zhaw icp research report 2010

Research Report 2010

Zurich Universities ofApplied Sciences and Arts www.engineering.zhaw.ch Research & Development

r d

Das Titelbild zeigt die magnetischen Feldlinien einesAsynchronmotors. Die simulierten Feldlinien wurden mit deram ICP entwickelten Multiphysik Software SESES erstellt.

The title graphic shows magnetic streamlines of anasynchronous motor. The simulation was performed with themultiphysics software SESES developed at ICP.

Contents

Einleitung 5

Introduction 7

1 Sensors, Actuators, and More 91.1 Analyse von thermisch induzierten Fertigungstoleranzen bei der Herstellung von

Schwingquarzen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101.2 FE software for the industry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111.3 Numerical modeling of laminates anisotropic plastic deep-drawing . . . . . . . . . . 121.4 Populationsbilanzmodellierung von Mahlprodukten . . . . . . . . . . . . . . . . . . . 131.5 Thermal-flow analysis of pre-tempered chocolate in cooling tunnels . . . . . . . . . . 14

2 Electrochemical Cells and Energy Systems 152.1 A new model for detailed simulation of multiple transport and conversion processes

in SOFC stack repeat units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162.2 Time-dependent analysis of Hexis 5-cell stack (U,i)-data to predict degradation be-

havior on the stack level and for natural gas as fuel . . . . . . . . . . . . . . . . . . . 172.3 Thermal management of a micro-fuel cell system . . . . . . . . . . . . . . . . . . . . 182.4 Modeling the reactive bonding process to connect silicon substrates . . . . . . . . . 192.5 Development of a thermo-mechanical model to predict the buckling behavior of fuel

cell membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202.6 Development and application of an electrode micromodel to optimize performance

and to predict degradation phenomena in mixed conductors Ni-YSZ anodes . . . . . 212.7 Holzvergasermodelle für dezentrale Kraft-Wärmekoppelung . . . . . . . . . . . . . . 22

3 Organic Electronics and Photovoltaics 233.1 Efficient areal organic solar cells via printing . . . . . . . . . . . . . . . . . . . . . . . 243.2 Numerical simulation and design of extremely thin absorber solar cells . . . . . . . . 253.3 AEVIOM-Advanced Experimentally Validated Integrated OLED Model . . . . . . . . 263.4 Cost efficient thin film photovoltaics for future electricity generation . . . . . . . . . . 273.5 Modeling, simulation and loss analysis of dye-sensitized solar cells . . . . . . . . . . 283.6 In-line inspection of photovoltaic modules . . . . . . . . . . . . . . . . . . . . . . . . 303.7 Optoelectronic research laboratory . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

4 Student Projects 334.1 Optimisation of new generation silicon solar cells by means of 2D simulations . . . . 344.2 SETFOS sweep data visualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.3 Mass- and heat-transfer with two-phase flow, project "earth tube": cooling a school

building in Tamil Nadu, India . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364.4 Complete List of Student Projects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37

Appendix 39A.1 Scientific Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40A.2 News Articles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40A.3 Exhibitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41A.4 Conferences and Workshops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

3

ICP Institute of Computational Physics Research Report 2010

A.5 Prizes and Awards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44A.6 Teaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44A.7 Spin-off Companies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45A.8 Team . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46A.9 Location . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

ZHAW 4


Einleitung

Am Institute of Computational Physics blicken wir auf ein ereignisreiches Jahr 2010 zurück. Wirbearbeiteten verschiedenste Forschungsprojekte im Bereich der numerischen Modellierung vonMultiphysik-Systemen sowie Messtechnik und durften in unserer Forschungstätigkeit viele span-nende Herausforderungen zusammen mit unseren Industrie- und Hochschulpartnern meistern.Thematisch haben wir uns verstärkt mit Forschungsfragen im Zusammenhang mit Energie ausein-andergesetzt. So trägt unsere Forschung auf dem Gebiet der organischen Leuchtdioden als spar-same Beleuchtungstechnologie zur Energieeffizienz bei. Der Energieerzeugung dienen die orga-nischen Solarzellen als auch die Farbstoff-Solarzellen, welche am ICP mit numerischen Modellenerforscht werden. Unsere Forschung an Brennstoffzellen und an Holzvergasersystemen widmetsich der effizienten Energieumwandlung. Schliesslich vermögen wir auch die Ressourceneffizienzzu steigern, indem wir photothermische Messmethoden für die zerstörungsfreie Schichtprüfungentwickeln. Denn die zuverlässige Messung von Schichtdicken ermöglicht Materialeinsparungen.Es ist zum einen eine gesellschaftspolitische Pflicht, dass wir uns mit Energie-relevanten Fragenauseinandersetzen. Zum anderen eröffnen sich in diesen Gebieten spannende und herausfordern-de Forschungsthemen.

Im vergangenen Jahr mussten wir uns von unserem lieben Kollegen Hansueli Schwarzenbach ver-abschieden. Er verstarb im September nach schwerer Krankheit. Seine berufliche Laufbahn standim Zeichen der Forschung. Er war ein Pionier der Forschung an Schweizer Fachhochschulen undverstand es hervorragend, die Brücke von der Theorie in die industrielle Praxis zu schlagen. Ausder von ihm initiierten Arbeitsgruppe für numerische Simulation von Sensoren und Aktoren hat erin unermüdlicher mehrjähriger Arbeit das Institute of Computational Physics aufgebaut und bis imJahr 2007 geleitet. Seine Schaffenskraft führte auch zur ersten Spin-off Firma unserer Arbeits-gruppe, der Numerical Modelling GmbH. Er wird uns als Freund und grossartiger, lieber Menschin bester Erinnerung bleiben.

In personeller Sicht kamen im vergangenen Jahr einige neue Mitarbeiter ans ICP. Simon Züfleund Remo Ritzmann arbeiten als wissenschaftliche Assistenten im Bereich organische Solarzellenbzw. angewandte Optik und Messtechnik. Dr. Matthias Bonmarin widmet sich der infrarot-basiertenMesstechnik in unserem Optoelectronic Research Lab (O-LAB). Verschiedene ZHAW Bachelorund Master Studenten fanden den Weg ans ICP und arbeiteten bei unseren Forschungsprojek-ten mit. Insgesamt 3 IAESTE Praktikanten aus Hong Kong, Japan bzw. Finnland kamen für einmehrmonatiges Praktikum zu uns. Zum Jahresende zählt das ICP fast zwei Dutzend Mitarbeiterund ist zu einem wichtigen Leistungsträger an der School of Engineering herangewachsen. In derLehre sind wir inzwischen in nahezu allen Ingenieurstudiengängen tätig, nämlich in Mathematik,Physik aber auch im Unterricht zu Thermodynamik oder Signalverarbeitung. Ebenfalls bestreitenwir mehrere Unterrichtsmodule im nationalen Master of Science in Engineering.

Wir durften im vergangenen Jahr wieder verschiedenste Erfolge in der Forschung feiern. ZumBeispiel wurde die ICP Spin-off Firma Fluxim AG im Frühling 2010 in den KTI Start-up CoachingProzess aufgenommen. Das im Rahmen eines KTI Projekts entwickelte photothermische Schicht-prüfsystem wurde mit dem ITG Innovationspreis 2010 des Electrosuisse Verbands ausgezeichnet.Nils Reinke und Andor Bariska initiierten gegen Jahresende die Gründung der Firma WinterthurInstruments GmbH und werden die Kommerzialisierung des Schichtprüfsystems nun vorantreiben.Ein Technologietransfer am ICP wurde erneut auch mit den Firmen Numerical Modeling GmbH so-

ZHAW 5


wie Hexis AG gepflegt. Der vorliegende Bericht gibt einen Überblick über die im Jahr 2010 am ICPbearbeiteten Forschungsprojekte. Am Ende sind auch alle Studentenarbeiten sowie Publikationenund Konferenzbeiträge zusammengefasst.

Im Namen des ICP Teams möchte ich an dieser Stelle der School of Engineering der ZHAW so-wie dem Kanton Zürich für die Unterstützung und insbesondere unseren Förderagenturen für dieerfolgreiche Zusammenarbeit danken. Ich möchte mich bei jedem einzelnen ICP Mitglied für denpersönlichen Einsatz im vergangenen Jahr bedanken und wünsche meinem Nachfolger ThomasHocker viel Freude und Erfolg bei der neuen Aufgabe.

Nun wünsche ich Ihnen eine spannende Lektüre.

Beat RuhstallerInstitutsleiter

ZHAW 6


Introduction

An eventful year at the Institute of Computational Physics has come to an end. We carried outnumerous research projects in the field of numerical modeling of multiphysics systems and mea-surement techniques and enjoyed tackling many challenges in our R&D activities together withacademic and industrial partners. We have been focusing ever more on energy related researchtopics. For instance, our research on organic light-emitting devices (OLEDs) eventually emergesinto an energy-saving lighting technology. Organic solar cells and dye-sensitized solar cells, whichwe are investigating, are contributing to the generation of electrical power. Our research in fuelcells and wood gas processing systems is concerned with efficient conversion of energy. Finally,we are also addressing the efficient use of resources by developing novel photothermal measure-ment methods for coating analysis: an accurate determination of coating thickness can reducematerial use. On the one hand it is a socio-political responsibility that we are addressing energyrelated research topics. On the other hand these topics open up many exciting and challengingresearch questions.

Last year in September our faithful colleague Hansueli Schwarzenbach passed away after a se-vere and long illness. His professional career was dedicated to research and he was a pioneer inresearch at Swiss Universities of Applied Sciences. He managed to efficiently bridge the gap be-tween theory and industrial application. He had initiated a working team for numerical simulationof sensors and actuators and eagerly promoted it to the Institute of Computational Physics thathe headed until 2007. His dedication and hard work has led to the first spin-off of our institute,Numerical Modeling GmbH. We will keep him in our memory as a great and faithful friend.

In terms of staffing, several new team members joined the ICP last year. Simon Züfle and RemoRitzmann work as scientific assistants in the field of organic solar cells and applied optics and dataacquisition, respectively. Dr. Matthias Bonmarin focuses on infrared based measurement tech-niques in the optoelectronic research lab (O-LAB) of the ICP. Several ZHAW bachelor and mastersstudents have joined the ICP temporarily and contributed to research projects. A total of threeIAESTE internship students from Hong Kong, Japan and Finland carried out an internship at ourinstitute. Towards the end of the year, the ICP counted almost two dozen staff members and hasbecome a crucial institute within School of Engineering. Regarding teaching we give lectures inalmost all of the engineering curricula of our school, namely lectures in mathematics, physics butalso in thermodynamics or signal processing. Moreover, we provide lectures in several nationalteaching modules of the Master of Science in Engineering.

In the past year, we celebrated several successes in research. For instance, the ICP spin-offcompany Fluxim AG entered the CTI startup coaching process. The photothermal coating anal-ysis system developed within a CTI project has received the ITG innovation award 2010 fromElectrosuisse. Nils Reinke and Andor Bariska have launched a start-up company towards the endof the year for commercializing this innovative measurement system. A technology transfer wasagain fostered with the companies Numerical Modeling GmbH and Hexis AG. The ICP researchreport gives an overview on the research projects at the ICP in 2010. The appendix informs aboutstudent projects, publications and conference contributions.

ZHAW 7


On behalf of the ICP team I would like to express our gratitute to the School of Engineering andthe Canton of Zurich for supporting us and the research funding agencies for the successful col-laboration. I would like to thank each ICP team member for her/his commitment in 2010 and wishmy successor Thomas Hocker a lot of success with the new mandate.

Enjoy reading this report!

Beat RuhstallerHead of the institute

ZHAW 8


Chapter 1

Sensors, Actuators, and More

ZHAW 9


1.1 Analyse von thermisch induzierten Fertigungstoleranzenbei der Herstellung von Schwingquarzen

Contributors: T. Hocker

Partners: Patric Jacques, Micro Crystal AG, GrenchenFunding: DirektfinanzierungDuration: 2010–2011

Die Micro Crystal AG, Grenchen entwickelt undproduziert piezoelektrische Schwingquarze, dieals Taktgeber in Uhren und Mobiltelefonen ver-wendet werden, siehe figure 1. Als Teil derSwatch Group gehört sie mit 850 Mitarbeiternzu den in diesem Bereich weltweit führendenFirmen.

Fig. 1: Aufbau einer Quarzuhr.

Herzstück ist eine Stimmgabel aus Quarz, die inihrer Eigenfrequenz (typischerweise mit32’768 Hz) schwingt. Die Quarzstimmgabel wirdüber aufgedampfte Goldfolien, die als Elektro-den fungieren, mit einer Wechselspannung an-geregt. Für die Genauigkeit dieses elektrisch-mechanischen Resonators ist entscheidend,dass die Fertigungstoleranzen bei der Herstel-lung der Quarzgabel so klein wie möglich sind.Beim Aufdampfprozess der Goldfolien treten je-doch als Nebeneffekt unerwünschte Diffusions-prozesse auf, die zu ungleichmässigen Breitender Quarzgabeln führen können. Dies ist sche-matisch in figure 2 dargestellt.

Fig. 2: Modellierung von Korn- und Korngrenzen Cr-Transport.

