Bulletin of the Transilvania University of BraşovCIBv 2015 • Vol. 8 (57) Special Issue No. 1 - 2015

ASPECTS ON DESIGNING THE TRUSSELEMENTS WELDED JOINTS

D. RADU1 TF. GALATANU1

Abstract: The structural advantages of tubular steel elements have becomeincreasingly obvious in recent decades because of research and experience inconstruction. Worldwide, these elements are frequently used especially atlarge span structures. In the past, the cheap permitted the manufacturing ofthe truss beams made of angle type elements joint in nodes with gussets. With21st century in our country appeared new structural solutions for latticesubassemblies, one of the solutions being the use of RHS (RectangularHollow Section) elements joined with welding directly onto the chords.

The present paper propose the applying of the component method [8] to acommon joint (truss beam), calibrating a finite element model and determinethe improvements possibilities.

Key words: truss beams, welded joints, steel structures

1 Department of Civil Engineering, Transilvania University of Braşov

1. Introduction

An important aspect in the developmentof these types of structural elements wasthe configuration of new design procedurebased on component method which ispresented in Eurocode provisions -EN1993-1-8 [8]. In terms of designing andexecution of steel structures, theimplementation of European normative inour country led to a design adapted to thenew standards complex calculation options(considering multiple failure modes) and tobetter execution control.

Node type plays an important role bothin manufacturing cost as well as theresistance values obtained. Also joiningtechnology greatly influences theperformance of hollow section structures.

In terms of the welding connection type,node elements can be connected directlythrough welded gussets or with welding of

the processed elements at the ends [4]. Themost economical way to combine commonand rectangular hollow section is by directwelding without gusset or other steelplates. This solution also provides the mosteffective way of corrosion protection andtime maintenance. Rectangular hollowsection joints can be easily achievedbecause the cutting only be provided withstraight end cuts. Thus can be achievednodes in T, K, V, X with space betweenelements or with overlapping (total orpartial).

2. EN1993-1-8 Design Procedure

In order to determine the parameterswhich are having an influence on tubularcross section elements behavior, are usedanalytical, numerical or experimentalmodels. The Eurocode adopted model is ananalytic type based on component method.

Bulletin of the Transilvania University of Brasov • Vol. 8 (57) Special Issue No.1 - 2015148

Thus a joint is made of a number ofcomponents depending on joint differentarea load type. In case of truss rectangularor circular hollow cross section elementsjoints, may be considered following failuremodes: chord face failure (plastic failure ofthe chord face) or chord plastification(plastic failure of the chord cross-section),chord side wall failure (or chord webfailure) by yielding, crushing or instability(crippling or buckling of the chord sidewall or chord web) under compressionbrace member, chord shear failure,punching shear failure of a hollow sectionchord wall (crack initiation leading torupture of the brace members from thechord member), brace failure with reducedeffective width (cracking in the welds or inthe brace members), local buckling failureof a brace member or of a hollow sectionchord member at the joint location [1].

Based on experimental tests combinedwith numerical models analysis results,Wardenier et al. [2], in 2008 proposed anew design formulation of the RHS/CHSelements welded joints design. Thusresulted the designing guide – CIDECT[1].

The EN normative [8] is presenting alsothe design procedures for the RHS truss toIPE/HEA chords welded joints, taken intoaccount several failure modes: failure ofthe chords web through yielding under thetruss compression - Figure 1 (a), sharefailure of the chords flange - Figure 1 (b),failure of the truss on an reduced effectivewidth (cracks in the welding or in trusselement) - Figure 1 (c), and local bucklingof the truss (truss wall) - Figure 1 (d).

Although the resistance of a joint withproperly formed welds is generally higherunder tension than under compression, thedesign resistance of the joint is generallybased on the resistance of the brace incompression to avoid the possibleexcessive local deformation or reduced

rotation capacity or deformation capacitywhich might otherwise occur.

(a)

(b)

(c)

(d)

Fig. 1. Failure modes of the joints fortruss beams with RHS/CHS diagonals and

I or H chord profile type.

D. RADU et al.: Aspects on designing the truss elements welded joints 149

The welds connecting the brace membersto the chords should be designed to havesufficient resistance to allow for non-uniform stress-distributions and sufficientdeformation capacity to allow forredistribution of bending moments [4], [8].

Giving the geometry presented in thefigure 2, for the K and/or N gap typejoints, the norm [8] requires a chord webyielding checking with the relation:

50sin Mwwyii btfN (1)

where rthbi

iw 05

sin

Fig. 2. K/N gap type joints geometry

The brace failure need to be taken intoaccount also. The relation

52 MeffiyiiRd ptfN (2)

where yiyfweff fftrtp 072

The chord share failure is consideredchecked by the relation (3).

50

,sin3

Mi

vyRdi

AfN

(3)

The failure of the weld is checked by theproper designing of the welded joint ([1]and [8]).

The local yielding of the truss elementthrough buckling is considered as form thebeam to column joint design

52 Meffiyii btfN (4)

where in case of beff>bi, conservatively istaken also the perimeter of the joint.

3. Case Study – 24m Span Truss BeamJoint

The present paper propose the applyingof the component method [8] to a commonjoints (truss beam), calibrating a finiteelement model and determine theimprovements possibilities.

