in Figure CS3.6a, the color did not change in most of the region, with a purple area where the alkaline concrete reacted to the indicator. When the indicator was applied after chipping out the 28 surface of concrete by 10 mm, a significant change in the color was immediately observed as shown in Figure CS3.6b. Therefore, the girder is judged not to have experienced a maximum temperature of more than 500˚C and, as a result, have remained sound. We infer that the concrete in the girders spalled during the fire not due to the degradation of the concrete itself, but to the sudden increase in temperature at the upper level in the room. The elevated temperature is deemed to have caused the expansion of the aggregate inside the concrete and to have weakened the so-called transition zone between the aggregate and the cement matrix. Figure CS3.6 Neutralization test on (a) the side surface and (b) inside of Beam #2 2. Strength Tests Two types of strength tests were performed on the concrete cylinders and rebar coupons. The purpose of the tests is to evaluate the remaining capacity of the structural element and, consequently, to determine the degree of retrofit required to recover the original structural capacity. For the strength test of the concrete cylinders, two cylinders with diameters of 100 mm were sampled from Girder #2 indicated in Figure CS3.7. Examples of the cored specimens are shown in Figure CS3.8. The compressive strength tests of these specimens were conducted per ASTM (American Society for Testing and Materials) C42. The results are summarized in Table CS3.1, and also per ACI 318-05. 29 Figure CS3.7 Locations of sampled concrete core and rebar coupons Figure CS3.8 (a) Cored specimens and (b) specimen 3-1 after the test CS3.1 Test results of sampled core cylinders Table Three rebar coupons 600 mm in length were sampled from each span of slab and also from one of the damaged girders, Girder #1. Their yield strength, tensile strength, and elongation were obtained as Table CS3.2 according to ASTM E8-09. As shown in the table, the yield and tensile strengths of the longitudinal rebar in Girder #1 are much higher than the designed yield strength of 400 MPa. The yield strength of the stirrup is slightly below the design strength but almost reached the level that would not degrade the shear strength of the girder. In the case of the reinforcing bars obtained from the slabs, the measured strengths are on the level of 60%–70% of the original strength, except for the Slab #2 where 30 the fire was first extinguished. The specimen ø14-2 showed sufficient strength levels, but the degree of elongation was relatively smaller than other rebar coupons. The results of the tests discussed in this section show that the compressive strength of concrete cores and the yield strength of longitudinal rebar were above the design strength, and the yield strength of the stirrups was slightly below the design strength. Only the residual strength of reinforcing bars obtained from the slab generally degraded from their original capacity. Table CS3.2. Test results of sampled rebar coupons C. Numerical Analysis This section discusses the results of the finite element analysis on a typical sample girder and two typical columns in the Battery Room #6 performed to investigate their remaining structural capacity. As target structural components, Girder #2 and Columns #3 and #4 were selected. The following assumptions are made in the course of modeling. They are based on the results of the on-site inspection and tests on the sampled core cylinders (see Table CS3.1) and rebar coupons (see Table CS3.2): I. Rebar is undamaged. II. The depth of damaged concrete is about 50 mm. III. The residual strength of the damaged concrete is approximately 60% compared with the original capacity. A material and geometric nonlinear finite element analysis was conducted on the target structural components using Abaqus/Standard v6.12. In this model, the behavior of concrete under compression is defined by Mander’s model, which is discussed in, while its behavior under tension is defined by a bi-linear tension softening curve with a peak strength corresponding to 10% of its compressive strength. D. Rehabilitation Plan Based on the results of the visual inspection, material tests, and finite element analysis discussed in the previous sections, the remaining structure needs to be rehabilitated. The scope of the rehabilitation can be summarized as follows: • Operation floor structure (slabs, girders): severe or partial damage to the concrete and reinforcing bars, thus retrofit is required. • Mezzanine floor structure: damage cannot be identified but no marked damage is inferred, thus inspection is required after removing the debris on the floor. The retrofit methods of the damaged girders can be summarized as: 31 • • Surface repair by epoxy-injection or mortar-grouting. Structural retrofit by bonding carbon fiber sheet or by anchoring steel plate around the damaged surface. Regarding the retrofit methods of the damaged slabs, there are two options depending on the handling method of the HVAC system located right above the battery room. The differences between the two options are analyzed in Table CS3.3. They are also illustrated in Figures CS3.9 and CS3.10 respectively. Table CS3.3 Comparison of retrofit plans for the damaged slabs Figure CS3.10 Figure