Roham Sadeghi Tabar, Kristina Wärmefjord, Rikard Söderberg, Lars Lindkvist
J. Mech. Des. Oct 2020, 142(10): 102001 (8 pages)
Paper No: MD-19-1377 https://doi.org/10.1115/1.4046436
J. Mech. Des. Oct 2020, 142(10): 102001 (8 pages)
Paper No: MD-19-1377 https://doi.org/10.1115/1.4046436
The availability of big data has made the role of digital twins in manufacturing more prominent. This paper introduces a geometry assurance digital twin – created from the scanned data of individual components – to define and improve an assembly’s geometrical quality. The joining sequence in a sheet metal assembly impacts geometrical quality and determining the optimal joining sequence is computationally expensive. Meta-heuristic optimization techniques like genetic algorithms often require many simulations, which can increase computational cost. This work improves the optimization process by combining a model-based heuristic algorithm – based on contact displacement minimization – with the meta-heuristic algorithm. Contact modeling avoids part penetration in adjacent areas, and the joining sequences that provide minimal penetration states are used to populate the initial solution for the meta-heuristic algorithm. This approach is demonstrated on two sheet metal assemblies and a reduction in sequence time of 60-80% is achieved. By using a digital twin, optimal joining solutions can be achieved with greater efficiency.
For the full article visit ASME’s Digital Collection.