Instruction manuals

5007-5017

Description
TT
Published
of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A novel evaluation method for interfacial adhesion strength in ductiledissimilar materials Shoji Kamiya a, * , Harunori Furuta a , Masaki Omiya b , Hiroshi Shimomura a a Department of Engineering Physics, Electronics and Mechanics, Nagoya Institute of Technology, Gokiso-cho, Showa-ku, Nagoya, 466-8555 Aichi, Japan b Department of Mechanical Engineering, Keio University 3-14-1, Hiyoshi, Kohoku-ku, Yokohama, 247-0006 Kanagawa, Japan a r t i c l e i n f o  Article history: Received 9 January 2007Received in revised form 15 January 2008Accepted 16 June 2008Available online 11 July 2008 Keywords: Flexible printed circuitInterfaceAdhesionStrengthEnergy a b s t r a c t The energy of interface adhesion between two elastic–plastic materials was directly eval-uatedasthemechanicalworksuppliedexclusivelytoseparatetheinterface.Interfacecrackextension was simulated by elastic plastic finite element models, where the nodes alongthe interface in the vicinity of crack tip were divided into two nodes and the nodal forceswere gradually decreased to zero. While further plastic deformation takes place in the vol-ume of materials during crack extension, the work done by these nodal forces againstmutualdisplacementofcracksurfacesshouldbeconsumedonthesurfacesandthusequalstotheinterfaceadhesionenergy. Thistechniquewas appliedtoacopper/polyimide systemfor flexible printed circuits in accordance with the new experimental results. In compari-sontotheresults obtained bytheconventional peel test, this techniqueyieldedfar smalleramount of interface energy successfully excluding the energy dissipated with bulk plasticdeformation without any insertion of cohesive strip along the interface in the model.   2008 Elsevier Ltd. All rights reserved. 1. Introduction Recently, thin films deposited on substrates are used as functional materials in various fields. Among those variety of combinations of films and substrates, demands for soft and flexible film–substrate systems have recently been emergingin many application fields. In electronics industries, such kind of materials systems are already widely used as flexibleprinted circuits and also expected to play important roles to realize flexible electronic devices such as wearable soft equip-ments. Because of the flexible features, however, mechanical reliability of these systems is an issue of great importance anddifficult to evaluate as well. Peel test [1,2] has become a de facto standard method in industries to evaluate the strength of adhesion in flexible film–substrate systems. This technique measures the force required to peel off the film per unit width,thusevaluatethestrengthintermsof N/mwhichisequivalenttotheenergyrequiredtopeel off unitarea, J/m 2 . Forapurelyelastic case, all the earlier works [3–13] identified that the peeling force  P   (N/m) corresponds to the energy used to separatetheinterface,i.e.,interfaceadhesionenergy C (J/m 2 ),withafactorofpeelingangleasdiscussedinthenextsection.However,in cases where plastic deformation takes place during peeling process, the work done by the peeling force is consumed notonly for separating the interface but also for the plastic deformation in peeled arm and substrate. Since the latter is usuallylarger than the former, the results of peeling tests are known to strongly depend on the thickness of peeled-off films.Extensive works [2,14–22] were devoted to the plastic deformation in the peeling process as the films detachedfromthesubstrate, and bend through the moment-curvature hysteresis loop (including plastic loading, elastic unloading, plasticreverse loading, and elastic reverse unloading). For the interfacial strength of materials used in flexible micro-electronic 0013-7944/$ - see front matter   2008 Elsevier Ltd. All rights reserved.doi:10.1016/j.engfracmech.2008.06.011 *  Corresponding author. Tel./fax: +81 052 735 5324. E-mail address:  kamiya.shoji@nitech.ac.jp (S. Kamiya).Engineering Fracture Mechanics 75 (2008) 5007–5017 Contents lists available at ScienceDirect Engineering Fracture Mechanics journal homepage: www.elsevier.com/locate/engfracmech  devices, Park et al. [23–25] measured the interfacial adhesion energy of Cu/Cr/polyimide system by subtracting energy