[corriolisforce]

Hi,
Is solidworks flow simulation a good software for CFD analysis? For Centrifugal pumps for example. I was just wondering if the results that would be obtained would be close to reality. Would CFDesign provide a better analysis?

Thank you,

Carlos

[Donovan]

SolidWorks is best-designed software. This software helps you to get your innovative products into the production faster. 2019 SolidWorks Free Download – Is There a Free Full Version is used by worldwide designers and engineers to bring designs to life.

[Gert-Jan]

You have to find out what method the software uses for its CFD analysis. If it uses Finite Volume Method, go ahead, Finite Element Method is a showstopper, especially for cenrifugal pumps.

[hokhay]

As far as i know, the CFD solver that is used in SolidWorks is from FloEFD. The advantage of this software is the ease of use but not accuracy. I used FloEFD for car drag and lift calculations, the results have 20-30% error. The reason is because it cannot make boundary mesh like other proper CFD software does, which affect the flow separation and hence the surface pressure and force. It also only has just one turbulence model to choose from, which is k-epsilon, this mean some kind of simulations may not be suitable for this.

For your case, if you just want to calculate the flow rate of the pump, the result should be fine, as this does not heavily affected by the wall boundary flow. If you want to calculate the force or torque, you would expect a noticeable amount of error.

[Time4Tea]

I've used SolidWorks Flow Simulation a fair bit - although mostly I use FLUENT. I've found SW Flow Sim to be 'ok', but I guess it depends on what you want/expect.


It does have advantages in that it is built into SolidWorks and eases some of the CAD cleanup and conversion that you would have to do with FLUENT. Also, it is more accessible to those that already have SolidWorks and is probably less costly to purchase a license.


However, in terms of drawbacks, the meshing is certainly not as good as FLUENT/ANSYS. I believe it is limited to a 'cut-cell' type of approach, which uses hex cells that are all aligned with the base cartesian coordinate system of the model. The main issue is that there is (as far as I am aware) no way to generate nice inflation layers on the wall boundaries. This can be an issue for accuracy, as the way the boundary layer is modeled can be very important for many analyses.


Then, on the solver side, compared with FLUENT, there is noticeably less visibility and control of what is going on 'under the hood'. Fewer knobs and buttons you can tweak. This is a double-edged sword: on one hand, it may be easier for people who are newer to CFD and/or don't have a deep background in Fluid Dynamics to learn and get to grips with. However, without the fine control that a more dedicated code like FLUENT gives you, it may be harder to achieve accurate results in some cases, which might frustrate more advanced users. For example, I believe there are fewer options in Flow Sim for types of solvers, turbulence models, discretization schemes, etc.



So, in summary I would say it's more accessible for less experienced users and is probably ok for getting some quick, rough numbers or comparing between different designs. But, I would be cautious about using it for an application where the CFD is critical for the design and where a good comparison to test data is needed or expected. At least, without doing a good amount of work up-front to validate it's results against test data.