Fire safety of concrete structures: Background to BS 8110 fire design

Transcription

1 Fire safety of concrete structures: Background to BS 811 fire design Tom Lennon FRS, the Fire Division of BRE

2 Fire safety of concrete structures: Background to BS 811 fire design Tom Lennon FRS, the Fire Division of BRE

3 BRE is committed to providing impartial and authoritative information on all aspects of the built environment for clients, designers, contractors, engineers, manufacturers, occupants, etc. We make every effort to ensure the accuracy and quality of information and guidance when it is first published. However, we can take no responsibility for the subsequent use of this information, nor for any errors or omissions it may contain. BRE is the UK s leading centre of expertise on building and construction, and the prevention and control of fire. Contact BRE for information about its services, or for technical advice, at: BRE Garston Watford WD25 9XX Tel: Fax: enquiries@bre.co.uk Details of BRE publications are available from: or IHS Rapidoc (BRE Bookshop) Willoughby Road Bracknell RG12 8DW Tel: Fax: brebookshop@ihsrapidoc.com Published by BRE Bookshop Building Research Establishment Watford WD25 9XX Tel: Fax: brebookshop@emap.com BR 468 Copyright The Concrete Centre First published 24 ISBN Requests to copy any part of this publication should be made to BRE Bookshop.

4 Contents Introduction 4 Description of the project 5 Historical development in national concrete codes 6 Comparison between tabulated values from different codes 14 Experimental background to tabulated values 15 Other relevant research 3 Discussion 33 Conclusions and recommendations 36 References 37 Appendix A Results from National Building Studies Research Paper No Appendix B Results from National Building Studies Research Paper No Appendix C - Results of fire resistance tests on elements of building construction 43 Appendix D - Results from Fire Research Note Executive summary This report has been prepared at the request of The Concrete Centre and the British Cement Association to investigate the background to the methods for establishing the fire resistance of concrete structures specified in the relevant parts of the UK concrete Code BS 811 1,2. The work focused on the original research and test results underpinning the tabulated data in BS 811, which have been revisited in order to assess the relevance of the approach to modern forms of concrete construction. This study is important in that it brings together in one document a body of information covering test results and research carried out over a number of years. There was a danger that much of the important work in support of the development of codes and standards would be lost. Hence a study was carried out to collate and assess all relevant information to ensure that the important lessons from the past are recorded and to help define the strategy for a new generation of codes and standards. The investigation shows that the experimental results used as the basis for developing the tabulated data in BS 811 support the provisions of the Code in relation to assumed periods of fire resistance. In many cases the provisions are very conservative as they are based on the assumption that structural elements are fully stressed at the fire limit state. Fire safety of concrete structures: Background to BS 811 fire design 3

5 Introduction For many years the most common method of ensuring compliance with the requirements of the Building Regulations in terms of the fire safety of concrete buildings has been to rely on tabulated values for minimum dimensions and minimum cover to reinforcement. Historically both reinforced and prestressed concrete have been shown to provide good resistance to fire. A study commissioned by the Fire Resistance Committee of The Concrete Society 3 investigated a large number and variety of fire-damaged concrete structures within the UK. The authors concluded that almost without exception the structures performed well during and after the fire, and that the majority of structures were repaired and re-used. However, the report emphasised the need to establish the circumstances under which spalling would have serious consequences. The information in the codes is based on the results from standard fire tests on elements of construction. Such tests generally assume that the structural element is fully stressed at the time of the fire. This is a conservative assumption. The provisions in terms of cover are based on limiting the temperature of the reinforcing or prestressing steel to a single critical value. The development of fire engineering methods has questioned the relevance of standard fire testing in relation to the performance of actual buildings subject to real fires. In recent years concrete construction has become more efficient with the use of chemical admixtures to improve workability, increase strength and reduce curing times. Modern concrete frames tend to consist of more slender members with all aspects of the design process rationalised to improve the speed and economy of construction. There is a need to assess the performance of modern concrete construction against the provisions of the Code, and to identify areas where the design can be made more efficient. Previous research 4 identified the lack of up-to-date data on the effect of fire on concrete structures. It pointed out that the industry is in danger of employing the material inefficiently and that design rules were based on research conducted many years ago. The report emphasised the need to conduct research in order to fill in gaps in the industry s knowledge and to keep abreast with the advances in concrete technology that had taken place in the preceding 1 to 15 years. The traditional means of ensuring compliance with the regulatory requirements for fire safety for elements of structure is to adopt the prescribed values set out in tables A1 and A2 of Approved Document B to the Building Regulations. The values relate to a minimum period for which the element must survive in the standard fire test measured against the relevant performance criteria of stability, integrity and insulation. Stability or loadbearing capacity relates to the period of time a structural element can maintain the appropriate design load during a fire test. Integrity measures the ability of an element (structural or non-structural) to prevent the passage of flames or hot gases during a test. Insulation is a measure of the ability of the material to prevent a prescribed temperature rise on the unexposed face during the prescribed period. For elements such as beams or columns the only relevant performance criteria is loadbearing capacity, whereas for loadbearing separating elements, such as compartment floors and walls, all three requirements have to be met for the prescribed period of fire resistance. 4 Fire safety of concrete structures: Background to BS 811 fire design

6 Description of the project This project set out to summarise the background to the design values in the current concrete Code, BS 811-2: in relation to fire resistance. This report details the background to the Code provisions informed by a comprehensive search of the available information from the Fire Research Station (FRS) archives. It provides information on the major changes since the publication of the 1948 version of CP as they relate to traditional structural elements (beams, columns, floor slabs and walls). The information covers both reinforced and prestressed concrete. Available test results are provided in the appendices with the information broken down, where possible, in terms of support conditions, period of fire resistance attained, type of aggregate used, moisture content, overall dimensions, load level and the presence and nature of spalling. The prescriptive design approach has served the profession well because of its inherent simplicity. However, in recent years fire engineering design has moved away from prescriptive design solutions towards a more performance-based approach. It is important that the concrete industry develops further guidance for designers and regulators to promote the use of a material with a history of very good performance in fire. There is now an opportunity to move away from the prescriptive approach in the UK Code and develop performance-based methods for the structural fire engineering design of concrete structures. The assistance of the following is gratefully acknowledged: Prof Colin Bailey, UMIST Dr Pal Chana, BCA Rohan Rupasinghe, BRE Fire safety of concrete structures: Background to BS 811 fire design 5

7 Historical development in national concrete codes This section considers the evolution of the concrete design codes in relation to the provisions for fire resistance. Table 1 Thickness (mm) of walls and floors for fire-resisting purposes (CP 114: 1948) CP 114: 1948 The starting point for this study is the provisions in the 1948 version of CP Although there was not a great deal of information on fire resistance within this Code it did include tabulated values for the fire resistance of walls and floors to achieve specific fire resistance periods. Table 14 from the Code is reproduced below with the critical dimensions converted to metric units. Grade of fire resistance 6 hours 4 hours 2 hours 1 hour 1 /2 hour Thickness of wall to attain grade: With class 1 aggregates With class 2 aggregates Thickness of solid reinforced concrete slab Thickness of concrete slab and solid material in tiles of hollow tile floor The 1948 version of the Code differentiated between two types of aggregate. Class 1 aggregates (foamed slag, pumice, blastfurnace slag, crushed brick and burnt clay products, well-burnt clinker and crushed limestone), which provide improved fire performance, and Class 2 aggregates (flint, gravel, granite and crushed natural stone other than limestone), which behave less well in fire situations. Consequently there were different provisions for thickness of walls to achieve a specified fire resistance. Although there were no tabulated values for columns, the Code recommended the use of Class 1 aggregates for fire resistance periods of two hours and above for columns with thicknesses in the range 25 3 mm. If Class 2 aggregates were used then a supplementary mesh placed centrally in the concrete cover was recommended for fire resistance periods up to 2 hours. For larger columns a 2-hour period could be obtained regardless of the aggregate used. Fire resistance periods up to 4 hours could be achieved by the use of Class 1 aggregates or a light mesh reinforcement. The tabulated values assumed a minimum cover of 25 mm for a 4-hour period, 19 mm for a 2- or 1-hour period and 13 mm for a half hour fire resistance in relation to hollow tile floors. No information was provided on the required levels of cover in relation to columns, walls or floors. The 1948 Code stated that the thicknesses used for structural reasons would normally lead to a sufficient degree of fire resistance. CP 114: 1957 The 1957 version of CP retained the provisions of the earlier version with respect to the requirements for walls and floors and provided additional tabulated data for the fire resistance of precast or in-situ inverted U sections where the minimum thickness occurred only at the crowns, hollow block construction and precast units of box or section, concrete beams and concrete columns. This current report is concerned only with the provisions in relation to commonly used structural forms. Tables 2 and 3 give minimum dimensions for columns and beams respectively. It is interesting to note that the beam provisions are expressed in terms of minimum cover rather than overall depth and are well in excess of the previously quoted values for hollow tile floors. This reflects the particular problems associated with the spalling of concrete beams. 6 Fire safety of concrete structures: Background to BS 811 fire design