dis-sipatedinplasticdeformationfromtheresultsof peeltest. Althoughtheircompensationtookintoaccounttheeffect ofplas-tic deformationinthe filmonthe basis of fundamental beambendingtheory, the extra plastic deformationin the vicinityof interfacecracktipwasignored.Thereforetheirevaluationresultsstillincludessomeamountofplasticallydissipatedenergy.Acohesivestripmodelinsertedalongtheinterfacegaveanewtwisttotheproblem.TheworksbyTvergaardandHutchinson[26] and Wei and Hutchinson [27] emphasized the influence of cohesive stress on the plastic energy dissipation in film and substrate. Other interface cohesive laws [28–30] elaborated several delicate issues of interface debonding, including the ef-fect of moderatioof crackopeningdisplacement. Asamatterof fact, cohesivemodel figuredout thetrendof peel forcewitha variety of cohesive stress applied to the interface crack surface. However, it did not give a possibility to measure thestrengthofinterfaceindependentofbulkplasticdeformationinthatitisdifficulttodeterminethecohesivestressitselffromthe experimental results.In order to overcome the problem mentioned above, development of a new technique is aimed in this study which en-ables direct evaluation of energy required exclusively to separate the interface between ductile thin films and substrates,i.e., the interface adhesion energy. Contrary to those trials to improve the evaluation within the framework of peel test, anew scheme is examined in this study on the basis of one of the author’s recent trials [31,32] for the evaluation of interfacetoughness in hard coating systems. The paper is composed of six sections as in the following. In Section 2, after the expla-nation of the sample examined in this study, details of the conventional peel tests are described along with the results ob-tained with the sample. The schemes of data compensation to exclude plastic deformation, proposed by Kinloch et al. [20]and Moidu et al. [21], are applied to the results and discussed in detail. In Section 3, the same sample is fabricated into a differentformofspecimenstoperformtheexperimentnewlyproposedinthisstudy.Theexperimentalresultsarecomparedwith the numerical simulation in Section 4. Determination of interface adhesion energy is discussed in detail in Section 5, where independent evaluation of the energy consumed to separate the interface and dissipated by plastic deformationwas confirmed. Section 6 summarizes the paper. 2. Samples and the results of peel test A commercial copper/polyimide laminate produced for printed flexible circuit fabrication was selected as the sampleexaminedinthisstudy.Thesubstratefilmwasa25 l mthickpolyimidefilm,Capton  suppliedbyDUPONT.Thestress–straincurve of this film obtained by uniaxial tensile test was presented in Fig. 1, which was supplied from DUPONT. Nickel was sputteredonthepolyimidefilmfor thefirst6nm, andthenCufor thenext100nm. Finally, Cuwaselectroplatedfortherestof the thickness. Samples with various total thickness of copper layer, from 7 to 22 l m, were subjected to the peel test ex-plained later. Yield stress and Young’s modulus of the copper film with 8.1 l m thickness was obtained by nanoindentation.Nanoindentation tests were performedwith a number of different depths from200 to 1500nmin order to exclude both theeffect of finite tip radius and soft substrate. Yield stress appeared stable with the indentation depth larger than 800nm andwas evaluated to be 500MPa. However, evaluated Young’s modulus varied more significantly from 90GPa to 160GPa anddid not reach an asymptotic value. This may be because elastic deformation was more sensitive to the soft substrate. It mayalso depend on the variation of grain size in depth direction. For this reason, the bulk value of 130GPa [33] was used in thesimulation models explained later. In the literature [34–36], Young’s modulus of electroplated copper filmwas found in therange between 100 and 160GPa. It is mentioned that because of the significant difference in elastic moduli of copper andpolyimide, variation in Young’s modulus of copper film as well as yield stress had little influence on the simulation resultsexplained in Section 4. The calculated amount of energy release rate changed only by roughly ±1% when 100 and 160GPawas used as Young’s modulus instead of 130GPa. Fig. 1.  