150 attendants. 4 short courses.
Proceedings on flash stick 978-1-4799-9950-7/15/$31.00 ©2015 IEEE IEEE catalog CFP15566-CDR
Authors vs sessions
A – AYED : 3 Abaidi : 4 AboRas : 2 Ackermann : 9 Adams : 10 Adli : 13 Adokanou : 11 Ahmar : 11 Albrecht : 18 Altieri-Weimar : 17 Amin-Akhlaghi : 6 Andersson : 7 8 Antretter : 18 Arbess : 3 Ardito : 14 Armin : 13 Arnold : 11 Arwidson : 2 Auersperg : 15 18 Azemard : 12 Azzopardi : 3 11 B – Baccar : 3 Bahrami : 11 Bailey : 6 Balaji : 13 Barink : 21 Barrière : 20 Basrour : 21 Bast : 2 Baum : 13 Bechtol : 3 Becker : 15 Belkhir : 21 Belov : 2 Berkan : 13 20 Bernard : 21 Bertarelli : 14 Bieler : 13 Blaudeck : 6 Bonitz : 6 Bonnet : 11 Bontempi : 4 Breuer : 15 Brinkfeldt : 7 8 Brunner : 18 Brunschwiler : 11 Buhl : 20 Bulashevich : 12 C – Cambieri : 5 Casset : 21 Ceric : 12 Chappaz : 21 Chaudhuri : 14 Chen : 10 12 Chernyakov : 12 Chiang : 8 14 Christian : 20 Christina : 11 13 Cirstea : 21 Colnago : 14 Comi : 14 Corbeij : 11 Corigliano : 14 Courtade : 11 Crécy : 11 D – DUBOIS : 3 DUCHAMP : 3 Dalsjø : 11 Danel : 21 Das : 3 Dasgupta : 3 13 Dauksevicius : 21 Dellaert : 14 Delshadmanesh : 9 Desloges : 21 Desmuliez : 6 Dompierre : 20 Dorum : 11 Doutreloigne : 14 Driel : 11 17 Dubini : 14 Ducharme : 12 Dudek : 9 18 Döring : 9 E – Edwards : 7 8 Ehrhardt : 9 Elata : 1 21 Ernst : 13 Ernstling : 9 Essen : 2 Eyden : 17 F – FREMONT : 3 Fan : 10 12 17 Fanget : 21 Fischer : 11 François : 20 G – Gaidys : 21 Gaiser : 11 Gakkestad : 11 Galles : 16 Galliere : 12 Gessner : 6 Gielen : 10 11 Gonzalez : 8 11 Grams : 5 Grasser : 12 Griffoni : 10 Gromala : 16 20 Guo : 8 Guyenot : 20 Gänser : 18 Günther : 9 H – Haag : 9 Hamon : 17 Hamou : 5 11 Han : 11 16 20 Hartmann : 6 Haupt : 5 Heggen : 6 Heilmann : 11 Hein : 9 Helin : 14 Helland : 5 11 Hermann : 6 Herzberger : 3 Hung : 8 Höfer : 5 Hölck : 6 I – Inal : 11 Iravani : 6 Iwamoto : 0 6 J – Jacobs : 17 Jakub : 16 Jansen : 13 Janssen : 15 21 Janus : 4 Jin : 18 Joffe : 1 Johannes : 11 13 Johannesson : 2 Jung : 7 Jürgen : 11 K – Kabakchiev : 20 Kang : 11 Karpov : 12 Karsten : 2 Kaspar : 13 20 Kaulfersch : 8 Khatibi : 9 Kheirallah : 12 Kiener : 18 Kijkanjanapaiboon : 12 Kim : 11 16 Kimpel : 11 Klingler : 11 Kludt : 9 Koehler : 4 Koffel : 5 Kolchuzhin : 3 Korvink : 3 Kozic : 18 Kozlov : 4 Kraemer : 4 5 Kraetschmer : 11 Kraft : 5 Kretschmer : 12 Kristiansen : 5 11 Kuczynska : 15 Kudryavtsev : 3 Kumar : 11 Kunal : 3 L – LEE : 7 LIU : 19 Lall : 18 19 Lang : 5 Lassnig : 9 Lederer : 9 11 Lee : 11 19 Leisner : 2 Lenczner : 4 21 Leo : 13 20 Li : 2 10 Liao : 8 Locker : 18 Lofrano : 8 Luczack : 2 Lunding : 2 Lürkens : 2 M – MAROT : 3 Machani : 15 Mackenzie : 10 Mandrillon : 11 Maniar : 20 Marin : 11 Martin : 13 Martineau : 11 Maus : 15 Mavinkurve : 21 Mayer : 20 McNally : 1 Mehner : 3 Mehr : 11 17 Melz : 20 Meneghini : 10 Merten : 2 Meszmer : 6 Metais : 20 Metasch : 20 Meuret : 11 Michel : 15 Middendorf : 5 Mike : 2 Minixhofer : 5 Montmitonnet : 11 Moreau : 11 Morgan : 16 Moujbani : 12 Muecklich : 4 Mukherjee : 13 Myllykoski : 2 Métais : 20 N – Nabi : 15 Natalie : 11 Neumaier : 7 8 Nguyen : 21 O – Oprins : 11 Otto : 2 8 Ottosson : 8 O’Reilly : 21 P – Palczynska : 16 20 Panpan : 20 Pantou : 11 18 Papaioannou : 21 Papathanassiou : 18 Park : 11 Patel : 6 Paulasto-Kröckel : 2 Pauly : 4 Pfost : 15 Pham : 15 Pichler : 5 Plas : 8 Poder : 2 Poshtan : 16 Przemyslaw : 13 20 Q – Qian : 17 R – Ratier : 21 Rencz : 19 Rettenmeier : 20 Rivlin : 1 Rochus : 14 Rodriguez : 3 Roellig : 20 Roessle : 11 Rottenberg : 14 Rovitto : 12 Rubin : 21 Rzepka : 2 9 15 16 18 Röllig : 20 Römer : 12 S – Sabath : 12 Sakalaukus : 19 Salahouelhadj : 11 Sasi : 16 Schafet : 15 Scheible : 7 Schindler-Saefkow : 11 Schlottig : 11 Schmadlak : 16 Schmitz : 5 Schuld : 17 Schulz : 6 Schweitzer : 15 Schwert : 20 Schöngrundner : 18 Seifikar : 21 Severi : 14 Shaporin : 6 Sheremet : 3 Shmulevich : 1 Silber : 7 16 Soestbergen : 15 Spraul : 15 Stefan : 20 Steller : 2 Stiebing : 2 Sturm : 6 Su : 8 14 Sun : 17 Suresh : 13 Swartjes : 15 Sylvestre : 1 T – TAO : 7 Taklo : 5 11 Tetzlaff : 4 Teyssieux : 4 Theolier : 3 Thomas : 13 Todri-Sanial : 12 Tonry : 6 Treml : 18 Tsai : 8 V – Vandevelde : 10 Vardøy : 5 Vogel : 2 Vu : 15 W – Wachmann : 5 Wang : 14 18 Wei : 19 Weide-Zaage : 9 12 Weiss : 15 Werner : 18 Wiejak : 7 Wiese : 4 5 11 Wilde : 11 Willems : 10 Witkowski : 4 Wittler : 5 Woirgard : 3 11 Wolf : 2 Wolter : 20 Wright : 5 Wu : 16 Wunderle : 2 6 11 13 16 Wyart : 20 Wymyslowski : 7 Y – Yadav : 18 Yadur : 16 Yang : 4 14 21 Yeung : 16 Youssef : 11 Yu : 6 Yuan : 17 Yuile : 4 Z – Zaal : 15 Zade : 3 Zahn : 3 Zakgeim : 12 Zamkotsian : 21 Zanon : 10 Zarbakhsh : 6 11 Zechner : 18 Zega : 14 Zerbini : 14 Zhang : 11 17 18 Zhou : 12 13 Zisser : 12 Zschenderlein : 11 13 Zschieschang : 7 8 Zukowski : 11 c – commissaris : 17 Å – Åklint : 7
|
Authors find your schedule
- A. Yuile: Monday April 20 2015, session 4 at 14:30 (#2)
- A.E. Chernyakov: Tuesday April 21 2015, session 12 at 10:40 (poster)
- A.R. Rezaie Adli: Tuesday April 21 2015, session 13 at 14:20 (#3)
- A.W.J. Gielen: Tuesday April 21 2015, session 10 at 09:00 (#1)
- Abdellah Salahouelhadj: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Alan Wright: Monday April 20 2015, session 5 at 17:10 (#3)
- Alexander Kozlov: Monday April 20 2015, session 4 at 15:10 (#4)
- Alicja Palczynska: Wednesday April 22 2015, session 20 at 12:30 (#5)
- Arian Grams: Monday April 20 2015, session 5 at 16:50 (#2)
- Arun Sasi: Tuesday April 21 2015, session 16 at 16:50 (#4)
- Aymen Moujbani: Tuesday April 21 2015, session 12 at 10:40 (poster)
- Bart Vandevelde: Tuesday April 21 2015, session 10 at 09:30 (#2)
- Ben Rivlin: Monday April 20 2015, session 1 at 11:00 (#1)
- Benjamin Métais: Wednesday April 22 2015, session 20 at 11:50 (#3)
- Benoît Dompierre: Wednesday April 22 2015, session 20 at 11:30 (#2)
- Bo Sun: Tuesday April 21 2015, session 17 at 16:50 (#4)
- Bulong Wu: Tuesday April 21 2015, session 16 at 15:40 (#1)
- C. Tonry: Monday April 20 2015, session 6 at 16:50 (#2)
- Carl Christoph Jung: Monday April 20 2015, session 7 at 17:30 (#4)
- Changwoon Han: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Chia-Chi Tsai: Tuesday April 21 2015, session 8 at 09:30 (#2)
- Claudia Comi: Tuesday April 21 2015, session 14 at 14:00 (#2)
- Darjan Kozic: Tuesday April 21 2015, session 18 at 16:10 (#2)
- Dr. Martin Lederer: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Dries Dellaert: Tuesday April 21 2015, session 14 at 13:30 (#1)
- Duy Duc Nguyen: Wednesday April 22 2015, session 21 at 11:00 (#1)
- E. Merten: Monday April 20 2015, session 2 at 14:30 (#2)
- Elena Zukowski: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Emad A. Poshtan: Tuesday April 21 2015, session 16 at 16:30 (#3)
- F. Casset: Wednesday April 22 2015, session 21 at 11:30 (#2)
- F. Kraemer: Monday April 20 2015, session 4 at 14:50 (#3)
- Fedia Baccar: Monday April 20 2015, session 3 at 15:10 (#4)
- Geneviève DUCHAMP: Monday April 20 2015, session 3 at 14:30 (#2)
- Gimi Pham: Tuesday April 21 2015, session 15 at 14:20 (#3)
- Gromala Przemyslaw: Tuesday April 21 2015, session 13 at 14:40 (#4)
- Hasan Sadat Nabi: Tuesday April 21 2015, session 15 at 14:40 (#4)
- Hung-Te Yang: Tuesday April 21 2015, session 14 at 15:00 (#5)
- Hyun Jin Kang: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Ilja Belov: Monday April 20 2015, session 2 at 15:10 (#4)
- Ilko Schmadlak: Tuesday April 21 2015, session 16 at 16:10 (#2)
- J. Auersperg: Tuesday April 21 2015, session 15 at 13:30 (#1)
- Jaemi L. Herzberger: Monday April 20 2015, session 3 at 14:00 (#1)
- Jan Albrecht: Tuesday April 21 2015, session 18 at 16:30 (#3)
- Javad Zarbakhsh: Monday April 20 2015, session 6 at 17:30 (#4)
- Jens Heilmann: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Jeremy Adams: Tuesday April 21 2015, session 10 at 09:50 (#3)
- Joseph Al Ahmar: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Jue Li: Monday April 20 2015, session 2 at 14:50 (#3)
- Julien Sylvestre: Monday April 20 2015, session 1 at 11:30 (#2)
- Kasemsak Kijkanjanapaiboon: Tuesday April 21 2015, session 12 at 10:40 (poster)
- Kirsten Weide-Zaage: Tuesday April 21 2015, session 9 at 09:00 (#1)
- Klas Brinkfeldt: Monday April 20 2015, session 7 at 17:10 (#3)
- Klas Brinkfeldt: Tuesday April 21 2015, session 8 at 09:00 (#1)
- Kokou Adokanou: Tuesday April 21 2015, session 11 at 10:40 (poster)
- M. Barink: Wednesday April 22 2015, session 21 at 11:50 (#3)
- M. Kudryavtsev: Monday April 20 2015, session 3 at 14:50 (#3)
- M. van Soestbergen: Tuesday April 21 2015, session 15 at 14:00 (#2)
- M. Yazdan Mehr: Tuesday April 21 2015, session 11 at 10:40 (poster)
- M. Yazdan Mehr: Tuesday April 21 2015, session 17 at 16:10 (#2)
- Marco Rovitto: Tuesday April 21 2015, session 12 at 10:40 (poster)
- Marta Kuczynska: Tuesday April 21 2015, session 15 at 15:00 (#5)
- Martin Lederer: Tuesday April 21 2015, session 9 at 09:50 (#3)
- Martin Stiebing: Monday April 20 2015, session 2 at 14:00 (#1)
- Meier, Karsten: Monday April 20 2015, session 2 at 15:30 (#5)
- Melina Lofrano: Tuesday April 21 2015, session 8 at 09:50 (#3)
- Mian TAO: Monday April 20 2015, session 7 at 16:50 (#2)
- Michel Lenczner: Monday April 20 2015, session 4 at 14:00 (#1)
- Nancy Iwamoto: Monday April 20 2015, session 6 at 17:10 (#3)
- Oeztuerk Berkan: Wednesday April 22 2015, session 20 at 11:00 (#1)
- Patrick Gaiser: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Patrick J. McNally: Monday April 20 2015, session 1 at 12:00 (#3)
- Pradeep Lall: Tuesday April 21 2015, session 18 at 15:40 (#1)
- Pradeep Lall: Wednesday April 22 2015, session 19 at 09:30 (#3)
- Prof Marta Rencz: Wednesday April 22 2015, session 19 at 10:00 (#4)
- R. Dudek: Tuesday April 21 2015, session 9 at 09:30 (#2)
- R. Fayҫal Hamou: Monday April 20 2015, session 5 at 17:30 (#4)
- R. Fayҫal Hamou: Tuesday April 21 2015, session 11 at 10:40 (poster)
- R. Metasch: Wednesday April 22 2015, session 20 at 12:10 (#4)
- Raffaele Ardito: Tuesday April 21 2015, session 14 at 14:20 (#3)
- Ricky Lee: Wednesday April 22 2015, session 19 at 09:00 (#2)
- Rida Kheirallah: Tuesday April 21 2015, session 12 at 10:40 (poster)
- Rolanas Dauksevicius: Wednesday April 22 2015, session 21 at 12:30 (#5)
- Rongsi Wang: Tuesday April 21 2015, session 18 at 16:50 (#4)
- Schicker Johannes: Tuesday April 21 2015, session 13 at 14:00 (#2)
- Sridhar Ganesh Kumar: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Steffen Hartmann: Monday April 20 2015, session 6 at 16:20 (#1)
- Steffen Wiese: Monday April 20 2015, session 5 at 16:20 (#1)
- Stéphane Moreau: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Subhasis Mukherjee: Tuesday April 21 2015, session 13 at 13:30 (#1)
- Tal Rubin: Wednesday April 22 2015, session 21 at 12:10 (#4)
- Thomas Tetzlaff: Monday April 20 2015, session 4 at 15:30 (#5)
- Toni Youssef: Tuesday April 21 2015, session 11 at 10:40 (poster)
- Uwe Zschenderlein: Tuesday April 21 2015, session 13 at 15:00 (#5)
- Veronique Rochus: Tuesday April 21 2015, session 14 at 14:40 (#4)
- Vladimir Kolchuzhin: Monday April 20 2015, session 3 at 15:30 (#5)
- Vollert, Natalie: Tuesday April 21 2015, session 11 at 10:40 (poster)
- W.D. van Driel: Tuesday April 21 2015, session 17 at 16:30 (#3)
- Waldemar Wiejak: Monday April 20 2015, session 7 at 16:20 (#1)
- Wenbo Yuan: Tuesday April 21 2015, session 17 at 15:40 (#1)
- Zhuangjian LIU: Wednesday April 22 2015, session 19 at 08:30 (#1)
|
Session 2 |
Reliability Investigations |
14:00 |
Monday April 20 2015 |
Chaired by Zoltán Hajnal, Ricky Lee |
14:00 |
30mn |
Keynote presentation – Challenges in the Reliability of 3D Integration using TSVs Martin Stiebing 1, Dietmar Vogel 2, Wolfram Steller 3, M. Jürgen Wolf 3, Bernhard Wunderle 1 1 Technische Universität Chemnitz, Germany 2 Fraunhofer ENAS, Micromaterials Center MMC, Germany 3 Fraunhofer IZM, ASSID, Germany abstract One of the possible target applications of 3Dintegrationis the high performance computing (HPC).The improved performance of 3D-integrating can baseon an Interposer with heterogeneous integration ofspecial high performance fluidic cooling, individualpower supply next to the microprocessor and a tailoredpackaging approach. This paper focuses on the reliabilityassessment of the interposer regarding the interactionof TSV, wiring and fluid channel integration based oncurrent publications in correlation with the selectedspecifications for Interposer manufacturing. A furtherpart is the correlation with initial results of stresssimulation and measurements applied to the Interposerdesign specifications of cavities and interconnect features(used for sealing the cooling channels and as electricalinterconnect). We give an overview of a new level ofcomplexity of 3D integration and discuss certain failuremechanisms that can influence the performance of 3Dintegrated devices. We show that the prediction of failurelocations becomes possible by the method of finiteelement modeling but with the necessity to obtain processspecific material data as an input for the simulation.An overview shows the possibilites of residual stressesanalyses on test vehicles by using special methods ofRaman spectroscopy and fibDAC including interpretationchallenges of gathered measurement data. Furthermorethe application of a new method for fatigue testing isproposed and discussed. |
14:30 |
20mn |
Reliability Investigation and Design of High Power Inverter Modules based on Material Characterization Simulation and experimental Verification. E. Merten 1, T. von Essen 1, F. Luczack 2, A. Otto 2, A. Lunding 3, P. Lürkens 4, M. Bast 5, M. AboRas 1, S. Rzepka 2 1 Berliner Nanotest u. Design GmbH, Berlin, Germany 2 Fraunhofer ENAS, Micro Materials Center, Chemnitz, Germany 3 Philips Medical Systems DMC GmbH, Hamburg, Germany 4 Philips Research Europe, Eindhoven, Netherlands 5 FuE-Zentrum FH Kiel GmbH, Kiel, Germany abstract A set of well suited methods is presented, which allow the quick acquisition of thermal and thermomechanicalrelevant data of electronics modules and components for the creation of profound simulation models within a virtual prototyping process. |
14:50 |
20mn |
Study on Thermomechanical Reliability of IGBT Power Module and Thermal Grease Pump-out Mechanism Jue Li 1, Pirkka Myllykoski 2, Mervi Paulasto-Kröckel 1 1 Aalto University, School of Electrical Engineering, Espoo, Finland 2 ABB Drives, Helsinki, Finland abstract This paper focuses on two major thermomechanical reliability topics related to high power devices such as insulated gate bipolar transistor (IGBT) modules. Firstly, the stress-free status and the thermal residual stress of a typical power module are investigated by finite element method (FEM) analysis. After determining the thermal residual stresses at room temperature, thermal cycling (TC) and power cycling (PC) tests are conducted and simulated by FEM. Secondly, the thermal grease pump-out phenomenon is explicitly simulated via a combined FEM and smoothed particle hydrodynamic (SPH) method for the first time. SPH method shows great potential for the thermal grease material selection and engineering, base plate optimization, and thermomechanical reliability optimization of power devices in general. The simulated contact opening results at the interface between copper base plate and heat sink indicates different pumping modes associated with different loading conditions. All modeling approaches presented in this work offer insight into our understanding of deformation, stress status, and thermal grease related failure mechanisms of high power devices. |
15:10 |
20mn |
The effect of variations in temperature cycling profile and mechanical properties of solder on thermo-mechanical reliability of a lead-free BGA package Ilja Belov 1, Jonas Arwidson 2, Ralf Poder 3, Pär Johannesson 4, Peter Leisner 4 1 School of Engineering, Jönköping University, Jönköping, Sweden 2 Saab AB publ.Avionics Systems, Jönköping, Sweden 3 SP Technical Research Institute of Sweden,Copenhagen, Denmark 4 SP Technical Research Institute of Sweden, Borås, Sweden abstract The paper investigates the effect of variations both in temperature cycling profile and in SAC305 solder Young’s modulus in the PBGA256 package on the thermo-mechanical reliability. FE simulations quantify the effect of cycle reduction and counting techniques by introducing different temperature profiles having identical dwell- and period time characteristics. A difference of 30% in predicted creep strain energy density per cycle has been determined for the studied profiles. Under the provided modelling assumptions and simplifications, the maximum variation of the thermal fatigue life of SAC305 solder joints is within 30% as the result of experimentally determined Young’s modulus variation in as-delivered packages. |
15:30 |
20mn |
