First CSC Meeting, Munich Nov 2005 - Initial topics

1. Materials Characterisation: Progress to Date and Future Plans Trevor Dutton, NAFEMS/Dutton Simulation Review of progress from project initialisation through the first three workshops plus plans for invited presentations for the next workshop

2. First CSC Meeting, Munich Nov 2005 - Initial topics • Proposed topics for consideration under Materials Characterisation (from FENET plus members input): • Composites, foams, new materials • Fracture mechanics & durability • Manufacturing simulation • Constitutive models and material data • Modelling connections • Initial suggestions for discussion at first workshop

3. Barcelona TW1, Jan 2006 - Way forward • Initial proposals for topics were presented • Use Materials theme to explore “gaps” determined from Confidence and Integration theme? • How to determine topics for further review / detailed presentations? • Questionnaire • Based on FENET topics – expanded following discussion • Based on FENET classification

4. Topics for Questionnaire • New Materials • Metals (including High Strength & Ultra High Strength Steels, Aluminium, Magnesium, Stainless Steel) as well as Polymers, Composites, Foams, Elastomerics • Material Data, especially • Strain rate sensitivity, Fracture / Failure, Fatigue life, Creep, Formability • Material Modelling • Constitutive Models, Multi-physics aspects, Micro-characterisation, Multi-scalar Modelling, Algorithms, Buckling / post-buckling, Fracture Mechanics, Damage / deterioration modelling • Modelling Connections • Spotwelds, Continuous welds, Self-piercing rivets, Clinching, Bonding • Manufacturing Simulation • Forming simulation, Effects of Forming, Assembly simulation, Paint simulation, Process Flow, Handling, Logistics, Cost analysis

5. TEHNOLOGY READINESS LEVEL Technological Readiness Level (TRL) indicates the state of the art or technological maturity.

6. TEHNOLOGY READINESS LEVEL cont.

7. STATE OF PRACTICE PRIORITY LEVEL INDEX State of Practice (SoP) gives a measure of industrial maturity. Priority Level Index (PLI) indicates the relevance for industry.

8. Classification of Indices Topics for best practice ↓ SoP (state of practice)  PLI (priority) Topics for breakthrough technologies ↓ TRL (tech. readiness) SoP Breakthrough technologies Best practice TRL

9. Munich TW2, May 2006 – Refine issues • Review of Questionnaire Results • Presentations • None specific to Materials Characterisation Topic; • However, papers from Integration and Confidence again emphasised the inter-relationship • Discussion points included: • Continued reference to overlap between themes, e.g., • Quality of material data (GIGO) impacts on Confidence • Effects of Forming impacts on Integration • Use Materials to help fill gaps / remove barriers identified by other two topics • Should we invite presentations from Material Suppliers? • What level of detail should we be working down to? • It does not seem sensible that we should discuss specific material data requirements? • What about selection of appropriate material models?

10. Questionnaire Results - Caveats • Number of Returns • Total of 23 Questionnaires returned, not all fully completed (NB blank fields were not included as “0” answers but any field set to “0” was included as a response) • Number of Responses to specific questions • maximum of 17, minimum of 10, average 14 • Range of Responses • A number of cases where TRL, SoP and PLI rated 0 by one person and 9 by another! • Of the 31 topics proposed only five scored a PLI below 4, and nine scored PLI below 5 • NB Includes two additional returns received immediately after the Workshop

11. Questionnaire Analysis • The following slides summarise, for each of the five categories, the responses received for each topic • The results are displayed as “bubble charts”; the location of the bubble is defined by TRL on the abscissa and SoP on the ordinand; the size of the bubble depends on the PLI score • Full results with tables were presented at Munich – presentation is available to download from AUTOSIM

12. New Materials – bubble chart SoP TRL

13. Material Data – bubble chart SoP TRL

14. Material Modelling – bubble chart SoP TRL

15. Modelling Connections – bubble chart SoP TRL

16. Manufacturing Simulation – bubble chart SoP TRL

17. All five category results together SoP No results group seems to stand out – and there are no obvious “off-axis” results Also, those results extending towards the lower left of the plot are not necessarily the higher priority scores TRL