8 Construction and materials Aggregates in accordance with BS 882 Minimum overall size (mm) for period of: 4 hours 2 hours 1 hour 1 /2 hour Table 2 Fire resistance of reinforced concrete columns (CP 114: 1957) Construction and materials Aggregates in accordance with BS 882 Minimum concrete cover to main reinforcement (mm) for period of: 4 hours 2 hours 1 hour 1 /2 hour Table 3 Fire resistance of reinforced concrete beams (CP 114: 1957) The Code stated that column thicknesses for 4 hours and 2 hours could be reduced to 35 mm and 229 mm where limestone aggregate is used or where mesh reinforcement is included within the concrete cover. The provisions in the Code are based substantially on an extensive series of fire tests carried out during the period 1936 to 1946 by the Building Research Station at the Fire Research Station test facility at Elstree (Borehamwood). The nature and extent of the test programme are documented in National Building Studies Research Paper No This important document also includes a summary of the significant results from the programme including the results of tests on walls and partitions, floors and roofs, columns and beams. It is this information that forms the basis for the provisions for concrete structures in fire in the UK. The information was incorporated into generic fire resistance tables for walls and partitions, floors and roofs, beams and columns. Table 4 below combines the information from these tables as they relate to reinforced concrete construction. Type of construction Fire resistance period (hours) 6 hours 4 hours 2 hours 1 hour 1 /2 hour Table 4 Minimum thicknesses (mm) to achieve indicated performance (NBS 12, 1953) Walls and partitions 1 Class 1 aggregates Class 2 aggregates Solid reinforced concrete slab Reinforced concrete columns Class 1 aggregates Class 2 aggregates Notes: 1 Walls to be reinforced vertically and horizontally at not more than 152 mm centres and reinforcement to be not less than.2% of volume. Walls less than 127 mm thick to have single layer of reinforcement in middle of wall. Walls more than 127 mm thick to have 2 layers of reinforcement, not less than 25 mm from each face. 2 Increased to 4 hours if light mesh reinforcement placed in cover 3 Increased to 2 hours if light mesh reinforcement placed in cover There is a clear recognition of the flexibility of reinforced concrete in providing fire resistance through the design and detailing process. Reference is made to the cover associated with reinforced concrete beams that would suggest the figures in Table 3 are conservative. Fire safety of concrete structures: Background to BS 811 fire design 7

9 CP 11: 1972 The next major change came with the publication of CP 11 8 in Anchor 9 has tabulated minimum section sizes and cover from CP 11 and compared the provisions to the Building Regulations requirements and alternative European specifications. Table 5 below summarises the code provisions from CP 11. Table 5 Provisions of CP 11: 1972 Type of Minimum dimensions (mm) for a fire resistance (hours) of: construction /2 1 1 /2 Reinforced concrete solid slab Thickness Cover Reinforced concrete hollow cored slabs Thickness Cover Prestressed solid slab Thickness Cover 65* 5* Prestressed hollow cored slab Thickness Cover 65* 5* Reinforced concrete simply supported beam Width 28 (25) 24 (2) 18 (16) 14 (13) 11 () 8 (8) Cover 65* (5) 55* (45) 45* (35) 35 (3) 25 (2) 15 (15) Prestressed simply supported beam Width 28 (25) 24 (2) 18 (16) 14 (13) 11 () 8 (8) Cover * (8) 85* (65) 65* (5) 5* (4) 4 (3) 25 (2) Columns (4-sided exposure) Minimum dimension 45 (3) 4 (275) 3 (225) 25 (2) 2 (15) 15 (15) Notes * Mesh reinforcement required in concrete cover. Figures in brackets refer to lightweight aggregate concrete, concrete cover is an average value. The Code indicated that for beams and slabs (both reinforced and prestressed) the influence of restraint could be incorporated into the design by reducing the requirements for cover to the next lowest category. For instance, a reinforced concrete beam simply supported would require a minimum width of 18 mm and an average cover to the reinforcement of 45 mm for a 2-hour fire resistance period. If the beam is built into a structure so as to provide restraint against thermal expansion at both ends then the requirement for cover was reduced to 35 mm, although the minimum width remains unchanged. The provisions of the Code related specifically to end restraint against thermal expansion rather than continuity over the supports. It is interesting to note that there was a distinction in CP 11 between solid and cored slabs, with cored slabs requiring minimum thicknesses in excess of that for solid slabs; this is presumably based on the insulation criteria. These separate provisions have been removed in Part 1 of the present Code and are not clearly presented in Part Clause of Part 2 mentions that an effective thickness should be used for cored slabs depending on the proportion of solid material per unit width of slab. It would be much clearer to include a separate table for cored slabs or at least to incorporate an additional note to Table 4.4 in the Code. 8 Fire safety of concrete structures: Background to BS 811 fire design

10 BS 811: 1985 and 1997 The current Code provisions in BS 811 refer to tabulated data published in Guidelines for the construction of fire resisting structural elements 1. As a document referenced in Approved Document B, the provisions in these BRE guidelines effectively become one way of achieving the requirements of the regulations in terms of minimum thickness and minimum cover. One of the major changes is the variation in requirements for normalweight and lightweight concrete and, for beams and floors, the recognition of the impact of continuity on the performance of reinforced concrete elements in fire. As there are extra provisions to allow for concrete density and continuity, no attempt is made to tabulate the provisions in a single table. The provisions relating to columns, beams, plain soffit (flat slab) floors and ribbed open soffit (downstand) floors are detailed in Tables 6 to 9. In relation to beams and floors the tables also provide information on prestressed concrete. Values from Part 1 of BS 811 are included in brackets in Tables 6 to 8 where appropriate. It should be noted that the covers in Part 1 relate to cover to all reinforcement (including links) while the values in Part 2 and the BRE guidelines are specified in terms of cover to the main reinforcement. For consistency the values from Table 3.4 of Part 1 have therefore been increased by 1 mm for beams and columns to allow for the inclusion of links to the main reinforcement. This approach is consistent with the guidance in the Code. Type of construction Minimum dimensions (mm) for a fire resistance (hours) of: /2 1 1 /2 Table 6 Minimum dimensions for reinforced concrete columns from BRE guidelines (BS 811) Fully exposed Dense concrete Width 45 (45) 4 (4) 3 (3) 25 (25) 2 (2) 15 (15) Cover 35 (35) 35 (35) 35 (35) 3 (3) 25 (3)* 2 (3)* Lightweight concrete Width Cover % exposed Dense concrete Width Cover Lightweight concrete Width Cover face exposed Dense concrete Width Cover Lightweight concrete Width Cover Notes The guidelines allow for a decrease in cover for a corresponding increase in width. * Reduced to 25 mm where the maximum aggregate size is less than or equal to 15 mm. Fire safety of concrete structures: Background to BS 811 fire design 9

11 Table 7 Minimum dimensions for concrete beams from BRE guidelines (BS 811) Type of construction Minimum dimensions (mm) for a fire resistance (hours) of: /2 1 1 /2 Reinforced simply supported Dense concrete Width 28 (28) 24 (24) 2 (2) 15 (2) 12 (2) 8 (2) Cover 8 (8) 7 (7) 5 (5) 4 (3) 3 (3)* 2 (3)* Lightweight concrete Width Cover Reinforced continuously supported Dense concrete Width 24 (28) 2 (24) 15 (2) 12 (2) 8 (2) 8 (2) Cover 7 (6) 6 (5) 5 (4) 35 (3)* 2 (3)* 2 (3)* Lightweight concrete Width Cover Prestressed simply supported Dense concrete Width Cover Lightweight concrete Width Cover Prestressed continuously supported Dense concrete Width Cover Lightweight concrete Width Cover Notes The guidelines allow for a decrease in cover for a corresponding increase in width. * Reduced to 25 mm where the maximum aggregate size is less than or equal to 15 mm. 1 Fire safety of concrete structures: Background to BS 811 fire design

12 Type of construction Minimum dimensions (mm) for a fire resistance (hours) of: /2 1 1 /2 Table 8 Minimum dimensions for plain soffit floors from BRE guidelines (BS 811) Reinforced simply supported Dense concrete Thickness 17 (17) 15 (15) 125 (125) 11 (11) 95 (95) 75 (75) Cover 55 (55) 45 (45) 35 (35) 25 (25) 2 (2) 15 (2)* Lightweight concrete Thickness Cover Reinforced continuously supported Dense concrete Thickness Cover 45 (45) 35 (35) 25 (25) 2 (2) 2 (2) 15 (2)* Lightweight concrete Thickness Cover Prestressed simply supported Dense concrete Thickness Cover Lightweight concrete Thickness Cover Prestressed continuously supported Dense concrete Thickness Cover Lightweight concrete Thickness Cover Note * Reduced to 15 mm where the maximum aggregate size is less than or equal to 15 mm. Fire safety of concrete structures: Background to BS 811 fire design 11

13 Table 9 Minimum dimensions for ribbed open soffit floors (BS 811) Type of construction Minimum dimensions (mm) for a fire resistance (hours) of: /2 1 1 /2 Reinforced simply supported Dense concrete Thickness Width (and cover) 175 (65) 15 (55) 125 (45) 11 (35) 9 (25) 75 (15) Lightweight concrete Thickness Width (and cover) 15 (55) 125 (45) (35) 85 (3) 75 (25) 6 (15) Reinforced continuously supported Dense concrete Thickness Width (and cover) 15 (55) 125 (45) (35) 9 (25) 8 (2) 75 (15) Lightweight concrete Thickness Width (and cover) 125 (45) (35) 9 (3) 8 (25) 75 (2) 7 (15) Prestressed simply supported Dense concrete Thickness Width (and cover) 175 (65) 15 (55) 125 (45) 11 (35) 75(25) 7 (2) Lightweight concrete Thickness Width (and cover) 15 (55) 125 (45) 11 (35) 9 (3) 75 (25) 7 (2) Note Width refers to the width of the downstand portion of the floor at the level of the lowest reinforcement. Table 1 Minimum dimensions for concrete walls with vertical reinforcement (BS 811) Type of construction Minimum dimensions thickness/cover (mm) for a fire resistance (hours) of: /2 1 1 /2 Walls made from dense aggregate with less than % reinforcement Walls made from dense aggregate with.4 1% 24/25 2/25 16/25 14/25 12/25 /25 reinforcement Walls made from 19/25 16/25 13/25 115/25 /2 /1 lightweight aggregate Walls made from dense aggregate with over 1% 18/25 15/25 /25 /25 75/15 75/15 reinforcement The relationship between the various documents referenced in this report and the corresponding national regulations and codes is illustrated in the flowchart Figure Fire safety of concrete structures: Background to BS 811 fire design

