Engineering Faculty Organization (EFO)

Jayathi Y. Murthy Seminar

Abstract

There has been great interest recently in predicting thermal transport at the nanoscale. Sensitivities or derivatives are used in thermal transport simulations to describe important crystal properties. For example, the Gruneisen parameter is the derivative of phonon frequency with respect to the volume of the crystal. The phonon group velocity is the derivative of the frequency with respect to the wave vector. A force constant is a derivative of the total potential energy of the system with respect to the displacement of a given atom. Accurate methods to compute these quantities are needed without the approximations inherent in finite difference methods. Derivatives are also needed for gradient-based optimization methods, for example, such as those used in topology optimization. Furthermore, engineers can benefit a great deal from understanding the sensitivity of inputs to outputs.

In this talk, we present the automatic code differentiation technique to perform unintrusive sensitivity analysis and derivative computation. This method exploits the concepts of templating and operator overloading in C++ and other similar programming languages to unintrusively convert existing codes into those yielding derivatives of arbitrary order. The idea is demonstrated through the computation of thermal properties in nanoscale heat transfer. Phonon properties such as second and third order force constants, the Gruneisen parameter, group velocities, and the temperature sensitivity of specific heat for graphene nanoribbons are computed. Derivative values so computed are compared with those obtained using finite difference approaches or with analytical values. The method is found to yield derivative values to machine accuracy, with none of the round-off issues associated with finite difference approaches. Use of automatic code differentiation is also demonstrated in topology optimization for fluid and heat transfer applications. Furthermore, templating and operator overloading approaches are demonstrated for polynomial chaos based uncertainty quantification in fluid flow applications.

 

Engineering Faculty Organization—Vision, Mission, and Goal

Virginia Tech faculty members Dushan Boroyevich and Fred Lee

The Engineering Faculty Organization (EFO) includes all members of the engineering teaching, research, and extension faculty.  The EFO holds a general assembly in the Fall of each academic year. The purpose of the EFO is to:

  1. To provide a forum for study, discussion, and formulation of policy in academic and personnel matters of particular interest to the College of Engineering;
  2. To provide meaningful two-way channels of communication between Faculty and Administration;
  3. To provide for the transaction of the routine business of the Faculty of the College of Engineering; and
  4. To discharge all responsibilities that may be delegated to the Engineering Faculty by the Faculty Senate, University Council, and/or Administration.

EFO Executive Committee Officers

Ran Jin, President
Ashley Johnson, Vice President
Lynn Abbott, Secretary

Executive Committee Members

Members by name, department, and term end date.

Gary Seidel (AOE) 2020
David Dillard (BEAM) 2021
Linbing Wang (CEE) 2020
Ashley Johnson (CEM) 2020
Danfeng Daphne Yao (CS) 2020
Sanket A. Deshmukh (CHE) 2020
Julia Ross Dean of Engineering
Lynn Abbott (ECE) 2022

Jeremi London (ENG-ED) 2021
Nathan Lau (ISE) 2021
Tafti Danesh (ME) 2021
Cheng Chen (MINE) 2019 
Kathy Lu (MSE) 2021
Open (CPES) 2021
Open (NCR) 2021
Open (SBES) 2021  

Alternates

Members by name and department.

Open (AOE)
Miguel Perez (BEAM)
Open (CEE)
Stephen Martin (CHE)
Adrian Sandu (CS)
Qiang Li (CPES)

Elena Lind (ECE)
Andrew Katz (Eng Ed)
Ran Jin (ISE)
Jan Helge Bohn (ME)
Emily Sarver (MINE)
Rebecca Cai (MSE)

We'd like your feedback

Do you have suggestions, or comments to bring forward in upcoming Engineering Faculty Organization Executive Committee Meetings? Please complete this survey.

Presentations at 2016 Annual Meeting