Multidisciplinary Optimisation of Geometric Variation and Dynamic Behaviour for Squeak & Rattle

Aswin Dhananjai Krishnaswamy & Chidambaram Sathappan present their master thesis project. On-line presentation.

Program: Product Development
Examiner: Lars Lindkvist, IMS
Supervisor: Mohsen Bayani, VCC
Opponent: Koushik Ravichandran

Squeak & Rattle (S&R) are undesired sounds that are caused when two components come in contact with each other. They give the user a feeling of a possibility of a quality issue. In modern premium cars, user comfort is of utmost importance and these undesired sounds must be eliminated. There are various reasons that causes the S&R of which geometric variations and dynamic behaviour are two major contributors. The variation and deviation in the assembly leads to undesired gap between the components making them to come in contact with each other. This also influences the dynamic behaviour of the component leading to resonance. However, varying the location of the attachment points could impact the cause of S&R and requires an optimisation. 
A benchmarking study was done with various cars in different segments from different manufacturers. Assemblies that are prone to have high squeak & rattle was identified and a simpler version of the assembly was modelled. The simplified models were categorized into different model configurations and the optimisation was carried out. A modified ISF program was written to generate DOE based on certain requirements of the work. The program was created by including all the constraints making the DOE design generation faster. A two-stage genetic algorithm-based optimisation approach was used to identify the right attachment scheme between components. The findings were documented, and the methodology was also extended for some industrial cases with models from real industrial applications like Instrument Panel assembly and Side door assembly.
​Link to presentation

Category Student project presentation
Location: On-line presentation
Starts: 17 September, 2020, 10:00
Ends: 17 September, 2020, 11:00

Published: Wed 09 Sep 2020.