CFD Online Logo CFD Online URL
Home > Jobs > Job Record #18391

CFD Jobs Database - Job Record #18391

Job Record #18391
TitleAdjoint-based hp-adaptation for high-order CFD simulations
CategoryPostDoc Position
EmployerIndian Institute of Science (IISc) Bangalore
LocationIndia, Bangalore
InternationalYes, international applications are welcome
Closure DateSunday, October 01, 2023
     Postdoc Position: Adjoint-based mesh and order (h/p) adaptation for high-
order CFD simulations   

Applications are invited for a Postdoctoral Researcher through the prestigious C V 
Raman fellowship or the Institute of Eminence (IOE) fellowship (depending on the 
qualifications) to join the CFD research group of Prof. Aravind Balan at the 
Department of Aerospace Engineering, Indian Institute of Science (IISc) Bangalore.

Project description: 
Higher order numerical methods that use local polynomials for solution 
representation such as Discontinuous Galerkin (DG), Flux Reconstruction etc are 
becoming increasingly popular in solving convection dominated flows due to their 
potential in giving accurate results more efficiently than lower order methods on 
unstructured meshes.  With these methods, to efficiently resolve the flow 
features, hp-adaptation can be done where the mesh size and the element-wise local 
polynomial degree get adapted based on error estimators in an iterative manner. In 
many situations, one may not be interested in the whole flow field, but only on 
certain output functionals such as lift, drag coefficients, etc. In such cases, 
using adjoint-based error estimators is known to be an effective approach. For 
compressible flows having many anisotropic features such as shocks and boundary 
layers, use of anisotropic (highly stretched) mesh elements that align with the 
flow features are much more efficient than using isotropic meshes. Metric-field 
based mesh adaptation provides a suitable framework for mesh adaptation since 
metric fields can embed both the size and the anisotropy (stretching ratio) of the 
mesh elements, and optimal metric fields can be obtained in an analytical manner 
[1]. The existing algorithms for adjoint and metric field-based adaptations for 
high order methods have shown excellent results, but are developed only for 
triangular/tetrahedral elements [2]. The proposed postdoctoral work is the 
extension of the adaptation algorithms to quads in 2D and hexahedra in 3D, and 
finally to include mixed element unstructured meshes. This includes a 
mathematically rigorous derivation of the metric fields, their implementation on a 
high order DG code, and numerical verification using various test cases from 
different flow regimes (subsonic, transonic and supersonic). The project also 
proposes extension of an inviscid adjoint-based adaptation algorithm to viscous 
RANS simulations. It is expected that, with a robust mesh adaptation technology, 
CFD works flows can be fully automated resulting in a much faster design process 
of the aerospace vehicles.

Applicants must have Ph.D in Mechanical/Aerospace Engineering or Applied 
Mathematics. Experience in code development for numerical simulation of fluid 
flows is mandatory.  

Fellowship amount:
C V Raman fellowship: Consolidated monthly package of Rs. 1 Lakh (inclusive of 
House Rent Allowance). In addition, they will receive an attractive research grant 
of Rs. 8 Lakhs per year.
IOE Fellowship: Selected candidates will receive Rs. 55,000+ HRA. In addition, 
they will receive a research grant of Rs. 2 Lakhs per year. 

Duration: The duration of the fellowship is one year, extendible to one more year. 

For any other information, please contact Prof. Aravind Balan at To apply, send an email with a brief expression of 
interest in this topic and the latest CV with the publication record, list of 
referees for letters of recommendation, and university transcripts. 

1. Aravind Balan, Michael Woopen, and Georg May. Adjoint-based Hp-Adaptation on 
Anisotropic Meshes for High-Order Compressible Flow Simulations. Computers and 
Fluids, 2016. Doi:

2. Aravind Balan, Michael A Park, Stephen L Wood, W Kyle Anderson, Ajay 
Rangarajan, Georg May, and Devina Sanjaya. A review and comparison of error 
estimators for anisotropic mesh adaptation for flow simulations. Computers and 
Fluids, 2022, Doi:

Contact Information:
Please mention the CFD Jobs Database, record #18391 when responding to this ad.
NameAravind Balan
Email ApplicationYes
Record Data:
Last Modified07:53:13, Tuesday, June 27, 2023

[Tell a Friend About this Job Advertisement]

Go to top Go to top