STW MuST Mi2M project proposal - Materials...

STW MuST – Multiscale Simulation Techniques

MuST Technologiestichting STW Administrative data

Project title

A multi‐scale approach towards integrated cohesive interface elements

Short title / Acronym

Multiscale integrated interface Modelling – Mi2M

Applicants

Prof.dr.ir. M.G.D. Geers Section Mechanics of Materials (TU/e‐MoM) Department of Mechanical Engineering (ME), Eindhoven University of Technology (TU/e) P.O. Box 513, 5600 MB Eindhoven, The Netherlands Tel: +31‐(0)40‐2475076; Fax: +31‐(0)40‐2447355; [email protected] Dr.ir. P.J.G. Schreurs, Dr.ir. L.E. Govaert Section Polymer Technology (TU/e‐PT) Department of Mechanical Engineering (ME), Eindhoven University of Technology (TU/e) P.O. Box 513, 5600 MB Eindhoven, The Netherlands Tel: +31‐(0)40‐2472778; Fax: +31‐(0)40‐2447355; [email protected] / [email protected] Prof.dr.ir. E. van der Giessen

Dept. of Applied Physics, University of Groningen (RuG‐Phys) Micromechanics of Materials, Nyenborgh 4, 9747 AG Groningen, The Netherlands. Tel. +31‐ (0)50‐3638046; Fax: +31‐(0)50‐363‐4886; [email protected]

Applications elsewhere

This proposal is original and no concurrent applications have been submitted elsewhere.

Project proposal

1 Summaries

1.1 Research

Cohesively bonded interfaces play a dominant role in a variety of products, commonly based on thin film coated substrates or layered thin film structures. The reliability of such products and devices is often compromised by the occurrence of interfacial separation processes (or delamination), for which dedicated experiments are being developed in industry and academia. However, results found and published, clearly reveal that a macroscopic approach to this problem does not enable the development of truly predictive insights under different loading conditions. The key problem resides in the fact that the total fracture energy encompasses contributions from physical debonding processes and micro‐scale dissipation processed in the bonded material, which cannot be separated without a multi‐scale approach. Due to this lack of insight, the industry is largely relying on trial‐and‐error procedures for each interfacial system to be designed or optimized.

This project focuses on a generic multi‐scale approach that bridges fine scale processes (de‐adhesion, chain pull‐out and scission and micro‐dissipation) to a coarse scale description by means of enriched cohesive zones. A fine scale representation of the interface will be establishes, including geometrical and physical defects, that will trigger the competition between de‐adhesion and micro‐deformation. Dedicated molecular mechanisms will be coarse grained to the fine scale model, where a temperature‐, pressure‐, rate‐ and time‐dependent continuum behaviour will be incorporated. The scale transition to the cohesive zone level will be established in a twofold manner, (i) by means of rigorous entirely computational homogenization scheme, from which (ii) a closed‐form description of the cohesive zone will be extracted. With the obtained generic structure of the cohesive zone model, more complex loading conditions can be handled, where individual contributions of adhesion and micro‐dissipation are to be identified. These results will be used in the context of the problems of each of the participating industrial partners in a separate workpackage, where the obtained model will be parameterized for particular interfaces, thereby ensuring a proper transfer of results.

Multi‐scale integrated cohesive interface Modelling 2

STW MuST – Multiscale Simulation Techniques Mi2M

1.2 Utilisation

Cohesive interfacial failure processes are encountered in coated systems, systems‐in‐foil, flexible/rollable/stretchable electronics, systems‐in‐foil, systems‐in‐package, adhesively bonded load‐carrying components and structures, etc. The interfacial adhesive properties and the mechanical properties are known to have a dominant influence on the reliability and lifetime of such products, thereby directly impacting on the economic revenues.

At present, many products under development are extensively tested before they are being released to the market, in several cases even with a sub‐optimal lifetime. By unravelling the competition between physicochemical adhesive properties and micro‐scale deformation driven dissipation, this project aims to make a strong contribution to 'interfacial engineering', which will help the industrial partners in reducing their time‐to‐market or in improving the reliability of engineered products.

Four industrial partners are directly involved in this research project, where the transfer of results is aligned with in house mirror activities in each of the companies. Because of the required generic nature of the project, a number of parallel tasks have been included in the application‐oriented workpackage that will enable us to make use of the results in dedicated applications (and particular interfaces) to the benefit of each partner. All partners contribute partially in kind and partially in cash to this project. Their in kind contribution ensures their active involvement needed to transfer results in an effective manner. Moreover, several of them will provide and produce samples and test materials (Helianthos, Polymer Vision, Corus), carry out dedicated experimental tests (Philips, Corus) and implement the resulting interface element in a commercial finite element environment (Philips, Corus).

Finally, utilisation will be enforced by involving a number of partners in the users committee, for which the generic approach in this project constitutes a major interest for other applications as well.

2 Research team & infrastructure

2.1 Research team

The research team applying for this research project consists of 2 academic and 4 industrial partners from different sectors. The academic partners are sharing their expertise, relying on their individual strengths on modeling at two distinct length scales, whereas the mutual co‐operation will provide the input needed to establish the scale transitions. The industrial partners unify a common interest in the behaviour of cohesive interfaces, each from their own application perspective. The proposed team therefore gathers the required scientific expertise and the generic industrial interest, which is needed to achieve the goals of this project as a part of the MuST programme. The role of each of the partners and staff members involved is listed below.