CS3.9 Rehabilitation plan with HVAC (Heating, Ventilation and Air Conditioning) removed during rehabilitation works Figure CS3.10 Alternative rehabilitation plan with HVAC remaining in position during rehabilitation work Case Study 4 : Distress Assessment & Rehabilitation of a Fire Damaged Building in Delhi 32 A. Overview To determine the extent of fire damage occurred in effected RCC Columns, Beams and Slabs of the building field assessment covering Quality assessment using UPV testing, measurement of concrete cover depth and carbonation depth, followed by laboratory scale assessment of residual strength of concrete, OM study, XRD study and DTA study is reported in this paper. Highlights on suitable materials & techniques to carryout repair and strengthening of the RCC Slabs are also briefly described in this paper. B. Field Investigation Quality of Concrete using UPV testing technique as per IS 13311 Part-I-1992: Quality of concrete was assessed by UPV testing (using Portable Ultrasonic Non-Destructive Digital Indicating Tester-PUNDIT) on the selected locations of RCC columns and Beams. Measurements were taken by direct method (cross-probing) as shown in figure CS4.3. Based on Ultrasonic Pulse Velocity obtained, the overall quality of concrete was graded as ‘Good’. Concrete cover study: The concrete cover study carried out by the Ferro-scanning Technique (using Ferro Scanner) to identify the thickness of concrete cover provided on the various identified locations of the fire damaged area of Building. The observations were made on the 8 numbers identified RCC members. The results indicated that the average concrete cover was varying from 35mm to 45mm in Slabs. Carbonation study: Carbonation study was carried out by spraying a pH indicator (solutions of 1%phenolphthalein in 70%ethyl alcohol) on freshly extracted concrete core samples from identified representative RCC members. The depth of carbonation in different samples was found to be varying from 0mm to 10mm. . Figure CS4.2: View of lift lobby in fire damaged area Figure CS4.1: Spalling of Plaster in RCC Beams & Slabs 33 Figure CS4.3: UPV measurement being recorded C. Lab Investigation Optical Microscopy (OM) Study: The petrographic studies had been carried out on three concrete core samples from identified locations of RCC Column C32, Slab SI & Slab S6 (Refer figure CS4.4) under Stereoscopic Microscope (NIKON SMZ-1500). Three thin slices (up to 5mm thick) from each of concrete core were drawn, one from top portion of core i.e. from exterior face /top of RCC member (0-30mm), second from middle portion of the core (30-60mm) and third from bottom portion of the core (beyond 60mm). The study was carried out to determine various micro structures, morphological features and pore distribution and development. Based on the results obtained on above parameters inferences were drawn to establish the distress development caused of the fire. Figure CS4.4: S-6-Bottom3: Stretching of coarse and fine aggregate due to temperature rise C. Discussion of Results 1. Visual observation had indicated cracking & spalling cement mortar plaster from various RCC Columns, Beams & Slab of fire damaged room of Building. In the area outside fire damaged room no spalling of cement mortar plaster was observed. 2. The UPV test results indicated that the Quality of concrete was “Good” in various RCC members, which were exposed to the fire. 3. The depth of carbonation assessed on various core samples extracted from identified location indicated that the carbonation depth was 0mm in 8 RCC members. 4. The average cover thickness in various tested RCC members was found to be 35mm to 45mm in slab. 5. The results of chemical analysis covering chloride content, sulphate content and pH value were found to be within the permissible limits as prescribed in IS: 456 – 2000. 6. The microscopic investigation of the fire damaged cores indicated that the distribution pattern of the grains and pores spaces changed drastically. Even the morphology of the grains of the fine aggregate component was also changed. Large variation was noticed in air void distribution pattern from the bottom to top portion of the roof. Bottom portion was directly exposed to fire, which had caused sealed walled air voids bigger 34 in size. Size of pores after fire might have increased but have very stable and firm walls, which had presently helped to achieve more strength to the concrete. D. Recommended Repair and Rehabilitation Measures 1. Removal of all cracked and loose cement plaster from the RCC members and walls in fire damaged area by chiseling out surface concrete up to 20mm depth in distressed structural elements and to remove the existing plaster over the wall also. 2. Applying bond coat of SBR (Styrene Butadiene Rubber) latex-based polymer modified cement slurry in proportions 1:1 (1 cement: 1 polymer) to be applied on the prepared surfaces of concrete/substrate. 