Affordable Access

Numerical modelling of parts distortion and beam supports breakage during selective laser melting (SLM) additive manufacturing

Authors
  • Bresson, Yves
  • Tongne, Amèvi
  • Selva, Pierre
  • Arnaud, Lionel
Publication Date
Jan 16, 2022
Source
Open Archive Toulouse Archive Ouverte
Keywords
Language
English
License
Green
External links

Abstract

The Selective Laser Melting (SLM) process has been progressively endorsed as an industrial manufacturing technique to produce high value-added components. However, one of the main obstacles to its wide application is the uncertainty regard- ing the successful completion of the manufacturing process. Mechanical stresses are generated and accumulated during the process, which may lead to the parts warping and cracking. Support structures may also detach from the part but it is not certain that these cracks conduct to the manufacturing failure. The process simulations currently available do not consider the cracking of the supports and the ongoing part’s deflection. The aim of this study is to investigate cone supports fracture behaviour comparing the results from a numerical model and the manufacturing of an industrial part. A model using 1D-beam elements to mesh the supports has been developed to consider the damage of the supports, their breakage and the ongoing deflection. Some numerical convergence issues are identified and solutions are proposed. Specific experimental set-ups are developed to characterise the behaviour of the supports individually and as a group. Significant improvements are denoted while injecting the measured characteristics within the model. Some key parameters of the supports damage behaviour are identified. It is shown that the supports mechanical characteristics are significantly different from the parts due to their manufacturing conditions and environment. Also, limitations regarding the characterisation of the supports as well as strong numerical convergence issues brought by multiple supports cracking are discussed.

Report this publication

Statistics

Seen <100 times