Name Discipline/ task University–department/

Industry

Prof.dr.ir. M.G.D. Geers Computational and constitutive homogenization of interfaces described by a micromechanical model

TU/e‐ MoM

Dr.ir. P.J.G. Schreurs Dr.ir. L.E. Govaert

Continuum modelling of debonding polymers 3D visco‐elastic constitutive modelling

TU/e–PT

Prof.dr.ir. E. van der Giessen Molecular debonding of polymers from an interface RuG–Phys

Dr. R. Schlatmann Ir. W. Scheerder

Workpackage 4: Experimental work on flexible solar cells for model verification, including reliability tests

Helianthos

Dr.ir. O. van der Sluis Workpackage 4: Testing, identification, application of interfacial model for stretchable and flexible electronics

Philips

Dr. E. van Veenendaal Workpackage 4: Display testing and reliability analyses for interfaces in rollable displays

Polymer Vision

Dr.ir. G. Nagy Workpackage 4: Experimental testing, identification and application of interface model for polymer‐coated steel

Corus

Post‐Doc Researcher workpackage 1, 1.0 fte, 2 years RuG– Phys

Ph.D. student Researcher workpackage 2/3, 1.0 fte, 4 years TU/e– MoM/PT

2.2 Infrastructure

This project focuses on the development of a dedicated multi‐scale computational tool, which heavily relies on the computational infrastructure present in each of the three groups at both universities. Each of the academic partners makes extensive use of a computer cluster to carry out most of the computational analyses.



The academic partners each rely on the infrastructure present in their R&D labs. The industrial partners enable their in kind contribution, by making use of their respective infrastructure to process the products of interest (Helianthos, Polymer Vision, Corus), to produce samples, carry out dedicated experimental tests (Philips, Polymer Vision, Corus) and to transfer the resulting interface element in a commercial finite element environment (Philips, Corus). For experimental contributions, use will be made of the multi‐scale lab at the TU/e (www.mate.tue.nl/mate/laboratories).

2.3 Researchers

Candidate researchers to carry out the academic workpackages were not yet identified. At the industrial partners, all research and technical support is present to carry out the contributions to this project.

3 Scientific project description

3.1 Introduction and objective

Cohesively bonded interfaces play a dominant role in a variety of products, commonly based on thin film coated substrates or layered thin film structures. The thickness of the films or layers may be in the range of 30 nm up to 100 µm and more. The cohesively bonded systems of interest for this proposal typically involve two different polymers, or a polymer and a substrate of an inorganic material e.g. a metal, glass or silicon. The adhesion between the layers in such systems has been the subject of great scientific and industrial interest over the past decades [Lane, 2003; Singhe and Gratien, 2003]. Most work has been focussed onto brittle interface failure in metal/metal and metal/ceramic interfaces. Ductile failure has been studied by taking into account the (visco)plastic dissipation mechanisms in the adjacent materials [Mishnaevsky and Gross,2005]. Ductile adhesive failure has been investigated using classical fracture mechanics with small scale yielding [Hui et al., 1992; Lane et al., 2000; Saulnier et al., 2004]. The interaction between adhesive and cohesive failure has only recently given some attention [Yao and Qu, 2002]. A systematic approach addressing the relevant scales separately and mutually seems to be missing however.

The classical reasoning in addressing the problem is to study the adhesion on the basis of the physicochemical bonding of the layers at the separating interface. Whereas this provides clear answers in brittle interfaces, it constitutes a major problem for cohesive interfaces. The main reason for this, is the fact that during interfacial separation (or delamination), a considerable amount of energy is dissipated in deforming the adjacent material. As a result, different experimental set‐ups do not provide consistent measurements of the work of adhesion, since the measured fracture energy typically consists of a contribution from (1) actual debonding at the surface, (2) dissipative deformation in the bonded material adjacent to the interface and (3) the interaction between both previous mechanisms. A typical example of this is the measured work of separation in e.g. a polymer‐coated metal, where values of 2 J/m2 [Fedorov et al., 2007], 30 J/m2 [van Tijum et al., 2007] and 194 J/m2 [van den Bosch et al., 2007] have been reported by different techniques on exactly the same material system. The physicochemical bonding was expected to be largely identical in each of the tests used, but the micro‐scale dissipation enforced in each was substantially different. At this point, the scientific community cannot provide predictive values for each situation and the industry is largely relying on trial‐and‐error procedures for each interfacial system to be designed or optimized.

In the particular example studied by Van den Bosch et al. [2006,2007] it has been demonstrated that the fine scale dissipation mechanism nearby the interface was fibrillation. Delamination comprised an interactive failure process, involving extending fibrils, de‐adhesion from the surface and even breakdown of fibrils. The last mechanism typically involves the pull‐out and scission of molecular chains. At the continuum interface level, all these mechanisms result in an overall work of separation, that is larger than the classical adhesion energy by an amount equal to the mechanically induced dissipation near the interface. Using a coarse scale description of an interface, the fine scale dissipation cannot be discriminated from the role of the adhesion, rendering the resulting models intrinsically case‐specific. In order to overcome this problem, a multi‐scale simulation method is inevitably required.

Objective This project focuses on the development of a multiscale tool that enables the identification and integration of the interaction between adhesion and micro‐scale dissipation in an upscaled cohesive interface description. To this purpose, a fine scale model will be made that integrates continuum defects and molecular contributions to the de‐adhesion and micro‐dissipation. The interfacial system of interest involves at least one polymer (e.g. polymer‐polymer or polymer‐metal). An upscaling strategy will be developed, relying on an interfacial computational homogenization approach, on the basis of which the interaction between adhesion and micro‐deformation processes can be studied. The resulting response will be mapped into a closed‐form constitutive description as well, which is of prime interest for the partners that wish to make use of it. The continuum averaged physicochemical adhesion will be taken as a starting point here, and will not be explored further, since this subject is receiving considerable attention elsewhere.



3.2 Scientific and industrial challenges

The development of a multi‐scale strategy for integrated cohesive interfaces presents a number of challenges that are to be addressed:

• Challenges at the molecular scale: - Coupled modelling of chain pull‐out and viscoelastic deformation processes inside the bonded polymer

bulk(s); - Coupling between normal and tangential separation; - Understanding the influence of temperature, possibly leading to scaling relations that can be directly

used at the next higher length scale; - Determination of competition between pulling‐out and chain scission.