Reliability Study on SMD Components on an Organic Substrate with a Thick Copper Core for Power Electronics Applications Meier, Karsten, Technische Universität Dresden, Electronics Packaging Laboratory Roellig, Mike, Fraunhofer Institute for Ceramic Technologies and Systems – Material Diagnostics abstract In this work we present the results on a reliability study on SMD components mounted on an organic power electronics substrate. The substrate technology has been developed to meet the needs of high ampacity and thermal conductivity with an organic substrate material. Therefore a thick structured copper core was introduced throughout the entire substrate size (see fig. 1). On top of the copper core common FR4 multilayer structures are realised to create a typical SMD PCB surface. The bottom of the copper core was covered with one prepreg with a high thermal conductivity and one copper layer to be able to connect to a heat sink. The copper core becomes structured to enable the use as the layer for conducting high currents. Therefore insulation trenches were manufactured and plugged with a polymer.Introducing a copper core as described changes the thermo-mechanical behaviour of the substrate. Hence, due to the higher stiffness of the copper core PCB, the thermo-mechanical reliability of SMD components can be affected as well.Thermal cycling and shock tests were conducted to reveal the life time of ceramic chip resistors with a size of 0603, 0805 and 1206 in the case of mounting position entirely over the copper core or FR4, over the copper core/FR4 interface or over the copper core insulation trenches (see fig. 2). A specific test board was designed and manufactured (see fig. 3).The achieved results indicate some differences in the reliability of the components depending on the mounting position. In case of component mounting over the copper much higher failure counts have been seen compared to components mounted over FR4 (see fig. 4). Mounting the components on top of the copper core/FR4 interface or insulation trenches changes the reliability in dependence of the component size. |
Session 3 |
Multi-Physics Simulations (including Moisture, Electrical, etc) |
14:00 |
Monday April 20 2015 |
Chaired by Alexander Gielen, David Elata |
14:00 |
30mn |
Keynote presentation – Multiphysics Study of Electromigration in Ceramic Capacitors Jaemi L. Herzberger, Abhijit Dasgupta, Siddhartha Das, University of Maryland Department of Mechanical Engineering, College Park, USA abstract Electrochemical migration across the surface of dielectric cracks in multilayer ceramic capacitors when exposed to humidity or condensed moisture can result in the growth of dendritic filaments, thus causing increased leakage currents or even short circuit failure. This study uses measured empirical data to demonstrate that the growth rate of the dendrite accelerates nonlinearly with time, due to the increased electric field strength and ionic flux that results from the continuously decreasing distance between the anode and the tip of the cathodic dendrite, as the dendrite grows with time. A simple 1D analytic predictive model is developed that incorporates the nonlinear growth kinetics by allowing the separation between the anode and effective cathode to vary with time. The failure time predicted by this model is calibrated with the help of the experimental data, and the dendrite growth kinetics are found to have a close qualitative and quantitative match with the experiments. A comparison is also made to a calibrated fixed-separation linear TTF prediction model in the literature and the comparison shows that the nonlinear model developed in this study produces results that are physically more meaningful than the linear model. |
14:30 |
20mn |
Measurement and Simulation of Electromagnetic Drift for Obsolescence Management in Electronics Geneviève DUCHAMP 1, Tristan DUBOIS 1, Ala AYED 1, Christian MAROT 2, Hélène FREMONT 1 1 University Bordeaux- IMS Laboratory, Talence, France 2 Airbus Group Innovation, Toulouse, France abstract This paper deals with a methodology based on both measurement and simulation approach to study the drift of system electromagnetic characteristics when modifications are made on an electronic board. The objective is to manage the obsolescence of components in electronic assemblies. The main applications involve aeronautical and automotive domains. |
14:50 |
20mn |
A Compact Parametric Model of Magnetic Resonance Micro Sensor M. Kudryavtsev 1, Sh. Gorgi Zade 1, J. G. Korvink 2, T. Bechtol 1 1 University of Rostock, Germany 2 Institute for Microsystems Engineering, Freiburg University, Germany abstract This work presents the application of mathematical methods of parametric model order reduction (pMOR) for automatic generation of the highly accurate, parametric compact models of radio-frequency micro-devices. More specifically, the miniaturized Faraday-induction-based magnetic resonance sensor is considered. Unlike conventional approaches, when magnetic resonance sensor is represented by a lumped-element-based compact model, mathematical pMOR methods are formal, robust and can be performed in an automated way. The introduced parametrization allows producing compact models that are valid over the range of desired parameter’s values without the need to repeat the reduction. |
15:10 |
20mn |
New Simulation Method for Deep Trench Termination Diode DT² using Mixed-mode TCAD Sentaurus Fedia Baccar, Houssam Arbess, Loic Theolier, Stephane Azzopardi, Eric Woirgard, IMS Laboratory, University of Bordeaux, France abstract This work presents a methodology using mixed-mode simulation with TCAD Sentaurus to model, analyze, and optimize the representation of the Deep Trench Termination Diode (DT2) without increasing the number of nodes and the computation time. Moreover, several convergence problems which can be found for many kinds of simulations have been resolved. |
15:30 |
20mn |
Understanding Tip-Enhanced Raman Spectroscopy by Multiphysics Finite Element Simulations Vladimir Kolchuzhin, Jan Mehner, Evgeniya Sheremet, Bhattacharya Kunal, Raul D. Rodriguez, Dietrich R.T. Zahn, Technische Universität Chemnitz, Chemnitz, Germany abstract This article deals with the models development and FE simulations for mechanical properties of all-metal AFM-TERS tips and electric field enhancement between the tip and the sample. The most important aspects in simulations, the parameters necessary in creating models, and the obtained results are presented and discussed in the article. |
Session 4 |
Thermal Behavior Modeling and Characterization |
14:00 |
Monday April 20 2015 |
Chaired by Peter Rodgers, Ilja Belov |
14:00 |
30mn |
Keynote presentation – Modeling and Model-based Control of Temperature in an SThM Probe Michel Lenczner 1, Bin Yang 2, Mohamed Abaidi 2, Alexia Bontempi 2, Damien Teyssieux 3, Bernd Koehler 4, Pawel Janus 5 1 FEMTO-ST – Technical University of Belfort Montbéliard, France 2 FEMTO-ST – University of Franche-Comté, France 3 FEMTO-ST – Ecole Nationale Supérieure de Mécanique et de Microtechnologie, France 4 Fraunhofer-Institute – IZFP, Germany 5 Instytut Technologii Elektronowej – Wroclaw University of Technology, Poland abstract We present a multi-scale model of a probe for scanningthermal microscopy. The probe is built by microfabricationtechniques. In active mode, it is supplied by a source of harmonicand/or continuous current and the tip temperature is measured aftera lock-in amplifier. The model distinguishes two time scales and twospace scales. Simulation results show the potential of the model interms of accuracy and computation speed and they are compared toexperimental results. Finally, a temperature control law constructedfrom this model is stated. |
14:30 |
20mn |
CFD Simulations of Wave Soldering on Through-Hole Printed Circuit Assemblies A. Yuile, S. Wiese, Saarland University, Germany abstract The main observations and results, which have been collected from wave soldering computational fluid dynamics (CFD) simulations for a lead-free SnAgCu (SAC) solder, are presented here in terms of solder shape, penetration and electrical continuity. The simulation models comprise of steady state transitional shear stress transport (SST) melting/solidification models of a single pin-through hole (PTH) configuration on a printed circuit board (PCB). The simulations make use of the commercially available ANSYS Fluent CFD solver. The simulation models have been developed to the extent that they are capable of capturing and investigating some of the physically salient features, which dominate wave soldering processes, such that improvements in efficiency/efficacy can potentially be pursued. The simulations also account for the influence of variations in solder material properties, such as viscosity, surface tension and density with respect to temperature.Furthermore, within this paper, areas are highlighted as to how to improve upon and extend the applicability of the models through future development. |
14:50 |
20mn |
Simulation of a Flip Chip Bonding Technique Using Reactive Foils F. Kraemer, C. Pauly, F. Muecklich, S. Wiese, Saarland University, Saarbrucken, Germany abstract This paper discusses the process requirements and opportunities for a flip chip bonding applying nano scale multilayer films, so called reactive foils. These reactive foils create heat by a highly exothermic and self-propagating reaction of its constituents which are sufficient to melt surrounding layers of solder. The assessment of this novel bonding technique is done by thermal FEM-simulations of a complete flip chip assembly. The goal of this analysis is the identification of processing limits and benefits applying this unconventional localized heat source in order to minimize mechanical stress in the entire flip chip assembly. The result of this analysis shows process requirements which shall be suited for a successful production step.The microelectronic packaging develops towards an increased integration of different functionalities into specialized components. This trend requires the 3-D stacking of different dies and the further miniaturization of interconnection structures. Especially the stacking of different die sizes as well as multiple die stacks require low temperature bonding techniques in order to prevent these assemblies from cracking. In case of tiny flip chip interconnections a localized temperature impact improves the solder reliability due to the missing thermal expansion of the bonded dies and the bonded die to substrate, respectively. Both advantages can be met by the application of a local heat source such as reactive foils.The presented assessment of the reactive foil capabilities is focused on a single flip chip which is soldered onto a substrate. The artificial test structure is represented by a 3-D FE-model. The analyzed interconnection area consists of 8 interconnections arranged in 2 rows including 4 columns each. The heat generated by the reactive Al/Ru foil is represented by heat fluxes which act for a short period in a small section of the substrate copper pads. These heat sources travel through the entire interconnection zone of the contact region.The paper discusses the simulation results of the flip chip bonding process using several assumptions for the heat flux generated by the reactive foil. It will be shown how different thicknesses of the reactive foil affect the fusing temperature inside the flip chip joints. It will be further discussed how different geometries of the flip chip joints (bump diameter, pitch) will influence the global heat flux during the joining process. |
15:10 |
20mn |
Thermal Analysis of Micro-Hotplates for Catalytic Gas Microsensors Alexander Kozlov, Omsk State University, Omsk, Russia abstract The analytical method is presented to conduct thermal analysis of micro-hot plates for catalytic gas microsensors. The 2D structure of a micro-hotplate is divided into the regions, the temperature distribution in which is found by the eigenfunction method. To take into account the temperature dependencies parameters the iteration procedure is used. The values of the parameters are determined using the weighted mean temperature in each region. The application of this method is considered on the example of the catalytic gas microsensors based on the square plates supported by two cantilever beams in various places: at centre of two opposite sides; at two opposite corners; at two adjacent corners. For these micro-hotplates the following characteristics were determined: the dependencies of the weighted mean overheating temperature, the difference between the maximum and minimum temperatures and the ratio of the weighted mean overheating temperature to the difference between the maximum and minimum temperatures on the length of the cantilever beams, the dimensions of the micro-hotplate and the supply voltage of the microsensor. |
15:30 |
20mn |
Current Load Capacity of Electrical Conductor Tracks Evaluated by Simulation and Thermographic Imaging Thomas Tetzlaff, Ulf Witkowski, Electronics and Circuit Technology, South Westphalia University of Applied Sciences, Soest, Germany abstract Today’s LEDs operate with high current, causing a significant increase in the device’s temperature despite the LEDs being highly energy efficient. Important for having a long LED lifetime is accurate temperature management which requires sensing the device temperature and knowing the temperature distribution. The underlying aim of the project is to simulate the temperature distribution in LED devices mounted on printed circuit boards and to compare the data with simulation results. The focus of this paper is the modelling and simulation of PCB tracks of different widths that are typically used in board designs and device interconnections. Based on the finite element method, the surface temperature of electrical tracks is simulated using the tool COMSOL. For comparison with real world measurements, the simulated structures have been fabricated and analysed via contact-free infrared thermography. The main challenge is the mechanical modelling of the track structure and the setup of the material parameters to have a good match between simulation results and measurements. In a future step, the simulation results will be used to allow for integrated temperature management of LED devices which include power supply structures to avoid overheating at the p-n junctions and at the light conversion material on top of the LED. |
Session 5 |
Interconnect Investigations |
16:20 |
Monday April 20 2015 |
Chaired by Steffen Hartmann, Patrick McNally |
16:20 |
30mn |
Keynote presentation – Adequate Mechanical Copper Modelling for 2nd Level Interconnect Structures Steffen Wiese, Frank Kraemer, Saarland University, Saarbruecken, Germany abstract This paper discusses the requirements for achieving adequate modelling of the mechanical behaviour of copper interconnect structures in electronic assemblies. In this context it focuses on an experimental approach to characterise the mechanical behaviour of representative copper specimens. For this purpose the general constitutive behaviour of pure copper will be discussed, in doing so covering the specifics of elastic response, the onset of plastic flow with respect to microstructure, as well as load history, and the resulting cyclic elastic-plastic behaviour. The discussion regarding an appropriate experimental methodology starts with considerations about specific deformations of representative thin stripe specimens during a standard tensile test. In the second part this discussion is developed into design ideas for appropriate experimental instrumentation, which enables one to determine the required properties of the representative specimens. |
16:50 |
20mn |
A Geometry-Independent Lifetime Modelling Method for Aluminum Heavy Wire Bond Joints Arian Grams 1, Jan Höfer 1, Andreas Middendorf 2, Stefan Schmitz 1, Olaf Wittler 1, Klaus-Dieter Lang 2 1 Fraunhofer IZM, Berlin, Germany 2 Technische Universität Berlin, Berlin, Germany abstract Wire bond degradation is a limiting factor for the lifetime of state of the art power modules. So, there is a need for widely applicable and proven modelling techniques to achieve a reliable design.In this paper, a new crack growth law has been developed and calibrated with experimental data. By defining a failure criterion and optimizing model parameters, good lifetime predictions have been achieved. In addition, further possibilities to use this modelling approach have been proposed, e.g. damage in interconnect layers as sinter silver or solder layers could be considered. |
17:10 |
20mn |
Thermo-mechanical ball bonding simulation with elasto-plastic parameters obtained from nanoindentation and atomic force measurements Alan Wright 1, Stephane Koffel 1, Silke Kraft 1, Peter Pichler 2, Juri Cambieri 3, Rainer Minixhofer 3, Ewald Wachmann 3 1 Fraunhofer Institute for Integrated Systems and Device Technology IISB, Erlangen, Germany 2 Fraunhofer IISB and Chair of Electron Devices, University of Erlangen-Nuremberg, Germany 3 ams AG, Unterpremstätten, Austria abstract A ball bonding process was simulated over a high-voltage isolation structure. The removal of an inter-dielectric metal crack-stop layer was investigated through 3D simulation. Material properties for the bonded gold ball were obtained using nanoindentation and atomic force microscopy with a methodology from the work of Ma et al. This yielded both elastic and plastic material parameters. The methodology was then evaluated by using the parameters in a nanoindentation simulation. Although the topography simulated only roughly agreed with measurement, the simulated and measured indenter curves closely overlapped. The parameters were then used in the bonding simulation. The deformation of the bond ball was also measured so that the equivalent deformation could be simulated. This was achieved following the incorporation of both ultrasonic motion and softening in the simulation. Two bonding process geometries were then set up: one with the crack-stop layer present and the other without. Both were simulated and the output was applied within a failure theory to evaluate the risk to the isolation oxide. |
