BS 5959 PT 8
BS 5959 PT 8
BS 5959 PT 8
G M Newman BSc(Eng),CEng,MIStructE,MlFS
ISBN 1 870004 88
4
Extracts from BS 5950: Part 8: 1990 are reproduced with the kind permissionof BSI.
Complete copies of the standard can be obtained by post from BSI Publications,
Linford Wood, Milton Keynes, M K 1 4 6LE.
..
11
CONTENTS
Page
SUMMARY iv
NOTATJON V
1. INTRODUCTION 1
3. METHOD OF DESIGN 3
5. PRACTICAL DETAILS 6
6. DESIGN TABLES 7
7. DESIGN EXAMPLE 17
REFERENCES 24
...
111
SUMMARY
Tests have shown that shelf angle floor beams supporting precast concrete floors havegood
inherent fire resistance. The shielded upper part of the section remains cool as do theupper parts
of the angles that support the precast slabs. Together, these cooler and hence stronger parts of
the section are capable of resisting applied m,oments in fire. Appendix C of BS 5950: Part 8
gives a method of evaluating the fireresistance of shelf angle floor beams. The method is based
on the moment capacity approach in which the plastic moment capacity of the section, including
the angles, is calculated at elevated temperatures.
Included in this report are design tables which give the highest position at which the angles can
be placed on the beam to achieve a certain momentcapacityin fire. This position therefore
relates to the portion of the beam that is exposed. Tables are given for a range of Universal
Beam sizes for grade 430 and 510 steels and for 30 and 60 minutes fire resistance. A design
example is also given.
iv
NOTATION
Width of exposed bottom flange
Temperature gradient
Minimum protected depth of section for a given load and fire resistance
End reaction
Thickness of element
Temperature at location x
V
l.
Shelf angle floor beams supporting precast concrete units are a practical structural option to
reduce construction depths. The precast units, instead of resting on the top flange of the beam,
are supported by angles bolted or welded to the web of the beam. In this way construction depths
can be reduced as the depth of the precast unit is contained within the beam depth. An additional
benefit of shelf angle floor construction is improved performance in fire where the leg of the
angle is embedded in the slab and the upper part of the beam is partially shielded by the concrete
floor. This fact alone will cause a reduction in the amount of fire protection required because
the exposed perimeter of the steel is reduced.
Research carried out by British Steel and SCI(') has shown that inmanyinstances the fire
protection can be totally eliminated by taking account ofthe strength retention of the cooler parts
in comparison to the exposed parts of the section. Fire resistances of 30 and 60 minutes can be
readily achieved depending on the applied load. It is appreciated that the analysis of shelf angle
floor beams in fire canberelativelycomplicatedbecauseof the considerable variation of
temperature through the cross-section. Therefore, emphasis has been placed on the preparation
of simplified design tables for shelf angle floor beams of different depths carrying different loads.
The guidance in this document is based on the calculation method described in BS 5950: Part 8
Code of Practice for5re resistant design(2)and in a Handbook to BS 5950: Part 8 published by
the
1
2.
A general arrangement of a shelf angle floor beam for fire conditions is shown in Figure 1. The
precast units may be standard proprietary units but must have reduced or solid ends in order that
they provide the necessary insulation to the steel. This point is further discussed in Section 4.
For adequate fire resistance it is essentialthat the gap between the web and concrete is filled with
grout or concrete to ensure that an effective heat sink is created around the section.
The design of an unprotected shelf angle floor beam is virtually the same as the traditional design
of a shelf angle floor beam. The only significant difference is the inversion of the angles so that
the short legs point upwards and are thus shielded from direct exposure to fire. The angles are
not normally designed to act structurally with the beam at normal design temperatures although
it is essential that they act with the beam in fire conditions. The angles may be bolted or welded
to the beam.
In fire, the precast units protect the upper part of the section resulting in only moderate rises in
temperature whilst the exposed lower part heats up rapidly and typically can exceed 900C after
60 minutes. The effective section resisting bending comprises the top flange, the upper web and
the vertical legs of the angles. The lower part of the section is only capable of contributing a
small amount to overall bending capacity.
