2009
DOI: 10.2514/1.41414
|View full text |Cite
|
Sign up to set email alerts
|

Dimensionality Reduction Approach for Response Surface Approximations: Application to Thermal Design

Abstract: Response surface approximations are a common engineering tool, often constructed based on finite element simulations. For some design problems, the finite element models can involve a large number of parameters. However, it is advantageous to construct the response surface approximations as a function of the smallest possible number of variables. The purpose of this paper is to demonstrate that a significant reduction in the number of variables needed for a response surface approximation is possible through ph… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
28
0

Year Published

2010
2010
2021
2021

Publication Types

Select...
6
1

Relationship

4
3

Authors

Journals

citations
Cited by 28 publications
(28 citation statements)
references
References 12 publications
0
28
0
Order By: Relevance
“…(8), where M and N are the number of capacity and response samples, respectively: (10) Using the computational and experimental results, along with the corresponding error estimates for the test article, we are able to refine the calculated value and its error for any design described by the design variables d and random variables r. In this way, the result of the single test can be used to calibrate calculations for other designs. We examine two methods, which take different approaches in using the test as calibration.…”
Section: Analyst-estimated Probability Of Failure Calculationmentioning
confidence: 99%
See 2 more Smart Citations
“…(8), where M and N are the number of capacity and response samples, respectively: (10) Using the computational and experimental results, along with the corresponding error estimates for the test article, we are able to refine the calculated value and its error for any design described by the design variables d and random variables r. In this way, the result of the single test can be used to calibrate calculations for other designs. We examine two methods, which take different approaches in using the test as calibration.…”
Section: Analyst-estimated Probability Of Failure Calculationmentioning
confidence: 99%
“…Next, the experimentally measured temperature is found using Eq. (10). The choice can be made to calibrate by the correction factor approach or the Bayesian-updating approach, and, further, the choice of deterministic or probabilistic redesign can be made.…”
Section: Monte Carlo Simulations Of a Future Test And Redesignmentioning
confidence: 99%
See 1 more Smart Citation
“…This simplified approach was earlier justified by comparing the results obtained from 1-D and 2-D heat transfer analyses by Bapanapalli [5]. The panel's core and webs were modeled as a thermally homogenous resultant material with its resultant conductivity and specific heat being determined through the rule of mixtures as [4][5] ( ) ) (…”
Section: A Heat Transfer Modelmentioning
confidence: 99%
“…The above panel concept was studied extensively in the past for structural optimization under thermal loading only [2][3][4][5]. The geometric parameters explored in the optimization studies for fixed areal panel dimensions are illustrated in Figure 2.…”
Section: Z Y Xmentioning
confidence: 99%