MEDEMA LABS CORP. — Department of Defense SBIR Phase I: N202-117

Description

While topology optimization can define parts that have minimal and nearly uniform strain energy, these solutions cannot be manufactured without modification. Numerous research papers have looked at ways to incorporate manufacturing constraints within the optimization. This has promise for additive manufacturing but has significant hurdles when subtractive manufacturing is used. There is not a clear sense of what is feasible and what is infeasible in subtractive machining; rather, one can spend a prohibitive amount of time removing minute amounts of material to get ever close to the organic curvy shapes produced by topology optimization. We propose that the best approach to arrive at optimal and manufacturable parts is to first use topology optimized parts without manufacturing constraints and then apply an artificial intelligence search on the result to find suitable solutions. Our approach is also distinguished by the fact that the results will not be a single solution but rather a range of solutions that show a trade-off in manufacturing cost and time versus how well the idealized topology optimized solution is created. The prime organization, Medema Labs, is a recent university spin-out company in Oregon focused on developing commercial software from ten years of academic research in automated manufacturing planning. For this project, the team will build upon a recently published journal article by academic associates, where material is added to the topology optimized part in varying amounts depending on the choice of tool and approach direction. This algorithm will be developed for Medema Labs new desktop tool, Medema PartAnalyzer, which can automate the planning of the entire manufacturing process including the choice of stock size, machine, setup orientations, and tools. The result will be a range of manufacturing plans with varying tradeoffs of added mass and manufacturing costs. We will be working closely with a local high-tech machine shop, Machine Sciences, to ensure that the automated modifications to the topology optimized parts produce can be realistically manufactured.  These solutions could then be used by engineering designers to make informed decisions about cost versus performance. During the Phase I Base Period, the team will demonstrate the effectiveness of our approach at modifying topology optimized parts and validate the resulting manufacturing plans by producing 3-axis CNC machined parts.  During the Phase I Option Period, the team will further calibrate the manufacturing plan results and developing the graphical user interface components necessary to make the tool truly useful to engineers. Future phases will focus on adding lathes and four+ axis CNC mills, as well as, enabling importing and exporting to commercial topology optimization and CAM software.