Chrom diffundiert durch Goldkörner und ent-lang von Korngrenzen und bildet an der Ober-fläche der Goldschicht (bei hohen Tempera-turen und in oxidierender Atmosphäre) Chro-moxid. Zur Simulation des transienten Chrom-transports in den aufgedampften Goldschichten

und der Bildung von Chromoxid an der Gold-oberfläche wurde ein 3D FE-Modell in unse-rer in-house Multi-Physik Software Seses ent-wickelt. Die darin enthaltenen Diffusionskoeffizi-enten wurden an Messdaten angepasst, siehefigure 3. Figure 4 zeigt typische Konzentrations-profile für Chrom in Gold zu verschiedenen Zei-ten.

Fig. 3: Dargestellt ist die Schichtdicke an Chrom,dCr, die während eines Temperierprozesses in eineangrenzende Goldschicht diffundiert. Die Messdaten(Punkte) wurden im Modell durch Anpassen des Dif-fusionskoeffzienten gefittet.

Fig. 4: 3D Seses FE-Modell für die Simulation destransienten Korn- und Korngrenzen Cr-Transports.

ZHAW 10


1.2 FE software for the industry

Contributors: G. Sartoris, H. Schwarzenbach, M. Roos

Partners: Numerical Modelling GmbHFunding: Commission for Technology and Innovation (CTI)Duration: 2007–2009

Goal of this project is to provide the industrialpartner NM Numerical modeling with up-to-datesoftware for the modeling of the Maxwell’s equa-tion in the low frequency regime where displace-ment currents may be neglected, i.e. the eddycurrent model. Of particular interest is the mod-eling of magnetic proximity sensors with com-mon PC hardware and computational time lim-itations of a few hours. The new software hasbeen implemented in the multi-physics numeri-cal simulation tool SESES developed at ICP, itsmajor strength being the possibility to solve andcouple almost any governing equation of clas-sical physics. Although SESES could alreadysolve eddy current problems, their computationin 3D was far from being efficient and this projectshould provide up-to-date numerical algorithms.On the other side, since even optimized 3D eddycurrent computations can be still computation-ally expensive, from the industrial partner therewas a need to know if reduced and less de-manding models may be used to replace fully-fledged 3D computations.

From the software point of view, the first mile-stone was the introduction of edge finite ele-ments within SESES. These elements are thenatural framework when modeling eddy cur-rents. Since they do not enforce the normalcomponent of the vector field to be continu-ous and different from Lagrangian or nodal el-ements, they are well-suited to model sharpcorners and sharp discontinuous material laws.The second milestone was the optimized solu-tion of the system of linear equations. For 3Dapplications, the memory and computational re-quirements are too large for robust direct solversto be used and one has to resort to iterativesolvers. This is always the case when usingedge elements, since here the eddy problemcannot be gauged, the assembled linear systemis singular and direct solvers fail. However iter-ative solvers can solve singular systems, if theequations are assembled in a consistent man-ner. For iterative solvers, optimizing the solu-tion consists in preconditioning the linear sys-tem, i.e. by solving a similar system much closerto the identity matrix so as to increase the con-vergence rate. In order to be effective, the as-

sociated similar system determined by the pre-conditioning matrix or simply the preconditioner,must be cheaply computed and a major draw-back is that optimized and very effective pre-conditioners are problem dependent and needto be discovered for each application. Here thechallange is to find robust preconditioners withconvergence rates independent from the choiceof the material parameters and the shape of theelements. For both nodal and edge elementsof first and second order, we have implementedup-to-date preconditioners. For first order nodalelements, robust and fast is the ILU(0) precondi-tioner. The convergence rate is so good that theproblems to be solved are generally bounded bythe large memory requirements and not by thecomputational time. For second order nodal el-ements, the ILU(0) preconditioner can fail andthe alternative is to use an auxiliary precondi-tioner based on the robust solution of first ordernodal elements. Due to the singularity of thesystem matrix, the ILU(0) preconditioner failsalso for edge elements. Here, the same auxil-iary preconditioning technique can be used butthe shape of the elements has a strong impacton its performance and we have not yet foundan overall robust preconditioner.

Fig. 1: Magnetic field streamlines of a rectangularcoil.

ZHAW 11


1.3 Numerical modeling of laminates anisotropic plastic deep-drawing

Contributors: G. Sartoris

Partners: Alcan Technology and Management AGFunding: Commission for Technology and Innovation (CTI)Duration: 2008–2010

Modern laminates for the production of pharma-ceutical blisters have a complex inner structure.Several layers of different plastic and aluminasheets are packed, stacked and glued togetherwith the intent to provide the best environmentalprotection from external agents. When formingthe blisters by deep-drawing, the overall func-tional role of the layers in the original laminateshould be preserved and no holes, cracks orlayer delamination are allowed to evolve duringdeformation. This is one of the major concernof the industrial partner, which provides its cus-tomers with laminates for the production of phar-maceutical blisters and where the mechanicalproperties of the laminate must be guaranteed.

From the material point of view, the major con-cern is the correct description of the anisotropicelasto-plastic behavior at large strains, sinceactual models are not always able to correctlyreproduce experimental data. More advancedand precise tools are required which was thefirst major goal of the project. The task is com-plex and includes both the writing and fitting ofnew material laws. From the numerical point ofview, the anisotropic nature of the deformationprocess requires a full 3D nonlinear analysis atlarge strains. These computations are still quiteexpensive and the second goal of the projectwas to provide various optimizations at the soft-ware level to speed-up the numerical analysiswithin the multi-physics software tool Seses.

Solid-shell nonlinear elements have been im-plemented. These elements with a lot of in-ternal degrees-of-freedom are little more sen-sitive to non-linearities than first order solid el-ements, they can be freely mixed with otherelements and are for the global solution also as

fast as first order elements, however, they aremore expensive to assemble. A new interface torigid-body contacts has also been implementedwhich can be used both for the algebraic andpenalty method. The contact surface is interpo-lated either by first order or cubic polynomials.Frictionless contacts and the penalty methodhave been tested. The method is robust forall types of mechanical elements and can workwith just a few elements. As an alternative toill-conditioned penalty methods, there are alge-braic methods. A method working directly withthe globally assembled linear system and usingthe most recent residual forces has been devel-oped. Not yet fully solved is the sticky problemand the writing of robust lock-out algorithms.

At project’s end, we have accomplished all ma-jor goals and to our opinion a competitive andinnovative framework for modeling the deep-drawing process of pharmaceutical blisters hasbeen established. During the project develop-ment, it has been noticed that friction laws andcontact modeling also play an important role.This aspect should further be considered to-gether with other minor numerical optimizationprocedures.

Fig. 1: Deep-drawing of a laminate to form a blister.

ZHAW 12


1.4 Populationsbilanzmodellierung von Mahlprodukten

Contributors: R. Axthelm

Partners: Novartis Pharma AGFunding: internDuration: 2010

Mahl- oder Zerteilprozesse zur Herstellung fei-ner Feststoffpartikel spielen in vielen industri-ellen Anwendungen eine grosse Rolle. Dabeiist häufig das Ziel, den Prozess so zu steuern,dass die Partikel am Ende des Mahlprozesseseine bestimmte Grösse bzw. Grössenverteilungaufweisen. Die Frage, die sich zunächst stelltist: Wie kann die Entwicklung der Grösenvertei-lungen analytisch beschrieben werden und wiehängt die gewonnene Darstellung von Einfluss-grössen des Prozesses ab? Zu den Einfluss-grössen gehören etwa Drücke und Geschwin-digkeiten, die auf die Partikel wirken oder Mate-rialeigenschaften wie Brüchigkeit der Partikel.Der Vorgang des Partikelzerkleinerns wirddurch ein Modell beschrieben, das zwei grund-sätzliche Fragen beinhaltet, nämlich: Bricht einPartikel? Und wenn ja, wie bricht es? DiesesModell führt auf ein System von linearen, inho-mogenen Differentialgleichungen, welche sichanalytisch berechnen lassen.Die Lösungen dienen als Ansatzfunktionen fürRegressionen von Messdaten. Diese wiederumliefern eine Parametrisierung nach verschiede-nen Einflussfaktoren. Figure 1 zeigt eine Über-blicksdarstellung dieser Vorgehensweise.

pw~

pw

. ..

.

..

.. . .

pw~

pw

Einflussfaktoren p

Modell

tw = A wp

Parametrisierung der Einfluss−faktoren p über Versuchsreihen

Messdaten

analytischeLösung

Regression der Datenmit den Ansatzfunktionen

Fig. 1: Die Grössenverteilung w der Partikel hängtvon den gegebenen Einflussfaktoren p ab.

Die Arbeitsschritte gliedern sich wie folgt: DerAnteil an Partikeln einer bestimmten Grössen-klasse i wird in wi zusammengefasst. BeimBruchprozess bricht ein Anteil si (selectivity)von wi in eine oder mehrere kleinere Klassenj. Die Verteilung auf kleinere Klassen wird in bij(breakage function) beschrieben. Für die Ände-rungsrate jeder Klasse i ergibt sich daraus

wi,t = �siwi +n∑

j=i+1bijsjwj :

wi,t beschreibt die Zeitableitung von wi. Wederdie breakage function noch die selectivity sinda-priori bekannt. Zunächst werden diese beidenAusdrücke zusammengefasst in eine Matrix Amit Aii = �si, Aij = bijsj , (i > j) und Aij = 0,(i < j). Damit stellt sich das Modell gleicher-massen aber in neuer Form dar als das Systemgewöhnlicher Differentialgleichungen

wt = A w ;

mit w = (w1; : : : ; wn). Die Lösungen diesesSystems ergeben sich durch die Rekursion:a11 = 0. Für k = n; : : : ; 1 berechne:

wk(t) =n∑l=k

�kl eall t

�kl =

∑nj=k+1

akj αjl

all−akk; l > k

wk,0 �∑ni=k+1 �

ki ; l = k > 1

100 ; l = k = 10 l < k

Die Unbekannten akj werden regressiv an dieDaten ermittelt (siehe figure 2). Anschliessendkönnen aus den akj die selectivity si und diebreakage function bij berechnet werden.

Fig. 2: Zeitliche Entwicklung der Partikelanteile jenach Grössenklasse. Dargestellt sind Messdaten (�)und Lösungen (�) der Differentialgleichungen.

Mit dieser Methode erhält man qualitativ guteErgebnisse. Es lassen sich, wie in figure 2 zusehen ist, ebenfalls die Ergebnisse des Mahl-prozesses extrapolieren.

ZHAW 13


1.5 Thermal-�ow analysis of pre-tempered chocolate in coo-ling tunnels

Contributors: T. Hocker, R. Axthelm

Partners: Manfred Suter, Max Felchlin AG, SchwyzFunding: Swiss Food Research NetworkDuration: 2010–2011

Max Felchlin AG, Schwyz, is well-known forthe production of high-quality couverture, pa-stries and semi-finished confectionery products.Felchlin offers its customers comprehensivesupport and know-how for the manufacture andmarketing of their end products. This support isbased on both experience and understanding.

For an optimal product quality, the achievedstate of crystallinity of the cocoa butter - a ma-jor ingredient of chocolate - plays a crucial ro-le. This crystallinity state is manipulated duringthe so-called tempering process, where (sta-ble) crystal seeds are generated. However, italso depends on the details of the subsequentcooling process, which so far has rather beenneglected. Therefore, this project aims at de-veloping methods to optimize the cooling pro-cess of pre-tempered chocolate in cooling tun-nels. Figure 1 gives an overview about the plan-ned combined experimental and modeling ap-proach.

To describe the cooling process on differentlength scales, three different types of models

need to be developed:

• a thermal-flow CFD model of the completecooling channel

• a thermo-mechanical FE model of a singlechocolate mould to simulate the crystalli-zation of chocolate

• a thermal crystallization model to predictthe heat release and consumption as-sociated with different crystallization andmelting processes

Based on the gained know-how, the followinggoals shall be reached:

• improving Felchlin’s internal cooling pro-cess, i. e. to tailor it to specific products

• supporting industry customers in the fur-ther processing of chocolate products

• offer complete solutions to confectionarycustomers e. g. about the mold removingcharacteristics

Fig. 1: Overview about the planned combined experimental and modeling approach.

ZHAW 14


Chapter 2

Electrochemical Cells and EnergySystems

ZHAW 15


2.1 A new model for detailed simulation of multiple transportand conversion processes in SOFC stack repeat units

Contributors: Y. Safa, T. Hocker

Partners: Hexis AG, HTceramix S.A., ETH-NIM, EPFL-LENI, EMPAFunding: SwissElectric Research, Swiss Federal Office of EnergyDuration: 2008–2010

In 1994 Achenbach was among the first topublish a comprehesive 3D model for a pla-nar repeat unit of a SOFC stack. A vast num-ber of publications on SOFC models have ap-peared since then. Almost all publications dis-cuss physical-chemical descriptions, model hi-erarchies (to capture phenomena on differentlength scales) and the resulting set of governingequations, but few deal with numerical issues.However, the robustness and speed of a SOFCmodel is critical for its usefulness as a designtool.

Progress in the study of local reactive transportphenomena in SOFC stacks has been achievedbased on both advanced physical and numer-ical approaches. Specifically, the numericallyunfavorable high aspect ratio of about 1’ 000 be-tween a typical stack diameter and a typical cellthickness has been successfully treated usingthe ADI (Alternating Direction Implicit) numeri-cal scheme. Unlike conventional methods, ADIallows one to predict local gradients of chemi-cal species and electrical charges with low com-puting costs and unconditional numerical stabil-ity. This is especially important in the vicinityof a current collector rib and under extreme op-eration conditions, e. g. when the fuel gets de-pleted. Furthermore, the convection-dominanttransport within the gas distribution channelshas been accurately calculated without using a(unphysical) transverse numerical diffusion.

