Concurrent CFD Explained (Part IV)
Last time I promised I’d drill down into more detail about the difference between the CAD-integrated up-front approach to CFD and CAD-embedded Concurrent CFD. CAD-integrated CFD tools are essentially stand-alone tools launched from within the CAD system whereas Concurrent CFD is fully built into the CAD system. So what’s the real difference between the two?
CAD-integrated CFD tools use the same approach for all CAD systems, so perhaps the best way to answer this is to look at what happens when the CFD tool gets launched from within the CAD system: First the original CAD model gets converted into the format required by the solid modeller built into the CFD tool, e.g. Parasolid. This converted geometry is then imported as parts and assemblies into the CFD program. So far, so good – so what’s the snag?
The geometry that gets transferred from the CAD system into an up-front CFD tool comes over ‘warts and all’. The implications of this are not apparent when dealing with rudimentary demo geometries, since these have very few parts with simple mates. However, real world CAD geometry contains dozens, sometimes hundreds of unwanted (and often very small) voids resulting from modeling simplifications that are common practice in CAD design.
All geometry parts including all new dummy parts that represent actual flow cavities in the model get some default material property assigned, e.g. that of air. The user has to manually select all the parts that should not belong to the computational domain and mark them as unwanted (otherwise they get meshed and treated as flow or solid domains). This includes marking small voids as unwanted to ensure that only the few desired fluid regions (most often only one) remain. Meshing these small voids would otherwise cause very high additional and unnecessary cell count, and, if not properly meshed, extreme solver convergence problems. Finally, the user has to manually assign the correct material properties to all fluids and solids remaining in the model and apply boundary conditions.
This all appears very quick and easy when working on a simple demo case with just a few solid blocks to represent something more complicated like an automobile – if you blink you’ll miss it. For a large number of parts is very laborious, time-consuming, and error-prone, because all solids are listed even if no heat transfer considered. Obviously, the user has to be meticulous in manually working through all the listed parts in the model and assign materials correctly.
Consequently, when using CAD-integrated CFD it is often much more efficient (especially for flow-only tasks) just to create the required flow domain in the CAD system using CAD functions for Boolean operations (“Cavity” function) and then transfer only this one part to the CFD tool for meshing and solving. This is the ‘Create Cavity’ step shown in my ‘Concurrent CFD Explained (Part II)’ post and something that you never have to do in Concurrent CFD.
A second issue arises when the user wants to modify the geometry. Quite rightly, this has to be done in the CAD system. Substantive design changes are common early in design when very little of the design has solidified, and so CFD can bring the most benefit as changes can be made freely in the pursuit of improved performance, reduced cost, etc. Moving geometry around is no problem, but making substantial changes that alter the topology of the model, for example by creating different parts, or building the assembly differently basically means that in CAD-integrated CFD the whole process has to be repeated.
Hopefully that’s given you a broad understanding of the state-of-the-art in up-front CAD-integrated CFD, described as being “fully associative” with the CAD geometry. Concurrent CFD resides fully within the CAD system so does not suffer from these issues. Full CAD embedding gives access to details of the native CAD geometry. This has allowed us to develop proprietary technology that automatically ignores all cavities in a CAD model that are not intended to be a computational domain for meshing and solving. It automatically handles solids and fluids appropriately, consistent with the user’s choice about whether to include heat transfer in the project or not, with no need for the kind of manual user intervention required by the CAD-integrated CFD approach.
One of the latest developments in up-front CFD is the concept of a ‘design study’, allowing different designs to be compared. This emulates the project management system that’s available in all CAD packages – and something that has been standard in Concurrent CFD for many years as Concurrent CFD utilises the CAD system’s capabilities. Another key advantage of Concurrent CFD is that the results can be displayed on the actual CAD geometry to aid understanding. Again, this is hard to achieve but delivers real benefit to the user in interpreting the analysis results.
I hope this provides more detail on the differences between the CAD-integrated upfront CFD and the CAD-embedded Concurrent CFD approaches illustrated in the animation below.
You can also check out the joint MAD/CIMdata presented web seminar entitled ‘CFD Analysis for MCAD Designs – Better Products Faster’. If you want to see Concurrent CFD in action register now for next week’s web seminar ‘Flow, Pressure, Cavitation! Use ‘X-Ray Vision’ to Avert a Design Disaster’.
Dr J, Hampton Court
More Blog Posts
Add Your Comment
About John Parry’s Blog
A mixed bag of things that interest me professionally -CFD technology and its use in education, cooling technologies and the place of thermal design in the overall design flow.
- Lies, Damned Lies, and “CFD Comparison Charts” – Part IV
- Lies, Damned Lies, and “CFD Comparison Charts” – Part III
- Lies, Damned Lies, and “CFD Comparison Charts” – Part II
- Lies, Damned Lies, and “CFD Comparison Charts” – Part I
- Mechanical Analysis Products Now in Mentor’s Higher Education Program
- Sony Vaio laptop in mass ‘recall’
- August 2010 (4)
- July 2010 (2)
- June 2010 (1)
- April 2010 (2)
- March 2010 (4)
- February 2010 (5)
- January 2010 (6)
- December 2009 (4)
- November 2009 (5)
- October 2009 (4)
- September 2009 (3)
- August 2009 (4)
- July 2009 (6)
- June 2009 (3)
- May 2009 (2)