Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Trouble modelling fluid flow (air) between two roofs of a biogas plant
Posted 19 set 2019, 04:41 GMT-4 Computational Fluid Dynamics (CFD), Heat Transfer Version 5.4 3 Replies
Please login with a confirmed email address before reporting spam
Hello!
I am currently modelling heat loss in a biogas plant (manure digester) placed on a farm. The digester is cyllindrical in shape, with a double domed roof. Between the two roofs, air is pumped through to assure an outflow of potentially dangerous gases leaking through the lower roof, where the inflowing air has the outside temperature, much lower than the temperatures inside the digester (10 degC for the inflow, 42 degC for the manure). I included a snapshot of the geometry under this post, as well as the model's MPH file. I use a combination of the CFD module (turbulent flow) and the possibilities for Heat Flow (Heat transfer in solids and fluids) in the basic licence (I do not have the Heat Flux module to my disposal). The following are the problems I encounter:
- I am having trouble modelling (or possibly, understanding the result of) the air flow between the roofs. As can be seen in the snapshots included (one in the plane of the inlet and outlet and one perpendicular to that), the airflow directly at the inlet and outlet is correct, but somehow the flow between the two roofs is in the direction of the inlet (negative x-components of velocity field), rather than the outlet, which seems odd.
- Also, it seems like the airflow from the inlet somehow penetrates the inner roof, which is a closed surface of a PVC foil, which by no means is permeable to such an extent.
- Thirdly, it seems like the velocity of the airflow between the roofs has the highest magnitude directly under the upper roof, where due to the no-slip condition imposed, I would expect the highest magnitude in the middle of the distance between the two roofs.
- On top of that, the result does not display a correct circulation of the biogas below the inner roof, or at least one would expect. The result shows a cold 'bubble' forming behind the lower roof near the inlet. This can be seen in the Temperature snapshot.
- Lastly, when integrating the Velocity magnitude over the surface of the inlet and doing the same for the outlet, shows a different value for the two. This should not be correct, because there is no other way for the inflowing air to exit the roof than the outlet on the other side.
Can anyone elaborate on what I might be doing incorrectly, or if the model is correct, where my thinking and expectations are off?
Any help would be greatly appreciated. Thank you in advance!
Attachments:
Please login with a confirmed email address before reporting spam
Hi Maarten,
Without looking at your file in detail, I can see that you are missing an "internal wall" node in the turbulent flow branch to tell the software that the flow cannot go through the inner roof. I didn't look further once I noticed that, so there could be other issues as well, but I'd suggest you start with that modification, re-solve and see how if the solution then makes more sense.
Best,
Jeff
-------------------Jeff Hiller
Please login with a confirmed email address before reporting spam
Hi Jeff,
Thank you very much. That ineed explains why the flow seems to go through the inner roof. I was not aware of the need of imposing such an "internal wall". I will apply it and hopefully that will give a good result. Thanks again!
Best, Maarten
Please login with a confirmed email address before reporting spam
Hi Jeff, or anyone else,
After adding an interior wall, I now encounter convergence problems. The solver initially converges, but after a few iterations it keeps going up and down. This can be seen in the picture I added.
I tried to simplify by using symmetry and seperating the study into two steps, for the Turbulent Flow and the Heat transfer, but this does not solve the problem. I also tried different Turbulence models, but to no avail. Using a Fully Coupled solver instead of the Segregated solver does not help either.
Is there a strategy you can recommend to overcome the problem, or at least localise where the problem might arise?
As I am fairly new to Comsol and only worked with simple models, I did not encounter a similar problem before.
All help would be welcome. Thank you in advance!