Using Simulation to Accelerate Autonomous Experimentation (AE): A Case Study Using Mechanics
TL;DR: A case study on the mechanics of additively manufactured polymer structures is used to investigate whether imperfect data from simulation can accelerate autonomous experimentation by highlighting multiple ways in which simulation can improve AE through transfer learning.
Abstract: Autonomous experimentation (AE) accelerates research by combining automation and machine learning to perform experiments intelligently and rapidly in a sequential fashion. While AE systems are most needed to study properties that cannot be predicted analytically or computationally, even imperfect predictions can in principle be useful. Here, we use a case study on the mechanics of additively manufactured polymer structures to investigate whether imperfect data from simulation can accelerate AE. Initially, we study resilience, a property that is well-predicted by finite element analysis (FEA), and find that FEA can be used to build a Bayesian prior, and that experimental data can be integrated using discrepancy modeling to reduce the number of needed experiments ten-fold. Next, we study toughness, which is not well predicted by FEA, and find that FEA can still improve learning by transforming experimental data and guiding experiment selection. These results highlight multiple ways in which simulation can improve AE through transfer learning.
Citations
More filters
TL;DR: In this article, the effect of bed temperature on the peak removal force for polylactic acid (PLA) parts printed on bare borosilicate glass and polyimide (PI)-coated beds was studied.
Abstract:
Additive manufacturing (AM) techniques, such as fused deposition modeling (FDM), are able to fabricate physical components from three-dimensional (3D) digital models through the sequential deposition of material onto a print bed in a layer-by-layer fashion. In FDM and many other AM techniques, it is critical that the part adheres to the bed during printing. After printing, however, excessive bed adhesion can lead to part damage or prevent automated part removal. In this work, we validate a novel testing method that quickly and cheaply evaluates bed adhesion without constraints on part geometry. Using this method, we study the effect of bed temperature on the peak removal force for polylactic acid (PLA) parts printed on bare borosilicate glass and polyimide (PI)-coated beds. In addition to validating conventional wisdom that bed adhesion is maximized between 60 and 70 °C (140 and 158 °F), we observe that cooling the bed below 40 °C (104 °F), as is commonly done to facilitate part removal, has minimal additional benefit. Counterintuitively, we find that heating the bed after printing is often a more efficient process for facile part removal. In addition to introducing a general method for measuring and optimizing bed adhesion via bed temperature modulation, these results can be used to accelerate the production and testing of AM components in printer farms and autonomous research systems.
4 citations
References
More filters
TL;DR: In this article, the effect of bed temperature on the peak removal force for polylactic acid (PLA) parts printed on bare borosilicate glass and polyimide (PI)-coated beds was studied.
Abstract:
Additive manufacturing (AM) techniques, such as fused deposition modeling (FDM), are able to fabricate physical components from three-dimensional (3D) digital models through the sequential deposition of material onto a print bed in a layer-by-layer fashion. In FDM and many other AM techniques, it is critical that the part adheres to the bed during printing. After printing, however, excessive bed adhesion can lead to part damage or prevent automated part removal. In this work, we validate a novel testing method that quickly and cheaply evaluates bed adhesion without constraints on part geometry. Using this method, we study the effect of bed temperature on the peak removal force for polylactic acid (PLA) parts printed on bare borosilicate glass and polyimide (PI)-coated beds. In addition to validating conventional wisdom that bed adhesion is maximized between 60 and 70 °C (140 and 158 °F), we observe that cooling the bed below 40 °C (104 °F), as is commonly done to facilitate part removal, has minimal additional benefit. Counterintuitively, we find that heating the bed after printing is often a more efficient process for facile part removal. In addition to introducing a general method for measuring and optimizing bed adhesion via bed temperature modulation, these results can be used to accelerate the production and testing of AM components in printer farms and autonomous research systems.
4 citations