3. To rehabilitate RCC roof slab by Polymer Modified Mortar (PMM) using emulsified SBR latex conforming to ASTM C1059-2013Type-I in damaged areas (1 Cement-3 Part graded cleaned riversand+15 %Latex by weight of cement) with 0.35w/c ratio, in 10-15 mm thick layers by applying bond coat between successive/each layers including leveling and profiling complete. Proper curing to be done for the repair work. Gunny bags should be used for effective curing. Polypropylene fibers to be added to reduce shrinkage. Case Study 5: Repair Case Study A. General information about the fire and building In 2005, a fire broke out at the "Cubrilovic" Cultural Centre in Gradiska. The fire caught all floors and was extinguished after 4.5 hours. During the fire, installations and the interior of the building as well as the timber roof structure were completely destroyed, while the bearing reinforced concrete structure was partially damaged (Figures CS5.1 and CS5.2). The "Cubrilovic" Cultural Centre in Gradiska was built in 1977. The part that was caught by fire is an independent structural entity that is separated by expansion joint from the movie theatre and "Simpo" furniture store. It consists of four stories (basement, ground floor, 1st and 2nd floor), and the dimensions at the base are 30 by 30 m. The Cultural Centre itself, taking the space between axis H-N and 1-7, is schematically represented in Figure CS5.3. Figure CS5.1 View of the front facade of the building after fire Figure CS5.2 Severely damaged RC slab after fire 35 Figure CS5.3 Disposition of building parts and adopted marks of the axis Structure of the object is the classic skeletal structure with RC walls as seismic stiffening. Basic elements of the load-bearing structure that was caught by fire are: - RC columns, 30 × 50 cm, - RC walls, width 20 and 30 cm, - RC partitions, width 10 cm, - RC beams, spanning 5 and 2.5 m (30 × 50 cm), - RC beams, spanning 15 m (30 × 110 cm), and - RC full slabs, 12, 13, 14 and 16 cm thick (continually two-way slabs, continual one-way slabs and cantilever slabs). In order to locate the position of individual elements of the structure, the axis markings are shown on the base of the ground floor (Fig. CS5.4). Main staircase is of reinforced concrete and was made as an elbowed one-way slab. Final supports are RC walls, and middle support is a cantilever beam. Auxiliary staircase is made as a double RC staircase with a half-landing. Designed concrete class for all elements of the bearing RC structure is C30/37. For reinforcing of the bearing structure ribbed reinforcement 400/500, mild reinforcement 240/360 and welded wire-mesh reinforcement 500/560, were used. Figure CS5.4 Base of ground floor After chase cutting, while checking all the elements of the structure, it is found that all slabs have ribbed reinforcement with different arrangement (R∅10/15, R∅10/20, R∅12/15 and R∅10/30), depending on the static system of the slabs. B. Quality of Concrete 36 In order to ascertain the quality of the concrete used in the load bearing RC structure and to ascertain the thickness of the concrete damaged in the fire, 35 cores were taken out (24 cores from the part of the building caught by fire, and 11 from the part that did not burn). Cutting drill equipment and samples prepared for the testing are presented in Figures CS5.5 and CS5.6. Testing results are presented in Table CS5.1 (part of the building caught by fire is colored). Figure CS5.5 Core cutting drill (RC beam with long span l=15 m) Figure CS5.6 View of a part of cores, prepared for testing Table CS5.1 Single and mean values of concrete compressive strength • 37 Based on the result of the analysis of the concrete compressive strength, the following conclusions have been made: a. In both sets of the results there are large differences in individual strength, within a single element of the structure, which shows the variations in the concrete quality, which are primarily the consequence of internal defects in the concrete structure caused by poor adhesion of cement stone and large grains of aggregate (Fig. CS5.7). b. Individual compressive strengths smaller than 20 MPa were gained on the concrete cores with defective and damaged structure. c. There is no significant difference in compressive strength between the part of the object being caught in the fire and the part that has not burned if cores with defective and damaged structure are excluded. d. All average values of concrete compressive strength in the part of the object that was caught by the fire are above 25 MPa, which proves that there is a "healthy" concrete core in the structure. Figure CS5.7 View of a plane fracture of concrete core C. Visual Inspection Reinforced concrete slabs originally had only a finishing decorative layer from the bottom side, and later a lowered ceiling was done, mostly from combustible materials. All the slabs contain numerous wirings, placed without a regular arrangement, whose heating and combusting made the surrounding concrete burn through, so that the fire damage to the concrete, at these places, is of higher degree. Due to high temperatures in those zones, there is also a local deformation of the reinforcement. On the bottom surface of the slabs there are also defects from the period when the structure was built (concrete honeycombing, linear segregation with de-levelling, insufficient thickness of the protective layer of the concrete, visible reinforcement and inappropriately performed breaking and continuations of concreting) that added to the damage caused by the fire and weather. All slabs show marks of water leaking, with sediment products of calcium-hydroxide wash-out and salt due to many months of exposure to rain and snow that enter through the damaged roof. Leaking of water from the roof and white sediments of washing out products