Thin-Walled Structures 49 (2011) 682690

Contents lists available at ScienceDirect

Thin-Walled Structures
journal homepage: www.elsevier.com/locate/tws

Web crippling behaviour of thin-walled lipped channel beams

M. Macdonald a, M.A. Heiyantuduwa Don a, M. Kote"ko b,n, J. Rhodes c
a b c

School of Engineering & Computing, Glasgow Caledonian University, Glasgow, UK Department of Strength of Materials & Structures, Technical University of Lodz, Stefanowskiego 1/15, 90-924 odz, Poland Department of Mechanical Engineering, University of Strathclyde, Glasgow, UK

a r t i c l e in f o
Available online 6 October 2010 Keywords: Web-crippling Load-capacity Code predictions FE analysis Plastic mechanism Experiment

a b s t r a c t
This paper presents the results of an investigation into web crippling behaviourconducted on cold-formed thin-walled steel lipped channel beams subjected to Interior-One-Flange (IOF), Interior-Two-Flange (ITF), End-One-Flange (EOF) and End-Two-Flange (ETF) loading conditions as dened by the American Iron and Steel Institute (AISI). An experimental program was designed to obtain the load-deformation characteristics of beam members with varying cross-sectional and loading parameters under the three web crippling loading conditions. The results obtained from the experiments comprised of the ultimate web crippling strength values and displacements of the thirty-six beam specimens tested. Nonlinear nite element models were developed to simulate web crippling failure of the two loading conditions considered in the experimental program. Also, a combination of elastic analysis with a plastic mechanism approach was employed to investigate the load-deformation characteristics of lipped channel members subjected to the IOF loading condition. The comparison of experimental, nite element and plastic mechanism approach results revealed that the nonlinear nite element models were best capable of closely simulating the web crippling failure behaviour observed in the experiments for all ranges of displacement. Web crippling strength predicted from the Eurocode 3, Part 1.3 [1], and the Polish PN-B-0327 [2] design specications were also compared with the experimental results and the comparisons indicated considerable underestimations for the range of specimens under EOF and ETF loading conditions. & 2010 Elsevier Ltd. All rights reserved.

1. Introduction Web crippling failure may occur at places where thin-walled exural members are subjected to high concentrated loadings or support reaction forces. Fig. 1 illustrates web crippling failure at a loading point. Web crippling deformation is dened as the decrease of cross-section height below the load-bearing plate [3]. Four different loading conditions, where web crippling may take place, have been dened by the AISI based on the number of loadings involved and the location of failure initiated, namely, Interior-One-Flange (IOF), Interior-Two-Flange (ITF), End-OneFlange (EOF) and End-Two-Flange (ETF) loading conditions [4]Fig. 2(a)(d), respectively. A considerable amount of research has been carried out on web crippling over many years by numerous researchers, particularly to validate various design rules for web crippling, and the majority were based on experimental investigations. The early research work conducted by Winter and Pian [5], Ratliff [6], Hetrakul and Yu [7], etc. provided the basis for web crippling design rules that appeared in the early versions of the AISI

Corresponding author. E-mail address: Maria.Kotelko@p.lodz.pl (M. Kote"ko).

Specication and is consequently adopted by the other major design codes including Eurocode 3, Part 1.3. In the recent past, number of investigations were carried out by Young and Hancock, Prabakaran and Schuster, and Shaojie, Yu and LaBoube, and these resulted in a more unied form of design rule which was adopted by the AISI Specication, 2001 edition. Web crippling (crushing behaviour) of hat section beams was investigated by Hofmeyer [3] and Hofmeyer et al. [8], who implemented the yield-line analysis (plastic mechanism analysis) to the investigation of the crushing behaviour of top hat-section beams subject to three-point bendinga similar approach was applied by Bakker and Stark [9]. A research program was initiated to investigate web crippling behaviour of cold-formed thin-walled lipped channel beams under the four loading conditions. The results of the preliminary experimental investigations and the nite element analysis of lipped channel beams under IOF and ITF loading conditions were reported in previous publications [10,11]. The aim of this paper is to present the results of experimental investigations and nite element analysis carried out to investigate the web crippling behaviour of lipped channel sections under IOF, EOF and ETF loading conditions. The experimental results were also compared with the web crippling strength predictions from Eurocode 3-1.3 and with the Polish code PN-B-0327, which adopted the Eurocode

0263-8231/$ - see front matter & 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.tws.2010.09.010

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

683

recommendations. The results for the IOF loading condition were also compared with the analytical solution of the plastic mechanism problem approach [12,13], leading to an evaluation of the unloading path.

