[Sammar Kazmi]

I have performed 1-way FSI simulation of compressor blade with two methods as per guidelines of ANSYS manual (The System Coupling system launches the System Coupling tab, allowing you to model one-way and two-way Multiphysics couplings for your project)

Method 1: With traditional 1-way FSI method by Importing direct pressure from CFD to Structure Module
Method 2: By 1-Way FSI System Coupling of CFD and Structure Module

Results:
Method 1 (direct pressure): Radial deformation of blade=0.19066mm
Method 2 (system coupling): Radial deformation of blade=0.20936mm
% change = 9%

I want to know what is the major difference between these two approaches? And how to justify the difference of 9% in final results in steady state?

[evcelica]

I'm not sure exactly, but when I used Method 1 in the past, it only imported the normal pressure forces, not any tangential shear forces. You should be able to check if this is the case with your model.
Compare your Force and Moment reactions from your fixed support in Mechanical to the Force and torque values from your fluid simulation in CFD Post, (or perhaps from the solver manager plots,) where in POST you should be able to look at (listing expressions for force in Z only):
1.) areaInt_z(Pressure)@Surface
2.) areaInt(Wall Shear Z)@Surface
3.) force_z()@Surface
Where 1+2 = 3.
Do this for the other two directions as well, and compare to your force results in mechanical. You can do the same thing for torque, though you may have to adjust the origin of the moment calc as well, which will be trickier, but you will probably already have your answer just from comparing the force values.

I'd guess that Method 1 in mechanical will equal expression 1.), and Method 2 in mechanical will equal expression 3.).
Let us know what you find. Good luck.

[ghorrocks]

When you start looking into the details of your results like this you need to make sure you know what is going on (eg: Erik's comment) and that the simulations you are comparing are accurate. It could be that the two approaches give the same results, it is just that your model is inaccurate and that error comes up as an apparent difference.

So I would do a sensitivity study on mesh size and convergence tolerance to check your results are accurate before comparing the details of the results.

[Sammar Kazmi]

Thank you very much Erik and Glenn for your prompt response.

1. As per the suggestions of Erik, we have calculated the forces and torques in all directions in CFD and forces and reaction moment in static structural. (Table is attached)

2. As far as the grid sensitivity analysis is concerned, highlighted by Glenn, we have performed a comprehensive mesh sensitivity analysis before running of this simulation for both CFD & Structural Simulations.

3- One important point to mention here is that similar differences (around 10%) in final results have been reported in peered review published research paper (Elsevier 2023) for 1-way FSI via System Coupling and Direct Pressure imported approach.

But unfortunately, the authors did not explain the reason of this discrepancy in published paper.

That is why we are interested to know the flow physics/formulation/equations behind this conceived difference (10%)?

[evcelica]

Your Force Reactions from mechanical are lower than the CFD by 25%.
Which mechanical model are those force reactions from? (Pressure import or system Coupling?)
Do they both use the same CFD results, or were those different as well?
Are you using the same mesh on all 3 (or 4) analyses?

Can you see the imported load in mechanical? Are those the same or different?
Sorry but it's tough to do troubleshooting on our end without the model, and these can be very tricky.