17:30 |
20mn |
Analyzing Thermo-Mechanical Reliability of an Interconnect Based on Metal Coated Polymer Spheres MPS R. Fayҫal Hamou 1, Daniel N. Wright 2, Astrid-Sofie B. Vardøy 2, Marie Haupt 3, Susanne Helland 4, Helge Kristiansen 4, Maaike M.Visser Taklo 2 1 SINTEF ICT, Microsystems and Nanotechnology MiNaLab, Oslo, Norway 2 SINTEF ICT, Instrumentation Dept., Oslo, Norway 3 AMIC Angewandte Micro-Messtechnik GmbH, Berlin, Germany 4 Conpart AS, Skjetten, Norway abstract In this study, we explore the thermo-mechanical stress distribution of a chip/substrate BGA interconnect based on metal coated polymer spheres (MPS) of 30 μm diameter. The bonding of the chip to a glass substrate with MPS is obtained by deposition and sintering of a silver nanoparticle suspension that forms a menisci needed for necking and metallic bonding of the MPS towards pads. The stand-off height is determined by the ball diameter, thus the necks and the MPS coating are considered to be the critical parameters of the system. The simulation study is focused on varying the shape and the size of the neck and the MPS coating thickness. The polymer core is modeled as a viscoelastic material using generalized Maxwell model with Prony series. The distribution of the relaxed thermal stress and strain within the MPS coating and necks is analyzed as a function of temperature and the identified critical parameters. Moreover, shear test measurements of single MPS and SEM images of the structures are presented and discussed, in order to expose the feasibility of this new interconnect technology. |
Session 6 |
Molecular Dynamics and Process Modelling |
16:20 |
Monday April 20 2015 |
Chaired by Li-Ling Liao, Emad Poshtan |
16:20 |
30mn |
Keynote presentation – Towards Nanoreliability of CNT based Sensor Applications: Investigations of CNT-Metal Interfaces Combining Molecular Dynamics Simulations Advanced In Situ Experiments and Analytics. Steffen Hartmann 1, Alexey Shaporin 1, Sascha Hermann 1, Jens Bonitz 1, Marc Heggen 2, Peter Meszmer 1, Heinz Sturm 3, Ole Hölck 4, Thomas Blaudeck 1, Stefan E. Schulz 5, Jan Mehner, Bernhard Wunderle 1, Thomas Gessner 5 1 Technische Universität Chemnitz, Germany 2 Ernst Ruska-Centrum und Peter Grünberg Institut, Forschungszentrum Jülich GmbH, Germany 3 Bundesanstalt für Materialforschung und -prüfung, Germany 4 Technische Universität Chemnitz Fraunhofer IZM Berlin, Germany 5 Fraunhofer ENAS Chemnitz, Germany abstract In this paper we present results of our recent efforts to understand the mechanical interface behaviour of single-walled carbon nanotubes (CNTs) embedded in metal matrices. We conducted experimental pull-out tests of CNTs embedded in Pd or Au and found maximum forces in the range 10-102 nN. These values are in good agreement with forces obtained from molecular dynamics simulations taking into account surface functional groups (SFGs) covalently linked to the CNT material. The dominant failure mode in experiment is a CNT rupture, which can be explained with the presence of SFGs. To qualify the existence of SFGs on our used CNT material, we pursue investigations by means of fluorescence labeling of surface species in combination with Raman imaging. We also report of a tensile test system to perform pull-out tests inside a transmission electron microscope to obtain in situ images of CNT-metal interfaces under mechanical loads at the atomic scale. |
16:50 |
20mn |
Computational Electrohydrodynamics in the Fabrication of Hollow Polymer Microstructures C. Tonry 1, M. Patel 1, M. Desmuliez 2, W. Yu 3, C. Bailey 1 1 Computational Mechanics and Reliability Group CMRG, School of Computing and Mathematical Sciences, University of Greenwich, UK 2 Microsystems Engineering Centre MISEC, School of Engineering Physical Sciences, Heriot Watt University, UK 3 State Key Laboratory of Applied Optics, Changchun Institute of Optics, Chinese Academy of Sciences, Changchun, China abstract Electric Field Assisted Capillarity is a novel process which has the potential for the fabrication of hollow polymer microstructures as a single step process. The process has been shown to work experimentally on a microscale using PDMS. The process makes use of both the electrohydrodynamics of polymers at a microscale and also the capillary force on the polymer caused by a low contact angle on a heavily wetted surface. Discussed in this paper are the results of a two-dimensional numerical simulation of the process. The results presented here are for the an angular mask producing microchannels and demonstrate how differing contact angles on the top mask effect the thickness of the top of the microstructures and also whether the fabrication of the microstructure is possible at all. |
17:10 |
20mn |
Diffusion Simulation thru Polymers using Coarse-grained Models Derived from Molecular Models. Nancy Iwamoto, Honeywell, USA |
17:30 |
20mn |
Submodeling FEM Analysis of 3D Printed Structures Javad Zarbakhsh 1, Armin Iravani 1, Zeinab Amin-Akhlaghi 2 1 Carinthia University of Applied Sciences, Department of Engineering IT, Villach, Austria 2 ZAMSTEC – Scientific Technology Engineering and Consulting, Riegersdorf, Austria abstract For the first time, nested sub-modeling approach and Finite Element Analysis have been used to analyze the structural mechanical analysis of 3D printed part, whereas the details of 3D printing patterns included in sub-model. The results present a general tool which can improve the quality of 3D printed parts, which have multidisciplinary application in various fields. It is found that the Maximum Principle stress is highly concentrated at 3D printed layers. For a specific 3D printing pattern, the stress intensity factor has been calculated to have the value of 4. Results have been discussed from theoretical, simulation and experimental observation point of view. |
Session 7 |
Thermal Modelling and Analysis |
16:20 |
Monday April 20 2015 |
Chaired by Marta Rencz, Frank Krämer |
16:20 |
30mn |
Keynote presentation – Analytical Experimental and Numerical Approach to Thermal Analysis and Design of a Travelling Wave Tube Waldemar Wiejak, PIT-RADWAR S.A., Wroclaw, Poland Artur Wymyslowski, Wroclaw University of Technology, Wroclaw, Poland abstract Simulation of thermal energy transport in complicated structures is usually a challenge. An example of such problem is heat generation and transfer through microwave delay structure in the travelling wave tube (TWT) device. A high level of microwave power and considerable energy of electrons intercepted by the delay line leads to the increase of local temperature of the microwave structure. Determination of the heat transfer from the hot spots is essential for the proper design of the delay line and assisting cooling system. This problem has been investigated by means of a combined: analytical, numerical and experimental approach. Such methodology does not require expensive equipment and is much faster than the pure experimental analysis. Presented analytical model is focused on evaluation of the electron beam power dissipation and microwave losses along the delay line, which is not uniform and most of the power is dissipated at terminal part of the delay line. The analytically evaluated power dissipation was used in numerical simulation in order to assess the temperature distribution. Finally the results were validated experimentally using a designed measuring setup. One of the final conclusions was that the temperature distribution has a nonuniform character and the resulting high temperatures at the delay line output can significantly influence the device reliability parameters. |
16:50 |
20mn |
Measurement of Non-uniform Junction Temperature Distribution of Large Light-Emitting Diode by Using a Modified Forward Voltage Method Mian TAO, Center for Advanced Microsystems Packaging, Hong Kong University of Science Technology, Hong Kong S. W. Ricky LEE, Department of Mechanical and Aerospace Engineering, Center for Advanced Microsystems Packaging, HKUST LED-FPD Technology RD Center at Foshan, Hong Kong University of Science Technology, Hong Kong abstract Large size light-emitting diodes (LED) are frequently used in the application that requires high luminous intensity. Among diverse types of LED chips, the wire-bonding LED chip has become the most common type for its simple and mature manufacturing processes. For the commercial product manufacturing, chip bonding is one of the most critical procedures. Conventional chip bonding uses adhesive filled with thermal conductive particles. As the size of the LED chips is being enlarged, it becomes much more challenging to achieve perfect bonding. Defects may often occur in the bonding layer in practical manufacturing. Previous studies have revealed that defects inside the bonding layer will block the heat flow from the LED junction to the carrier and create non-uniform junction temperature distribution. The ordinary method to measure the junction temperature is the forward voltage method which uses the negative forward voltage-junction temperature characteristic of LEDs. Nevertheless, this standard method could not offer any information about the non-uniformity of junction temperature. As it is well understood that the junction temperature is critical to the performance of an LED device, it is necessary to find out a method to evaluate this non-uniform junction temperature phenomenon. In this study, we prepared several custom-made LED samples with artificial bonding defects to generate non-uniform junction temperature. An improved forward voltage method was developed to detect the junction temperature non-uniformity. The introduced method was experimentally validated. |
17:10 |
20mn |
Model Verification of Heat Exchangers in a Flow Test Rig Klas Brinkfeldt 1, Thorbjörn Åklint 1, Klaus Neumaier 2, Olaf Zschieschang 2, Michael Edwards 3, Dag Andersson 1 1 Swerea IVF, Mölndal, Sweden 2 Fairchild Semiconductor GmbH, Aschheim, Germany 3 Chalmers University of Technology, Göteborg, Sweden abstract In power electronics, more efficient removal of heat from the junction of power devices leads to a higher power rating per die, which in turn leads to fewer die and reduced system volume. Since temperature is a main driver in expected failure modes an increase in cooling capability can also enhance margins of the device reliability. Previously, CFD simulations of two novel heat exchanger designs that will be used in a power module with double sided cooling have been reported on. The heat exchangers are fabricated by direct 3D manufacturing of copper structures, which allows almost complete freedom in geometric design. Two novel geometries of heat exchanger cooling structures have previously been modeled in terms of thermal performance and expected pressure drop. A flow rig has been designed and calibrated to measure thermal performance and pressure drops of these heat sinks. For calibration purposes, measurements of the thermal response of wave structured and unstructured heat sinks are reported here. The results show that, as expected, the heat sink temperatures are lower for all flow rates in the wave-structured geometry. A thermal CFD model accurately predicts the behavior of the temperature difference between inlet and outlet versus flow rate, but predicts higher absolute temperature values. It was also found that the model underestimates the pressure drop over the tested heat sinks. The pressure drop across a novel pine cone geometry heat sink fabricated by additive manufacturing methods was also measured. Comparisons to a reduced model, which neglects everything before the inlet and after the outlet of the tested device, showed that the behavior of this pine structured heat sink was not predicted correctly. The pressure drop increased more rapidly with flow rates in the model than in the measurements. The main source of error in the measurements and simulations comes from a lack of thermal loading. Future work to improve the flow rig includes possibilities to increase the temperature loading at the bottom of the heat sink under test. |
17:30 |
20mn |
Temperature Profiles along Bonding Wires Revealed by the Bond Calculator, a New Thermo-Electrical Simulation Tool Carl Christoph Jung 1, Christian Silber 2, Jürgen Scheible 1 1 Robert Bosch Center for Power Electronics, Reutlingen University, Germany 2 Robert Bosch GmbH, Reutlingen, Germany abstract When a bonding wire becomes too hot, it fuses and fails. The ohmic heat that is generated in the wire can be partially dissipated to a mold package. For this cooling effect the thermal contact between wire and package is an important parameter. Because this parameter can degrade over lifetime, the fusing of a bonding wire can also occur as a long-term effect. Another important factor is the thermal power generated in the vicinity of the bond pads. Nowadays, the reliability of bond wires relies on robust dimensioning based on estimations. Smaller package sizes increase the need for better predictive methods.The Bond Calculator, a new thermo-electrical simulation tool, is able to predict the temperature profiles along bond wires of arbitrary dimensions in dependence on the applied arbitrary transient current profile, the mold surrounding the wire, and the thermal contact between wire and mold.In this paper we closely investigated the spatial temperature profiles along different bond wires in air in order to make a first step towards the experimental verification of the simulation model. We are using infrared microscopy in order to measure the thermal radiation generated along the bond wire. This is easier to perform quantitatively in air than in the mold package, because of the non-negligible absorbance of the mold material in the infrared wavelength region.The Bond Calculator is a fast and exact tool to help designers to choose the apt bond wire, which does not fuse under the conditions of operation. The Bond Calculator is orders of magnitude faster than FEM and Easy-to-use.The Bond Calculator also helps designers to estimate the temperature changes at the bond connection itself, by calculating the time and space dependence of the power delivered from the bond wire to the chip. These temperature changes can affect the durability of the bond connection.The Bond Calculator uses a simplified simulation model, which allows the numerical calculation of partial differential equations, to calculate the temperature profile in a bond wire from the induced current profile. This software tool has been validated by FEM and measurement.The Bond Calculator is easier to apply and much faster in calculation than FEM software, it can be for example 1000 times faster for transient simulations. The temporal and spatial power profiles delivered to the chip are easily estimated. |
Session 8 |
Modeling for Advanced Package Development |
09:00 |
Tuesday April 21 2015 |
Chaired by Dag Andersson, Wenbo Yuan |
09:00 |
30mn |
Keynote presentation – Thermo-Mechanical Simulations of SiC Power Modules with Single and Double Sided Cooling Klas Brinkfeldt 1, Michael Edwards 2, Jonas Ottosson 3, Klaus Neumaier 4, Olaf Zschieschang 4, Alexander Otto 5, Eberhard Kaulfersch 6, Dag Andersson 1 1 Swerea IVF, Mölndal, Sweden 2 Chalmers University of Technology, Göteborg, Sweden 3 Volvo Group Truck Technology, Göteborg, Sweden 4 Fairchild Semiconductor GmbH, Aschheim, Germany 5 Fraunhofer ENAS, Chemnitz, Germany 6 Berliner Nanotest und Design GmbH, Berlin, Germany abstract Effectively removing dissipated heat from the switching devices enables a higher current carrying capability per chip area ratio, thus leading to smaller or fewer devices for a given power requirement specification. Further, the use of SiC based devices has proven to increase the efficiency of the system thereby reducing the dissipated heat. Thermal models have been used to compare SiC power modules. Single and double sided cooling have been simulated. The simulated maximum temperatures were 141 °C for the single sided version and 119.7 °C for the double sided version. In addition, the reliability of the single sided module has been investigated. A local model of the wire bond interface to the transistor metallization shows a 3‰ maximum increase in plastic strain during the power cycle. Simulations of the creep strain rates in the die attach solder layer for a power cycling loads also shows a 3‰ increase in creep strain per cycle. |
09:30 |
20mn |
Reliability Analysis of Copper Bump Interconnection in Double-sided Power Module Chia-Chi Tsai, Li-Ling Liao, Yen-Fu Su, Tuan-Yu Hung, Kuo-Ning Chiang, Dept. of Power Mechanical Engineering National Tsing Hua University, HsinChu, Taiwan, R.O.C. abstract A high current load may cause the Joule heating, subsequently raising the chip temperature in a conventional power module. Temperature excursion in power chip may generate thermal stress, induce failure and reduce its reliability. Double-sided power module is a crucial structure to provide another heat dissipation path and efficiently reduce chip temperature. This study estimate the thermal and reliability analysis of double-sided power module by using copper bump as an interconnection under different cooling condition. The connection layout can be designed more flexible by using bump interconnection in double-sided power module. The concept of dummy ball also utilized to reduce the mechanical strain or stress of copper bump and improve its reliability in a power module. |
