Implementation of a Material Model for Adhesives in Abaqus
| dc.contributor.author | Aalto, Robin | |
| dc.contributor.author | Käll, Daniel | |
| dc.contributor.department | Chalmers tekniska högskola / Institutionen för industri- och materialvetenskap | sv |
| dc.contributor.examiner | Fagerström, Martin | |
| dc.contributor.supervisor | Shetty, Sandeep | |
| dc.date.accessioned | 2022-06-15T11:23:58Z | |
| dc.date.available | 2022-06-15T11:23:58Z | |
| dc.date.issued | 2022 | sv |
| dc.date.submitted | 2020 | |
| dc.description.abstract | In recent years it has become more common to use adhesive bonded multi-material structures in the automotive industry. These structures are however prone to distortions during the Electrocoat-oven process. These distortions are partly generated due to the curing of the adhesive, in combination with different coefficients of thermal expansion of the materials joined together (Δα). One way to predict these distortions is through FE-simulation. To be able to get accurate results from FE-simulations, a material model that accurately describes the degree of cure and its impact on the mechanical behaviour is required. Volvo Cars is therefore part of the research project MADBOND, in which a material model for adhesives has been developed and implemented in the FE-solver LS-DYNA. In this master’s thesis, the adhesive material model developed in the MADBOND project has been implemented as a user material model in the FE-solver Abaqus, which is the preferred solver for Δα-simulations at Volvo Cars. The Abaqus implementation was evaluated against both the existing implementation in LS-DYNA and a physical test. In the evaluation, the results from Abaqus were within the defined error tolerance of 10% for the residual stresses and residual deformations, when compared to LS-DYNA. In the comparison with the physical test, the same buckling behaviour was noted in the results from both the Abaqus simulation and the test. Significant differences in the magnitudes of the residual displacements was however seen, which in some regions was 10 times larger in the test case. The differences were to a large extent explained by adhesive failure in the test, which was an aspect not included in the simulation model. A convergence study was conducted for the material model, and it showed that the time increment size had a significant impact on the results, both in LS-DYNA and Abaqus. It was also discovered that the Abaqus implementation was not compatible with parallel execution. It was therefore recommended that these aspects are taken into consideration in any future work at Volvo Cars in order to perform accurate Δα-simulations for large structures. | sv |
| dc.identifier.coursecode | IMSX30 | sv |
| dc.identifier.uri | https://hdl.handle.net/20.500.12380/304711 | |
| dc.language.iso | eng | sv |
| dc.setspec.uppsok | Technology | |
| dc.subject | Multi-material structures | sv |
| dc.subject | Adhesive | sv |
| dc.subject | Δα | sv |
| dc.subject | UMAT | sv |
| dc.subject | Abaqus | sv |
| dc.subject | LS-DYNA | sv |
| dc.subject | CAE | sv |
| dc.subject | FEA | sv |
| dc.title | Implementation of a Material Model for Adhesives in Abaqus | sv |
| dc.type.degree | Examensarbete för masterexamen | sv |
| dc.type.uppsok | H | |
| local.programme | Applied mechanics (MPAME), MSc |