Senvol machine learning software recognized by U.S. Air Force, demonstrating robust 3D printing material identification capabilities

Introduction: Generally speaking, in order to identify a3D printingThe pros and cons of the practical performance of materials often need to go through a long application effect evaluation and rich data accumulation to determine. Today, machine learning technology is greatly shortening this process. The Antarctic Bear believes that if the technology matures, the traditional “material trial and error method” can even be eliminated.

On February 22, 2022, additive manufacturing data specialist Senvol demonstrated how its machine learning (ML) technology can be used to accelerate new
aviation
3D printingMaterial identification process.
The company has already deployed Senvol ML under a contract issued by US 3D printing accelerator AmericaMakes
software
, in order to expedite the determination of whether the material development process is economical. The project is sponsored by the United States Air Force (USAF)
funding
which is considered key to producing a new generation of durable, lightweight components for aerospace applications
aerospace
and commercial fields.

Dr. William E. Frazier, retired chief scientist and project participant at NAVAIR, said: “I have been involved in several in-flight qualifications of additive manufacturing processes and materials, and in my opinion, further development of this technology will be a great addition to the Department of Defense (DoD) ) and the cost, schedule and performance of commercial platforms. Senvol’s Mil-enabled approach directly addresses a major industry challenge: faster, cost-effective3D printing materialsdevelopment. “

△ The US Air Force is using Senvol’s AM software to optimizeEOS 3D printersetup to produce aerospace parts. Image via Mikayla Heineck of the U.S. Air Force.

Military potential of Senvol ML

In fact, Senvol ML is a data-driven software that can be used to analyze the relationship between 3D printing parameters and the final properties of materials. Unconstrained by any one technology, the program is compatible with all known AM processes, enabling users to generate a set of parameters for their system or material based on pre-set target mechanical properties.

Senvol ML’s predictive capabilities are largely attributable to its modular Integrated Computational Materials Engineering or “ICME” framework. Divided into four modules, this ICME quantifies the relationship between collected data on parameters, material properties, process characteristics and mechanical properties to generate predictions of how any resulting part will perform at rapid speed.

In doing so, Senvol ML provides aerospace, defense, consumer,medical,carand oil and gas field adopters provide an important quality assurance tool, and the technology is now increasingly used in the military field as well. Since 2020 alone, Senvol has used its software to help print large aerospace parts and developed a missile qualification workflow, while receiving further research and development funding from the Department of Defense.

“Senvol’s data-driven machine learning technology has the potential to dramatically reduce the cost of developing usable materials,” said Zach Simkin, President of Senvol. “By demonstrating a new, significantly more efficient, and quantifiable development approach, Senvol offers the U.S. Air Force, Made in America members, as well as the additive manufacturing industry as a whole, are proving its tremendous value.”

△ Analysis example from SenvolML – stainless steel powderFinal and laser powder bed fusion. Image via Senvol.

Development of Materials Analysis Technology

Sunwall’s “Made in America” research program revolves around the process of determining a material’s “Performance Allowance.” Also known as “Design Allowances”, these values basically represent the minimum durability properties required for a statistically determined material to make them viable in particularly demanding applications.

Specifically, the project calls for showing how Senvol ML software’s machine learning algorithms can speed up the process, while reducing the amount of empirical data typically required, which could be a barrier to 3D printing’s wider military deployment.

Dr. Brandon Ribic, Technical Director of Made in America, explained: “Additive manufacturing is a modern and digital manufacturing method with rapid customizable processing. Continuing the traditional trial-and-error approach to materials is a constraint on additive manufacturing for a wider range of materials. and process options and capacity bottlenecks.”

According to Ribic, Senvol has successfully overcome this bottleneck by developing a data analysis-driven identification method, demonstrating the capabilities of its software in a very powerful way. In fact, by partnering with Northrop Grumman, the National Institute of Aeronautics and Astronautics (NIAR), Stratasys Direct Manufacturing, and Pilgrim Consulting, the company is said to be able to determine proven material properties while optimizing its data generation Require.

Although Senvol has focused its efforts on finding allowable values for properties of flame-retardant nylon 11 processed by powder bed fusion machining in the project, it says its approach can be applied to “any AM process, machine, and material” in order to be more efficient than traditional methods. Determine its parameters quickly, economically and accurately.

Simkin concluded, “The results of this Made in America initiative have been incredibly successful. In addition, we have identified several other areas of opportunity to delve deeper into machine learning capabilities to address key needs in this industry. We look forward to continuing to work with Industry cooperation to advance the development of this cutting-edge field.”

△Alchemite software interface. Image via Intellegens

A new identification method?

Compared to traditional “trial and error” identification methods for materials, machine learning-based algorithms offer a means of eliminating the human error behind failed prints. While many of these programs are still experimental, the benefits of software developed by companies like Senvol ML and Intellegens are now starting to make their way onto the factory floor.

Just last year, the latter teamed up with the University of Sheffield’s AMRC and Boeing to come up with a new and optimised method for 3D printing aerospace parts. Leveraging the company’s own Alchemite ML platform, the project aims to enhance the powder bed fusion process to ensure parts are produced faster, cheaper and with more material than before.

Similarly, engineers at the Massachusetts Institute of Technology (MIT) have developed an open-source ML algorithm designed to help more people speed up the process of identifying 3D printable materials. In their “AutoOED” software, the team’s program is able to automatically identify viable materials with desired qualities, such as toughness, stiffness and strength.

(responsible editor: admin)

Senvol machine learning software recognized by U.S. Air Force, demonstrating robust 3D printing material identification capabilities

0 Comments for “Senvol machine learning software recognized by U.S. Air Force, demonstrating robust 3D printing material identification capabilities”

Leave a Reply Cancel reply