2. Experimental investigations Experimental investigations were designed to examine the inuence of various cross-sectional and loading parameters on web crippling strength. Experimental tests for IOF loading conditions (Fig. 2a) were performed by Heiyantuduwa and others and described in detail in [12,14]. Two separate series of tests were performed considering EOF and ETF loading conditions. The test specimens were xed on to load bearing plates during both series of tests to prevent ange rotations and possible lateral movements of specimens during

loading. Each series comprised of eighteen test specimens manufactured from 0.78 mm thickness carbon steel sheets. The test specimens were designed to have three different corner radii and two different web heights and were loaded with three different sizes of load bearing plate. Fig. 3 illustrates the crosssectional and loading parameters used in the specimen design. A separate series of tensile tests was carried out prior to specimen manufacture in order to obtain the material properties of the individual steel sheets. During the web crippling tests, applied load, displacement at the loading point and the displacement at a number of other critical points were measured. Results of the experimental investigations were used to validate the nite element models, to check the validity of web crippling strength predictions obtained from design codes and to compare with theoretical unloading paths obtained from a plastic mechanism analysis. Detailed results of material tests are given in [10,14].

2.1. EOF loading tests EOF loading tests were performed as three-point bending tests; however, the failure was intended to occur at the end of the beam (at supports) and the loading was applied to the mid-point of the beam. The load bearing plate was fully xed at the mid-point in order to prevent failure around this area. The test rig used in the EOF loading tests is shown in Fig. 4.

2.2. ETF loading tests ETF loading tests were performed by applying a load which was directly above the support. Hence, the failure was initiated at the end of the beam due to the heavy loading and the support reaction force. The test rig used in ETF loading tests is shown in Fig. 5.

Fig. 1. Web crippling at loading point.

IOF Loading.

ITF Loading

EOF Loading.

ETF Loading.

Fig. 2. (a) IOF loading; (b) ITF loading; (c) EOF loading; and (d) ETF loading.

684

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

Fig. 3. Cross-sectional and loading parameters.

Fig. 4. Test rig for EOF loading tests.

Fig. 5. Test rig for ETF loading tests.

3. Finite element (FE) models Finite element models were developed to simulate the tests conducted in the experimental investigations. The nite element analysis package ANSYSs was employed for the modelling and analysis procedure. Nonlinear characteristics such as material nonlinearity, geometric nonlinearity and contact situations were considered to accurately represent web crippling failure. Two different nite element models were developed to represent EOF and ETF loading tests described in the experimental investigations.

FE model for IOF loading conditions was developed by Heiyantuduwa and others and described in detail in [12,14].

3.1. FE models for EOF loading condition (EOF-FE models) EOF-FE models were developed to simulate the EOF loading tests carried out in the experimental investigations. The geometric model for the EOF-FE models was similar to the test setup used in the EOF loading tests. However, the advantage of vertical

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

685

Support Block (Solid Elements)

Fig. 6. Finite element mesh for EOF-FE model.

Fig. 7. Element mesh for EOF-FE model: boundary conditions.

with each other. The geometric model for the ETF-FE models was similar to the test setup used in the ETF loading tests. The ETF loading setup was symmetrical about the horizontal plane passing through the centre line of the beam. Therefore, only one-half of the setup was modelled to use the advantage of symmetry. Fig. 8 shows the element mesh generated for the ETF-FE models. Web crippling failure was identied to occur around the central area of the web under the load bearing plates. Thus, the mesh was created to have relatively small elements around the central part of the web and larger, coarser elements further away from the failure region. The loading was applied through a load bearing plate using the displacement control method. Contact elements were used between the load bearing plate and the top ange of the beam. Fig. 9 shows the boundary conditions used in the ETF-FE models.