Concerning the gas flow in the channels alonga porous electrode, another important featureof our model is the non-zero slip velocity at theelectrode surface. In conventional approaches,this slip velocity is often estimated empirically.Alternatively, the hydrodynamic flow field is cal-culated from the Navier-Stokes equations whichare simultaneously solved in the open flowchannel and the porous electrode. Within the

electrode, the velocity field is then penalizedartificially by adding the Karman-Kozeny term.However, such a penalization suffers from inac-curate velocities in the vicinity of the electrodesurface. In our approach, the compressible flowin both regions is treated in a unified mannerin which the slip velocity is not known a pri-ori, but follows from requiring continuous shearstresses at the electrode surface.

Fig. 1: Modeled repeat unit geometry including theanode, the open flow channel and the current collec-tion ribs.

As an example, simulation results for the influ-ence of the current-collection pattern (displayedin figure 1) on the hydrogen and ionic poten-tial distributions within an SOFC repeat unit areshown in figure 2.

Fig. 2: Simulation results in a vertical cross-sectionof a SOFC repeat unit. Left: hydrogen distribution inchannel and electrode, right: Ionic potential in elec-trode.

ZHAW 16


2.2 Time-dependent analysis of Hexis 5-cell stack (U,i)-data topredict degradation behavior on the stack level and fornatural gas as fuel

Contributors: M. Linder, T. Hocker

Partners: Roland Denzler, Hexis AG, HTceramix S.A., ETH-NIM, EPFL-LENI, EMPAFunding: AccelenT, Swiss Federal Office of EnergyDuration: 2009–2011

A simplified, Mathematica-based model toanalyze the current-voltage characteristics ofSOFC-stacks has been developed and is in reg-ular use at Hexis. The model predicts the start-ing performance for a large number of current-voltage stack data without the need for any fit-ting parameters. It has been used to study thesensitivity of various input parameters such aschanges in mass-flows, fuel composition, tem-perature, and humidity. The model is currentlyextended to the Hexis Galileo system to per-form model-based analysis of field tests. Here,the goal is to explain unusual system behaviorthat, for example, could result from compositionchanges in the natural gas supply, and to assesslong-term degradation from the field test results.

Fig. 1: Three-step procedure to adjust the model toUj-data obtained from SOFC stacks measurements.

The model extracts the internal stack repeatunit resistance from an experimentally availablecurrent-voltage (Uj)-curve in a three-step proce-dure. This is shown in figure 1. As step one,an ”ideal“ Uj-curve is calculated under the as-sumption that the fuel reforming process andthe subsequent electrochemical fuel conversionproceed under thermodynamic equilibrium con-ditions. Hence, for a given fuel composition atthe reformer inlet and a given reformer temper-ature, the outlet composition is calculated us-ing a thermodynamics software-package suchas Cantera. To match the experimental open cir-cuit voltage by the model, in step two, the corre-

sponding leakage flux of hydrogen is removedand water is added to the anode gas mixture.Finally, in step three, we analyze the remainingdifferences between the experimental Uj-dataand the simulated, dotted curve. These differ-ences are caused by internal ohmic and polar-ization losses within the stack repeat unit.

Fig. 2: Total energy balance for a Hexis SOFC stackat 920 ◦C that runs on CPO-reformed natural gas.

Besides analyzing Uj-data, the model also per-forms a full energy balance. Figure 2 showshow the various energy fluxes change with cur-rent density for a typical Hexis SOFC-stack. Themaximum electrical efficiency (DC) at320 mA cm−2 is around 40 % - relative to theheating value of the fuel input. This correspondswith a power-density of roughly 200 mW cm−2.At this operation point, about 9 % is lost by thereforming of the natural gas, and 7 % by fuelleakages. An additional 7 % of the availablefuel is burned in the post combustion unit (pcu)and another 7 % is lost by Joule’s heat relatedto internal repeat unit resistances. The largestloss comes from unavoidable reversible thermo-dynamic losses and is roughly 30 %. It also fol-lows from the results that the maximum achiev-able electrical efficiency of a Hexis stack withelectrolyte supported cell technology running onCPO-reformed natural gas is around 45 %.

ZHAW 17


2.3 Thermal management of a micro-fuel cell system

Contributors: C. Meier, T. Hocker

Partners: ETH-NIM, ETH-LTNT, EPFL-SAMLAB, CSEM, NTBFunding: Swiss National Science Foundation SinergiaDuration: 2010–2012

A simple but yet accurate model based on globalmass and energy balances has been devel-oped to predict the thermal characteristics of theONEBAT micro-SOFC and similar systems thataim at powering portable devices. Our motiva-tion was to assess the basic requirements onthe thermal design of such a system that lead toa thermally self-sustaining operation. Figure 1and table 1 give an overview of the main systemcomponents and specifications.

dition, a simplified model to characterize the ther-mal insulation is presented that links the energybalance to geometrical aspects of the system (e. gthe stack size to the heat loss and to the insulationvolume).

2. Description of the system

The components and specifications relevant tothis work are summarized in Figure 1 and Table 1.A catalytic partial oxidation fuel reforming unit(CPO), the SOFC stack and a post combustionunit (PCU) are integrated in a micro-fabricated,isothermally operated module named hot module(HM). The cell membranes are mounted as cell ar-ray on a Silicon or Foturan substrate with 0.3 to 0.5mm thickness. A number of stacked repetitive units(RUs) with about 2.5Wel each make up the SOFCstack, dimensioned to meet the targeted power out-put.

The assumed isothermal conditions are achiev-able by a compact design with high thermal con-ductance between the components. Sufficient heattransfer between hot module and the incomingair must be ensured to reduce thermally inducedstresses on the substrate and the membranes.

Two counteractive mechanisms are selected tocontrol the hot module temperature. A cooling in-terface disposes excess heat from the hot moduleand lowers the operating temperature, while a re-cuperator preheats the incoming air flow and risesthe operating temperature.

Additional system components not further con-sidered here are start-up heater, system control unitand flow management devices for both fuel and airsupply (i.e. micro-valves, pumps, fans and a fuelcartridge). The auxiliary power consumption forthese components is thus not accounted for in thiswork. Cooling of the exhaust gas can be accom-plished by either dilution with ambient air or by anadditional heat sink.

Heat losses to the surroundings (caused by non-adiabatic thermal insulations) become a key issuefor portable micro-SOFC systems. First, the insu-lation needs to ensure a low surface temperature fora save handling of the device. Second, the operat-ing temperature is strongly influenced as heat lossesare in the same order of magnitude as the electricalpower output. Third, the insulation is expected toincrease the system volume significantly and thusbecomes subject to an optimization regarding thepower density of the system.

Fuel Reforming

Stack

Post Combustion

Hot

Mod

ule

Coo

ling

Inte

rfac

e

Exhaust Gas

Cathode Air

Butane / Air

Recuperator

Hot Module

Thermal Insulation Isothermal Connector

TSO

FC

=55

0◦C

Figure 1: Proposed scheme for a micro-SOFC system with

isothermally operated hot module, cooling interface, recu-

perator and thermal insulation. Not shown are devices for

the flow management, a fuel cartridge and additional heat

sinks for the exhaust gas.

Proposed micro-SOFC system specificationssystem electrical power, Pel 2.5 ... 20Wel

repetitive unit power, Pel,RU 2.5Wel

fuel n-butane C4H10

air as oxidant 21% O2 / 79% N2

operating temperature, TSOFC 550◦Creformer catalyst [15] Rh/Ce0.5Zr0.5O2

post combustion catalyst [2] Pt/Ce0.5Zr0.5O2

hot module substrate Silicon or Foturan

Table 1: Proposed specifications for a MEMS based micro-

SOFC system.

3. Energy and mass balance

3.1. Overview

If we assume that all the unreacted combustiblesundergo complete oxidation in the PCU, the en-ergy conservation for the micro-SOFC system thenstates that

Qth + Pel = Pchem (1)

where Qth is the heat released in CPO, stack andPCU. Pel is the electrical power withdrawn fromthe stack and Pchem the chemical power deliveredby the fuel,

Pchem = nC4H10LHVC4H10

(2)

with n as the molar flow rate and LHV as lowerheating value of the fuel. If we further define theelectrical efficiency ηel as ratio between electricaland chemical energy,

ηel =Pel

Pchem(3)

2

Fig. 1: Schematic overview of the MEMS basedONEBAT micro-SOFC with an isothermally (self-sustaining) operated hot module, a cooling interface,a recuperator and a thermal insulation.

The thermal characteristics of systems with anelectrical power output between 2.5 and 20 Welwere obtained by dimensioning the hot modulesand the surrounding insulations, calculate heatlosses to the environment and solve the energyand the mass balance. The model thus links theenergy balance to geometrical aspects of thesystem like the insulation volume.

dition, a simplified model to characterize the ther-mal insulation is presented that links the energybalance to geometrical aspects of the system (e. gthe stack size to the heat loss and to the insulationvolume).

2. Description of the system

The components and specifications relevant tothis work are summarized in Figure 1 and Table 1.A catalytic partial oxidation fuel reforming unit(CPO), the SOFC stack and a post combustionunit (PCU) are integrated in a micro-fabricated,isothermally operated module named hot module(HM). The cell membranes are mounted as cell ar-ray on a Silicon or Foturan substrate with 0.3 to 0.5mm thickness. A number of stacked repetitive units(RUs) with about 2.5Wel each make up the SOFCstack, dimensioned to meet the targeted power out-put.

The assumed isothermal conditions are achiev-able by a compact design with high thermal con-ductance between the components. Sufficient heattransfer between hot module and the incomingair must be ensured to reduce thermally inducedstresses on the substrate and the membranes.

Two counteractive mechanisms are selected tocontrol the hot module temperature. A cooling in-terface disposes excess heat from the hot moduleand lowers the operating temperature, while a re-cuperator preheats the incoming air flow and risesthe operating temperature.

Additional system components not further con-sidered here are start-up heater, system control unitand flow management devices for both fuel and airsupply (i.e. micro-valves, pumps, fans and a fuelcartridge). The auxiliary power consumption forthese components is thus not accounted for in thiswork. Cooling of the exhaust gas can be accom-plished by either dilution with ambient air or by anadditional heat sink.

Heat losses to the surroundings (caused by non-adiabatic thermal insulations) become a key issuefor portable micro-SOFC systems. First, the insu-lation needs to ensure a low surface temperature fora save handling of the device. Second, the operat-ing temperature is strongly influenced as heat lossesare in the same order of magnitude as the electricalpower output. Third, the insulation is expected toincrease the system volume significantly and thusbecomes subject to an optimization regarding thepower density of the system.

FuelReforming

Stack

PostCombustion

Hot

Mod

ule

Coo

ling

Inte

rfac

e

Exhaust Gas

Cathode Air

Butane / Air

Recuperator

Hot Module

Thermal Insulation IsothermalConnector

TSO

FC

=55

0◦C

Figure 1: Proposed scheme for a micro-SOFC system with

isothermally operated hot module, cooling interface, recu-

perator and thermal insulation. Not shown are devices for

the flow management, a fuel cartridge and additional heat

sinks for the exhaust gas.

Proposed micro-SOFC system specificationssystem electrical power, Pel 2.5 ... 20Wel

repetitive unit power, Pel,RU 2.5Wel

fuel n-butane C4H10

air as oxidant 21% O2 / 79% N2

operating temperature, TSOFC 550◦Creformer catalyst [15] Rh/Ce0.5Zr0.5O2

post combustion catalyst [2] Pt/Ce0.5Zr0.5O2

hot module substrate Silicon or Foturan

Table 1: Proposed specifications for a MEMS based micro-

SOFC system.

3. Energy and mass balance

3.1. Overview

If we assume that all the unreacted combustiblesundergo complete oxidation in the PCU, the en-ergy conservation for the micro-SOFC system thenstates that

Qth + Pel = Pchem (1)

where Qth is the heat released in CPO, stack andPCU. Pel is the electrical power withdrawn fromthe stack and Pchem the chemical power deliveredby the fuel,

Pchem = nC4H10LHVC4H10

(2)

with n as the molar flow rate and LHV as lowerheating value of the fuel. If we further define theelectrical efficiency ηel as ratio between electricaland chemical energy,

ηel =Pel

Pchem(3)

2

Tab. 1: Main system specifications.

The air-to-fuel ratio λ is an appropriate param-eter to control the operating temperature. How-ever, a minimal λ of ≈ 2 is required by the elec-trochemical and post-combustion process (stoi-chiometric oxidation: λ = 1). At high electricalefficiencies and hence low heat release rates orin case of high insulation heat losses, a recuper-ator is required to achieve the desired operat-ing temperature (→recuperator mode). In con-trast, at lower electrical efficiencies, disposingthe released heat becomes an issue (→coolingmode).The basic thermal requirements for non-adiabatic systems are visualized in figure 2. Incooling mode (a), the required air-to-fuel ratio λto not exceed an operating temperature of550 ◦C is shown. In recuperator mode (b), theheat to be recuperated Qrecu and the dimen-sionless airflow preheating Θrecu are shown thatmaintain the operating temperature.

