Towards automated guidance for helping novices design for sustainable additive manufacturing and CNC machining

Design for manufacturing;Environmental impacts;Failure analysis;Additive manufacturing;Machining;Novices

Song, Ruoyu ; Telenko, Cassandra Mechanical Engineering Durso, Frank Kurfess, Thomas Saldana, Christopher Wang, Yan ; Telenko, Cassandra

University:Georgia Institute of Technology

Department:Mechanical Engineering

关键词: Design for manufacturing;    Environmental impacts;    Failure analysis;    Additive manufacturing;    Machining;    Novices;   

Others  :  https://smartech.gatech.edu/bitstream/1853/61731/1/SONG-DISSERTATION-2019.pdf

美国|英语

来源: SMARTech Repository

【 摘 要 】

Thanks to computer-aided design and computer-aided manufacturing software, novice engineering designers can engage in product design and production more easily, increasing opportunities for innovation. Despite such increase in computer support, novice designers still make improper design decisions which unnecessarily increase the fabrication failures that lead to higher environmental impacts. For fused deposition modeling, the baseline waste rate for material consumption was 35-45% which increased the energy usage by 45%. Therefore, this research aims to discover what feedback content should be provided to novice designers, and what strategies best communicate the content. The feedback content and strategies were developed from existing databases, benchmarking studies, interviews and observation studies in a university machining mall. The feedback content identified includes the design for manufacturing guidelines and design suggestions to inform novice designers about how to make changes to fix problems with their designs. The effective strategies identified include visualization of problematic features by highlighting the features and providing example pictures to show high and low manufacturability features. A novice DFM prototype was developed and tested. From the test, the prototype was able to assist novice designers better than the benchmarking software, in terms of average number of problems identified and fixed, and average time spent. For the pawn piece with one problematic feature used in the test, the novice prototype could assist the participants in identifying 0.97 ± 0.05 problems, and fix 0.94 ± 0.06 problems. When using the benchmarking software, the participants only identified 0.78 ± 0.12 problems, and fixed 0.70 ± 0.13 problems.