4. Post-failure behaviourmodelling and test results symmetry was used to create a half-model in this case. The geometry was initially created using the solid modelling techniques within ANSYS. Fig. 6 shows the element mesh generated for the EOF-FE models. In this case, web crippling failure was expected to occur at the support reaction point. Thus, the mesh was controlled to have relatively ner elements closer to the support area and coarser elements further away from the support area. The support reaction force was applied using a support block modelled with solid elements and appropriate boundary conditions were employed to simulate the actual supports used in the experiments. Contact elements were employed in between the support block and the lipped channel beam to represent the actual loading situation. Furthermore, the ange-xed condition was represented using a set of nodes with coupled degrees-offreedom. The loading was applied with displacement control on to a set of nodes selected along the bottom centre line of the support block. The rotation about the Z-axis was restrained along the centre line to represent the actual support conditions in the test setup. Fig. 7 shows the boundary conditions used for the EOF-FE models. A set of nodes around the mid-span of the beam was fully restrained against translations and rotations in all directions. A very competitive and time-saving method (in comparison with FEM), which enables the determination of the ultimate load and the post-failure behaviour of a thin-walled prole, is the compilation of the post-buckling analysis (performed using FS, effective width approach or other analytical method) with the plastic mechanism solution. The advantage of using this approach is that it provides not only a very simple algorithm of the ultimate load determination but also a very quick answer of the question about the post-failure behaviour of the structure. On the other hand, the main drawback may be a relatively high error depending on the level of approximation of the real plastic mechanism of failure. The yield-line analysis leading to the determination of the plastic mechanism of failure was performed by Bakker and Hofmeyer, who used this approach to investigate the web crippling of hat section beams [3,8,9]. On the basis of experimental results obtained for IOF loading conditions [12,14], the theoretical model of the plastic mechanism was developed (Fig. 10a), close to the failure pattern observed in the experiment (Fig. 10b). Post-failure behaviour of beams subject to EOF and ITF loading conditions was substantially different (Fig. 11). Thus, results of theoretical yield-line analysis can be compared with the test results for IOF conditions only. The geometry of W-shaped plastic mechanism and details of the theoretical solution based on the energy method are given in [12]. After performing a numerical derivation of the energy of plastic deformation with respect to the angle of rotation at the global plastic hinge, the load vs. deection curve is obtained.

3.2. FE models for ETF loading condition (ETF-FE models) ETF-FE models were developed to simulate the carried out ETF loading tests. In this case, web crippling failure was expected to occur at the end of the beam under two opposite forces in line

686

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

Load Bearing Plate (Solid Elements)

Beam (Shell Elements Shell 181)

Fig. 8. Finite element mesh for ETF-FE model.

Fig. 9. Element mesh for ETF-FE model: boundary conditions.

Fig. 10. Web crippling post-failure modes: (a) W-shaped theoretical plastic mechanism model and (b) post-failure mode for IOF test specimen.

Fig. 11. Web crippling post-failure modes for: (a) EOF and (b) ITF loading conditions.

An intersection point of the latter with the elastic (pre- or postbuckling) path is generally considered as an upper-bound estimation of the ultimate load.

Comparisons of load vs. deection curves obtained from plastic mechanism analysis, FE simulation and test results for IOF loading conditions are shown in Fig. 12.

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

687

2,000 1,800 1,600 1,400 Load [kN] 1,200 1,000 0,800 0,600 0,400 0,200 0,000 -2,000 0,000 2,000 4,000 6,000 8,000 10,000 Deflection [mm] 12,000 14,000 Experimental FEA Mechanism

4.5 4.0 3.5 3.0

Load (kN)
Collapse Curve - Test 43 FE Curve Test 43

2.5 2.0 1.5 1.0 0.5 0.0 0 10 20 30 Displacement (mm) 40 50

Fig. 12. Comparison of load vs. displacement curve from plastic mechanism approach with experimental and nite element analysis results for a beam member with 100 mm web height, corner radius 1.25 mm and loaded with 100 mm wide bearing plate.

Fig. 14. Comparison of collapse curves obtained from plastic mechanism approach and a nonlinear nite element analysis

5. Eurocode and Polish code (PN) web crippling strength predictions, test results and comparisons
Curve-A Curve-B

Load

Curve-C

K
Displacment of Load
Fig. 13. Typical load vs. deection curves observed in web crippling under IOF loading conditions.