09:50 |
20mn |
Chip Package Interaction: A Stress analysis on 3D IC’s packages Melina Lofrano, Mario Gonzalez, Wei Guo, Geert Van der Plas, imec, Leuven, Belgium abstract In this work CPI induced mechanical stress for 3D stacks and 3D interposer packages is studied. The stress built during package assembly has been obtained using finite element modeling (FEM). For the package layout and materials properties chosen for this work, the results shown that the stresses induced during the processing of a 3D stacks and 3D interposer configuration are similar when they are assembled in a Flip Chip Ball Grid Array (fcBGA) package. Furthermore, the interconnection between the different silicon dies assured with the use of µbumps were analyzed with different interconnect densities and configurations. Results shown that stress induced around the µbumps increases by increasing the µbump pitch. Different molding configurations for the fcBGA packages were investigated, including high power (exposed die) and low power (embedded dies) packages. The results showed that exposed die packages present lower out of plane deformation due to a reduction of the epoxy mold compound (EMC) thickness. It is very important to accurately calculate the residual stresses that each processing steps of the assembly induced on the die. Mass reflow and thermo compression bonding process assembly have been investigated. Results showed that solder joint reflow is the bottleneck for mass reflow process assembly, high stress in this step indicate that failures can occur. In this work we showed that low CTE laminate is a good alternative to reduce until 60% stress at flip chip reflow step. |
Session 9 |
Multi-Physics Modelling of Interconnects |
09:00 |
Tuesday April 21 2015 |
Chaired by Véronique Rochus, Pradeep Lall |
09:00 |
30mn |
Keynote presentation – Life Time Characterization for a Highly Robust Metallization Kirsten Weide-Zaage 1, Jörg Kludt 1, Markus Ackermann 2, Verena Hein 2, Marco Ernstling 2 1 Reliability: Simulation and Risk Analysis Group, Institute of Microelectronic Systems – AS, Leibniz Universität Hannover, Germany 2 X-FAB Semiconductor Foundries AG, Germany abstract For mixed signal applications it is necessary to have metallization which are able to carry high currents. Also the on chip integration leads to special requirements on the metallization concerning their robustness. A common method for the determination of interconnect lifetime is described in JP001A and based on Black’s law and the measurement of time to failure, medium stress current density and medium stress temperature. The highly robust metallization presented here, which was developed for higher current and temperature applications shows more complicated shapes than presently used metallization systems with metal line tracks and via. To determine a realistic life time of highly robust metallization the used method is not applicable anymore. A more suitable determination of the variables current density and temperature for AlCu metallization with W-plug can be achieved by simulations. In the metal line layout the most critical locations regarding mass flux are chosen. The results are validated by measurements. |
09:30 |
20mn |
Electro-Thermo-Mechanical Analyses on Silver Sintered IGBT-Module Reliability in Power Cycling R. Dudek 1, R. Döring 1, S. Rzepka 1, C. Ehrhardt 2, M. Günther 3, M. Haag 4 1 Fraunhofer ENAS, Micro Materials Center, Chemnitz, Germany 2 Fraunhofer IZM, Berlin, Germany 3 Robert Bosch GmbH, CRAPJ3, Stuttgart, Germany 4 Bertrandt Ingenieurbuero GmbH, Ingolstadt, Germany abstract New demands on the thermo-me¬chanical design of sintered silver interconnections emerge. Development of this inter-connection technology and both experimental and theoretical studies on their reliability were subjects of the project “PROPOWER”. The focus of this paper is on theoretical analysis of thermo-mechanical reliability risks of a project demonstrator, an IGBT module, subjected to power cycling loadings. Coupled electro-thermal-mechanical analyses have been carried out using the finite element method (FEM). Introduction of a new interconnect material means at the same time introduction of a new constitutive behavior and new failure modes. As the material stiffness increases, the decoupling effect of compliant solder layers reduces and intrinsic mechanical stresses increase in the whole pow-er stack. This leads on one hand to less low cycle fatigue in the interconnect, as plastic dissipation is reduced, but on the other hand to higher failure risks like brittle cracking and sub-critical crack growth. However, if early brittle fail-ure can be avoided by appropriate designs, the new interconnection technology allows an increase in fatigue relia-bility of several hundred percent. Based on the complex theoretical framework simulation results are validated by testing in order to achieve trustworthy thermo-mechanical reliability predictions. Failures like chip metallization damage and the different damage mechanisms of the die bond if either solder or sinter silver is used are related to the different stress situations in the module. |
09:50 |
20mn |
Reliability Assessment of Copper Ball Bonds by Combination of Simulation and Accelerated Mechanical Testing Martin Lederer 1, Alice Lassnig 2, Golta Khatibi 1, Mitra Delshadmanesh 1 1 Vienna University of Technology, Austria 2 University of Vienna, Austria abstract We have investigated the fatigue strength of copper ball bonds attached to silicon chips covered with an aluminum metallization layer. Under service conditions, the chips are exposed to repeated temperature changes leading to thermo-mechanical stresses. In order to predict their lifetime, accelerated mechanical tests were performed at ultrasonic frequency. Thereby, different setups were tested to optimize the method. The stresses resulting from mechanical tests were compared to deviatoric stresses originating from thermal mismatch of the different materials involved. For this purpose, FEM computer simulations were performed. It was found that the stress distributions of the different setups are quite different. Nevertheless, the weakest link of the construction was always the aluminum metallization layer. In consequence, the number of loading cycles to failure could be related to the von Mises stress observed in the aluminum film. |
Session 10 |
Thermomechanical Behavior Modeling and Characterization |
09:00 |
Tuesday April 21 2015 |
Chaired by Jeroen Zaal, Ilko Schmadlak |
09:00 |
30mn |
Keynote presentation – Thermo-Mechanical Effects in Majorana Type Quantum Devices A.W.J. Gielen, F.O. Valega Mackenzie, TNO Technical Sciences Materials Solutions, Eindhoven, the Netherlands abstract We have developed a multi-scale model, consisting of an atomistic model in LAMMPS of an InSb nano-wire, and a continuum model in COMSOL of a so-called Majorana research device, to study the effects of thermo-mechanical deformations during the cool down from room temperature to the operating temperature of about 50 mK. For the simulation of the InSb nano-wire suitable potentials were implemented in LAMMPS. The simulation results of the nano-wire show size dependent Young’s moduli and gradients in the radial lattice spacing during uniaxial straining. The materials properties that were derived from the atomistic model, were introduced in the continuum model. Cool down of the device from room temperature to it’s operating temperature introduced significant deformation. However, the stresses in the system are moderate and no fracture or damage is expected. Still, deformation of the device will induce shifts in band gap behavior of the device. Band gap shifts using a simple approximation are estimated to be about 34%. |
09:30 |
20mn |
Impact of Solder-Joint Tilting on the Reliability of LED-based PCB Assemblies: A Combined Experimental and FEM Analysis Bart Vandevelde 1, Franco Zanon 2, Alessio Griffoni 2, Xiaolong Li 3, Geert Willems 1, Matteo Meneghini 4 1 imec, Leuven, Belgium 2 Osram, Treviso, Italy 3 Osram, Treviso; University of Padova, Italy 4 University of Padova, Italy abstract The impact of solder-joint tilting on the reliability of high-power LEDs soldered on PCBs is investigated by means of FEM simulations correlated with thermal cycling experiments. A non-uniform solder joint stand-off height is implemented into the FEM and, using crack propagation modelling approach, the number of cycles to complete fracture are predicted. |
09:50 |
20mn |
Vapor Pressure Prediction in Reflow for Stacked-Chip Packages by a Convection-Diffusion Model Jeremy Adams, Liangbiao Chen, Xuejun Fan, Lamar University, USA abstract Moisture plays a critical role in the reliability of electronic devices, especially in the desorption process at reflow temperatures (around 270° C) when severe damages may occur due to high-pressure vapor concerted from condensed moisture. Such pressure-driven vapor flow, however, could not be described by conventional Fick’s Law. Furthermore, using conventional Fick’s Law for multi-materials always encounters interface discontinuity issues. Therefore, this paper adopts a Convection-Diffusion Model that is able to describe complex desorption behavior in a multi-material media without the discontinuity issue. Both pressure gradient-driven (convection) and concentration-gradient driven (diffusion) moisture transports are considered in the model. To achieve this, absorbed moisture is partitioned into vapor phase and liquid phase (condensed water), with the vapor flux governed by Darcy’s Law and the water flux by Fick’s Law. Henry’s Law is also implemented so that the Fickian term is converted to pressure, resulting in a unified vapor pressure model. The model is applied to analyze a stacked-chip package by two numerical cases: desorption under 2 typical reflow temperature profiles. Numerical validations are also performed to show that the Convection-Diffusion Model can be reduced to traditional Fickian Model and Convection-Only Model as special cases. The numerical results show that the concentration desorption rate is much faster than that of the traditional Fickian diffusion, and somewhat faster than the Convection Model, this results in a much lower pressure in the material. However, the desorption profile with time and the pressures at low temperatures of the different models– the Convection-Only, Diffusion-only and the Convection-Diffusion Model are indistinguishable which can be seen in both reflow profiles. The sensitivity of the CD Model to the gas permeability k and the reflow temperature profiles governs the maximum pressure that is predicted as well as the concentration content. |
Session 11 |
Thermo-Mechanical Issues in Microelectronics |
10:40 |
Tuesday April 21 2015 |
Chaired by Bernhard Wunderle, Kaspar Jansen |
|
|
Multi-physics Modelling of Thin Films : Optimization for Finite Elements Simulations Tools Toni Youssef 1, Eric Woirgard 2, Stephane Azzopardi 2, Donatien Martineau 1, Regis Meuret 1 1 Labinal Power Systems, Moissy-Cramayel, France 2 IMS Laboratory, Talence, France abstract This paper focuses on the thin Nickel layer thicknesses in power modules. Thermal and mechanical behaviors of these thin layers in a power module are investigated. An approach is shown in order to present an improved modeling by considering the effect of these thin layers and by reducing time computation. |
|
|
An Overview of Scanning Acoustic Microscope a Reliable Method for Non-destructive Failure Analysis of Microelectronic Components M. Yazdan Mehr 1, A. Bahrami 2, H. Fischer 2, S. Gielen 2, R. Corbeij 2, W.D. van Driel 3, G.Q. Zhang 4 1 Material innovation institute M2i, Delft, The Netherlands; Delft University of Technology, EEMCS Faculty, Delft, The Netherlands 2 Netherlands Organization for Applied Scientific Research, TNO, The Netherlands 3 Delft University of Technology, EEMCS Faculty, Delft, The Netherlands; Philips Lighting, Eindhoven, The Netherland 4 Delft University of Technology, EEMCS Faculty, Delft, The Netherlands abstract In a highly competitive and demanding microelectronics market, reliable non-destructive methods for quality control and failure analysis of electronic components are highly demanded. Any robust non-destructive method should be capable of dealing with the complexity of miniaturized assemblies such as chip-scale packages and 3D IC stacks. Scanning acoustic microscopy (SAM) is indeed one the best non-destructive tools for failure analysis purposes. It is also a useful technique for imaging the morphology, location and size distribution of defects in different microelectronics components. SAM can detect delaminations at sub-micron thicknesses. It is also one of the only available techniques capable of efficiently evaluating popcorning in PBGA’s and is a also useful device to detect sub-micron air gaps. SAM can also be used to measure the thickness of an internal layer of material. Overall, SAM is an efficient tool for evaluating such a wide range of different defects in printed circuit boards, underfills, BGAs, wire bonds, discrete components, and wafers. In SAM a focused sound is directed from a transducer at a small point on a target object, as is schematically shown here. Sound, hitting a defect, inhomogeneity or a boundary inside material, is partly scatted and will be detected. The transducer transforms the reflected sound pulses into electromagnetic pulses which are displayed as pixels with defined gray values thereby creating an image. This article aims at giving an overview of scanning acoustic microscope (SAM) and explaining its operating principles and its limitations. A few examples are also given for further clarification. |
|
|
Application of Design of Computer Experiments DoCE method for the extraction of the elasto-plastic behavior law of ECD copper through nano-indentation tests Stéphane Moreau, François de Crécy, Vincent Mandrillon, Univ. Grenoble Alpes, Grenoble, France. CEA, LETI, MINATEC Campus, Grenoble, France abstract This paper report a new methodology combining finite element modeling (FEM) and nanoindentation tests in order to extract the plasticity law of ECD copper film. Thanks to this approach, we demonstrate that simple plasticity law (ANSYS Multilinear ISOtropic hardening (MISO)) reaches the goal of fitting instead of more complex ones (Hollomon-like models). In addition, there is no unique solution but a set of good solutions! |
|
|
Designing Efficient Computer Experiments – The Step Beyond Finite Element Modelling Vollert, Natalie 1, Schicker, Johannes 1, Hirschl, Christina 1, Kraft, Marin 1, Pilz, Jürgen 2 1 CTR Carinthian Tech Research AG, Villach, Austria 2 Alpen-Adria University of Klagenfurt, Klagenfurt, Austria abstract A common challenge in modern multi-physicssimulations like FEM is that the more complex theunderlying problems become, the more the simulationdepends on a range of not or just poorly understoodparameters. At the same time, the increase of FEMcomputing time with the complexity of the underlyingproblem makes it impossible to explore the wholeparameter space with FE simulations. Gaining as muchinformation as possible from a manageable number ofruns clearly requires involving some form of Design ofExperiments (DOE), referred to as Design of ComputerExperiments (DOCE) for simulation studies. In additionto the decision for which parameter sets simulationsshould be performed, the results of these simulations areused as data for constructing a statistical “metamodel”.By enabling the calculation of any variable of interestfrom arbitrary parameter sets without having to runnew simulations, these metamodels facilitate an efficientexploration of the entire parameter space with optimaleffort. Hence, the DOCE approach is indeed capableof expanding and optimizing the possibilities alreadyachievable by simulation studies alone. For demonstratingthe method on a relatively simple example, this workis focused on designing and validating a metamodel forcalculating linear, one-directional stresses in rectangularmonocrystalline (100) samples. It will be shown that thedifferences between FEM and the metamodel are alwayssmaller than ≈ 4 MPa for different stress states up to amaximum stress of 215 MPa. |
|
|
Die Thickness Impact on Thermo-mechanical Stress in 3D Packages Abdellah Salahouelhadj, Mario Gonzalez, Herman Oprins, IMEC, Leuven, Belgium abstract In this study, Finite Element Modeling (FEM) is used to predict the stress and deformation induced by packaging and temperature hot spots for 3D-IC packages. The studied packages consist of a stack of two Si dies attached with flip chip technology to a laminate in a ball grid array (BGA) configuration. Three packages were considered in this paper: two molded packages with different epoxy mold compounds (EMCs) and one bare die package without EMC. The impact of the bottom die thickness on the stress and package deformation is investigated.The finite element simulation results indicate that thinning the bottom die will cause larger stress and more warpage induced by packaging. Moreover, temperature hot spots cause larger stress and more deformation for thinner bottom dies. Furthermore, the results show that the stress and deformation caused by processing are much higher than those induced by temperature hot spots. |
|
|