2.00

2.25

2.50

2.75

3.00

3.25

3.50

λ(-

)

cool

ing

mode

Qre

cu

=0

(a)

INSUL A(fumed silica)

INSUL B(evacuated)

constant temperature operating modes at 550◦Cfor 2.5Wel systems with varying pel,net

10% 20% 30% 40% 50%

electrical efficiency ηel

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Qre

cu

/P

el(W

th/W

el)

recu

per

ator

mode

λ=

2

(b)

0.2

Θrecu =0.4 0.6 0.8 0.99

HM I(37.5mW)

HM II(75mW) HM III

(150mW)

9.1

2.00

2.25

2.50

2.75

3.00

3.25

3.50

λ(-

)

cool

ing

mode

Qre

cu

=0

(a)

HMVI(20Wel)

HMV(10Wel)

HM IV(5Wel)

HM II(2.5Wel)

INSUL A (fumed silica)

constant temperature operating modes at 550◦Cfor systems from 2.5 to 20Wel

10% 20% 30% 40% 50%

electrical efficiency ηel

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Qre

cu

/P

el(W

th/W

el)

recu

per

ator

mode

λ=

2

(b)

0.2

Θrecu =0.4 0.6 0.8 0.99

9.2

Figure 9: Resulting thermal design requirements for a selection of the case study scenarios. Cooling mode (a) with Qcool perWel to be disposed by a heat sink, λ-mode (b) with pure convective heat disposal (except for insulation heat losses), andrecuperator mode (c) with Qrecu per Wel to be transferred from the exhaust gas to the inflow. 9.1 shows the right-shift ofthe characteristics obtained by compactly built hot modules (i. e. with a high cell power density) and by an improved thermalinsulation. 9.2 shows the right-shift of the characteristics due to lower surface-to-power ratios of stacked systems.

10

Fig. 2: Constant temperature operating plot for sys-tems with 2.5 to 20 Wel. The increasing heat-loss-to-power ratio for systems with small electrical outputsshifts the characteristics to the left dashed line in (a):adiabatic operated system.

ZHAW 18


2.4 Modeling the reactive bonding process to connect siliconsubstrates

Contributors: C. Meier

Partners: ETH-NIM, ETH-LTNT, EPFL-SAMLAB, CSEM, NTBFunding: Swiss National Science FoundationDuration: 2010–2012

Reactive multilayer foils are used as local heatsource in MEMS fabrication to melt a solder. Af-ter ignition on one end, an exothermic reactionfront (Al+Ni→ AlNi) propagates through the foil,see figure 1. The bilayer thickness influencesthe propagation velocity, the reaction durationand the specific heat of reaction. The thicknessof the foil determines the total amount of heat re-leased per bonding area. The whole solder layerabove and below the multilayer foil has to melt toachieve mechanical strength and gas tightness.

Fig. 1: Self-propagating reaction of a reactive multi-layer foil (from Wang et al., 2004).

In an ideal (adiabatic) system, the required ther-mal energy is determined by the mass andthe heat capacity of the multilayer foil andthe solder, respectively, as well as the melt-ing temperature and the heat of fusion of thesolder, see e. g. Wang et al., "Joining ofstainless-steel specimens with nano-structuredAl/Ni foils", Journal of Appl. Physics (2004). Ina real system, the heat diffusion away from thesolder requires more heat to completely melt thesolder.To compute the temperature distribution withinthe bond and in its surrounds, a 1-dimensional

model of the bonding system has been devel-oped. The transient heat conduction equationis solved in our in-house FE-tool Seses. Onlyone half of the bonding is modeled, the adiabaticboundary at the reactive foil reflects symmetry,see figure 2. The heat is evenly released intothe AlNi layer.Figure 3 shows calculated maximum tempera-tures at different locations. A 2 µm thick sili-con oxide layer on top of the substrate has beentaken into account, which causes a steepT-gradient between the solder and the wafer dueto the low thermal conductivity of SiO2. Thanksto this thermal barrier, the temperature in the Si-wafer does not exceed the solder’s melting tem-perature of 280 ◦C. As the assumed reaction du-ration goes toward small numbers, its influenceon the temperature field around the reactive foilbecomes small.

− 25 0 25 50 75 100

distance z from interface solder/reactive foil ( µ m)

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

Tm

ax

(◦C

)

AlNi AuSn 2µ m SiO2 + 498 µ m Si

Tm (280 ◦ C)

1000.0 µ s

500.0 µ s

250.0 µ s

125.0 µ s

62.5 µ s

31.25 µ sduration of heat releasebetween 31.25 and 1000.0 µ s

symmetry at z = -25 µ m

solder phase change: noSiO2 layer: yes

profi le of local temperature maxima during bonding process

Fig. 3: Finite-element simulation results of maximumtemperatures at different locations in the vicinity ofthe reactive foil, based on the transient heat conduc-tion equation.

Boundary condition: adiabatic (symmetry)

25um AlNi

25um AuSn 498um Si 10mm Stainless Steel

(working table)

2um SiO2

Boundary condition: Isothermal 20°C

heat source q‘

Fig. 2: Materials, layer thicknesses and boundary conditions of the 1D model.

ZHAW 19


2.5 Development of a thermo-mechanical model to predict thebuckling behavior of fuel cell membranes

Contributors: Y. Safa, T. Hocker

Partners: ETH-NIM, ETH-LTNT, EPFL-SAMLAB, CSEM, NTBFunding: Swiss National Science Foundation SinergiaDuration: 2010–2012

The presented model development is con-cerned with mechanical integrity issues. Henceit involves the analysis of different thermo-mechanical failure modes and the developmentof strategies to avoid them. At different stagesduring manufacturing and operation, both com-pressive and tensile stresses are exerted ontothe fuel cell membranes. The membranes arethin ceramic layers with thicknesses of around1 µm. Thus they represent the mechanicallyweakest component within the system. Com-pressive stresses can cause buckling and sub-sequent fracture. Tensile stresses can initi-ate cracks, which might as well cause fracture.Hence both types of stresses can severely dam-age or completely break the membranes. Itis therefore mandatory to identify save ”designspaces“ as guidelines for the manufacturing andoperation of the fuel cell membranes.As main task, a complete buckling modelhas been implemented into the Mathematicasoftware package. This model is based onthe Railey-Ritz method as described by Ya-mamoto et al., ”Nonlinear thermomechanicaldesign of microfabricated thin plate devices inthe post-buckling regime“, J. Micromech Micro-eng, vol. 20, pp. 035027 (2010). This approachuses trial functions for the buckling shapes thatsatisfy certain symmetry conditions. The mem-brane dimensions, its stress state caused bythe manufacturing, its material properties andan applied thermal load are used to calculatethe stress and displacement fields under com-pression. Note that the behavior of the mem-brane under compression is highly nonlinearsince buckling is the result of a instability transi-tion that is accompanied by a sudden energy re-lease. Therefore standard mechanical theoriessuch as Hooke’s theory for linear-elastic mate-

rials are not applicable. In practice, several dif-ferent buckling modes with different shapes canbe observed.Figure 1 (top left) shows schematically the al-lowed design space for a fuel cell membranein dependence of the applied thermal load andthe residual stresses imposed by the manufac-turing.

Fig. 1: top left: Schematic representation of the de-sign space to ensure mechanical integrity, top right:simulation results of a buckled membrane under com-pression, bottom: experimental results for buckledmembranes with similar shapes.

The design space is bound by too much com-pression (which causes buckling) and too muchtension (which causes fracture). The diagramon the top right shows simulation results ob-tained by Y. Safa of a buckled membrane un-der compression. Note that the membrane isclamped at its boundaries. Note also that thewarping exhibits four characteristic ridges. Thisis similar to the experimentally obtained bucklingshapes shown on the bottom.

ZHAW 20


2.6 Development and application of an electrode micromodelto optimize performance and to predict degradation phe-nomena in mixed conductors Ni-YSZ anodes

Contributors: T. Hocker

Partners: Hexis AG, HTceramix S.A., ETH-NIM, EPFL-LENI, EMPAFunding: SwissElectric Research, Swiss Federal Office of EnergyDuration: 2008–2010

The developed electrode model mimics thereal 3D microstructure of mixed ion-/electron-conducting anodes and cathodes. It is espe-cially suited for investigating the impact of mi-crostructural details on electrode performanceand aging. For example, it allows one to as-sess the effect of microstruture changes suchNi coarsening and electrode poisoning on theohmic and polarization resistances of the elec-trode. The electrode model is part of a larger3D local repeat unit model, which representsthe full cell, contact layers as well as non-homogeneous current collectors (i. e. intercon-nects) on both sides. Hence the impact of elec-trode degradation on the local repeat unit per-formance can be predicted as well. To real-istically mimic the electrodes solid-phase net-works, up to eight different particle sizes canbe specified for each phase. This allows one toadjust synthetic microstructures to real ones bymatching their experimentally-obtained continu-ous particle size distributions (PSDs). Further-more, the limiting effect of sinter necks on thesolid-phase ion and electron transport is takeninto account via additional contact resistancesbetween neighboring particles. When com-pared with literature data, the electrode modelpossesses all main characteristics reported forreal electrodes, see S. Sunde, ”Simulations ofComposite Electrodes in Fuel Cells”, Journal ofElectroceramics, 5, pp. 153-182, 2000.Figure 1 shows typical SEM pictures - on the leftfor a new Ni-8YSZ anode and on the right forthe same anode after eight redox cycles. Theelectrode samples were provided by Boris Iwan-schitz from Hexis, the image analysis was per-formed by Lorenz Holzer from EMPA. One cansee that the redox cycles significantly increasedthe size of the Ni-particles, but left the 8YSZ-particles unchanged. Note also that the redoxcycles caused a significant increase of pores.To adjust the synthetic microstructures used inthe model to the ones shown in figure 1, con-tinuous particle size distributions (PSDs) havebeen extracted from the SEM image data.

Fig. 1: SEM pictures for a new Ni-8YSZ anode (left)and the same anode after eight redox cycles (right).

This is shown in figure 2, where three an-odes with initially fine, medium-fine, and coarsemicrostructures are compared to each other.The red bars represent experimental resultsfor the new anodes as obtained from EIS-measurements. The corresponding simulationresults based on artificial microstructures areshown as gray bars. They were obtained for20 µm thick anodes with synthetic microstruc-tures exhibiting PSDs similar to the experimen-tal ones. It was observed that the model welldescribes the change of the polarization resis-tance with mircrostructure. The deviations fromthe experimental data lie within 20 %. The ex-perimental and simulation results for the agedanodes after eight redox cycles are shown aswell.

Fig. 2: Polarization resistances for new and aged Ni-8YSZ anodes (after eight redox cycles).

ZHAW 21


2.7 Holzvergasermodelle für dezentrale Kraft-Wärmekoppelung

Contributors: C. Meier, T. Hocker, R. Axthelm

Partners: Woodpower AG, Wila.Funding: Gebert Rüf StiftungDuration: 2010–2012

Ziel des Projekts ist die Entwicklung von pra-xistauglichen Computermodellen, die die Um-wandlung von Holz in Holzgas in Gleichstrom-holzvergasern abbilden. Die Modelle sollen diewichtigsten Prozessgrössen zueinander in Be-ziehung setzen und so helfen, Fragen zum (i)Reaktordesign, zum (ii) anfahren und (iii) konti-nuierlichen Betrieb des Holzvergasers sowie zur(iv) Auswirkung von unvermeidlichen Schwan-kungen in den Brennstoffeigenschaften im Vor-aus zu klären.

Aufgrund der grossen Anzahl an Einflussgrös-sen und der Komplexität der im Vergaser ablau-fenden physikalisch-chemischen Prozesse er-folgt die Validierung der Modelle über Mess-daten, die hauptsächlich mithilfe eines eigensentwickelten Experimentalvergasers gewonnenwerden.

Der für die Modellvalidierung benötigte Experi-mentalvergaser mit einer thermischen Leistungvon rund 10 kW konnte 2009 gefertigt und 2010in Betrieb genommen werden, siehe figure 1.

Fig. 1: ZHAW Experimentalvergaser im Aufbau undim Betrieb auf dem Dach des Maschinenlaboratori-ums.

Figure 2 gibt einen Überblick der im Projektverfolgten Modellansätze zur Beschreibung derUmsetzung von Holz zu Holzgas in Festbettre-aktoren.

Fig. 2: Überblick Modellansätze zur Beschreibung derHolzvergasung.

ZHAW 22


Chapter 3

Organic Electronics andPhotovoltaics

ZHAW 23


3.1 Efficient areal organic solar cells via printing

Contributors: M. T. Neukom, R. Ritzmann, T. Lanz, N. A. Reinke, B. Ruhstaller

Partners: TU/e Eindhoven University of Technology (NL), University Jaume I (E), CSEM, BASFFunding: Swiss Federal Office of EnergyDuration: 2008–2011

This international research project with aca-demic and industrial partners aims at creatingmore efficient and stable organic solar cells.The ICP is simulation partner and develops acomprehensive electro-optical model to numeri-cally characterize organic solar cells.An accurate determination of the charge carriermobilities is essential to model and improve or-ganic solar cells. A frequently used method todetermine charge carrier mobilities is the CELIVtechnique (charge extraction by linearly increas-ing voltage). In this technique a voltage rampis applied to the device in order to extract freecharge carriers inside the bulk.We simulate the photo-CELIV experiment with afully-coupled electro-optical model and measureCELIV currents for several voltage slopes a andtransient photo-currents of an organic bulk het-erojunction solar cell. By fitting our numericalmodel to multiple curves we show that importantmaterial parameters like the charge mobilities,charge generation and recombination efficiencycan be determined using numerical parameterextraction.Figure 1 shows measured CELIV photocurrentswith different voltage slopes and the corre-sponding simulations. For all voltage slopes thesame set of simulation parameters is used. Thetransient currents can in general be reproducedwith a numerical drift-diffusion model that simplyassumes constant charge mobilities.