are most noticeable on the slabs above the second floor. Due to occasional presence of moisture, some reinforcement bars have started to corrode. Upper surfaces of the slabs were protected via the cement screed and mineral wool, so they showed no damage from the fire. By analyzing the results of visual inspection of RC slabs (above ground floor, 1st and 2nd floor), from the bottom side, that were exposed to multiple effects of the fire, it has been concluded that: - Thin net-like fissures in the surface layer of the concrete were recorded in about 20 % of the total number of slabs. - Surface layer of concrete that is burnt through, dilapidated or fallen off was recorded in about 60 % of the total number of slabs. The thickness of this layer is 20 cm. At the places where wiring is set, the concrete has burned through deeper. The described damage is also more pronounced at the places of defects, especially at concrete honeycombing and breaking of concreting. On slabs with smaller span this damage is local and exists in several places. On slabs with the span of 15 m (J-M/3-5) the described damage has caught most of the downside surface of the slab, and white, broken grains of the aggregate were recorded, apart from the change in color of the concrete. For these slabs, falling off of concrete up to and behind the reinforcement is characteristic. 38 - Longitudinal, transverse and slanting fissures and cracks were recorded in about 60 % of the total number of slabs. On most recorded fissures, marks and products of water leaking are noticeable. A number of fissures are located at the area of wiring in the slab. -On cantilever slabs (1-1’ and 7-7’) transverse fissures are characteristic, probably located at places where there was a break in applying of the concrete layer. On square slabs, spanning 5 m, fissures appear in both ways, and on some there are also diagonal fissures. On slabs spanning 15 m, longitudinal fissures are characteristic (at the third and in the middle of the span). They are located at places of badly performed breaking and continuations of concreting. The width of the fissures is from 0.1 to 0.3 mm. - A small protection layer (bared – visible reinforcement) was recorded on about 35 % of the total number of slabs. - Surface reinforcement corrosion was recorded on about 10 % of slabs. Poor adhesion between the aggregate and the cement stone is characteristic for the concrete built-in in reinforced concrete slabs, as well as the concrete built-in in other elements of the structure. Characteristic appearance of the bottom side of RC slabs after the fire can be seen in Figures CS5.8 and CS5.9. Figure CS5.8 RC slab above the first floor (K-LM/3-5) Figure CS5.9 RC slab above the first floor (L-M/3-5), detail D. Assessment of the Structure Based on the analysis of all the test results, both on the field and in the laboratory, and the data gathered through visual inspection, it has been concluded that: 39 • Reinforced concrete structure was done with many defects (honeycombing, unequal and often not sufficient protective layer of concrete, irregularly done break of concreting). The mentioned defects have all contributed to the faster appearance and greater propagation of damage due to the fire. • Principle of wiring guidance through the bearing elements of RC structure, which was applied on this building, caused significant damage to the surrounding concrete and reinforcement during the fire. These defects were especially pronounced on 5 m RC slabs. • Due to great variations in the concrete compressive strength, it has been suggested that for the control calculation of the structure, for all the bearing elements, in the repair project, lower concrete class has to be used (C25 instead of C30). This would partially include the defects in the concrete structure, mostly poor adhesion between the large grained aggregate and the cement stone. • The built-in reinforcements in the bearing elements of RC structure have mechanical characteristics corresponding to the quality of mild reinforcement (240/360) and the ribbed reinforcement (400/500). • During the fire, installations and the interior of the object as well as the wooden roof structure were completely devastated, while the bearing RC structure was partially damaged. The degree and the character of the damage done to the elements of the bearing reinforced concrete structure, caught in the fire, is such that the entire load bearing capacity of the structure has decreased, but the global stability of the structure has not been jeopardized. It was concluded that with the appropriate repair and strengthening measures, this part of the structure can be brought to a designed state of load bearing capacity and stability. E. Repair and Strengthening of the Structure Strengthening of two-way RC slabs (5 × 5 m and 5 × 2.5 m) was executed by adding new reinforcement and a new concrete layer on the bottom side of the slab. Strengthening of the continuous RC slabs was done by adding new reinforcement and a new protective layer of concrete on the bottom side of the slab, (Figures CS5.10 and CS5.11). The new concrete layer was done as a shotcrete. Strengthening in supporting zones was also done by adding new reinforcement and a new protective layer of concrete on the upper side of the slab. It refers only to the slabs for which it was established by calculation that they lack reinforcement due to bending moments. Figure CS5.12 shows a characteristic detail which served to solve the anchorage problem of the additional reinforcement in RC wall zone, and Figure CS5.13 execution of the strengthening. 