2
3. METHOD OF DESIGN
BS 5950: Part 8 states that the strength of a beam in fire may be considered adequate if the
moment capacity at elevated temperature is greater than the applied moment in fire. The Code
describes a method for calculating the momentcapacitywhichisbased on rectangular stress
blocks comprising the reduced strength of the various elements of the beam in fire. The method
is used to determine the reduced plastic moment capacity of the section in fire as a function of
the temperatures given in the Code. By comparison with the performance in actual fire resistance
tests it can be demonstrated that the moment capacity method is conservative.
SCI and other research organisations have developed alternative mathematical models which are
less conservative than the moment capacity method butstill conservative when compared to tests.
These calculations can only be practically carried out using a computer and are not strictly in
accordance with the Code. The design tables in this publication are based on the method given
in Appendix C of BS 5950: Part 8 and are conservative because of the many beneficial factors
that are ignored.
At elevated temperatures, the strength retention of steel to be used in design is based on 1.5%
strain at all points in the cross-section ( Table l below, also Table 1 ofBS 5950: Part 8).
However, in fire, strains greater than 5 % , are often experienced in bending members. This leads
to a significant increase in strength of the highly strained elements due to strain hardening. The
1.5% strain value is therefore conservative. It may be noted that Eurocode 3: Part l .2(4)will use
a limiting strain of 2% giving slightly higher strengths than in BS 5950: Part 8.
In addition to a moment capacity calculation, BS 5950: Part 8 recommends that the transverse
bending of the angles as well as the connection ofthe angles to the beam are checked. Both these
points are illustrated in the design example in Section 7 of this publication.
Table 1 Strength reduction factors for steel to BS 4360 grades 430 and 510
/%l
STRAIN
%
300 650 600
TEMPERATURE OF STEEL
500 550 400
450 650
O
700
600
I
I 750
I
I 800
0.854
0.5 0.798 0.854 0.492 0.622 0.269
0.721 0.186
0.378 I 0.127 I 0.071 I 1
BS 5950: Part 8 recommends a tabular method for obtaining the temperature distribution across
the section. This methodisbasedontestmeasurementsandis a close fit to the measured
temperatures. The general form of the distribution is shown in Figure 2.
3
Figure 2 Temperature distribution in a shelf angle floor beam
4
4. RECOMMENDATIONS OF BS 5950: PART 8
BS 5950: Part 8 Appendix C recommends the following design conditions:
a) Precast concrete slabs should be made of normal weight concrete and should not have any
deliberately designed voids in the end 75 mm of their length.
b) The voidbetween the precast slab and the beam should be filled with grout.
d) The steel angles shouldbeof grade 510 steel, not less than 125 X 75 X 12 mm,fixed
with the longer legs supporting the concrete slabs, and the vertical leg upwards as shown
in Figure 1 .
0 The moments due to the loadstransmittedvia the slab at the fire limit state, shouldnot
exceed the transverse momentcapacityof the angles at the required periodof fire
resistance ( M q )given by:
where
t = thickness of
angle
leg;
kR = the strength reduction factor from Table l ( also Table 1 in BS 5950: Part 8 )
for 1.5 % strain, for the temperature of the angle at the fire limit state.
g)The angles may hewelded or boltedto the beam.In addition to resisting the applied
vertical loads at the fire limit state, the connection of the angles to the beam should be
capable of transmitting the longitudinal shear force necessary to develop the required
axial forces in the angles at the point of maximum moment. Any weld below the angle
should be ignored. In these calculations the strengths of welds and bolts should be taken
as 80% of the relevantdesignstrengthatelevated temperature, derived using the
appropriate strength reduction factor for steel from Table 1 using a 0.5% strain limit.