14 Figure 1 Relationship between documents referenced and national regulations and codes National Building Studies 12 (1953) 7 and National Building Studies 18 (1953) 11 : Investigations on building fires CP 114: CP 11: FIP/CEB Recommendations for the design of reinforced and prestressed concrete structural members for fire resistance (1975) 12 Design and detailing of concrete structures for fire resistance: Interim guidance by a Joint Committee of The Institution of Structural Engineers and The Concrete Society (1978). 14 Tables produced by the Fire Research Station Schedule 8: Building Regulations FIP/CEB Report on methods of assessment of the fire resistance of concrete structural members (1978) 13 Guidelines for the construction of fire resisting structural elements: BRE Report (1982) 1 Approved Document ADB BS (1985 and 1997) 2 BS (1985) 1 Guidelines for the construction of fire resisting structural elements: BRE Report (1988) 1 Approved Document ADB , ADB 2 18 Fire safety of concrete structures: Background to BS 811 fire design 13

15 Comparison between tabulated values from different codes The provisions according to the various standards are summarised in Table 11 with reference to examples of specific structural elements. Table 11 Comparison of provisions for common forms of construction Type of structural element Minimum dimensions thickness/cover (mm) for a fire resistance period (hours) of: /2 1 1 /2 Reinforced column CP 114: 1948 > CP 114: NBS Paper No NBS Paper No CP 11 (1972) BS (1985) 45/35 4/35 3/35 25/3 2/25 15/2 BS (1985) 36/35 32/35 24/35 2/25 16/2 15/2 EC Solid reinforced slab CP (1948) CP (1948) CP 114 (1957) 152/25 127/13 12/13 89/13 NBS Paper No CP 11 (1972) 15/25 15/25 125/2 125/2 /15 /15 BS (1985) 17/55 15/45 125/35 11/25 95/2 75/15 BS (1985) 15/45 135/35 115/25 15/2 9/15 7/15 BS (1985) 17/45 15/35 125/25 11/2 95/2 75/15 BS (1985) 15/35 135/25 115/2 15/2 9/15 7/15 EC /65 15/55 12/4 /3 8/2 6/1 Notes Class 1 aggregates Class 2 aggregates 3 Increased to 4 hours with light mesh in cover 4 Increased to 2 hours with light mesh in cover 5 Dense concrete 6 Lightweight concrete 7 Thickness and cover dependent on load ratio at ambient temperature and reinforcement ratio 8 Simply supported/dense concrete 9 Simply supported/lightweight 1 Continuous/dense concrete 11 Continuous/lightweight 12 Simply supported one way spanning The extent of the notes required for use with Table 11 provide some indication of the care to be taken in the use of tabulated data. The table itself illustrates the gradual development of knowledge related to the performance of concrete structures in fire. The close correlation between the UK codified values and the values taken from National Building Studies Research Paper No. 12 show the importance of this body of work on the design of concrete structures in fire. The results from this publication are investigated in greater detail later in this report. What is also significant is the development of specific provisions for different types of aggregate, for lightweight concrete and for the effects of continuity. 14 Fire safety of concrete structures: Background to BS 811 fire design

16 Experimental background to tabulated values Unless otherwise specified, comparison between measured data and assumed fire periods are on the basis of both minimum dimension and minimum cover. National Building Studies research papers It is widely acknowledged that the main source of data for the derivation of the tabulated values in successive revisions of the concrete codes of practice comes from fire resistance tests carried out at the Fire Research Station, Borehamwood between 1936 and The results from these tests are summarised in National Building Studies Research Paper No (NBS 12) for a range of different structural elements and different construction materials. Additional source material related to reinforced concrete columns is contained in National Building Studies Research Paper No The test results have been revisited as part of this project and analysed with respect to the concept of load ratio. This performance-based approach is not included in BS 811 for columns and walls. The concept of load ratio relates the load applied at the fire limit state to the capacity of the element at ambient temperature. This is the basis of the provisions for columns set out in the fire part of the Eurocode for concrete structures 19. Reinforced concrete floors Table A1 of Appendix A is a summary of test data in relation to reinforced concrete floors. One of the most significant aspects is the extent of spalling to the specimens and the impact of spalling on the overall fire resistance attained. The experimental values used as the basis of the Code provisions may be classified according to the support conditions adopted during the standard fire tests. Figure 2 illustrates the results from simply supported slabs in terms of fire resistance period attained against the load ratio. The load ratio has been calculated using the design formulae from BS 811: 1985 incorporating the partial safety factors but using the actual concrete material properties as measured during the experimental programme (Table A1). No information was available for the measured values of the steel yield stress so an assumption that fy = 25 N/mm 2 has been made for all cases. It could be argued that the calculation should be performed without safety factors for the concrete strength. The measured values shown in the figures are therefore worst-case values in terms of the load ratio or degree of utilisation. Figure 2 indicates the fire resistance period attained or time to failure as well as the mode of failure and the critical parameters of overall thickness and cover to the main reinforcement compared with the tabulated values in the Code. Generally the Code provisions are very conservative when compared with the experimental results. The specific combinations of thickness and cover make a direct comparison difficult. In general the mode of failure was cracking through the full depth of the slab leading to collapse into the furnace. Spalling was not particularly significant with the exception of one specimen made from limestone aggregate, which collapsed after 35 minutes. Loadbearing failure in a standard test is generally a function of the temperature of the reinforcement. The most effective way to delay such a failure is to increase the cover to the main steel. The thickness of the specimen will determine the temperature of the unexposed face (insulation criteria). If cover is used as the means of assessment then all test specimens achieved at least the requirement of the Code and in many cases the measured performance was well in excess of that assumed using the tabulated values. This is illustrated in Figure 3. Fire safety of concrete structures: Background to BS 811 fire design 15

17 Figure 2 Applied load ratio and failure time for simply supported floor slabs (NBS 12, 1953) Current Code requirements /25 Failure mode Insulation Loadbearing Did not fail during test Time (minutes) /13 Plaster finish Plaster finish 127/25 114/13 127/13 Plaster finish 114/13 114/13 114/13 114/ Load ratio 125/35 11/25 95/2 75/15 Figure 3 Comparison of measured and assumed design values based on depth of cover for simply supported floor slabs (NBS 12, 1953) Fire endurance/resistance (minutes) Plaster finish Plaster finish Plaster finish Fire endurance measured Fire resistance assumed 25 F2 F16 F21 F19 F17 F22 F18 F23 F24 Sample reference It is clear from the test results that loadbearing failure is the critical mode for all the slabs where the concrete is directly exposed to the fire. The increased cover will, in the absence of spalling, lead to increased fire resistance periods. The provisions in relation to simply supported slabs are therefore conservative due to insufficient variation in the value of applied load to determine the sensitivity of the results to the load ratio. However, as the predominant mode of failure is collapse into the furnace it could be assumed that lower values of imposed load (such as are used for the fire limit state) would increase the fire resistance. 16 Fire safety of concrete structures: Background to BS 811 fire design

18 Figure 4 is a comparison of measured values and assumed fire resistance periods from the tabulated data based purely on the minimum thickness of the floor slab. Apart from one rogue value (F2) the results support the Code provisions. Although it is useful to investigate the assumptions in terms of loadbearing capacity (minimum cover) and insulation (minimum thickness) separately they need to be taken together. For simply supported slabs the measured results, taken together, support the provisions of the current Code Figure 4 Comparison of measured and assumed design values based on minimum thickness for simply supported floor slabs Fire endurance/resistance (minutes) F2 F16 F21 F19 F17 F22 F18 F23 F24 Sample reference Fire endurance measured Fire resistance assumed For continuous slabs the Code provisions maintain similar values for minimum thickness based on the insulation criteria and allow for a reduction in the cover to the main reinforcement. Figure 5 shows the results for continuous (BS 811) or restrained (NBS 12) slabs. The different terminology is important as the reduction in cover in CP 11 was based on the influence of restraint to thermal expansion while structural continuity implies a rotational restraint such as that found over supports. One of the most significant results is the noticeable increase in spalling when compared with the results from the simply supported slabs. Fire safety of concrete structures: Background to BS 811 fire design 17

19 Figure 5 Applied load ratio and failure time for restrained floor slabs Failure mode Insulation Loadbearing Integrity Did not fail during test Results for specimens that spalled are shown thus: 152/13 and others thus: 114/13 Time (minutes) /13 Apart from one test all specimens had a cover of only 13 mm. In terms of the provisions of the Code in relation to minimum thickness and minimum cover, all the specimens tested (with the exception of the specimen attaining a fire resistance period of 6 hours) would be expected to have a fire resistance of only 3 minutes. The comparison between measured and assumed resistance is shown in Figure /38 152/13 182/13 152/13 152/13 152/13 114/13 12/13 114/13 12/13 Sprayed asbestos Load ratio 82/13 82/13 82/13 82/13 82/13 Plaster covered 12/13 Current Code requirements 17/45 15/35 125/25 11/2 95/2 75/15 Figure 6 Comparison of measured and assumed design values based on depth of cover for restrained floor slabs (NBS 12, 1953) Fire endurance measured Fire resistance assumed Fire endurance/resistance (minutes) Sprayed asbestos F53 F34 F33 F49 F48 F73 F71 F25 F77 F74 F45 F68 F72 F76 E3/S1 F67 F63 Sample reference 18 Fire safety of concrete structures: Background to BS 811 fire design

20 Only one specimen lies below the assumed design value and in many cases the measured values show the tabulated data to be extremely conservative with 4 hour fire resistance periods attained for specimens with only 13 mm cover. The load ratios used in the test are very high. From the data available the performance of the floor slabs does not seem to be significantly influenced by the value of the imposed load. Of much greater significance is the degree of spalling observed. Where the specimens failed to attain a fire resistance period of 2 hours the reason for this, and for the subsequent integrity failure, was due to spalling. Figure 6 indicates that spalling has been taken into account in the development of the tabulated data and accounts in large part for the degree of conservatism present. The tabulated values have been set according to a mode of failure driven largely by spalling. Reinforced concrete walls The information on reinforced concrete walls from the National Building Studies report is included in Table A2 of Appendix A. Although there are only three tests in this category the results indicate the improved performance of crushed brick aggregate compared with gravel. Figure 7 shows the results for reinforced concrete walls in terms of the Code provisions with respect to medium (.4 to 1.%) and high (greater than 1.%) reinforcement. Although there are obviously not many tests on which to base conclusions, it is clear that the results available show that the Code provisions are reasonable and that load ratio does not have a significant effect on performance as the failure criteria is governed by the insulation properties of the concrete. European research also supports this conclusion /25 Current Code requirements 18/25 high r/f Figure 7 Applied load ratio and failure time for reinforced concrete walls /25 low r/f Time (minutes) /25 12/25 /25 high r/f Load ratio /25 low r/f Failure mode Insulation Did not fail during test Fire safety of concrete structures: Background to BS 811 fire design 19

