International Concrete Abstracts Portal

This report describes a five-phase test program. In Phase I, 13 small size specimens were tested. Included were flat plates, flat plates with edge beams, and ribbed slabs. Horizontal and vertical temperature distributions, expansions, and deflections were measured during heating periods. In Phases II and III, computer programs for heat flow cal-culations and stress analysis were developed. Subsequently, six 14x18-ft floor slabs were fire tested. During the tests, slab expansions in both directions were controlled to follow computed time-expansion relationships. Four more 14x18-ft specimens were fire tested in Phases IV and V to verify results of studies to develop methods of simula-ting realistic restraints in a fixed frame furnace through the use of pads made with compressible materials. Expansions and restraining forces measured when compressible pads were used were compared with those obtained for companion slabs tested in a furnace with hydraulically controlled restraining frames. The comparisons show reasonable agreement indicating that it is possible to use fixed frame furnaces to simulate realistic thermal restraints during fire tests of floor slabs.

The thermal and structural responses of prestressed concrete elements under ASTM E-119 fire test exposure were studied analytically. Comparisons of analytical results with experimentally recorded local temperatures and deflections for specimens tested by the Portland Cement Association showed good agreement. In addition, stress histories in concrete and in steel tendons and extent of cracking in concrete during the fire test exposures were determined analytically. Sensitivities in material characterization at elevated temperatures and other modeling uncertainties are discussed.

Knowledge in fire science is sufficiently advanced to provide fire safety in buildings on the basis of engineering design rather than authoritative decisions. Although two important factors in the design considerations, fire load and ventilation, are random variables, conservative design values for them can be found by stochastic and extreme value studies. With these values the design for fire safety is fully deterministic. The design for fire safety has two components: countering the destructive spread potential of fire by the use of fire-resistant compartment boundaries, and countering its convective spread potential by the use of self-closing doors, fire drainage, and flame deflectors. The fire resistance requirements are determined by the normalized heat load on the compartment boundaries which is derived in the course of the design procedure. When dealing with the performance in fire of some key elements of the building, extra precautions are necessary.

The behaviour of concrete and reinforced concrete are examined in the context of rational design for fire resistance. Links with plasticity theory are noted together with the requirements for ductility in Structures designed using these concepts. It is concluded that current methods of rational design are most applicable when structural behaviour is determined by the action of the steel in tension at high temperatures. In considering the limitations of the method, doubts are expressed on extensions to compression loading and shear where the problems of detailing the reinforcement for adequate ductility are more difficult. The problem of local failure by spalling is also considered with different mechanisms being identified and results presented for pore pressures developed during heating. The paper concludes with a pragmatic philosophy for design in which a defensive view of bond, local damage and compression failure is to be adopted within an overall framework of rational design and knowledge from full scale tests.

This paper describes the evalution of the necessary fireprecaution measures for the inner surface of a box girder bridge used for subway traffic, where a considerable fire risk can occur. The study shows in a concise form the evaluation of the relevant amount of fire load caused by the train and gives a summarizing view of a heat balance calculation, which leads to non-steady tem-perature conditions inside the box girder bridge, depending on the fire load and ventilation conditions. Taking into account the decrease in stiffness caused by temperature and cracking, and their combination with the load effects under service conditions in transverse as well as in longitudinal direction, the thermal restraining effects are evaluated. In consequence of this study it was shown that the bridge girder can not resist a fire attack without severe damage, if no inside insulation to the box girder is applied.