[NachtErfinder]

Hello all,

I am new to CFD, so bear with me.

I am simulating ejecting a contaminant (a gas) at a certain concentration into the ambient at a high velocity. The purpose of the analysis is to determine the minimum distance at which the maximum concentration of the contaminant is a certain value.

I have run this simulation with a small ambient (to reduce computation time), and it ran perfectly. The results were in line with what has been empirically observed. The only problem is that the results needed can't be obtained from a small simulation.

I have greatly increased the size of the modeled ambient, and applied exactly the same BC and IC. The results I'm getting are not physically possible. The inlet has a contaminant concentration of .01. In the transient simulation, after a few seconds of the simulation have been run, the max concentration in the ambient jumps to .75, the average being about .25, which clearly isn't possible. The simulation doesn't fail due to divergence, however.

Are there guidelines of how to work with large models? It is odd that I applied the exact same BC and IC and got completely different results with a similar shape of CAD model. Any help would be greatly appreciated. If you need more details about the simulation, I will gladly give them.

[FMDenaro]

Originally Posted by NachtErfinder

Hello all,

I am new to CFD, so bear with me.

I am simulating ejecting a contaminant (a gas) at a certain concentration into the ambient at a high velocity. The purpose of the analysis is to determine the minimum distance at which the maximum concentration of the contaminant is a certain value.

I have run this simulation with a small ambient (to reduce computation time), and it ran perfectly. The results were in line with what has been empirically observed. The only problem is that the results needed can't be obtained from a small simulation.

I have greatly increased the size of the modeled ambient, and applied exactly the same BC and IC. The results I'm getting are not physically possible. The inlet has a contaminant concentration of .01. In the transient simulation, after a few seconds of the simulation have been run, the max concentration in the ambient jumps to .75, the average being about .25, which clearly isn't possible. The simulation doesn't fail due to divergence, however.

Are there guidelines of how to work with large models? It is odd that I applied the exact same BC and IC and got completely different results with a similar shape of CAD model. Any help would be greatly appreciated. If you need more details about the simulation, I will gladly give them.

If you increases the lenght scales of your simulation, the Reynolds number increases accordingly and the physics is very different from the low-scale simulation.

You should consider the difficulties encountering simulation of turbulent flows. I suggest to check what you can measure from the experiments and the Re number before to decide the type of formulation you can use (RANS/URANS, LES, DNS).