Fig. 1: Measurement (black) and simulation (red) oftransient CELIV currents for different voltage slopes.

Measured and simulated IV-curves of a spin-coated cell are plotted in figure 2 as a function ofillumination intensities. The measured current-voltage curves can be reproduced successfullyby the simulations With the introduction of a lightsource with calibrated illumination intensity weobtain an additional scaling variable to investi-gate.

Fig. 2: Measured (blue) and simulated (red) current-voltage curves for spin-coated cells measured at dif-ferent illumination intensities.

In the coming (last) project year, a series ofbulk heterojunction blends dissolved in differ-ent solvents will be used for solar cell fabri-cation. The cell performance will be evalu-ated by model-based data analysis. The cor-relation of the cell performance with the blendmorphology will be studied. Moreover, furtherexperiments with time-dependent characterisa-tion techniques (e.g. CELIV, impedance spec-troscopy) will be carried out.

ZHAW 24


3.2 Numerical simulation and design of extremely thin absorbersolar cells

Contributors: M. Loeser, B. Ruhstaller

Partners: EMPA Thun, ICP, FLUXIMFunding: Swiss Federal Office of EnergyDuration: 2011–2012

This project is a joint research effort amongthe Laboratory for Mechanics of Materials andNanostructures of the EMPA Thun and the Insti-tute of Computational Physics of the Zurich Uni-versity of Applied Sciences, to investigate andimprove - both numerically and experimentally- the optical characteristics of nano-structured,extremely-thin absorber solar cells.In recent years extremely thin absorber (ETA)solar cells have received considerable attention,and despite their yet low efficiencies, rangingabout 2.5 %, they are considered as promisingnovel devices for converting sunlight into elec-tric energy. The use of extremely thin absorberlayers has numerous advantages. The probabil-ity that two optically generated carriers recom-bine before they reach the contact electrodesand are thus lost for energy conversion stronglydecreases with the width of the absorber layer.In traditional photovoltaic (PV) applications thisissue was dealt with by employing very expen-sive absorber materials with excellent electroniccharacteristics (i.e. long diffusion lengths). Thehigh price prevents these solar cells from large-scale market penetration. In contrast, ETA solarcells do not only require a strongly decreasedamount of absorber materials but also allow forthe use of absorbers with significantly poorerelectronic properties that come at a strongly re-duced price. Together with rather inexpensiveprocessing procedures ETA solar cells are verypromising from both an academic and an eco-nomic perspective.However, the efficiency of ETA solar cells is notonly determined by their electrical, but also bytheir optical characteristics. The ability to ef-ficiently trap and convert incident sunlight is akey feature that decides about economic suc-cess. In many cases the percentage of the inci-dent light that is converted into free charges de-pends exponentially on the absorber thickness.In order to combine the advantages inherent tothin absorbers with the necessity of high opticalabsorption efforts must be made to enlarge theoptical path in such way that the solar cell canabsorb a significant fraction of the incident sun-light. To gain deeper understanding of the un-

derlying complex optical processes and to ob-tain optimal device designs numerical simula-tions of the complete three-dimensional (3-D)problem are indispensible, as ill-chosen micro-structures can even decrease the device perfor-mance.

Fig. 1: Simulated three-dimensional full-wave emis-sion beam of an LED featuring a micro-structuredsurface.

Yet, due to the combination of large devicesand small micro-structures the accurate multi-dimensional simulation of such structures ismost challenging. At the ICP we developed a3-D simulation software that solves Maxwell’sequations at greatly reduced computational costthanks to a novel numerical method, the Ultra-Weak Variational Formulation. This software al-lows to obtain full-wave solutions for geometriesthat so far could not be numerically analyzeddue to their large size.A strong asset of this research project is thecombination of both experimental and numeri-cal methods. Whereas the EMPA Thun has pro-found knowledge in the field of device design,fabrication, and experimental characterizationthe Institute of Computational Physics providesprofound expertise for numerical device simula-tion.

ZHAW 25


3.3 AEVIOM-Advanced Experimentally Validated Integrated OLEDModel

Contributors: B. Perucco, E. Knapp, B. Ruhstaller

Partners: Philips Research Aachen and Eindhoven, Technical University Dresden,University of Cambridge, University of Groningen,Eindhoven University of Technology, Sim4tec, Fluxim

Funding: EU FP7Duration: 2008–2011

The overall goal of the project AEVIOM is thedevelopment of a mathematical model for or-ganic light-emitting devices (OLEDs) that is ex-perimentally validated. This is what AEVIOM -Advanced Experimentally Validated IntegratedOLED Model - stands for. In the long run anaccurate numerical model is crucial for the im-provement and optimization of OLEDs.Simulations range from three dimensionalMonte-Carlo simulations to our one dimensionalcontinuum model which however integrates thedisorder physics observed in the Monte-Carlocalculations. Thus, the Extended Gaussian Dis-order model (EGDM) accounts for the disorderin organic semiconductors by a field, densityand temperature dependent mobility model.In the course of the third and last year of theEU-funded project the numerical solver, devel-oped by the ZHAW in the previous two projectyears, has been extended for impedance mod-eling. With this approach we complete the setof simulations to be compared with measure-ments, i. e. we can now simulate steady-statecurrent-voltage curves, dark-injection transientsand impedance spectroscopy. Further, meth-ods to extract model parameters from measure-ments were implemented.Within the framework of the AEVIOM project,the interest was primarily on electrical param-eter extraction from measured current-voltagecurves to characterize the materials of an OLEDstack. The algorithm to extract these parame-ters (mainly EGDM parameters) is based on anonlinear least-square method which minimizesthe error between the simulated and measuredcurrent-voltage curve. We observed that themethod only needs a few iterations to solve thenonlinear least-square problem, typically around10 iterations and further, we noticed a large con-vergence radius. The method was developed incollaboration between the ZHAW and FLUXiMand is available in the commercial semicon-ducting thin film optics simulator SETFOS. Themethod not only provides the model parameters

which reproduce the measured current-voltagecurve best, but also calculates other importantkey figures like the correlation between the ex-tracted parameters, the landscape of the errorfunction or the confidence region of the param-eters. As an example, Figure 1 illustrates theerror landscape as a function of the number andwidth of the density of states (DOS) for the elec-trons, N0 and �. The device analyzed is a singlelayer made of a polymer film which is embeddedbetween the electrodes where the charges areinjected. The mobility of the charges is parame-terized according to EGDM.

0.05 0.06 0.07 0.08 0.09 0.1

σ (eV)

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

N0 (

10

21 c

m-3

)

10-5

10-4

10-3

10-2

10-1

100

Fig. 1: The error landscape as a function of the num-ber and the width of the DOS, N0 and σ. The mini-mum is at N0 = 1 · 1021 cm−3 and σ = 0.078 eV

In order to reduce the correlation of the ex-tracted parameters and thus provide a uniqueset of extracted parameters, we extended themethod so that multiple measured current-voltage curves can be fitted simultaneously.Typically in the AEVIOM project, the current-voltage curves were measured for different layerthicknesses and for different temperatures ofthe device. We also introduced the possibilityto specify parameters which must fit a singlecurrent-voltage curve and in parallel, we can op-timize a parameter set which must fit multiplecurrent-voltage curves. Moreover, SETFOS canbe used to fit every physical quantity that can besimulated. This makes SETFOS a very flexibleand useful tool in the application of parameterextraction.

ZHAW 26


3.4 Cost efficient thin film photovoltaics for future electricitygeneration

Contributors: T. Lanz, B. Ruhstaller

Partners: EPFL, EMPA, PSI, Oerlikon Solar, Greatcell, Solaronix, FlisomFunding: SwissElectric ResearchDuration: 2007–2010

In thin film silicon solar cells light scattering atrough layer interfaces is employed to increasethe absorption. The roughness can be con-trolled by varying the deposition conditions ofthe transparent electrodes and the absorbermaterials. To perform an optical simulation ofthe light absorption in these devices it is neces-sary to take into account the scattering of thelight at the rough interfaces as well as interfer-ence effects that may arise in thin layers. Our re-cently developed angular finite element method(aFEM) is capable of taking both effects into ac-count. It is not as computationally expensive asfull-wave methods and can treat - in contrast toray-tracing approaches - light scattering in a di-rect, non-iterative procedure. Thin layers aretreated coherently with wave optics (transfer-matrix formalism). In thick layers propagationof the intensity is employed. The light scatter-ing at rough layer interfaces can be described interms of angular distribution functions that alsodetermine the fraction of scattered light. Fig-ure 1 shows a block schematic of the individualsteps in the computation.

define multilayer structure

formulate interface equations

for every wavelength

compute short circuit current

and generation profile

set up block-matrix equation

extract layer absorbances,

reflection and transmission and Poynting vector

solve block-matrix equation

Fig. 1: Angular finite element method (aFEM) blockschematic.

The necessary input parameters for the simu-lation are the layer thicknesses of all the indi-vidual layers, the refractive indices spectra ofthe materials and the scattering properties of

the interfaces. In figures 2 and 3 the capa-bilities of aFEM are demonstrated for a micro-crystalline silicon solar cell developed at EPFL-IMT Neuchatel. The front and back contact aremade of ZnO with a thickness of 4.8 um. Themicro-crystalline absorber layer has a thicknessof 1.1 µm. The ZnO front contact of this cell hasbeen smoothed after deposition to limit the lightscattering at this interface.

400 500 600 700 800 900 10000.0

0.2

0.4

0.6

0.8

1.0

Wavelength @nmD

EQ

E Μc-Si case C

Μc-Si case D

Μc-Si exp.

Fig. 2: aFEM calculations for the external quantumefficiency (EQE) of a micro-crystalline silicon (uc-Si)solar cell and measurement. Simulation cases C andD assume flat interfaces (no scattering). In case Cthe absorber layer is treated coherently, in case Dincoherently (intensity only).

400 500 600 700 800 900 10000.0

0.2

0.4

0.6

0.8

1.0

Wavelength @nmD

EQ

E Μc-Si case B

Μc-Si case A

Μc-Si exp.

Fig. 3: aFEM EQE calculation of a uc-Si solar celland measurement. Simulation case A assumes a co-herent absorber layer and little scattering. Simulationcase B assumes an incoherent absorber layer and aLambertian scattering distribution.

ZHAW 27


3.5 Modeling, simulation and loss analysis of dye-sensitizedsolar cells

Contributors: M. Schmid, A. Gentsch, J. O. Schumacher

Partners: Laboratoire de Photonique et Interfaces (LPI), EPFLFunding: Gebert Rüf StiftungDuration: 2008–2010

From September 2008 to December 2010 theICP carried out a research project for the mod-eling of dye-sensitized solar cells (DSCs). Theobjective of this project was to develop validatedmathematical models for the DSC.

Steady-state parameter extraction

In the first period of the project we developed avalidated model for the DSC. The model con-sists of an optical part, which simulates re-flectance, transmittance and absorbance of theincoming light within the different layers of theDSC. The optical model is based on forwardray-tracing and thin-film optics. It allows to cal-culate the fraction of absorbed light in the ac-tive layer as well as the spatially resolved rateof generated excited dye states. The gener-ation rate serves as an input for the electricalmodel, which simulates the injection of electronsinto the TiO2, the subsequent charge transportthrough the nanoporous layer and possible re-combination with electrolyte species. In addi-tion, the diffusive transport of redox species inthe electrolyte is taken into account.

1.0

0.8

0.6

0.4

0.2

0.0

EQ

E

800700600500400l / nm

C101, 11.2 µm0.05 M LiI

fabs

EE

SE

Fig. 1: Measured and simulated (bold red lines) EQEsfor illumination from SE side (open circles) and EEside (open triangles) on a 11.2 µm thick test cell

The coupled optical and electrical DSC modelis compared to steady-state measurement datain order to extract the essential parameters forthe steady-state performance of the cell. Twoimportant parameters characterizing the steady-state performance of the DSC are the electron

injection efficiency, which describes how effec-tively the excited dye molecules insert electronsinto the nanoporous TiO2 layer, and the electrondiffusion length, which characterizes the chargetransport properties inside the semiconductor.By measuring the short-circuit current densityand external quantum efficiency (EQE) for il-lumination from either the substrate electrode(SE) side or the electrolyte electrode (EE) sideand comparing these measurement data to sim-ulations, the influence of these two parameterscan be disentangled. In this way injection effi-ciency and diffusion length of the DSC could bedetermined independently. This analysis (an ex-ample is shown in figure 1) was carried out fordifferent types of DSCs (i.e. with different dyes,electrolytes with varying additives).

Transient simulations

Steady-state measurements are not sufficientto understand the detailed physical propertiesof the DSC. Two different DSCs may have thesame photovoltaic parameters (IV-curve, poweroutput, efficiency etc.), but behave differently,when the cells are perturbed by a small time-dependent signal around the steady-state.

Fig. 2: Comparison of measured and simulated smallperturbation current decay allows to extract trappingparameters.

Time-dependent measurement techniques suchas electrochemical impedance spectroscopyand small amplitude photovoltage/photocurrentdecay are crucial diagnostic tools. The de-veloped simulation software allows to simulate

ZHAW 28


these measurement techniques. Then, by com-paring experimental data to simulated data, theinfluence of the different model parameters onthe response can be disentangled. Figure 2shows exemplary a measured and simulatedsmall perturbation photocurrent decay.

