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

Equation 4.1 from NZS 3404 is much less conservative as it gives temperatures
that are lower than Equation 4.2 from ECCS, and deviates from the spreadsheet
curve more throughout the temperature range than the ECCS formula does. The
accuracy of the spreadsheet, therefore, is important to show the accuracy of the
equations. From Section 4.2.2, the spreadsheet gives very close results to those
found from the SAFIR programme, which itself has been proven to model the
thermal response of a steel section from comparison with experimental data,
(Gilvery and Dexter 1996).
The largest deviation from the spreadsheet curve occurs when the steel is reaching
the limiting temperature range of between 550 – 800 °C. The equations generally
overestimate the times it takes for these temperatures to be reached which is
unsafe, see Section 4.1.2.
For example, if an unprotected 310 UB 40.4 beam was being considered for the
rate of temperature increase, equation 4.1 from NZS 3404, gives a time of nearly
16 minutes until a temperature of 600 °C is reached, and equation 4.2 from ECCS,
suggests a time of just over 13 minutes. According to the spreadsheet calculations,
however, the temperature will reach 600 °C in around 11.5 minutes. Although
these times are not substantially different, they are non-conservative compared
with the spreadsheet temperature curve. The ECCS formula generally gives times
that are closer to the spreadsheet curve than the New Zealand Code equations are.
The differences between the spreadsheet results and the formulae are relatively
constant for the different sizes of the beams, when the beam sizes are within the
recommended range for the formula. From Figure 4.5 a however, the formulas are
significantly further from the spreadsheet temperature curve than as seen in Figure
4.5 b and c. The section factor for the 180 UB 16.1 beam for four sided exposure
is 334 m -1 which is outside of the section factor range recommended by both
ECCS and the NZS 3404/AS 4100 standards. Both lines from the formulas are a
significant way from the spreadsheet curve in this case, so it appears that the steel
section size limitations of the formulas are valid.
62