40 Figure CS5.10 Repair of RC slab (additional reinforcement and executing of new concrete layer) Figure CS5.11 A detail of strengthening of RC slabs (additional reinforcement and new concrete layer) Figure CS5.12 Details of strengthening of RC slabs in the zone of existing RC wall Figure CS5.13 Execution of strengthening of RC slabs in the zone of existing RC wall Strengthening of cantilever slabs was executed by increasing the critical cross-section, i.e. by adding new concrete in supporting zones (Figures CS5.14, CS5.15 and CS5.16). Figure CS5.14 Detail of strengthening of RC cantilever slabs Figure CS5.15 Execution of strengthening of RC cantilever slabs 41 4. Additional Concrete Applications according to Fire Resistance in Slabs A. Improving Fire Safety in Road Tunnels Europe is served by over 15,000 kilometers of road and rail tunnels; these are part of our transport infrastructure and are particularly important in mountainous regions, but increasingly so in major cities where tunnels can relieve traffic congestion and free up urban spaces. The problem is that accidents involving vehicles can cause extremely severe fires; tunnel fires tend to reach very high temperatures due to the burning fuel and vehicles, reportedly up to 1350oC, but more usually around 1000 – 1200oC. Peak temperatures are reached more quickly in tunnels compared with building fires, mainly because of the hydrocarbons in petrol and diesel fuel, but also because of the confined spaces (see Figure EX2.1). Figure EX2.1: Tunnel fires burn at very high temperatures. (Courtesy J-F Denoël/FEBELCEM, Belgium) It must be remembered that tunnel fires are likely to be some of the most extreme fires experienced. With these very high temperatures, some spalling from concrete surfaces is to be expected. Much research effort has gone into developing lining materials to minimize the effects of spalling from concrete surfaces when exposed to severe fires (e.g. Khoury, 2000). There is clear evidence that the addition of monofilament polypropylene fibers to the concrete mix is an effective solution and creates a concrete that can ‘breathe’ in a fire situation, making it less likely to spall. 42 B. Fire Safety in the residential Buildings In a comparison of fire safety in concrete and timber frame construction, Professor Ulrich Schneider of Vienna University of Technology identified that seven specific risks arise from the use of a combustible construction material (such as timber) within a building structure and envelope (Schneider and Oswald, 2005); these are listed below: 1. 2. 3. 4. 5. 6. 7. An increase in fire load. An increase in smoke and pyrolysis products. Higher amounts of carbon monoxide. Fire ignition of structural elements. Fire ignition inside construction cavities. Danger of smouldering combustion and imperceptible glowing (pockets of embers). Increasing occurrence of flashovers. Schneider went on to examine fire death statistics from various countries and established a clear link between the number of fire victims and construction materials used in buildings, as shown in Figure EX3.1. His detailed study of typical timber construction details showed that failure in a fire could occur through ignition and collapse of structural or non-structural elements and via metallic connectors within the timber structure, which soften on exposure to fire and lose their loadbearing capacity. Schneider also found that fire spread between adjacent rooms and/or apartments was accelerated significantly in buildings where timber materials or cladding had been used as part of the external wall. Figure EX3.1: Fire deaths compared with construction type in five major countries (1994 – 1996). (TUW, Vienna, Schneider and Oswald 2005) 43 C. Independent Fire Damage Assessment In Sweden, Olle Lundberg undertook an independent investigation of the cost of fire damage in relation to the building material with which the houses are constructed, based on statistics from the insurance association in Sweden (Forsakringsforbundet). The study was limited to larger fires in multi-family buildings in which the value of the structure insured exceeded €150k; it covered 125 fires that occurred between 1995 and 2004. (These amounted to 10% of the fires in multi-family homes, but 56% of the major fires.) The results showed that: • • • • • The average insurance payout per fire and per apartment in timber houses is around five times that of fires in concrete/masonry houses (approx. €50,000 compared with €10,000). A major fire is more than 11 times more likely to develop in a timber house than in one built from concrete/masonry. Of the burned houses, 50% of the timber houses had to be demolished, compared with just 9% of the concrete ones. In only three of the 55 fires in concrete houses did the fire spread to neighboring apartments. Of the 55 fires, 45 were in attics and roofing; typically the fire starts in the upper dwelling, it spreads to the attic and roofing (wood). Example 4: Timber construction site fire, Colindale, London (2006) During the construction of a major new residential complex in North London, a fire broke out and