Speeding up additive manufacturing with fine-tuned software, and help from GE

A multidisciplinary team, led in part by GE Research, is working to cut the design and validation process in 3D printed turbomachinery components by two thirds - from 2-5 years to a 1-2 years.

Advanced Research Projects Agency (ARPA-E), a division of the U.S. Energy Department that seeds early-stage tech research, is funding a new push to develop software and standards to accelerate 3D-printing design and prototype development. The agency has granted $1.3 million to a consortium of GE, the PARC subsidiary of Xerox and the Oak Ridge National Laboratory (ORNL). The teams believe they can cut the timeline for designing and validating 3D printed components by two-thirds.

3D printing allows engineers and designers to print complex metal parts straight from a computer file. Traditionally, this kind of manufacturing has meant casting parts from molten alloys or pounding them out on massive forging machines. Additive manufacturing, by contrast, takes place in large, humming machines that layer tiny metal particles and fuse them with lasers. This allows for the creation of stronger, lighter and far more intricate pieces than what has been possible with traditional methods.

From left to right: Naresh Iyer, GE Research principal scientist for machine learning; Sathyanarayanan Raghavan, lead engineer for structural mechanics/dynamics; Dean Robinson, principal engineer for additive research; and Carlos Velez, lead engineer in the Thermosciences group. They are part of a multidisciplinary research team, working to reduce the timeline for designing and validating 3D-printed turbomachinery components from 2-5 years to a quicker 1-2 years. Here, they are in discussing 3D design tools in GE Research’s Information Works Lab. Top: A sampling of the types of design tools GE, ORNL and PARC researchers will utilize to accelerate 3D part design of turbomachinery parts. This GIF illustrates the design evolution of a thermal optimum heat exchanger, using an enhanced multiphysics topology optimization tool. Images credit: GE Research.[/caption]

With the support of the APRA-E grant, GE and partners aim to streamline the painstaking process of prototype development that precedes mass production. Typically, prototype teams will jump into building brand new parts and continue to iterate the material and design throughout the process.

It’s not unusual for new-parts development to take three to five years, and a string of 10 to 30 prototype variations, before engineers nail down the manufacturing process and transfer it to a factory for mass production.

Under the ARPA-E program, GE and partners plan to compile a database of materials that could be used in 3D printing based on their various physical characteristics. Researchers will also look at other factors in the production process — like laser speed, airflow inside the machines, and levels of energy applied in layering — to better understand how to prevent circumstances that can warp prototypes. When it comes to shaping the actual designs, ORNL’s supercomputers will help speed up prototype design by precisely modeling the variables a part may experience during the build process. These physical factors, such as thermal profile, stresses and strains, have been difficult to estimate precisely.

The trial project will involve redesigning a natural gas turbine component called the shroud, which directs the airflow into the hottest part of the machine. The goal: a revamped shroud with fewer parts, lighter weight and better airflow. This approach has the potential to significantly boost GE turbine efficiency, which would lead to real customer savings in fuel and other costs.

Overall, the consortium expects to shorten development time to just one to two years for new prototypes, while refining existing designs would move even faster.