Show simple item record

contributor authorHanna B. Opdahl
contributor authorDavid W. Jensen
contributor authorAndrew Ning
date accessioned2022-08-18T12:10:00Z
date available2022-08-18T12:10:00Z
date issued2022/05/30
identifier other%28ASCE%29AS.1943-5525.0001425.pdf
identifier urihttp://yetl.yabesh.ir/yetl1/handle/yetl/4286121
description abstractThis study examines two techniques to optimize the design of lightweight composite IsoTruss columns subject to uniaxial compression: gradient-based (interior-point with algorithmic differentiation and Latin hypercube sampling) and gradient-free (nonsorting genetic algorithm). Both methods used analytical expressions to minimize mass with respect to the number of bays, number of carbon tows in each longitudinal member, and the outer diameter. The general geometric configuration was based on experimental specimens, tested previously, that were approximately 3 m (10 ft) long and 0.14 kg (0.3 lb). The analytical constraint functions used to predict material failure and buckling were verified by finite-element analysis and validated with experimental data in preceding works. The gradient-based and gradient-free approaches both produced similar results, indicating an optimal design 14%–16% lighter than the experimental configuration used as the initial design criteria. Sensitivity derivatives produced in the gradient-based analysis indicated that (1) increasing any of the design variables increases the mass of the structure; (2) increasing the outer diameter increases the global buckling capacity and decreases the shell-like buckling capacity; (3) increasing the number of bays increases the local buckling capacity with negligible impact on the global buckling capacity; and (4) increasing the number of tows in each longitudinal member increases both the global and local buckling capacities. Although both methods produced similar results and demonstrated satisfactory convergence, the gradient-based framework was selected for a subsequent study to evaluate novel IsoTruss configurations because it provides additional insights to the design space via outputs such as Lagrange multipliers and Jacobian matrices. The gradient-free framework could be augmented and used in future studies to optimize IsoTruss structures for specific applications and design criteria.
publisherASCE
titleImplementing Optimization Methods in the Design of IsoTruss Structures in Uniaxial Compression
typeJournal Article
journal volume35
journal issue5
journal titleJournal of Aerospace Engineering
identifier doi10.1061/(ASCE)AS.1943-5525.0001425
journal fristpage04022062
journal lastpage04022062-11
page11
treeJournal of Aerospace Engineering:;2022:;Volume ( 035 ):;issue: 005
contenttypeFulltext


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record