• Challenges at the fine scale: - Accurate and detailed modelling of the deformation and de‐adhesion in the interface region, where

distributed physical and geometrical imperfections in the surface bonding trigger local failure initiation; - Modelling the contribution of continuum debonding processes and implementation thereof in the micro‐

scale model; - Integrating a state‐of‐the‐art description of the bonded polymer system, inducing a temperature‐,

pressure‐, rate‐ and time‐dependency (T,p,e,t) [the choice for the polymer systems of interest will be made at the start of the project jointly with all industrial partners involved].

• Challenges for the scale transition: - Establishing (i) a computational homogenization scheme that maps the response of an interfacial volume

into a coarse scale interface response (in a rigorous, yet entirely numerical manner), accounting for large deformations and preserving the intrinsic load‐dependent properties of the bonded polymer (T,p,e,t), (ii) which will be used to extract the coarse scale closed‐form description of a cohesive zone.

• Challenges at the coarse scale: - Identify all coarse grained cohesive mechanisms and integration into a closed‐form cohesive description

that highlights the individual role of the adhesive properties, micro‐dissipation and all load‐dependent characteristics of the polymer;

- Identify interfacial properties in engineering applications as a function of different loading conditions, to the benefit of the participating industrial partners.

3.3 Multi‐scale character

As emphasized in the introduction, the interplay between the loss of adhesion and micro‐scale dissipative mechanisms cannot be tackled without making use of an integrated multi‐scale approach. This project essentially addresses three interacting length scales:

• The molecular level, needed to assess the dissipative contribution of bonding structures that are too small to fit in a continuum approach.

• The fine scale interface level, at which a detailed continuum model will be used to capture the influence of a distributed adhesion, and irregularities and defects at the level of the bonded interface.

• The macroscopic level, in which the mechanisms identified at the fine scale will be modelled at a constitutive level in terms of bulk constitutive properties and cohesive zone models.

Figure 1. Scales, methodologies and scale transitions up to the applications level, considered in this proposal

Two scale transitions are required. The first one relies on the coarse graining of the molecular results, which will be used in the failure model at the fine scale model. The fine scale model, with a considerable level of physical and



geometrical complexity, will be homogenized to the coarse scale, using a direct computational method [Kouznetsova et al., 2001, 2002, 2004], specific for the type of interface studied here.

3.4 Multi‐disciplinary character & expertise of the research team

The project unifies disciplines in mechanics, physics and materials science, which is a prerequisite to address all the scales of interest. Each of the participating groups has its scientific strengths in one of these disciplines, which are to be integrated through the scale transitions to be made:

• Polymer Technology (Dr.ir. P.J.G. Schreurs, Dr.ir. L.E. Govaert): This group has a long‐standing expertise in the industrial arts of manufacturing polymer‐based products, with a special emphasis on bridging the gap between science and technology in the area of polymer processing and design. The three‐dimensional non‐linear viscoelastic constitutive equations are derived and improved on the basis of experimental observations and include yield, intrinsic strain softening, strain hardening and the important dependence on strain rate and temperature [van Melick et al., 2002; Meijer et al., 2003, 2005; Klompen et al., 2005]. These models are here required at the level of the fine scale interfacial model.

• Mechanics of Materials (prof.dr.ir. M. Geers): A solid expertise has been built up in this group in a number of topics that are relevant for this proposal, i.e. (1) numerical‐experimental characterization of interface models for polymer‐coated metal sheet [van den Bosch et al., 2006, 2007]; (2) modelling of damage in continua, both using quasi‐brittle and ductile enriched models (3) computational homogenization (coarse graining) of continua across scales [Kouznetsova et al., 2001, 2002, 2004; Geers et al., 2001, 2003].

• Physically‐based Micromechanics (prof.dr.ir. E. Van der Giessen): This group has been working on various aspects of polymer modelling inspired directly by physical mechanisms. Landmark results relevant for this proposal include the full‐network model for rubbers [Wu and Van der Giessen, 1993], the first physically‐based cohesive zone model for crazing in polymers [Tijssens et al., 2000, 2002; Basu et al., 2005], micromechanical studies of void growth in polymers and blends [Steenbrink et al., 1997, 1998, 1999; Pijnenburg et al., 1999, 2005] and extensive study of the fracture of homopolymers [Estevez et al., 2005]. In more recent years, the group has been involved in coarse‐grained molecular dynamics simulations of glassy polymers, including in particular the analyses of de‐bonding of two glassy polymers glued together by stitching (bi and tri‐block) co‐polymers [Bulacu, 2008; Bulacu and Van der Giessen, 2005, 2007].

Moreover, the industrial partners make use of the results in applications touching on different disciplines as well.

• Corus R&D is active and interested in interfacial failure processes of polymer‐coated packaging materials. • Philips Applied Technologies is actively involved in the analysis of delamination problems in flexible and

stretchable electronics, which is a major bottleneck in the development of these new technologies. Quantitative models are a key requirement to facilitate the design and ensure a high reliability.

• Helianthos is producing rollable solar cells, the lifetime of which is amongst others determined by the interface properties and loading. The adhesion between different polymer layers and external factors influencing this are of great relevance here.

• Polymer Vision develops and manufactures rollable displays. These displays are large and thin, which makes them vulnerable in reliability. The displays are multilayer systems, which contain many, possibly weak, interfaces and diverse local failure mechanisms. In practical use, various different deformation modes are of interest, such as cyclic loading in rolling and unrolling of the rollable displays. As the application area for rollable displays is novel, Polymer Vision has established an extensive reliability test program, including mechanical tests.

3.5 Description of workpackages and tasks

The work in this project is organised in four workpackages (WPs).

