The challenges in process control of resistance spot welding

The challenges of spot welding, with consistently reproducible and high-quality welding results in modern body construction as well as in the industrial production of metalworking, arise from the following factors:

• The use of new materials (hardened, high-strength, coated, several and different thick sheets).
• The different state of the materials on site at a time (possibly partial contacting by structural adhesive, paint residues, soiling, general oxidation, different surface properties).
• The control of the resistance welding process requires, due to the non-constant process impedance and in part stochastic process history (see above), a perfect impedance measurement, which is, however, dealing with different problems during the running welding process. These problems can be traced back to the required voltage measurement in the immediate vicinity of the process at the location of the spot weld.
• A process-based control of spot welding therefore requires a "virtual" measurement of the voltage to reconstruct the original waveform. For this ELMA-Tech has developed a corresponding reconstruction algorithm, which allows a technically perfect measurement of the process impedance.

In combination with the use of virtual machine and virtual measuring technology, a very fast reacting process control is made possible.

The general process model of the virtual machine for spot welding

It is possible to roughly distinguish four phases of a completely automatic spot welding, i.e. the setting of a single welding spot:

• A targeted energy input tests the initial state of the materials to be spotted. A distinction between total insulation, partial insulation and ideal initial conditions takes place.
• Second, a first formation of the welding nugget takes place by introducing a targeted energy input. In the most frequently encountered case of partial insulation, the introduction of a targeted amount of energy brings about an approximation of the sheets to be welded in the joining area.
• The actual welding process is then carried out using the classic current-time program. In doing so, the amount of energy introduced is continuously added up in order to end the welding process, when a desired energy is reached.
• Finally, there is a determination of design-related shunts or shunts resulting from the joining process itself. The energy loss of the welding nugget formation e.g. through shunts is now being added.

The Plug & Weld principle ensures that this process runs completely automatically. Incidentally, this also includes the automatic measurement of the sheet thicknesses to be welded and the recognition of the types of material of the used sheets. For these reasons, it is not necessary for the user to enter welding parameters before a welding process. The aspects just described are implemented, for example, in our mobile spot welder MIDIspot VISION.

The extended process sequence for robot-guided spot welding

In this process, the virtual machine takes over other tasks of process control, which are in other concepts usually performed by the robot controller:

• Closing the gun arms with automatic balancing (float mode)
• Force generation and detecting the overall sheet thickness
• Calculation of the spot energy to be inserted
• Start of the welding process with phase-dependent generator characteristic
• Determining the quality of steel (normal / high-strength)
• Terminating the process with the determined point energy
• Qualification of the process
• Opening of the gun (7th axis function), disabling the balance of the gun
• Output of the FK switch contact to the roboter

A more technical description of the fully automatic spot welding system for application in automotive industry download here.

Back