New York-based database management company Senvol recently demonstrated a machine learning approach to material tolerance development. This approach has been shown to be more flexible, more cost-effective, and more time-efficient than traditional approaches (in this case metal material characterization). Standardization (MMPDS).
This work was awarded to the US government contract W911NF-20-9-0009 to enable Senvol to apply machine learning software Senvol ML to enable a path to rapidly develop acceptable material properties for additive manufacturing (AM). was done as part of
Partners in Senvol’s program included EWI and Pilgrim Consulting. Battelle and Lockheed Martin Fellow Hector Sandoval also served as technical advisors to the program. This contract was administered by the National Center for Manufacturing Science (NCMS) through the AMMP Other Trade Agreement (OTA) program.
The machine learning approach is very flexible and adaptable to any changes to the AM process, making this approach ideal for long-term sustainability. This program focused on demonstrating the approach using 17-4 PH stainless steel material processed in a powder bed fusion AM machine.
“Developing material tolerances is a very expensive and time-consuming undertaking,” said Annie Wang, president of Senvol. “Senvol’s program has been very successful in demonstrating a new approach to additive manufacturing tolerance development leveraging machine learning. I look forward to your research.”
The high cost of tolerance development is primarily due to the fact that material tolerance development requires the generation of vast amounts of empirical data at fixed processing points. This usually means that all empirical data must be regenerated from scratch each time. When there is a big change in the process. As a result, AM processes are not only costly and time consuming to initially implement, but also costly and time consuming to maintain in the long term if changes to the AM process are unavoidable.
Hector Sandoval, who reviewed the AMMP program’s technical approach and test results, added: The current process works well, but has some limitations. It was exciting to see him supporting the AMMP program by reviewing technical approaches, test results and final presentations. It was great to see firsthand the potential of leveraging machine learning-based approaches to establish material tolerances. “
Senvol ML software was used in the program to support AM process qualification and develop statistically validated material properties similar to material tolerances. Moreover, we have achieved this while optimizing the data generation requirements at the same time. Importantly, the software is flexible and adaptable to any AM process, any AM machine, or any AM material. Also note that this project has not developed true tolerance. Due to budgetary and programmatic constraints, the project team had to make several decisions to simplify.
“The use of machine learning in additive manufacturing processes and materials development is very mature. The use of learning is still in progress, and I’m happy to have two successful demonstrations of the machine learning approach to tolerance: once using metal alloys in this program and comparing it to MMPDS, and once again have used and compared polymeric materials in previous programs funded by America Makes to CMH-17 – but more research is needed The benefits are significant and we will continue to work with government and industry on this We look forward to continuing our work in the field,” said Zach Simkin, president of Senvol.
“I was very pleased to join the Senvol team for this program. Addressing the Challenge Directly: I have been involved in the qualification of several additive manufacturing processes and materials for flight, but in my opinion the further development of this technology will be useful for both defense and commercial platforms. It will have a positive impact on cost, schedule and performance.” William E. Frazier, Ph.D., is a former Chief Aeronautical Engineer and Naval Materials Engineering Senior Scientist at NAVAIR and is currently President of Pilgrim He Consulting LLC.
Users of Senvol ML software include organizations in the aerospace, defense, oil and gas, consumer products, medical and automotive industries, as well as AM machine builders and AM materials suppliers.
