Foreword
TCVN 9311-3:2012 is completely equivalent to ISO/TR 834-3:1994.
TCVN 9311-3:2012 is converted from TCXDVN 343:2005 (ISO/TR 834-3:1994) in accordance with the provisions in Clause 1, Article 69 of the Law on Standards and Technical Regulations and Point a), Clause 1, Article 7 of the Government’s Decree No. 127/2007/ND-CP dated August 1, 2007 detailing the implementation of a number of articles of the Law on Standards and Technical Regulations.
TCVN 9311 under the general title “Fire-resistance tests – Elements of building construction” consists of the following parts:
– TCVN 9311-1:2012, Part 1: General requirements.
– TCVN 9311-3:2012, Part 3: Guidance on test methods and application of test data.
– TCVN 9311-4:2012, Part 4: Specific requirements for loadbearing vertical separating elements.
– TCVN 9311-5:2012, Part 5: Specific requirements for loadbearing horizontal separating elements.
– TCVN 9311-6:2012, Part 6: Specific requirements for beams.
– TCVN 9311-7:2012, Part 7: Specific requirements for columns.
– TCVN 9311-8:2012, Part 8: Specific requirements for non-loadbearing vertical separating elements.
The ISO 834 series “Fire-resistance tests – Elements of building construction” also includes the following parts:
– ISO 834-9:2003, Fire-resistance tests – Elements of building construction – Part 9: Specific requirements for non-loadbearing ceiling elements
– ISO/DIS 834-10, Fire resistance tests – Elements of building construction – Part 10: Specific requirements to determine the contribution of applied fire protection materials to structural elements
– ISO/DIS 834-11, Fire resistance tests – Elements of building construction – Part 11: Specific requirements for the assessment of fire protection to structural steel elements
TCVN 9311-3:2012 is compiled by the Institute of Architecture, Urban and Rural Planning, proposed by the Ministry of Construction, appraised by the Directorate for Standards, Metrology and Quality, and published by the Ministry of Science and Technology.
1. Scope
Information is provided for explaining the nature and guiding the use of fire resistance test methods and for applying the data obtained. Areas are identified where research relevant to the performance of test specimen assemblies and their relationship to building construction in practice could usefully be conducted; and to instrumentation and testing techniques.
2. Normative references
The following normative references are indispensable for the application of this standard. For dated references, only the edition cited applies. For undated references, the latest edition of the referenced document (including any amendments) applies.
TCVN 9311-1:2012 1), Fire-resistance tests – Elements of building construction – Part 1: General requirements.
ISO/TR 3956 2), Principles of structural fire-engineering design with special regard to the connection between real fire exposure and the heating conditions of the standard fire-resistance test (ISO 834).
ISO/TR 10158, Principles and rationale underlying calculation methods in relation to fire resistance of structural elements.
3. Standard test procedure
Practical experience shows that it is necessary to make a number of simplifications in the standard test procedure in order to make it operable under controlled conditions in any laboratory with the expectation of obtaining reproducible and repeatable results.
Some of the factors that lead to a degree of variability are outside the scope of the test procedure, notably differences in materials and variations in workmanship are potentially large. Other factors, which are identified in this Technical Report, are within the capability of the user to control. If these factors are given due attention, the reproducibility and repeatability of the test procedure can be improved to acceptable levels.
3.1. Heating conditions
The standard time/temperature curve of the test furnace as described in 6.1.1 of TCVN 9311-1:2012 is essentially unchanged from the original time/temperature curve first used over 70 years ago to control the heating environment in fire resistance tests. It clearly bears a relationship to observed temperatures of fires in practice in terms of, for example, the times for melting of materials at known melting points.
The fundamental purpose of this standard time/temperature curve is to provide a reasonably representative standard heating environment to allow inter-comparison of the load-bearing capacity and separating performance of representative forms of construction. It is important, however, that the standard heating conditions are not regarded as necessarily providing either a reproduction of a real fire exposure or an indication of the anticipated performance of the construction in a real fire. Nevertheless, the extent of testing of both separating and load-bearing elements of buildings on a common basis is such that useful engineering guidance can be derived from the data concerned. Attention is also drawn to the fact that the fire resistance relates to the test duration and not to real time.
ISO/TR 3956 refers to relationships between heating conditions produced by various time/temperature curves commonly occurring in the case of real fires and those embodied in standard fire resistance tests. A range of decay curves is also referred to.
Note that the standard time/temperature furnace curve of the test may also be represented by an exponential function that closely fits the curve defined by the equation T = 345 log10(480t + 1) and may be considered for specific calculation purposes. The curve equation is then:
T = 1325(1 – 0,325 e-0,2t – 0,204 e-1,7t – 0,471 e-19t)
where
T is the temperature rise, in degrees Celsius;
t is the time at which the temperature rise occurs, in hours.
To establish the percentage deviation d defined in 6.1.2 of TCVN 9311-1:2012, the areas under the average time/temperature curve of the furnace and the standard temperature curve above can be compared using either a planimeter on the chart record or by calculation using Simpson’s rule or the trapezoidal rule.
