With the advent in Additive Manufacturing (AM) technologies and Computational Sciences, design algorithms such as Topology Optimization (TO) have garnered the interest of academia and industry. TO aims to generate optimum structures by maximizing the stiffness of the structure, given a set of geometric, loading and boundary conditions. However, these approaches are computationally expensive as it requires many iterations to converge to an optimum solution. The purpose of this work is to explore the effectiveness of deep generative models on a diverse range of topology optimization problems with varying design constraints, loading and boundary conditions. Specifically, four distinctive models were successfully developed, trained, and evaluated to generate rapid designs with comparable results to that of conventional algorithms. Our findings highlight the effectiveness of the novel design problem representation and proposed generative models in rapid topology optimization.
In this research, we collected eye-tracking data from nine engineering graduate students as they redesigned a traditionally manufactured part for Additive Manufacturing. Final artifacts were assessed for manufacturability and quality of final design, and design behaviors were captured via the eye-tracking data. Statistical analysis of design behavior duration shows that participants with more than three years of industry experience spend significantly less time removing material and revising than those with less experience. Hidden Markov Modeling (HMM) analysis of the design behaviors gives insight to the transitions between behaviors through which designers proceed. Findings show that high-performing designers proceeded through four behavioral states, smoothly transitioning between states. In contrast, low-performing designers roughly transitioned between states, with moderate transition probabilities back and forth between multiple states.
As additive manufacturing (3D printing) technology becomes increasingly mainstream, new tools are required to enhance creative endeavors in disciplines such as fashion design, which have been slow to embrace digital technologies and computer-aided design software. Additionally, new techniques are required based on an understanding of the limitations of common desktop extrusion-based 3D printers, with 2.5D printing presenting new opportunities to rapidly produce complex forms, such as fashion, which can be assembled by hand as a hybrid approach to digital manufacturing. This paper presents a parametric system through which a new type of digitally fabricated, hand-assembled knit can be customized using a constrained selection of interactive controls. Novice users may rapidly iterate a pattern of loops and floats at a scale matching their Fused Deposition Modeling (FDM) 3D printer, creating a series of knit courses which can be assembled into a textile of expandable dimensions. More advanced users may modify the geometry or logic of the system, building new forms of knits that could not be manufactured through traditional means. This paper will guide designers through the process of developing a new typology of textile appropriate for production on ubiquitous FDM machines, following a workflow from 2D to 2.5D to 3D. Examples were printed at various scales and contribute to discourse on customization, parametric design and visual programming languages for additive manufacturing.
She earned her B.S. in aerospace engineering from Syracuse University and her Ph.D. in engineering education from the School of Engineering Education at Purdue University. She is particularly interested in teaching conceptions and methods and graduate level engineering education.
While calculating intercoder reliability (ICR) is straightforward for text-based data, such as for interview transcript excerpts, determining ICR for naturalistic observational video data is much more complex. To date, there have been few methods proposed in literature that are robust enough to handle complexities such as the occurrence of simultaneous event complexity and partial agreement by the raters. This is especially important with the emergence of high-resolution video data, which collects nearly continuous or continuous observational data in naturalistic settings. In this paper, we present three approaches to calculating ICR. First, we present the technical approach to clean and compare two coders’ results such that traditional metrics of ICR (e.g., Cohen’s κ, Krippendorff’s α, Scott’s Π) can be calculated, methods previously unarticulated in literature. However, these calculations are intensive, requiring significant data manipulation. As an alternative, this paper also proposes two novel methods to calculate ICR by algorithmically comparing visual representations of each coders’ results. To demonstrate efficacy of the approaches, we employ all three methods on data from two separate ongoing research contexts using observational data. We find that the visual methods perform as well as the traditional measures of ICR and offer significant reduction in the work required to calculate ICR, with an added advantage of allowing the researcher to set thresholds for acceptable agreement in lag time. These methods may transform the consideration of ICR in other studies across disciplines that employ observational data.
The layered fabrication approach induces directional anisotropy and impacts mechanical strength of FDM components significantly. This paper proposes generalized machine learning based parameter optimization framework to determine optimal build orientation for FDM components. The algorithm determines ideal build orientation by maximizing the minimum Factor of Safety (FoS) for the component under prescribed loading conditions ensuring its even distribution. An Artificial Neural Network (ANN) coupled with Bayesian algorithm has been employed to accelerate the optimization process. The algorithm begins with an initial sample data collected using brute force approach; uses single layered ANN for approximation and optimization is achieved using Bayesian algorithm. A series of computational experiments considering five different test components has been devised to evaluate the performance and efficacy of the proposed algorithm. These experiments demonstrated that the proposed algorithm can determine the optimum building orientation effectively with certain limitations
Additive manufacturing (AM) offers access to the entire volume of a printed artifact during the build operation. This makes it possible to embedding foreign components (e.g. sensors, motors, actuators) into AM parts, thus enabling multifunctional products directly from the build tray. However, the process of designing for embedding currently requires extensive designer expertise in AM. Current methods rely on a designer to select an orientation for the embedded component and design a cavity such that the component can be successfully embedded without compromising the print quality of the final part. For irregular geometries, additional design knowledge is required to prepare a shape converter: a secondary piece to ensure a flush deposition surface on top of the embedded component. This research aims to develop a tool to automate these different design decisions for in-situ embedding, thus reducing the need for expert design knowledge. A three-stage process is proposed to 1) find the optimum orientation based on cavity volume and cross-section area, 2) create the necessary cavity geometry to successfully insert the component, and 3) perform a Boolean operation to create the digital design for any requisite shape converter. Performance of the tool is demonstrated with four test cases with varying levels of geometric complexity. These test cases show that the proposed process successfully handles arbitrary embedded geometries, though several limitations are noted for future work.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.