18. All five category results together (loci) SoP No results group seems to stand out – and there are no obvious “off-axis” results Also, those results extending towards the lower left of the plot are not necessarily the higher priority scores New Materials Material Data Connections Manufacturing Simulation Material Models TRL

19. Questionnaire Results – Additional topics

20. Questionnaire Results – Top Three Topics mentioned more than once include Fatigue, Multi-physics, Foams, Failure and Modelling Connections (Spotwelds, Bonding, Continuous Welds)

21. Observations • New Materials • All similar (high) priorities; High Strength Steels highest, Stainless Steel lowest • Polymers, Foams and Composites are at lower TRL/SoP • Of the metals, Magnesium noted as high PLI, lower SoP • Material Data • Fatigue Life show highest priority, then Fracture/Failure and Strain Rate • Creep and Formability lower PLI but also lower TRL/SoP • Material Models • Constitutive Models has highest PLI (but also quite high TRL/SoP?) • Multi-physics, Micro-characterisation, Multi-scalar Modelling and Algorithms all have low PLI and TRL/SoP? • Fracture Mechanics, Cracking & Damage/Deterioration have high PLI and low TRL/SoP • Modelling Connections • Spotwelds have highest PLI, though all are above 5, but Continuous Weld, SPR (especially) & Bonding have lower TRL/SoP than Spotwelds • Manufacturing Simulation • Effects of Forming and then Forming Simulation have highest PLIs here • Assembly, Paint (especially), Process Flow, Cost all low PLI and TRL/SoP

22. Conclusions • High priority topics also seem to have higher Technology Readiness Level and/or State of Practice ratings; so the results have not automatically highlighted the key topics for attention • Perhaps the results are too dependent on the type of analysis (e.g., fatigue of little priority in crash analysis, strain rate of little priority in durability) • Hence: • All Materials seem to have some interest • Material Data requirements for all topics have some interest • Material models in general noted for attention; fracture & damage aspects noted • Modelling connections in general all noted • Forming simulation & effects of forming highlighted, other manufacturing issues of lesser interest

23. Breakout Group – Munich TW2 • Taking the results of the questionnaire as a starting point • Goal: to identify areas for definition of Best Practice • Matrix Concept – Materials in columns vs. Topics in Rows • Further broken down into Loading Types: • Quasi-static • Cyclic (frequency dependent) • Short term transient (i.e., impact) • Long term transient • First matrix (quasi-static) completed at the meeting; the rest were circulated and finalised over the summer

24. Matrix interpretation • Tick mark identifies a given topic as important for a particular material, for the current load case • Then, any “ticked” cell in the matrix that is felt to be a topic where Best Practice is not yet established is highlighted in yellow • Shaded cells indicate that a change was proposed during the review process

25. Matrix - Quasi-static Load

26. Matrix – Short Duration Transient Load

27. Matrix – Long Duration Transient Load

28. Matrix – Cyclic Load

29. Distillation of Issues from Matrices • 1. New ultra high strength steels - these require new constitutive models which must include forming and assembly effects (strain rate sensitivity and failure are also seen as especially important for these materials - see below) • 2. Strain rate sensitivity, for all materials (short duration transient loading) - although one reply suggested that Best Practice is already established here for all but high strength steels; perhaps a good topic for debate therefore? • 3. Composites (for all load cases) - material properties, models, failure/damage, connections and even effects of forming (e.g., need for draping simulation) all seen as areas for more work • 4. Failure / Fracture / Damage - most materials across all load cases; understanding the differences and how to best model them in CAE, choice of parameters, etc

30. Distillation of Issues from Matrices cont. • 5. Effects of Forming for all materials for short duration loading and non-metals for other cases - Forming Simulation as a topic in itself was eliminated at Munich although it is inherently required for calculating the forming effects; effect of assembly also highlighted, particularly for metals.  Perhaps this is one topic - Effects of Manufacturing? • 6. Best choice of constitutive models, for non-metals - complexity vs. ease of use & efficiency (CAE code specific?) • 7. Modelling connections - all materials and most load cases

