CFD Simulations AC 7 01
Thoughts of a poor old sod
Flummery & Flannery
Application Challenge AC7-01
Overview of CFD Simulations
If CFD simulations of the AC have been performed then provide a brief overview. This should cover the scope of the calculations and the main aspects of the modelling strategy e.g. equations solved, turbulence and other physical models employed). All computational domain simplifications/idealisations, and the treatment of subgrid features should also be identified (e.g. imposed symmetry plane, omission of detailed features, simplification of complex/small scale features, i.e. porous media, use of equivalent wall roughness). If important details of the geometry representation are uncertain then the impact of these uncertainties on the DOAPs should be discussed, including possible ways for managing their effect.
It is left to the discretion of each author to decide the most appropriate way for structuring and summarising the CFD results. Ideally, the data structures used should be consistent with those used for the test data.
A summary table for all CFD simulation results should be included, as shown below in Table CFD-A. Available data should be clearly identified, (e.g. UVW, k, concentration, etc). As with test data, a distinction should be drawn between detailed local data (e.g. p(x,y,z)) and data relating to DOAPs which are likely to be global/summary parameters (e.g. coefficient of lift, CL).
All available detailed data should be stored in separate electronic datafiles (according to guidelines set out by the Knowledge Base team at the University of Surrey). These should be summarised as shown below in CFD-B, with links to each of the datafiles.
|Name||GNDPs||PDPs (problem definition parameters)||MPs (measured parameters)|
|Re||Fr||Wind direction||Source rate (kg/s)||Release density(kg/m3)||Detailed data||DOAPs|
|CFD 1 (dense gas dispersion)||0, 30, 45, 90, 180|
|Re||Wind direction||Building geometry||Detailed data||DOAPs|
|CFD 2 (passive gas releases)||0, 30, 45||A, B, C, D|
|SP1||SP2||SP3||DOAPs, or other miscellaneous data|
|CFD 1||✔✔✔ cfd11.dat||✔||✔ cfd13.dat||cfd14.dat, cfd15.dat|
|CFD 2||✔✔✔ cfd21.dat||✔ cfd22.dat||✔ cfd23.dat||cfd24.dat, cfd25.dat,|
SIMULATION CASE CFD1
A self-contained description of the solution strategy employed. This should include:
- the name of the code (including version number)
- all equations solved, including the turbulence model, and any other physical model used (e.g. chemistry, pollutant transport)
- the numerical discretisation scheme used in both space and time (this may depend on the equation set solved, e.g. momentum equations may be different from the turbulence equations)
- the solution algorithm used
Describe the geometry of the computational domain including the location of all boundaries. Does this domain coincide with the test geometry or does it represent a simplification? Discuss the impact of any simplification on the DOAPs. In particular justify any 2-D idealisation. Describe in detail the mesh (or meshes) used, including the total number of cells/grid points and mesh density distribution.
Describe the numerical boundary conditions at all boundaries including inflows, symmetry surfaces, walls, far-fields, free surfaces and outflows. In each case, comment on how these replicate conditions in the test rig (e.g. inflow conditions based on measured data taken at a rig measurement station?). Have sensitivity tests been carried out to explore the effect of uncertainties in boundary conditions (e.g. turbulence data at inflow, position of far field boundaries, etc.)
Application of Physical Models
Describe the details of how turbulence models, and any other physical models, were applied For example if RANS turbulence models have been used what is the near wall set-up? What is y+ at the first grid point from the wall (y is normal distance to the wall) and how many grid points lie in the wall boundary layer? (the type of wall treatment employed, i.e. wall functions or low Reynolds number turbulence model is described in the Boundary Conditions subsection above).
What steps have been taken to estimate or demonstrate the numerical accuracy of the results (e.g. mesh and time-step refinement studies, increasing the order of accuracy of the scheme)? What measure of iteration convergence was used, and was this achieved?
List data included, with links to all data files. Describe the data storage format used, and specify conventions (define reference coordinate system, and state if data are vertex, cell-centered, or interpolated).
CFD1 Dense gas release simulation: RANS calculations, standard k-ε, coordinate system as per experiments, cell-centred data
cfd11.dat binary file; headers: Re, Fr, wind direction, source rate, release density;
cfd13.dat binary file; headers: Re, Fr, wind direction, source rate, release density;
Any references in the scientific literature that are relevant; if possible use references that can be accessed via the web and add hyperlink.
- Reference 1
- Reference 2 ...
SIMULATION CASE CFD2
(as per CFD 1)
Contributed by: JB Priestley — Yorkshire
© copyright ERCOFTAC 2023