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

More documents

Recommendations

Info

Since both programmes use the same principles of heat transfer to analyse a cross section of a steel beam, these results are as expected. 4.4.2 Comparison with Firecalc: This programme is a simple heat transfer package with capabilities to analyse protected and unprotected steel. The background information on the Firecalc programme is given in Section 2.3.2. When plotting the standard fire as already in the Firecalc package against the ISO 834 fire used in this report, it was found that the fire time-temperature curve was not the same. On examining the raw data that is input into the Firecalc programme, it was found that the temperatures were that of the increase above ambient temperature with time as for the ISO 834 fire, but that the ambient temperature had not been added on to this increase as the formula requires. This means that the standard fire in Firecalc is always 20 °C lower than the temperatures calculated in the spreadsheet method. This data can be updated and this was done to be able to more accurately compare the data output. The results of a comparison with the spreadsheet method, SAFIR and Firecalc for an unprotected beam of size 530UB82.0 are shown in Figure 4.14. 1200 1000 Spreadsheet Temperature ( o C) 800 600 400 200 SAFIR Firecalc 0 0 50 100 150 200 Time (min) Spreadsheet Firecalc SAFIR with slab Figure 4.14: Comparison between the results from the spreadsheet method, SAFIR and Firecalc for an unprotected 180UB16.1 beam exposed on three sides to ISO 834 fire. 77
When comparing Firecalc and the Spreadsheet results, it appears that the Firecalc programme gives temperatures that are a lot lower which means these are rather non conservative results. When these results are compared with SAFIR, however, the time-temperature curves are very close. The Firecalc manual instructs that the programme is designed for three sided exposure, although when using the programme this information is not provided. The SAFIR simulation is one with three sided exposure with a concrete slab modelled on top of the beam as in Section 4.3.1. The SAFIR results are much closer to the Firecalc results than the Spreadsheet results are because the Firecalc programme models a concrete slab on top of the beam as does SAFIR. The properties of the concrete are listed in the Firecalc programme with density = 2400 kg/m 3 , thermal conductivity = 1.83 W/mK, and specific heat = 960 J/kgK. The thickness or width of the concrete slab are not provided in the instruction manual or the computer ‘help’ file. This package has many uncertainties, and assumptions have to be made concerning the affect of the slab to use the information found from the analysis. This application has been removed from the updated version of the Firecalc menu, Firewind, so it assumed that the programming assumptions and methods have been deemed inaccurate and will therefore seldom be in use in the future. This comparison has been included in this report for completeness. 4.5 CONCLUSIONS: The results in this section show that the empirical equations from the ECCS recommendations give better results than those equations presently found in NZS 3404, when comparisons are made with the temperatures curves from the spreadsheet method and SAFIR. This is definitely the case for four sided exposure of unprotected steel members. For three sided exposure, even though the NZS 3404 equations give a time temperature relationship that is closer to the average temperature with a concrete slab, the ECCS equations give results closer to the maximum temperature over the cross section, which is more likely to cause failure. 78
Page 1 and 2:
Fire Design of Steel Members K R Le
Page 4:
ABSTRACT: The New Zealand Steel Cod
Page 8 and 9:
TABLE OF CONTENTS: ABSTRACT:.......
Page 10 and 11:
5.4 RESULTS FOR FOUR SIDED EXPOSURE
Page 12 and 13:
LIST OF FIGURES: FIGURE 1.1: VARIAT
Page 14:
