I'm sketching on a new software section in CFD Online. In relation to this I'm thinking about different ways of classifying CFD related software. At least two different kinds of classifications are needed - one concerning functionality and one concerning application areas (see below for further details). Can you think of any more important classification criteria that is important? Did I miss any functionality classes or application areas? I appreciate any comments. You can post here or email me directly.

Functionality Classes:

• <LI>Design <LI>Grid Generation <LI>Integrated CFD <LI>Software Library <LI>System Analysis <LI>Visualization

What did I miss?

Application Areas:

• <LI>Aerospace <LI>Airfoils <LI>Astrophysics <LI>Automotive <LI>Biomedical <LI>Casting <LI>Chemical Processes <LI>Education <LI>Electronics Cooling <LI>Environmental <LI>Fire and Safety <LI>Geophysics <LI>Maritime <LI>Micro Electro Mechanical Systems <LI>Piping <LI>Power Generation <LI>Research <LI>Semiconductor Processing <LI>Turbomachinery <LI>Ventilation

What did I miss?

Hello Jonas,

I don't know if you should extend the application areas so as to include industrial processing, such as glass industry in which combustion and heat transfer involved, plastics industry, heat transfer and non-Newtonian flow involved.

Please consider the future of CFD simulation for processing of glass and plastics.

Fred Kang

Hi Jonas! How about adding solver speed, solver type etc? Or is this included in software library?

Regards, Anders

Hi Anders! Yes, I thought about adding specific fields for things like grid (structured, unstructured, ...), solver technology (coupled, uncoupled, explict, implict, ...) etc.

However, I think this will become too specific. The classification critera should be applicable to most kinds of CFD software - grid generators, solvers, post-processors, system analysis tools, integrated packages, ... - how do you specify which solver technology a grid generator uses (sic!) ? Of course I could make this a "CFD Software Yahoo" - with sub-classes and sub-sub-classes ... This is not my intention though and it would also demand too much manual work to create all cross-links etc. that would be necessary.

The purpose of the "classification" I'm looking at now is to help people locate the kind of software they are looking for. Then further information will be available in a detailed description field for each code. This general field can be used more freely and for CFD solvers it should naturally give information about solver technology, grid etc. This description field can also be searched for keywords.

Aside from the two classifications I mentioned previously, that is:

functionality - solver, grid generator / pre-proc., post-processor, integrated CFD, software library ...

applications - aerospace, automotive, biomedical, ...

there will also be fields for obvious things like:

availability - pulic domain, shareware, commercial, ...

source code - yes, no

origin - company name, organization, individual

Hi Jonas,

Please could you tell me first what you understand by "Integrated CFD" and "System Analysis" ?

Serge

Integrated CFD = A whole package including grid generator, solver and post-processing...

System analysis = simplified (not full CFD, although some parts can include CFD) anaylsis of complex systems (for example large piping systems, jet-engine cycle analysis etc.)

Dear Jonas

I am a graduate student at NC State (working on Numerical Simulation of Plasma Flow) and a big fan of CFD. I didn't find either plasma flow (electro-gas dynamics) nor computational reactor thermalhydraulics in your application areas. Incidentally, the the term CFD is not very popular in the nuclear engineering.

Jacob

Ventilation -- The broader industrial sector would be HVAC (heating, ventilation and air conditioning).

I forgot several important application areas. Thanks to everyone who pointed this out. This is a new and updated list. If you still miss any CFD application areas please let me know:

• <LI>Aerospace<LI>Architecture and Wind-Loading<LI>Astrophysics<LI>Automotive<LI>Biomedica l<LI>Casting and Metal Processing<LI>Coating, Painting and Printing<LI>Education<LI>Electronics Cooling<LI>Environmental<LI>Fire and Safety<LI>Geophysical<LI>Heating, Ventilation and Airconditioning<LI>Maritime and Ship Hydrodynamics<LI>Micro Electro Mechanical Systems<LI>Nuclear Reactor Thermohydralics<LI>Piping<LI>Plasma Physics<LI>Power Generation<LI>Process Industry, Chemical Reactions<LI>Process Industry, Other<LI>Process Industry, Polymers and Glass<LI>Research<LI>River and Coastal Engineering<LI>Semiconductor Processing<LI>Turbomachinery<LI>Virtual Reality and Entertainment

and

Misc

I would suggest an additional class called "physical models": viscous, compressible, chemically-reacting gas, unsteady, potential flow, boundary layer, etc.

An alternative to the above could be "mathematical models": steady or time-dependent, Laplace, Full-potential, Euler, TLNS, RANS, LES, DNS, etc.

Presumably, these give information on the potential applications of a code.

These may have already been included in the software library or integrated CFD, but it is not obvious for me.