Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 172 (2017) 665 – 672

Modern Building Materials, Structures and Techniques, MBMST 2016

Fire Resistance of Unprotected Steel Beams – Comparison between

Fire Tests and Calculation Models
Marek Łukomskia*, Piotr Turkowskia, Paweł Roszkowskia, Bartłomiej Papisa
a
Building Research Institute (ITB), Fire Research Department, 1, Filtrowa Str., 00-611 Warsaw, Poland

Abstract

The paper presents results of fire resistance test of unprotected steel beams, compared with simple and advanced calculation models
given in EN 1993-1-2. The comparison shows differences between temperatures recorded during tests, temperatures calculated in
accordance with EN 1993-1-2 and temperatures calculated in FEM analysis. The resulting loadbearing capacity in fire situation
and criteria used for its assessment in each method is also being discussed.

©
© 2017
2016TheTheAuthors. Published
Authors. by Elsevier
Published Ltd. Ltd.
by Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of MBMST 2016.
Peer-review under responsibility of the organizing committee of MBMST 2016
Keywords: fire resistance, steel structures, steel beams, fire tests, simple calculation model, advanced calculation model.

1. Introduction

Methods for assessing the fire resistance of steel structures are among best reported in the literature and they are
generally focused on analytical methods. Simplified calculation model given in EN 1993-1-1 [1] and EN 1993-1-2 [2]
is easy to use and give a good estimate of the actual fire resistance. This is done either by comparing temperature of
the element with its critical temperature or by direct calculations of element’s loadbearing capacity in fire.
At the same time, fire resistance tests are still often used to assess the fire resistance, applying the criteria of
maximum allowed deflection and rate of deflection. Comparison of both methods, with addition of advanced FEM
analysis, shows that all these models are suitable for fire resistance assessment.

* Corresponding author. Tel.: +485664244; fax: +48228472311.

E-mail address: m.lukomski@itb.pl

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of MBMST 2016
doi:10.1016/j.proeng.2017.02.078
666 Marek Łukomski et al. / Procedia Engineering 172 (2017) 665 – 672

2. Test procedure

Fire tests described further in this paper, where conducted in November 2014 in the Fire Testing Laboratory of
Building Research Institute (ITB) as a part of the EGOLF (European Group of Organisations for Fire Testing,
Inspection and Certification) round robin comparative studies. The same tests were performed in 16 participating fire
testing laboratories.
Tests were performed on two identical HEB 300 hot rolled steel beams made of carbon steel of grade S355,
reinforced by 8 welded stiffeners, in accordance with EN 1365-3 [3]. The real elastic limit has been determined by
tensile tests at an average value of 448 MPa. The stiffeners were provided at the supports and at the load application
points on both sides of the web. Aerated concrete blocks were placed on the upper flange of the beam, as required in
EN 1365-3 [3] in order to simulate the floor and provide three-sided heating.
Total length of each beam was Lspec = 4400 mm, with span between supports Lsup = 4200 mm and length subjected
to heating Lexp = 4000 mm, see Fig. 1.
The load was applied with two hydraulic jacks, spaced 140 cm, each applying force P = 100 kN.
No fire protection material was used to insulate the beams from the heating.
Tests were performed under standard fire heating conditions described in EN 1363-1 [4] and using the limiting
value of deflection Dlim = 147 mm, expressed as (1) and rate of deflection dD/dtlim = 6,53 mm/min, expressed as (2),
where L is the clear span of the test specimen, in millimeters and d is the distance from the extreme fibre of the cold
design compression zone to the extreme fibre of the cold design tension zone of the structural section, in millimeters,
for assessing the fire loadbearing capacity of beams.
View of the tests specimens, prior to and after the test, are given in Fig. 2.

L2
Dlim mm (1)
400d

L2
dD dt lim mm/min (2)
9000d

Fig. 1. Test specimen design and thermocouples locations.

 Marek Łukomski et al. / Procedia Engineering 172 (2017) 665 – 672 667

a b

c d

Fig. 2. (a) view of the test specimen on the exposed side, prior to the test; (b) loading setup; (c) view of the test specimen on the exposed side,
after to the test; (d) beam at the end of test.