3.5.1 WP1: Coarse‐grained molecular dynamics analysis Recent work by Ph.D. student Bulacu in Groningen has concentrated on computing the work of adhesion provided by block co‐polymers that stitch together two immiscible polymers [Bulacu, 2008]. The coarse‐grained molecular dynamics model was designed so that only the work is computed that is done while the stitching molecules are being pulled out. This has yielded an improved understanding of forced reptation and some interesting scaling relations. In the present workpackage, this will be extended by allowing for concurrent relaxation processes in the bulks to take place. This will couple chain pull‐out to large scale viscoelasticity, including the possibility of shear yielding and crazing, while comparison with the previous results provides information in how the total fracture energy can be attributed to the various mechanisms. The coarse‐grained results will be used in the fine scale continuum model, in particular where the molecular mechanisms described above are not well addressed through the bulk equations used. This applies in particular to tasks T2.4 and T2.5. Two types of systems are relevant here: (1) a polymer–polymer system,



and (2) a polymer perfectly adhering to a rigid substrate (e.g. metal), where pull‐out and chain scission occurs due to its strong bonding to the substrate (the latter is typically observed in polymer‐coated steel, as used by Corus).

The work will be carried out in the following tasks:

• T1.1: Extension of Bulacu’s [2008] simulation tool to include viscoelastic deformation of the bulks. • T1.2: Set up similar simulation tool for chains that penetrate a bulk polymer on one side and are rigidly connected to a stiff

substrate on the other side. • T1.3: Extend the simulation tool of T1.2 with the ability of chain scission. • T1.4: Perform parametric study of tensile debonding of a polymer from a substrate, varying chain stiffness, chain length,

separation rate and temperature; separate dissipation by chain pull‐out or scission from dissipation inside bulk. • T1.5: Formulate cohesive zone traction–separation law for tensile debonding (pass on to WP2). • T1.6: Extend the simulation tool of T1.2 with loading under mixed–mode conditions (tension plus shear). • T1.7: Extend simulations in T1.4 and cohesive law from T1.5 for mixed–mode loading (pass on to WP2). • T1.8: Wrap‐up and write articles and manual of the code. 3.5.2 WP2: Fine scale interface modelling (Part 1 of PhD assignment)

While WP1 focuses on the molecular dissipation processes, the fine scale model will address the dissipation and delamination processes on the length scales of fibrils (figure 1), including void formation, fibril extension, etc. The failure behaviour of individual fibrils, either through brittle fracture or stable necking, will be determined making use of the advanced constitutive models developed (and experimentally verified) in previous [Klompen et al., 2005; Meijer et al., 2003,2005] and ongoing work.

Figure 1. Typical example of an interfacial separation problem that emphasizes the interaction between the loss of adhesion and dissipation near the interfaces [van den Bosch et al.].

Model features:

• The interface will be modelled with a varying degree of geometrical imperfections (roughness, spatially distributed bonding).

• Different loading configurations will be applied to the interface (tension/compression/shear, with different pressures, temperatures, strain rates and ageing history).

• State‐of‐the‐art 3D nonlinear visco‐elasto‐plastic continuum models will be applied at the fine scale model, enriched with a damage model to describe the failure and dissipative processes, e.g. fibrillation will be modelled explicitly by accounting for the visco‐elasto‐plastic deformation of a fibril being pulled out from the polymer matrix.

• The physicochemical adhesion of individual bonds will be considered as a known physical quantity. Nevertheless, its influence will be studied by varying the mean value and its spatial distribution.

• The nucleation and growth of voids at the interface will be incorporated. Nucleation of voids will be incorporated by discrete releasing the adhesion upon exceeding a threshold value for the local traction. To this purpose discretely failing adhesion mechanism will be used at the inhomogeneous interface, for which interfacial elements at a smaller scale need to be used. The constitutive behaviour of these small‐scale interface elements is to be obtained from WP1 (tasks 1.5 and 1.7).

• The constitutive description of failure processes in the bonded polymer or fibrils, which is not accounted for in existing bulk models, will be determined by coarse graining the molecular results of WP1, see task 1.7.

A short description of resulting tasks is given below:

• T2.1: Literature study on latest developments not covered in this proposal. • T2.2: Set‐up of the fine scale geometrical model and associated boundary conditions. • T2.3: Incorporation of a spatially distributed and statistically non‐homogeneous adhesion. Small‐scale interface elements will

be used, for which the constitutive response is determined in WP1.



• T2.4: Implementation of a 3D non‐linear visco‐elastoplastic model that accounts for the pressure, temperature, rate and time‐dependency of the considered polymer .

• T2.5: Study of fibrillation, associated energy dissipation and failure of individual fibrils – update of governing equations (making use of input from WP 1).

• T2.6: 3D fine scale modelling of the interaction between adhesion and micro‐scale dissipation – connection to WP 3.

3.5.3 WP3: Interfacial homogenization & coarse scale cohesive description (Part 2 of PhD assignment)

A fine scale model of the type suggested in WP1 provides a detailed insight in the intrinsic role of several interfacial characteristics. Yet, it cannot be used for practical engineering applications, where the interfaces of interest are subjected to real macroscopic loads (e.g. rolling/bending/shearing for systems‐in‐foil, displays and solar cells; bending/deep‐drawing/wall‐ironing for polymer coated metal sheet, extension for stretchable electronics). WP3 therefore addresses the scale bridging to this macroscopic scale.

At this level, we depart from the recent experience in the PhD work of Van den Bosch, where cohesive zones were developed and characterized with peel tests for PET‐coated steel (figure 2).

Figure 2. Typical example of a coarse scale cohesive zone model used to capture the entire fracture energy for particular loading conditions [van den Bosch et al.]: fibrillation at the interface (left); constitutive description by means of an interconnecting traction vector (middle); 3D implementation and traction‐opening equation (2 most right subfigures).

The interfacial fracture energy was determined as a function of the interfacial roughness (resulting from pre‐deformation). The large value (194 J/m2) obtained was largely due to a huge contribution of micro‐scale dissipation. However, separating the interfacial contributions in the PhD work of van den Bosch was not possible at that scale nor was upscaling the particular complex load‐ and temperature‐dependent properties of the polymer. This workpackage will apply and enrich the computational homogenization model developed in the Mechanics of Materials group, to extract a realistic macroscopic behaviour. The result thereof will be used to develop an updated 3D cohesive zone model, which carries the main fine scale characteristics that impact the coarse scale.