Quantitative loss analysis

A particularly instructive feature of the coupledoptical and electrical DSC model is the quan-tification of optical and electric losses. As illus-trated in figure 3, only a small fraction of the totalincident solar spectrum is converted into elec-tric power (10 - 11 % in high-efficiency DSCs).A large part of the solar flux is lost due to opti-cal losses. Excited dye states might relax backto the ground state (injection loss). After injec-tion, at least half of the photon energy is lost dueto thermalization of the injected electron to theTiO2 conduction band level, and due to the off-set between the dye ground state and the redoxenergy level (potential loss). Also, the electronsmay recombine with I−3 in the electrolyte or withdye cations.

Reflection!

Absorption excl. dye!

Transmittance!

Dye!absorption!~ 30-40 %!

Injection losses!

Potential losses!

Recombination!

Electrical output ~ 10 %!

DSC!

T!

R!

Aʼ + Adye!

1!

ECB!EFn!

EF0!

S+ / S!

S+ / S* !

Eredox!

hν!

Solar irradiation!(1000 W m-2)!

eVOC!

Fig. 3: Schematic of the various optical and electricallosses in a DSC.

We show a sample calculation for a DSC cellin table 1 below. A large fraction (53 %) of theincident power is lost due to optical losses. In-jection losses can constitute a significant losschannel (5.5 %). After injection, a major frac-tion of the photon energy is lost due to potentiallosses (20 %). Surprisingly, only a small frac-tion (0.7 %) is lost due to recombination. An-

other small fraction (1 %) is lost to electrical re-sistances in the electrodes. Finally, we find apower output of 75.9 Wm−2 (efficiency of 7.6 %)for the specific DSC considered her. In our opin-ion, the comprehensive loss analysis is an im-portant tool for the optimization of the DSC.

Loss channel Power Current(W m−2) (mA cm−2)

Incident spectrum 872.0 51.5

Reflectance 85.2 5.2Transmittance 317.3 22.4Absorptance 126.7 9.3Injection 54.9 2.3Recombination 6.8 0.9Potential 196.0 —Series Resistance 9.6 —

Output 75.9 11.4

Tab. 1: Sample calculation for a DSC.

Graphical user interface

We implemented the developed DSC modelinto a simulation software called PECSIM(PECSIM=photoelectrochemical cell simula-tor)1. PECSIM is written in Mathematica andis equipped with a GUI. Currently, four simula-tion modes are available:

1. Steady-state analysis

2. Electrochemical impedance spectroscopy(EIS) simulation

3. Transient photovoltage and photocurrentdecay simulation

4. Quantitative loss analysis

Outlook

The present project allowed the ICP to extendits experience of simulating optical devices todye-sensitized solar cells and to broaden theknow-how in the simulation of optical and elec-trochemical processes. The ICP currently evalu-ates the acquisition of several follow-up projects.

1http://www.pecsim.ch

ZHAW 29


3.6 In-line inspection of photovoltaic modules

Contributors: M. T. Neukom, T. Lanz, B. Ruhstaller, N. A. Reinke

Partners: Komax Solar, AltatecFunding: Commission for Technology and InnovationDuration: 2010–2011

This project is concerned with a new in-line in-spection system for the quality control of pho-tovoltaic modules during production. The pho-tovoltaic modules shall be optically excited andelectrically measured in order to determine keyfigures such as the maximum generated powerunder one sun condition. In this project, an inno-vative LED-based excitation is being developedthat allows for a compact, high-intensity and ho-mogeneous source of light suitable for solar cellcharacterization.The industrial partner Komax Solar developsand sells in-line, high-throughput production andtesting systems for industrial solar cell fabrica-tion. Altatec is a microelectronics company witha strong competence in solutions for packag-

ing of optoelectronic components. The Instituteof Computational Physics provides its compe-tence in the development of experimental char-acterization equipment and physical modeling.For characterizing the spectral photocurrent re-sponse of solar cells under test, a dedicatedmeasurement system has been developed thatservers as a benchmarking system. It makesuse of a broadband light source, a monochro-mator and the lock-in technique for a low-noisemeasurement of the spectral photocurrent. Apicture of the developed setup is shown in fig-ure 1.The project also deals with numerical modelingmainly of the optical part of the system and isstill ongoing.

Fig. 1: Experimental setup for spectral photocurrent characterization of solar cells.

ZHAW 30


3.7 Optoelectronic research laboratory

Contributors: M. Bonmarin, K. A. Brossi, M. T. Neukom, N. A. Reinke, R. Ritzmann, S. Züfle

Partners: Various project partnersFunding: ICP, School of EngineeringDuration: ongoing

True-to-life modeling always is based on accu-rate input parameters from the real world. Thepurpose of the optoelectronic research labora-tory (O-LAB) is both providing such input pa-rameters and experimentally validating simula-tion results. Combining advanced computa-tional methods and state-of-the-art experimen-tal methods allows us to develop innovations inmeasurement engineering.

Fig. 1: Optical microscopy image of an organic pho-tovoltaic cell.

The O-LAB is a consequent expansion of ourexisting computational toolbox, facilitating thecollaboration with experimentally-oriented part-ners and industry. Our current investigationscomprise the following fields of interest:

• Ultra-fast photo-thermalheating and detection

• Thin film optical spectroscopy

• Transient optoelectronicdevice characterization

Figure 1 shows a microscopical photography ofelectrode organic solar cell. An impedance an-alyzer used for the characterization of organic

semiconducting devices is depicted in figure 2.Our experimental toolbox comprises fast oscil-loscopes, signal generators, impedance analyz-ers, IR-sources, detectors and cameras, opticalflashers and spectrometers covering the spec-tral range from 0.35 and 14 µm. A measuringsetup for time resolved fluorescence and phos-phorescence spectra is currently under con-struction.The O-LAB supports industrial needs in thecharacterization of sensors and actuators andseveral ICP projects related to thermal labeling,thin-film spectroscopy, organic light emitting de-vices and photovoltaics. In addition, our labora-tory is ideally suited for educational purposes inteaching B.Sc. and M.Sc. students in mecha-tronics, electronics and informatics. The O-LABgives young scientists and engineers the possi-bility of getting in contact with R&D and workingon exciting issues in ongoing projects.

Fig. 2: Impedance analysing instruments for organicsemiconducting device characterization.

ZHAW 31


ZHAW 32


Chapter 4

Student Projects

ZHAW 33


4.1 Optimisation of new generation silicon solar cells by meansof 2D simulations

Students: Andreas Tiefenauer

Category: MSE Master ThesisMentoring: J. John (IMEC), B. RuhstallerHand In: April 2010

This project has been conducted in the frame-work of a master thesis and internship at theBelgian interuniversity microelectronic center(IMEC) in Leuven. Solar cells are large areasemiconductor devices. Other than for mostsemiconductor industry of microelectronic de-vices, the market driver for multi crystalline sili-con solar cells is its cost structure. Since the sili-con price makes about 30 % of the costs dealingwith low quality silicon and with thinner and thin-ner wavers is imperative in the multi cristallinesilicon solar cell industry. However the thinningof wafers brings new difficulties and asks for fur-ther optimisation in various domains of solar cellresearch.Nevertheless, it emerges that reducing the cellthickness is not as straightforward as it seems.For technological and operational reasons ityields into structurally more complex solar cellconcepts than the conventional thicker standardtype cells. Thereby this increase of complex-ity naturally leads to more extensive Design ofExperiments (DoE) which hardly could be donesolely by experimental screening of parameters.At this point the need of simulations is essen-tial. By virtually identifying insensitive param-eters one can finally decrease the size of theexperimental DoE matrix in the lab.Since already existing 1D simulation tools such

as PC1D are not sufficient anymore for morecomplex structures, it was the purpose of thiswork to develop a 2D model that is dedicatedto simulate the opto-electrical behaviour of nextgeneration solar cells. A first version of a 2D FE-model has been created and validated. By cal-culating IV characteristics under various param-eter setups we have successfully shown that the2D model is able to correctly reproduce trendsand optimums for various cell parameters. Themodel showed in general good agreement withexperimental data and some occurring discrep-ancies could be explained by the lack of certain- at that time - not yet implemented features.The model has then been widely extended withmany further features. To mention a few: A re-sistive circuit model has been added in order totake ohmic losses into account. In order to cal-culate the effect of textured front side an addi-tional sub model had to be coupled to the 2Dmodel. And a spectral characterisation modehas been added in order to also calculate IQE,EQE and spectral response. The final modelcontains thus a big variety of features and there-fore allows extensive virtual characterisation ofsolar cells. The model is believed to have agreat potential for availability in reasearch anddevelopement and is further used and extendedin the solar cell group of IMEC.

Fig. 1: Scheme of the 2D FE model coupled with a circuit model accounting for resistive losses.

ZHAW 34


4.2 SETFOS sweep data visualization

Students: André Götz

Category: Projektarbeit, Studiengang Systeminformatik UIMentoring: B. RuhstallerHand In: Dezember 2010

Die ZHAW Spin-off Firma Fluxim AG(www.fluxim.com) entwickelt und vertreibtwissenschaftlich-technische Software (SET-FOS) für organische Bauelemente wie Leucht-dioden und Solarzellen. Eine Anwendung derSoftware betrifft die systematische Variation vonEingabeparametern für die Simulation. Die Sim-ulationen können so umfangreich sein, dassdie Visualisierung der Rechenresultate in SET-FOS und der Vergleich mit Messresultaten er-schwert ist. In dieser Projektarbeit wurde eineeigene Java Applikation zur Visualisierung dervorliegenden Rechenresultate entwickelt. DieApplikation basiert auf Java-Bibliotheken und-Klassen, welche von der Hauptanwendung(SETFOS) stammen. Dies sind Bibliothekenund Klassen für wissenschaftliche Graphen,die interaktiv angepasst werden können. DieApplikation dient somit als Datenbrowser undeine beliebte Anwendung besteht darin, dassAnwender ihre Messdaten mit den simuliertenDaten direkt vergleichen.

Diese Projektarbeit beinhaltete das Design unddie Programmierung einer eigenständigen JavaApplikation zur Visualisierung der durch denSETFOS-Kernel berechneten Simulationsresul-tate. Basierend auf einem Prototyp in Mat-lab wurde eine Oberfläche entworfen, welchees dem Benutzer ermöglicht, intuitiv Simula-tionsdaten mit mehreren Parametern mit Mess-daten zu vergleichen. Der Benutzer kann dabeiDatenspalten per Drag and Drop den Achsender Graphen zuweisen, wo sie in zwei Dimen-sionen dargestellt werden. Eingangsparame-ter, wie zum Beispiel Schichtdicken oder dieangelegte Spannung, können durch Schiebere-gler fixiert werden. Als weiteres Werkzeugwurde eine Kurvenschar-Funktion eingebaut, sodass durch Variation der Eingangsparametermehrere Kurven in einen Graph geplottet wer-den können. Die Hauptfunktionalitäten, diegegenüber der Standard-SETFOS-GUI einenMehrwert darstellen, konnten in dieser Arbeitalle erfolgreich umgesetzt werden.

Fig. 1: Zeigt einen Screenshot des erzielten Software Prototyps.

ZHAW 35


4.3 Mass- and heat-transfer with two-phase �ow, project "earthtube": cooling a school building in Tamil Nadu, India

Students: Remo Ritzmann

Category: Spare time project, MSE Lecture "Heat and Mass Transfer with Two-Phase Flow"Mentoring: T. HockerHand In: 2010–2011

Cooling buildings in hot summertime near theequator is an energy intense task. Two sum-mer months with up to 40 ◦C are seen alongsidethree rainy season months around 30 ◦C. Ad-ditionaly there is a temperature gradient of ap-prox. 10 ◦C between daily minimum and max-imum. Air conditioning with an undergroundheat exchanger offers a low priced solution.This project shows how much cooling effect thisbrings.

Fig. 1: Earth tube system with 20 tubes in India.

First of all a 1D transient heat coduction prob-lem has been solved to calculate heat waves inthe earth. The warm-up of soil around the earthtube was calculated by solving the global energybalance. Then a 2D stationary mass and energybalance lead to an understanding of the coolingprocess inside the earth tube.The simulation showed the heat flow from theair into the soil. The temperatures are calcu-lated from the given (turbulent) velocity-profile.With 20 m long pipes in a 30 ◦C ground the airtemperature in the tube falls from 40 ◦C down to32 ◦C. Since there is a climate with lots of hu-midity the simulation has shown that every hour2 dl of Water is condensed per tube per hour. Soin total 4 l=h can be pumped out for watering theplants. Water condensation happens only nearthe tube walls.

Fig. 2: Condensation conditions along earth tube.

The temperature T(x,t) in the ground at depthx and at time t has been simulated. The resultsshowed that the air temperature can be reducedby 8 degrees. Another significant effect on hu-man comfort will be that humidity is beeing re-duced from 90 % to 60 %. Therefore one canperspire more easily in the airconditioned housewith this low priced technique.

Fig. 3: Condensation & heat production rates at tubeend in 3 m depth of sandy ground.

ZHAW 36

http://www.pfunzle.ch/index.php?page=college_passive_cooling_811


4.4 Complete List of Student Projects

1. Peter Bauer, Stephan Ernst, ”ICE-LED, Entwicklung eines LED-Screens für Werbung unterEisflächen in Hockeyarenas“, Projektarbeit Studiengang Elektrotechnik, Betreuer: N. A. Reinke,R. Ritzmann, B. Ruhstaller, Firmenpartner: Martin Altorfer (Altira GmbH), Winterthur (2010).