3. Steel temperature

3.1. Fire test

The temperature of steel in tests was measured with 11 K-type thermocouples mounted on the web and both flanges,
see Fig. 1. Furnace temperature was measured with 8 plate thermometers, in accordance with EN 1363-1 [4]. Furnace
pressure was set in such a way, as to achieve 20 Pa on the bottom of the test specimen, as required in EN 1363-1 [4].
Average steel temperature was calculated as the arithmetic mean temperature of the lower flange, the web and the
upper flange.

3.2. Simplified calculation model

The formula (3) given in EN 1993-1-2 [2] was used to calculate the temperature of unprotected steel elements,
where ksh is the correction factor for the shadow effect, equal to 0,5625 (calculations were also made with ksh = 1,0),
Am/V is the nominal section factor of unprotected steel beam equal to 96 m-1 for HEB 300, ca is the temperature
dependent specific heat of steel, as described in EN 1993-1-2 [2], hnet,d is the design value of the net heat flux per unit
area, as described in EN 1991-1-2 [5], Δt is the time step of 1 second, ρa is the unit mass of steel equal to 7850 kg/m3.

Am V
'Ta,t ksh hnet ,d 't (3)
ca Ua
668 Marek Łukomski et al. / Procedia Engineering 172 (2017) 665 – 672

3.3. Advanced calculation model

The FEM analysis was performed on 3D beam model in MSC.Marc 2013.1 by MSC.Software. The model
incorporated temperature dependent mechanical and thermal properties of steel, based on EN 1993-1-2 [2].
Thermomechanical couples analysis was performed in 120 steps, each of which responded to 15 seconds of actions.
The heat flux on the exposed surfaces was described by both convection and radiation, using rules given in EN
1991-1-2 [5].

3.4. Results comparison

Measured and calculated temperature of steel is given in Fig. 3 and in Table 1. The graphs shows that in both fire
tests, the average measured steel temperature did not differ by more than 2%. At the same time the temperature
calculated using simplified calculation model, with shadow effect, given in EN 1993-1-2 [2] does not differ by more
than 20% from the values recorded in tests.
The FEM analysis give highest values of temperatures as the model didn’t incorporate the shadow effect. Moreover,
the temperatures in the top flange and in the web, recorded in tests are also lower than the ones calculated in FEM
analysis, for the same reason – no shadow effect.

900
800 EN 1363-1 standard fire
curve
700
Temperature (°C)

Beam "1" (fire test)

600
500
Beam "2" (fire test)
400
300
EN 1993-1-2 simplified
200 with shadow effect
100 EN 1993-1-2 simplified
0 without shadow effect
0 5 10 15 20 25 30 35
MSC.Marc advanced
Time (min) model

Fig. 3. (a) view of the test specimen on the exposed side, prior to the test; (b) loading setup; (c) view of the test specimen on the exposed side,
after to the test; (d) beam at the end of test.

4. Fire resistance classification

4.1. Fire test

The time to reach the values of limiting deflection and limiting rate of deflection was almost the same for both steel
beams – about 33 minutes and 15 seconds for the first criterion and 28 minutes for the second. The testing standard
EN 1363-1 [4] and the fire resistance classification standard EN 13501-2 [6] will therefore give two separate
assessments, as the first standard states the fire loadbearing capacity R is reached as soon as one of the criteria is
exceeded, and the other requires both criteria to be met. The exact classification is given in Table 2.
Results of the fire tests in other participating laboratories are summarized report [7]. The general mean value for
all laboratories, for reaching the limiting deflection criterion was 31:37 minutes, and 27:17 minutes for limiting rate
of deflection criterion.
 Marek Łukomski et al. / Procedia Engineering 172 (2017) 665 – 672 669

Table 1. Steel temperature as measured in tests and calculated with simplified and advanced models.
Fire Steel temperature (°C)
exposure
Measured in fire tests Calculated in MSC.Marc (advanced model) Calculated in
time
accordance with
(min) Top Web Bottom Average Top Web Bottom Average
EN 1993-1-2
flange flange flange flange
(with shadow
effect)
5 112 128 271 196 198 90 198 178 155
10 256 292 520 422 411 187 373 354 305
15 403 451 661 590 567 303 524 516 448
20 532 577 726 691 665 423 631 630 561
25 633 669 760 734 721 530 702 704 645
26 650 683 768 740 730 549 713 716 659
27 666 696 777 746 740 567 723 725 672
28 680 707 788 752 749 585 732 734 683
29 693 716 799 759 758 601 740 742 694
30 705 722 812 766 767 617 748 750 705
31 715 728 822 775 775 631 757 760 716
32 724 733 831 784 783 645 767 769 727
33 730 738 839 795 791 657 776 778 737

Table 2. Fire resistance classification based on fire tests.

