Automakers and parts suppliers typically inspect welds using destructive processes. A sample of parts is pulled from the line, taken apart and examined. Not every weld is inspected, and companies waste money on the sampled parts.
Oak Ridge National Laboratory (ORNL) has developed a new portable, infrared weld inspection system that can be used on the production line to inspect welds as they are completed. The system can also be used in post-production for inspections, and could potentially be implemented in collision shops.
“The idea is to measure every weld, or at least every critical weld,” said ORNL’s Zhili Feng, one of the technology’s developers. “It gives automakers an efficient method to immediately send feedback to the production lines.”
ORNL has licensed the technology to a Tennessee company, APLAIR Manufacturing Systems, which is already in talks with OEMs to refine the system and deploy it on the production line. APLAIR and ORNL staff plan to collaborate to improve and validate the technology under a proposed cooperative research and development agreement, including tests on an industrial-scale assembly line. The company hopes to have a commercial version of the system available within two years.
"Right now, parts are typically inspected either using a sonic device or destructively," says Robert Watts, CEO at APLAIR. "The sonic method takes extra labor and some degree of skill on the part of the operator to get reliable results. With destructive testing, you are essentially taking a good part and making it a bad part to prove it's a good part, and paying extra money to have it done. We've had one Tier 1 supplier tell us that eliminating destructive testing at a high-volume plant would save them upwards of a high six-figures in parts lost to the testing process."
The system was originally developed via U.S. Department of Energy Vehicle Technologies Office project. It uses an infrared imaging system and quality analysis software to analyze welds within a second. The heat created during the welding process is used to examine the weld using an infrared camera. The heat signature indicates whether or not the weld was completed properly. In post production, the system can be used with an induction heater to raise the temperature of a finished weld to 122 degrees F before imaging the weld for analysis.
"One of the purposes of the project was to find a way to analyze welds for new high-strength steels," says Jian Chen, ORNL researcher. "Manually opening the welds to check them is very difficult and time consuming. The OEMs need a more effective way to know the quality of the weld."
According to Chen, the solution is relatively easy to deploy. The infrared camera simply needs to be positioned to take images during the welding process. In the post-production scenario, users only need to add the small induction heater.
In collision shops, the system could be used to ensure the quality of new welds or to evaluate existing welds on a damaged vehicle. "We have had some interest in the collision repair industry, and you could use it either after the weld is done or instantly as the weld is being done in the shop," Watts says.
"The environment of the repair shop is actually very similar to what we have developed in our lab," Chen adds. "I don't think it would be a problem to use in that situation."
Watts says the system should be cost effective for deployment in a collision shop. "This can really be a liability mitigator," Watts says. "Right now, shops don't know for sure that every weld is a good weld. You can tell if the machine the welder used did what it was supposed to do during the process, but you don't really know if the weld is good. That's what this system does."
Presently, the analysis system can be used with a variety of steel welds, and ORNL plans to expand the heat signature database so it can be used with other materials. "We have tested this with a lot of high-strength steels and some aluminum welds," Chen says. "In theory it should work with other types of materials as long as you are using spot welding techniques."
Chen says the lab has begun testing other welding techniques, including different material combinations (aluminum to steel, carbon fiber, and reinforced steel).
According to Watts, APLAIR and ORNL are working to reduce the time it takes to achieve an analysis. "right now, depending on the type of weld and the metal, we can have an answer in 1 to 1.5 seconds. That seems fast until you get into a high-speed production line, where you have fractions of a second t complete the analysis."
APLAIR is rewriting the software so that it will be more useful in a production environment and more user friendly, as well as improving the cycle time. APLAIR and ORNL also hope to expand the system to analyze projection welds.
Watts expects a collision industry version of the solution to be ready first, because body shops don't require sub-second response times. "We hope to have something ready to go by the end of the year for that market," Watts says. "Right now we're looking for a partner to help move this forward on the collision repair side."