Although the heating regime described in 6.1.1 of TCVN 9311-1:2012 is the fire exposure condition which is prescribed in this Technical Report, it is recognized that this regime is not appropriate to represent fire exposures such as those from hydrocarbon fuels. Such exposures will be addressed more appropriately in other standards relating to fire resistance tests for structures other than buildings. An example of a heating regime which has recently been proposed to represent hydrocarbon fires is as follows:
T= 1100(1 – 0,325 e-0,1667t – 0,204 e-1,417t – 0,471 e-15,833t)
where
T is the temperature rise, in degrees °C;
t is the time at which the temperature rise occurs, in hours.
Or in its convenient form:
T = 1100(1 – 0,33 e-0,17t)
where
t is the time in hours.
3.2. Test furnace
The heating conditions alone as described in 6.1.1 of TCVN 9311-1:2012 are not sufficient to ensure that differently designed furnaces meeting the temperature requirements will impose the same heating regime on test specimens and will, therefore, yield the same test result between them.
The thermocouples used for controlling the furnace temperature are essentially in a state of thermodynamic equilibrium with an environment governed by convective and radiative heat transfer conditions prevailing in the furnace. Convective heat transfer to an object exposed to fire will be a function of its geometry and size and will in general be higher for objects smaller than the thermocouple bead than for larger ones such as the test specimen. Convection will therefore tend to have a greater influence on the thermocouple temperatures while heat transfer to the test specimen will be primarily influenced by radiation from the hot furnace wall and from the flames.
Within the furnace there are both gas and surface-to-surface radiations. The gas radiation depends on the temperature and absorptive properties of the furnace gases and also depends heavily on the visible emissivity of the flames.
The surface-to-surface radiation depends on the temperature of the furnace walls, their absorptivity and emissivity and also on the size and shape of the furnace. The wall temperatures in turn depend on their thermal properties.
The convective heat transfer to an object depends on the local difference in temperature between the gases and the surface of the object and on the velocity of the gases.
The radiation from the gases corresponds to their temperature and that from the specimen of the combined effects of gas and furnace wall radiation. The wall radiation is at first lower but increases as the walls become hotter. The thermocouples prescribed in this document are small and will follow the gas temperature. Test specimens, by contrast, will be more sensitive to radiation.
From the above, it is clear that an ideal solution to achieving stability between tests organized in accordance with this document would only be achieved if generally accepted “ideal” designs for the test furnace(s) were specified precisely regarding size, shape, materials, construction techniques and the nature of the fuel employed.
One method for mitigating the problems described above, which may be applied to existing furnaces of widely differing design, is to line the furnace walls with a low thermal inertia insulating material which follows more closely the gas temperature in the furnace, for example a material having properties prescribed in 5.2 of TCVN 9311-1:2012. The difference between the gas temperature and the wall temperature will be reduced and the increasing heat input to the specimen from the furnace walls by radiation will therefore improve the similarity between the results obtained from widely differing furnace designs.
Where possible, the design of existing test furnaces should also be reviewed with regard to furnace geometry and location of flues to avoid swirling flow and the resulting pressure fluctuations causing uneven heating of the surface of the test specimen.
5.5.1.1 of TCVN 9311-1:2012 specifies the design of thermocouples used for measuring and thus controlling the furnace environment and testing may be conducted using more responsive thermocouples to the combined effects of radiation and convection as an alternative means for reducing problems arising from different thermal characteristics of furnaces [1].
Ultimately, one of the most effective tools in adjusting existing furnace designs in order to improve the stability between furnaces is regular calibration (see 3.11).
3.3. Conditioning of test specimens
3.3.1. Sub-standard moisture content in concrete
At the time of the test, as required by 7.4 of TCVN 9311-1:2012, the moisture content in the test specimen shall be as close as possible to that expected in normal use.
Building elements exposed to the ambient environment tend to follow a cyclic variation in temperature and/or humidity of the atmosphere unless in continuously air-conditioned and heated buildings. The nature of the constituent materials and the dimensions of the elements will determine the degree of fluctuation of the moisture content of the element around a mean value.
Relating the state of the test specimen to normal service conditions may lead to variations in moisture content of structural elements of test specimens, notably of constituent materials having a high moisture absorption capacity from the air such as Portland cement, gypsum and wood. However, after conditioning in accordance with 7.4 of TCVN 9311-1:2012, of the common inorganic building materials, only hydrated Portland cement products are likely to retain sufficient moisture to significantly affect the result of the fire resistance test.
For comparative purposes, discrepancies in moisture content are best adjusted by using the moisture content established at the equilibrium state by drying in an ambient atmosphere having a relative humidity of 50 % at a temperature of 20 °C as a standard reference condition.
If the fire resistance associated with the insulation performance of the specimen at a particular known moisture content is known, then the fire resistance at any other moisture content can be calculated from:
T2d + Td (4 + 4bF – TF) – 4TF = 0
where
F is the moisture content, in g/m3;
TF is the fire resistance at moisture content F, in h;
Td is the fire resistance in the oven-dry condition, in h;
b is a coefficient which varies with the porosity.
(For bricks, dense concretes and gunite, b may be taken as 5.5; for lightweight concretes as 8.0 and for aerated concretes as 10.0).
