An official website of the United States government
Abstract Additive manufacturing (AM) has revolutionized the way we design, prototype, and produce complex parts with unprecedented geometries. However, the lack of understanding of the functional properties of 3D-printed parts has hindered their adoption in critical applications where reliability and durability are paramount. This paper proposes a novel approach to the functional qualification of 3D-printed parts via physical and digital twins. Physical twins are parts that are printed under the same process conditions as the functional parts and undergo a wide range of (destructive) tests to determine their mechanical, thermal, and chemical properties. Digital twins are virtual replicas of the physical twins that are generated using finite element analysis (FEA) simulations based on the 3D shape of the part of interest. We propose a novel approach to transfer learning, specifically designed for the fusion of diverse, unstructured 3D shape data and process inputs from multiple sources. The proposed approach has demonstrated remarkable results in predicting the functional properties of 3D-printed lattice structures. From an engineering standpoint, this paper introduces a comprehensive and innovative methodology for the functional qualification of 3D-printed parts. By combining the strengths of physical and digital twins with transfer learning, our approach opens up possibilities for the widespread adoption of 3D printing in safety-critical applications. Methodologically, this work presents a significant advancement in transfer learning techniques, specifically addressing the challenges of multi-source (e.g., digital and physical twins) and multi-input (e.g., 3D shapes and process variables) transfer learning.
more »
« less
Mitigating Scattering Effects in Light-Based Three-Dimensional Printing Using Machine Learning
Abstract When using light-based three-dimensional (3D) printing methods to fabricate functional micro-devices, unwanted light scattering during the printing process is a significant challenge to achieve high-resolution fabrication. We report the use of a deep neural network (NN)-based machine learning (ML) technique to mitigate the scattering effect, where our NN was employed to study the highly sophisticated relationship between the input digital masks and their corresponding output 3D printed structures. Furthermore, the NN was used to model an inverse 3D printing process, where it took desired printed structures as inputs and subsequently generated grayscale digital masks that optimized the light exposure dose according to the desired structures’ local features. Verification results showed that using NN-generated digital masks yielded significant improvements in printing fidelity when compared with using masks identical to the desired structures.
more »
« less
- Award ID(s):
- 1907434
- PAR ID:
- 10258096
- Date Published:
- Journal Name:
- Journal of Manufacturing Science and Engineering
- Volume:
- 142
- Issue:
- 8
- ISSN:
- 1087-1357
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
3D‐printed polymer blends with programmable mechanical and compositional heterogeneity were fabricated using grayscale digital light processing by spatially modulating the intensity of light during printing and swelling the resulting part with a second monomer. A rubbery poly(ethylene glycol) diacrylate functionally graded print is variably swollen with acrylamide monomer as a function of crosslinking density. Following a secondary polymerization, a 3D‐printed functionally graded blend with regions of varying composition and stiffness was formed. A deterministic model for polymer conversion informs printing conditions to correspond with predicted material properties based upon local volume fractions of the two materials. Upon the secondary polymerization, two networks are present within the printed structure including glassy and rubbery regions. The compressive moduli of local regions within prints ranges from 76 to 200 MPa and measured moduli of the structures agree with predicted values acquired using finite element analysis. A lattice structure with prescribed local stiffness printed using grayscale exposures deforms differentially when compressed. Advantageously, local dimensional deformations caused by the removal of the unreacted printing monomer are eliminated due to the introduction of the second polymer. This method provides predictive control over local mechanical properties and high shape precision while maintaining the simplicity of vat photopolymerization.more » « less
-
Abstract Liquid crystalline elastomers (LCEs) are anisotropic soft materials capable of large dimensional changes when subjected to a stimulus. The magnitude and directionality of the stimuli‐induced thermomechanical response is associated with the alignment of the LCE. Recent reports detail the preparation of LCEs by additive manufacturing (AM) techniques, predominately using direct ink write printing. Another AM technique, digital light process (DLP) 3D printing, has generated significant interest as it affords LCE free‐forms with high fidelity and resolution. However, one challenge of printing LCEs using vat polymerization methods such as DLP is enforcing alignment. Here, we document the preparation of aligned, main‐chain LCEs via DLP 3D printing using a 100 mT magnetic field. Systematic examination isolates the contribution of magnetic field strength, alignment time, and build layer thickness on the degree of orientation in 3D printed LCEs. Informed by this fundamental understanding, DLP is used to print complex LCE free‐forms with through‐thickness variation in both spatial orientations. The hierarchical variation in spatial orientation within LCE free‐forms is used to produce objects that exhibit mechanical instabilities upon heating. DLP printing of aligned LCEs opens new opportunities to fabricate stimuli‐responsive materials in form factors optimized for functional use in soft robotics and energy absorption.more » « less
-
Abstract The majority of 3D‐printed biodegradable biomaterials are brittle, limiting their application to compliant tissues. Poly(glycerol sebacate) acrylate (PGSA) is a synthetic biocompatible elastomer and compatible with light‐based 3D printing. In this article, digital‐light‐processing (DLP)‐based 3D printing is employed to create a complex PGSA network structure. Nature‐inspired double network (DN) structures consisting of interconnected segments with different mechanical properties are printed from the same material in a single shot. Such capability has not been demonstrated by any other fabrication techniques so far. The biocompatibility of PGSA is confirmed via cell‐viability analysis. Furthermore, a finite‐element analysis (FEA) model is used to predict the failure of the DN structure under uniaxial tension. FEA confirms that the DN structure absorbs 100% more energy before rupture by using the soft segments as sacrificial elements while the hard segments retain structural integrity. Using the FEA‐informed design, a new DN structure is printed and tensile test results agree with the simulation. This article demonstrates how geometrically‐optimized material design can be easily and rapidly constructed by DLP‐based 3D printing, where well‐defined patterns of different stiffnesses can be simultaneously formed using the same elastic biomaterial, and overall mechanical properties can be specifically optimized for different biomedical applications.more » « less
-
Abstract Two-photon lithography (TPL) is an attractive technique for nanoscale additive manufacturing of functional three-dimensional (3D) structures due to its ability to print subdiffraction features with light. Despite its advantages, it has not been widely adopted due to its slow point-by-point writing mechanism. Projection TPL (P-TPL) is a high-throughput variant that overcomes this limitation by enabling the printing of entire two-dimensional (2D) layers at once. However, printing the desired 3D structures is challenging due to the lack of fast and accurate process models. Here, we present a fast and accurate physics-based model of P-TPL to predict the printed geometry and the degree of curing. Our model implements a finite difference method (FDM) enabled by operator splitting to solve the reaction–diffusion rate equations that govern photopolymerization. When compared with finite element simulations, our model is at least a hundred times faster and its predictions lie within 5% of the predictions of the finite element simulations. This rapid modeling capability enabled performing high-fidelity simulations of printing of arbitrarily complex 3D structures for the first time. We demonstrate how these 3D simulations can predict those aspects of the 3D printing behavior that cannot be captured by simulating the printing of individual 2D layers. Thus, our models provide a resource-efficient and knowledge-based predictive capability that can significantly reduce the need for guesswork-based iterations during process planning and optimization.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1115/1.4046986
