3D EM Simulation of a resistance spot welding gun

This article is concerned with the 3D electromagnetic modelling and equivalent circuit parameter extraction of a Resistance Spot Welding Gun (RSWG) . The parameters for the gun are not as easily established for the gun as for other RSWG system components.  The current transformer and electronic converter electrical parameters are either known or can be obtained by standard test procesdures. This is not the case for the gun where the electrical parameters are unknown and difficult to measure. A 3D EM simulation using CST EM STUDIO® (CST EMS) can be used to establish these parameters and complete the system characterisation. The static current and magnetoquasistatic solvers were used for the simulations.

This article is a summary of the work of A. Canova, B. Vusini, Dipartimento di Ingegneria Elettrica-Politecnico di Torino and G. Gruosso, Dipartimento di Elettronica e Informazione-Politecnico di Milano [1]

The CST EMS model of the welding gun [Figure 1] was imported via the SAT CAD format and consists of terminal connections, the gun arms and the electrodes. This model forms the basis for the ensuing static current and magnetoquasistatic simulations reported later in this article.

The behavior of an RSWG depends on whether it is supplied by an alternating current (AC) or a medium frequency direct current (MFDC).

In an AC welding system the supply is usually a single phase voltage at 50 Hz. The system [Figure 2] consists of an AC supply system connected to a step-down current transformer through a couple of controlled diodes (SCR). These diodes allow the regulation of the rms current value and so of the welding power through their duty cycle variation.

Similarly, for the MFDC case [Figure 2], the primary current is regulated by a three phase inverter and the secondary current is rectified by a diode bridge.

The current transformer and electronic converter electrical parameters are either known or can be obtained by standard test procedures. This is not the case for the gun where the electrical parameters are unknown and difficult to measure. A critical point of both technologies is the electrical efficiency, assumed as the ratio between the output RSW power and the input power supplied to the transformer, which is about 20-30%. Responsible for the power losses are the ohmic parts of each component: transformer, diodes, RSW parts in the AC technology plus the inverter losses in the MFDC one. It should be noted that in the AC case inductance effects lead to a power factor reduction, to an increase of the apparent power and of the current which correspond to and increase of the ohmic losses. The improvement of the global efficiency initially requires an analysis in order to estimate the ohmic losses due to the different parts. In this paper the attention has been devoted to the resistance evaluation of the RSW parts (connections to the transformer, arms and electrodes) which are usually unknown, while the other component: transformer, inverter and diode bridge, are considered commercial components and are usually evaluated by standard measurements. From preliminary estimation the loss contribution of the gun parts can be up to 50% of the RSW output power.

In order to obtain the electrical resistance and inductance of the different parts constituting the RSW a set of stationary 3D current solver simulations is carried our where each gun component is analyzed as a massive conductor supplied by a known current injected under real working conditions. When the electrical parameters are identified it is possible to evaluate the energising performance of the gun. The study of RSW is then completed by a 3D low frequency solver of the whole system which allows the estimation of the external magnetic field pollution and the thermal behaviour.

The welding gun electrical parameter estimation, devoted to the MFDC modeling, requires the solution of a set of static current simulations where each part is analyzed separately. The static current simulation requires the identification of two terminals for the current injection. The model of each gun part takes into account the massive conductivity of the material while the terminals are ideal conductors ( Perfect Electric Conductor ( PEC ) ) with infinite conductivity.

An example of the simulation performed for the flexible part of the connections is shown [Figure 3]. The simulation can be also performed on a set of parts as for example the upper or the lower ones or the total RSW structure. In this simulation the contact between the different parts is considered ideal (contact resistance negligible) but the current distribution takes into account the real shape of the interface between the different parts.

In [Figure 4] the potential distribution of the whole system when the two electrodes are put in short circuit is shown. The electrical parameters obtained with this simulation have to be equal to the sum of the singular values of the different parts. The calculation of the electric resistance of each part leads to the results shown in the table in [Figure 5]. The total value is also given in the table.

The magnetoquasistatic solver allows the evaluation of the RSWG resistance and inductance under AC supply conditions. In this section the analysis have been conducted on the whole RSWG but a similar methodology as proposed in the previous section can be carried on. Using the magnetoquasistatic solver the supply frequency and current amplitude can be chosen. The magnetic behavior of the RSWG can be considered linear and so the evaluation of the electrical parameters depends only from the supply frequency. The simulation under very low frequency (e.g. 1Hz) allows a verification of the previous calculations provided by the static current formulation when the skin effect is negligible.

The 3D Eddy current (log scale) and 2D plane H-Field distributions at 50Hz are shown in [Figure 6]. The H-field magnitude gives an indication of the field pollution created by the gun.

The magnetoquasistatic simulation provide the values reported in [Figure 7]. The simulation at 1 Hz allows to verify that the resistance are practically equal to the previous static current simulation. The simulation at 50 Hz shows that a significant eddy current effect occurs and the resistance increases.

The calculated and measured resistance values obtained at DC are shown in [Figure 8] for the most critical component, the flexible connection.

The impedance was evaluated by an AC test. As a result of the very low resistance values, active and reactive power measurements, necessary for the determination of the active and reactive parts of the impedance, were not possible. The impedance at 20 Hz and 300 Hz is shown in [Figure 9].

This article has served to demonstrate the application of CST EMS to the evaluation of equivalent circuit parameters for a spot welding gun. The equivalent circuit allows the system efficiency to be obtained. As parts of CST's Complete Technology approach, CST EMS allows the construction of a single model which, with minor modifications, can be solved using the stationary current and magnetoquasistatic (low frequency) solvers.

[1] "Electromagnetic Modelling of Resistance Spot Welding System" , A. Canova, G Gruosso, B Vuisini, ISEF 2007 - XIII International Symposium on Electromagnetic Fields in Mechatronics, Electrical and Electronic Engineering, Prague, Czech Republic September 13-15,2008