etal behaningualeldsles.tion.

theilaringfaceure

lap

ded[1]

tingTheom

beheyand

thethe

CIRP Annals - Manufacturing Technology 67 (2018) 301–304

cesss inerial

laps ofand

IRP.

A new joining by forming process to produce lap joints in metal sheets

João P.M. Pragana a, Carlos M.A. Silva a, Ivo M.F. Bragança b, Luis M. Alves a,Paulo A.F. Martins (2)a,*a Instituto Superior Técnico, Universidade de Lisboa, Portugalb Instituto Superior de Engenharia de Lisboa, Instituto Politécnico de Lisboa, Portugal

1. Introduction

Lap joints are widely used to assemble surfaces from individualsheets or plates partially placed over one another (Fig. 1).

Arc and resistance welding (Fig.1a and b), brazing (or soldering)(Fig. 1c) and friction stir welding (Fig. 1d) are among the most

commonly used welding processes to produce lap joints in msheets. However, its utilization is limited when the sheets tojoined are made from dissimilar materials and is also costlier talternative processes if weld inspections are required. Weldprocesses are additionally constrained by distortion and residstresses arising from the expansion and contraction of the wand adjacent base metals during the heating-cooling cycClamps, jigs and fixtures that lock and hold the sheets in posiduring welding are commonly utilized to minimize distortion

Lap joints produced by adhesive bonding (Fig. 1e) overcomewelding difficulties in joining sheets made from dissimmaterials but its utilization is constrained by temperature, agedue to weathering, long curing times and cautious surpreparation. Clamps, jigs and fixtures are also employed to ensa uniform pressure across the adhesive bonded area of thejoints during curing time.

Mechanical fastening comprises the utilization of threafasteners (or rivets) (Fig. 1f) whereas joining by forming

comprises a wide range of processes such as, self-piercing rive(Fig. 1g), clinching (Fig. 1h) and bending of tabs (Fig. 1i).

fastened joints are easy to assemble and disassemble, are free frthermal after-effects and curing time requirements and canused to connect metallic and non-metallic sheets. However, tare limited by the maximum force that they can safely support

by aesthetic and dimensional requirements.Self-piercing riveting, clinching and bending of tabs avoid

forces to be concentrated at the points of fastening but

A R T I C L E I N F O

Article history:

Available online 1 May 2018

Keywords:JoiningFormingSheet metal

A B S T R A C T

This paper proposes a new joining by forming process to produce lap joints in metal sheets. The procombines partial cutting and bending with mechanical interlocking by sheet-bulk compression of tabthe direction perpendicular to thickness. The lap joints are flat with all the plastically deforming matcontained within the thickness of the two sheets partially placed over one another. The design of thejoints is performed by a simple analytical model and the overall concept is validated by meannumerical modelling and experimentation. Destructive shear tests demonstrate the effectiveness

performance of the new proposed lap joints.© 2018 Published by Elsevier Ltd on behalf of C

Contents lists available at ScienceDirect

CIRP Annals - Manufacturing Technology

journal homepage: http: / /ees.elsevier.com/cirp/default .asp

pen beids.canxa-clic

Fig. 1. Processes for producing lap joints in sheets or plates: (a) arc welding, (b)resistance welding, (c) brazing (or soldering), (d) friction stir welding, (e) adhesivebonding, (f) fastening (or riveting), (g) self-piercing riveting, (h) clinching and (i)bending of tabs.

* Corresponding author.E-mail address: pmartins@tecnico.ulisboa.pt (Paulo A.F. Martins).

https://doi.org/10.1016/j.cirp.2018.04.1210007-8506/© 2018 Published by Elsevier Ltd on behalf of CIRP.

resulting joints are not hermetic upon loading (due to the onature of the joints) and, therefore, are not recommended toused in environments with water, moisture and other fluMoreover, some of the lap joints produced by forming

experience loosening during impact or material stress relation and can also be sensitive to fatigue failure upon cyloading.