The most important of the above points relates to the precast slab. The recommendation that
there should be no deliberately designed voids in the end 75 mm is very onerous and may not be
easy to achieve economically. The purpose of the recommendation is to ensure that the slab acts
as an adequate heat sink to prevent excessive heating of the steel beam. Manufacturers can press
the ends ofeach precast unitwhilst the concrete is stillgreen. This willhave the effect of
closing the circular holes which are normally cast into the units. While this may not strictly meet
the 75 mm recommendation of a) above, it will achieve the intent of providing an effective heat
sink, in conjunction with the grout infill between the precast unit and the steel beam.
5
Tests carried out since the publication of BS 5950: Part 8 have shown that grout is much more
able to absorb heat than the dry sand which was used to fill the area between the beam and the
precast units in the original test programme. Alternatively the holes maybe filled, prior to
erection, with grout but this may prove expensive.
5. PRACTICAL DETAILS
Shelf angle floors may present a problem during construction due to difficulties in locating the
precast units. In order to place the units the angles need to be relatively wide, but to reduce the
transverse bending of the angle the units should be as long as possible. The units should also be
located with approximately equal gaps atboth ends. It is strongly recommendedthat the
recommendations of the manufacturers or suppliers are followed or that any design is carried out
by an engineer experienced in using shelf angle floor construction.
The design information presented here is based on the assumption that the beams are adequately
restrained during construction.
The design tables given do not include anysteelbeams of depths less than 300 mm. It is
considered that beams below this depth cannot practically be used as shelf angle beams.
6
6.
The moment capacity of a shelf angle tloor beam increases as the proportion of the section that
is protected by the concrete floor units increases, i.e. it increases as the angles are placed lower
down the section. For a range of universal beam sizes and load ratios the highest position that
the shelf angle maybeplacedbelow the top of the beam has been computed and is given in
Tables 2, 3, 4 and 5. The position, H, of the angle is measured from the top of the steel beam
to the top surface of the horizontal leg of the angle ( Figure 3 ).
A
I
Maximum height H
v I / ,
I I
\ Connectordesign force
Transverse bending 6,7,8,9. Table6,7,8,9.
capacity,Table 10
JL,
Figure 3 Design factors
Typical load ratios for well designed beams are in the range 0.45 to 0.55. For convenience the
moment capacity at normal temperature is also given. This has been calculated as py S,.
The connection between the angles and the beam should be designed to transmit the horizontal
load given in Tables 6 to 9, as appropriate. (Also, see Section 4).
The angles must be checked for transverse bending. The bending capacity of 12 mm thick angles
is given in Table 10. (Also, see Section 4).
The use of all the design tables is illustrated in the design example.
7
Table 2 Design table for shelf angle floor beams in fire
FireResistance 30 mins
Beam
Grade Fe430
Angle
Grade Fe510
Section size MP
(kNm)
TH or1 of the ant]le below toD of beam for load ratio of:
0.4 0.5 0.55 0.6 0.65 0.7
305 x 102 x 25 93 82 82 82 91 108 123 137
x 28 112 84 84 84 101 118 133 148
x 33 132 86 86 90 108 125 141 156
3 0 5 x 1 2 7 ~37 149 86 86 94 112 124 140 155
x 42 168 87 87 98 116 128 144 160
x 48 194 89 89 104 122 133 150 166
3 0 5 x 1 6 5 ~40 172 85 85 97 115 128 145 161