The tasks to be carried out in this WP are:

• T3.1: Computational coarse graining: determine the boundary conditions for different types of macroscopic loading, using appropriate mathematical averaging theorems.

• T3.2: Set up and solve fine scale boundary value problems (BVP) that incorporate the fine scale model developed in WP 2. • T3.3: Computational coarse graining: extract the homogenized properties of the interface on the basis of the BVP solution,

involving tractions, rate of tractions and possibly spatial distribution of the tractions on the interface. • T3.4: Study the coarse grained interfacial response in different loading conditions and analysis of the interaction between

adhesion and micro‐scale dissipation. • T3.5: Development of the structure of a generic interfacial model that captures the phenomena revealed in task 3.5. • T3.6: Parameterization of the interfacial model and identifying individual contributions to the overall fracture energy. • T3.7: Finalize scientific output and wrap up PhD thesis.

3.5.4 WP4: Industrial support – experimental data / validation / implementation This workpackage integrates the industrial support in the project and defines the transfer of results.

• T4.1: (Helianthos) Carry out reliability tests (including thermal cycling, UV exposure tests) identifying the behaviour of individual interfaces in produced solar cells. Modify the microstructures of the adhesive zone for higher strength and adhesive property using different additives.

• T4.2: (Helianthos) Provide materials, test samples and carry out durability and reliability tests, which will be used as input in task T4.3.

• T4.3: (TU/e) Apply the model developed in WP3 to a particular interface of interest for Helianthos, with an additional effort towards determining the parameters (from task T3.6), making use of data provided in task T4.2. This task will be carried out in a MSc project.

• T4.4: (Polymer Vision) Mechanical reliability tests on display cells with the aim of identifying the behaviour of individual interfaces in rollable displays (roll‐up tests, peel tests).

• T4.5: (Polymer Vision) Provide materials, test samples and carry out durability and reliability tests, which will be used as input in task T4.6. A particular focus is made on factors that influence the interfacial failure mechanisms in a display.



• T4.6: (TU/e) Making use of the model developed in WP3, apply it to a particular interface of interest for Polymer Vision, with an additional effort towards determining the parameters (from task 3.6), making use of data provided in task T4.5. This task will be carried out in a MSc project

• T4.7: (Philips) Provide industrial test samples (related to stretchable and flexible electronics). • T4.8: (Philips) Application of the developed numerical models from WP2 and WP3 on real carriers in an industrial environment. • T4.9: (TU/e) Parameter identification and evaluation of the model developed in WP3 for stretchable and flexible electronics.

Close interaction with task T4.6 is here possible. This task will be carried out in a MSc project. • T4.10: (Corus) Provide materials, samples and experimental tests on polymer‐coated steels, related to different contributions

of the interfacial fracture energy. • T4.11: (Corus) Application of the developed numerical models from WP3 on real carriers in an industrial environment, in

support of task T4.12 • T4.12: (TU/e) Parameter identification and evaluation of the model developed in WP3 for polymer‐coated steel. This task will

be carried out in a MSc project.

3.6 Required personnel and equipment

The scientific staff of the academic partners listed in the Table in section 2, will ensure the supervision of the research activities and the transfer of knowledge to the industry as described in the proposal. The senior researchers of the participating industrial partners will ensure the execution of their respective tasks as well as a part of the supervision of the MSc students that will work on implementing the results. All scientific and permanent staff is in position.

We therefore only request funding for 1 Ph.D. student, 1 Post‐Doc and a limited investment.

3.7 Project planning

Time schedule

The planning of all tasks in all workpackages and their mutual relations are indicated in the time table below. Milestones are indicated upon transfer of deliverables (D in Table) during the project and at the end.

Task Year 1 2 3 4Month 1 6 7 12 13 18 19 24 25 30 31 36 37 42 43 48

WP1.1: Visco‐elastic extension of Bulacu's model

WP1.2: Extension to polymer‐substrate interfaces

WP1.3: Include chain scission

WP1.4: Parametric study of tensile debonding

WP1.5: Small‐scale cohesive zone formulation D

WP1.6: Extend T1.2 to mixed‐mode loading conditions

WP1.7: Extend T1.4‐5 to mixed‐mode D

WP1.8: Wrap‐up and write articles and manual of the code D

WP2.1: Literature update

WP2.2: Set‐up fine scale geometrical model & BCs

WP2.3: Incorporation of adhesion & defects

WP2.4: Implementation of 3D visco‐elastoplasticity D

WP2.5: Study of fibrillation and fibril behaviourWP2.6: Analysis of interaction adhesion/dissipation D

WP3.1: Computational homogenization: determine BC'sWP3.2: Set up and solve fine scale BVPWP3.3: Extract homogenized propertiesWP3.4: Coarse scale interfacial competition adh.‐dissip. DWP3.5: Develop structure of generic interfacial model DWP3.6: Parameterization of interfacial model DWP3.7: Publications and PhD thesis D

WP4.1: Testing & adhesion solar cells (Helianthos)WP4.2: Materials, samples & model input (Helianthos)WP4.3: Qualify model (WP3) for application Helianthos DWP4.4: Global testing flexible displays (Polymer Vision)WP4.5: Materials, samples & model input (Polymer Vision)WP4.6: Qualify model (WP3) for application Polymer Vision DWP4.7: Industrial test samples (Philips)WP4.8: Application numerical model WP3 (Philips)WP4.9: Qualify model (WP3) for application Philips DWP4.10: Polymer‐coated steel interfacial tests (Corus)WP4.11: Application numerical model WP3 (Corus)WP4.12: Qualify model (WP3) for application Corus D

3.8 Research elsewhere, national and international cooperation

Considerable research efforts are conducted worldwide related to the behaviour of interfaces between different materials. Within the context of the present proposal, only limited attempts are being made to address the problem in a truly multi‐scale and generic approach. Within the scope of this project, we wish to emphasize the ongoing work in



the group of prof. Geubelle (University of Illinois at Urbana‐Champaign), where numerical homogenization of interface behaviour is being explored. Co‐operation plans are being discussed. Likewise a connection with the CG‐MD work on the fracture of polymers exists with Dr. Basu at IIT Kanpur (former post‐doc and collaboration with group Van der Giessen). This work is considered useful as additional input, but continued collaboration ensures no overlap.