21 Reinforced concrete columns Although there may be few test results to assess the provisions for walls the same cannot be said for reinforced columns. A total of 95 individual tests have been analysed and the results processed according to section size and applied load. Column sizes: 152, 23 and 229 mm Figure 8 shows the results from tests on 152 mm, 23 mm and 229 mm square columns in terms of load ratio while Figure 9 is a comparison between measured and assumed fire endurance/resistance periods. The results generally support the provisions of the Code and indicate a relationship between the amount of load present and the fire endurance period attained. Figure 8 Applied load ratio and failure time for small reinforced concrete columns Time (minutes) mm cover 25 mm cover 25 mm cover 28 mm cover 4 Column size 152 mm 23 mm 229 mm Load ratio Figure 9 Measured and assumed values of fire endurance/resistance for small reinforced concrete columns Fire endurance/resistance (minutes) Fire endurance measured Fire resistance assumed mm 152 mm 23 mm 229 mm Column size The results above show the Code provisions to be generally conservative and indicate a relationship between applied load and fire endurance period attained. 2 Fire safety of concrete structures: Background to BS 811 fire design

22 Column size 254 mm Figure 1 shows the results for 254 mm square columns while Figure 11 shows the relationship between the measured values and the assumed fire resistance period taken from the tabulated data in the Code bars, cover = 25.4 mm Figure 1 Applied load ratio and failure time for 254 mm square columns 175 Time (minutes) Load ratio Figure 11 Comparison of failure time with assumed fire resistance period 2 Fire endurance/resistance (minutes) A9 A1 C23 C24 C3 C31 C32 C35 C11 C13 C15 C23 C24 C3 C31 C32 C35 Sample reference Fire endurance measured Fire resistance assumed Again the results indicate that the Code provides a generally conservative approach. Figure 1 also shows some correlation between load ratio and decreasing fire performance of columns. Although the Code provides a conservative approach it does not take into account the complex interaction between applied load and fire resistance. Fire safety of concrete structures: Background to BS 811 fire design 21

23 Column size 28 mm Figure 12 shows the relationship between applied load and failure time for 28 mm square columns. Here the relationship between applied load and time to failure is not so straightforward. The comparison between fire endurance period achieved in the tests and fire resistance assumed in design is shown in Figure 13 below. Figure 12 Applied load ratio and failure time for 28 mm square columns bars, cover = 38.1 mm 175 Time (minutes) Load ratio Figure 13 Comparison between failure time and fire resistance period assumed (28 mm square columns) Fire endurance/resistance (minutes) Fire endurance measured Fire resistance assumed 25 C33 C34 C82 C86 C87 C88 C89 C9 E25/S3 C2 C36 C37 C38 C42 Sample reference There are some instances where the assumed period of fire resistance is less than that achieved during the test for this particular size of column. However, the only areas where there is a significant shortfall in performance relates to a specimen where the load ratio is in excess of.5. This is a higher ratio than is generally seen in practice for the fire limit state. 22 Fire safety of concrete structures: Background to BS 811 fire design

24 Column size 35 mm The corresponding values for 35 mm square columns are shown in Figures 14 and 15 below. Again this shows the relationship between applied load and fire endurance period and indicates that the test data generally supports the tabulated values from the Code. There is an important relationship between minimum dimensions for width and cover. If cover is seen to be the governing factor in terms of slowing the increase in the temperature rise of the reinforcement then the results should be compared against a presumed performance of 1 hour. However, if the overall size of the column is the main determinant then the comparison should more effectively be made against a fire resistance period of 2 hours bars, cover = 25.4 mm Figure 14 Applied load ratio and failure time for 35 mm square columns Time (minutes) bars, cover = 25.4mm bars, cover = 22 mm 4 bars, cover = 25.4 mm Load ratio Figure 15 Comparison between failure time and fire resistance period assumed (35 mm square columns) Fire endurance/resitance (minutes) Fire endurance measured 8 A7 A8 1 Sample reference Fire resistance assumed Fire safety of concrete structures: Background to BS 811 fire design 23

25 Column size 355 mm The plot for the 355 mm square columns is shown in Figure 16. This indicates a significant reduction in fire performance with increasing values of load ratio. However, the applied load is much higher than typically applied in practice and, in some cases, could have led to a structural collapse even before the start of the fire test. It would only take a small eccentricity of loading to exceed the ambient temperature capacity at a load ratio of.9. Figure 16 Applied load ratio and failure time for 356 mm square columns Time (minutes) mm cover 38 mm cover 35 mm cover 25 mm cover Figure 17 Comparison between failure time and fire resistance period assumed (356 mm square columns) Load ratio Figure 17 shows the comparison between assumed fire resistance and measured failure times. Although these results do not support the Code provisions they need to be viewed in the light of the comments related to load ratio above Fire endurance/resistance (minutes) Fire endurance measured 1 Fire resistance assumed E16/S8 Sample reference From observations during the testing, the failures were attributed to spalling of the corners of the columns. This behaviour is accounted for in the Code through the requirement for additional measures (such as the use of supplementary mesh within the cover zone) once the cover exceeds 4 mm. As the cover increases the fire performance of the concrete member will increase, assuming no spalling takes place. However, increase in cover increases the susceptibility of the member to spalling. There is therefore a need to either limit the cover (the tabulated values for columns do not exceed 35 mm) or take additional measures. The incidence of spalling means that there is no direct correlation between increasing cover and increasing periods of fire resistance. The provisions of the Code take this into account. 24 Fire safety of concrete structures: Background to BS 811 fire design

26 Column size 381 mm The results for the 381 mm square columns are shown in Figure 18. There is a clear correlation between the load ratio and the performance in fire. Here the relationship is almost linear. The values above a load ratio of about.7 need to be viewed in the light of the comments made above in relation to practical load ratios in operation at the fire limit state. Time (minutes) bars, cover = 28.6 mm Load ratio Figure 18 Applied load ratio and failure time for 381 mm square columns The comparison with assumed fire resistance period is shown in Figure 19. Again this needs to be viewed in the light of the comments about the importance of applied load and also that the comparison has been made for a cover of 25 mm. The actual column size is much greater than the minimum width for the 1 hour category (width 2 mm, cover 25 mm) and it may be that a 1 1 /2 hour fire resistance period (width 25 mm, cover 3 mm) would be a more accurate basis for comparison. Fire endurance/resistance (minutes) A1 A2 A3 A4 A5 A6 Sample reference Figure 19 Comparison between failure time and fire resistance period assumed (381 mm square columns) Fire endurance measured Fire resistance assumed However on the basis of a direct comparison with the provisions of the Code, the experimental results support the tabulated values. The only member that failed to achieve the prescribed period of fire resistance had a load ratio in excess of.9, which is considered to be unrealistic for both the fire and ambient temperature limit states. Fire safety of concrete structures: Background to BS 811 fire design 25

27 Column size 46 mm Figure 2 is the plot for load ratio against failure time for the three tests on 46 mm square columns. What is again quite apparent is the relationship between applied load and fire endurance period. Effectively the failure period of 46 minutes can be discounted from the comparison because of the excessive load imposed at the time of the test. Again the comparison with assumed fire resistance periods based on the tabular data will be made for the 1 hour period because the cover is still only 25 mm. The fire resistance period attained in the tests for the two columns is therefore twice the assumed value. Figure 2 Applied load and failure time for 46 mm square columns bars, cover = 25.4 mm 8 Time (minutes) Load ratio 26 Fire safety of concrete structures: Background to BS 811 fire design

28 Column size 483 mm The results from the tests on 483 mm square columns are shown in Figure 21 and the comparison with the assumed values from the tabulated data is shown in Figure 22. Time (minutes) Cover = 25 mm Cover = 32 mm Cover = 25 mm Cover = 32 mm Cover = 32 mm Cover = 25 mm Load ratio Figure 21 Applied load ratio and failure time for 483 mm square columns Fire endurance/resistance (minutes) A11 A12 15 A13 A14 Sample reference The results from tests on loaded reinforced concrete columns are contained in Table A3 of Appendix A and in Appendix B. The comprehensive test programme includes a large range of sizes, aggregate types and levels of reinforcement. Again the impact of spalling on the performance of reinforced concrete members is highlighted. In many cases the spalling did not lead to collapse of the columns and the commentary often points to corner spalling or sloughing off rather than explosive spalling. Spalling generally occurred at the corners of the columns, In many cases the corners were chamfered, but this did not appear to make a great deal of difference to the overall fire resistance attained nor did it prevent significant spalling taking place. Moisture content does not appear to have had a significant influence. The measured values of moisture content for this size of columns varied from 2.3% to 3.48% by weight and there is no variation in the extent or nature of the spalling observed that could be related to the difference in measured moisture content. Figure 22 Comparison between failure time and fire resistance period assumed (483 mm square columns) Fire endurance measured Fire resistance assumed Fire safety of concrete structures: Background to BS 811 fire design 27

