Journal
ENGINEERING ANALYSIS WITH BOUNDARY ELEMENTS
Volume 142, Issue -, Pages 93-116Publisher
ELSEVIER SCI LTD
DOI: 10.1016/j.enganabound.2022.05.019
Keywords
Additive Manufacturing; Artificial Intelligence; Machine Learning; Lattice Structure; Mechanical Properties
Funding
- Taif University, Taif, Saudi Arabia [TURSP - 2020/14]
Ask authors/readers for more resources
The article firstly provides a short review on the assessment of mechanical properties of lattice structures and introduces some numerical models for predicting their mechanical properties. Secondly, a deep learning model is proposed for highly accurate assessment of stress-strain behavior of various lattice structures, which has simple, quick and precise predictability.
Lattice structures fabricated by additive manufacturing (AM) technology have many excellent properties, such as lightweight, high strength, energy absorption, and vibration reduction, which have been extensively researched and made a breakthrough. Lattice structures have been commonly used in aviation, bioengineering, robotics, and other industrial fiber because of their outstanding properties. The first part of this article provides a short review on the assessment of mechanical properties of various lattice structures in terms of their classification, applications, materials and fabrication techniques, and complexity of designing, fabrication, and post-processing as well as some of the numerical models to predict the mechanical properties of the lattice structures. The second part of the article proposes a deep learning (DL) model for a highly accurate stress-strain behavior assessment of numerous lattice structures such as namely: the octet, face center-cubic, body-centered cubic, diamond, rhombic, cubic, truncated cube, and truncated cuboctahedron, etc, which were fabricated using many different materials via various approaches and methods. Using the proposed DL model, an accuracy in terms of R-2 = 0.999 (correlation coefficient), MSE = 0.0017 (mean squared error), and MAE = 0.0312 (mean absolute error) can be achieved for the prediction of the deemed mechanical property of the lattice structures. The model contains simple, quick and precise predictability that makes it ideal for the use of lattice structures in various practical applications, including heater and heat exchangers, engine hood, biomedical implant, wings, gas turbine, vibration absorber, robotic device, etc.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available