Learning Forum

[mcolom]

Hi,

 

I try to do a simple calculation of a NACA profile, and the lift results have an error higher than 5000%. Same calculations in fluent and compering with wing tunnel results are so similar, but with Discovery fluid the results does not make sense. And I don't know why. Can you help me?

[Subashni Ravichandran]

Hi  Miquel Colom  
Discovery Live is intended more to show a trend of results rather than accurate results. Although, can you please give us the configuration of your graphics card? Meanwhile, you can try solving at maximum fidelity to possibly, see a better result.  Furthur, I would suggest you perform some basic checks, such as consistency of units.

[mcolom]

Subashni NVIDIA Quadro M5000

[Subashni Ravichandran]

Hi Miquel Colom 
Can you share the model so that we can help you better? Also, can you give us more details regarding the boundary conditions and the lift that you are getting?

[mcolom]

Hi Subashni I try to use small domain, big ones, to use or not use ground, different speeds, but the result is always so far away from reality

🛈 This post originally contained file attachments which have been removed in compliance with the updated Ansys">https://www.ansys.com/legal/terms-and-conditions">Ansys Learning Forum Terms & Conditions

[Subashni Ravichandran]

Miquel Colom 
We encountered some issues while solving your model. I will consult the development team and get back to you soon.

[Subashni Ravichandran]

Miquel Colom Firstly, we modified the control volume in your model to a more suitable size and ran a simulation to get very fluctuating results. This is because the lift at a very low angle of attack is very much a function of how the fluid behaves at and along the walls of the airfoil. Discovery Live, currently, does not handle walls as smooth surfaces, when curved. So, we tried again by rotating the airfoil, as shown in the below picture, by 3 degrees and we got better results. It is to be noted that Discovery Live supports only weakly compressible flows (up to Mach 0.3). Also, this tool is intended to give a trend of the result rather than accurate results. For solving Compressible flow problems or get more accuracy in results we recommend using Discovery AIM.

🛈 This post originally contained file attachments which have been removed in compliance with the updated Ansys Learning Forum Terms & Conditions

[mcolom]

Subashni Thank you I run the simulation that you attach, but the lift results are 500% lower with my computer  

[Subashni Ravichandran]

Miquel Colom 
As discussed already, Discovery Live does not handle walls as smooth surfaces, when curved and also graphics capacity plays a major role in the result that you get. I'm using a NVIDIA Quadro P5000 16Gb graphics card to simulate the models as opposed to your NVIDIA Quadro M5000 8Gb.
As mentioned earlier Discovery Live is meant to show you a trend of results rather than accurate results. You can vary the angle of attack and observe the trend in results to be consistent. To get more accuracy in your results, I strongly recommend using AIM.