FIRE DESIGN OF STEEL MEMBERS - Civil and Natural Resources ...

Figure 4.13 shows that using the same heat transfer techniques in a computer
model gives very similar results. The beam that the simulation of FIRES-T2 was
performed on was a section from the British Steel Sections Product Brochure, and
is close to the size of the BHP 530 UB 82.0 used throughout this report. The two
beams have close to exactly the same dimensions of each other with rounding, and
the largest difference in dimension is the root radius. Both beams have been
subjected to the ISO 834 fire with three sided exposure.
Differences in these two curves could be the result of different sized elements in
the finite element modelling system and nodal points could be located at different
places around the beam cross section. The two curves show the average
temperature of the provided locations of the beam. The averaging system in place
here has been a simple mean of all temperatures at each time step, but a weighted
mean with regards to the area of the beam cross section at that temperature may
provide slightly more accurate or consistent results between the two programmes.
The mechanical properties of steel, concrete and insulation, namely density,
specific heat and thermal conductivity vary with time in both computer
programmes. There are slight differences between the programmes, as to the exact
values of these properties and when the fluctuations occur. These variations can
account for the small deviation between the two curves above. The values of the
emissivity for the radiation component of heat transfer, or the convective heat
transfer co-efficient used in the FIRES-T2 are not known, but it is assumed that
they would not vary too much from the accepted values discussed in Section 2.1.
FIRES-T2 uses a 130 mm slab and this simulation performed in SAFIR also has a
130 mm slab resting on the top flange of the steel section. The concrete is
calcareous concrete with properties as stated in EC2 as simulated in SAFIR. The
properties of the concrete in the FIRES-T2 programme are not known, but it is
assumed the thermal differences between types of concrete would not affect the
average temperature too significantly.
76