Alternatively, calculation procedures such as those described in [2] and [3] may be used.
If artificial drying techniques are used to obtain the moisture content appropriate to the standard reference condition, the laboratory undertaking the test should avoid methods that may significantly alter the characteristics of the materials constituting the test specimen.
3.3.2. Determining the moisture content of hardened concrete by relative humidity
A method for determining the relative moisture content in hardened concrete test specimens may employ electrical sensing elements as described in [4]. A similar procedure using electrical sensing elements may be applied to determine the relative moisture content in fire resistance test specimens made from other materials.
For timber structures, where appropriate, an electrical resistance moisture content meter may be used as a means of measuring relative moisture content to determine when the timber has reached the required moisture content. The electrical resistance method of moisture content measurement is described in [5] and [6].
3.4. Fuel supply and heat distribution
At present, fuel consumption measurements are not among the data required to be recorded during a standard fire resistance test, although this parameter is often measured by laboratories and users of this document are encouraged to gather such data for further development.
The following guidance may be applied to the recording of fuel consumption in the test procedure.
Every 10 min or less the cumulative fuel supplied to the combustion chamber shall be recorded. The total amount of fuel supplied throughout the test shall also be determined. The use of a continuous recording flow meter is preferable to periodic readings on an instantaneous flow meter or a totalizing flow meter. The metering and recording system shall be selected so that the accuracy of flow measurement is within ± 5 %. The nature of the fuel, its gross calorific value and the cumulative consumption (corrected to standard conditions of 15 °C and 100 kPa) during each period shall be reported.
Where fuel consumption measurements are made, the measurements indicate the heat input to the furnace environment during the later stages of tests on specimen assemblies containing combustible components. This is not normally a matter of concern in national regulations which prescribe functional requirements for combustible constructions based upon a deemed to satisfy the basis of classification related to functional requirements and the height and area limits of the building to which they are applied.
It shall also be noted that fuel consumption measurements may be significantly affected when testing water-cooled steel constructions or steel constructions with large cross-sections.
3.5. Pressure measurement techniques
When setting up the pipework used in pressure sensors, the sensor and reference pipe shall always be treated as a pair, and the leads (connecting together) taken off horizontally in any direction with respect to the measuring instrument. Where no reference pipe is present, it should still be considered to be notionally present (the air in the room between the two points of measurement symbolizing the reference pipe in this instance).
When the reference and sensing pipes are at the same elevation, the pipes may be at different temperatures.
When the reference and sensing pipes are curved from one elevation to another, then the pipes shall be at the same temperature. They may be hot at the top and cold at the bottom but the temperature at each elevation shall be the same.
Attention shall be paid to the location of sensors within the furnace to avoid subjecting them to aerodynamic effects caused by velocity and turbulence of the furnace gases (see also [8]).
3.6. Post-heating procedures
TCVN 9311-1:2012 does not specify requirements to be applied or referred to for a post-heating procedure. However, in practice in some countries, the test load or a multiple of the test load is maintained for a period commonly of 24 h after the test. The objective of this procedure is to gain some overall information concerning the post-fire strength and stiffness of the building structure represented by the specimen. As this information is difficult to relate to a fire (or post-fire) situation, it was concluded that any requirements for a post-heating procedure were outside the scope of this document.
Some countries follow a tradition of assessing the additional performance of separating constructions by subjecting them to some form of impact test, immediately after the fire test. This is intended to simulate the effect of falling debris or a firefighter’s hose stream on a fire separating construction in situations where the separating construction is required to maintain its effectiveness throughout a burnout or after the period of firefighting. This impact test may be applied after the full completion of the fire test period or after some fraction (e.g. one half) of the prescribed period and is commonly regarded as a measure of stability, irrespective of any implications as regards the simulation of the extinguishment of fires by firefighters’ hose streams.
In most instances, both these tests preclude the possibility of continuing the fire test beyond the prescribed fire period. With the increasing need to provide data for extrapolation and other calculation purposes, testing organizations should be encouraged to maintain the fire test period until the appropriate failure criteria can be safely exceeded.
3.7. Size of test specimen
TCVN 9311-1:2012 prescribes in general terms that fire resistance tests shall be conducted on specimens of full size. It also recognizes that this is frequently not practicable because of limitations of the size of testing facilities. In cases where a full-size specimen cannot be used, a reduced scale simulation may be constructed to the minimum dimensions standardized for a representative specimen for a room of 3 m height and having a cross-section of 3 m × 4 m.
The use of full-size test specimens is encouraged for several reasons arising from the difficulties of achieving wholly scaled fire performance of most load-bearing elements and some separating elements.
For the majority of non-load-bearing elements, the overall dimensional reductions to convenient sizes for testing purposes do not pose any serious problems particularly where modular constructions are concerned.
For load-bearing systems, it is important to emphasize the need to maintain functional similarity when reducing the scale of the fire test specimen. For example, aspect ratios should not be changed when the real dimensions of floors are scaled down. In other words, it is necessary to maintain the balance between the various stress forms which the specimen represents in the reduced-scale replica and to determine representative stresses on the basis of the scale of the building element considered.