29 Reinforced concrete beams Table A4 of Appendix A contains the results from tests on reinforced concrete beams taken from the National Building Studies Research report. From these particular tests moisture content appears to have had no specific influence on whether the specimen spalled or not. Spalling occurred on one of the specimens with a measured moisture content of 4% by weight whereas no spalling was observed for the duration of the test for a similar section with a measured moisture content of 4.7%. However, it should be noted that this contradicts other published research which shows that moisture content has a critical influence on susceptibly to spalling. The results also indicate that spalling does not necessarily lead to a reduction in the fire resistance period attained. Figure 23 shows the results for reinforced concrete beams in terms of applied load and time to failure. The results are plotted using the BS 811 design equations both with and without the safety factors for material variability. The effect of ignoring the material safety factors is to reduce the load ratio. However, the performance seems to be largely independent of the applied load. The Code requirements in terms of minimum dimensions for width of beams and cover are indicated on the right of the figure. Two things are immediately apparent from the figure. Firstly, the load ratio is very high and, secondly, the Code provisions are extremely conservative. In the case of beams it is loadbearing that is the critical factor, not integrity. Therefore the critical parameter is the minimum cover to the reinforcement. Figure 23 Applied load ratio and failure time for reinforced concrete beams /38 165/38 165/38 Current Code requirements 28/8 Applied load 25 N/mm 2, includes BS 811 safety factors 25 N/mm 2, excludes BS 811 safety factors 272 N/mm 2, excludes BS 811 safety factors Time (minutes) /25 23/25 114/13 24/7 2/5 15/4 Failure mode 5 12/3 Loadbearing 25 8/2 Did not fail during test Load ratio 28 Fire safety of concrete structures: Background to BS 811 fire design

30 Figure 24 is a comparison between the assumed period of fire resistance and the test results Figure 24 Comparison between failure times and assumed period of fire resistance (reinforced concrete beams) 2 Fire endurance/resistance (minutes) (F42) 2 (F43) 3 (F44) 4 (F51) 5 (F52) 6 (E16/S9) Sample reference Fire endurance measured Fire resistance assumed Fire safety of concrete structures: Background to BS 811 fire design 29

31 Other relevant research Although the National Building Studies reports provided a large part of the background data on which the Code provisions are based, there is insufficient data in the reports to extend the provisions to lightweight concrete, prestressed concrete and ribbed open soffit floors. The additional provisions for these have arisen as a result of a number of research programmes and commercial tests undertaken over a number of years. Some of the most significant research projects are briefly reviewed here, although it is impossible to determine the extent to which each project influenced the development of the Code provisions. A report (in two volumes) published by BRE 2 details the results from sponsored fire tests undertaken between 195 and A summary of the test results is given in Appendix C, covering floors, columns and beams. What should be immediately clear is the difference between published research results and the results from sponsored commercial tests. Commercial tests were designed to provide a justification for the stated period of fire resistance. Hence little information is provided on the types of aggregate used, moisture contents, reinforcement details, restraint conditions or whether spalling occurred. In general the mode of failure (stability, integrity and insulation) is not made clear. Where the information provides only the fire resistance period attained it has been assumed that this was the target value and the test was stopped at this point. However, the results do provide some justification for the stated fire resistance periods quoted by manufacturers and the tabulated provisions relating to lightweight concrete and hollow core slabs. The results for reinforced columns come largely from work commissioned through the Department of the Environment and therefore more details of these research-based tests are available for scrutiny. Fire Research Notes Fire Research Note No. 46, Malhotra, H L. Effect of restraint on fire resistance of concrete floors 21 Malhotra reported the results from tests to determine the effects of restraint on the fire resistance of concrete floors and compared the results with similar tests carried out in Holland. The tests were carried out on precast hollow beams with a concrete topping. Three different levels of restraint were investigated, ranging from simply supported to longitudinal and angular restraint. Some limited spalling occurred during the third test. The first two tests provided a direct comparison of the effects of longitudinal restraint on the fire resistance of reinforced concrete floors. The inclusion of longitudinal restraint resulted in an increase of fire resistance of approximately 5% based on a displacement criteria of span/3. Malhotra pointed out that damage by spalling may, however, be greater with increasing restraint. Fire Research Note No. 38, The fire resistance of prestressed concrete 22 This Note claims that spalling of concrete leading to premature failure only occurs with small prestressed beams of slender section directly exposed to the fire. It is estimated that a fire resistance of 2 hours can be obtained with a concrete cover to the steel of approximately 63 mm. This is consistent with the Eurocode values and the BS 811 values for prestressed beams. Fire Research Note 54, Ashton, L A, Prestressed concrete during and after fires. Comparative tests on composite floors in prestressed and reinforced concrete 23 This study showed very little difference in performance between reinforced and prestressed construction for short periods of exposure ( 1 /2 hour). The maximum deflections were of the same order, with the prestressed floor recovering more quickly indicating no loss of prestress. However for 1 hour exposure the reinforced concrete floor exhibited 3 Fire safety of concrete structures: Background to BS 811 fire design

32 lower deflections and a better rate of recovery. Both types of floor had similar cover to the steel, based on the requirements of the London County Council Byelaws for 1 hours fire resistance. Fire Research Note 65, Ashton, L A and Malhotra, H L, The fire resistance of prestressed concrete beams 24 An experimental study was undertaken on scaled-down specimens representative of larger post-tensioned beams used in buildings of high fire risk. The results indicate that full-scale beams of the type tested should give a fire resistance of 2 hours without recourse to special measures, but for 4 hours resistance, extra protection would be necessary for the tendons. This study was initiated due to a sudden brittle failure in a standard fire test of a small prestressed concrete floor unit due to spalling. Fire Research Note 741, Malhotra, H L, Fire resistance of structural concrete beams 25 Malhotra pointed out the scarcity of data in relation to the performance of reinforced concrete beams when compared with prestressed beams (see above). The programme involved tests on 24 beams including 13 simply supported beams with a clear span of 7.3 m and an overall length of 7.6 m. Additional specimens were cast with an overall length of 11.3 m supported over the same span as the smaller beams and with loaded cantilever ends to produce negative bending moments at the supports to simulate continuous construction. The overall test programme consisted of 24 tests including prestressed and encased steel beams as well as reinforced concrete. The factors studied were: Type of beam reinforced, prestressed, steel encased Type of aggregate normal (gravel), lightweight (expanded clay and foam slag) Type of steel mild steel, cold worked steel, hot rolled alloy steel Thickness of cover varied between 25 and 63 mm Use of supplementary reinforcement End conditions simply supported, simply supported with continuity Leaving aside the encased steel beams, Tables 12 and 13 describe the main test parameters for prestressed and reinforced beams respectively. No. Type of Type of beam Shape of Cover Supplementary aggregate x-section (mm) reinforcement (Y/N) 7.6 m long specimens 1 Gravel Post-tensioned with tendons Rectangular Y 2 Gravel Pre-tensioned with tendons Rectangular Y 3 Gravel Pre-tensioned with strands I section 5 N 4 Gravel Pre-tensioned with strands I section 5 Y 5 Gravel Pre-tensioned with strands I section 5 N* 11.3 m long specimens 6 Gravel Post-tensioned with strands Rectangular Y 7 Gravel Post-tensioned with strands Rectangular Y Table 12 Test parameters for prestressed concrete beams Note * Encasement of 13 mm plaster Fire safety of concrete structures: Background to BS 811 fire design 31

33 Table 13 Test parameters for reinforced concrete beams No. Type of Type of beam Cover Supplementary aggregate (mm) reinforcement (Y/N) 7.6 m long specimens 8 Gravel Mild steel 63 N 9 Gravel Cold worked deformed 63 N 1 Gravel Cold worked twisted 63 N 11 Gravel Hot rolled alloy 63 N 12 Expanded clay Mild steel 63 N 13 Foamed slag Mild steel 63 N 14 Gravel Mild steel 63 Y 15 Gravel Hot rolled alloy 63 Y 16 Gravel Cold worked twisted 63 Y 17 Gravel Hot rolled alloy 38 Y 18 Gravel Hot rolled alloy 38 N 19 Gravel Hot rolled alloy 25 N 11.3 m long specimens 2 Gravel Mild steel 63 N 21 Gravel Cold worked deformed 63 N The results from these tests are summarised in Appendix D. The specimens were stored for periods up to three years to ensure a stable moisture content. The tests were generally terminated prior to collapse by carefully monitoring deflections. However, in three cases (3, 6 and 7) collapse occurred before the test could be terminated. For the reinforced concrete beams made with gravel aggregate the occurrence of spalling greatly reduced the fire resistance achieved. Additional specimens were therefore made which incorporated supplementary reinforcement in cases where cover thicknesses were large. The tests confirmed the improved performance in fire of beams made using lightweight aggregates. The specimens made from lightweight concrete withstood heating for 6 hours without showing any signs of spalling or reduction in concrete cover, whereas the siliceous aggregates showed signs of damage by spalling within the first 3 minutes of the test, with the extent of the damage varying from specimen to specimen. The report was quite specific about the effect of spalling on fire resistance of concretes made from siliceous aggregates, stating that premature failure was due to spalling. The inclusion of a supplementary mesh 25 mm below the exposed surface gave at least the expected performance and in some cases better than expected. The report emphasised the need to adopt mitigating measures when dealing with concretes made with siliceous aggregates. The report recommended the use of supplementary reinforcement for covers in excess of 4 mm. This is consistent with the provisions of CP 11: 1972 and included in BS 811: (clause 4.3.4). The current requirements of 4 mm for dense concrete and 5 mm for lightweight concrete are conservative in relation to lightweight aggregate where it is suggested supplementary reinforcement is not required for covers up to 63 mm. The time taken for the reinforcement to reach a temperature of 55ºC was 36 minutes for the lightweight aggregate and 26 minutes for the gravel aggregate. The thickness of concrete cover to limit the rise in temperature for a given size of beam is inversely proportional to the square root of thermal diffusivity. It is concluded that, for lightweight aggregates a reduction in cover of about 2% is possible for similar performance in fire. This is consistent with the provisions in BS Fire safety of concrete structures: Background to BS 811 fire design