The nominal web crippling strength of thirty-six specimens under EOF and ETF loading conditions was determined using Eurocode 3, Part 1.3, and the equivalent Polish code. It should be highlighted that the Polish Code incorporates the Eurocode recommendations entirely, and that the Eurocode provides individual design equations to predict the nominal web crippling strength of each loading condition. However, these equations can be rearranged in terms of the factors corresponding to the contribution of individual parameters such as corner radii, web height and load bearing length. The local transverse resistance of a web may be determined from: For EOF loading: Rw,Rd k1 k2 k3 9:04hw =t=601 0:01ss =tt 2 fyb =gM1 For ETF loading: 1

Experimental results indicated (Fig. 13) that the loaddisplacement curves for IOF loading conditions can have various shapes depending on the section parameters and the load bearing length. The type of failure shown in Curve-A is exhibited in sections with small corner radii and loaded with wide load bearing plates. In this case, the loaddisplacement curve shows a relatively sharp transition from elastic to plastic failure due to the small top corner and the web is subjected to more direct loading. In the literature, such a failure mode is referred to as a brittle failure. Loaddisplacement curves similar to Curve-C were observed in members with large corner radii and small bearing lengths. In this case, the top corner and the top portion of the web both undergo high stress levels due to the bending produced by the offset of loading. This effect is represented in the gradual change from elastic to plastic collapse through the elasto-plastic transition of the loaddisplacement curves. This failure mode is termed by some researchers as a ductile failure. Curve-B is typical for intermediate values of corner radii and bearing lengths. The theoretical model of W-shaped mechanism coincides with failure mode displayed by sections with small corner radii and loaded with wide bearing plate (Curve-A). Fig. 14 shows the comparison of FE simulation and the failure curve obtained from the plastic mechanism solution for a beam with larger corner radius (2.5 mm) and length of bearing plate (100 mm). In this case the plastic mechanism curve is only capable of representing the post-failure behaviour well beyond the initial collapse stage.

Rw,Rd k1 k2 k3 6:66hw =t=641 0:01ss =tt 2 fyb =gM1

All terms are as dened in the Eurocode 3, Part 1.3. The loaddisplacement graphs obtained from the tests and nite element analysis were used to determine the ultimate web crippling strength of the specimens under EOF and ETF loading conditions. Tables 1 and 2 show the web crippling strength results obtained from the tests (Pexp:ult.), nite element analysis (PFE:ult.) and Eurocode (PEuro) for these loading conditions. Figs. 15 and 16 illustrate sample loaddisplacement graphs obtained from tests and nite element analysis along with the corresponding Eurocode web crippling strength predictions. The nite element strength and the nominal web crippling strength predicted from Eurocode were compared with the experimental web crippling strength results. Table 3 shows the mean and standard deviation of ratios between nite element strength and experimental strength (PFE:ult./Pexp:ult.), as well as Eurocode predictions and experimental strength (PEuro/Pexp:ult.). Figs. 17 and 18 show graphs comparing PEuro/Pexp:ult. with varying lipped channel specimen corner radii for the EOF loading condition, with web depths of 75 and 100 mm, and load bearing plate lengths varying from 25 to 100 mm. Figs. 19 and 20 show graphs comparing PEuro/Pexp:ult. with varying lipped channel specimen corner radii for the ETF loading condition, with web depths of 75 and 100 mm, and load bearing plate lengths varying from 25 to 100 mm.

688

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

Table 1 Web crippling strength results for EOF loading condition. Test No. h (mm) ri (mm) N (mm) t (mm) Span Length Ls (mm) 600 600 600 600 600 600 600 600 600 600 600 600 600 600 600 600 600 600 0.2% Prf stress (MPa) 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 Pexp:ult (kN) PFE:ult. (kN) PEuro (kN)

EOF-1 EOF-2 EOF-3 EOF-4 EOF-5 EOF-6 EOF-7 EOF-8 EOF-9 EOF-10 EOF-11 EOF-12 EOF-13 EOF-14 EOF-15 EOF-16 EOF-17 EOF-18

95.2 95.5 97.3 95.2 95.5 97.3 95.5 97.3 95.2 73.3 70.0 69.2 73.3 70.0 69.2 73.3 70.0 69.2

4.0 2.6 1.2 4.0 2.6 1.2 2.6 1.2 4.0 1.2 2.6 4.0 1.2 2.6 4.0 1.2 2.6 4.0

25 25 25 100 100 100 50 50 50 50 50 50 25 25 25 100 100 100

0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78

1.18 1.24 1.46 1.74 2.00 2.25 1.43 1.70 1.34 1.80 1.44 1.32 1.44 1.12 1.10 2.35 1.90 1.62

1.14 1.19 1.41 1.61 1.92 2.39 1.52 1.93 1.34 2.09 1.57 1.40 1.43 1.23 1.20 2.61 2.17 1.75

0.46 0.80 1.13 0.80 1.38 1.96 0.99 1.41 0.58 1.51 1.07 0.78 1.22 0.86 0.50 2.10 1.49 0.87