ignited several six-story timber frame blocks (see Figures EX4.1). The fire burned for five hours; it took 100 firefighters and 20 fire engines to bring it under control. Eyewitnesses reported that the blocks were destroyed within minutes. Shortly after the fire, an air quality monitoring station nearby recorded a significant rise in toxic PM10 particulates, which can have serious health implications for people with breathing difficulties. About 2,500 people were evacuated from the surrounding area, a major road was closed for two hours and a local college hall of residence was affected so badly that students could not return. Fortunately the housing development had not been occupied by new residents and the college was largely empty during the summer holidays. Nevertheless, the disruption was significant. Local building control officers expressed concern, noting that “if you have concrete floor design and there’s a fire, then it’s going to compartmentalize. If you have timber, it’s going to burn right through”. At the time of writing, at least one block of the development was due to be rebuilt – this time using concrete. 44 Figure EX4.1 The fire at Colindale raged for five hours in the partially constructed timber frame residential blocks and took 100 firefighters with 20 fire engines to control it (Courtesy John-Macdonald-Fulton, UK) D. Concrete prevents fire spread following earthquakes The seismic design considerations that apply in some countries require designers to pay attention to the specific problem of fires following earthquakes. This has been given due consideration in countries such as New Zealand, where concrete structures have been identified as having a low level of vulnerability to the spread of fire following earthquakes (Wellington Lifelines Group, 2002). E. Lower insurance premiums with concrete Every fire causes an economic loss and in most cases it is insurers that have to pay for the damage caused by fires. For this reason, insurance companies maintain comprehensive and accurate databases on the performance of all construction materials in fire – they know that concrete offers excellent fire protection and this is reflected in reduced insurance premiums. Across Europe, insurance premiums for concrete buildings tend to be less than for buildings made from other materials (which are more often affected badly or even destroyed by fire). In most cases, concrete buildings are classified in the most favorable category for fire insurance due to their proven fire protection and resistance. Of course, every insurance company will have its own individual prescriptions and premium lists; these differ between countries, but because of concrete’s good track record, most offer benefits to owners of concrete buildings. When calculating a policy premium, insurers will take the following factors into account: • • • • • • • • 45 Material of construction Type of roof material Type of activity/building use Distance to neighboring buildings Nature of construction elements Type of heating system Electric installation(s) Protection and anticipation (preparedness) Example 5: Insurance premiums for warehouses in France Unfortunately, very little data on insurance costs is made publicly available, but some comparative studies do exist. In France, CIMbéton (2006) published a summary and insurance cost model based on insurers’ views of single-story warehousing/industrial buildings. The study explains that insurance premiums are based on a number of factors, including the activity within the building and construction material. The building material is certainly important – the structure, exterior walls, number of floors, roof covering and furnishings are all taken into account in the calculations. The results show clearly the extent to which concrete is preferable to other materials, such as steel and timber, for all parts of the building. For example, by selecting a concrete frame and walls for a single story warehouse means a possible 20% reduction on the ‘standard’/average premium paid. Changing this for a steel frame and cladding option would add 10 to 12% to the ‘standard’ premium, therefore making at least a 30%. Case Study 5 Insurance premiums for warehouses in France 23 difference in total. In deciding the final premium, the insurers also take into account security equipment, fire prevention and suppression measures, which includes compartmentation – a fire prevention option in which concrete excels. Table EX5.1: Insurance premiums for a 10,000 m2 warehouse (single story, no furnishings); total insured = EUR 25 million (CIMbéton, 2006) Repair Methods & Techniques: - Suggested remedies: Formwork should be placed to support slabs Figure (3) 46 - New rebar with the same count and diameters as the existing rebar should be connected to the existing rebar, in case lap splices corresponding to bar diameters according to ACI 318 requirements are not met new rebar should be drilled for in the existing columns with a minimum of 150mm deep hole, then hole should be cleaned and new rebar installed then the hole should be filled with epoxy. OR new rebar could be welded to the existing rebar according to AWS standards. - Existing concrete surfaces should be cleaned and free from any loose concrete and paint or plaster. New concrete grade (C30) is to be poured. Suggested remedies: Existing concrete surfaces should