Band

prodpermof sh

� thor

� thor

� thta

� thcothlo

Iprocfixintee jis tolap

withaestpoinrive

2. A

Tcuttcomone

by mand

lanc

Tplasundthe

Fig. 2(c) mcross

J.P.M. Pragana et al. / CIRP Annals - Manufacturing Technology 67 (2018) 301–304302

ending of tabs (Fig. 1h) require no additional filler materialsaccessories and is the easiest and most economical way touce lap joints in sheets that will permanently or semi-anently attach to one another [2]. However, there are a coupleortcomings that prevent its widespread utilization:

e thickness of the sheets is usually in the range of 1 to 5 mm inder to facilitate bending of tabs;e sheets must have good ductility and fracture toughness inder to avoid cracking when the bending radius is too small;e elastic spring back of the sheets must be small so that thebs will not recover after being bent;e fixtures should not have aesthetic and dimensionalnstraints because the lap joint surfaces are not flat due toe protrusion of the bent tab beyond the upper (and sometimeswer) sheet surfaces.

n a recent investigation authors proposed a joining by formingess that is capable of avoiding material protrusions wheng two metal sheets perpendicular to one another by means ofoints [3]. Under these circumstances, the purpose of this paper

present a new joining by forming process for producing strongjoints with all the plastically deformed material containedin the thickness of the two sheets. The process avoids thehetic and dimensional limitations that are often negativelyted to fasteners and to material protrusions of self-piercingting, clinching and bending of tabs.

nalytical modelling

he new proposed joining by forming process combines partialing and bending with mechanical interlocking by sheet-bulkpression [4] of tabs from the two sheets partially placed overanother (Fig. 2). In this work, the tabs were cut from the sheetseans of wire electro discharge machining. However, cuttingbending of tabs can be performed faster and cheaper bying in a single step.

neglected. Since the volume of the tabs does not change duringplastic deformation, the cutting length L to be performed on the twosheetscanbewrittenasafunctionof their thicknesses tandclearancec between the bent tabs, as follows, (refer to Fig. 3b for notation),

L ¼ c þ ti þ tj ð1Þ

Eq. (1) is plotted in Fig. 3a as a black solid line with a slope equalto ‘ + 1’ and a vertical-axis intercept equal to the thickness ti + tj ofthe lap joint. Conversely, the tab heights h to be compressed alongthe thickness direction are given by,

hi ¼ hfi þ tj ¼ c þ tj ð2Þ

where hfi is the free tab height of a sheet with thickness ti.

Because, compression of the tab heights hi is limited by plasticinstability (buckling) of the thinner tab (say, ti),

hi ¼ c þ tj < h ðtiÞbuckling ð3Þ

It is possible to write the clearance c between the bent tabs as afunction of the critical height h ðtiÞbuckling to buckle of the thinnertab, as follows,

c < h ðtiÞbuckling � tj ð4Þ

The above equation provides the workability limit due tobuckling and sets the process window to the grey region(assuming, h ðtiÞbuckling > tj) depicted in Fig. 3a. The critical heighth ðtiÞbuckling is to be determined experimentally.

Under these circumstances the design values for the newproposed lap joints are given by the portion of the black solid linethat is located inside the grey region (e.g. point ‘P’ in Fig. 3a).

To conclude, it is worth noting that in case the bent radius r istaken into consideration (Fig. 3c) there will be a shifting of theneutral axis towards the inner bent radius and, therefore, thecorresponding tab height hr

i of a sheet with thickness ti becomeslarger than hi given by Eq. (2). The consequence of this is that thefree tab height hf

i will increase and point P will become closer tothe right-hand side (i.e. closer to the edge of the process window).

3. Experimental and numerical modelling

Fig. 3. Process window: (a) cutting length L of the tabs as a function of the clearancec between the two bent tabs, (b) main notation of the plane strain analytical modeland (c) additional notation when considering the bending radius.

. The new proposed joining by forming process: (a) partial cutting, (b) bending,echanical interlocking by sheet-bulk compression and (d) photograph with a

section detail of the joint.

he analytical model for designing the new lap joints considerstic deformation of the tabs to be homogeneous and isotropicerplanestrain conditions. It isalsoassumedthatthethickness toftabs remains constant and that their bent radius r can be