Test Time to reach loadbearing capacity criteria Fire resistance classification
number
Limiting deflection Limiting rate of deflection In acc. with EN 13501-2 [6] In acc. with EN 1363-1 [4]
“1” 33 minutes and 15 seconds 28 minutes and 15 seconds R 30 R 20
“2” 33 minutes and 15 seconds 28 minutes and 0 seconds R 30 R 20

4.2. Simplified calculation model

The design plastic moment resistance and plastic shear resistance of the beam, calculated in accordance with
EN 1993-1-1 [1], EN 1993-1-2 [2] and [9], in fire situation at time t = 0 were Mfi,0,Rd = 693 kNm and Vfi,0,Rd = 1260
kN. The actions applied to beam were M = 145 kNm and V = 103 kN. Therefore the utilization factor at t = 0 was μ0 =
0,209 (in bending) and μ0 = 0,082 (in shear).
The calculations for critical temperature of the beam, due to temperature dependent parameters of steel were carried
out iteratively until convergence was achieved and are given in Table 3.

Table 3. Critical temperature of the beam.

θa ky,θ kE,θ k y ,T k E ,T O LT ,T ϕLT,fi χLT,fi Mb,fi,0,Rd μ0 θa,cr
(°C) (-) (-) (-) (-) (-) (-) (kNm) (-) (°C)
20 1,000 1,000 1,000 0,437 0,697 0,806 693,2 0,209 718
718 0,208 0,123 1,300 0,569 0,794 0,742 637,9 0,227 706
706 0,223 0,128 1,320 0,577 0,801 0,738 634,2 0,229 705
705 0,224 0,128 1,323 0,579 0,802 0,737 633,6 0,229 705

Time to reach the critical temperature of the beam θa,cr = 705°C is exactly 30 minutes, see Table 1. Exact calculations
of the fire loadbearing capacity shows that the loadbearing capacity of the beam at time t = 30 minutes is Mb,fi,t,Rd =
141,9 kNm, see Table 4.
670 Marek Łukomski et al. / Procedia Engineering 172 (2017) 665 – 672

Table 4. Critical temperature of the beam.

t θcom = θa ky,θ kE,θ k y ,T k E ,T O LT ,T ϕLT,fi χLT,fi Mb,fi,t,Rd
(min) (°C) (-) (-) (-) (-) (-) (-) (kNm)
0 20 1,000 1,000 1,000 0,437 0,697 0,806 693,2
5 112 1,000 0,988 1,006 0,440 0,699 0,805 692,1
10 256 1,000 0,844 1,089 0,476 0,724 0,788 677,3
15 403 0,993 0,697 1,194 0,522 0,758 0,765 653,5
20 532 0,681 0,507 1,159 0,507 0,746 0,773 452,5
25 633 0,391 0,251 1,248 0,546 0,776 0,754 253,3
26 650 0,350 0,220 1,261 0,552 0,780 0,751 225,9
27 666 0,312 0,191 1,278 0,559 0,786 0,747 200,4
28 680 0,278 0,191 1,206 0,528 0,762 0,763 182,3
29 693 0,247 0,143 1,314 0,575 0,799 0,739 156,9
30 706 0,224 0,128 1,323 0,579 0,802 0,737 141,9
31 715 0,212 0,128 1,287 0,563 0,789 0,745 135,8
32 724 0,201 0,120 1,294 0,566 0,792 0,743 128,5
33 730 0,194 0,118 1,282 0,561 0,788 0,746 124,4

The final fire resistance of the beam, assessed with simplified calculation model in accordance with EN 1993-1-1
[1] and EN 1993-1-2 [2], is 29 minutes and 55 seconds.

4.3. Advanced calculation model

In the FEM analysis, the criteria for fire resistance were adopted from the fire testing procedure and the deformation
criterion was applied. The failure time was not taken as the time when the limiting deflection or rate of deflection, but
as the time of reaching the fracture in steel.
In Fig. 4, the displacement of the beam in numerical model is shown at the time t = 29 minutes, one calculation step
before fracture.

Fig. 4. Displacement of the beam, at failure time t = 29 min.