Efficient Modeling of Printed Circuit Boards Structures for Dynamic Simulations Elena Zukowski 1, Thomas Kimpel 2, Daniel Kraetschmer 3, Andreas Roessle 1 1 AEEDT3, Robert Bosch GmbH, Germany 2 AEEDS1, Robert Bosch GmbH, Germany 3 AEPJ-SOP, Robert Bosch GmbH, Germany abstract Printed circuit boards (PCB) are complex geometrical and functional systems that may be exposed to a combination of external and internal loads. In order to evaluate the dynamic behavior of PCBs in early stages of the development process, modal finite element (FE) simulations are used. Realistic results for a wide frequency range can only be achieved if all the geometrical features, such as PCB assembly, copper layer thicknesses, prepreg structures, etc. with the appropriate material properties are taken into account. To model a printed circuit board including all details such as glass fiber-epoxy compounds and copper traces is possible, but is found to be very time-consuming.A method to model PCBs was developed taking into account the corresponding functional board layout and assembly. In order to ensure an appropriate representation of the layout-dependent local material properties for FE applications without considering the geometry in full detail, a simplified approach based on general composite theory, domain-specific mixture rules and generalized laminate theory was developed. The analytically calculated material property distributions of the PCB such as local stiffness values and densities can be transferred to the meshed geometry. To verify the developed method by comparison with experimentally achieved results, operational modal analysis (OMA) for a frequency up to 25 kHz was carried out by piezo patch transducer. It can be shown that both simulated mode shapes and natural frequencies of the non-assembled board show a very good agreement with the experimental results.The quality of the method based on an analytical approach to represent PCB layout and assembly allows an easy and fast application for more complex simulations like harmonic response and power spectrum density (psd) analysis with the focus on damage modes of components and APT. |
|
|
Advances in Percolated Thermal Underfill PTU Simulations for 3D- Integration Sridhar Ganesh Kumar 1, Uwe Zschenderlein 1, Remi Pantou 2, Thomas Brunschwiler 3, Gerd Schlottig 3, Florian Schindler-Saefkow 4, Bernhard Wunderle 1 1 Technische Universität Chemnitz, Chair Materials and Reliability of Microsystems, Chemnitz, Germany 2 Fraunhofer Institute for Electronic Nanosystems ENAS, Chemnitz, Germany 3 IBM Research, Zurich, Switzerland 4 AMIC Angewandte Micro-Messtechnik GmbH, Berlin, Germany abstract To satisfy the increasing need in today’s industry for high performance, more complex chips are being designed. These chips, when integrated in 3D packages, have a high energy density and require new and innovative cooling strategies as many of them are designed as flip-chip assemblies, usually requiring back-side cooling. Classical underfills currently used offer poor thermal conductivity. But cooling through the underfill would enable cost-efficient and low complexity cooling solutions. For this purpose, thermal underfills with percolating fillers and necks are currently under development. They are to provide a significant improvement in thermal conductivity to classical capillary underfills and will find applications in, for example, 3D integrated packages to improve heat dissipation. The idea behind the percolating thermal underfill (PTU) comprises a sequential joint forming process ensuring a high fill fraction.Although flip chip technology has been well described, the addition of the neck based percolating underfill could entail several new thermo-mechanical reliability concerns that need to be studied using a physics of failure approach, since the PTU exhibits vastly different thermo-mechanical behavior, giving rise to possible new failure mechanisms and locations. This paper in particular deals with FE simulations carried out to understand different key aspects of the thermal underfill and to study the effects of the increased underfill stiffness at these locations. The simulations are implemented using detailed elastic, plastic, visco-elastic and visco-plastic material data. In case of larger models a complexity reduction is required and implemented by using effective material data to improve computational time. |
|
|
Fracture Mechanical Modeling for the Stress Analysis of DBC Ceramics Patrick Gaiser 1, Markus Klingler 1, Juergen Wilde 2 1 Robert Bosch GmbH, Automotive Electronics, Kusterdingen, Germany 2 Department of Microsystem Engineering IMTEK, University of Freiburg, Freiburg, Germany abstract Nowadays, the progress in power electronics requires the improvement of the reliability of DBC ceramics. The well-documented phenomenon of conchoidal cracking initiates failures at the metallization-ceramic interface. It is a result of the CTE mismatch between metallization and ceramics. Thermal cycling stresses lead to crack propagation which can consequently lead to failure in power devices due to diminished heat dissipation. In this paper, a novel concept was used in order to analyze the thermo-mechanical stresses in DBC ceramics under passive thermal cycling conditions by combining the Finite Element Method and fracture mechanics. Fracture mechanical parameters such as stress intensity factors and the J-integral were calculated with regard to the variation of the dimple depth, the topology of the etched metal edge and the ceramic thickness. Furthermore, this concept was applied to optimize the edge geometry of the metallization with the criterion of stress reduction at the metal-ceramic interface. The concept to minimize local stresses as a basis for reliability improvement will have to be validated experimentally. By this methodology, improvements in substrate technology for future power electronic assembly are made possible. The principle of this study presented here is the basis for a future lifetime prediction. |
|
|
Material Characterization and Process Optimization of Dye-sensitized Solar Cell Sealant Changwoon Han, Seungil Park, Korea Electronics Technology Institute abstract Large-scaled dye-sensitized solar cell (DSC) modules are recently developed for building-integrated photovoltaic (BIPV) applications. In the modules, two glasses with electrodes and dye are sealed together to prevent the leakage of liquid electrolyte. It is known that DSC modules deteriorate rapidly under high temperature conditions. Previous studies showed that expansion of liquid electrolyte in the module is the main reason for the degradation; the expansion of electrolyte induces the breakage of sealant material of DSC module in high temperature. This study investigates how the sealant curing process affects the integrity of DSC module in high temperature. Sealant samples are made up by several UV curing times. Shadow moiré technique is used to measure the coefficient of thermal expansion (CTE) of the sealant samples. With the test results, finite element analyses are conducted to optimize the curing process time. It is finally suggested that the longer the curing time, the more robust the DSC module. |
|
|
Constitutive Modelling of Copper Films on Silicon Substrate Dr. Martin Lederer, Vienna University of Technology, Vienna, Austria Dr. Javad Zarbakhsh, KAI Kompetenzzentrum Automobil- und Industrieelektronik GmbH and Carinthia University of Applied Science, Department of Engineering and IT, Villach, Austria abstract In order to characterize the material behavior of copper films deposited on silicon substrate, wafer curvature experiments were performed. The samples were exposed to repeated cycles in the range between -50°C to 400°C. The diagrams of film stress versus temperature show linear film behavior followed by plastic flow. In fact, a pronounced Bauschinger effect was observed which is attributed to back-stress arising from the dislocation structure in copper films. For better understanding of the underlying mechanisms, a new statistical dislocation model was developed which can nicely be fitted to experiments. However, the algorithm of the dislocation model appeared to be very time consuming during computation. Therefore, a second model was developed which can refit the experimental data with high accuracy using a fast algorithm. We call this model pressure dependent combined isotropic and kinematic hardening. This model was implemented in ANSYS with user-subroutine usermat. |
|
|
Thermo-mechanical Analysis of GaAs Devices under Temperature-Humidity-Bias testing Kokou Adokanou 1, Karim Inal 1, Pierre Montmitonnet 1, Frédéric Courtade 2, Barbara Bonnet 3 1 CEMEF MINES Paristech, Sophia Antipolis, France 2 Centre National d’Etudes Spatiales, Toulouse, France 3 Thales Alenia Space, Toulouse, France abstract Accelerated life tests on microelectronic devices are needed to estimate their degradation under severe environment. THB (Temperature Humidity Bias) at 85°C and 85%RH (relative humidity) is commonly used for reliability studies. Empirical acceleration laws, used for THB test take into account the temperature change (from 22°C to 85°C), but they do not quantify its impact of the corresponding thermo-elastic stress which it adds to the residual stress in the die and of possible microstructure changes. The aim of this work is to determine the thermo-mechanical stresses induced in the active layer of a Gallium Arsenide (GaAs) chip by the THB test. They are due to the mismatch in Coefficients of Thermal Expansion (CTE) between the stack of thin film materials used as metallurgic interconnection and the intermediate dielectric layers above the active area of the chip. To estimate this stress, fist layers thicknesses measurement have been made with various techniques; second few configurations have been used to simulate heating and finally “complete” 2D Finite Element Analysis (FEA) has been performed. Elastic and thermo-physical materials data come from the literature. The results indicate compression of metal gate (Ti/Al/Au) and tensile stress concentration in the SiNx passivation layer. The outcomes is compared with THB test results from previous works and suggests that stress induced by heating must be considered to explain failure during THB test. |
|
|
Numerical Investigation of Ceramic Package Interposer Interconnects using Isotropic Conductive Adhesive R. Fayҫal Hamou 1, Per Dalsjø 2, Cato Dorum 2, Susanne Helland 3, Helge Kristiansen 3, Maaike M.Visser Taklo 4, Jakob Gakkestad 2 1 SINTEF ICT, Microsystems and Nanotechnology MiNaLab, Oslo, Norway 2 Norwegian Defence Research Establishment FFI, Kjeller, Norway 3 Conpart AS, Skjetten, Norway 4 SINTEF ICT, Instrumentation Dept., Oslo, Norway abstract The aim of this study is to investigate the applicability of using an epoxy based isotropic conductive adhesive (ICA), to mount a silicon 3D system-on chip (SoC) in a ceramic 16 pad leadless chip carrier (LCC). We present and discuss thermo-mechanical FEA simulation results obtained by implementing the viscoelastic properties of the adhesive. A generalized Maxwell model using Prony series was considered in this study. We analyse the generated strain and stress in the ICA interconnect as a function of the adhesive geometry for two high and low operating temperatures. The goal is to define an optimal volume and geometry of the cured adhesive with respect to minimized stress at the interfaces between the pads and the adhesive. The target is to reduce the risk of crack initiation and propagation caused by thermal stress by careful design. |
|
|
A Finite Element Modelling and Fracture Mechanical Approach of Multilayer Ceramic Capacitors Joseph Al Ahmar, Steffen Wiese, Saarland University, Germany |
|
|
Acceleration of Lifetime Modeling by Isothermal Bending Fatigue Tests Jens Heilmann, Joerg Arnold, Bernhard Wunderle, Technische Universität Chemnitz, Chemnitz, Germany abstract The generation of meaningful lifetime-models is a seriousand time-consuming challenge throughout the field ofpackaging. Wherever different materials are joined, the CTEmismatch will usually lead to thermo-mechanical fatigue dueto the temperature cycles during the usage of the system [1–3].As a result, the fatigue of interconnections is the limiting factorfor reliability of electronic systems [4]. Usually lifetime investigationsare executed as active or passive thermal cycles usingthe final systems with fixed amplitudes. The main objective israther the validation that the system will exceed a minimumthreshold than the developing of a full lifetime-model. Detailedinvestigations are often bypassed due to time and financiallimitations not realizing the future benefits of a lifetime-model,i.e. by gaining understanding of failure mechanisms and thepossibility to predict them by modelling [5–9].Especially for interfaces based on new developed andmostly insufficiently examined materials like sintered (porous)or composite with their predicted time-depending or highlyanisotropic behavior, more detailed experiments are necessaryto understand the physics of failure. Such results are requiredfor the technology developing and optimization of fatiguebehavior. Therefor more experiments with samples of differenttechnology-parameters as well as different amplitudes orload-regimes are necessary to examine the stability of failuremechanisms and the damage accumulation. New conceptsto conduct such lifetime investigations faster are urgentlyneeded [5]. The idea presented in this paper is to show asuitable method to substitute lengthy thermal cycling tests byresults obtained by rapid isothermal fatigue tests at differenttemperatures and how to establish a correlation between bothof them. For now, samples based on galvanically depositedcopper are used as common reference-material. Based onphysics of failure principles, the applicability and viability ofsuch a concept then is evaluated and discussed. In conclusion,this work shows a approach for a significant acceleration ofthe design for reliability procedure in system integration. It isbased on the now possible rapid generation of a lifetime modelby thin metal layer samples which are easily manufacturablewith the same technology as the thermal cycling test (TCT)samples and should show the same failure mechanism. Detailedinvestigations are still needed to confirm an applicability ofthe method also to other metal layers used in the electronicpackaging industry. |
|
|
Analyses on a Directly Attached Airbag Sensor Packaging System Hyun Jin Kang 1, Jong Dae Lee 2, Joon Ki Kim 2, Yeong K. Kim 3 1 Inha University, Incheon South Korea 2 Korea Institute of Industrial Technology,Incheon South Korea 3 Inha University,Incheon South Korea abstract We propose a directly attached sensor-mounted flexible PCB using an adhesive on the automobile frame for airbag sensor. For performance analyses, experiments and numerical calculations were carried out to simulate the impact signals using an aluminum channel, and the results were characterized. The preliminary results showed that the signals from the directly attached sensor represented high potential to replace the conventional module sensor with better signal transfer. |
Session 12 |
Thermal and Multi-Physics Issues in Microelectronics |
10:40 |
Tuesday April 21 2015 |
Chaired by Bernhard Wunderle, Kaspar Jansen |
|
|
GEANT4 Simulations in Terms of Radiation Hardness of Commercially Available SRAM Aymen Moujbani 1, Kirsten Weide-Zaage 1, Berthold Römer 2, Frank Sabath 2 1 RESRI Group, Institute of Microelectronic Systems IMS, Leibniz Universität Hannover, Germany 2 Bundeswehr Research Institute for Protective Technologies and NBC Protection WIS,Germany abstract In case of the characterization of commercial memory regarding radiation hardness we investigated neutron-induced failure on standard commercial SRAM component with different neutron energy. A Geant4 simulation structure is developed and implemented. The basic idea of the simulation was to applicate the test condition on a single transistor cell to determine the interaction with the matter. A simplified structure was modelled and the interaction determined. The simulations were correlated with the measurements. |
|
|
Electromigration Modelling of Void Nucleation in Open Cu-TSVs Marco Rovitto, Wolfhard H. Zisser, Hajdin Ceric, Tibor Grasser, Vienna University of Technology TU Wien, Vienna, Austria abstract Recently, Through Silicon Vias (TSVs) have attracted much attention in three-dimensional (3D) integration technology due to their function as vertical connections of the different stacked semiconductor dies. Since electromigration (EM) will continue to be a key reliability issue in modern structures, the prediction of the EM failure behavior is a crucial necessity. Traditionally, Black’s equation has been used from the early times of EM investigations for the estimation of the interconnect time to failure. In this work we investigate the applicability of Black’s equation to open copper TSV structures using TCAD. TCAD can significantly contribute to the comprehension of EM failure mechanisms, in particular for the understanding of the early failure mode dominated by the void nucleation mechanism. The simulation procedure is applied to an open copper TSV technology in order to find the sites where void formation is most likely to occur. The time to failure is determined as the time needed to reach the stress threshold for void nucleation. Simulations are carried out for different current densities and successfully fitted to Black’s equation. In this way, we have shown that failure development in studied TSV structures obeys Black’s equation. |