FIGURE 5.11: TEMPERATURE CONTOUR LI
Page 17 and 18:
designed with confidence. Full scal
Page 19 and 20:
eams and struts when subjected to a
Page 21 and 22:
Gilvery and Dexter, (1997), prepare
Page 23 and 24:
with conditions specified to ensure
Page 25 and 26:
1.6 BACKGROUND INFORMATION: 1.6.1 F
Page 27 and 28:
experienced in fire situations, but
Page 29 and 30:
The thermal conductivity is not imp
Page 31 and 32:
[ The heated perimeter of the steel
Page 33 and 34:
The advantages of board systems are
Page 35 and 36:
The values for the thermal properti
Page 37 and 38:
insulation has on the rise of the s
Page 39 and 40:
( T − T ) k H p f s ∆t ∆T
Page 41 and 42: spreadsheet method very closely, wi
Page 43 and 44: protection, and for user defined fi
Page 45 and 46: This equation originates from the y
Page 47 and 48: 1400 1200 1000 800 600 400 200 0 0.
Page 49 and 50: time. The value of the wall lining
Page 51 and 52: The Eurocode does not define a meth
Page 53 and 54: f f y y ( T ) (20) = 1.0 0 < T < 21
Page 55 and 56: temperature is chosen because it is
Page 57 and 58: The variation of yield stress data
Page 59 and 60: 15 m -1 < H p /A < 275 m -1 in NZS
Page 61 and 62: = H p A (m -1 ) 7.85 The cons
Page 63 and 64: Connections: NZS 3404 requires that
Page 66 and 67: 4 CALCULATION OF STEEL TEMPERATURES
Page 68 and 69: 4.1.2 Analysis Between Methods of T
Page 70 and 71: Figure 4.2 shows the variation of t
Page 72 and 73: Minimum temperatures Maximum temper
Page 74 and 75: For all three sizes of beam, the sp
Page 76 and 77: Temperature ( o C) 1200 1000 800 60
Page 78 and 79: The ECCS equation has a time period
Page 80 and 81: 4.3.1 Results from simulations with
Page 82 and 83: From Figure 4.5 a-c, the formula ra
Page 86 and 87: average temperature of the beam, be
Page 90 and 91: significantly decreased from a cool
Page 94 and 95: Therefore, the ECCS equations provi
Page 96 and 97: 5 CALCULATION OF STEEL TEMPERATURES
Page 98 and 99: calculations made by the spreadshee
Page 100 and 101: susceptible to heating, heat up mor
Page 102 and 103: Figure 5.3 a-c show the results fro
Page 104 and 105: 7850*600*10500 > 2*775*1100*(1869*2
Page 106 and 107: As can be seen from Figure 5.4 a-c,
Page 108 and 109: 310 UB 40.4 beam H p /A = 241 m -1
Page 110 and 111: The SAFIR results fit between the t
Page 112 and 113: constant specific heat being used i
Page 114 and 115: From Figure 5.7 a-c, it appears tha
Page 116 and 117: 800 Temperature ( o C) 600 400 200
Page 118 and 119: To examine the limits of the thickn
Page 122 and 123: Thermal conductivity, k i = 0.19 W/
Page 126 and 127: the slab except to reduce the secti
Page 128 and 129: Figure 5.14 shows the correlation b
Page 130 and 131: 5.8 CONCLUSIONS: The thermal behavi
Page 132 and 133: 6 COMPARISON OF METHODS USING OTHER
Page 136 and 137: 6.2.4 Results for protected steel:
Page 138 and 139: Figures 6.2 b shows the comparison
Page 140 and 141: Figure 6.3 a shows that the average
Page 142 and 143:
The fire that the test beams were e
Page 144 and 145:
7 ADDITIONAL MATERIAL IN EUROCODE 3
Page 146 and 147:
For thermal analysis in Eurocode 3,
Page 148 and 149:
To determine the performance of a s
Page 150 and 151:
The plastic neutral axis is located
Page 152 and 153:
It is therefore recommended that th
Page 154 and 155:
These methods may be used to provid
Page 156:
Eurocode has a large annex with gui
Page 159 and 160:
8.3 CHANGES TO NZS 3404: The follow
Page 161 and 162:
H v height of the ventilation (m) I
Page 163 and 164:
EC3, 1995. Eurocode 3: Design of St
Page 165 and 166:
Temperature Conditions. National Re
Page 168 and 169:
APPENDIX A - SPREADSHEET METHOD: Be
Page 170 and 171:
APPENDIX B - SAFIR: In Figure B.2 i
Page 172 and 173:
*.str file developed by the pre pro
Page 174 and 175:
22 FISO 23 FISO FISO 24 FISO FISO 2
Page 176:
APPENDIX C - FIRECALC Below in Figu
show all

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

Create successful ePaper yourself

Delete template?

Save as template?