Burst test analysis of metal sandwich wall panels
Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
In this report, an investigation of some of the most common damage models available in literature is performed. Their ability to predict ductile failure of metals will be compared and a specific focus is put on burst failure prediction of the cooling channels inside a sandwich wall structure. This structure is intended to be used at future versions of the Ariane rocket nozzles developed at GKN-Aerospace in Trollhättan, Sweden. Currently GKN-Aerospace uses a prediction method of burst failure based on the Rice and Tracey damage model but it has been shown to sometimes deviate from experimental results, why they seek for a new method to predict ductile failure. Hence the objective of this project is to find a more accurate method to predict burst failure of the sandwich wall than the one used today at GKN-Aerospace. This will be done by both a literature research and performing simulations with the studied damage models. The project was demarcated to investigate four damage models; the Gurson-Tvergaard-Needleman, Johnson-Cook, the Extended Mohr Coulomb criterion and the currently used damage model at GKN-Aerospace. Further the simulations were delimited to the softwares ANSYS APDL and LS-DYNA, for which burst test simulations was only performed in ANSYS APDL. It was found that only the Gurson-Tvergaard-Needleman and the Rice and Tracey based model could be used for prediction of burst failure in ANSYS APDL, why only those damage models could be evaluated against the objective of the project. The burst test simulation results showed that the Rice and Tracey based model gave predictions of burst pressure closest to experimental results, why the objective of the project was not met since a more accurate prediction method was not found. However important conclusions could be drawn and valuable recommendations for future work could be given.
Place, publisher, year, edition, pages
2015. , 113 p.
TRITA-AVE, ISSN 1651-7660 ; 2015:42
Engineering and Technology
IdentifiersURN: urn:nbn:se:kth:diva-180456OAI: oai:DiVA.org:kth-180456DiVA: diva2:894052
Subject / course