3.8. Specimen construction
TCVN 9311-1:2012 prescribes that the materials used in the test specimen, the method of construction and the workmanship shall be representative of those used in practice.
This means that characteristic features such as joints in the test specimen, expansion provisions, or fixing details shall be incorporated in the test specimen in a representative manner.
It should be noted that, except in special circumstances, test specimens may be constructed to a higher standard than may be found in practice. On the other hand, it is also important to maintain consistency in specimen construction so as not to produce anomalous results from defects in this respect.
It is therefore essential for the test data to be supplemented by accurate and detailed descriptions of the test specimen and its condition at the time of the test and, when necessary, for these features to be highlighted in explaining any anomalies in the test results.
3.9. Loading
The load applied to a specimen during a fire resistance test has a significant effect upon its performance. This is also an important consideration for the continuing application of the resulting test data and its relation to other tests or similar constructions.
6.3 of TCVN 9311-1:2012 provides several alternative bases for selecting the load. The most widely applicable basis is one which relates to the determination of an appropriate test load and the resulting stresses, to the nominal properties of the material(s) incorporated within the element(s) used in the specimen construction and which results in the stresses approaching the appropriate limit states in these elements, in which the maximum permissible stresses are defined by an acceptable national structural code. This defines the test load application quite rigorously, as well as providing a reliable basis for extrapolation of test data and application in calculation methods.
The second basis relates the required test load to properties of the materials incorporated in the specimen. These values are often provided by the material manufacturer or obtained by reference to documentation relating to the standard properties of the materials concerned (often presented in terms of a range). In most instances, these will tend to produce safe values for the test load since actual values will generally be higher than characteristic values and the elements will not be stressed to the limit states defined in the design methods. On the other hand, this practice relates closely to typical design methods and the corresponding specifications for materials used in building construction. The utility of the results obtained from such tests may be enhanced if the actual properties of the materials are determined or if the actual stresses in the elements of the specimen are measured at some point during the fire test.
The third basis differs from those described above in that the load derived relates to a specific and limited application. The test load will always be lower than that normally applicable where the elements have been selected on the basis of limit state serviceability requirements defined in the relevant structural codes and a margin of safety and fire resistance will be available when compared with the performance of specimens loaded on the basis of the two previously described bases. Again, the utility of the test results may be enhanced if data relating the actual physical properties of the materials incorporated in the test specimen with the stress levels obtained in these elements under the prescribed load are available.
Apart from the relevant basis for developing the load to be applied during the test, it should be noted that the structural codes, applied to the design of the building, may themselves prescribe different design parameters which do not always involve a comparable assessment in different countries. There are considerable variations in the analysis of the characteristic features of wind, snow and seismic loading.
It is important to note that, whatever method is used to develop the load during the fire resistance test, it will be related to the limiting load on the test element before heating, and it is important that the report should clearly state the basis of load development and any information relevant to the interpretation and application of the test results such as the material properties and the stress levels concerned.
From the above, it is clear that at points of concentrated loading, an accurate simulation can be made of the stress conditions as tested for beams and columns. More attention needs to be paid to simulating the effect of uniformly distributed loading for floors and walls. A maximum number of load points shall be applied and the loading system shall be adapted to the expected deflection during the test and maintain the required load distribution.
3.10. Restraint and boundary conditions
3.10.1. Introduction
6.4 of TCVN 9311-1:2012 provides a number of options for the application of restraint, in terms of resistance to thermal expansion or rotation, to load-bearing systems. This clause reflects the original philosophy of the test method described in TCVN 9311-1:2012 of testing specimens under conditions as close as possible to their use in practice.
The following principles should be applied to requirements concerning the restraint of test specimens to the conditions pertaining in actual building construction:
– Floors, roofs, wall structures, columns and isolated beams in buildings shall be considered for their ability to provide restraint against thermal expansion and/or rotation when the separating construction and supporting construction can withstand the forces generated throughout the range of elevated temperatures represented by the standard time/ temperature curve;
– In carrying out any such technical assessment to determine the restraining ability of building elements, it shall be recognized that this ability may be due either to the lateral stiffness of the supports to floor, roof and isolated beam assemblies forming part of the building, or to the self-weight of the supporting construction. There shall also be sufficient connections to transfer the forces generated by thermal expansion and/or rotation into the supports or supporting construction. The stiffness of any adjacent slabs or structures also needs to be considered for their ability to provide restraint against thermal expansion. Continuity arising from continuous beams, for example, spanning more than two supports, also has the potential for inducing rotational restraint;
– It is well known from test results that variations in restraint can have a significant effect on the fire resistance period of an element or assembly. In most cases, the application of restraint during the fire test is beneficial to the performance of the test specimen. However, in some instances axial restraint in excess of allowable limits may accelerate the onset of instability or fracture occurring in concrete structures. In other cases, for concrete slabs with reinforcement on the side remote from the fire, the fixing moments may induce serious tensile cracking in under-reinforced or unreinforced regions, leading to tensile failure.
