Licentiate thesis, 2019

Disturbances in the manufacturing and assembly processes cause deviation and geometrical variation from the ideal geometry. This variation eventually results in functional and aesthetic problems in the final product. Being able to control the disturbances is the desire of the manufacturing industry. This, in other words, means turning the noise factors to control factors, in a robust design perspective.

With the recent breakthroughs in the technology, the new digitalization reform, and availability of big data from the manufacturing processes, the concepts of digital twins have grasped the attention of the researchers and the practitioners.
In line with this trend, Söderberg et al. have introduced the geometry assurance digital twin and the concept of the self-compensating individualized assembly line. Steering the assembly process with online real-time optimization, through the digital twin medium is the vision of such a concept.

Joining sequences impact the final geometrical outcome in an assembly considerably. To optimize the sequence for the optimal geometrical outcome is both experimentally and computationally expensive. In the simulation-based approaches, several sequences need to be evaluated together with the finite element method and Monte Carlo simulations.

In this thesis, the simulation-based joining sequence optimization, using compliant variation simulation is studied. Initially, the limitations of the formulations and the applied algorithms in the literature have been addressed. Two evolutionary algorithms have been introduced to compare the computational performances to the genetic algorithm. Secondly, a reduced formulation of the sequence optimization is introduced through the identification of the critical points to lock the geometry, geometry joints. A rule-based method has been proposed to initiate the evolutionary algorithm and thereby to increase the algorithm’s computational efficiency. This approach has been further improved by a contact displacement minimization approach to generate model-dependent rules. Finally, a surrogate-assisted approach has been introduced to parallelize the computation process, saving computation time drastically. The approach also unveiled the potential of the simulation-based geometry joint identification, simultaneous to complete sequence determination.

The results achieved from the presented studies indicate that the simulation-based real-time optimization of the joining sequences is achievable through a parallelized search algorithm, to be implemented in the geometry assurance digital twin concept. The results can help to control the joining sequence in the assembly process, improving the geometrical quality in a cost-effective manner, and saving significant computational time.

Compliant Variation Simulation




Virtual Development Laboratory (VDL), Hörsalsvägen 7A
Opponent: Dr. Alexander Govik, Volvo Car Group, Sweden


Roham Sadeghi Tabar

Chalmers, Industrial and Materials Science, Product Development

A method for identification and sequence optimisation of geometry spot welds in a digital twin context

Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science,; Vol. 223(2019)p. 5610-5621

Journal article

A Novel Rule-Based Method For Individualized Spot Welding Sequence Optimization With Respect to Geometrical Quality

Journal of Manufacturing Science and Engineering, Transactions of the ASME,; Vol. 141(2019)

Journal article

Tabar R.S., Wärmefjord K., Söderberg R., Lindkvist L., Efficient spot welding sequence optimization in a geometry assurance digital twin

Tabar R.S., Wärmefjord K., Söderberg R., A new surrogate model based method for individualized spot welding sequence optimization with respect to geometrical quality

Smart Assembly 4.0

Swedish Foundation for Strategic Research (SSF) (RIT15-0025), 2016-05-01 -- 2021-06-30.

Subject Categories

Production Engineering, Human Work Science and Ergonomics

Control Engineering

Computer Science

Areas of Advance




Virtual Development Laboratory (VDL), Hörsalsvägen 7A

Opponent: Dr. Alexander Govik, Volvo Car Group, Sweden

More information

Latest update