Resistance spot welding is a widely used joining process in the automotive indus-
try. The forming of the joint is based on two main physical principles. Firstly,
the joining parts are pressed together during the whole process by a mechanical
force. Secondly, electrical energy is applied through electrodes and is converted
to thermal energy by joule heating. For the simulation of the welding process
it is necessary to consider thermal, electrical and mechanical fields and their
interaction.
The physical quantities which can be determined in welding experiments are
strongly restricted. This is due to the short process times and covered up re-
gions of interest. Only a view quantities can be measured during the welding
process and offside the region, where the welding nugget is formed, e.g. voltage
drops, movements of the electrode and local temperatures at the electrode or
joining parts. Only after the welding internal quantities like nugget shape can
be determined.
Using these few quantities available for validation is not sufficient for ensur-
ing reliance of a complete multiphysical process simulation. This is due to the
strong interaction of the multiple physical domains. FEM inherent errors might
superpose. The consequence is, that the validation procedure has to be assisted
by a detailed verification process. This accounts especially, if the simulation is
used to investigate quantities, which can not be validated through direct mea-
surement in the experiment.
This contribution shows how different sources of errors may lead to significant
errors in simulation results if they are not reviewed by both validation and ver-
ification. In the worst case the errors are so big, that a use of the simulation
model is highly restricted because of not trustworthy results. Main sources of
error are: numerical solving process, variations in model input and modelling
approach. In order improve and ensure model accuracy this contribution shows,
quantifies and discusses the mentioned errors in resistance spot welding simu-
lations, for the first time. As a novelty, a strategy of a structured verification
process for FEM simulations of resistance spot welding is developed and pre-
sented. Therefore, this presentation aims for displaying an method of quality
assurance in a field of high practical interest and high requirements in numerical
simulation.