A novel quality evaluation method for resistance spot welding based on the electrode displacement signal and the Chernoff faces technique

• A novel quality evaluation method for resistance spot welding is discussed.
• Some statistical features are extracted from the electrode displacement curve.
• The multi-dimensional extracted features are used to sketch Chernoff faces.
• The face reflects the weld quality through different facial expressions.
• The visual feature makes quality assessment for weld easily understood.

To develop a visual and reliable weld quality assessment method for resistant spot welding, the electrode displacement signal was measured and analyzed. Some statistical features closely related to the weld quality were extracted. The multi-dimensional extracted features from the signal were sketched as Chernoff face which characterizes the welding quality through the different facial expressions. The weld quality was summarized as four levels, namely poor, good, excellent and expulsion, and each level corresponded to one kind of facial expression. By visual judgment, good or bad welds were identified accurately and rapidly even though the weld was made under abnormal welding conditions. In particular, the visual characteristic makes the diagnostic procedure for welding quality easily understood and interpreted.