x 46 199 87 87 100 118 131 148 165
x 54 232 89 89 104 124 136 154 171
3 5 6 x 1 2 7 ~33 148 84 92 113 132 149 163 179
x 39 180 86 97 119 139 157 170 187
356 x 171 x 45 21 3 86 101 123 144 162 176 194
x 51 246 88 103 126 147 167 180 199
x 57 277 90 103 127 149 169 183 202
x 67 333 96 110 134 157 174 192 21 1
4 0 6 x 1 4 0 ~39 198 98 123 145 166 185 200 21 7
x 46 244 99 126 149 171 189 207 225
406 x 178 x 54 288 110 135 152 174 196 21 1 23 1
x 60 328 109 135 150 175 197 213 234
x 67 370 111 138 154 178 20 1 21 7 238
x 74 41 4 115 143 159 184 207 223 244
457x152~ 52 301 1 24 153 178 20 1 221 239 259
x 60 353 121 151 179 204 225 243 264
x 67 396 124 155 183 208 229 247 269
x 74 430 128 159 187 21 2 232 252 274
x 82 477 125 158 187 214 235 255 278
457 x 191 x 67 40 5 130 159 176 202 227 245 266
x 74 456 130 161 178 205 230 248 27 1
x 82 504 134 165 182 210 235 252 27 6
x 89 534 135 166 184 21 2 238 255 279
x 98 59 1 129 162 180 210 237 256 28 1
533x210~ 82 565 161 195 21 5 244 272 298 317
x 92 65 1 163 197 21 9 249 278 303 324
x 101 694 164 199 222 25 1 28 l 306 328
x 109 748 161 198 223 251 282 308 330
x 122 849 154 194 227 2 50 282 31 2 334
610 x229 x 101 793 189 228 261 285 317 347 368
x113 87 1 191 23 1 265 289 322 354 376
x 125 974 181 224 263 285 32 1 355 379
x 140 1099 179 224 265 287 326 363 387
610x305~149 1212 163 21 1 255 280 323 363 366
x 179 1463 157 175 230 280 303 352 372
x 238 1976 106 106 106 166 214 288 315
I I
I Fire Resistance I 30 mins I
I
10
Table 5 Design table for shelf angle floor beams in fire
Angle
Grade Fe510
T l
Section size or I of the iangle below top of beam for load ratio of: I
MP
(kNm) r 0.5 1 0.55 -- 0.6
-_ 0.65 I 0.7
305 x 102 x 25 120 129 144 158 172 184 196 208
x 28 145 137 152 167 180 193 205 217
x 33 170 144 159 174 188 20 1 214 227
3 0 5 x 1 2 7 ~37 192 145 160 174 188 202 209 222
x 42 21 7 1 50 166 181 196 207 21 6 230
x 48 25 1 158 174 190 206 21 2 235 227
3 0 5 x 1 6 5 ~40 222 149 166 174 189 204 224 210
x 46 257 159 175 183 200 208 220 235
x 54 300 171 179 196 21 1 21 8 234 249
3 5 6 x 1 2 7 ~33 192 158 175 191 207 22 1 23 6 248
x 39 232 166 184 20 1 21 8 234 250 257
3 5 6 x 1 7 1 x 45 275 173 192 21 1 21 9 236 259 249
x 51 318 184 204 21 8 23 1 249 272 255
x 57 358 194 214 223 242 255 267 284
x 67 430 21 1 225 241 2 60 267 305 287
4 0 6 x 1 4 0 ~39 256 185 204 223 241 259 284 276
x 46 31 5 196 218 239 258 278 288 304
406 x 178 x 54 372 207 229 251 261 28 1 29 6 307
x 60 424 21 9 243 264 275 296 323 303
x 67 478 230 254 266 287 302 31 5 336
x 74 534 24 l 265 276 299 307 328 3 50
4 5 7 x 1 5 2 ~52 388 224 248 27 1 293 315 344 328
x60 456 238 263 287 310 332 363 343
x 67 51 2 247 273 298 322 336 366 355
x 74 560 257 283 308 333 343 374 366
x 82 621 266 294 319 342 355 387 373
4 5 7 x 1 9 1 x 67 522 248 275 300 31 1 335 366 343
x 74 588 26 1 288 305 326 343 359 382
x 82 651 271 299 312 337 347 395 371
x 89 695 279 308 3 20 347 356 38 1 405
x 98 770 290 313 332 352 369 41 9 394
533 x 210 x 82 730 299 3 30 3 60 374 40 2 41 2 439
x 92 8 40 317 349 366 394 408 462 434
x 101 904 326 359 375 40 5 41 7 474 446
x 109 974 335 368 384 41 3 427 485 457
x 122 1105 349 375 399 41 7 444 503 474
610x229~101 1023 3 57 394 420 446 466 49 1 522
x113 1134 373 41 0 429 464 47 9 51 1 543
x125 1269 388 426 446 47 5 497 563 530
x 140 1430 40 2 433 462 479 514 58 1 548
610x305~149 1577 414 432 472 510 547 ___ --_
x 179
x 238
1905
2572
e H is the highest permitted position of the uppersurfaceof the horizontalleg of the angle,
as measured from the top ofthe beam to the upper face of the angle. Hence the exposed
depth of the section is given by subtracting H from the section depth and adding the angle
thickness.