For the required local data on adhesive properties, we will make use of the literature and concurrent work being done in the Department of Chemistry and Physics at the TU/e and the RuG.

3.9 Outreach plan

The work in this project will be integrated in teaching at the M.Sc. and graduate level at each of the institutes involved. Prof. Van der Giessen’s Micromechanics course at RuG will be extended with an optional set of lectures on multiscale modeling of cohesive interfaces. Prof. Geers coordinates the course Micromechanics within the graduate school Engineering Mechanics, where developed concepts of this project can be incorporated in the module on Multi‐scale modeling.

4 Utilisation 4.1 The challenge from the practice and the proposed solution

The experimental characterization and numerical modeling of interfaces involving dissipative layers is seriously compromised by a lack of fine scale information on the different contributions of the overall fracture energy. Depending on the loading conditions, experimental values for the work of separation φ are found that may differ an order of magnitude. Coarse scale models cannot adequately discriminate between the different mechanisms that contribute to φ. This problem is generic for a multitude of applications that involve soft or dissipative layers and for which interfacial failure constitutes a major reliability issue. To solve this, a multi‐scale simulation tool is needed that provides both the details on the fine scale and the coupling to the coarse scale. On this basis, a refined interfacial model can be developed for engineering analyses, in which the essence of the coarse grained fine scale sources of dissipation are preserved. The use of a rigorous computational homogenization scheme as a reference solution, provides the modeling approach its solid and reliable characteristics.

High‐tech character

The problem of interest for this proposal is generic in a great number of high‐tech industrial applications. Interfacial failure of the type studied here is encountered in coated systems, systems‐in‐foil, flexible/rollable/stretchable electronics, adhesively bonded load‐carrying components and structures, etc.

Economic impact

Since interfacial failure processes have a dominant influence on the reliability and lifetime of many products (both existing and still under development), they directly impact on the economic revenues. Many products under development have to be extensively tested before they are being release to the market, in several cases even with a sub‐optimal lifetime. The contribution of this project to ‘interfacial engineering’ is therefore a clear step forward in reducing the time‐to‐market and improving the reliability of engineered products.

Environmental impact

Among some of the applications, clear environmental issues matter. Some of the coated systems are used to seal off substances that may not leak in the environment. In other cases (e.g. polymer coated steel), the use of soft adhering layers prohibits the use of lubricants in forming processes, which considerably reduces the environmental load of the production process. Understanding the behavior of adhesive zones will help the production development of low lost solar cells. This can help reducing the use of fossil fuels which have a serious environmental impact.

4.2 Generic value

The project must constitute a robust approach to analyze interface failure, starting from a very detailed level and condensing results in a macroscopic model. A generic numerical model will come available for the continuum analysis of polymer/substrate interfaces. The generic value is apparent from the various applications of interest:

Applications of the participating industrial partners

• The integrity of flexible and stretchable electronics (Philips Applied Technologies) is largely compromised by the occurrence of undesired delamination and buckling phenomena. Philips Applied Technologies makes extensive use of modeling methodologies based on cohesive zone elements to analyse and understand these phenomena, and ultimately, to prevent these from occurring. Within this context, a clear need has been expressed to discriminate between all dissipative phenomena to yield predictive insights in a variety of loading conditions.

• Polymer coated steel sheet (Corus): Polymeric coatings are familiar technology for the surface modification of metals, from corrosion resistant coatings for industrial equipment to polymeric coatings on vascular stents to improve bio‐compatibility. Corus supports its clients who use polymer coated metals for food packaging. These materials are subjected to a series of forming steps, possibly compromising the polymer‐metal interface.



• Amorphous solar cells (Helianthos): Helianthos (NV Nuon company) is successfully developing a patented roll‐to‐roll (R2R) production technique for amorphous silicon based solar cells on a flexible laminate. This R2R thin film technology has a high potential for low cost production of solar energy. The lifetime and power output of a solar cell is depending on the reliability of the solar cell and polymer encapsulation materials. Therefore the interfacial integrity during thermal and mechanical loading is highly important. Both the manufacturer and the end‐user will benefit from the durability and a reliable performance of the solar cell.

• Rollable displays (Polymer Vision): Polymer Vision is founded in 2006. Polymer Vision’s goal is to provide the leading technologies for large rollable displays in small mobile devices. Within one year of its launch, Polymer Vision demonstrated the world’s thinnest, highest resolution, largest diagonal rollable display. In September 2004, a pilot production line was started with sufficient capacity to demonstrate the commercial viability of the technology. The company has established strategic partnerships with a number of other leading companies specializing in films, organic materials and advanced display technologies. The Readius, as the first commercial mobile device with rollable display is coined, will first be marketed by Telecom Italia Mobile (TIM) in Italy and will be available early 2008. Mechanical considerations are very important in the design and manufacturing of this rollable display device and its successors, since they are large, very thin multilayer stacks that are continuously mechanically stressed when using the devices.

Other applications

• Adhesively bonded systems: Fixation of all kind of different materials is commonly done by polymeric adhesives. Because the two substrates may have widely different properties, special requirements are posed on the adhesive material. Conductive adhesives will also be used in micro‐electronics to replace the solder materials in mechanic‐electrical joints.

• Releasable adhesives: Many applications require adhesives which provide an adequate bonding between two materials, which can be released easily. Food containers, for example, are often closed by glueing an aluminum foil on the upper edge. This seal has to stay during sterilization, and easy to open by the customer afterwards. These conflicting requirements call for dedicated solutions, where optimal use should be made on the load‐dependent interplay between adhesion and micro‐scale dissipation in the adhesive.