3.1. Material stress-strain curve

The experiments were performed in aluminium EN AW5754 H111 sheets with 2.5 and 5 mm thickness in the ‘as-supplied’condition. The stress-strain curve of the material was determinedby means of tensile and stack compression tests. The tensile testspecimens were cut out from the supplied sheets at 0, 45, and 90�

with respect to the rolling direction and the stack compression testspecimens were assembled by pilling up discs with 10 mmdiameter that were also cut out from the supplied sheets. The

icaluseion.ed

e ofoss-telyigidearniteulktheimecity. on

g inthe

the.ina-ted

lapom

allolesets.rossd 5

r asring

t the

J.P.M. Pragana et al. / CIRP Annals - Manufacturing Technology 67 (2018) 301–304 303

average stress-strain curve resulting from the entire set of testswas approximated by the following Ludwik–Hollomon’s equation,

s ¼ 325e0:18ðMPaÞ ð5Þ

3.2. Experimental methods and procedures

The experimental work plan involved three different tests. Thefirst type of tests was focused on determining the critical height forthe occurrence of buckling in sheet-bulk compression of tabs in thedirection perpendicular to thickness. This was done by confiningone end of the tab in a die and compressing the other end in orderto force material to plastically deform (Table 1). The tests wereperformed for different values of the tab height-to-width hi=wi andof the tab height-to-thickness hi=ti ratios.

The influence of the bending radius was not taken into accountbecause during compression of the tab heights in the directionperpendicular to the sheet thickness there will be a contact betweenthe outer bending radius of the tab and the adjacent tool surface. As aresult of this, both the inner and outer bending radius become fixedand, therefore, will not influence the overall buckling performance.

The second type of tests was focused on the validation of thenew proposed joining process. The sequence of forming operationsis schematically illustrated in Fig. 2. The tests were performed inunit cells for different sheet thicknesses t, cutting lengths L andclearances c between the bent tabs. Table 2 provides details andshows a schematic representation of the laboratory modular toolthat was utilized in the experiments.

The third type of tests consisted of shear tests for determiningthe maximum force that the new proposed lap joints canwithstand without detachment or failure. Results are compared

consideration. These operative conditions allowed numermodelling of the process to be performed with an in-hocomputer program built upon the finite element flow formulat

The numerical simulation of the forming operations performon the tabs (bending and sheet-bulk compression) made ustwo-dimensional plane strain deformation models. The crsection of the sheets was discretized by means of approxima4000 quadrilateral elements and the tools were modelled as robjects with their geometries discretized by means of lincontact-friction elements. Fig. 4 shows the initial and final fi

element meshes of the mechanical locking by sheet-bcompression in the direction perpendicular to thickness of

two previously bent tabs. The central processing unit (CPU) tfor a typical analysis using convergence criteria for the velofield and residual force equal to 10�3 was approximately 10 mina computer equipped with an Intel i7-5930K CPU processor.

4. Result and discussion

4.1. Two-stage joining by forming process

The first set of tests aimed at identifying the onset of bucklinthe sheet-bulk compression of tabs allowed determining

critical heights hbuckling of Eq. (4) that are needed to establishprocess window of the new proposed joining process (Fig. 5)

Three process conditions with different thicknesses combtions corresponding to test cases ‘A’, ‘B’ and ‘C’ were selec(Fig. 5). As seen in Fig. 5b all the test cases give rise to successfuljoints in good agreement with the process window derived frthe analytical model despite the overall assumption thatplastically deforming material will completely fill the hresulting from partial cutting and bending of tabs in the two she

In fact, both the experimental and finite element predicted csections of the lap joint corresponding to test case ‘A’ (Figs. 4 an(top)) reveal empty regions derived from taking the bent radiuswell as the material that is lost along the cutting perimeter duwire electro discharge machining, into consideration.

Fig. 4. Finite element model of the sheet-bulk compression of the bent tabs ainitial and final instants of deformation.

Table 1Summary of the experiments for determining the occurrence of buckling in thesheet-bulk compression of tabs (adapted from [3]).

hi (mm) 5, 10, 15, 20

wi (mm) 10

ti (mm) 1, 2.5, 5

hi=wi 2, 3, 4

hi=ti 1, 1.5, 2

Table 2Summary of the experiments for producing the new lap joints.

ti (mm) 2.5, 5

L (mm) 7.5, 10, 15

c (mm) 2.5, 5

Fig. 5. Joining two aluminium EN AW 5754 H111 sheets partially placed over oneanother. (a) Process window based on the analytical model and (b) cross sectionalphotographs of the lap joints corresponding to cases A (t1 = t2 = 5 mm), B (t1 = 5 mm,t2 = 2.5 mm) and C (t1 = t2 = 2.5 mm).

against those obtained from alternative lap joints with mechanicalinterlock produced by bending of tabs (Fig. 1i).