Experience of fire testing restrained structures may anticipate some of the abnormal effects noted above. It may also relate in a more general way the conditions of restrained test specimens to the conditions of actual building construction. However, a great deal remains to be done and, in the absence of an ability to relate the required boundary conditions of a specimen to the boundary conditions to which the element will be subjected in actual building construction, it is common practice to conduct the experiment under conditions of minimal or no resistance to expansion or rotation.
3.10.2. Flexural elements (beams, floors, roofs)
Specimens incorporating flexural elements are either exposed to fire whilst supported on roller bearings or tested within a restraint frame. In the case of restraint, resistance to thermal expansion, axially or rotationally, may be applied in a number of ways. In the simplest of these arrangements, the specimen is mounted in a restraint frame of such dimensions that it will resist the axial thrust imposed by the elements in the specimen without significant bowing. In some instances, this axial thrust has been measured in terms of the size of the restraint frame. In other instances, some adjustment is made by means of an expansion gap between the ends of the elements and the restraint frame. Such an arrangement will also provide rotational restraint by contact and thereby effectively fix the ends of the elements to the full depth of their section and the section of the restraint frame. In more elaborate arrangements, restraint and measurement of the degree of restraint is provided by hydraulic jacks arranged axially and at right angles to the element(s).
In those instances, where resistance to thermal expansion occurs, the heating during the fire test generates an axial compressive force in the elements concerned. In most cases, this force will act at a position in the cross-section of the element where the corresponding bending moment will tend to oppose the effect of the bending moment due to the applied load and thereby increase the load-bearing capacity and fire resistance, unless fracture or instability due to potential overloading negates this beneficial influence. In most instances, should a flexural element be tested under conditions of minimal restraint, the resulting data will tend to be safe when applied to an element in a building where some restraint to expansion in fire will occur.
3.10.3. Axially-loaded elements (columns, loadbearing walls)
Fire tests on columns and load-bearing walls are carried out under laboratory conditions idealizing the stresses in a real fire. For example, in a test it is not practicable to reproduce the end-moments that may occur in a real fire. The effect of fixing moments in practice depends on the nature, severity and location of the fire, in the compartment. In the event of even heating conditions occurring over a compartment, the effect of restraint against expansion may be considerably reduced.
The load-bearing capacity and the related test load of columns and loadbearing walls are considerably dependent on the supporting conditions. For such elements of a strut type, assumed to be hinged, even small forces developing due to friction at the bearings may considerably increase the load-bearing capacity. In a fire test, the unintentional application of restraint to the ends of the test specimen may lead to a considerable increase in the test endurance. Experience in some laboratories indicates that it is very difficult to produce purely axial reactions (or loading) concentric with a column, even when using spherical seats, and the suggestion is made to incorporate a known small eccentricity.
For the reasons given above, tests on columns or loadbearing walls are conducted either without restraint to expansion (elongation) or with the ends fully restrained.
3.10.4. Non-loadbearing walls and partitions
Logically, all non-loadbearing walls and partitions should be tested without the imposition of external forces. However, in practice, such elements may be subjected to load transferred from other elements of construction or to reactions resulting from the expansion of the element itself when exposed to fire. It may, therefore, be appropriate to conduct tests on such elements within a restraint frame of sufficient rigidity to interact with the expansive forces generated in the specimen during the test with minimal or no deflection.
3.10.5. Laboratory measurements
In view of the present lack of information on the effects of restraint to thermal expansion or rotation, laboratories should endeavour to determine the magnitude and direction of the restraint forces when testing specimens restrained in any form.
3.11. Calibration
Calibration is the means of ensuring that identical test specimens tested in accordance with this document, in different furnaces or in the same furnace at different times, will yield comparable results. If this objective were achieved, the time at which the defined performance of the specimen is reached should be reproducible within acceptably small limits in respect of both load-bearing capacity and insulation.
The principal feature of calibration in fire resistance testing relates to the methods and instrumentation for controlling and measuring the temperature, pressure and air movement in the test furnace. The objective of a furnace calibration experiment is to establish uniform heating conditions over the whole of the hot face of the test specimen and to achieve the prescribed level of heat exposure. The purpose of such an experiment is also to ensure that a linear static pressure gradient exists over the hot face of the specimen in the vertical direction, and that an even distribution of static pressure exists over the hot face of specimens in the horizontal mode.
A method of calibration concentrating upon the temperature and pressure conditions within the test furnace is described in the relevant referenced document.
The load-bearing capacity of a specimen may also be affected by factors such as the manner of supporting the specimen; the boundary and restraint conditions, the application of the design load; and the measurement of load magnitude, deformation and deflection by instruments compared with reference standards. No method of direct calibration exists which evaluates the above features and reliability shall depend upon consistency in the specification of these parameters in the test method and upon the achievement of the pressure and temperature conditions based upon the method described in [9].
4. Fire resistance criteria
4.1. Objective
The objective of a fire resistance determination, as prescribed in TCVN 9311-1:2012, is to assess how an element of building construction will perform under standard conditions of heating and pressure. The test method described in this document provides a means of quantifying the ability of a building element to maintain its designed function under defined elevated temperature conditions in terms of performance criteria. These criteria are used to ensure that, under these test conditions, the test specimen will continue to perform its designed function either as a load-bearing element or as a separating element, or both. The criteria are based upon load-bearing capacity and the transmission of fire. Fire may be transmitted from one compartment to another in two ways, either integrity failure or by the transfer of heat in quantities sufficient to cause unacceptably high temperatures at the surface remote from the fire.