4.3 Users committee

Core members (participating industrial partners)

• Helianthos – contact person: Dr. R. Schlatmann, ir. W. Scheerder • Philips – contact person: Dr.ir. O. van der Sluis • Corus – contact persons: Dr.ir. G. Nagy • Polymer Vision – contact person: Dr. E. van Veenendaal

Confirmed additional members

• HOLST (systems‐in‐foil) – contact person: Prof.dr. A. Dietzel • DSM – contact person: Dr.ir. H. Van Melick • Hauzer Techno Coating – contact person: Dr.ir. P. Peeters • Abaqus (Simulia) – contact person: Dr.ir. G‐J. Kloosterman • MSC Software – contact person: Dhr. M. Edelkamp

4.3 Implementation of results at industrial partners

The implementation of the results is largely integrated in workpackage 4 of the proposal. Each partner contributes in kind to the project, which guarantees the partner’s involvement and contributions to the project. Interfacial test data will be gathered for each specific case, making use of mechanical and environmental durability and reliability tests. Workpackage 4 integrates 4 tasks which will facilitate the transfer of results from WP2 and WP3 to the individual interests of each of the partners. These tasks will be supported by MSc students, as indicated.

4.4 Past performance

The proposed research team has been successful in implementing results of academic research into the industrial practice in a number of projects. Relevant to the present MuST proposal, we wish to emphasize: (1) the PhD work of Dr. Kouznetsova on first‐order computational homogenization, which has been transferred to Philips and presently to Corus; (2) the PhD work of Dr. van den Bosch on cohesive zone models, which has been implemented at Corus packaging and Philips; (3) the PhD work of Dr. Mediavilla on modeling ductile damage and crack propagation, which has been implemented at TNO, Philips and Corus; (4) the Post‐Doc work of Bas van Hal, where MSC implemented his cohesive zone control algorithm in MSC.Marc to the benefit of Philips; (5) the PhD work of Dr. Matin and ir. Erinc, where interface models for fatigue in Sn‐Ag‐Cu solders have been implemented in a commercial FE platform, which is



transferred to Philips (STW project EWT.4924); (6) the PhD work of Dr.E.T.J. Klompen, who's 3D nonlinear viscoelastic model for polymers has been made available for the community by software routines, linked to the commercial FE‐program MSC.Marc, and are currently used at DSM and the Max Planck institute.

5. Contracts & patents To the best of our knowledge, we are not aware of any conflicting contracts or patents filed for the proposed research.

6. Requested financial support

6.1 Personnel

1 Ph.D. student (1.0 fte) is requested for a period of 4 years and 1 Post‐Doc (1.0 fte) for 2 years.

6.2 Materials

Specification Amount National travel (1 PhD x 4 years x k€ 2/yr + 1 Post‐Doc x 2 years x k€ 2/yr) k€ 12

6.3 External travel expenses

Specification Amount 1‐2 International visits per year (researchers and their supervisors: 2 x 4 years x k€ 5 / year) k€ 40

6.4 Investments

Specification Amount (€) 20 new high‐performance nodes for a computing cluster at TU/e (20 x k€ 3) k€ 60 8 new high‐performance nodes at RuG (8 x k€ 3) k€ 24 High‐performance PC for each academic workpackage (2 x k€ 4) k€ 8 The total amount requested for investments equals k€ 92

6.5 Contributions of partners

Specification Amount (€) Helianthos: in kind 300 hours (240 hr senior staff/ 60 hr technician); cash k€ 10 k€ 40.0

Philips Applied Technologies: in kind 235 hours; materials k€ 10; cash k€ 5 k€ 39.9

Polymer Vision: in kind 189 hours; materials k€ 10; cash k€ 10 k€ 40.0

Corus : in kind 188 hours; materials&testing k€ 12; cash k€ 8 k€ 39.9

The total contribution of the industrial partners equals k€ 159.8, comprising a cash contribution of k€ 33.

6.5 Overview of the total project costs

The total financial costs for this project is summarized in the table below. All amounts are in euro, exclusive VAT.

Specification Year 1 Year 2 Year 3 Year 4 Total expenses

Personnel (1 fte) PhD. 1 k€ 33.6 k€ 40.0 k€ 42.8 k€ 45.8 k€ 162.2

Personnel (1 fte) Post‐Doc k€ 52.6 k€ 52.6 ‐ ‐ k€ 105.2

Helianthos – senior researcher k€ 9 k€ 6 k€ 6 k€ 9 k€ 30.0

Philips – senior researcher k€ 9 k€ 8 k€ 8 k€ 9.9 k€ 34.9

Polymer Vision – senior researcher k€ 8 k€ 7 k€ 7 k€ 8 k€ 30.0

Corus – senior researcher k€ 9 k€ 7 k€ 7 k€ 8.9 k€ 31.9

Materials k€ 3 k€ 3 k€ 3 k€ 3 k€ 12

Foreign travel k€ 12 k€ 12 k€ 8 k€ 8 k€ 40

Investments k€ 92 ‐ ‐ ‐ k€ 92

Totals k€ 228.2 k€ 135.6 k€ 81.8 k€ 92.6 k€ 538.2 Based on STW salary information dd. July 1st 2007