Table 2 Web crippling strength results for ETF loading condition. Test No. h (mm) ri (mm) N (mm) t (mm) Span length Ls (mm) 400 400 400 400 400 400 400 400 400 400 400 400 400 400 400 400 400 400 0.2% Prf stress (MPa) 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 220 Pexp:ult (kN) PFE:ult. (kN) PEuro (kN)

ETF-1 ETF-2 ETF-3 ETF-4 ETF-5 ETF-6 ETF-7 ETF-8 ETF-9 ETF-10 ETF-11 ETF-12 ETF-13 ETF-14 ETF-15 ETF-16 ETF-17 ETF-18

73.0 73.4 65.2 98.2 96.2 89.8 73.0 73.4 65.2 98.2 96.2 89.8 73.0 73.4 65.2 98.2 96.2 89.8

1.6 2.4 5.0 1.6 2.4 5.0 1.6 2.4 5.0 1.6 2.4 5.0 1.6 2.4 5.0 1.6 2.4 5.0

25 25 25 25 25 25 50 50 50 50 50 50 100 100 100 100 100 100

0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78 0.78

0.87 0.81 0.76 1.25 0.98 0.80 1.38 0.92 1.14 1.24 1.22 0.98 1.84 1.76 1.58 1.72 1.56 1.28

0.91 0.95 0.82 0.98 0.92 0.81 1.29 1.18 0.95 1.20 1.12 0.96 1.90 1.72 1.34 1.64 1.52 1.32

0.77 0.62 0.17 0.69 0.57 0.16 0.96 0.78 0.21 0.86 0.71 0.19 1.33 1.08 0.23 1.20 0.98 0.27

1.4

Eurocode
1.2 1.0

Experimental

FEA

Load (kN)

0.8 0.6 0.4 0.2 0.0

0.0

5.0

10.0

15.0

20.0

25.0

30.0

Displacement (mm)
Fig. 15. Experimental, FE loaddisplacement graphs compared with Eurocode web crippling strength predictionEOF-1.

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

689

1.2

Experimental
1.0

FEA

Eurocode

0.8

Load (kN)

0.6

0.4

0.2

0.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0

Displacement (mm)
Fig. 16. Experimental and FE loaddisplacement graphs compared with Eurocode web crippling strength predictionETF-5.

Table 3 Summary of comparisons. Mean of strength ratios PFE:ult./Pexp:ult. EOF ETF 1.01 0.98 PEuro/Pexp:ult. 0.66 0.52 Standard deviation of strength ratios PFE:ult./Pexp:ult. 0.13 0.11 PEuro/Pexp:ult. 0.16 0.26

Loading condition

1.2 n=25 mm 1.0 PEuro /Pexp.ult. 0.8 0.6 0.4 0.2 0.0 n=50 mm n=100 mm

1.2 1.0 PEuro /Pexp.ult. 0.8 0.6 0.4 0.2 0.0

0.0 1.0 2.0 3.0 Corner Radii, ri (mm) 4.0 5.0

n=25 mm

n=50 mm

n=100 mm

0.0

1.0

2.0 3.0 4.0 Corner Radii, ri (mm)

5.0

6.0

Fig. 17. Comparison of Eurocode web crippling strength predictions with experimental ultimate strength values for specimens of 75 mm web height under EOF loading condition (tests: EOF-10 to EOF-18).

Fig. 19. Comparison of Eurocode web crippling strength predictions with experimental ultimate strength values for specimens of 75 mm web height under ETF loading condition (tests: ETF-1 to ETF-9).

1.2 n=25 mm 1.0

PEuro /Pexp.ult.

1.2
n=50 mm n=100 mm

n=25 mm 1.0 PEuro / Pexp. ult. 0.8 0.6 0.4 0.2 0.0

n=50 mm

n=100 mm

0.8 0.6 0.4 0.2 0.0 0.0 1.0 2.0 3.0 Corner Radii, ri (mm) 4.0 5.0

0.0

1.0

2.0 3.0 4.0 Corner Radii, ri (mm)

5.0

6.0

Fig. 18. Comparison of Eurocode web crippling strength predictions with experimental ultimate strength values for specimens of 100 mm web height under EOF loading condition (tests: EOF-1 to EOF-9).

Fig. 20. Comparison of Eurocode web crippling strength predictions with experimental ultimate strength values for specimens of 100 mm web height under ETF loading condition (tests: ETF-10 to ETF-18).