The time/temperature curve prescribed in this document represents only some of the heating conditions which may occur in the growth period of a fire and the test method does not attempt to quantify the performance of an element in a real fire situation over a specific time period (see 3.1).
4.2. Load-bearing capacity
This criterion is used to determine the ability of a load-bearing element to support its test load, throughout a fire resistance test, without collapse. To measure load-bearing capacity without the necessity to continue the test to actual collapse, limits of deflection and rate of deflection for floors, beams and roofs are laid down. It is not practicable to specify limits for walls because experience indicates that the rate of deformation recorded immediately prior to failure for walls of various types of construction can vary over a considerable range.
4.3. Integrity
This criterion may be applied to separating elements and provides a measure of the ability to prevent the transmission of fire and hot gases from the fire-exposed side to the non-exposed side of the specimen; it is dependent upon the time which elapses before flaming of a cotton pad occurs at any location of cracks or fissures. The tendency for the pad to ignite will depend upon the size of the gap, the pressure within the furnace at the position of the gap, the temperature and the oxygen content.
Surface flaming on the fire side of the element may constitute a hazard that is not permissible and, therefore, the site where a pad might ignite also represents the location where the integrity criterion is not fulfilled.
4.4. Insulation
This criterion may be applied to separating elements and provides a measure of the ability of the specimen to restrict the temperature rise of the non-exposed face to defined limits.
When a separating element under test is uninsulated or exceeds the prescribed temperature limits, the heat radiated from the non-fire-exposed face may be sufficient to cause ignition of cotton pads.
The prescribed levels are set to ensure that, at temperatures below these levels, any combustible materials in proximity to the non-fire-exposed face will be unlikely to ignite at temperatures below the limits. The maximum temperature rise limits make allowance for potential areas on the element which may provide paths of direct heat transfer and cause hot spots on the non-exposed face, where test specimens are in accordance with the requirements given in 5.5.1.2 of TCVN 9311-1:2012.
It has been suggested that the limiting temperature rise values are conservative, in that they are based on the assumption that the temperature of the non-fire-exposed face continues to rise after the fire has been removed from the test assembly. Tests have been carried out [10] using containers filled with wood-wool or wood shavings placed against the non-fire-exposed face of brick walls exposed to fire in accordance with the standard fire resistance test. No evidence of ignition of wood or cotton was observed at temperatures below 204 °C (or a temperature rise of 163 °C) for fire exposure periods of between 1.5 h and 12 h. Signs of ignition were observed when the temperature was in the range of 204 °C to 232 °C and definite evidence of ignition was observed when the temperature reached 232 °C to 260 °C.
4.5. Other characteristics
When applying the test methods prescribed in this document to constituent materials of the test specimen, a number of other characteristics, which are undesirable in service, may arise during the testing, for example, the production of smoke. These phenomena are not covered in this document and are more appropriately assessed by separate test methods.
5. Classification
Buildings are commonly regulated in terms of limits of height, area, occupancy and separation by dividing them primarily on a deemed to satisfy basis of fire resistance of structural and separating elements, determined from the results of standard fire resistance tests applied to specimens representative of elements of such buildings.
This document provides a system for expressing the performance of building elements when subjected to the heating and pressure conditions of the test, by reference to the duration of exposure over which the relevant criteria, for example, structural, integrity and insulation, continue to be satisfied.
National regulations and requirements use a variety of methods for expressing the fire resistance requirements in practice. Some countries require absolutely that the element shall be considered to have satisfied all the relevant performance criteria for the necessary time duration. Other national standards and in different circumstances require only one or two specific performance characteristics to be satisfied for all or part of the fire test duration. It is, therefore, appropriate for regulations and requirements to refer to systems by clear and specific citation when satisfaction of a particular set of criteria is necessary.
Requirements for fire resistance are referenced by a fire resistance classification and fire resistance level. Classification and time periods are frequently designed in half-hour or one-hour steps, ranging from 0.5 h to 6 h. For design evaluation, it is necessary that a system satisfy the criteria for a period at least equal to the classification period cited, with a minimum of 1 h. In some countries, letters of the alphabet are used to indicate characteristic fire resistance periods; in other countries the time set down in appropriate standards may be used as the designation code.
It should also be noted that some countries distinguish between classifications of combustible and non-combustible construction. It is common practice in some countries to add a coded letter or some other form of designation referring to the relevant element of construction to the fire resistance classification level of the building.
6. Repeatability and reproducibility
Although this document has been reviewed with a view to increasing repeatability and reproducibility, to date no comprehensive programme of testing has been undertaken to explore statistically the repeatability and reproducibility of the fire resistance test defined above. As repeat testing of identical specimens is not a requirement nor a common practice, statistical data exists for relatively few variations. However, a number of sources of collected data do exist.
Repeatability and reproducibility are most commonly expressed in terms of standard deviation or coefficient of variation (the ratio of the standard deviation to the overall mean expressed as a percentage); it may also be expressed in terms of repeatability limits or relative accuracy (repeatability limits within which two averages of tests can be said to agree 95 % of the time).
