In performance-based fire safety design of buildings, design fires are assumed based on relatively coarse information of potential combustible objects described by the object name, weight, size and constituent materials. The fire effect is calculated by using assumed design heat release rate (HRR) to judge appropriateness of fire safety provisions. For this purpose, a simple estimation method to estimate design HRR curve (full time history of HRR) was developed. The method is based on summary of available experimental datasets of various combustibles, which were categorized into groups by their names and constituent materials. As a nature of categorized groups, there are two types of variations of HRR characteristics within each category group. One type is that the object size is almost constant within each category group. For this type, HRR curves are simply averaged over the available datasets in order to obtain the characteristic design HRR curve. The method was applied to chairs and sofas. The other type is that the object size varies within each category group. To develop size-dependent HRR curves, rates of fire growth and decay were averaged over datasets in the group. Maximum HRR and total heat release (THR) were correlated with surface area and weight, respectively.The method was applied to Christmas trees. By using the proposed methods, it is possible to establish a simple formula to represent full HRR diagrams by their category names.

It is difficult to create a fire risk evaluation method that is applicable to different map scales from a single building to the whole region of Japan. Conventional simulation methods use city indices or mesh-data, but in the proposed method the fire destruction probability of each single building can be calculated. Thus the method reflects the actual spatial characteristics of a city and can be used for general purposes with no margin of error. It is possible to rearrange the results in different area units and to obtain the numbers of burned out houses and areas. In addition, the method provides appropriate information for conveying the fire risk to the public by showing them cluster data of the simulation. This cluster data is sometimes referred to as the Fatalistic Collaboration Community Unit.

No abstract

In this study, it is pointed out that the safety level that can be achieved by the Verification Method for Fire Resistance (hereinafter, the Verification Method) is not necessarily made clear in engineering sense. Using the theory of structural reliability engineering, the safety level that can be achieved by complying with the Verification Method was estimated. By using a sofa set in a hotel lounge as the fire source for a localized fire and a typical office room as the fire source for a fully-developed fire, the steel members exposed to fire were subjected to analysis.The limit state conditions were defined by the conditions where the critical temperature for which a steel member withstands the applied load becomes equal to the maximum steel temperature developed by the exposure to fire. Then, (1) the input parameters were extracted, and (1-1) with respect to the margin of the steel temperature, which is defined as the difference of these temperatures, the input parameters associated with the calculation formula were categorized in order to quantify the variability and uncertainties. (1-2) In the case of a localized fires, the history of the heat release rate of combustible materials and its range of variation are quantified and taken into consideration for establishing design fire. (1-3) In the case of a fully-developed

SUMMARYStructural fire resistance design method came into effect due to the revision of Japan's building code (building standards law of Japan) in June 2001. The method includes standard methods to calculate (1) fire exposure to structural elements, (2) temperature rise of steel and RC elements during fire exposure and (3) structural end points such as ultimate steel temperature for buckling of columns, bending failure of beams and so on. This paper discusses the technical basis for design methods especially focused on steel framed buildings. The calculated values by design equations were compared with experimental values in order to examine the redundancies implied. In the final stage, all the redundancies were combined by Monte-Carlo method and first-order moment method (AFORM). Target safety index and corresponding partial safety factors were discussed.