Lonnie J. Love received his B.S. and M.S. degree in mechanical engineering from Old Dominion University, and a Ph.D. in mechanical engineering from the Georgia Institute of Technology. He is currently a distinguished research scientist in the Energy and Transportation ScieDivision and group leader of the Manufacturing Systems Research Group at the Department of Energy’s (DOE’s) Oak Ridge National Laboratory (ORNL). He has made major contributions at ORNL as a researcher, a leader, and an innovator in advanced robotics and additive manufacturing (AM). His research has most recently focused on largescale and highspeed advanced AM and 3D printing.
What are some of the biggest bottlenecks slowing AM from becoming a mainstream manufacturing technology and what is the DOE’s Manufacturing Demonstration Facility (MDF) at ORNL doing to address these challenges?
Reliability, cost and being user friendly.
• Reliability and being user friendly because what you design, can be printed. Design rules are different from machine to machine, technology to technology. We need tools to aid the engineer in designing for different AM processes.
• Cost, especially for metals, is prohibitive for a lot of applications. Material costs can be as high a $50 to $200/lb for some materials. But what’s worse is the machines are expensive (~$500K to $2M) with often low production rates (1 to 5 ci/hr., or 1000 lb./yr. to 2000 lb./ yr.). In general, net cost of printing is closer to $500/lb. to $2000/lb. for final parts. ORNL is focusing on large scale, high deposition rate systems that can use commodity grade feedstocks. Our goal is to get to $20/lb. for final parts.
When the MDF research team printed the first ever composites car using Big Area Additive Manufacturing (BAAM), a groundbreaking 3D printing system developed jointly between ORNL and Cincinnati Incorporated, there was a lot of excitement in the industry on the possibilities for the technology. How is that research maturing?
The Local Motors Strati car showed the scalability and impact of composites on AM. However, it also showed some of the challenges. A big one is surface finish. There has been a lot of activity on low cost, rapid coating technologies. Also, the killer application is tooling, being able to print molds in hours rather than weeks at costs of thousands of dollars rather than $10K’s to $100K’s. We have fabricated tooling for the aerospace, automotive, appliance, marine (boats), and precast concrete industries…tooling impacts every industry.
Do you anticipate large-scale metal systems as a way of moving away from small build volumes and moving to larger equipment and parts? Will it be combined with hybrid machining methods so large parts can be printed and machined at the same time?
Yes, large scale metal systems will evolve to multi-material (low cost inner core with hard outer surface) with integrated machining. ORNL will be doing a demonstration of a production hot stamping die later this year with conformal cooling.
What role do you see simulation playing in developing the large-scale metal systems method?
Simulation is critical. We need to be able to predict whether a part is ‘printable’ and the expected properties. First, we will use modeling and simulation to guide us on the design of the parts. Next, we will use it to help guide us on the toolpaths that minimize stresses in the part. The simulation tools need to be directly integrated into the design tools.
Do you see simulation being used to control and optimize the manufacturing process itself someday? If so, what are some of the simulation based analytics you see necessary to drive such controls?
Yes. But simulation is only animation until you validate the models. I believe that there will be different levels of simulation.
• As above, we need tools that may be reduced model to help us design the parts.
• All AM systems are limited by thermodynamics and heat transfer. I believe there is a big need for modeling and simulation to design AM systems
• We need higher fidelity AM modeling to help optimize the process and systems (e.g. what infill patterns are best for different geometries…).
In this work, you’ve spent a lot of effort in careful experimentation of parts and then validated them with simulation. For mainstream usage, do you expect every manufactured part to have to go through the same rigor, or, do you see good benchmarks as a way to help build confidence in the predictive nature and accuracy of your simulation models?
As above, I believe the mainstream engineer needs something simple and intuitive that gives them the confidence that the part can be printed successfully and will have predictable properties. Fifteen years ago, Finite Element Analysis (FEA) and design were two different pieces of software. Dassault Systèmes started bundling them together to make it more efficient and easier to use. Today, we have the same problem with slicing software. A similar approach needs to be taken for the AM process itself. The way you print the part will impact material properties. Therefore, we need to have slicing software directly integrated into the design package. You design, slice, and THEN analyze before manufacturing.
This article was published in the September 2017 issue of SIMULIA Community News magazine. Coming soon, stay tuned for the second half of this interview with Lonnie Love!