3.3. Numerical modelling

Because the experiments were carried out at room temperatureunder a very small displacement rate of the cross head speed of themechanical testing machine, no inertial effects on formingmechanisms are likely to occur and, therefore, no dynamic effectsin the deformation mechanics were needed to be taken into

Tthicpeak

Tsimuanalshowmacheig

Telemmat

Fevolsheeformand

both

4.2.

Tperfsheathatlap jcom

Fbothexpesheaadvawiththic

Fmorby s

Fig. 6case

sheet

J.P.M. Pragana et al. / CIRP Annals - Manufacturing Technology 67 (2018) 301–304304

he contact pressure between the tabs varies with sheetkness but finite element modelling of case A allows estimating

values of approximately 500 MPa.aking the bent radius r into account during finite elementlation means that the theoretical tab heights hi of theytical model need to be replaced by the real tab heights hr

in in Fig. 3c. Taking material lost during wire electro discharge

hining into account is accomplished by reducing the real tabht hri of the finite element models by approximately 5%.he plane strain deformation assumption utilized in the finiteent models reveals adequate because there is no significanterial flow along the width direction of the tabs.ig. 6 shows the experimental and finite element predictedutions of the force with displacement during bending andt-bulk compression stages of the new proposed joining bying process. The maximum compression force is below 30 kNthe enclosed pictures allow understanding material flow in forming stages.

Destructive testing of the joints

he final section of this paper is focused on evaluating theormance of the new proposed lap joint by means of destructiver tests. The objective is to determine the maximum shear force

the lap joints can withstand without detachment or failure. Aoint produced by conventional bending of tabs is included forparison purposes (Fig. 7a and b ).ig. 7c shows the experimental force-displacement curves for

types of joints and a schematic representation of the

5. Conclusions

A new joining by forming process that is capable of producingstrong lap joints with all the plastically deformed materialcontained within the thickness on the two sheets was presented.The process combines cutting and bending with mechanicalinterlocking by sheet-bulk compression of the tabs from the twosheets partially placed over one another.

An analytical model based on plastic deformation andinstability under plane strain deformation proved adequate fordesigning the joints and selecting the major operating variables.The model was successfully validated against experimental andfinite element simulation data.

Destructive testing revealed that the maximum tensile loadthat is safely withstood by the new lap joint is approximately twiceof that of a conventional lap joint produced by bending of tabs.

Acknowledgments

The authors would like to acknowledge the support provided byFundação para a Ciência e a Tecnologia of Portugal and IDMEC underLAETA-UID/EMS/50022/2013 and PDTC/EMS-TEC/0626/2014.

References

. Experimental and finite element predicted force-displacement curves for testA (t1 = t2 = 5 mm, L = 15 mm, c = 5 mm) of Fig. 5 during (a) bending and (b)-bulk compression of a tab.

Fig. 7. Destructive shear tests on lap joints produced (a) by the new proposedjoining by forming process (test case ‘A’ of Fig. 5) and (b) by conventional bending oftabs. The experimental evolution of the force vs. displacement for the two types ofjoints are shown in (c) and a picture of the corresponding joints and tabs afterfailure is provided in (d).

rimental setup that was utilized to perform the destructiver tests. Results show that the new proposed joint has a dualntage of being stronger (10.7 kN vs. 5.4 kN) and perfectly flat, all the plastically deformed material contained within thekness of the two sheets partially placed over one another.ig. 7d shows the lap joints after destructive testing. Thephology of the cracked surfaces reveals triggering of the crackshear followed by opening in tension due to bending.

[1] Mori K, Bay N, Fratini L, Micari F, Tekkaya AE (2013) Joining by Plastic Defor-mation. CIRP Annals - Manufacturing Technology 62:673–694.

[2] Mraz S (2015) Methods for Fastening Sheet Metal without Fasteners. MachineDesign);(August).

[3] Silva CMA, Bragança IMF, Alves LM, Martins PAF (2018) Two-stage Joining ofSheets Perpendicular to One Another by Sheet-bulk Forming. Journal of Materi-als Processing Technology 253:109–120.

[4] Merklein M, Allwood JM, Behrens BA, Brosius A, Hagenah H, Kuzman K, Mori K,Tekkaya AE, Weckenmann A (2012) Bulk Forming of Sheet Metal. CIRP Annals -Manufacturing Technology 61:725–745.