© The developed test measurement system and hardware – a control module for acoustic detection of welding defects

Finding Defects Acoustically

A Steinbeis test measurement system can detect welding defects using acoustic emission analysis

Welding is applied as a production technique in many areas of industry. The advantages offered by the method are often tarnished, however, by one major disadvantage: Quality testing is complex, because welding defects are so difficult to detect. This even applies to gas tungsten arc welding (GTAW), a process in widespread use in industry and a method that enjoys particular popularity because of the quality of the weld seam. This type of welding is typically used when speed is not as important as quality requirements – for example when making pipelines, or equipment for power stations or the chemicals industry. Conducting quality checks on weld seams can be expensive when they are carried out after production, so to find a way to avoid high outlays the Steinbeis Innovation Center for Intelligent Functional Materials, Welding and Joining Techniques, Implementation (Dresden) has been working with the welding macfhine producer Merkle as part of a sponsored research project. Together, the experts have developed a testing measurement system capable of preventing weld defects in the first place by carrying out quality inspections during the actual welding process.

Monitoring welding processes is not only advantageous when it comes to avoiding welding defects. It also improves the quality of seams, raises productivity, reduces production costs, and enhances automation levels. At the same time, monitoring is an important instrument of professional quality management – a crucial activity in most sectors of industry.

Key problem: Correcting errors is expensive

Until now, checking for welding defects has been based on the DIN standard ISO 5817. This standard is applied to determine whether it will be possible or even necessary to correct a defect after the welding process. Making such decisions has been an extremely expensive part of quality inspections until now. There are a number of non-destructive and destructive testing options for conducting checks. The non-destructive methods range from straightforward visual inspections to dye penetrant inspection, magnetic particle examination, X-ray testing, ultrasonic testing, and eddy current testing. The destructive examination procedures include metallographic tests, strength testing, and notch impact bending tests.

Key solution: Avoid errors during the production process!

All of these testing methods have something in common: They are conducted downstream after production, so they do not actually prevent errors, they (at best) rectify them and this is extremely costly. This is where the research alliance comes in between the Dresdner-based Steinbeis Innovation Center Intelligent Functional Materials, Welding and Joining Techniques, Implementation and its partner in industry, Merkle Schweißanlagen-Technik GmbH. The test measurement system they have developed controls and monitors the welding process in line, thus avoiding potential welding defects and safeguarding the required seam quality and welding properties. Their innovative system detects welding errors by spotting correlations between acoustic emissions and process phenomena.

“We’ve developed a measurement concept with our project partner Merkle and a prototype for a test measurement system. This has made it possible for us to detect and assess seam weld defects like pores, incomplete joint penetration, internal shrinkage, cracks, and other faults affected by the process parameters and the materials used,” explains Associate Professor Dr.-Ing. habil. Khaled Alaluss, who is co-director of the Steinbeis Innovation Center alongside Dr. Lars Kulke and Prof. Dr.-Ing. Gunnar Bürkner. The measurement modules on the prototype provide process data on the defined seam lengths and volumes in more detail than was previously possible. This has now made it possible for the project team to examine and analyze different factors affecting the welding process.

Putting the system to use

The measurements provided the team with important confirmation that the new system makes it possible to detect welding seams using non-contact measurement methods and match the resulting signatures to detected weld defects. The airborne sound detector – a microphone – works from 0.005 to 200 kHz and the structure-borne sensor from 0.1 to 500 kHz. The current state of technology will only work to a maximum of 20 kHz. Irrespective of sensor positions, the welding experts now have a method for detecting sound signals during processes with measurement times of up to 30 seconds per measurement. Airborne and structure-borne sound detectors can be mounted in any kind of position on welding equipment and welding samples. This makes it possible to determine and identify welding defects with greater accuracy and sensitivity.

Now that the project has been completed it is becoming obvious just how much value the new test measurement system is delivering. The system can already evaluate airborne and structure-borne sound signals during the welding process, making it much easier to detect, analyze, and assess existing welding defects. This has resulted in a new measurement process for use in inline welding process monitoring, one that has also been evaluated in terms of process technology. The project team has built on the modular system to come up with an innovative control technology for the GTAW process, and this allows the process to be monitored for possible welding defects. Extensive examination of welding technology has shown that the desired functionality of the prototype has clearly been achieved, also thanks to the control module for detecting weld defects using airborne and structure-borne noise sensors. Detecting and identifying welding defects using the detected acoustic emissions worked without any problems with gas tungsten arc welding. The monitoring system is also capable of using detected acoustic emissions to determine the specific nature and location of welding defects. The developed error prevention strategies were used afterwards to determine and show the impact that certain process parameters (welding current, welding speed, filler metal, shielding gas flow, surface texture, contamination, grease, swarf, etc.) have on welding defects and quality.

Once a correlation was worked out between acoustic emissions and process parameters, the team also developed the control strategies that will be required, in a format that will be suitable for the market, and this was also evaluated in terms of process technology. How suitable the monitoring system prototype is in actual practice was then tested by Merkle by putting the solution through its paces. The company used the system successfully with its own control module on a GTAW machine. It had no problems detecting weld defects by looking at the correlations between acoustic emissions and process phenomena.

Contact

Associate Professor Dr.-Ing. habil. Khaled Alaluss, Hayder Al-Mashhadani, Prof. Dr.-Ing. Gunnar Bürkner, Dr. Lars Kulke
Steinbeis Innovation Center Intelligent Functional Materials, Welding and Joining Techniques, Implementation (Dresden)

Wilhelm Merkle, Dr.-Ing. Ulrich Prank
Merkle Schweißanlagen-Technik GmbH (Kötz)