Session: 17-01-01: Research Posters

Paper Number: 150142

150142 - Using Topological Interlocked Tessellations to Enable Cooperative 3d Printing 

With the rapid growth of additive manufacturing, several challenges have appeared. Printing large parts in a time-efficient manner can be particularly problematic. To address this issue, cooperative 3D printing (C3DP) has gained interest, as multiple robots can work together to print at the same time. However, C3DP introduces several new challenges, including a) segmentation (partitioning), b) collision avoidance, c) printing schedule optimization, and d) preserving the mechanical integrity of the part. In this work we use topologically interlocked tessellations to segment the part to aid in the cohesion between the parts printed by different robots. To create topologically interlocked tessellations, we use interlocked space-filling helical shapes. We specifically focus on creating tiles which result from stacking 2-honeycombs (i.e., tessellations of the plane with a single ‘prototile’) along a helical trajectory in the vertical direction. The amount of interlocking can then be tuned using parameters of the helix generating the structure, as well as the initial pattern of the 2-honeycomb.

In order to demonstrate that topologically interlocked tessellations are advantageous for C3DP as a potent method of segmentation and achieving mechanical robustness of the parts, we conduct finite element simulations on assemblies of topologically interlocked tessellations. Through simulations and physical tests, we determine that the amount of interlocking is highly correlated with the helix parameters. To increase the amount of interlocking, the radius or number of turns of the helix used for translation can be increased. This approach allows for a parameterizable design space, wherein the amount of interlocking can be tuned based on the requirements of the designer. Within the context of C3DP this approach results in helix parameters that can be varied to increase (or decrease) the amount of cohesion between parts printed by different robots.
Once the part has been segmented into tiles, the printing cells can be assigned on a layer-by-layer basis depending on which robot is closest. To avoid collision between robots, we label the printing cells based on the distance to cells assigned to other robots. If a cell is near cells assigned to other robots it is labelled ‘interfacing’. If only one robot is allowed to print the interfacing cells at a time, then collision is completely avoided.
To minimize overall printing time, we introduce the Concurrence Measure (CM), which quantifies the amount of parallel printing that occurs during the printing process. Using CM as the objective function, the orientation and position of a part can be optimally placed to allow for equal printing work as well as minimal interfacing areas. Through testing of various helix parameters for interlocking, it has been determined that the amount of interlocking is independent of CM, meaning the degree of interlocking can be independently adjusted without affecting the time required to print the part.

The viability of the proposed approach combining the three methods introduced, interlocking-based segmentation, collision avoidance, and optimization of part placement, has been demonstrated through the printing of arbitrary geometries.

Presenting Author: Matthew Ebert Texas A&M University

Presenting Author Biography: Matthew Ebert is a PhD student working under Dr. Vinayak Krishnamurthy at Texas A&M University in Mechanical Engineering. Matthew graduated from Texas A&M University with his undergraduate degree in Fall of 2021. His research involves how to segment a part for multi-robot additive manufacturing in order to aid in adhesion between pieces printed by different robots as well as avoid collision between the multiple robots.

Authors:

Matthew Ebert Texas A&M University
Ronnie Stone University of Texas
John Koithan Texas A&M University
Ergun Akleman Texas A&M University
Yuri Estrin Monash University
Zhenghui Sha University of Texas
Matt Pharr Texas A&M University
Vinayak Krishnamurthy Texas A&M University