34 Discussion What is of primary interest is how the experimental evidence discussed above relates to the provisions of the national standard. There is a direct link between the groundbreaking studies reported in the National Building Studies report and the tabulated values in BS 811. One useful means of looking at the various inter-relationships between the provisions of the Codes and the available research results is to start with a standard solution from the 1948 tabulated values and see how this changes with time. If we take the simple examples of a solid reinforced concrete slab and a reinforced concrete column subjected to a four-sided exposure then in some instances very little has changed over the years. Where no restrictions on the type of aggregate apply and the effect of continuity is not allowed for the situation is summarised in Table 14 in terms of specified minimum dimensions and (where appropriate) minimum cover for a fire resistance period of 2 hours. Source document Solid reinforced concrete Reinforced concrete column slab minimum (4 sided exposure) minimum dimension/cover (mm) dimension/cover (mm) CP 114: CP 114: /13 35 National Building Studies CP 11: /2 3 BRE guidelines 125/35 3/35 (BS 811: 1985/1997) EN /4* 3/45** Table 14 Code provisions Notes * Axis distance rather than cover to main steel ** Axis distance rather than cover, based on a load ratio of.5 Over the years the codes have been extended to provide more information on the effects of continuity, the inclusion of prestressed concrete, the use of lightweight concrete, the choice of aggregate and the depth of cover. From the data analysis undertaken in this project a number of general conclusions can be drawn: For reinforced concrete walls, the test data supports the provisions of the Code. On the basis of the limited data available, load ratio does not have a significant impact on the performance in fire of reinforced concrete walls. For reinforced concrete columns, the test data has highlighted the important influence of applied load on their performance in fire. This is discussed below in relation to Figure 25. For reinforced concrete beams, the test data indicates that the provisions for their fire resistance are extremely conservative. In some cases measured performance is more than three times greater than the assumed design value based on the tabulated approach. For restrained (continuous) floor slabs the test data supports the provisions of the Code in relation to their fire performance For simply supported reinforced concrete floor slabs the test data supports the provisions of the Code in relation to their fire performance. For all floor slabs, load ratio does not have a significant influence on the slabs performance in fire. The provisions of the Code in terms of restrained floor slabs take spalling into account through a reduction in the minimum value for cover to the main reinforcement. However, there is some confusion over the use of the terms restrained and continuous. There is evidence from standard tests to support the CP 11 approach based on restraint to thermal expansion (lateral restraint). However, the limited experimental evidence available 19 does not support the view that the same benefits can be achieved through structural continuity (rotational restraint). Fire safety of concrete structures: Background to BS 811 fire design 33

35 Spalling severely limits the fire resistance periods for reinforced concrete restrained floor slabs. Where spalling can be prevented, either through protection to the soffit or a suitable choice of dimensions, fire resistance periods of up to 6 hours can be achieved for flat slabs. The prescribed values include the effect of spalling in that they have been based on the results of specimens where significant spalling took place. The provisions for spalling account for a large part of the conservatism inherent in the Code provisions. Figure 25 Test data for reinforced concrete columns Section size and cover 152/25 254/25 279/25 35/25 356/25 381/29 46/25 483/25 Time (minutes) Load ratio Current Code requirements 45/35 4/35 3/35 25/3 2/25 15/2 Figure 25 shows the relationship between load ratio and fire endurance period for a range of different section sizes and a small variation in cover. It is important to consider the results in light of typical values of imposed load on columns in buildings and the reduced partial factors to be adopted for design for the fire limit state. There are no discrepancies between the provisions of the Code and the test results at load ratios below approximately.5. For column sizes of 254, 279, 35 and 46 mm only one specimen failed before achieving the required fire resistance period and all these specimens were loaded to values higher than would be typically in place in a building during a fire. At first sight the performance of the 381 mm column (29 mm cover) looks of particular concern with six specimens out of a total of 13 failing to achieve the prescribed level of fire resistance. However, the values of load ratio for these specimens range from.77 to.96 (i.e. they fall within the shaded area). It is highly unlikely that this level of applied load would be imposed for columns in buildings. There is therefore no concern over the performance of existing structures based on the test data used to develop the provisions of the Code. However, this project has highlighted the need for levels of applied load to be incorporated into the design procedures for reinforced concrete columns subject to fire. Such an approach is already included in the fire part of EC2 and will form the basis of fire design procedures in the years to come. The simplicity of the BS 811 tabular data, in particular figure 3.2 of BS 811 Part 1 2, is of great benefit to designers. However, an attempt to cover provisions for such a wide range of concrete members in a single table may lead to unnecessary conservatism. For hollowcore slabs the results from standard fire tests 2 suggest that dimensions 34 Fire safety of concrete structures: Background to BS 811 fire design

36 significantly below the tabulated values would provide adequate levels of fire resistance. As the fire resistance of floor slabs with adequate cover is related to insulation values one might expect the same to be true of solid slabs. If the units are sized according to the ambient temperature loading requirements and simply checked against the provisions of the Code this is a reasonable approach. However, if they are sized initially on the requirements for fire resistance then they are likely to be conservative and the potential exists for a reduction in the dimensions to fulfil normal temperature structural requirements. This applies not only to prestressed hollowcore units but also to reinforced units cast from lightweight aggregate. Solid floor units tests 2 have shown that 2 hours fire resistance can be obtained for unprotected floor slabs made using lightweight aggregate with an overall thickness of 12 mm. This could not be achieved from a simple reliance on tabular data, which would require a minimum dimension of 115 mm to achieve the 2 hour requirement. This would suggest that the tabular data is generally conservative. The work by Malhotra 21 suggests that a reduction in overall thickness could be justified, based on the effects of structural continuity and restraint to thermal expansion. Again this is not allowed for in the Code, which does take into account the increased likelihood of spalling through a reduction in the minimum value for cover. For downstand beams a fire resistance period of 2 hours has been achieved with a slab thickness of just 89 mm using lightweight aggregate. For floors the critical criteria in terms of minimum dimensions relates to the insulation properties of the material. In general this will be the critical factor in the choice of the minimum dimension for floor slabs provided spalling does not take place. A number of obstacles remain to the development of a more rational approach to the fire engineering design of concrete structures. There is a wealth of information on the performance of concrete elements subject to standard fire tests. However, little research information is available on the performance of concrete structures subject to realistic (natural) fire exposures. The prescriptive approach adopted in BS 811 has proved very effective over a number of years as indicated by the performance of real buildings in real fires. However, the design methods available in the structural Eurocodes will enable a performance-based approach to be taken to the design of concrete structures in fire, leading to a more efficient construction. Fire safety of concrete structures: Background to BS 811 fire design 35

37 Conclusions and recommendations The current Code provisions are broadly based on standard fire tests carried out by the Joint Fire Research Organisation (later FRS) at Borehamwood between 1936 and The results from research programmes have been augmented by the results from tests sponsored by the concrete industry on specific products. When any departure is required from the tabulated data assessments have to be made from data that has not been updated or from data obtained from international sources. Whilst in many cases a reliable assessment can be made, there are a number of areas where the lack of knowledge may be restricting the economic use of concrete. Calcareous aggregates such as crushed brick, which give the best performance in the National Building Studies research, are rarely used nowadays. The majority of concrete is made from siliceous aggregates such as river gravel or granite. It is therefore appropriate for the tabulated data to relate to siliceous aggregates as the standard case. There is evidence that the use of lightweight aggregates gives a much improved performance in terms of insulation and the incidence of spalling. This is adequately reflected in the current Code. Evidence of the effect of limestone aggregates compared with siliceous aggregates is not very clear. Although the Code states that concretes made from limestone aggregates are less susceptible to spalling, this is not borne out from the results of the National Building Studies report. Unlike previous versions, the current Code values do not allow for any reduction in requirements based on the use of limestone aggregates. The BS 811 approach is therefore valid in this respect. There is clear evidence from performance in real fires over a number of years that the tabular approach has proved effective. This study has highlighted a number of design issues that need to be looked at in a little more detail. In general the provisions of the Code in terms of minimum cover to the reinforcement are based on limiting the temperature rise, and therefore the reduction in strength, to values of 55 C and 45 C for reinforcing bars and prestressing tendons respectively for the prescribed period of fire resistance. The assumption is that the elements are supporting the full design load at the fire limit state. This is a conservative assumption and may lead to the inefficient use of concrete in buildings. The relationship between load ratio and fire resistance should be explored further. This concept is already used in European standards for columns and walls through the introduction of a factor (µ fi ) to take account of the load level in the fire situation. The results from the National Building Studies research reports highlight the important influence of applied load on the fire endurance of columns. This study of the background information underpinning the provisions of BS 811 has shown that the experimental results support the tabulated data in the Code in relation to assumed periods of fire resistance. In many cases the provisions are very conservative, largely due to the requirements related to the prevention of spalling and to the assumption that concrete elements are fully stressed at the time of the fire. The provisions of the Code, although drawing extensively on the experimental results discussed in this report, are also informed by engineering judgement based largely on observed performance of concrete structures in real fires. The beneficial aspects of structural continuity provide an enhanced level of safety above that derived from the results of standard fire tests. The importance of adequate detailing, particularly at connections between structural members, has been identified as a crucial aspect of the fire engineering design of concrete structures. The development of the structural Eurocodes has provided an opportunity for UK designers to adopt a performance-based approach to designing concrete structures for the effects of real fires, taking into account the beneficial aspects of whole building behaviour and the inherent continuity and robustness of properly detailed concrete buildings. 36 Fire safety of concrete structures: Background to BS 811 fire design

38 References 1. BS 811: Part 2: Structural use of concrete, Part 2, Code of practice for special circumstances. British Standards Institution, London, BS 811: Part 1: Structural use of concrete. Part 1, Code of practice for design and construction. British Standards Institution, London, Tovey, A K and Crook, R N. Experience of fires in concrete structures, in ACI Symposium on evaluation and repair of fire damage to concrete. San Francisco, March, American Concrete Institute, Detroit, Ref SP Hopkinson, J S. Concrete research and fire. Building Research Establishment Note N14/81 BRE, Garston, CP 114. The structural use of reinforced concrete in buildings. British Standards Institution, CP 114. The structural use of concrete in buildings. British Standards Institution, Davey, N and Ashton, L A. Investigations on building fires, Part V. Fire tests on structural elements, National Building Studies Research Paper No. 12, Department of Scientific and Industrial Research (Building Research Station), HMSO, London, CP 11. Code of practice for the structural use of concrete, British Standards Institution, Anchor, R D. The design of concrete structures for resistance to fire. Building Research Establishment Note N7/77. BRE, Garston Morris, W A, Read, R E H and Cooke, G M E. Guidelines for the construction of fireresisting structural elements. Building Research Establishment Report, BR128 BRE, Garston, 1988 (revised). 11. Thomas, F G and Webster, C T. National Building Studies Research Paper No. 18, Investigations on building fires, Part VI. The fire resistance of reinforced concrete columns. Department of Scientific and Industrial Research (Building Research Station), HMSO, FIP/CEB. Recommendations for the design of reinforced and prestressed concrete structural members for fire resistance. FIP Commission on Fire Resistance. Slough, Cement and Concrete Association for FIP, 1975, FIP Guide to good practice. 19 pp. 13. FIP / CEB. Report on methods of assessment of the fire resistance of concrete structural members. FIP Commission on Fire Resistance of Prestressed Concrete Structures. Slough, Cement and Concrete Association for FIP, 1978, 91 pp. 14. Design and detailing of concrete structures for fire resistance: interim guidance by a joint committee of the Institution of Structural Engineers and The Concrete Society. London, The Institution, 1978, 59 pp. 15. Building Regulations, Schedule Building Regulations, Approved Document B, Fire Safety. 17. Building Regulations, Approved Document B, Fire Safety. 18. Building Regulations, 2. Approved Document B, Fire Safety. 19. pren Eurocode 2: Design of concrete structures Part 1.2: General rules Structural fire design, CEN, Brussels, July 23. Fire safety of concrete structures: Background to BS 811 fire design 37