11
Table 6 Angle connection force
Note:
The table provides the total longitudinal force transferred to each angle along half the span of the
beam. This force is then used to determine the number of connecting bolts or the weld size.
12
Table 7 Angle connection force
Note:
The table provides the total longitudinal force transferred to each angle along half the span of the
beam. This force is then used to determine the number of connecting bolts or the weld size.
13
Table 8 Angle connection force
Note:
The table provides the total longitudinal force transferred to each angle along half the span of the
beam. This force is then used to determine the number of connecting bolts or the weld size.
14
Table 9 Angle connection force
E
I Fire Resistance I 60 mins I
I BeamGrade I Fe510 I
Angle Grade Fe510
Section Size T Load (kN) in each angle to be resisted by bolts or welds for a load ratio of:
0.45
1
0.4 0.5 0.55 0.6 0.65 0.7
305 x 102 x 25 149 162 175 186 197 207 21 7
x 28 185 199 21 2 224 236 246 257
x 33 21 8 233 247 260 261 261 261
305 x 127 x 37 249 261 261 261 261 287 287
x 42 261 261 26 l 26 1 287 287 287
x 48 261 261 26 1 26 1 287 287 318
305 x 165 x 40 261 261 287 287 287 31 8 318
x 46 261 287 287 287 318 318 318
x 54 261 287 287 318 318 318 318
356 x 127 x 33 223 238 252 261 261 261 287
x 39 261 261 261 261 261 261 287
356 x 171 x 45 261 261 261 287 287 318 318
x 51 261 261 287 287 287 318 318
x 57 261 261 287 287 31 8 318 318
x 67 261 287 287 318 318 318 318
406 x 140 x 39 261 26 1 26 1 261 261 26 1 287
x 46 261 26 1 261 26 1 261 287 287
406 x 178 x 54 261 261 261 287 287 318 318
x 60 26 1 26 1 287 287 287 318 318
x 67 261 261 287 287 318 318 318
x 74 261 261 287 287 31 8 31 8 31 8
457 x 152 x 52 261 261 26 1 26 1 261 287 287
x 60 261 261 261 261 287 287 318
x 67 261 261 261 261 287 287 318
x 74 26 1 261 261 261 287 287 318
x 82 261 261 261 287 287 318 318
457 x 191 x 67 261 261 261 287 287 318 318
x 74 261 26 1 287 287 31 8 318 318
x 82 261 26 1 287 287 318 318 318
x 89 261 26 1 287 287 318 318 318
x 98 261 287 287 318 318 318 318
533 x 210 x 82 26 l 261 261 287 287 318 318
x 92 26 1 261 287 287 31 8 31 8 318
x 101 26 1 261 287 287 31 8 31 8 318
x 109 261 261 287 318 318 318 31 8
x 122 261 287 287 31 8 318 318 31 8
510 x 229 x 101 261 26 l 287 287 318 318 318
x 113 261 261 287 287 318 318 318
x 125 26 1 261 287 31 8 318 318 318
x 140 261 287 287 318 318 318 318
510 x 305 x 149 287 318 318 318 318 ... ___
x 179 318 318 318 318 318 ___ ___
x 238 318 318 318 318 ___ ___ ___
Note:
The table provides the total longitudinal force transferred to each angle along half the span of the
beam. This force is then used to determine the number of connecting bolts or the weld size.
