Adjoint CFD code development
This paper addresses the concerns of CFD code developers who are facing the task of creating a discrete adjoint CFD code for design optimisation. It discusses how the development of such a code can be greatly eased through the selective use of Automatic Differentiation, and how each stage of the software development process can be subjected to a sequence of checks to ensure the correctness of the final software.
The complete source code can be download from the webpage. This includes a Makefile which uses TAPENADE to generate the linear (forward mode) and adjoint (reverse mode) subroutines.
Related AD-Tool:
TAPENADE
Reference:
People involved:
Reference:
M. B. Giles, D. Ghate, M. C. Duta
Using automatic differentiation for adjoint CFD code development
Recent Trends in Aerospace Design and Optimization, Tata McGraw-Hill, New Delhi, 2006
Using automatic differentiation for adjoint CFD code development
Recent Trends in Aerospace Design and Optimization, Tata McGraw-Hill, New Delhi, 2006
People involved:
- Mike Giles
,
Oxford University - Devendra Ghate
,
Oxford University - Mihai Duta
,
Oxford University