39 2. Fisher, R W and Smart, P M T. Results of fire resistance tests on elements of building construction, Volumes 1 and 2. Department of the Environment and Fire Office s Committee, Joint Fire Research Organisation, Fire Research Station, Building Research Establishment, HMSO, 1975 and Malhotra, H L. Effect of restraint on fire resistance of concrete floors. Fire Research Note 46, Department of Scientific and Industrial Research and Fire Officer s Committee, Joint Fire Research Organisation, Borehamwood The fire resistance of prestressed concrete, Fire Research Note 38, Department of Scientific and Industrial Research and Fire Officer s Committee, Joint Fire Research Organisation, Borehamwood Ashton, L A. Prestressed concrete during and after fires. Comparative tests on composite floors in prestressed and reinforced concrete, Fire Research Note 54, Department of Scientific and Industrial Research and Fire Officer s Committee, Joint Fire Research Organisation, Borehamwood Ashton, L A and Malhotra, H L. The fire resistance of prestressed concrete beams. Fire Research Note 65, Department of Scientific and Industrial Research and Fire Officer s Committee, Joint Fire Research Organisation, Borehamwood Malhotra, H L. Fire resistance of structural concrete beams, Fire Research Note 741, Ministry of Technology and Fire Offices Committee, Joint Fire Research Organisation, Borehamwood Fire safety of concrete structures: Background to BS 811 fire design

40 Appendix A Results from National Building Studies Research Paper No. 12 Table A1 Test results for reinforced concrete floors - from investigations on building fires No. Fire resistance Failure Moisture Thickness Aggregate Cover Spalling Comments Support Span Test load Ref. period attained criteria content (%) (mm) (mm) (Y/N) (m) (kn/m 2 ) (minutes) n/a Crushed brick 38 N Restraint F n/a Limestone 13 Y Explosive spalling Restraint F Insulation n/k 152 Limestone 13 Y Explosive spalling Restraint F Insulation n/k 152 Limestone 13 Y Explosive spalling Restraint F n/a River gravel 13 Y Restraint F Insulation River gravel 25 N Plaster finish Simple F Insulation River gravel 13 N Plaster finish Simple 3.66 F n/a River gravel 13 N Plaster finish Simple F Stability River gravel 25 N Simple F Stability River gravel 13 N Simple F Stability Limestone 13 Y Explosive spalling Simple F Stability Crushed brick 13 Y Simple F Stability River gravel 13 N Simple F Stability River gravel 13 Y Simple F Integrity River gravel 13 Y Restraint F Stability River gravel 13 N Simple F Integrity n/k 12 River gravel 13 Y Explosive spalling Restraint F n/a n/k 114 Limestone 13 Y Restraint F Insulation Limestone 13 Y Explosive spalling Restraint F Insulation Limestone 13 Y Explosive spalling Restraint F n/a River gravel 13 Y Restraint F n/a River gravel 13 N Asbestos covering Restraint E3/S1 23 Integrity River gravel 13 Y Plaster finish Restraint F Integrity Crushed whinstone 13 Y Explosive spalling Restraint F Insulation Torphin whinstone 13 Y Restraint F n/a Crushed brick 13 N Restraint F Integrity River gravel 13 Y Explosive spalling Restraint F Insulation /63 River gravel 25 N Downstand beams Restraint F54 Table A2 Test results for reinforced concrete walls - from investigations on building fires Fire resistance Failure Moisture Thickness Aggregate Cover Spalling Comments Support Failure Test load Ref. Size (m) period attained criteria content (mm) (mm) (Y/N) time (kn) (minutes) 1 (%) (minutes) 12 Insulation Gravel 25 Y Explosive spalling Not known W x Insulation Gravel 48 Y Surface spalling Restraint - W x n/a Crushed brick 25 N No spalling Not known W x 3.5 Note 1 Rounded down to nearest full period Fire safety of concrete structures: Background to BS 811 fire design 39

41 Table A3 Test results for reinforced concrete columns - from investigations on building fires No. Fire resistance period Failure Moisture Size Aggregate Cover Spalling Comments Loaded attained (minutes) 1 criteria content (%) (mm) (mm) (Y/N) (Y/N) 1 6 Not known x 152 River gravel 25 N Cracking Y 2 6 Stability x 152 River gravel 25 N Y 3 12 Stability x 254 River gravel 25 N Y 4 12 Not known x 254 Torphin whinstone 25 N Y 5 12 Stability x 254 Hillhouse whinstone 25 N Y 6 9 Stability x 254 River gravel 29 Y Steel exposed Y 7 9 Stability x 254 River gravel 25 Y Steel exposed Y 8 9 Stability x 254 River gravel 25 N Cracking Y 9 12 Stability x 254 River gravel 25 N Cracking Y 1 12 n/a x 254 River gravel 25 Y Y n/a x 254 River gravel 25 Y Y 12 9 Stability x 254 River gravel 25 N Y 13 6 Stability x 254 River gravel 32 Y Y 14 3 Stability x 279 River gravel 38 Y Y 15 9 Stability x 279 River gravel 38 Y Y 16 6 Stability x 279 River gravel 38 Y Y n/a x 279 Hillhouse whinstone 38 Y Y n/a x 279 Torphin whinstone 38 Y Y 19 6 Stability x 279 River gravel 38 Y Restrained Y 2 12 n/a x 279 River gravel 38 N Plastered Y n/a x 279 River gravel 38 Y Y n/a n/k 279 x 279 Cheddar limestone 38 N Y Stability x 279 Matlock limestone 38 Y Y Stability x 279 Clitheroe limestone 38 N Y n/a x 279 Blast furnace slag 38 N Y Stability n/k 279 x 279 River gravel 38 Y Steel mesh Y 27 6 Stability x 279 River gravel 38 Y Y 28 6 Stability x 279 River gravel 38 Y Y 29 3 Stability x 35 River gravel 51 N Cover removed Y 3 6 Not known x 356 Gravel 38 Y Y n/a x 46 River gravel 25 Y Y n/a x 46 River gravel 35 Y Y 33 6 n/a x 46 River gravel 35 Y Y n/a x 58 River gravel 44 Y Y n/a x 58 River gravel 44 Y Y Stability x 514 River gravel 29 Y Y 37 6 n/a x 35 River gravel 13 N Hexagonal Y 38 3 Stability x 35 River gravel 13 Y Hexagonal Y n/a x 356 River gravel 13 Y Hexagonal Y 4 12 n/a x 46 River gravel 13 Y Hexagonal Y 41 9 Stability x 46 River gravel 13 N Hexagonal Y n/a x 58 River gravel 13 Y Hexagonal Y Stability x 58 River gravel 13 Y Hexagonal Y Note 1 Rounded down to nearest full period 4 Fire safety of concrete structures: Background to BS 811 fire design

42 Table A4 Test results for reinforced concrete beams - from investigations on building fires No. Fire resistance period Failure Moisture Size Aggregate Cover Spalling Comments Loaded attained (minutes) 1 criteria content (%) (mm) (mm) (Y/N) (Y/N) 1 18 Stability /14 River gravel 51 N Downstand Restraint 2 18 Stability /14 River gravel 38 N Downstand Restraint 3 12 Stability 4. 12/14 River gravel 25 Y Downstand Restraint 4 9 Stability /14 River gravel 13 N Downstand Restraint 5 24 n/a /14 Torphin whinstone 38 Y Downstand Restraint 6 18 Stability /14 Brick/gravel 38 N Downstand Restraint 7 12 n/a River gravel 25 Y Rectangular Restraint Note 1 Rounded down to nearest full period Fire safety of concrete structures: Background to BS 811 fire design 41

43 Appendix B Results from National Building Studies Research Paper No. 18 Table B1 Test results for reinforced concrete columns - from investigations on building fires No. Fire endurance period Failure Moisture Thickness Cube strength at Age Cover Reinforcement Comments Load attained (minutes) 1 criteria content (%) 2 (mm) time of test (N/mm 2 ) (days) (mm) ratio (kn) 1 56 Spalling 35 x Sloughing of corners Cracking & spalling 28 x Sloughing of corners Cracking & spalling 254 x Sloughing of corners Cracking & spalling 23 x Sloughing of corners Cracking & spalling 381 x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling 35 x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling 381 x Sloughing of corners Cracking & spalling 381 x Sloughing of corners Cracking & spalling 381 x Sloughing of corners Cracking & spalling 381 x Sloughing of corners Cracking & spalling 381 x Sloughing of corners Cracking & spalling 381 x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling 35 x Sloughing of corners Cracking & spalling 254 x Sloughing of corners Cracking & spalling 254 x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x Sloughing of corners Cracking & spalling x * Sloughing of corners Cracking x * Chamfered edges Explosive spalling x 46 27* Chamfered edges Cracking x * Chamfered edges Cracking & spalling x * Chamfered edges Cracking & spalling x * Chamfered edges Cracking & spalling x * Chamfered edges Cracking & spalling x * Chamfered edges Cracking x * Chamfered edges Cracking & spalling x * Chamfered edges Did not fail 279 x * Chamfered edges Cracking & spalling x * Chamfered edges Cracking x * Chamfered edges Cracking & spalling x Chamfered edges Cracking & spalling 279 x * Chamfered edges Cracking & spalling x 279 3* Chamfered edges Did not fail x * Chamfered edges 97 Notes 1 Rounded down to nearest full period 2 Where measured * Cube strength at 28 days 42 Fire safety of concrete structures: Background to BS 811 fire design