|
|
Comparison of Electrical, Thermal, and Optical characteristics of High-Power LEDs operating in various spectral ranges: from UV to green A.E. Chernyakov 1, A.L. Zakgeim 1, K.A. Bulashevich 2, S.Yu. Karpov 2 1 Submicron Heterostructures for Microelectronics Research Engineering Center, RAS, St. Petersburg, Russia 2 STR Group – Soft-Impact, Ltd., St.Petersburg, Russia abstract The paper reports on influence of the emission wavelength on characteristics and performance of highpower light-emitting diodes (LEDs) emitting light in the near-UV (370 nm) and green (530 nm) spectral ranges and having a similar chip design. Similarity and difference in operation of the LEDs is revealed by their detailed characterization. The correlations between the device characteristics and properties of materials used for the LED fabrication are discussed. |
|
|
LED’s Luminous Flux Lifetime Prediction Using a Hybrid Numerical Approach Kasemsak Kijkanjanapaiboon, Theodore Wagner Kretschmer, Liangbiao Chen, Xuejun Fan, Jiang Zhou, Lamar University, USA abstract Light-emitting diodes (LEDs) have several advantages over traditional incandescent bulbs and compact fluorescent lamps, such as superior energy efficiency, environmental friendliness, and particularly long lifetime (between 25,000 to 100,000 hours). However, this long lifetime of LED proves inconvenient to manufacturers for conducting reliability tests which require the same amount of time to conclude. To overcome such inconvenience, this paper presents a hybrid numerical approach that combines numerical modeling with analytical analysis to predict the lifetime of LEDs. In this paper, a 60W-equivalent 10W phosphor-converted white LED bulb is studied by two numerical approaches. A one-dimensional (1-D) thermal-resistance circuit analysis and a three-dimensional (3-D) hybrid finite element analysis (FEA) are employed to estimate the LEDs’ junction temperature in accord to the data obtained through the experiment. The numerical results showed that both 1-D thermal-resistance circuit and the hybrid FEA model are in agreement with the experiment data, thus invaluable to manufacturers who need to carry out reliability testing. Then the estimated junction temperature is used to determine the LED luminaire’s lifetime according to the known LM-80 database and TM-21 method. |
|
|
Statistical Energy Study for 28nm FDSOI Devices Rida Kheirallah 1, Jean-Marc Galliere 1, Aida Todri-Sanial 1, Gilles Ducharme 2, Nadine Azemard 1 1 LIRMM, UMR CNRS-University of Montpellier, France 2 Dept. of Mathematics, University of Montpellier, France abstract Due to the effects of the Moore’s law, the process variations in current technologies are increasing and have a major impact on power and performance which results in parametric yield loss. Due to this, process variability and the difficulty of modeling accurately transistor behavior impede the dimensions scaling benefits. The Fully Depleted Silicon-On-Insulator (FDSOI) technology is one of the main contenders for deep submicron devices as they can operate at low voltage with superior energy efficiency compared with bulk CMOS. In this paper, we study the static energy on 28nm FDSOI devices to implement sub-threshold circuits. Study of delay vs. static power trade-off reveals the FDSOI robustness with respect to process variations. |
Session 13 |
Materials Modeling |
13:30 |
Tuesday April 21 2015 |
Chaired by Steffen Wiese, Kirsten Weide-Zaage |
13:30 |
30mn |
Keynote presentation – Multiscale modeling of the Anisotropic Transient Creep Response of Heterogeneous SAC Single Crystals Subhasis Mukherjee 1, Bite Zhou 2, Abhijit Dasgupta 1, Thomas Bieler 2 1 University of Maryland, USA 2 Michigan State University, USA abstract This paper provides fundamental mechanistic insights into the significant piece-to-piece variability that many researchers have reported in the creep response of micronscale high-Sn SAC solder joints in the as-fabricated state, due to coarse-grained microstructure and the anisotropy of Sn. A multiscale mechanistic creep modeling approach is proposed, by combining the individual contributions of the eutectic Sn-Ag phase and the pro-eutectic Sn dendritic phase. The anisotropic transient creep deformation in the eutectic Sn-Ag phase is termed Tier 1 and is modeledwith dislocation mechanics. The creep rate of the pure Sn dendritic phase is similarly modeled with dislocationmechanics and combined with that of the eutectic phase, in Tier 2, using an anisotropic load-sharing scheme thatutilizes Eshelby methods and Mori-Tanaka homogenization. The creep rate calculations are performed along the dominant slip systems of the Sn grain in a single crystal of SAC solder material, to obtainthe transient creep response of a SAC305 single crystal along global loading directions. This model has beencalibrated using experimentally obtained transient creep response of a SAC305 single crystal specimen of aparticular orientation and then verified against a second SAC305 single crystal specimen of a different orientation.The effect of grain orientation () on the transient creep response of SAC305 single crystal is parametricallydemonstrated by varying one of the Euler angles of the grain. |
14:00 |
20mn |
Simulation of the Deformation Behaviour of Large Thin Silicon Wafers and Comparison with Experimental Findings Schicker Johannes, Arnold Thomas, Hirschl Christina, Iravani Armin, Kraft Martin, CTR Carinthian Tech Research AG, Villach, Austria abstract The deformation of large thin uncoated silicon wafers without remaining intrinsic misfit stresses resting on a ring is investigated. We use both, FiniteElement simulations and THz tomography mapping. Specific attention is given the scaling of the warping for increasing slenderness of those wafers. We followthe approach of starting with a known solution for a compact wafer and increase the slenderness, i.e. increase the radius and decrease the thickness, usingsimulation models. Then, we measure the warping by THz mapping for some slender wafers and compare the data to simulation results. We compare the maximumwarpage for given loadings and we compare the deflected shapes. Due to the geometric ratio radius/thickness of over 1000:1 and the anisotropic material behaviour, simulations can only be done effectively using shell element modelling of a spatial plate. And due to large warpages in the order of 10 times of the thickness, only incremental update Lagrange nonlinear calculations give reliable results. Simulations using the available shell elements overestimate slightly the values measured by tomography, but still yield acceptable values with errors less than 10% for very slender wafers and below for more compact ones. For invariable loading conditions, a logarithmic scaling function gives an acceptable estimate for the maximum warpage for increasing slenderness. An additional important observation was that the warpage of thin wafers is heavily affected by the size of the contact radius of a weight. |
14:20 |
20mn |
Numerical Analysis of Cure Induced Residual Stress in Die Packaging A.R. Rezaie Adli, K.M.B. Jansen, L.J. Ernst, TU Delft, the Netherlands abstract This paper comprises a numerical constitutive model for evaluating residual stresses generated during encapsulation of integrated circuits. Residual stress is a consequence of molding process which can be divided in cure and thermal induced parts. Cure originated stress had been mostly neglected in literature and a special attention had always been given to detection of thermal induced stresses. In this study, both encapsulation resulted stresses are studied independently and a numerical methodology has been developed based on the applied boundary conditions during each stage of molding and the established process dependent mechanical models. A two dimensional numerical model is implemented in a commercially available software package. The numerically predicted stress results are experimentally validated by implementing a piezoresistive stress measuring chip in a transfer molding process. |
14:40 |
20mn |
Material Characterization and Nonlinear Viscoelastic Modelling of Epoxy Based Thermosets for Automotive Application Gromala Przemyslaw 1, Muthuraman Balaji 1, Oeztuerk Berkan 1, Jansen Kaspar 2, Ernst Leo 3 1 Robert Bosch GmbH, Automotive Electronics, Tuebingerstr. 123, 72762, Reutlingen, Germany 2 Delft University of Technology, IO, Landbergstraat15, 2628 CE Delft, The Netherlands 3 Emeritus Professor of Delft University of Technology abstract This paper presents material characterization utilizing static tensile tests until failure and static tests with relaxation segments until failure for commercially available molding compound. In order to model the material behavior quantitatively a non-linear viscoelastic (NLVE) Bergstrom-Boyce model (BB) is proposed. Material constants of the BB model are optimized utilizing commercially available code optiSLang. In this paper a detailed optimization scheme is presented, including regression and sensitivity analysis. The NLVE model is shown to improve the predictions of the experimental results compared to state-of-the-art linear viscoelastic (LVE) material model. Thus, the BB model is proposed to be used for time, temperature and stress dependent behavior of polymers. |
15:00 |
20mn |
Experimental Thermal and Mechanical Characterisation of Percolating Thermal Underfills PTU for Thermally Enhanced Flip-Chip Packages Uwe Zschenderlein 1, Karthik Suresh 1, Mario Baum 2, Bernhard Wunderle 1 1 Technische Universität Chemnitz, Germany 2 Fraunhofer Institut ENAS, Chemnitz, Germany abstract This paper covers a detailed preliminary study to micro bending tests used to obtain the mechanical properties of nanoparticles as used in advanced electrical or thermal joints. A method for direct, easy, fast and highly accurate transformation of an experimental force-displacement-curve into the corresponding σ-ε-curve is presented. That gives access to the elastic and plastic properties of the material. No additional profiling measurements are necessary. The authors focus on the theoretical background of both, the bending test itself as well as the transformation. Both is discussed in detail by utilizing the results of a finite elements model which simulates a micro bending test. |
Session 14 |
MEMS – Switches, Accelerometers, Sensors |
13:30 |
Tuesday April 21 2015 |
Chaired by Rainer Dudek, Nancy Iwamoto |
13:30 |
30mn |
Keynote presentation – Latching MEMS Switch Matrices for Telecommunication Applications Dries Dellaert, Jan Doutreloigne, CMST, Ghent University – IMEC, Ghent, Belgium abstract This paper describes the design and characterization of a 2×2 latching MEMS switch matrix for switching twisted pairs in telecommunication applications. The RF performance of this fabricated prototype was very promising so larger virtual prototypes of 16×16 were designed and their performance was simulated. From these simulations it is shown that the RF performance of a crossbar switch matrix is not suitable for high frequency access technologies. In order to improve these RF characteristics, a switch matrix with disconnectable stubs is presented and its characteristics are simulated. From these results, the designed 16×16 switch matrix could be used for FTTdp (fiber to the distribution point) deployments using the next generation access technology: G.fast. |
14:00 |
20mn |
Optimal Design and Nonlinearities in a Z-axis Resonant Micro-Accelerometer Claudia Comi 1, Alberto Corigliano 1, Valentina Zega 1, Sarah Zerbini 2 1 Politecnico di Milano, Milano, Italy 2 ST-Microelectronics, Cornaredo Milano, Italy abstract Micro-Electro-Mechanical Systems (MEMS) accelerometers are micro-sized devices largely used for detecting accelerations in the consumer and automotive market. Both capacitive and resonant sensing have been successfully employed in these devices.In the present work, we focus on a z-axis resonant accelerometer recently proposed, fabricated by the Thelma© surface-micromachining technique developed by STMicroelectronics.After a full non-linear dynamic study, an optimization of the design of the device is carried out. The main goal of the optimization process is to increase the sensitivity of the device together with the reliability and the linearity. |
14:20 |
20mn |
Modelling and Characterization of Circular Electrostatic Actuators Raffaele Ardito, Emanuele Bertarelli, Alice Colnago, Alberto Corigliano, Gabriele Dubini, Politecnico di Milano, Milan, Italy abstract In the last decades, an increasing interest is being directed towards micropumps, the main component of active microfluidic systems. A current research aim is to pursue compact dimensions and low power consumption; a viable alternative is to exploit MEMS manufacturing techniques. This work presents the modelling and the characterization of microplate electrostatic actuators for micropump applications. A one degree-of-freedom model is proposed to describe deformable plate electro-mechanics. The well-established ThELMA technique (STMicroelectronics) is adopted to manufacture the actuator prototypes. The effective plate flexural stiffness is extracted from pull-in tests on a series of plate actuators. The structural model is calibrated accordingly and successfully used to describe the electromechanical response of actuators where an annular electrode is introduced in order to obtain a higher actuator stroke. |
14:40 |
20mn |
Fast Analytical Design of Poly-SiGe MEMS Pressure Sensors Veronique Rochus, Bo Wang, Ashesh Ray Chaudhuri, Philippe Helin, Simone Severi, Xavier Rottenberg, imec, Leuven, Belgium abstract This paper presents a fast design strategy for Poly-SiGe MEMS pressure sensors, based on circular Kirchhoff-Love plate theory. The underlying analytical model allows for a rapid and accurate evaluation of the sensitivity of the sensors, crucial for improving their design. The accuracy of the new model is demonstrated by comparing its predictions with more computationally expensive simulation techniques (high-order parametric element and three-dimensional finite element models) and with experimental measurements performed on a 300µm membrane fabricated using the Poly-SiGe platform developed at imec. |
15:00 |
20mn |
Packaging Effects of Three-axis SOI MEMS Accelerometer Hung-Te Yang, Yen-Fu Su, Kuo-Ning Chiang, Advanced Microsystem Packaging and Nano-Mechanics Research Lab Dept. of Power Mechanical Engineering, National Tsing Hua University, Hsinchu, Taiwan, R.O.C abstract This paper presents the packaging and residual stress effects on three-axis silicon-on-insulator (SOI) micro-electro-mechanical system (MEMS) accelerometer by using finite element method (FEM). The 3D FEM model was established and the resonance frequency was obtained by modal analysis method. This paper also developed a simple compensation model for trimming the offset of capacitance differentiation by measuring resonance frequency. It can be trimmed by adjusting application-specific integrated circuit (ASIC) gain. The capacitance differentiation offset which is caused by packaging effect can be effectively compensated to the standard capacitance differentiation. |
Session 15 |
Thermo-Mechanical Simulations |
13:30 |
Tuesday April 21 2015 |
Chaired by René Metasch, Jan Albrecht |
13:30 |
30mn |
Keynote presentation – FEA Study of Damage and Cracking Risks in BEoL Structures under Copper Wirebonding Impact J. Auersperg 1, D. Breuer 2, K.V. Machani 2, S. Rzepka 1, B. Michel 1 1 Micro Materials Center at Fraunhofer ENAS, Chemnitz, Germany 2 GLOBALFOUNDRIES Dresden Module One LLC Co. KG, Dresden, Germany abstract With the recent increase in Gold (Au) wire cost; Copper (Cu) wire becomes an attractive way to manage overall package cost. On the other hand, Copper wire bonding introduces much higher mechanical impact to underlying BEoLstructures and actives because of the higher stiffness and lower ductility of Copper compared to Gold. These trends are accompanied by the application of new porous or nano-particle filled materials like low-k and ultra low-k materials for Back-end of line (BEoL) layers of advanced CMOS technologies. As a result, higher dela¬mination and cracking risks in BEoLstructures underneath bonded areas represent an increasing challenge for the thermo-mechanical reliability require-ments. To overcome the related reliability issues the authors performed a two level nonlinear FEM-simulation approach. Initially nonlinear axisymmetric modeling and simulation of the copper bonding process are coupled with a spatial simulation model of the whole BeoL and bond pad structure. Cracking and delamination risks are estimated by a surface based cohesive contact approach and the utilization of a crushing foam constitu¬tive material model for ultra low-k materials. |
14:00 |
20mn |
