1 INTRODUCTION Hollow structural sections of all types are widely used in truss construction due to their attractive ap- pearance, light weight and structural advantages. They are commonly used in onshore and offshore structures e.g. bridges, towers, stadiums, airports, railway stations, offshore platforms etc. For these structures, fire presents one of the most severe de- sign conditions, because the mechanical properties of the steel degrade as the temperature increases. For truss design, both the members and the joints should be checked. Truss member design at ambient temperature is relatively easy, involving mainly de- sign checks for tension and compression resistance after performing static analysis to obtain the member forces. There is abundant amount of literature on the behavior and strength and of truss joints at ambient temperature. Indeed, the CIDECT (2010) design guide and Eurocode EN 1993-1-8 present design equations to calculate the ambient temperature static strength of practically all tubular truss joints. Under fire condition, the current method for truss member design involves calculating the member force using static analysis at ambient temperature and then finding the critical temperatures of the member using the ambient temperature member force. The member force – critical temperature rela- tionship can be evaluated using design methods such as those in BS 5950 Part 8 and EN 1993-1-2. How- ever, the member force obtained from truss static analysis at ambient temperature may not be correct at elevated temperatures due to large deformations of the truss. There has been little study to investigate how truss member forces change at elevated temper- atures and how such changes affect the member crit- ical temperatures. The specific scope of this paper is to investigate whether the member-based fire resistance design ap- proach is safe, and if not, to develop a modified member-based method to take into consideration truss behavior. 2 VALIDATION OF FINITE ELEMENT MODEL For validation, the fire tests of Edwards (2004) and Liu et al. (2010) were simulated and compared with the test results. Figure 1 and Figure 2 show the tested trusses. Failure modes and displacement- temperature curves were compared. Figure 1. Test Girder B of Edwards (2004) (dimensions in mm) 200 2300 335 2835 800 350 1150 1150 1015 3665 400kN TC8 TC7 TC6 TC5 TC11 TC4 TC3 TC2 TC1 TC9 TC10 Figure 2. Test specimen SP1 of Liu et al. (2010) (dimensions in mm) Effects of Truss Behavior on Critical Temperatures of Welded Steel Tubular Truss Members Exposed to Fire E. Ozyurt & Y.C. Wang School of Mechanical, Aerospace and Civil Engineering, University of Manchester, UK ABSTRACT: This paper presents the results of a numerical investigation into the behavior of welded steel tubular truss at elevated temperatures. The purpose is to assess whether the current method of calculating truss member limiting temperature, based on considering each individual truss member and using the member force from ambient temperature analysis, is suitable. Finite Element (FE) simulations were carried out for Circular Hollow Section (CHS) trusses using the commercial Finite Element software ABAQUS. The results of the numerical parametric study indicate that due to truss undergoing large displacements at elevated tem- peratures, some truss members (compression brace members near the truss centre) experience large increases in member forces. Finally, this paper proposes and validates an analytical method to take into consideration the additional compression force due to large truss displacement. This is based on assuming a maximum truss displacement of span over 30. 1125 3375 1125 1125 1 1125 1125 1125 1125 4500 78 kN 78 kN 216 kN 562,5 2 5 6 7 8 13 14 3 4 11 12 15 9 10