44 Appendix C Results of fire resistance tests on elements of building construction Table C1 Test results for concrete floors - from results of fire resistance tests on elements of construction No. Fire resistance Failure Thickness Comments Support period attained criteria (mm) (minutes) Not applicable 51/254 Prestressed concrete T beam incorporating 76 mm structural topping Restraint 2 3 Not known 12 Hollow prestressed Aglite beams incorporating 38 mm structural topping Not known 3 3 Insulation 51/254 Reinforced concrete T form floor Not known 4 3 Insulation 56/254 Reinforced concrete T form floor Not known 5 6 Not known 127 Hollow prestressed floor units incorporating 32 mm structural topping Not known 6 6 Not known 159 Hollow prestressed floor units incorporating 32 mm structural topping Not known 7 6 Not known 152 Reinforced aerated concrete slabs Not known 8 6 Not known 152 Prestressed concrete hollow units Not known 9 6 Not known 12 Prestressed concrete hollow units with 51 mm structural topping Not known 1 6 Not known 89 Concrete rib floor (rib depth not known) Not known 11 6 Not known 35 Prestressed concrete jointed double T beam incorporating 51 mm structural topping Not known Not known 14 Prestressed concrete plank floor including structural topping and screed Not known Not known 152 Aerated concrete slab floor incorporating 38 mm screed Not known Not known 229 Reinforced concrete floor slab and beam Not known Not known 178 Reinforced aerated concrete floor slabs Not known Not known 12 Aglite concrete floor slab Not known Not known 159 Prestressed hollow floor slab incorporating 51 mm structural topping Not known Not known 152 Prestressed hollow floor slab Not known Not known 178 Reinforced hollow concrete slabs Not known 2 12 Not known 15 Reinforced concrete units Not known Not known 159 Prestressed hollow units with suspended ceiling Not known 22 6 Not known 89 Reinforced concrete floor slab Not known Table C2 Test results for concrete columns - from results of fire resistance tests on elements of construction No. Fire resistance period Failure Thickness Comments Support attained (minutes) 1 criteria (mm) 1 Failed reload test Reload 229 x 229 Achieved 12 minutes stability but failed reload test Not known 2 Reload test could not be applied Stability 18 x 18 Achieved 9 minutes stability prior to failure Not known 3 Reload test could not be applied Stability 18 x 18 Achieved 3 minutes stability prior to failure Not known 4 Failed reload test Reload 18 x 18 Achieved 9 minutes stability but failed reload test Not known 5 Reload test could not be applied Stability 18 x 18 Achieved 6 minutes stability prior to failure Not known 6 Failed reload test Reload 28 x 18 Achieved 6 minutes stability but failed reload test Not known 7 12 Stability 229 x 229 Lytag aggregate Not known 8 12 Stability 292 x 292 Prestressed concrete Not known 9 12 Stability 23 x 23 Capstone column Not known Table C3 Test results for reinforced concrete beams - from investigations on building fires No. Fire resistance Failure Thickness Comments Support period attained criteria (mm) (minutes) 1 1 Failed reload test Reload 51/254 Prestressed concrete double T beam - achieved 9 minutes stability but failed reload test Not known Note 1 Rounded down to nearest full period Fire safety of concrete structures: Background to BS 811 fire design 43

45 Appendix D Results from Fire Research Note 741 Table D1 Summary of test results for prestressed concrete beams No. Time (minutes) Beam type Age Cover Design fire Load Spalling Mean Time to to reach critical (months) (mm) resistance (kn) (Y/N) reinforcement max temp deflection l/3 (minutes) temperature (deg C) (minutes) R/PO/W/SR Y R/PO/W/SR Y /PR/W/X Y /PR/S/SR Y /PR/S/X(VG) N R/PO/S/SR (52.5) Y R/PO/S/SR (46) Y Note All 7.6 m span, except for 6 and 7, which are 11.3 m Key R Rectangular PR Pre-tensioned section W Wire tendons S Strands VG Vermiculite-gypsum plaster SR Supplementary reinforcement Collapse X No supplementary reinforcement PO Post-tensioned Figures in brackets are end loads Table D2 Summary of test results for reinforced concrete beams No. Time (minutes) Beam type Age Cover Design fire Load Spalling Mean to reach critical (months) (mm) resistance (kn) (Y/N) reinforcement deflection l/3 (minutes) temperature (deg C) 8 DG/MS/X Y DG/CD/X Y DG/CT/X Y DG/HR/X Y LC/MS/X N LS/MS/X N DG/MS/SR Y DG/HR/SR Y DG/CT/SR Y DG/HR/SR Y DG/HR/X Y DG/HR/X Y 6 2 DG/MS/X (83) Y DG/CD/X (79) Y 5 Note All 7.6 m spans, except for 2 and 21, which are 11.3 m Key DG Dense gravel HR Hot rolled alloy steel LC Lightweight expanded clay SR Supplementary reinforcement LS Lightweight foamed slag X No supplementary reinforcement MS Mild steel Collapse CD Cold worked deformed steel CT Cold worked twisted steel Figures in brackets are end loads 44 Fire safety of concrete structures: Background to BS 811 fire design

46 Related titles on concrete and fire Approaches to the design of reinforced concrete flat slabs R M Moss. BRE Report BR 422, 21, 47 pages. The choice of design method should be based on what is appropriate for the structure, on the designer s experience, and on what is likely to benefit the client most. This report gives pointers as to how existing design guidance and methods could be developed and made more user-friendly, particularly with the advent of Eurocode 2. It points out issues for the Permanent Works Designer to consider as a result of the desire to strike slabs earlier and speed up the construction process. Backprop forces and deflections in flat slabs: construction at St George Wharf R Vollum. BRE Report BR 463, 24, 35 pages. Analysis of backprop forces measured at the European Concrete Building Project s in-situ concrete frame building at Cardington showed that the upper floor carried a greater proportion of the load from casting the slab above than that conventionally assumed. This was investigated at St George Wharf by measuring backprop forces during construction. The work confirms that most of the conclusions from research into construction loading and deflection at Cardington are valid for practical purposes. Best practice in concrete frame construction: practical application at St George Wharf R M Moss. BRE Report 462, 23, 2 pages. This report demonstrates the practical benefits of adopting many of the innovative features and techniques used in the design and construction of the in-situ concrete frame building at Cardington, for which a series of Best Practice guides and companion reports is available. Details are given of the application of innovations to a live project involving the construction of a large residential and mixeduse development. Effect of polypropylene fibres on performance in fire of high grade concrete N Clayton and T Lennon. BRE Report BR 395, 2, 32 pages. Based on fire tests of concrete columns, this report gives information and recommendations on the use of polypropylene fibres to improve the performance of high grade concrete in fire. Incorporating polypropylene fibres into the mix prevented spalling but led to no difference in the ability of the columns to survive the fire test. Separate tests showed that addition of fibres to the concrete led to a reduced cube compressive strength, which needs to be allowed for in design. Fire safety engineering: a reference guide R Chitty and J Fraser-Mitchell. BRE Report 459, 23, 19 pages. This guide provides basic descriptions for key topics in fire safety engineering and aspects that should be considered by designers, enforcers and other responsible persons. It is a reference for those who only occasionally encounter fire safety engineering or are new to the subject. For experienced practitioners it provides a useful short cut to information in detailed references. The text is formatted so that users can annotate their copy to build a personalised reference on fire safety engineering. Performance of high grade concrete containing polypropylene fibres for fire resistance: the effect on strength N Clayton. BRE Report BR 384, 2, 16 pages. This study accompanies research into the possible enhancement of fire resistance of high grade concrete by polypropylene fibres. Addition of fibres caused a small reduction of cube strength, which was related to the reduction in density due to the addition of the fibres. It may affect the design of structural elements, but there is no significant loss of flexural or cylinder splitting strength from the addition of polypropylene fibres. Precast hollowcore slabs in fire T Lennon. Information Paper 5/3, 23, 4 pages. Two full-scale fire tests were carried out at BRE s Large Building Test Facility at Cardington. The objectives were to assess the adequacy of precast hollowcore slabs in terms of the functional requirements of Approved Document B of the Building Regulations. This paper explains the background, describes the test parameters in detail and summarises the results and conclusions. For more information, visit or brebookshop@ihsrapidoc.com.

47 Fire safety of concrete structures: Background to BS 811 fire design Tom Lennon FRS, the Fire Division of BRE The background to methods for establishing the fire resistance of concrete structures specified in BS 811, the British Standard code of practice for structural concrete, has been investigated. In particular, research and test results dating back some 6 years, which have underpinned the tabulated data in all the codes of practice since 1948, have been examined and revisited This publication brings together information derived from testing and research carried out over a number of years. There was a danger that the work supporting the development of codes and standards could have been lost. Collating and assessing the relevant information means that the lessons from the past are recorded and used to help define the strategy for the next generation of codes and standards. The investigations underpinning this publication found that the experimental results used as data for developing the tabulated approach in BS 811 fully supported the provisions of the code in relation to assumed periods of fire resistance. Furthermore in many cases these provisions were found to be conservative as they were based on the assumption that structural elements were fully stressed at the fire limit state and take into account the spalling characteristics of concrete. Evidence from performance in real fires over a number of years demonstrates that the tabular approach to determining fire resistance of concrete elements has been effective. It is suggested that further research could result in greater economies in construction and cost for concrete structures. BRE Bookshop Building Research Establisment Watford WD25 9XX Tel: Fax: brebookshop@emap.com BR 468 ISBN