Laser Cutting of Semiconductor Wafers: comparing a Simplified Numerical Approach with Experiments M. van Soestbergen, J.J.M. Zaal, F.H.M. Swartjes, J.H.J. Janssen, NXP Semiconductors, Nijmegen, the Netherlands abstract Laser grooving is used for the singulation of advanced CMOS wafers since it is believed that it exerts lower mechanical stress than traditional blade dicing. The very local heating of wafers, however, might result in high thermal stress around the heat affected zone. In this work we present a model to predict the temperature distribution, material removal, and the resulting stress, in a sandwiched structure of metals and dielectric materials that are commonly found in the back-end of line of semiconductor wafers. Simulation results on realistic three dimensional back-end structures reveal that the presence of metals clearly affects both the ablation depth, and the stress in the material. Experiments showed a similar observation for the ablation depth. The shape of the crater, however, was found to be more uniform than predicted by simulations, which is probably due to the redistribution of molten metal. |
14:20 |
20mn |
Efficient Simulation of Thermo-Mechanical Stress in the On-Chip Metallization of Power Semiconductors Gimi Pham, Martin Pfost, Robert Bosch Center for Power Electronics, Reutlingen University, Reutlingen, Germany abstract Large power semiconductors are complex structures, their metallization usually containing many thousands of contacts or vias. Because of this, detailed FEM simulations of the whole device are nowadays not possible because of excessive simulation time. This paper introduces a simulation approach which allows quick identification of critical regions with respect to lifetime by a simplified simulation. For this, the complex layers are replaced by a much simpler equivalent layer, allowing a simulation of the whole device even including its package. In a second step, precise simulations taking all details of the structure into account are carried out, but only for the critical regions of interest. Thus, this approach gives detailed results where required with consideration of the whole structure including packaging. Further, the simulation time requirements are very moderate. |
14:40 |
20mn |
Prediction of Package Delamination Based on µMMT and BST Experiments Hasan Sadat Nabi 1, Dirk Schweitzer 2, Duc-Khoi Vu 2, Ingrid Maus 1, Laurens Weiss 2 1 Infineon Technologies AG, Regensburg, Germany 2 Infineon Technologies AG, Neubiberg, Germany abstract The numerical simulation and prediction of interfacial delamination in electronic packages using the finite element method requires a correct understanding of the failure and an accurate characterization of the materials involved at the interface. In this work, experiments were realized to characterize the interfacial adhesion and fracture toughness of copper-to-epoxy-molding-compound interfaces. Adhesion parameters for different combinations of coppers and molding compounds were extracted from the so-called micro mixed mode test (µMMT) and button shear test (BST). Finite element simulations using cohesive zone modeling (CZM) were used to predict the delamination behavior of different micro-electronic packages being subjected to half temperature cycle tests. The results of the numerical simulation and prediction were verified experimentally. |
15:00 |
20mn |
The Influence of PBGAs Post-Cure Warpage Simulation Viscoelastic Material Properties and Evaluation Methodology on Accuracy of Solder Joints Damage Prediction Marta Kuczynska, Natalja Schafet, Manfred Spraul, Ulrich Becker, Robert Bosch GmbH AEEDE3, Schwieberdingen, Germany abstract Two Plastic Ball Grid Array (PBGA) designs were stressed with Temperature Cycles (TC) in the Board Level Reliability (BLR) test, and simulated using FEM. The test result for one of the PBGAs was not in line with original expectation, but it was confirmed by FE-simulation. The simulation incorporated manufacturing process (the real post-cure shape verified by measurement) and viscoelastic material properties. The impact of model simplifications (i.e. omitting manufacturing process simulation or viscoelastic material properties) on the thermo-mechanical load of the solder joints was investigated. The opportunities and restrictions of these simplifications are shown and discussed.The current work also describes an improvement of the strategy for evaluation of the simulation results. The correlation between the test result (crack lengths observed via cross sections) and the local damage-related state variables from the FE-simulation was improved when effects of hydrostatic stress and cross section direction were included in the evaluation. |
Session 16 |
Failure Mode Extraction |
15:40 |
Tuesday April 21 2015 |
Chaired by Marco Barink, Michel Lenczner |
15:40 |
30mn |
Keynote presentation – Thermal Deformation Analysis of Automotive Electronic Control Units Subjected to Passive and Active Thermal Cycles Bulong Wu 1, Dae-Suk Kim 1, Bongtae Han 1, Alicja Palczynska 2, Przemyslaw Gromala 2 1 Mechanical Engineering Department, University of Maryland College Park, USA 2 Robert Bosch GmbH, AEEDT3, Reutlingen, Germany abstract Transfer molding as a widespread and cost effective encapsulation process technology can be implemented to protect the automotive electronic control unit (ECU). An outer EMC can cover entire ECU which is supposed to ensure reliability under harsh environment to replace the conventional metal case. In this work, effects of outer EMC on the thermal deformation of ECU were studied using moiré interferometry and FEM modeling. Two sets of ECU specimens molded with and without outer EMC were subjected to passive and active thermal conditions, respectively. Two orthogonal in-plane displacement fields on cross sections of specimens were documented at various temperatures experimentally. The results of passive condition case were used to verify the complex FEA modeling of the units and the initial material properties were calibrated. The accuracy of FEA modeling after calibration was applied and confirmed by the results from experiment under active thermal condition. The verified modeling and stress data obtained from silicon based sensor inside ECU during experiments can be further utilized to assess the reliability of ECU molded with outer EMC. |
16:10 |
20mn |
Fracture Risk Assessment of Laser Marked Die by Means of Simulation and Test Ilko Schmadlak 1, Betty Yeung 2, Derek Morgan 2, Paul Galles 3 1 Freescale Halbleiter Deutschland GmbH, Munich, Germany 2 Freescale Semiconductor Inc., Phoenix, Arizona, USA 3 Freescale Semiconductor Inc., Austin, Texas, USA abstract A presentation of new methodology that allows obtaining stress concentration results by simulating laser marked die in great detail. |
16:30 |
20mn |
An In-situ Numerical-experimental Approach for Fatigue Delamination Characterization in Microelectronic Packages Emad A. Poshtan 1, Sven Rzepka 2, Christian Silber 1, Bernhard Wunderle 3 1 Robert Bosch GmbH, Automotive Electronics, Reutlingen, Germany 2 Fraunhofer Institute ENAS, Dept. Micro Material Center, Chemnitz, Germany 3 Chemnitz University of Technology, Chemnitz, Germany abstract An in-situ and cost-effective numerical-experimental approach for fatigue characterization of bi-material interfaces in Microelectronic Packages is presented. In this method using a sample-centered approach a Miniaturized Sub-Critical Bending (MSCB) test setup is designed and fabricated based on the samples that are acquired directly from production-line. The accuracy of the results and stiffness of the test-set up is validated using digital image correlation method. The delamination growth is measured using a compliance-based numericalexperimentalmethod under sub-critical cyclic loading. The critical and threshold toughness values are measured. The sample are examined after tests using EDX and SEM measurements. The fractographical study of samples shows that, although some of the molding compound particles are left on the LF surface, the interfacial fracture is the dominant failure mode. |
16:50 |
20mn |
Simulation Driven Design of Novel Integrated Circuits -Part 1: Selection of the materials based on the Virtual DoE Arun Sasi 1, Arjun Yadur 1, Gromala Przemyslaw Jakub 2 1 Robert Bosch Engineering and Business Solutions Private Limited RBEI, Bangalore, India 2 Robert Bosch GmbH, Automotive Electronics, Tuebingerstr. 123, Reutlingen, Germany abstract In the present paper an example is taken up to illustrate the selection of the molding compound for simplified DPAK model using Virtual Design of Experiment (VDoE) technique. Five parameters: coefficient of thermal expansion below and above glass transition temperature (Tg), glass transition temperature, modulus of elasticity and thermal conductivity; are investigated. Using central composite faced (CCF) plan DoE plan, 52 simulation legs are defined. This covers the entire range of the molding compounds available in the market. The results of the 52 simulation cases are evaluated using ANSYS and later on regression analysis is conducted in Cornerstone tool. The coorelation co-eficients from the regression anaysis is imported to MS Excel to recreate the response surface. The graphical representation of simulation and DoE results in the easy to use Excel tool enables the process engineers to obtain a better understanding of the implications of the varying material properties on the design. This principle can be extended to selecting optimum geometric configuration from many design alternatives. This technique & method applied is an enabling tool in simulation driven design concept resulting in reduced time to market with improved development efficiency. |
Session 17 |
Solid State Lighting |
15:40 |
Tuesday April 21 2015 |
Chaired by Xuejun Fan, Zhuangjian Liu |
15:40 |
30mn |
Keynote presentation – Modeling of LED Solder Joint Cracking during Temperature Cycling with Finite Element FE Wenbo Yuan, Paola Altieri-Weimar, Osram Opto Semiconductors GmbH Regensburg, Germany abstract In this study a new simulative and analytic method is developed to determine the failure of LED package solder joints at temperature cycling (TC). The solder joint reliability is calculated using a crack growth model, which is based on a combination of cohesive zone modeling and solder creep simulation in FE-model. The crack growth model is calibrated using shear test data after TC loading. The reliability model is validated by means of TC experimental results for three different TC conditions. |
16:10 |
20mn |
Accelerated Reliability Test method for Optics in LED Luminaire Applications M. Yazdan Mehr, Materials innovation institute M2i, Delft, The Netherlands and Delft University of Technology, EEMCS Faculty, Delft, The Netherlands W.D. van Driel, Delft University of Technology, EEMCS Faculty, Delft, The Netherlands 3 Philips Lighting, Eindhoven, The Netherlands G.Q. Zhang, Delft University of Technology, EEMCS Faculty, Delft, The Netherlands abstract A high accelerated stress testing (HAST) system is introduced to study the photo-thermal stability and reliability of remote phosphor plates, made from Bisphenol-A polycarbonate (BPA-PC) and YAG. Remote phosphor plates, combined with a blue-light LED source, are used to produce white light with a correlated colour temperature (CCT) of 4000 K. In this study, the remote-phosphor BPA-PC samples of 3 mm thickness were photo-thermally aged at temperature range 80 to 120 ºC. The blue light is radiated on the sample with light intensity of 13200 W/m2. Thermal quenching of the YAG samples is also studied. It is shown that crystallographic structure of phosphor is stable during thermal ageing. |
16:30 |
20mn |
Lumen Maintenance Predictions for LED Packages using LM80 Data W.D. van Driel 1, M. Schuld 2, B. Jacobs 1, F. commissaris 1, J. van der Eyden 1, B. Hamon 1 1 Philips Lighting, The Netherlands 2 CQM, The Netherlands abstract Per today, commercial claims for LED-based products in terms of lumen maintenance are fully based on TM21 extrapolations using LM80 data. There may be a risk in doing this as TM21 only relies on the behavior of the average LED degradation, instead of taking into account the degradation of all individual LEDs. A more profound statistical analysis is required to make the step from TM21 extrapolation to lumen maintenance on product level. This is needed as the commercial claims are used as input for service bids up to periods of 20 to 25 years of operation. This paper describes the different approaches currently known to perform lumen maintenance extrapolations. We have analyzed several LM80 data sets from a statistical point of view. The main question is if there is sufficient evidence for lumen maintenance claims up to and/or beyond 100khrs. |
16:50 |
20mn |
A Degradation Model of Aluminum Electrolytic Capacitors for LED Power Supplies Bo Sun 1, Xuejun Fan 2, C.A. Yuan 3, Cheng Qian 3, Guoqi Zhang 3 1 Beijing Research Center, Delft University of Technology, The Netherlands 2 Lamar University, USA 3 Institute of Semiconductors, Chinese Academy of Sciences, China abstract The failure of aluminum electrolytic capacitors is considered as one of major failure modes of the LED drivers. This paper propose a degradation model of aluminum electrolytic capacitors considers impacts of operation time and temperature. |
Session 18 |
Experimental Investigations |
15:40 |
Tuesday April 21 2015 |
Chaired by Jürgen Auersperg, Michiel van Soestbergen |
15:40 |
30mn |
Keynote presentation – High Strain-Rate Constitutive Behavior of SAC105 and SAC305 Leadfree Solder During Operation at High Temperature Pradeep Lall 1, Di Zhang 1, Vikas Yadav 1, David Locker 2 1 Auburn University, NSF CAVE3 Electronics Research Center, USA 2 US Army AMRDEC, Hunstville, USA abstract Industry migration to leadfree solders has resulted in a proliferation of a wide variety of solder alloy compositions. The most popular amongst these are the Sn-Ag-Cu family of alloys like SAC105 and SAC305. Electronics subjected to shock and vibration may experience strain rates of 1-100 per sec. Electronic product may often be exposed to high temperature during storage, operation and handling in addition to high strain rate transient dynamic loads during drop-impact, shock and vibration. Properties of leadfree solder alloys at high strain rates at low and high temperatures experienced by the solder joint during typical mechanical shock events are scarce. Previous studies have showed the effect of high strain rates and thermal aging on the mechanical properties of leadfree alloys including elastic modulus and the ultimate tensile strength. The ANAND viscoplastic constitutive model has been widely used to describe the inelastic deformation behavior of solders in electronic components. In this study, SAC105 and SAC305 leadfree alloys have been tested at strain rates of 10, 35, 50 and 75 per sec at various operating temperatures of 50°C, 75°C, 100°C and 125°C. Full-field strain in the specimen have been measured using high speed imaging at frame rates up to 75,000 fps in combination with digital image correlation. The cross-head velocity has been measured prior-to, during, and after deformation to ensure the constancy of cross-head velocity. Stress-Strain curves have been plotted over a wide range of strain rates and temperatures. Experimental data for the pristine specimen has been fit to the ANAND’s viscoplastic model. |
16:10 |
20mn |
Fracture Mechanics of Thin Film Systems on the Sub-Micron Scale Darjan Kozic 1, Ruth Treml 2, Ronald Schöngrundner 1, Roland Brunner 1, Daniel Kiener 2, Johannes Zechner 3, Thomas Antretter 4, Hans-Peter Gänser 1 1 Materials Center Leoben, Leoben Austria 2 Department Materials Physics, Montanuniversität Leoben, Leoben Austria 3 EMPA, Swiss Federal Laboratories of Materials Science and Technologz, Thun, Switzerland 4 Institute of Mechanics, Montanuniversität Leoben, Leoben, Austria abstract Novel design of microelectronic components creates new issues concerning their reliability. Internal mechanical loading, e.g. from residual stresses, or external loading when the component is assembled into a microelectronic device, can cause failure via cracking or delamination. In this work, finite element simulations of micro-beam bending experiments for testing the fracture behavior of thin film metal composites deposited on a silicon substrate are presented. Due to the lattice mismatch between the materials, residual stresses are generated. Calculating the magnitude and distribution of these stresses is very important, as they add to the stresses produced by the external loads. Consequently, a stress free setting will behave differently compared to a structure with residual stresses. Additionally, crack propagation is affected by an interface to a material with different characteristics. In what follows, the variation of typical fracture parameters will be shown, depending on the residual stresses in the composite and on the crack position relative to the interface. |
16:30 |
20mn |
Thermal and Mechanical Behavior of an RFID Based Smart System Embedded in a Transmission Belt Determined by FEM Simulations for Industry 4.0 Applications Jan Albrecht, Rainer Dudek, Jürgen Auersperg, Remi Pantou, Sven Rzepka, Fraunhofer ENAS, Micro Materials Center, Chemnitz, Germany abstract The determination of the mechanical and thermo-mechanical behaviour of a UHF-RFID-based smart system has been the goal of the work reported in this paper. Complex bending and thermal loads are taken into account by finite element simulations of fabrication and operation scenarios using ABAQUS standard TM. In order to achieve quantitatively correct results, dynamic mechanical analyses using DMA Q800, DMA 2000+ as well as thermo-mechanical analyses using TMA Q400 have been performed to characterize the behaviour of the different materials. The computational results obtained using finite element analyses match the experimental observations very well. Based on the results, full validation of the simulation approach could be achieved and recommendations for design optimization could be deduced, which help preventing early and fatigue failures of the smart system inside the transmission belt. |