be cleaned and free from any loose concrete and paint or plaster. Form work should be installed. New rebar with the same count and diameters as the existing rebar should be connected to the existing rebar, in case lap splices corresponding to bar diameters according to ACI 318 requirements are not met new rebar should be drilled for in the existing columns with a minimum of 150mm deep hole, then hole should be cleaned and new rebar installed then the hole should be filled with epoxy figure (4). - OR new rebar could be welded to the existing rebar according to AWS standards. - New reinforced concrete slab with (160mm) thickness and 5 diameter 12mm per meter bottom rebar both directions, and 5 diameter 12mm per meter top reinforcement on top of beams is to be poured. Figure 4 47 - Suggested remedies: Concrete surface should be prepared by removing Plastering and paint. - - Removed works should be done carefully without introducing any additional stresses in structural elements, including the affected slabs themselves. Concrete surface should be cleaned with high pressure water blasting (10 to 35Mpa). New rebar with the same count and diameters as the existing rebar should be connected to the existing rebar, in case lap splices corresponding to bar diameters according to ACI 318 requirements are not met new rebar should be drilled for in the existing columns with a minimum of 150mm deep hole, then hole should be cleaned and new rebar installed then the hole should be filled with epoxy. OR new rebar could be welded to the existing rebar according to AWS standards. - Formwork should be installed. Concrete surfaces should be painted with sikadur 3-2 epoxy. New concrete grade (C30) arch with thickness 250mm –to match existing width- should be poured. - Suggested remedies: Existing concrete surfaces around the hole should be cleaned high pressure water blasting (10 to 35Mpa). Old surfaces should be roughened. Formwork below the hole should be placed. Incase rebar does not exist in the hole area new rebar should be drilled for in existing concrete with 100mm depth hole, 5 bars diameter 10mm in both directions rebar are then installed in holes and the holes are epoxy filled. A bonding material between existing and new concrete should be applied on concrete surfaces. New concrete slab grade C30 should be poured. - - d- Stairs and its beams destroyed 48 - Suggested remedies: All remaining steps of the stairs must be demolishes - New beam at floor mid height should be constructed. - Steel brackets should be placed on both sides of the damaged beams in connection with column as per the detail provided by the consultant figure (3) with dimensions (1400x500x20)mm with 8 diameter 18 hilti bolts anchored to column - New rebar with the same count and diameters as the existing rebar should be connected to the existing rebar, in case lap splices corresponding to bar diameters according to ACI 318 requirements are not met new rebar should be drilled for in the existing columns with a minimum of 150mm deep hole, then hole should be cleaned and new rebar installed then the hole should be filled with epoxy. OR new rebar could be welded to the existing rebar according to AWS D1.1 standards. - Existing concrete surfaces should be cleaned and free from any loose concrete and paint or plaster. - Dowels for new stairs must be installed in the new beam (6 bars with 16mm diameter per meter bottom and top bars) - Formwork for new beam is to be installed. - New concrete grade (C30) is to be poured. - Inclined form work for stairs is to be installed - rebar for stairs are connected to the existing rebar at basement floor, at stair mid height and at ground floor level. - New rebar with the same count and diameters as the existing rebar should be connected to the existing rebar, in case lap splices corresponding to bar diameters according to ACI 318 requirements are not met new rebar should be drilled for in the existing columns with a minimum of 150mm deep hole, then hole should be cleaned and new rebar installed then the hole should be filled with epoxy. OR new rebar could be welded to the existing rebar according to AWS D1.1 standards. 49 - Existing concrete surfaces should be cleaned and free from any loose concrete and paint or plaster. - A bonding material between existing and new concrete should be applied on concrete surfaces. - Reinforcements for stair steps are installed as per figure (6) - New concrete slab 200mm thick grade C30 should be poured. Figure 6 Small cracks in slabs 50 - Suggested remedies: Concrete surface should be prepared by removing plastering and paint. - Removed works should be done carefully without introducing any additional tresses in structural elements, including the affected slabs themselves - Concrete surface should be cleaned with high pressure water blasting ( 10 to 35Mpa). - Wire fabric (cheken wire) to be installed on slab damaged surface connected by shear connectors to the existing slabs. - Form work should be installed. - Then cementitious mortar with (3:1) ratio and sika latex or equivalent with (1:1) ratio should be used. - New rebar with the same count and diameters as the existing rebar should be connected to the existing rebar, in case lap splices corresponding to bar diameters according to ACI 318 requirements are not met new rebar should be drilled for in the existing columns with a minimum of 150mm deep hole, then hole should be cleaned and new rebar installed then the hole should be filled with epoxy. OR new rebar could be welded to the existing rebar according to AWS standards. - 51 Existing concrete surfaces should be cleaned and free from any loose concrete and paint or plaster. New concrete grade (C30) is to be poured. 