2. S. Brunner, L. Imholz, ”Analyse von thermisch induzierten Fertigungstoleranzen bei derHerstellung von Schwingquarzen“, Projektarbeit Studiengang Maschinentechnik, Betreuer:T. Hocker, Firmenpartner: Micro Crystal AG (Swatch), Grenchen (2010).

3. L. Candrian, O. Hasler, ”Validierung einer Nernst-Sonde zur variablen Regelung der Erdgas-reformierung im Hexis Brennstoffzellen-System“, Projektarbeit Studiengang Maschinentech-nik, Betreuer: M. Linder, T. Hocker, Firmenpartner: Hexis AG, Winterthur (2010).

4. Chantal Eschbach, Ivo Günther, ”Charakterisierung von Solarzellen mit einem spektralen IU-Kennlinien-Messplatz“, Projektarbeit Studiengang UI, Betreuer: N. A. Reinke, M. T. Neukom(2010).

5. A. Götz, ”SETFOS sweep data visualization“, Studiengang Systeminformatik UI, Betreuer:B. Ruhstaller, Firmenpartner: Fluxim AG, Winterthur (2010).

6. Homajun Ahadyar, Pascal Schori, ”Temperatur-geregelter Gasstrom für die Untersuchung or-ganischer Leuchtdioden und Solarzellen“, Projektarbeit Studiengang , Betreuer: N. A. Reinke,M. T. Neukom, S. Züfle (2010).

7. Fridolin Hösli, Stefan Kuch, Simon Lehmann, ”Algenreaktor“, Projektarbeit Studiengang UI,Betreuer: N. A. Reinke, K. A. Brossi (2010).

8. L. Kaufmann, R. Schmid, ”Optimierung eines Prüfstands für thermo-mechanische Belas-tungstests von keramischen Brennstoffzellen“, Projektarbeit Studiengang Maschinentechnik,Betreuer: T. Hocker, Firmenpartner: Hexis AG, Winterthur (2010).

9. Adrian Koller, Robin Jergen, ”ASURO Rasenmäher (ASMÄHRO), RPK - Rasenmähen PerKnopfdruck“, Projektarbeit Studiengang MT, Betreuer: A. Fassbind, R. Ritzmann (2010).

10. M. Linder, ”Modellbasierte Analyse der Spannungs- Strom-Kennlinie von SOFC Hochtem-peratur Brennstoffzellen“, MSE Masterarbeit, Betreuer: T. Hocker, Firmenpartner: Hexis AG,Winterthur (2010).

11. Alexander Pohl, Marco Bär, ”1DES One Doctor Endoscopic Surgery, Foot operated laparo-scopic telescope“, Projektarbeit Studiengang Mechatronik, Betreuer: T. Järmann, R. Ritz-mann, Firmenpartner: Dr. J. Gnanaraj, India (2010).

12. Juan Romero, Fabio Lambreschi, ”Raytracing Applikation für die Simulation optoelektronis-cher Bauelemente“, Projektarbeit Studiengang UI, Betreuer: N. A. Reinke, M. T. Neukom,(2010).

13. R. Ringenberg, E. Schmidhauser, ”Entwicklung eines Injektors für die Brenngaszufuhr inBrennstoffzellensystemen“, Projektarbeit Studiengang Maschinentechnik, Betreuer: M. Lin-der, T. Hocker, Firmenpartner: Hexis AG, Winterthur (2010).

14. R. Ritzmann, ”Mass- and heat-transfer with two-phase flow, project "earth tube": cooling aschool building in Tamil Nadu, India“, Freizeitprojekt, Betreuer: T. Hocker, Firmenpartner:Chance für Morgen, Winterthur (2010–2011).

15. Jonas Schumacher, Toni-Ante, ”LED-Blitz 2“, Projektarbeit Studiengang UI, Betreuer: N. A. Reinke(2010).

16. P. Specker, S. Wolff, ”Stationärer Betrieb eines Labor-Holzvergasers“, Projektarbeit Studien-gang Maschinentechnik, Betreuer: C. Meier, T. Hocker (2010).

ZHAW 37

http://www.karunakarya.org/


17. M. Steidinger, R. Christinger, ”Populationsbilanzen zur Modellierung des Bruchverhaltensvon pharmazeutischen Wirkstoffen“, Projektarbeit Studienrichtung Material- und Verfahren-stechnik, Betreuer: R. Axthelm, T. Hocker, Firmenpartner: Novartis Pharma AG, Basel(2010).

18. A. Tiefenauer, ”Optimisation new generation silicon solar cells by means of 2D simulations“,Master thesis and internship at the belgian interuniversity microelectronic center (IMEC),Betreuer: J. John, B. Ruhstaller, Leuven (2010).

19. D. Winter, V. Beer, ”Untersuchung des Rissbildungsmechanismus in Brennstoffzellen“, Pro-jektarbeit Studienrichtung Material- und Verfahrenstechnik : Betreuer: D. Penner (IMPE),T. Hocker, Firmenpartner: Hexis AG, Winterthur (2010).

ZHAW 38


Appendix

ZHAW 39


A.1 Scientific Publications

1. J. Eller, J. O. Schumacher, G. Sartoris, B. Seyfang, and T. Colinart: A 2+1D model of a protonexchange membrane fuel cell with glassy-carbon micro-structures, Special Issue of Mathe-matical and Computer Modelling of Dynamical Systems (MCMDS) (accepted for publication,2010).

2. L. Holzer, B. Iwanschitz, T. Hocker, B. Münch, M. Prestat, D. Wiedenmann, U. Vogt, P. Holtap-pels, J. Sfeir, A. Mai, and T. Graule: Microstructural degradation in SOFC anodes: quanti�-cation of nickel grain growth in dry and in humid atmospheres, Journal of Power Sources,vol. 196, pp. 1279–1294 (2011).

3. L. Holzer, B. Münch; B. Iwanschitz, M. Cantoni, and T. Hocker: Quantitative relationships be-tween composition, particle size, triple phase boundary length and surface area in Ni-cermetanodes for Solid Oxide Fuel Cells, Journal of Power Sources, doi:10.1016/j.jpowsour.2010.08.006(in press).

4. E. Knapp, R. Häusermann, H. U. Schwarzenbach, and B. Ruhstaller: Numerical simulationof charge transport in disordered organic semiconductor devices, J. Appl. Phys. 108, 054504(2010).

5. E. Knapp and B. Ruhstaller: Numerical Analysis of Steady-State and Transient Charge Trans-port in Organic Semiconductor Devices, OQE, accepted for publication (2010).

6. J. Kuebler, U. Vogt, D. Haberstock, J. Sfeir, A. Mai, T. Hocker, M. Roos, and U. Harnisch: Sim-ulation and Validation of Thermo-mechanical Stresses in Planar SOFCs, Fuel Cells, vol. 10,pp. 1066–1073 (2010).

7. M.T. Neukom, N. A. Reinke, K. A. Brossi, and B. Ruhstaller: Transient Photocurrent Re-sponse of Organic Bulk Heterojunction Solar Cells, Proc. SPIE 7722 30 (2010).

8. B. Perucco, N. A. Reinke, F. Müller, D. Rezzonico, and B. Ruhstaller: The in�uence of theoptical environment on the emission pro�le and methods of its determination, Proc. SPIE7722 14 (2010).

9. B. Perucco, N. A. Reinke, D. Rezzonico, M. Moos, and B. Ruhstaller: Analysis of the emissionpro�le in organic light-emitting devices, Optics Express, Vol. 18, Issue S2, pp. A246-A260(2010).

10. D. Rezzonico, B. Perucco, E. Knapp, R. Häusermann, N. A. Reinke, F. Müller, and B. Ruh-staller: Numerical analysis of exciton dynamics in organic light-emitting devices and solarcells, J. Photon. Energy 1, 011005 (2011).

11. B. Ruhstaller, E. Knapp, N. A. Reinke, M. Moos, B. Perucco, D. Rezzonico, and M. H. Lu:Analysis of Exciton Distributions in OLEDs: The In�uence of Optical Modes and EnergeticDisorder, SID Symposium Digest of Technical Papers, Volume 41, Issue 1, pp. 1894-1898(May 2010).

12. S. Wenger, M. Schmid, G. Rothenberger, M. Grätzel, and J. O. Schumacher: Understand-ing steady-state behavior in dye-sensitized solar cells using a validated coupled optical andelectrical model, Journal of Physical Chemistry C (in press).

A.2 News Articles

1. E. Heinzelmann: Anti-Aging für Brennstoffzellen, In Auftrag von Swisselectric Research(SWISS ENGINEERING STZ). Mai 2010

2. N. A. Reinke and A. Bariska: Einfache Beschichtungskontrolle mit thermischer Schichtprü-fung, Bulletin 10s/2010 itg-Sonderausgabe / numéro spécial itg (2010).

ZHAW 40


3. N. A. Reinke and B. Ruhstaller: Organische Elektronik, Swiss Engineering STZ (Januar/Februar2010). Organische Elektronik

4. N. A. Reinke: Sicherheit dank Winterthurer Technik, Winterthurer Stadtanzeiger, Ausgabevom 21.12.2010 (21.12.2010).

5. N. A. Reinke: Soirée Electrique, blitz 04 Fachzeitschrift des AMIV an der ETH, 44. Jahrgang(17. November 2010).

6. N. A. Reinke: Soirée Electrique, Bulletin (2010).

7. N. A. Reinke: Soirée Electrique, Impact (2010).

8. N. A. Reinke: Thermische Schichtprüfung, Swiss Innovation Guide 2011 HandelszeitungBeilage im Handelsblatt (November 2010).

9. B. Ruhstaller and N. A. Reinke: Organische Elektronik mit Perspektive, Bulletin 9/2010 Ju-biläumsausgabe/numéro du centenaire (2010).

A.3 Exhibitions

1. Control, Stuttgart, 2010.

2. Nationale Photovoltaik Tagung, Winterthur, February 2010.

3. Swiss Innovation Forum, Basel, 2010.

4. Symposium für virtuelle Produktentwicklung Schweiz, HSR, Rapperswil, April 2010.

5. Thurgauer Technologietag, Schönenberg, 19. März 2010.

6. Winterthurer Nacht der Technik, ZHAW, Winterthur, 2010.

A.4 Conferences and Workshops

1. K. A. Brossi and M. T. Neukom: Fundamental Properties of Devices - Sensors, Transistorsand Solar Cells, Winterschool on Organic Electronics, Donnersbach (Austria), 6.–12.03.2010.

2. T. Hocker: Model-based analyses of the Hexis SOFC system: from electrode degradationto overall system behavior, invited talk, IEA International Energy Agency Annex 25 Meeting,Winterthur, April 2010.

3. T. Hocker, B. Iwanschitz, and L. Holzer: Analysing microstructures to predict the performanceand degradation of Ni-8YSZ anodes, invited talk, Workshop on Image Analysis in SOFCDegradation Research, Brussels (Belgium), September 2010.

4. T. Hocker, B. Iwanschitz, and L. Holzer: Assessing the effect of electrode microstructureon repeat unit performance and cell degradation, Modval7 - 7th Symposium on Fuel CellModelling and Experimental Validation, Morges, March 2010.

5. T. Hocker, B. Iwanschitz, and L. Holzer: Assessing the effect of SOFC electrode microstruc-ture on its performance and degradation, invited talk, Symposium on Solid Oxide Cell Elec-trodes in 3D, Risø, National Labs (Denmark), July 2010.

6. T. Hocker, B. Iwanschitz, and L. Holzer: Using arti�cial, computer generated electrode mi-crostructures to predict the performance and degradation of SOFC anodes, invited talk,Transpore 2010 - International Symposium on Transport in Porous Materials, PSI Villigen,August 2010.

ZHAW 41

http://www.icp.zhaw.ch/fileadmin/user_upload/engineering/_Institute_und_Zentren/ICP/inthemedia/stz0120103814-1.pdf


7. L. Holzer et al.: The microstructure of porous SOFC electrodes: new methodologies for reli-able quanti�cation and link with performance, 9th European SOFC Forum, Lucerne, June/July2010.

8. B. Iwanschitz, A. Mai, T. Hocker, L. Holzer, and M. Schütze: Degradation of Ni-Cermet An-odes in Solid Oxide Fuel Cells, 9th European SOFC Forum, Lucerne, June/July 2010.

9. E. Knapp: Impedance Analysis, International Simulation Workshop on Organic Electronicsand Photovoltaics 2010, Winterthur, Switzerland, July (2010).

10. E. Knapp: Impedance Analysis of Organic Light-Emitting Devices with Disorder, SOLEDOptics and Photonics Congress, Karlsruhe, 22/06/2010.

11. E. Knapp and B. Ruhstaller: Numerical Analysis of Steady-State and Transient Charge Trans-port in Organic Semiconductor Devices, NUSOD, Atlanta, USA, 07/09/2010.

12. T. Lanz and B. Ruhstaller, C. Battaglia, F.-J. Haug, and C. Ballif: Light scattering model foroptical simulation of thin �lm silicon solar cells, 25th European Photovoltaic Solar EnergyConference and Exhibition (EU-PV-SEC), Valencia, Spain, 2010.

13. T. Lanz and B. Ruhstaller: Coherent and incoherent optical simulation of thin �lm solar cells,12th Euregional Workshop on Novel Concepts for Future Thin-film Silicon Solar Cells, Delft,The Netherlands, 28.1.2010.

14. T. Lanz and B. Ruhstaller: Light Scattering Simulation for Thin Film Silicon Solar Cells,SOLED Optics and Photonics Congress, Karlsruhe, 22/6/2010.