15
Table 10 Transverse bending and angle connection design table
Aspect 30 minutes
60
Ratio
DJB, I9.6
Key t o table:
D, Exposeddepth
of
steel
B, Width of bottomflange
M,, Transversebendingstrength(kNmJmetre)
8, Designtemperatureforwelds (OC)
8, Design
temperature for
bolts ( O C )
Ratio
Aspect I of Period Fire Resistance I
30 minutes 60 minutes
02 03 0, (32 83 0,
DJB, I0.6 600 e, - 140
725 350 e, - 90
475
0.6 D,IB, I0.8 e, - 90 510 385 e, - 60 620 745
0.8 < D,IB, I 1.1 e, - 45 e, - 30 425 550640 765
1.I D,IB, I 1.5 e, - 25 550 425 61 640 765
1.5 < D,/B, 61 425 5 5 08 , 640 765
Key to table:
8, Bottom
flange
temperature ('C)
8, Modifiedbottom
flange
temperature ("C)
8, Exposedangle
temperature ( O C )
8, Angle
root
temperature ( O C )
16
7. DESIGN EXAMPLE
Job No. Rev.
l'he
Steel Construction
---
.
--
F
&
I--
- --
-
Job Title
BCF 4880 Sheet 1 of 5
[nstitute -
--
--
-
-
-
- fire Resistance of Sherf Angle n o o r Beams
- Subiect
Mf
-
-
7.22 x (0.8 x 4 + 1.0 x 4.5)
8
= 299.4 kNm
Loadmtio =
299.4 -
- 0.53
566
b..For load m0 of
. 0.53, H = 232 mm (by linear interpolation)
Thus the upper surface of the angles must be positioned not less than 232 mm below
the upper surface of the beam.
17
The
Steel Construction
Institute
---
---
-
-
-
-
-
-
-
.
-
-
p
--
--
-
-
-
--
I Job No.
'
Job Title
Suhiect
BCF 4880
I
I Sheet 2
Fire Resistance of Shelf Angle Floor Beams
Of 5
I Rev.
I
The bending moment will depend on the exact details selected and the assumed
centre of the applied load. For fire, it is reasonable to assume that the load acts
10 mm in from the end of the precast unit. The clearance between the end of the
unit and the upstand of the angle will be approximately 40 mm giving a lever arm of
50 mm.
Reactiodmetre, R
R -
- -
6 x (0.8 x 4 + 1.0 x 4.5)
2
- 23.1 kN/m
M -
- 50 x 23.1 x lo-, kNm/m
- l .155 kNm/m
The bending capacity may be obtained directly from Table 10. However, to
illustrate the derivation of the table the fullcalculation is presented here.
Z -
- t 2/6
-
- l 22 x 1000 mm3/m
-
6
-
- 24000 mm3/m
The temperature of the exposed flange of the angle is obtained from BS 5950:
Part 8, Table 18. This Table is reproduced here as Table 11.
The strength reduction factor is obtained from BS 5950: Part 8, Table l at 1.5%
stmin. This is reproduced here as Table 1.
kR
-
- 0.612
.*.Mcs -
-
1.2 x 355 x 24000 x 0.612 x
6.26 kNmlm
CONNECTION OF ANGLES
The angles may be bolted or welded to the beam. This connection must be able to
tmnsmit the vertical loads and the horizontal load induced by bending. In this
example the vertical load is 23.1 kNlm, as calculated previously. The horizontal
load is not as easily calculated ( by hand ) as it depends upon the position of the
plastic neutml axis in fire. whilst using the moment capacity method to calculate
the angle positions (H) in Tables 2 to 5, the load in each angle was also calculated.
This calculation took account of the neutral axis position and whether it fell within
the depth of the angle. This load is given in Tables 6 to 9 where Table 6
corresponds to design Table 2, Table 7 to Table 3, Table 8 to Table 4 and Table 9 to
Table 5. The temperature at which any weld or bolt must be assessed is given in
Table 10.
19
t I Job No. I I Rev.