At present no estimates are available for a reproducibility factor, but experience indicates that reproducibility between laboratories may be two or three times that of within-laboratory repeatability.
Repeatability and reproducibility may be improved by investigation of the following factors.
6.1. Repeatability
Repeatability is a measure of the variability associated with repeat tests of a nominally identical assembly in a single laboratory. The variability of measured fire resistance time may be due to either random or systematic factors, and may involve:
a) The test assembly;
b) The equipment (in terms of the furnace and loading equipment);
c) The instrumentation;
d) The operator (testing or supervision);
e) Environmental effects.
Random factors include variability in materials and workmanship; magnitude of load and load distribution (e.g. degree of fixity, stability of end conditions, eccentricity of loading); variability of instrumentation and measuring equipment; effects of operator, changes in environment (temperature, humidity, etc.).
Systematic factors include such items as the above, i.e. different operators, equipment assemblies, personnel, test assemblies; systematic shifts (increase or decrease) in furnace temperatures and pressure; shifts in calibration in instrumentation and measuring equipment.
In some instances, a limiting factor may involve both random and systematic aspects. For example, the magnitude (and variability) of furnace pressure may cause premature failure of a suspended ceiling component of a floor-ceiling assembly. This may happen randomly at a pressure level (being controlled) and systematically at a slightly higher pressure level.
6.2. Reproducibility
Reproducibility is a measure of the variability associated with tests of a nominally identical assembly in different laboratories. The random and systematic factors noted above may also apply to inter-laboratory variations. Characteristic systematic factors which may increase the variability include:
– Differences in furnaces (e.g. specimen size, nature of fuel, number, shape and direction of flues);
– Structural loading (e.g. method of loading, load distribution, eccentricity of loading);
– Boundary conditions (e.g. restraint, cooling around furnace);
– Instrumentation used for control and measurement (e.g. automatic/manual; temperature; pressure);
– Interpretation of test conditions and criteria.
7. Interpolation and extrapolation
7.1. Interpolation
Interpolation is the determination of the effect of a change in an element of construction which has previously undergone a series of fire resistance tests and has been accepted in a fire resistance classification within the established range of the testing. Interpolation requires mathematical or empirical relationships, developed on the basis of at least two test results. The factors that may be considered are: changes in dimensions, materials or design within the range of variations checked by test.
7.2. Extrapolation
Extrapolation is the effect of a change in an element of construction that has been fire tested and accepted in a fire resistance classification with the aim of extending a classification beyond the range established by tests. Extrapolation requires a fire model developed on the basis of one or more tests and the relevant fire performance data. The factors that may be considered are: changes in dimensions, materials or design which are usually outside the range of variations validated by the tests. The reliability of the extrapolation depends on the validity of the fire model used and this should be stated when the procedure is applied.
Several other factors influence the potential for establishing interpolations and extrapolations. When the need for such data is anticipated, all the relevant parameters must be monitored and if necessary additional measurements must be made to facilitate this. There are three main parameters to be considered for these purposes:
a) Changes in dimensions: length, width, thickness, etc.;
b) Changes in materials: strength, density, thermal properties, moisture content;
c) Changes in loading and design – load, boundary conditions, jointing and stability modes.
The relevance of the above parameters depends upon the type of specimen and the variations considered. Only some of the factors involved can be cited in a few typical cases. To do this, test specimens can be divided into limiting considerations of load-bearing and separating function. In the former case, it is primarily necessary to ensure that the variant will effectively support the loads and in the latter that insulation and integrity are maintained. In some instances, both considerations apply.
The load-bearing assemblies most amenable to simple principles are insulated steel systems, concrete constructions protecting reinforcing steel and timber constructions where the charring rate is a limiting factor. In the case of steel elements, differences in dimensions, loading and design concept will lead to a changed limiting temperature for the insulation material. For concrete elements, a similar approach may be used for simple systems where either the temperature of the steel in the concrete needs to be limited, or with more complex assemblies, redistribution of stresses and strains also has to be taken into account. Most timber structures can be analysed on the basis of the initial strength of the uncharred cross-section. Several published documents provide guidance for some typical structural systems in the above materials.
Interpolations and extrapolations may be divided into 4 categories, each of increasing complexity. The exact principles and the limits of application will need to be agreed by the national bodies using such procedures:
a) Quantitative design principles based upon fire resistance tests and general concepts. These principles are only useful to specialists in this field.
b) Quantitative design principles (or rules of thumb) based upon fire resistance tests where a specific value of fire resistance of materials, products has been provided to eliminate unrealistic outcomes.
c) Regression techniques: The monitoring of a number of parameters in a series of tests and the determination of the most suitable relationship derived using regression techniques.
d) Physical model: The development of a physical model relating fire resistance to the appropriate material properties, either from first principles or by using test data. Once the model has been rationalized, fire resistance may be determined from appropriate input data.
Careful consideration should be given to the use of interpolation or extrapolation techniques to determine fire resistance classifications in cases where the data is incomplete or when the construction being considered is not substantially represented by the construction fire tested, on which the interpolation or extrapolation is based.
