Resistance Weld Monitors Improve Process Consistency

Using Resistance Weld Monitoring to Improve Process Consistency

What do you do when your established welding process suddenly begins to yield weak or inconsistent welds that require time-consuming rework or cause expensive scrap down the line? In a prior blog, we addressed troubleshooting resistance welding problems, but in this post, we’re going to talk about identifying them and using data to prevent them. We’ll look at the basics of resistance weld process monitoring and how to use it to ensure quality welds. We will also touch briefly on the brave new world of applying artificial intelligence (AI) and machine learning (ML) to resistance welding processes.

Process monitoring provides critical manufacturing data

Resistance weld monitors can measure many parameters including current during the weld, the voltage between electrodes, electrode force, and electrode movement. Once dialed in on a successful process, variations in these parameters can tell a lot about the welding conditions. A change in any of them could indicate issues like improper part or electrode alignment, dirt/debris, change in material, inconsistent plating, and other defects.

Why monitor the process?

Why? Simple: the data collected by weld monitors tells a story that even the most well-trained eyes and ears of an experienced operator cannot. Even when it is obvious that a weld is not working, the operator may not be able to pinpoint the root of the problem, whereas the data from a monitor provides that type of insight.

Know which model suits your purposes

There are two types of monitors: basic and advanced. Basic weld monitors output a numeric aggregated value (min or max) for one or more parameters. For simple analysis and flagging suspected manufacturing defects, these monitors are sufficient. Advanced monitors, on the other hand can capture and analyze the entire, high-resolution waveform of every parameter, which provides a deeper insight into the dynamics of the weld process. In addition, the pass / fail signal can be further refined by looking at different regions of interest, rather than a simple min/max value.

Do your due diligence to determine your range of acceptable values or limits

It’s important to note that a monitor will not definitively say whether a weld is bad or good. It can, however, compare it to a known good weld. If the parameters are similar, it will be passed as “good”; if outside the range, it will be flagged as “different.” Different is “bad” in a production setting and the part would be tagged for rework or scrapped.

To determine the range of acceptable values or limits, users should run a design of experiments (DoE) using the different weld equipment settings that are known to affect weld quality. The experiments should be recorded and then reviewed with weld quality results. At the end of the study, set upper and lower parameter limits so the range includes “good” welds and excludes “bad” welds. The optimal limits settings reflect a balance between the successes at accepting a “good” weld and avoiding too many “false positives.”

Sounds time-consuming, right? It certainly can be. But it will generate robust process settings that will accurately identify a good weld within statistical limits. This is the de facto procedure to follow for such safety-critical products as automotive or aerospace components or medical devices.

Use monitors to get data for use in other processes

In addition to quality assurance, weld monitors are used to gather data for process development, maintenance troubleshooting, equipment certification, and data storage.

Data can be stored locally in the process monitor or on a networked server. The waveform data stored on the server in a database can be imported into SPC, equipment health, or OEE software programs for further analysis. The output from these programs can be used to make manufacturing and business decisions.

The future in welding: artificial intelligence and machine learning

AI and ML algorithms are starting to be applied in the welding field to create smart adaptive limits and to predict weld quality. The AI/ML algorithms look for trends in the recorded data that aren’t revealed in standard analysis of graphs and extract important data features that affect the welding quality. This goes beyond the use of simple limits by today’s monitors and can set “smart adaptive limits” that update over time, further improving throughput and output product quality.

In short, AI/ML may be able to achieve the holy grail of process monitoring and finally answer the question, “Was that last weld good or bad?” and offer a path to weld quality prediction.

See AMADA WELD TECH’s full range of resistance weld monitors and checkers.