 Marek Łukomski et al. / Procedia Engineering 172 (2017) 665 – 672 671

4.4. Results comparison

Deformation of the beam, as measured in fire tests and calculated in advanced FEM model is given in Fig. 5
(deflection) and Fig 6 (rate of deflection). Fire resistance of the beam, assessed with deformation criteria (fire tests),
steel strength (advanced FEM model) and temperature criteria (simplified calculations) are summarized in Table 5,
following [8]

160
140
120 Beam "1" (fire test)
Deflection (mm)

100
Beam "2" (fire test)
80
60 MSC.Marc advanced
model
40
EN 1363-1 limiting
20 deflection
0
0 5 10 15 20 25 30 35
Time (min)

Fig. 5. (a) view of the test specimen on the exposed side, prior to the test; (b) loading setup; (c) view of the test specimen on the exposed side,
after to the test; (d) beam at the end of test.

35
Rate of deflection (mm/min)

30
Beam "1" (fire test)
25

20 Beam "2" (fire test)

15
MSC.Marc advanced
10 model
EN 1363-1 limiting
5
rate of deflection
0
0 5 10 15 20 25 30 35
Time (min)

Fig. 6. (a) view of the test specimen on the exposed side, prior to the test; (b) loading setup; (c) view of the test specimen on the exposed side,
after to the test; (d) beam at the end of test.
672 Marek Łukomski et al. / Procedia Engineering 172 (2017) 665 – 672

Table 5. Assessment methods comparison.

Assessment method Time of reach beam failure in fire situation Fire resistance classification
Fire tests with EN 13501-2 [6] classification 33 minutes 15 seconds R 30
Simplified calculation model (EN 1993-1-2) 29 minutes 55 seconds R 20
Advanced FEM analysis 29 minutes 0 seconds R 20

5. Conclusions

All three methods presented in the paper lead to similar results in terms of beam failure time, yet only the fire
resistance tests allowed for R 30 classification where the other two methods allowed only for R 20. Nonetheless,
assessment methods given in Eurocode 3 are very efficient at predicting the fire resistance and they are on the safe
side. Fire tests have also demonstrated the shadow effect, which must be considered in calculations of unprotected
steel elements. Moreover, the top flange was connected to aerated autoclaved concrete blocks, with almost no heat
transfer between these elements. In practice, such beams would have been connected to a reinforced concrete slab,
what would cause bigger temperature distribution within the cross-section of the beam. Such elements would also
have different mechanical properties.

References

[1] EN 1993-1-1:2005. Eurocode 3: Design of steel structures – Part 1-1: General rules and rules for buildings. European Committee for
Standardization, Brussels, Belgium, May 2005.
[2] EN 1993-1-2:2005. Eurocode 3: Design of steel structures – Part 1-2: General rules – Structural fire design. European Committee for
Standardization, Brussels, Belgium, April 2005.
[3] EN 1365-3:1999. Fire resistance tests for loadbearing elements – Part 3: Beams. European Committee for Standardization, Brussels, Belgium,
December 1999.
[4] EN 1363-1:2012. Fire resistance tests. General requirements. European Committee for Standardization, Brussels, Belgium, July 2012.
[5] EN 1991-1-2:2002. Eurocode 1: Actions on structures – Part 1-2: General actions – Actions on structures exposed to fire. European Committee
for Standardization, Brussels, Belgium, November 2002.
[6] EN 13501-2:2007+A1. Fire classification of construction products and building elements – Part 2: Classification using data from fire resistance
tests, excluding ventilation services. European Committee for Standardization, Brussels, Belgium, September 2009.
[7] F. Dumont, L. Boström, M. Łukomski, G. van den Berg, Summary report of the EGOLF round-robin nr. TC2 14-1 in fire resistance testing,
[http://egolf.org.uk/reference-documents/2333/TC2 N766 public summary report RR TC2 14-1 steel beam to EN 1363-1 and EN 1365-3.pdf],
EGOLF, March 2015.
[8] P. Sulik, P. Turkowski, M. Łukomski, Porównanie metod oceny odporności ogniowej konstrukcji stalowych, Materiały Budowlane 11 (2015)
62-64.
[9] P. Turkowski, P. Sulik, Projektowanie konstrukcji stalowych z uwagi na warunki pożarowe według Eurokodu 3, Instytut Techniki Budowlanej,
Warszawa, 2015.