[mechengrkj]

I have a fairly complicated conjugate heat transfer model with fins on the outer surface. After setting up and running the model, I was asked by a colleague to provide the total heat loss from each fin individually. Is there a way to create a report of the Total Boundary Heat Transfer for individual faces/facets contained in a boundary AFTER having created the regions and boundaries? The model takes a while to run so I'd like to do this without having to rerun everything. Hopefully that makes sense. Thanks!

[me3840]

If you are clever you can make a series of thresholds with only the boundary faces for each fin in them, then run reports on them all.

[bluebase]

Hi Mechengrkj,

another solution would be to split the respective fins boundary by angle, such as 89degree ( context menu of a boundary object -> split by angle).
If your fins are fairly simple, you would receive a large number of new boundaries - one for each separatable face. You then would identify, group boundaries for each fin by selecting the respective new boundaries and recombine them ( context menu of a boundary object -> combine). This step might get laborious.
In the end, you would have restructered your region's boundaries. Possibly you would need to run another (time) step. But then, you should be able to get the desired results without a major rerun.
Be aware, that this modification is not persistent! If you remesh, those changes will be lost - unless you perform the same modifications to the source geometry, too.

Best regards,
Sebastian

[mechengrkj]

Originally Posted by bluebase

Hi Mechengrkj,

another solution would be to split the respective fins boundary by angle, such as 89degree ( context menu of a boundary object -> split by angle).
If your fins are fairly simple, you would receive a large number of new boundaries - one for each separatable face. You then would identify, group boundaries for each fin by selecting the respective new boundaries and recombine them ( context menu of a boundary object -> combine). This step might get laborious.
In the end, you would have restructered your region's boundaries. Possibly you would need to run another (time) step. But then, you should be able to get the desired results without a major rerun.
Be aware, that this modification is not persistent! If you remesh, those changes will be lost - unless you perform the same modifications to the source geometry, too.

Best regards,
Sebastian

Thanks for the response! I'll give this a shot.

[mechengrkj]

Originally Posted by bluebase

Hi Mechengrkj,

another solution would be to split the respective fins boundary by angle, such as 89degree ( context menu of a boundary object -> split by angle).
If your fins are fairly simple, you would receive a large number of new boundaries - one for each separatable face. You then would identify, group boundaries for each fin by selecting the respective new boundaries and recombine them ( context menu of a boundary object -> combine). This step might get laborious.
In the end, you would have restructered your region's boundaries. Possibly you would need to run another (time) step. But then, you should be able to get the desired results without a major rerun.
Be aware, that this modification is not persistent! If you remesh, those changes will be lost - unless you perform the same modifications to the source geometry, too.

Best regards,
Sebastian

This doesn't seem to work. It splits the faces but doesn't provide any data on the boundary without rerunning. I even tried running a few iterations to see if it just needed to update but no luck. Any other ideas?

[me3840]

Using thresholds should not be tedious at all if you're generating them with a quick Java macro, which is my recommendation.

[bluebase]

Have you updated your interfaces, too?

[mechengrkj]

Originally Posted by bluebase

Have you updated your interfaces, too?

Yes I made sure all the interfaces were updated.

[cwl]

In general you can create only one fin in 3D-CAD module, then Linear Pattern them in Parts, then Unite, Subtract or Extract Volume - and thus you'll get separate surface for each fin which can be used in Histogram Plot to determine the most effective fins, - but that would require rebuilding the whole project.

But again - all that in general looks a bit pointless because numbering the fins and finding out which is which is confusing and loses sense if you're running simulations with varying amount of fins.

[mechengrkj]

Originally Posted by me3840

If you are clever you can make a series of thresholds with only the boundary faces for each fin in them, then run reports on them all.

I was hoping to avoid this because given the geometry it would be very tedious.