(a)

(b)

Fig. 3. Designed joints geometry:(a) Lower chord joint (b) Upper chord

joint

Bulletin of the Transilvania University of Brasov • Vol. 8 (57) Special Issue No.1 - 2015150

The designed joints are a 24m span beamtruss welded joint with RHS 120x6,3compressed truss, RHS 90x6,3 tensionedtruss, a HEB140 chord for the lower chordjoint and RHS 120x6,3 , RHS 90x6,3 andHEA200 chord for the upper chord joint.The trusses are made of S235JR steel andthe chords are made of S355 steel grade.

Additional chord web stiffeners areneeded due to the shear force in the area ofthe web.

Following the component methoddesign, for the lower chord, the higheststress ratio is reached in tensionedRHS90x6,3 diagonal = 0.68,corresponding to a stress of 244,78 MPa.The upper chord is having almost the samebehaviour, the values of the stresses beingcomparable to the lower chord. The valueof the stress dimensioned stress is 251,20MPa.

In order to decrease the stress value, it isproposed an additional stiffener in the areaof the centre of the joints and aredistribution of the other two stiffeners,with the aim to diminish the influence ofthe tension diagonal effort/stress.

Considering that according withEurocodes [6], [7] [8], the designimplication of the additional elementcannot be assessed, it was done a FEManalysis for the in case joints.

4. Finite Element Analysis

The general purpose finite elementsoftware Ansys [5] was used for thenumerical investigation. The modellingwas done using solid finite elements with15 and 20 nodes. Meshing was madeautomatically by the software, resultingHEX20 (quadratic hexahedral- 20 nodeshexahedron is very robust, but verymemory intensive) and WED 15(15 nodeslinear wedge-prism) [5].

In the contact between elements in thejoint, was required thickening of the mesh

for accuracy results.After analyzing truss beams-diagonal tolower chord joint it was obtained the stressdistribution. It should be noted that thestress by web of chord was significantincreased (Figure 5).

Fig. 4. Finite element mesh of joints

Fig. 5. Equivalent stress in the lower chordjoint

The value of maximum equivalent (von-Mises) stress it is 253N/mm2 for the lowerchord joint and 262 N/mm2 for the upperchord joint (Figures 5 and 6).

Considering the comparable values ofthe stresses – FEM analysis vs. componentmethod [8], the model was consideredcalibrated.

The next step was to redesign thegeometry of the joint model – applying anadditional stiffener in the centre of thejoint with a reconsidered position of the

D. RADU et al.: Aspects on designing the truss elements welded joints 151

other stiffeners – position correlated withthe walls of the truss diagonals.

Fig. 6. Equivalent stress in the upper chordjoint

By rearrangement of the stiffeners andby adding the new stiffener, it wasobtained an improvement of stressdistribution in the joint.

Fig. 7. Equivalent stress in the lowerchord joint proposal

We can conclude that following therearrangement of the chord stiffeners bythe proposed geometry, the equivalentstress in the joint can be reduced up to 30-35% (Figures 7 and 8). So, the position andnumber of stiffeners is important.

Fig. 8. Equivalent stress in the upper chordjoint proposal

4. Conclusions

The Eurocode [8] is presenting adetailed design procedure for the trussbeams welded joints – the componentmethod. The present article isunderlining that can be achieved betterjoint behaviour by adding additionalelements with geometries that cannot betaken into account by means of presentnormative. The position and the numberof the stiffeners is important, taken intoaccount the differences found outthrough our study. Decreasing the jointstress value can lead to better jointbehaviour and optimized design.

References

1. Design guide for rectangular hollowsection (RHS) joints underpredominantly static loading,CIDECT-Comité International pour leDéveloppement et l’Étude de laConstruction Tubulaire, Jeffrey A.Packer, 2009 ISBN 978-3-938817-04-9.

2. Wardenier, J, Kurobane, Y, Packer,JA, van der Vegte, GJ and Zhao XL,“Design Guide - For Circular Hollow

Bulletin of the Transilvania University of Brasov • Vol. 8 (57) Special Issue No.1 - 2015152

Section (CHS) Joints UnderPredominantly Static Loading” – 2ndEdition, CIDECT, 2008.

3. D. Radu, A. Sedmak.: Failure ModesAnd Designing Procedures Of TheTubular Truss Beams Welded JointsAccording With En 1993-1-8,INTERNATIONAL SCIENTIFICCONFERENCE CIBV 2014.

4. Taus, D. – „The buckling behavior ofdouble T steel sections according theinfluence of welding tehnology” -Bulletin of the Transilvania Universityof Braşov • Vol. 6 (55) Special IssueNo. 1 - 2013, pp. 145-156,Transilvania University PublishingHouse Brasov,

5. ANSYS® Academic Research,Release 15, Help System, ANSYSMechanical User's Guide, ANSYS,Inc.

6. ***SR EN 1090-1+A1:2012:Requirements for conformityassessment for structural components(CE marking). ASRO Romania.

7. ***SR EN 1090-2+A1:2012: Technicalrequirements for the execution of steelstructures; ASRO Romania.

8. ***SR EN 1993-1-8 Proiectareastructurilor din otel – Partea 1-8Proiectarea imbinarilor(Design ofsteel structures- Part 1-8: Design ofjoints), ASRO Romania.