Additive Manufacturing Of Functional Constructs Under Process Uncertainty
| dc.contributor.author | Cohen, Daniel | en_US |
| dc.date.accessioned | 2010-08-05T16:25:33Z | |
| dc.date.available | 2015-08-05T06:22:49Z | |
| dc.date.issued | 2010-08-05T16:25:33Z | |
| dc.description.abstract | In recent years, development of novel material-sets has enabled additive manufacturing (AM) to transform from being used purely for model-making applications, to production of functional constructs. What was once only a rapid prototyping technology is now being used to print functional components, including batteries, actuators, transistors, and for tissue engineering applications, living tissue (Chapter 1). Associated with these new printing inks, however, is a drastic increase in the complexity of AM materials, and consequently, in the process uncertainty related to deposition of these materials. New applications enabled by functional printing capabilities, in particular in situ AM, also have associated process uncertainties, including situational and environmental uncertainties. That is, uncertainty in the shape of the substrate and in environmental parameters, such as temperature and humidity. As additive manufacturing makes the transition from a prototyping technology to more of a functional-object fabrication platform, these new associated process uncertainties must be addressed to yield sufficient geometric fidelity. Existing control schemes largely relied upon open-loop control and did not handle uncertainty through control algorithms, but instead avoided them by limiting their material-sets, printing on trivially shaped substrates, and restricting environmental conditions. A few techniques used geometric feedback to handle materials uncertainty, but these techniques did so on a process-parameter-level, and did not monitor/manipulate on the whole-part level. As a result, these techniques could not detect high-level errors such as whole-part deformation. The technique proposed herein, Greedy Geometric Feedback (GGF), closes the loop on the whole-part level and therefore can detect/correct types of errors that were previously un-addressable. Simulations and physical experiments were employed to validate and study the GGF algorithm. Not only was GGF effective at handling materials uncertainties, but it also has potential for situational and environmental uncertainties. Additional work focused on situational uncertainty and alternative control schemes were developed that effectively handled this type of uncertainty with less computational and data collection overhead. A novel differencebased planning approach was employed to explore in situ AM repair of osteochondral defects, and repair/adaptation of a four-legged robot. These proof-of-concept prints are the only known examples of generalized in situ AM, to date, in which the AM system was not provided a priori hard-coded substrate geometric information. The contributions of the work presented herein fall into three categories: 1) development of functional printing materials, 2) development of novel methodologies for quantitatively optimizing the printing qualities of functional printing inks, and 3) development of novel generalized control schemes for handling AM process uncertainty. | en_US |
| dc.identifier.other | bibid: 6980463 | |
| dc.identifier.uri | https://hdl.handle.net/1813/17206 | |
| dc.language.iso | en_US | en_US |
| dc.title | Additive Manufacturing Of Functional Constructs Under Process Uncertainty | en_US |
| dc.type | dissertation or thesis | en_US |
Files
Original bundle
1 - 1 of 1