PAVE: Probabilistic analysis and verification process for composite structures
Reference number | |
Coordinator | Swerea SICOMP AB - Swerea SICOMP AB, Piteå |
Funding from Vinnova | SEK 1 870 800 |
Project duration | June 2015 - May 2018 |
Status | Completed |
Venture | The strategic innovation programme SIP LIGHTer |
Purpose and goal
Composite materials are created at the same time as the structure: ´the material does not exist before the structure is manufactured´. Design data must therefore take into account the chosen manufacturing process and how it affects the performance of a particular component. On PAVE project, this made industrial and implementation of invariant and probabilistic methods developed by SICOMP, KTH, CAEValue, helped develop a cost-effective and time-efficient method that allows a weight loss of more than 20% and reduced material testing requirements at OEMs (VCC, IKEA, Gestamp, SCANIA)
Expected results and effects
The probabilistic method is expected to be implemented at industry (VCC, Gestamp, IKEA, SCANIA) in the design of composite components. The method is then expected to contribute to robust designs and to lighter products. The results from the experimental series show that the invariant method is an appropriate methodology for simply predicting mechanical properties of any composite laminate. As a basis, only a few tests on coupon level are required. However, the method needs some further development before it can be used in industry in a timely manner.
Planned approach and implementation
Implementation of methods on the demonstrators at VCC, IKEA, SCANIA was successful as both methods showed good results and all project goals could be met. The implementation of the probabilistic method could have been made deeper if material parameters could be varied based on actual measured values. Invariant methodology has been experimentally evaluated, both at coupon level and with two physical component-level demonstrators. The results of the experimental tests by the demonstrators, by Exova and Marström, were evaluated against calculations with good results.