16:50 |
20mn |
Investigation of Sensitive Parameters in the Structural Simulation Tool Chain for Fiber Reinforced Plastics in Automotive Electronic Control Units Rongsi Wang 1, Dorothea Papathanassiou 1, Matthias Werner 2, Jing Jin 3 1 Bosch Automotive Products Suzhou Co., Ltd, Suzhou, China 2 Robert Bosch GmbH, Reutlingen, Germany 3 BASF China Co., Ltd., Shanghai, China abstract In automotive electronic control units (ECU) the use of fiber reinforced plastics is widely spread. In contrast to the common utilization of these materials their influence on the reliability of electronic or interconnecting components on Printed Circuit Board (PCB) is not yet sufficiently accounted for in design decisions in early stages of the product development. Fiber reinforced plastic parts exhibit a strong anisotropic macroscopic material behavior. Their deformation under thermal and mechanical load cannot be correctly predicted by simulation if oversimplified material properties are applied. However, this deformation behavior can dominate the strain and stress induced in electronic or interconnecting components on PCB and hence their reliability over lifetime. By means of an integrative simulation approach the influence of the microscopic fiber orientation on the macroscopic behavior of plastic parts can be evaluated. In this paper a study of sensitive parameters of the integrative simulation and their influence on the simulation quality and accuracy is presented. For this purpose test specimen have been cut from an ECU cover which is made of reinforced polybutylene terephthalate with 30 wt% short glass fiber (PBT-GF30). Subsequently, uni-axial, quasi-static tensile tests as well as structural finite element (FE) simulations, comparable to the experiments and including the mapping of fiber orientation information from injection molding simulation were performed. |
Session 20 |
Solders |
11:00 |
Wednesday April 22 2015 |
Chaired by Alberto Corigliano, Nadine Azemard |
11:00 |
30mn |
Keynote presentation – Characterization and Simulation of LTCC Adhesive and Alloy42 Adhesive Interface Strength for Automotive Applications Oeztuerk Berkan 1, Lou Panpan 1, Gromala Przemyslaw 1, Silber Christian 1, Jansen Kaspar 2, Ernst Leo 3 1 Robert Bosch GmbH, Automotive Electronics, Reutlingen, Germany 2 Delft University of Technology, IO, CE Delft, The Netherlands 3 Emeritus Professor of Delft University of Technology, The Netherlands abstract Thermoset-based adhesives are used as thermal and electrical interfaces. In automotive applications, they are required to have excellent adhesion since delamination may precipitate other electrical, thermal or mechanical failure mechanisms. A vast amount of literature is available on the investigation of molding compounds and various material interfaces. However, only very few studies focus on delamination of adhesive interfaces. The reason is that apparently it was not possible to initiate an interface crack in a delamination sample. In various attempts, random cracking in the adhesive was obtained instead. Yet interface cracks are found in real products and really form a reliability issue. But so far the absence of adequate interface strength data makes it hardly possible to design for reliability of products with adhesive interfaces.The present paper solves the above problem. We succeeded to get an interface delamination between the adhesive and two different materials (e.g. Low temperature cofired ceramic (LTCC) and alloy 42). The specimens are made by identical fabrications processes as during the fabrication of the electronic control unit under study. The interface to be investigated is preconditioned for delamination initiation, by adding a single step to the fabrication process, thus enabling the investigation of different interfaces that have the same processing conditions as the real product. The presented specimen preparation method and the testing methodology can be used for determination of critical adhesion properties of different interfaces (including brittle materials like LTCC) in electronic control units. Specimens are investigated by delamination experiments near Mode-I loading conditions at room temperature. The obtained interface data is interpreted via image processing and finite element modeling of the J-integral method. In particular, cohesive zone modeling is used to validate the critical energy release rates for different interfaces. |
11:30 |
20mn |
Surrogate Model Based Mechanical Characterization of Lead-free Soldered Joint Material Exhibiting Ratcheting Behavior: an Advanced Methodology Benoît Dompierre, Ludovic Barrière, Arnaud François, Eric Wyart, Cenaero ASBL, Gosselies, Belgium abstract This study is focused on the methodology dedicated to the identification of parameters of a general Chaboche model for a nanosilver sintered material. This material is used, in particular, for die-attachment in power electronics applications under harsh temperature conditions. This material model is an elastic-viscous-plastic constitutive model accounting for creep and kinematic hardening mechanisms which underlie ratcheting behavior. The proposed methodology relies on an evolutionary algorithm which is coupled with state-of-the-art surrogate modeling. The implemented procedure enables an efficient and robust identification of the parameters of the constitutive model. The methodology is applied to the characterization of a sintered nanosilver joint. Experimental data are extracted from literature in the case of a classical lap-shear test. The Chaboche model is identified for three temperatures from 25°C to 325°C. A monotonic variation of the parameters in function of the temperature is imposed in order to ease the interpolation of parameters inside the temperature range. The parameters identified for the Chaboche model present a better correlation with experiments for each temperature than the Ohno-Wang and Anand constitutive models identified in literature. |
11:50 |
20mn |
A Viscoplastic-Fatigue-Creep-Damage Model for Tin-based Solder Alloy Benjamin Métais 1, Alexander Kabakchiev 2, Youssef Maniar 3, Michael Guyenot 1, René Metasch 4, Mike Roellig 4, Philipp Rettenmeier 5, Patrick Buhl 5, Weihe Stefan 5 1 Robert Bosch GmbH, Stuttgart, MPA Stuttgart University, Germany 2 Robert Bosch GmbH, Stuttgart, Germany 3 Robert Bosch GmbH, Stuttgart, IMWF Stuttgart University, Germany 4 Fraunhofer IKTS, Dresden, Germany 5 MPA Stuttgart University, Germany abstract During the past decade the demand for high performance automotive electronics is steadily increasing. An efficient development of such products requires the use of durability assessment techniques throughout the whole design optimization process. Since typical components comprise a large number of different materials and complex geometrical structures, Finite Element (FE) analysis is preferably used for durability evaluation and continuously replaces analytical calculations. However, a direct lifetime calculation by means of FE-techniques is still not achieved, partly due to the lack of material models capable of mapping the intrinsic material degradation under the relevant thermo-mechanical loads. Here, we propose a material model for a tin-based solder alloy which describes the non-linear mechanical behavior at the beginning of deformation as well as during continuous cyclic aging. We investigated the evolution of the mechanical properties and microstructure of the solder alloy Sn$_{96.5}$Ag$_{3.5}$ by cyclic strain-rate controlled fatigue- and creep-tests on as-casted standardized specimens. Material modeling is focused on the description of the complex interplay between viscoplastic, fatigue and creep processes observed in the experiment. Finally, a very good agreement is obtained between the measurements and the numerical model, which can offer new opportunities for lifetime simulations of lead-free solder joints. |
12:10 |
20mn |
Experimental Investigation on the Influence of Microstructure on Visco-Plastic Mechanical Properties of a Sn-based Solder Alloy for Material Modelling in Finite Element Calculations R. Metasch 1, R. Schwert 1, M. Röllig 1, A. Kabakchiev 2, B. Metais 2, K.-J. Wolter 3 1 Fraunhofer Institute for Ceramic Technology and Systems, Branch Material Diagnostics, Germany 2 Robert Bosch GmbH, Germany 3 Technische Universität Dresden, Electronics Packaging Laboratory, Germany abstract The paper presents experimental results on tin-based solder alloys to their mechanical visco-plastic deformation behaviour under systematically investigation of cooling rates and their micro-structural solidification.We developed a novel process to produce solder bulk specimens in a re-melting process under specific cooling rates up to -300 K/min. The paper shows a comparison of SnAg3.5 and SnAg3.8cu0.75 solidified with -20 K/min and -200 K/min as well as a SnSbCu alloy solidified with -20 K/min and -100 K/min. By contrast to a commonly used passive cooling solution the increased cooling rates are closer to an actual industrial soldering process.The metallographic investigation shows significant changes of the micro-structure with increasing grain quantity while their size decreased. The intermetallic sizes are reduced and the surface roughness of the specimens overall decreased with higher cooling rates.The mechanical comparison of the different produced specimens uses an advanced experimental procedure to determine the material properties for a unified visco-plastic constitutive model initially proposed by Chaboche et al. The constitutive model describes the time-dependent material behaviour in the strain range of primary creep under cyclic load and isothermal conditions. This progress is performed in a temperature range between -40 °C up to 150 °C, with varying strain rates between 1E-3 to 1E-6 per second and relaxation steps. The detailed characterization procedure has been presented in 2014. In two separate chapters the paper explains the advantages of this modelling approach on lifetime prediction using finite-element simulations. |
12:30 |
20mn |
Investigation of Relaxation Behavior by Piezoresistive Stress Sensor Alicja Palczynska 1, Przemyslaw Jakub Gromala 1, Dirk Mayer 2, Bongtae Han 3, Tobias Melz 2 1 Robert Bosch GmbH, Reliability Modeling and System Optimization AEEDT3 Reutlingen, Germany 2 Fraunhofer-Institut für Betriebsfestigkeit und Systemzuverlässigkeit LBF, Darmstadt, Germany 3 Mechanical Engineering Department, University of Maryland College Park, USA abstract In this study, a piezoresistive silicon-based stress sensor is used to understand the stress state in an ECU, more specifically the relaxation behavior of EMC caused by the storage time of an ECU (i.e., duration between production and actual usage). The experimental results clearly indicate that the proposed approach can be used for better understanding of the evolution of stresses in molded packages during their lifetime, especially during storage, which in turn can lead to more optimal designs in the future. |
Session 21 |
Multi-Scale/Physics Modelling of MEMS Devices |
11:00 |
Wednesday April 22 2015 |
Chaired by Abhijit Dasgupta, Karsten Meier |
11:00 |
30mn |
Keynote presentation – A Multi-scale Model of Micro-Mirror Array Duy Duc Nguyen 1, Walid Belkhir 1, Nicolas Ratier 2, Bin Yang 1, Michel Lenczner 3, Frédéric Zamkotsian 4, Horatiu Cirstea 5 1 FEMTO-ST, Time and Frequency Department, University of Franche-Comté, France 2 FEMTO-ST, Time and Frequency Department, Ecole Nationale Supérieure de Mécanique et de Microtechniques, France 3 FEMTO-ST, Time and Frequency Department, Technical University of Belfort-Monbéliard, France 4 LAM-CNRS, France 5 University of Lorraine – LORIA, France abstract This paper reports recent advances in the development of a symbolic asymptotic modeling software package MEMSALab which will be used for automatic generation of asymptotic models for arrays of micro and nanosystems. First, an asymptotic model for the stationary heat equation in a Micro-Mirror Array developed for astrophysics is presented together with some key elements of its derivation. This illustrates the mathematical operations that need to be implemented in MEMSALab. The principle of operation of this software is to construct models incrementally so that model features can be included step by step. This idea conceptualized under the name “by extension-combination” is presented for the first time after having recalled the general functioning principles. A friendly user language recently introduced is also shortly discussed. |
11:30 |
20mn |
Design of thin-film AlN actuators for 4 inches transparent plates for haptic applications F. Casset 1, JS. Danel 1, C. Chappaz 2, F. Bernard 3, S. Basrour 3, B. Desloges 1, S. Fanget 1 1 CEA-LETI, MINATEC Campus, France 2 STMicroelectronics, France 3 TIMA, France abstract Numerous applications require tactile interfaces today. In particular, many customers’ applications such as Smartphone, tablet PC or touch pad can be concerned by high performances, low voltage haptic interfaces which allow the user to interact with its environment by the sense of touch. This technology is already used but with limitations such as high power consumption and limited feedback effect (simple vibration). We chose to work on the squeeze-film effect. It consists in changing the friction between the finger and a plate resonator. It provides high granularity level of haptic sensation. This paper deals with the design of high performances actuators in order to promote the squeeze-film effect on a 4-inch plate (diagonal of the plate). Using predictive models, we select the best design, able to generate the highest substrate displacement amplitude as possible. We built demonstrators using a generic technology based on thin-film AlN actuators. Electromechanical characterization is ongoing before the integration of the thin-film actuator plate in a haptic demonstrator in a close future. |
11:50 |
20mn |
Predicting Non-Fickian Moisture Diffusion in EMCs for Application in Micro-electronic Devices M. Barink 1, A. Mavinkurve 1, J. Janssen 2 1 TNO, Materials Solutions, Eindhoven, The Netherlands 2 NXP, Package Innovation, Nijmegen, The Netherlands abstract This study made an attempt to predict the temperaturedependent moisture diffusion of an epoxy molding compound with 3 different diffusion models: Fickian, dual stage and Langmuir diffusion. The Langmuir model provided the best prediction of the moisture diffusion when simulating the input experiments. Beyond the temperature range of the input experiments, the Langmuir model was still able to provide a fair prediction. Hence, the Langmuir model also provides better predictions for the moisture distribution in general. This allows for building on existing prediction models, enablingsimulations of reliability tests like UHST. |
12:10 |
20mn |
A Simple Constitutive Model for Dielectric Charging Based on Frenkel-Poole Mechanism Tal Rubin 1, George Papaioannou 2, David Elata 1 1 Technion – Israel Institute of Technology, Haifa, Israel 2 University of Athens, Athens, Greece abstract We present a new simple constitutive model fordielectric charging. The model is motivated by thefunctional form of the Frenkel-Poole conduction, but itenables to track the time evolution of charge distributionwithin the dielectric. The prediction of charge distributionis used to deduce the voltage buildup in the dielectric.Based on this model, we present a prediction of thecurrent through a Metal-Insulator-Metal (MIM) structurewhen it is subjected to voltage loading, and present aprediction of the discharge process and its limitations. Wepropose a voltage-control measurement protocol for theMIM structure, to calibrate the two material parameters ofthe model. |
12:30 |
20mn |
Finite Element Modeling of ZnO Nanowire with Different Configurations of Electrodes Connected to External Capacitive Circuit for Pressure Sensing Applications Rolanas Dauksevicius 1, Rimvydas Gaidys 1, Eoin O’Reilly 2, Masoud Seifikar 2 1 Kaunas University of Technology, Kaunas, Lithuania 2 Tyndall National Institute. Department of Physics, University College Cork, Ireland abstract This paper reports the results of finite element modeling and analysis of a vertically-aligned ZnO nanowire including surrounding chip components (seed layer, insulating top layer and metal electrodes), taking into account the influence of external capacitance and considering different nanowire morphologies and electrode topographies in order to predict magnitude of electrical outputs as a function of applied dynamic load (compression and/or bending). The length and diameter of the modeled nanowire is in the µm and sub-nm range, respectively and it is intended to function as a single “piezo-pixel” in a matrix of interconnected ZnO nanowires performing dynamic pressure sensing, which could be used for ultraprecise reconstruction of the smallest fingerprint features in highly-reliable security and ID applications. |
compiled on Thu, 16 Apr 2015 11:16:54 +0200 after template 2015_conference_program.htm