31. Lisbon TW3, Breakout Sessions, November 23rd – 24th • Plastics (or more broadly polymers, composites, foams, elastomers …) • every company produces unique materials • vital to get relevant material tested • Use of elasto-plastic models (intended for steels) is not appropriate • Shorter history for plastics cf. steels • Automotive CAE most heavily used for crash • More and more cases • Legal and liability issues • Cost of testing prototypes • Interior and pedestrian cases are the greatest challenges • Performance of non-metals critical

32. Open Discussion (cont.) • Set up a standard database, push suppliers to provide data in specified format? • Material manufacturers don’t know what to measure; analysts need to define requirements • Tests are expensive, suppliers want to keep data proprietary • Even OEMs struggle to get data • Do we always know what we have and how it was derived? STEP, QA/QC, traceability, longevity (NS paper) • Data market • should data be free? • Lack of understanding of costs involved in generating data or appreciation of added value – cost-benefit • Plastics vs. Metals • Is case simpler for metals? – perhaps not because formulations are changing all the time – but generally agreement that polymer materials have more issues • Effect of mis-representation of a connection seen as more significant than issues with metal material data

33. Front-loading (integration) – materials aspects • Discussion on “lower quality” but quick vs. high quality but slow – quality of data and/or models • Concern here is how front loaded CAE’s requirements for material data and modelling affect the value of the analysis • If we know the effect of a parameter then we can compensate to some extent in concept phase but need more precise data for detailed work • Balance of these challenges with commercial pressure for early CAE, zero prototype? • Constrained to use existing materials with known properties • How precise does data need to be for early concept work?

34. Conclusions from Matrix + Discussion • Need to select topics for focussed attention: • Proposed that three issues stand out: • Material Characterisation for polymers, composites, foams • Modelling Connections • Fracture/Failure (all aspects, as well as for the above topics) • … and then effects of manufacturing • because of complexity, breadth; perhaps isn’t as significant in general CAE results accuracy (however it is probably essential for certain materials) • NB Fibre-reinforced plastics – forming process – but; complexity & cost of simulation • These are all important for crash but are also highlighted across the other load cases

35. Illustration of Best Practice vs. Breakthrough Technologies • Two illustrations of material characterisation for automotive CAE considering relative cost (in both material data requirements and CAE model) vs. confidence (or perhaps accuracy) in end result • Left graph shows a scenario where a low cost solution gives a relatively high amount of confidence and is hence good for upfront analysis • Right graph shows, even at high cost, confidence remains low and a breakthrough technology is required 1 1 BT 2 Cost Cost 2 100% 100% Confidence Confidence

36. Final Conclusions – Lisbon TW3 • Seek input from industry against this definition – how? • Questionnaire; little more to be gained? • Invited papers; e.g., along the lines of the Connections paper from Thomas Münz focussing on a specific topic and detailing all the key issues • Need to be quite specific on what we are asking for and who we are asking • What do you do now? • What would you like to do? • Opinion of leading CAE users, in OEM and their Tier 1 suppliers • Draft of a Paper Invitation with specific questions to be addressed to be prepared by rapporteur, circulated for comment before issue to all AUTOSIM correspondents for presentations for the next Workshop (Paris, June/July 2007)

37. Appendix – initial points for paper invitation for Paris TW4 • Proposal:- invite papers with targeted questions under a specific subject (as a test case to see if the method is valid): • Fibre-reinforced plastics under crash loading (could be exterior or interior) • How is the material chosen for the component • How is material data obtained (test, existing database) • How is it validated • QA/QC for data • Which CAE model is used (visco-plastic etc.) • What parameters are required • Is data fit to a Material Law or is it used directly (as data points) • What techniques are used to fit the data to the material model parameters (optimisation?) • Is failure / fracture / damage modelled; if so, how • How are connections modelled – strength, stiffness, failure • Is the manufacturing process considered • Cost (of material characterisation as well as CAE effort) • Accuracy (ideally report verification / validation results) • What are the limitations of the current method