Simulation of Flow Around a Surface-Mounted Square-Section Cylinder of Aspect Ratio 4

Purpose: This challenge aims to motivate an open discussion on the state-of-the-art in CFD simulation and strategies for analysis of complex, separated flows at moderate Reynolds numbers. An experimental benchmark will provide a context for presenting results. This challenge will focus on identifying the strengths and the cost/benefit of the simulation/analysis approaches.

Simulations based on URANS, LES, DNS, hybrid or discontinuous Galerkin methods are welcome. Discussions addressing simulation challenges, post-processing strategies or feature extraction techniques are highly desirable.

Presentation and Workshop: Participants to the challenge are expected to contribute a full length paper and oral presentation during the general meeting. A special workshop is planned following the general meeting on Saturday, May 12, 2012. Participants are asked to share simulation details, challenges and/or technique advances in a 10 minute presentation followed by a question period. A brief comparative overview of submitted results will be provided by the organisers leading to an open forum discussion. Negotiations are presently being conducted for dissemination of the Challenge results in a journal special issue.

More information is offered in the 2012 Challenge Preparation Guidelines.

Overview of Experimental: Experiments were conducted in an open-section (zero streamwise pressure gradient) wind tunnel. The obstacle was a square-section cylinder of side dimension d = 12.7mm and height h = 50.4mm (aspect ratio h/d = 4). The obstacle was mounted vertically at a distance of 4h from the leading edge of the plate. The unperturbed boundary layer (obstacle removed) at 4h was developing turbulent with a height of approximately 0.18h. The free stream velocity was U = 15m/s, the free stream turbulence intensity approximately 0.8% and the obstacle Reynolds number was Re ≈ 11,000 based on d and U. The flow is characterised by periodic shedding vortices corresponding to a Strouhal number, based on d and U, of St = 0.010 ± 0.003.

Velocity measurements were made with high frame-rate PIV (500Hz or 1000Hz) and hot-wire anemometry. Reference pressure measurements on the side of the obstacle were used for the purposes of phase-averaging. Results are complemented by surface oil-film visualisations. Initial measurements indicated that St and the macroscopic flow features were insensitive to the Reynolds number in the range 6,000 < Re < 50,000.

Experimental measurements are available at the 2012 Challenge Data page.

Timeline:

July 15, 2011: Base flow information will be made available through the society website to allow initial problem set-up. The information provided will include:
• Nomenclature and geometry definition
• Unperturbed boundary layer profiles
• Time (Reynolds) averaged velocity field the plane of symmetry, three horizontal planes parallel to plate and one plane normal to the flow

Sept. 30, 2011: Participants are asked to register their intent to participate at the challenge. This registration will assist the organising committee in coordinating and preparing the workshop. Initial participant input in the preparation of the workshop will be requested.

The registration form is available at the 2012 Challenge Registration page.

March 1, 2012: Phase average data in selected planes will be provided. Format and selected comparison formats and data selection will be provided.

April 1, 2012: Participants will be asked to provide sample results in prescribed format to allow the organising committee to prepare a comparative overview. At this time, participant suggestions for the organisation of the workshop are strongly encouraged.

Challenge Organising Committee:

Carlos Lange (carlos.lange@ualberta.ca)
David E. Rival (derival@ucalgary.ca)
Robert J. Martinuzzi (rmartinu@ucalgary.ca)