690

M. Macdonald et al. / Thin-Walled Structures 49 (2011) 682690

6. Conclusions The results showed that the nonlinear nite elements models developed were capable of closely representing the web crippling failure of the specimens considered in this research. An average deviation of 72% of nite element strength from experimental results was observed. The nominal web crippling strength of thirty-six specimens subject to EOF and ETF loading conditions predicted using Eurocode 3, Part 1.3 (equivalent to the Polish code), were compared with the experimental results. The comparisons indicated averages of 34% and 48% underestimations of Eurocode web crippling strength predictions for the EOF and ETF loading conditions, respectively. It was also observed that the length of the load bearing plate, along with the value of corner radii and web depth, all had an effect on the web crippling strength of the lipped channels testedparticularly noted for the IOF and EOF loading conditions. However, no denite trends could be observed for the ETF loading condition. Unloading paths for beams under IOF loading conditions obtained from the plastic mechanism analysis represented the same character of brittle failure as experimental results for specimens of small corner radii. Quantitative agreement of theoretical and experimental results was rather unsatisfactory. Theoretical failure curves used in this case in compilation with elastic paths to the upper-bound estimation of the beam ultimate load would give an over-estimation of about 40% in comparison with FE and experimental results. One of the reasons for discrepancies was that the analysis did not account for any strain hardening effect. For sections with large corner radii, failure curves obtained from the plastic mechanism analysis were only capable of representing the global collapse mechanism, which normally occurs well beyond the initial collapse stage.

References
[1] ENV 1993-1-3, Eurocode 3: design of steel structures; Part 1.3: supplementary rules for cold formed thin gauge members and sheeting, Brussels, Belgium; February 1996. [2] PN-B-0327: steel structuresstructures of cold formed thin gauge members and sheetingdesign and execution. [3] Hofmeyer H. Cross-section crushing behaviour of hat sections (Part I: analytical modelling). International Journal of Thin-Walled Structures 2005;43:115565. [4] Rhodes J. Design of cold formed steel members. England: Elsevier Applied Science; 1991. [5] Winter G, Pian RHJ. Crushing strength of thin steel webs. Engineering Experiment Station, Cornell University, Bulletin No. 35, Part 1; 1946. [6] Ratliff GD. Interaction of concentrated loads and bending in C-shaped beams. In: Proceedings of the 3rd international specialty conference on cold-formed steel structures, St. Louis, Missouri, U.S.A., November 1975. [7] Hetrakul N, Yu WW. Cold-formed steel I-beams subjected to combined bending and web crippling. In: Proceedings of the international conference on thin-walled structures. Glasgow: University of Strathclyde; April 1979. [8] Hofmeyer H, et al. New prediction model for failure of steel sheeting subject to concentrated load (web crippling and bending moment). International Journal of Thin-Walled Structures 2001;39:77396. [9] Bakker MCM, Stark JWB. Theoretical and experimental research on web crippling of cold-formed exural steel members. International Journal of Thin-Walled Structures 1994;18:261350. [10] Heiyantuduwa MA, Macdonald, M, Rhodes J. Investigation of web crippling behaviour of thin-walled lipped channel beam members. In: Proceedings of the 3rd international conference on structural engineering, mechanics and computation, Cape Town, South Africa; September 2007. p. 37576. [11] Macdonald M, Heiyantuduwa MA, Rhodes, J. Finite element analysis of web crippling behaviour of cold-formed steel exural members. In: Proceedings of the 18th specialty conference on cold-formed steel structures, Orlando, Florida, USA; October 2006. [12] Heiyantuduwa MA, Macdonald M, Kote"ko, M., Rhodes, J. Plastic mechanism approach to web crippling behaviour of cold formed channel section beams. In: Proceedings of the 6th international conference on steel and aluminium structures (ICSAS2007). Oxford Brookes University; July 2007. [13] Heiyantuduwa MA, Macdonald M, Rhodes J, Kote"ko M. Theoretical investigation of web crippling behaviour of thin-walled lipped channel beams. In: Proceedings of the 5th international conference on coupled instabilities in metal structures (CIMS 2008) vol. 1. Australia: University of Sydney; June 2008. [14] Heiyantuduwa MA. Web crippling behaviour of cold formed thin-walled steel lipped channel beams. PhD Thesis. Glasgow Caledonian University; 2008.