Dynamic resistance in spot welding: why there is no universal OK curve
An analysis of more than 1.8 million spot welds shows how material, stack thickness and electrode tip-face diameter change process-signal morphology.
A signal must be linked to a specific weld point
Dynamic resistance is one of the most useful resistance-spot-welding signals. Its evolution reflects changing contact resistance, heating, coating breakdown, melting and nugget growth. This does not mean that one correct curve shape exists for every stack-up.
A new industrial study analysed more than 1.8 million welds and derived 15,096 representative signal profiles. The dataset included different steels, coatings, thicknesses, two- to four-sheet stacks and different welding guns.
Main observations
- higher alloying content was associated with faster local heating and earlier resistance peaks,
- the secondary peak was often less pronounced in press-hardened steels,
- greater total thickness increased resistance, electrode-displacement and force amplitudes,
- reducing tip-face diameter from 8 to 6 mm increased the initial resistance peak by about 9%,
- at constant thickness, the number of internal faying surfaces was not a significant predictor of macroscopic resistance.
A smaller electrode face changes both current density and contact pressure. A change from 8 to 6 mm is therefore not merely a maintenance-related geometric adjustment; it can alter the complete reference signature of the process.
What the result does not prove
The absence of a pronounced secondary peak does not automatically indicate a nonconforming weld. Conversely, a curve inside a monitoring envelope does not by itself prove the required nugget diameter. The monitoring system must be validated for the specific stack-up and linked to destructive or other qualification testing.
Source
- Schulz, H.; Stender, M.: Causal discovery of signal morphology drivers in steel resistance spot welding, *Welding in the World*, 16 October 2026.
This article is an original interpretation of the reported results. No figures, tables or passages from the paper have been reproduced.