Further reference can be made to ISO/TR 10158.
8. Relationship between fire resistance and building fires
In considering this relationship, it is necessary to appreciate that the fire resistance determination is carried out under a complete test procedure. When compared with building fires, attention is focused on the time/temperature curve and its relationship to the temperatures and rates of temperature rise that may be achieved in real fires under different fire situations.
The test is used to assess building constructions and thereby postulate the necessary level of fire safety. This is achieved by the application of fire resistance test results through regulations or codes of practice, which specify the performance considered necessary in given situations. The adequacy of the approach is monitored by feedback from experience which is, in general, able to avoid unacceptable levels of damage.
The test result expressed as a fire resistance classification or a fire resistance level represents the period for which defined criteria continue to be satisfied.
This period represents a relative classification of performance and cannot be related directly to the situation in a building fire. It is important that the transition from an expression of time to the technical characteristics of a building in fire is established through building regulations.
The actual performance achieved in a fire resistance test is closely related to the test conditions, to the degree to which the building is simulated in the test and to the criteria applied to detect failure under the test. A small change in the test conditions regarding failure, particularly when related to integrity and insulation, can also significantly influence the assessment result.
In particular, the time recorded in a fire resistance test, for these criteria, does not bear a direct relationship to the time of failure, in a real fire situation. This has been recognized as a principle since the beginning of testing [12], [13].
Performance testing by means of fire resistance tests has existed for more than a century. The early tests created fires using gas, oil, wood and even combinations of the above. The wide variation in test conditions caused difficulty in comparing and assessing the results obtained.
The first real moves towards a more unified approach were made in the USA when a Committee of the ASTM in 1918 proposed a time/temperature relationship very close to that of the present international standards. The temporal resolution of the early test furnaces seems to have been very dependent upon the original time/temperature curve. This curve was established which could be used for a range of furnaces even in different countries. In consequence of this a furnace, within the scope of the standard, tends to be self-operating, i.e. following the standard curve with very little intervention of the operator.
A system of classification was introduced, in which elements sustained for a longer period in the fire test furnace were deemed to provide a better performance when an actual building fire occurred. Ingberg, using the equal area concept first, attempted and represented the standard test in terms of real fire exposure, deriving equivalent relationships between the fire load simulated and the required fire resistance period.
Many more recent attempts have been and are being made to reinforce the link between test method and real building fires. These attempts have been extended to include factors such as ventilation, compartment size, fire load and the thermal characteristics of the compartment. The object of such attempts is to quantify the level of severity that may occur in a fire, and thereby, through relationships derived from experimental experience, prescribe a fire resistance period achievable in a test which will ensure safety. Much of this work has been reviewed by Odeen.
Fire resistance testing is considered as a means of measuring the relative performance of elements of construction under fire conditions in which an approximation is made to both real fire and prototype conditions.
Attempts to make tests more realistic need to be viewed with caution. Any measure which significantly alters established fire resistance classifications shall be validated by experience in which the test result is used, and only implemented if changes in the level of safety are considered necessary and appropriate.
REFERENCES
[1] Wickstrom, U. The plate thermometer- A simple instrument for reaching harmonized fire resistance tests. SP-RAPP 1986: D. Swedish National Testing Institute, Boras 1983 : 03.
[2] Harmathy, T.Z. Experimental Study on moisture and fire endurance, Fire technology, I 91), 1986
[3] ASTM E 119 Standard method of fire tests of building construction and materials
[4] Menzel, C.A, A. Method for determining the moisture condition of hardened concrete in term of relative humidity. In Proceeding American society for testing and materials ASTM .55.1955, page 1 085.
[5] Wood handbook of the forest products laboratory, US Department of Agriculture, pp. 14-2,14-3, 1987.
[6] ASTM D 4444, Standard test methods for use and calibration of hand-held moisture meters
[7] NBSIR 81-2415, Furnance pressure probe investigation, National Bureau of Standards.
[8] OLSSON, S, Swedish National Testing Institute technical report of standards, SP-RAPP 1985-2
[9] CEN/TC 127/WI 57, Fire resistance testing- Commissioning and calibration of furnaces.
[10] INGBERG, SH. Fire tests brick wall, building material and structures report 143 . US Department of Commerce, National Bureau of Standards, 1954.
[11] Task group report on repeatability and reproducibility of ASTM E119 fire test. ASTM Research report RR : 05-1003 (1981)
[12] BLETZACKER, RW. The Role of research and testing in Building Code Regulation. News in Engineering. The Ohio State University, 1962
[13] BS 476-10:1983, Guide to the Principles and application of fire testing.
[14] ODEEN, K Standard fire endurance test- Discussion, Criticism, and Alternatives. ASTM STP 464, 1970
[15] ROBERTSON, A.F. and GROSS D. Fire load, fire severity, and fire endurance. ASTM STP 464, 1970.
[16] HARMATHY,T.Z. The fire resistance test and its relation to real- world fires. Fire and materials 5(3), 1981.
[1] Will be published.
[2] ISO/TR 3956, ISO/TR 10158 are now cancelled.