Required industrial contribution = 25% x k€ 538.2 = k€ 134.6

Realised Industrial contribution = k€ 159.8

Cash contribution from industry = k€ 33

Requested subsidy from STW = k€ 378.4



7. Literature

7.1 Relevant publications from the research group (last 6 years) M.A.H. van der Aa, P.J.G. Schreurs, F.P.T. Baaijens, Mech. Mat., 33, 555‐572, (2001) S. Basu, D.K. Mahajan and E. Van der Giessen, Polymer 46, 7504‐7518 (2005). M. Bulacu and E. Van der Giessen, Journal of Chemical Physics 123, 114901 (2005). M. Bulacu and E. Van der Giessen, Phys. Rev. E.76, 011807 (2007). M. Bulacu, University of Groningen, Ph.D. thesis (E. Van der Giessen, promotor) 2008. M.J. van den Bosch, P.J.G. Schreurs, M.G.D. Geers, Engng.Fracture Mech., 73(9), 1220‐1234, (2006) M.J. van den Bosch, P.J.G. Schreurs, M.G.D. Geers, Eur.J.Mech., 26(1), 1‐19, (2007) M.J. van den Bosch, P.J.G. Schreurs, M.G.D. Geers, Comp. Mech., accep, x, (2007) R. Estevez and E. Van der Giessen, Adv. Poly. Sci. 188, 195‐234 (2005). R. Estevez, S. Basu and E. Van der Giessen, Int. J. Fract. 132, 249‐273 (2005). M.G.D. Geers, V.G. Kouznetsova, W.A.M. Brekelmans, Journal de Physique IV, 11, 145‐152, (2001). M.G.D. Geers, V.G. Kouznetsova, W.A.M. Brekelmans, IJMCE, 1 (4), 371‐386 (2003).

E.T.J. Klompen, T.A.P. Engels, L.E. Govaert, H.E.H. Meijer, H.E.H, Macromolecules, 38, 6997‐7008 (2005).

E.T.J. Klompen, T.A.P. Engels, L. Van Breemen, P. Schreurs, L. Govaert, H.E.H. Meijer, Macromolecules, Vol 38, pp 7009‐7017, (2005) V.G. Kouznetsova, W.A.M. Brekelmans, F.P.T. Baaijens, Comp. Mech., 27, 37‐48, (2001). V.G. Kouznetsova, M.G.D. Geers, W.A.M. Brekelmans, Int. J. Numer. Meth. Engng, 54, 1235‐1260, (2002). V.G. Kouznetsova, M.G.D. Geers, W.A.M. Brekelmans, Comput. Methods. Appl. Mech. Engrg., 193(48‐51), 5525‐5550, (2004). V.G. Kouznetsova, M.G.D. Geers, W.A.M. Brekelmans, Int. Jnl. Multiscale Comp. Eng., 2(4), 575‐598, (2004). H.E.H. Meijer, L.E. Govaert, Macromol. Chem. Phys., 204(2), 274‐288, (2003) H.E.H. Meijer, L.E. Govaert, Prog. Polym. Sci., 30(8), 915‐938, (2005) H.G.H. van Melick, L.E. Govaert, H.E.H. Meijer, Polymer, 44, 2493‐2502 (2003). H.G.H. van Melick, A. van Dijken, J.M.J. den Toonder, L.E. Govaert, H.E.H. Meijer, Philos. Mag. A, 82(10), 2093‐2102, (2002) K.G.W. Pijnenburg, Th. Seelig and E. Van der Giessen, Eur. J. Mech. A/Solids 24 (2005) 740‐756. B.A.G. Schrauwen, R.P.M. Janssen, L.E. Govaert, H.E.H. Meijer, Macromolecules, 37, 6069‐6078 (2004). M.G.A. Tijssens and E. van der Giessen, Polymer 43, 831‐838 (2002).

7.2 Other references related to this research proposal A.V. Fedorov, R. van Tuijm, W.P. Vellinga and J.Th.M. De Hosson, Progress in Organic Coatings, 58, 2007, pp. 180‐186.

C‐Y Hui, D‐B Xu, E.J. Kramer, J. Appl. Phys., 72 (8), 1992, pp. 3294‐3304

M. Lane, Interface fracture, Annu. Rev. Mater. Res., 33, 2003, pp. 29‐54

M. Lane, R.H. Dauskardt, A. Vainchtein, H. Gao, J. Mater. Res., 15, 2000, pp. 2758‐2769

L.J. Mishnaevsky, D. Gross,Transactions of the ASME, 58, 2005, pp. 338‐353 K.G.W. Pijnenburg, A.C. Steenbrink, E. van der Giessen, Polymer 40 (1999) 5761‐5771.

F. Saulnier, T. Ondarcuhu, A. Aradian, E. Raphael, Macromolecules, 37, 2004, pp. 1067‐1075

R.P. Singhe, A. Gratien, J. Adhesion Sci. Technol., 17, 2003, pp. 871‐888 A.C. Steenbrink, E. van der Giessen and P.D. Wu, Journal Mech. Phys. Solids 45 (1997), 405‐437. A.C. Steenbrink and E. van der Giessen, J. Mech. Phys. Solids 47 (1999) 843‐876. A.C. Steenbrink, E. van der Giessen and P.D. Wu, J. Mat. Science 33 (1998) 3163‐3175. R. van Tijum, W.P. Vellinga, J.Th.M. De Hosson, J. Mater Sci., 42, 2007, pp. 3529‐3536.

R. Vayeda, J. Wang, Int. J. of Adhesion and Adhesives, 27, 2007, pp. 480‐492 P.D. Wu and E. van der Giessen, J. Mech. Phys. Solids 41 (1993) 427‐456. M.G.A. Tijssens, E. van der Giessen and L.J. Sluys, Mech. Mater. 32 (2000) 19‐35. R. Estevez, M.G.A. Tijssens and E. Van der Giessen, J. Mech. Phys. Solids 48 (2000) 2585‐2617.

Q. Yao, J. Qu, Journal of Electronic Packaging, 124, 2002, pp. 127‐134

8. Keywords, abbreviations & acronyms

Keywords

Cohesive zone, adhesion, delamination, interface

Definitions and abbreviations

CZ Cohesive Zone BC Boundary Conditions CG Coarse‐Grained BVP Boundary‐Value Problem MD Molecular Dynamics FEM Finite Element Method

Appendix. Potential referees : A list of 5 world renowned experts on the topic of this proposal.

Appendix. Confirmation letters: Three letters written by competent partner authorities, officially stating their technical and financial contributions.

STW MuST Mi2M project proposal - Materials...

Documents

Transcript of STW MuST Mi2M project proposal - Materials...