The
Steel Construction
---
-S-
- - - ,
-
--
-
---
Job Title
BCF 4880 I Sheet 4 of 5
Fire Resistance of Shelf Angle Floor Beams
I
Institute -
-
-
-
-
--
Subiect
8,
- e, - G.X
OR
- 425 "C Table l 1
G - 2.3"C/mm Table 12 (Table 19, BS 5950: Part 8)
At mid-height
X
-
- 31.5 mm
The average temperature is therefore 352 "C. If it is assumed that the bolts are
positioned at mid-height of the vertical leg will be at a this temperature( 8, ). Bolts
are assumed to lose strength in a manner similar to welds.
Shear
strength - 0.66 x 91.9
-
- 60.6 kN
Total
effective load =
-
[ 4102 + (23.1 X 7.2 / 2)2]'
(per halfspan ) - 419 kN
Thus a minimum of 7 bolts will be required per half fullspan making 14 bolts on
the fullspan at a pitch of about 520 mm.
For welded construction BS 5950: Part 8 states that only weld above the angle may
be used and that any weld below must be discounted. The weld temperature, 8 , , is
taken to be the temperature at the top of the angle leg. A linear variation up the
height of the angle is again assumed.
X
- 63 mm
The strength of weM is assumed to be 80% of the strength of steel based on 0.5%
strain.
20
rhe
Steel Construction
---
---
=
-
=
-
-
--
=
Job Title
BCF 4880 I Sheet 5 of 5
-
--- --
Institute -
-
-
-
-
-- - Subiect
Fire Resistance of Shelf Angle Noor
Beams
jilwood Park Ascot Berks SL5 7QN
Design Example
Date Client by Made
relephone:(0344) 23345
Zax:(0344) 22944
B. S . GMN Feb 1993
Checked by Date
SHEET SALCULATION BKC Feb 1993
Thus at 69% of normal design strength weld must be provided to transmit 410 kN
over half the span (114 kN/m) and a vettical load of 23.1 kN/m.
Strength/m -
- 4
- x 1000 x 215 x x 0.69
h-
- 419 kN/m
21
8. COMPARISON WITH TESTS
Atleast 8 fire resistance tests havebeen carried on shelf angle floor beamsand these are
summarised in Table 13. The tests were onboth grade 430 and 510 universal beams all with
125 X 75 X 12 grade 510 angles. The span in each case was 4.5 m, which is the standard span
used for fire resistance tests. The angle position is measured from the top of the beam to the
upper surface of the horizontal legof the angle. The tests were carried out by BritishSteel
Technical, Swinden Laboratories.
Test
2
0.58
3
1
Section
4 0 6 x 178 x 5 4
305 x 165 x 40
4 0 6 x 178 x 5 42 2
Grade
430
430
510
1 ::: 1 1
Angle
position
(mm)
1 94
Load
0.56
0.23
Fire
resistance
( mins 1
,"j
Ratio
43 I I I I
8 254 x 146 x 430 148 0.55 I 69
It is not possible to make a direct comparison between all the tests and the design tables but a
reasonable comparison can be made for Tests 1, 4 and 8. These tests achieved fire resistances
close to some of the tabulated design data. In Table 14, a comparison of both angle position and
the temperature of the horizontal leg of the angle is shown. The temperature for the comparison
is taken from BS 5950: Part 8, Table 18 which is reproduced on page 16 as Table 11.
1 x 178 x 5 4 0 . 5 64 0 6
I
I
4 1406
I
X 178 x 5 4 ) 0.52
I
~~
8 I 2 5 4 x 1 4 6 x 43 1 0 . 5 5
22
In each case the design method is conservative in that it results in lower angle positions with more
of the beam being shielded from the tire. In Tests 1 and 4 the predicted temperature of the angle,
making allowances for the small time differences, compares well withthe measured temperatures.
In Test 4 the design temperature is somewhat conservative. As stated earlier, more exact
calculation methods exist which would give results closer to those achieved in the tests but these
are very complex and beyond the scope of BS 5950: Part 8.
23
REFERENCES
1. WAINMAN, D.E. and KIRBY, B.R.
Compendium of UK Standard Fire Test Data
Unprotected Structural Steel
British Steel Technical, Swinden Laboratories, 1988
24