Stress–strain curve of polyimide film.5008  S. Kamiya et al./Engineering Fracture Mechanics 75 (2008) 5007–5017   Fig. 2 shows a schematic illustration of peel test. For purely elastic case, the interface adhesion energy C is correlated tothe peeling force  P   and the peeling angle  /  in the following equation [3–13] C ¼  P  ð 1  cos / Þ :  ð 1 Þ In case plastic deformation takes place during the peeling off process, Eq. (1) gives the whole sum of interface adhesion en-ergy, elastic strain energy of peeled film and the energy dissipated in plastic deformation. A 90  peel test, where  /  = p /2 inEq.(1),wasperformedwithanumberofdifferentcopperfilmthicknessesaccordingtothestandardofJISC5012(Astandard of 90  peel test for flexible printed circuit, equivalent to IEC 346-2,4,5,6 [1]). In this test, copper films were peeled off in thedirection perpendicular to the substrate as illustrated in Fig. 2. The results are shown in Fig. 3 with circular symbols. Although the interface is expected to be identical because the thickness of the specimen was changed only by changingthethicknessof theelectroplatedsecondcopper layer, thepeel strengthobviouslydepends onthethicknessof copper films.There might be an idea [27] that the results could be extrapolated to zero thickness to exclude the effect of plastic deforma-tion in copper films, which yields roughly 250J/m 2 .ThetheoryproposedbyKinlochetal.[20]andMoiduetal.[21]wereappliedtotheresultsofpeeltestinFig.3tosubtract theamountofenergydissipatedbyplasticdeformation. InKinloch’stheory,thefilmwasmodeledasanelastic–plasticbeamadhered to a rigid substrate and the energy dissipated by plastic deformation during the peel off test was estimated on thebasis of simple beam bending theory. The compensated results were plotted in Fig. 3 with triangular symbols. Moidu et al.improved Kinloch’s model by considering the substrate as an elastic body. The results compensated by Moidu’s model were P FilmSubstrate φ  Fig. 2.  Schematic illustration of peel test. 14001200100080060040020002520151050 Cu thickness ( µ m)  Peel strength Kinloch [20] Moidu [21]       Γ      (   J   /  m    2    ) Fig. 3.  Behavior of peel strength with respect to copper film thickness. S. Kamiya et al./Engineering Fracture Mechanics 75 (2008) 5007–5017   5009  alsoplottedinFig.3withsquaresymbols.However,boththeresultsstilltendtograduallydecreasewithsmallercopperfilmthickness. This fact suggests that there still is remaining plastic deformation which is not yet considered in those theories.When those results are again extrapolated to zero thickness, Kinloch’s theory yielded the interface adhesion energy of approximately100J/m 2 andMoidu’stheoryyielded200J/m 2 . Theenergydissipatedbytheplasticdeformationinthecopperfilm and polyimide substrate in the vicinity of interface crack tip must still be included in these amounts. 3. Experiment  3.1. Specimen preparation Among the samples in the previous section, the laminate with 8.1 l m thick copper film was subjected to a new experi-ment for the evaluation of interface adhesion energy. According to the technique established by Kamiya et al. [31,32], thepolyimidesubstratefilmwas cut intothespecimensof squarebricks inthefollowingprocedures. First of all, thecopper filmsideofthelaminatewasgluedonto2mmthickaluminumplatewithaoneliquidepoxyresinsuppliedbySanyuRecCo.Ltd.The polyimide film exposed on the surface side was cut into about 100 l m wide strips by using two parallel razor bladescombined together. Another set of strips were then cut in the direction perpendicular to the previous cuttings to makesquare bricks, as shown in Fig. 4a. Finally, the brick specimens of polyimide left on the copper film were made by removingtherestofpolyimidefilmsurroundingthebrick,asillustratedinFig.4b.Asshownlaterinthemicrographsofthespecimens, Polyimide brick AlCuPolyimide filmScratch off Polyimide brick  Slit a b Fig. 4.  Specimen of polyimide film in a brick like shape: (a) fabrication process and (b) appearance of a finished specimen. Strain gaugeDiamond needleCantilever beamHorizontal stageAlCuPolyimide brick Optical microscopeVertical stagePolyimide brick Diamond needleCuHorizontal loadVertical load ab Fig. 5.  Experimental setup.5010  S. Kamiya et al./Engineering Fracture Mechanics 75 (2008) 5007–5017 
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks