[pcg]

The position has been filled; thank you to all applicants.

Introduction

The SU2 Foundation is an educational and scientific not-for-profit entity that brings computational scientists and engineers together through the SU2 platform. The SU2 Foundation develops, maintains, and supports a collection of C++ and Python-based open-source software tools for performing Partial Differential Equation (PDE) analysis and solving PDE-constrained optimization problems.

To this end, we have identified key areas of the SU2 code for development and improvement to ensure that the cutting-edge optimization capabilities of the code are backed by equally capable analysis capabilities. Therefore, we strive to maintain high standards regarding the accuracy and robustness of the code on benchmark cases that are relevant to various industries. In order to meet the objective, the Foundation has decided to fund a series of projects targeting the improvement of the computational performance of the tool.

The first group of projects focuses on external aerodynamic applications; their goal is to improve the accuracy and robustness of the SU2 code on NASA’s Drag and High-Lift Prediction Workshop cases (DPW and HLPW, respectively). The initial project is structured as follows.

Project Description

The main objective of the first project is to assess the current state of the code, in particular:
Identify whether the main challenge is the robustness of the iterative solver or the accuracy of the results obtained at convergence.
Identify what terms (viscous fluxes, convective fluxes, source terms) are responsible for either challenge.
The findings from this project will inform the next steps that will be funded by the SU2 Foundation.

The suggested methodology is to first verify the code (assess correctness/accuracy) on simpler cases (for example the 2D DSMA661 case from HLPW-3 and 3D CRM-HL-WB test case from HLPW-5) and then attempt the configuration build-up cases from HLPW-5 to assess robustness.
Identification of the numerical terms responsible for accuracy issues can be done by comparing viscous and pressure force components with other codes. For robustness, careful instrumentation of the code can be used to determine sources of instability, or numerical experiments that freeze suspected terms (locally or globally) to assess their impact on convergence (regardless of accuracy). Candidates can propose other methodologies.

The deliverables for this project are:
Publish an additional V&V case in SU2’s V&V collection (https://su2code.github.io/vandv/home/) based on the verification part of this project.
Produce working SU2 configuration files for the 2D DSMA661 case and the CRM-HL-WB case (if accuracy is acceptable, this should be the case added to the V&V collection).
Produce a report for the investigation on the accuracy and robustness of the different terms.

The SU2 Foundation estimates a commitment of 160 hours for this project remunerated with €2600. Access to computational resources for large calculations and support from the Foundation members throughout the project will be provided. The project should be completed in 4 to 8 weeks (full- or part-time). Remote work is expected.

Candidate Profile

Candidates to execute the project should have experience with scientific software development, preferably with SU2, namely
Be familiar with open-source contribution processes,
Know the fundamental steps involved in a CFD solver,
Be able to refer to the areas of SU2 where they happen,
Be able to cite some literature relevant to the project,
Know relevant verification and validation cases for the project.

To apply, please send a short motivation letter (half page) and a short CV (two pages maximum) to info@su2foundation.org (start the title with “Project 1.1 Application”). Suitable candidates will be invited for an interview.

[airwarriorg91]

Hello !
Thank you for the sharing the opportunity ! I wanted to know the tentative dates for the start date of work as well is there any deadline for the application. I have been using SU2 for external aerodynamics specifically RANS of transition flow over airfoils and bluff body along with other codes like NEK5000 and xcompact3D for the DNS. I am interested for the project.

Thanks,
Gaurav.

[bigfootedrockmidget]

The starting date is flexible, depending on the candidate availability. Preferably we would like the candidate to start in the next 2-3 months and work on and finalize the project in a period of 4-8 weeks.

[airwarriorg91]

Sure, thank you for the information ! I will share my application through email by the end of this week.