Numerical experiences with parallel clusters for generating Pareto surfaces: Application in structural topology optimization

S. Wuppalapati, Ashok D. Belegundu, A. Aziz, V. Agarwala

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    Generating Pareto surfaces is a well-accepted technique in multi-attribute decision making. For computationally intensive applications like finite element based optimization, it can become very expensive to generate the complete Pareto surface. Hence, using parallel computer clusters in these scenarios becomes very attractive. Pareto surfaces are generated using two different clusters with a structural topology problem as a test problem and the performance gains realized are analyzed. Near linear speed-ups and high efficiencies are observed on both the clusters. It is possible to integrate this methodology into commercial software applications, leading to less turn around times to make critical decisions in various applications.

    Original languageEnglish (US)
    Title of host publicationComputer Aided Optimum Design in Engineering X
    Pages3-12
    Number of pages10
    DOIs
    StatePublished - Dec 1 2007
    Event10th International Conference on Optimum Design in Engineering, OPTI 2007 - Myrtle Beach, SC, United States
    Duration: May 7 2007May 9 2007

    Publication series

    NameWIT Transactions on the Built Environment
    Volume91
    ISSN (Print)1743-3509

    Other

    Other10th International Conference on Optimum Design in Engineering, OPTI 2007
    CountryUnited States
    CityMyrtle Beach, SC
    Period5/7/075/9/07

    Fingerprint

    Structural optimization
    Shape optimization
    Turnaround time
    Application programs
    Decision making
    Topology

    All Science Journal Classification (ASJC) codes

    • Architecture
    • Civil and Structural Engineering
    • Building and Construction
    • Automotive Engineering
    • Safety, Risk, Reliability and Quality
    • Arts and Humanities (miscellaneous)
    • Transportation
    • Safety Research
    • Computer Science Applications

    Cite this

    Wuppalapati, S., Belegundu, A. D., Aziz, A., & Agarwala, V. (2007). Numerical experiences with parallel clusters for generating Pareto surfaces: Application in structural topology optimization. In Computer Aided Optimum Design in Engineering X (pp. 3-12). (WIT Transactions on the Built Environment; Vol. 91). https://doi.org/10.2495/OP070011
    Wuppalapati, S. ; Belegundu, Ashok D. ; Aziz, A. ; Agarwala, V. / Numerical experiences with parallel clusters for generating Pareto surfaces : Application in structural topology optimization. Computer Aided Optimum Design in Engineering X. 2007. pp. 3-12 (WIT Transactions on the Built Environment).
    @inproceedings{6ae62e4180bc491cb345a16d42ef7809,
    title = "Numerical experiences with parallel clusters for generating Pareto surfaces: Application in structural topology optimization",
    abstract = "Generating Pareto surfaces is a well-accepted technique in multi-attribute decision making. For computationally intensive applications like finite element based optimization, it can become very expensive to generate the complete Pareto surface. Hence, using parallel computer clusters in these scenarios becomes very attractive. Pareto surfaces are generated using two different clusters with a structural topology problem as a test problem and the performance gains realized are analyzed. Near linear speed-ups and high efficiencies are observed on both the clusters. It is possible to integrate this methodology into commercial software applications, leading to less turn around times to make critical decisions in various applications.",
    author = "S. Wuppalapati and Belegundu, {Ashok D.} and A. Aziz and V. Agarwala",
    year = "2007",
    month = "12",
    day = "1",
    doi = "10.2495/OP070011",
    language = "English (US)",
    isbn = "9781845640705",
    series = "WIT Transactions on the Built Environment",
    pages = "3--12",
    booktitle = "Computer Aided Optimum Design in Engineering X",

    }

    Wuppalapati, S, Belegundu, AD, Aziz, A & Agarwala, V 2007, Numerical experiences with parallel clusters for generating Pareto surfaces: Application in structural topology optimization. in Computer Aided Optimum Design in Engineering X. WIT Transactions on the Built Environment, vol. 91, pp. 3-12, 10th International Conference on Optimum Design in Engineering, OPTI 2007, Myrtle Beach, SC, United States, 5/7/07. https://doi.org/10.2495/OP070011

    Numerical experiences with parallel clusters for generating Pareto surfaces : Application in structural topology optimization. / Wuppalapati, S.; Belegundu, Ashok D.; Aziz, A.; Agarwala, V.

    Computer Aided Optimum Design in Engineering X. 2007. p. 3-12 (WIT Transactions on the Built Environment; Vol. 91).

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    TY - GEN

    T1 - Numerical experiences with parallel clusters for generating Pareto surfaces

    T2 - Application in structural topology optimization

    AU - Wuppalapati, S.

    AU - Belegundu, Ashok D.

    AU - Aziz, A.

    AU - Agarwala, V.

    PY - 2007/12/1

    Y1 - 2007/12/1

    N2 - Generating Pareto surfaces is a well-accepted technique in multi-attribute decision making. For computationally intensive applications like finite element based optimization, it can become very expensive to generate the complete Pareto surface. Hence, using parallel computer clusters in these scenarios becomes very attractive. Pareto surfaces are generated using two different clusters with a structural topology problem as a test problem and the performance gains realized are analyzed. Near linear speed-ups and high efficiencies are observed on both the clusters. It is possible to integrate this methodology into commercial software applications, leading to less turn around times to make critical decisions in various applications.

    AB - Generating Pareto surfaces is a well-accepted technique in multi-attribute decision making. For computationally intensive applications like finite element based optimization, it can become very expensive to generate the complete Pareto surface. Hence, using parallel computer clusters in these scenarios becomes very attractive. Pareto surfaces are generated using two different clusters with a structural topology problem as a test problem and the performance gains realized are analyzed. Near linear speed-ups and high efficiencies are observed on both the clusters. It is possible to integrate this methodology into commercial software applications, leading to less turn around times to make critical decisions in various applications.

    UR - http://www.scopus.com/inward/record.url?scp=38849103038&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=38849103038&partnerID=8YFLogxK

    U2 - 10.2495/OP070011

    DO - 10.2495/OP070011

    M3 - Conference contribution

    AN - SCOPUS:38849103038

    SN - 9781845640705

    T3 - WIT Transactions on the Built Environment

    SP - 3

    EP - 12

    BT - Computer Aided Optimum Design in Engineering X

    ER -

    Wuppalapati S, Belegundu AD, Aziz A, Agarwala V. Numerical experiences with parallel clusters for generating Pareto surfaces: Application in structural topology optimization. In Computer Aided Optimum Design in Engineering X. 2007. p. 3-12. (WIT Transactions on the Built Environment). https://doi.org/10.2495/OP070011