52 53 Conclusion and Recommendations ❖ Concrete structures completely demolished by fire are rare in practice. Most of the facilities with RC structure have been repaired and used again, even those which have been exposed to great fires. ❖ This report presents more than one case study on establishing a rehabilitation plan for a fire-damaged reinforced concrete structure. A team of structural and material researchers/engineers in concrete did lots of studies in this field and carried out both on-site and laboratory tests on the damaged concrete and rebar. A finite element analysis can be also performed to estimate the residual capacity of the structure. The following steps are the conclusion of these studies and the recommended rehabilitation methods: (1) Through visual inspection of the structures, the mostly damaged structural elements can be gotten and evaluated the degree of fire. (2) The judgment from the visual inspection is supported by the material test using the indicator method (3) Strength tests on concrete cores and rebar coupons obtained from the slabs showed the degree of concrete and the reinforcing bars damage. (4) Then the recommended solutions can be produced for slabs: replaced, or repaired. (5) Finally a plan for this solution can be put. ❖ The design procedure used in fire safety engineering takes into account the following factors to calculate the design value of the fire load, from which individual structural members can be assessed and the overall probability of a fire causing structural damage can be established: • The characteristic fire load density per unit of floor area. • The expected fire load caused by combustion of the contents (combustion factor). • Fire risk due to the size of the compartment (large compartments are given a higher risk factor). • The likelihood of a fire starting, based on occupants and type of use (use factor). • Ventilation conditions and heat release. 54 ❖ Common rules for fire safety engineering methods do not exist, user-friendly software is still under development and there are significant variations in approach, experience and levels of acceptance by authorities. FSE has to be used with care through appropriate experts and proper evaluation of its assumptions. Serious concerns have been raised about the validity and accuracy of the probability based calculations, with critics noting that a faulty FSE calculation could lead to a catastrophe. Others have voiced fears that inexperienced, inexpert attempts to use FSE could lead to misunderstandings in calculations and the wrong results. Large variability of parameters within the assumptions underpinning the calculations could include, but are not limited to, the following aspects: • Fire brigade success rates: again, average values are provided, but are clearly not applicable to all buildings; there will be significant variation in performance. • Human behavior: assumptions are made on how people will behave in an emergency, but there is a very high degree of variability here related to crowd behavior and means of escape. • Reliability of sprinkler systems: average values are given, but there are many types of systems to suit all types of buildings. • Arson or deliberate fires (i.e. caused by criminal intent) – these are not really covered sufficiently. Some building types and locations will naturally be more vulnerable to crime. 55 References • Holistic behavior of concrete buildings in fire. Professor Colin Bailey. Manchester Centre for Civil and Construction Engineering Published in the Proceedings of the Institution of Civil Engineers, Structures and Buildings 152, August 2002, Issue 3, pp 199-212 http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.534.3296&rep=rep1&type =pdf • Fire and Concrete Structures Authors: David N. Bilow, P.E., S.E., Director, Engineered Structures, Portland Cement Association 5420 Old Orchard Road, Skokie, IL 60077, Phone 847-972-9064, email: dbilow@cement.org Mahmoud E. Kamara, PhD., Senior Structural Engineer, Portland cement Association 5420 Old Orchard Road, Skokie, IL 60077, Phone 847-972-9012, email: mkamara@cement.org http://www.cement.org/docs/default-source/th-buildings-structures-pdfs/fire-concretestruc-sei-08.pdf • Comprehensive fire protection and safety with concrete http://www.bibm.eu/Documenten/ecpComprehensive_fire_protection_and_safety_with_concrete.pdf • V P Chatterjee, Satish Sharma, Adarsh Kumar, Rizwan Anwar, Y.N. Daniel National Council for Cement and Building Materials, India • A Case Study on the Rehabilitation of a Fire-Damaged Structure www.mdpi.com/journal/applsci 56 • Behavior of Reinforced Concrete Slab Subjected To Fire International Journal of Computational Engineering Research (ijceronline.com) Vol. 3 Issue. 1 • Turning up the heat – full service fire safety engineering for concrete structures • ASSESSMENT AND REPAIR OF THE BEARING STRUCTURE OF THE GRADISKA CULTURAL CENTRE AFTER FIRE Vlastimir Radonjanin, Mirjana Malešev, Radomir Folić, Ivan Lukić Tehnički vjesnik 21, 2(2014), 435-445 57