Scribd
Upload
14 views1 page

CFD Shape Optimization via Adjoint Method

The document discusses CFD shape optimization using adjoint methodology. It notes that an optimization problem in CFD is defined by objective functions like minimizing force or maximizing lift to drag ratio, with design variables such as angle of attack. Shape optimization methods can be classified as CAD-based, surface mesh-based, or volume mesh-based. Each optimization cycle has four steps: solving the CFD equations, obtaining adjoint variables, computing gradients, and updating the shape. Continuous and discrete adjoint techniques can be used and have advantages like computing gradients efficiently. FEM performs better than FVM for diffusion-dominated problems while FVM is better for advection-dominated problems.

Uploaded by

suhannshetty
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
14 views1 page

CFD Shape Optimization via Adjoint Method

The document discusses CFD shape optimization using adjoint methodology. It notes that an optimization problem in CFD is defined by objective functions like minimizing force or maximizing lift to drag ratio, with design variables such as angle of attack. Shape optimization methods can be classified as CAD-based, surface mesh-based, or volume mesh-based. Each optimization cycle has four steps: solving the CFD equations, obtaining adjoint variables, computing gradients, and updating the shape. Continuous and discrete adjoint techniques can be used and have advantages like computing gradients efficiently. FEM performs better than FVM for diffusion-dominated problems while FVM is better for advection-dominated problems.

Uploaded by

suhannshetty
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Date: 17/10/14

CFD SHAPE OPTIMIZATION USING


ADJOINT METHODOLOGY
-By Dr. Srinivas Chippada
An optimization problem in CFD is defined in terms of objective functions, such as minimize
force, minimize pressure drop, maximize lift to drag ratio, etc. It will have design variables such as
angle of attack, pressure at an inlet,etc. It will also consist constraints that need to be satisfied by the
optimized solution.
Shape optimization methods could be classified as CAD based optimization, surface mesh
based optimization, and volume mesh based optimization. The surface mesh based shape optimization
involves design variables of order 10^4 and volume mesh based optimization involves design variables
of order 10^6.
Each optimization cycle has four steps : CFD step, adjoint step, gradient step, and shape update
step. In the CFD step flow governing equations are solved. In the next step, adjoint step, the adjoint
variables are obtained. In the gradient step gradients are obtained. The shape update step involves
smoothening of external surface based on shape sensitivities, re meshing of the interior volume , and
step size selection to ensure objective function reduction and mesh quality preservation.
Continuous adjoint technique and discrete adjoint techniques could be used in the adjoint step
of optimization process. These adjoint techniques have several advantageous. They are very easy to
implement and they can compute gradients in a cost effective manner.
FEM and Finite Volume Method(FVM) have several differences. FEM does a better job than
FVM in diffusion dominated physics. In such cases, FEM gives results faster and with better accuracy.
FVM does a better job than FEM in advection dominates physics.

Suhan N Shetty
SR NO 11256
Mechanical Engineering
ME 2014-16

You might also like