15. M. Linder, T. Hocker, R. Denzler, A. Mai, and B. Iwanschitz: Automated model-based analysisof i-V data for Hexis SOFC short stacks running on CPO-reformed natural gas, Modval7 - 7thSymposium on Fuel Cell Modelling and Experimental Validation, Morges, March 2010.

16. M. Loeser and B. Ruhstaller: Three-Dimensional Full-Wave Optical Simulation of LEDs,SOLED Optics and Photonics Congress, Karlsruhe, 22/6/2010.

17. M.H. Lu, B. Perucco, D. Rezzonico, E. Knapp, N. A. Reinke, and B. Ruhstaller: ExcitonDistributions in OLEDs: In�uence of Optical Modes and Energetic Disorder, SID Symposium,Seattle, USA, 25/5/2010.

18. C. Meier and T. Hocker: Modeling of Fixed-Bed Wood Gasi�ers for Combined Heat andPower Applications, Modval7 - 7th Symposium on Fuel Cell Modelling and ExperimentalValidation, Morges, March 2010.

19. M. T. Neukom and K. A. Brossi: Fundamental properties of devices - sensors, transistors andsolar cells, Winterschool on organic electronics, Donnersbach Austria, 6–12 March (2010).

20. M. T. Neukom, N. A. Reinke, K. A. Brossi, and B. Ruhstaller: Transient Photocurrent Re-sponse of Organic Bulk Heterojunction Solar Cells, SPIE Photonics Europe, Brussels, 12/4/2010.

21. B. Perucco: Analysis of Exciton Distributions in OLEDs, International Simulation Workshopon Organic Electronics and Photovoltaics 2010, Winterthur, Switzerland, July (2010).

22. B. Perucco: The in�uence of the optical environment on the emission pro�le and methods ofits determination, SPIE Photonics Europe, Brussels, Belgium, April (2010).

23. B. Perucco, N. A. Reinke, F. Müller, D. Rezzonico, and B. Ruhstaller: The in�uence of the opti-cal environment and energetic disorder on the shape of emission zones in OLEDs and meth-ods of their determination, SOLED Optics and Photonics Congress, Karlsruhe, 22/6/2010.

24. B. Perucco, N. A. Reinke, F. Müller, D. Rezzonico, and B. Ruhstaller: The in�uence of theoptical environment and energetic disorder on the shape of emission zones in OLEDs andmethods of their determination, SPIE Photonics Europe, Brussels, 12/4/2010.

ZHAW 42


25. D. Rezzonico, B. Perucco, E. Knapp, R. Häusermann, N. A. Reinke, F. Müller, and B. Ruh-staller: Numerical analysis of exciton dynamics in organic light-emitting devices and solarcells, SPIE Optics + Photonics, San Diego, USA, 2/8/2010.

26. D. Rezzonico, B. Perucco, F. Mueller, N. A. Reinke, E. Knapp, and B. Ruhstaller: Design ofMultilayer Organic Solar Cells and Light-emitting Devices, LOPE-C, Frankfurt, 31/5/2010.

27. B. Ruhstaller: Advanced simulation of OLEDs and organic solar cells, Intl. Workshop on Or-ganic Optoelectronic Devices: Simulation and Models of Operation, Castello, Spain, 5/10/2010.

28. B. Ruhstaller: Simulation Software for Organic Electronics, SLN/Opera Optoelectronics Meet-ing, Basel, 25/6/2010.

29. M. Schmid, A. Gentsch, S. Wenger, G. Rothenberger, and J. O. Schumacher: Time-dependentcoupled optical and electric modeling of dye-sensitized solar cells, Hybrid and Organic Pho-tovoltaics Conference HOPV 2010, Assisi, Italy, 23–27 May 2010.

30. M. Schmid, A. Gentsch, S. Wenger, G. Rothenberger, and J. O. Schumacher: Time-dependentcoupled optical and electric model for dye-sensitized solar cells, International SimulationWorkshop on Organic Electronics and Photovoltaics (ISW’10), Winterthur Switzerland, June30–July 2 (2010).

31. J. O. Schumacher, and Y. Safa: Modeling two-phase �ow in the gas diffusion layer of aproton exchange membrane fuel cell, 7th Symposium on fuel cell modeling and experimentalvalidation, Morges, Switzerland, 23–24 March 2010.

32. S. Wenger, M. Schmid, G. Rothenberger, M. Grätzel, and J. O. Schumacher: Understandingcharge transport and recombination in dye-sensitized solar cells using an experimentallyvalidated coupled optical and electrical model, 2010 MRS Spring Meeting, San Francisco,USA, 5–9 April 2010.

ZHAW 43


A.5 Prizes and Awards

1. M. T. Neukom received the Thales - Preis for his Bachelor Thesis Characterisation of opticalpulsed solar cells.

2. N. A. Reinke, Ausgewählter Aussteller auf dem Swiss Innovation Forum 2010

3. N. A. Reinke and A. Bariska, ITG Innovationspreis

4. N. A. Reinke and A. Bariska, Winterthur Instruments, Venture Kick Stage I ICP/IDP-SpinOff

5. B. Ruhstaller, Fluxim AG, CTI Startup Coaching Acceptance

A.6 Teaching

Bachelor of Science

• "Mathematik für Ingenieure 1" (MAE 1) - R. Axthelm

• "Mathematik für Ingenieure 2" (MAE 2) - R. Axthelm

• "Mensch, Technik, Umwelt" für Maschinentechniker (MTU) - T. Hocker, C. Meier

• "Modellbildung und Simulation" für Material- und Verfahrenstechniker (MBS) - T. Hocker,R. Axthelm

• "Physik 1" für Elektrotechnik, Maschinentechnik und Systemtechnik (PHEMS1) - M. Schmid,M. Loeser

• "Physik 2" für Maschinentechnik (PHMT2) - M. Schmid

• "Physik 1" für Unternehmensinformatiker (PhUI 1) - N. A. Reinke

• "Physik 2" für Unternehmensinformatiker (PhUI 2) - N. A. Reinke

• "Reaktions- und Prozesstechnik" für Material- und Verfahrenstechniker (PRTMV) - T. Hocker

• "Signale und Systeme 1" (SiSy 1) - J. O. Schumacher

• "Signale und Systeme 2" (SiSy 2) - J. O. Schumacher

Master of Science

• "Applied Photonics" (Mirco and Nano Technolgoy) - B. Ruhstaller

• "Heat and Mass Transfer with Two-Phase Flow" (Engineering) - T. Hocker

• "Material Properties of Crystals / Tensors for Engineers" (Engineering) - M. Roos

• "Multiphysics Modelling and Simulation" (Engineering) - J. O. Schumacher

ZHAW 44

http://www.electrosuisse.ch/g3.cms/s_page/92930/s_name/itginnovationspreis2010


A.7 Spin-off Companies

Numerical Modelling GmbH

www.nmtec.ch

Numerical Modelling GmbH works in the field of Computer Aided Engineering (CAE) and offersservices and simulation tools for small and medium enterprises. Our core competence is knowl-edge transfer: we bridge the gap between scientific know-how and its application in the industry.With our knowledge from physics, chemistry and the engineering sciences we are able to pro-foundly support your product development cycle. Numerical Modelling speaks your language andis able to conform to given constraints with respect to time and budget.We often create so-called customer specific CAE tools in which the scientific knowledge requiredfor your product is embedded. In this form, it is easily deployed within your R&D departmentand supports actual projects as well as improving the skills of your staff. Ask for our individualconsulting service which covers all areas of scientific knowledge transfer without obligation.

FLUXiM AG

www.fluxim.com

FLUXiM is a provider of device simulation software to the display, lighting, photovoltaics and elec-tronics industries worldwide. Our principal activity is the development and the marketing of thesimulation software SETFOS which was designed to simulate light emission from thin film devicessuch as organic light-emitting diodes (OLEDs), thin film solar cells (organic and inorganic) andorganic semiconducting multilayer systems.Our company name FLUXiM is derived from flux simulation. Our software products are usedworldwide in industrial and academic research labs for the study of device physics and productdevelopment. Check out our references and testimonials for more info. We develop swiss-madesoftware in Switzerland and in addition also provide services such as consulting, training andsoftware development, see services page for more details.

Winterthur Instuments

www.winterthurinstruments.ch

Winterthur Instruments GmbH develops measurement systems for fast non-contact and non-de-structive testing of industrial coatings. These measurement systems can be used to determinecoating thicknesses, material parameters (e.g. porosity) and contact quality (e.g. to detect delami-nation). The system is based on optical-thermal measurements and works with all types of coatingand substrate materials. Our measurement systems provide the unique opportunity of non-contactand non-destructive testing of arbitrary coatings on substrates.

ZHAW 45

http://www.nmtec.ch/

http://www.nmtec.ch/

http://www.fluxim.com/

http://www.fluxim.com/

http://www.winterthurinstruments.ch/

http://www.winterthurinstruments.ch/


A.8 Team

The ICP team members as of December 2010 are listed below.

Name Title Function

Axthelm, Rebekka Dr. rer. nat., Dipl. Math. Research AssociateBonmarin, Mathias Dr. sc. nat., Dipl. Ing. Research AssociateBrossi, Kai B.Sc. in Systems Engineering Research AssistantGentsch, Adrian B.Sc. in Systems Engineering Research AssistantHocker, Thomas Prof. Dr. Chem. Eng. LecturerKaufmann, Lukas B. Sc. in Systems Engineering Research AssistantKnapp, Evelyne Dipl. Rech. Wiss. ETH Research Assistant, PhD StudentLanz, Thomas M.Sc. Physics ETH Research Assistant, PhD StudentLinder, Markus M.Sc. in Engineering Research Assistant, MSE StudentLoeser, Martin Dr. ès. sc. ETH, Dipl. Ing. TUM Research AssociateMeier, Christoph Dipl. Ing. FH ZHAW Research AssistantNeukom, Martin B.Sc. in Systems Engineering Research AssistantPerucco, Benjamin M.Sc. in Engineering Research Assistant, MSE StudentReinke, Nils A. Dr. rer. nat., Dipl.-Phys. LecturerRitzmann, Remo M.Sc. in Engineering Research Assistant, MSE StudentRoos, Markus Prof. Dr., Dipl. Phys. ETH LecturerRuhstaller, Beat Prof. Dr., Dipl. Phys. ETH, e-MBA Lecturer, Head of the ICPSafa, Yasser Dr. sc., M.Sc. Research AssociateSartoris, Guido Dr., Dipl. Phys. ETH Research AssociateSchmid, Matthias Dr. ès. sc., Dipl. Phys. ETH Research AssociateSchumacher, Jürgen Dr. rer. nat., Dipl.-Phys. LecturerSpiess, Esther Administrative AssistantToniolo, Lilian Administrative AssistantZüfle, Simon Dipl. Phys. Research Assistant

The following visiting scientists were staying at the ICP for some time during 2010:

• Alfred Blaum, Dipl. Phys., Switzerland

• Moe Shibita, IAESTE Exchange Student, Japan

• Pekka Luomo, IAESTE Exchange Student, Turku University, Finnland

• Sung Man (Jack) Lai, IAESTE Exchange Student, The Hong Kong Polytechnic University,Hong Kong, China

ZHAW 46


A.9 Location

Building TKInstitute of Computational PhysicsWildbachstrasse 21CH-8401 Winterthur

ContactProf. Dr. Thomas HockerTel.: +41 58 934 78 37Fax: +41 58 935 78 37E-Mail: [email protected]

AdministrationLilian TonioloTel.: +41 58 934 73 06E-Mail: [email protected]

TL

TA

TK

TF

TPTH

TB

TE

TC

TR

TV

TG

TW

TN

TSTO

Mensa

TM

N

Technikumstrasse

Langgasse

Rosenstrasse

Zeughausstrasse

Büelrainstrasse

stra

sse

Eulach

Steinberggasse

Neu

mar

kt

Metzg

gasse

Ho

lderp

l.

Ob

ergasse

Ob

ere Kirch

g.

Un

t. Kirch

gasse

Wild

bach

strasse

P

Winterthur

Campus Technikumstrasse

Departement GesundheitTN EulachpassageTO EulachpassageTS Eulachpassage

School of EngineeringTA MathematikgebäudeTB BibliotheksgebäudeTC ChemiegebäudeTE OstgebäudeTF IMS Institut für Mechatronische Systeme TG InIT Institut für angewandte Informationstechnologie TH HauptgebäudeTL LaborgebäudeTK ICP Institute of Computational Physics TM MaschinenlaborTP PhysikgebäudeTR IDP Institut für Datenanalyse und Prozessdesign TV VerfahrenstechnikTW InES Institute of Embedded Systems

Winterthur

BUS

Winterthur − Campus Technikumstrasse

Technikum

School of Engineering

Tel.: +41 58 934 75 02, [email protected], www.engineering.zhaw.ch

Departement Gesundheit

Tel.: +41 58 934 63 02, [email protected], www.gesundheit.zhaw.ch

Adressdetails auf der Rückseite

Turm

hald

en

strasse

Lagerhausstrasse

P

P

SB

B

BUS

Arc

h-S

tras

se

Mei

sens

tr.

UntertorMarktgasse

BUS

ZHAW 47

http://map.search.ch/winterthur/wildbachstr.21

Zurich University of Applied Sciences

School of EngineeringICP Institute of Computational Physics

Technikumstrasse 9 P.O. BoxCH-8401 Winterthur

Phone +41 58 934 71 [email protected]

zhaw icp research report 2010

Documents