I have solved a laminar flow problem. The Re number is 0.2. The objective is to calculate the pressure loss in the system. But the CFD simulation is giving around 17% errors. Sometimes the errors are positive and sometimes negative. I tried different mesh densities but in vain.

What could be the reason? Has anybody faced such problem?

Hi

Can you give more detail about the system? And what did you use for the length when calculating the Reynolds number. Are you using double presicion? Did you check/adapt the velocity gradients?

Regards

AAA can you tell us what the relation between adapting the velocity gradient and predicting the pressure drop?

Salam salam

In general, having smooth gradients for all parameters (p, v, etc.) by using the adaption tool offers better results.

Initially I calculated the Re number at inlet by hydraulic diameter. Then I checked the Re number at various points calculated by Fluent (from display->contours->velocity->Re).

I have not tried grid adaptation. I have used velocity inlet and pressure outlet. Now I am trying pr-pr BC with target mass flow rate. Let us see.

Let me know if you have any inputs.

Thanks for your comments.

OK can you guide us to how can did this with more details?

Salam salam

Just grab any CFD book and see how pressure and velocity are coupled. The smoothness of the staggered grid dictates how accurate the final solution is. If you have a more specific question, I'll try my best to dig up the answer for you or at least refer you to a reference.

Kind regards

AAA

Hi

Please try adaption and tell me if you obtain better results.

Hi,

I tried the grid adaptation method. But the error has has not gone and there is not much difference in results in this case. Now I am doubting the fluid properties at that temperature.

Let me have your comments.

Hi

When you mentioned a 17% error, are you comparing your results with other people? What are your BCs? Is it an article you are comparing your results with?

If the case is simple 2D, I'll try to simulate it, If you like. Maybe I can find what is wrong.

Regards

Hi AAA,

The 17% error is with experimental results. The case is 3D. I have used the 4 different meshes and grid adaptation to track the error. Flow is laminar Re=0.2. Fluid is very viscuss. mu ~ 1 kg/m-s, rho ~ 0.8.

I have taken the fluid properties from its data sheet. This can be one source of error. Actual properties might be slightly different.

What do you think?

hi, should we try with outflow as BC? & as you say fluid properties can be one of the reasons. Is it assumed the walls to be smooth?