I am currently conducting a research study using Simulation CFD to perform a thermal heating analysis. The context of this simulation is the construction of a simple 6mx6m square house in Autodesk Revit, using generic models, which has then been loaded into Simulation CFD for a solar heating thermal analysis. I hope to understand the heating effects of the internal brick partition walls, with the sun heating the building through a north facing window and a thermally insulated ceiling preventing downward heating through the roof.
At present, the Model solves well, however I have noticed that the temperature only increases, despite specifying a transient analysis, a location, and inputting data as a boundary condition (which is a piecewise linear transient set of data of external air recorded over a period of 7 days). I have tried both temperature boundary conditions and radiation boundary conditions (both transient with the piecewise data inputted into the temperature section of the form), however I cannot see any noticable difference.
I need to obtain results that I can use to see the temperature of the internal air fluctuation over the 7 day period. I intend to then modify the material used for the internal partition (L-shaped) walls, and repeat the process, and determine which material, concrete, brick or timber frame is best for reducing the variation in internal temperature.
Please see the .cfz file at my dropbox link https://www.dropbox.com/s/bc5jz2ltqfozgmi/Insulate
This is very new research, and I have a fast approaching deadline so any assistance would be immensly appreciated.
Also, regarding the recording of results, I understand at present that the best way to obtain the average internal air temperature of my model is to set up the solver to save results each hour (presently tried 3 hours to save solving time), then open parts from the results menu, and request a calculation of the temperature results for the internal air volume. I will then put this average temperature into a spreadsheet application, which will then be used to graph the results.
Could I please ask if there is an easier way to graph the results for one component during the simulation?
Thank you in advance,
I think there is a misunderstnading here of what the radiation Boundary Condition (BC) is doing.
Take a look at the description here: http://help.autodesk.com/view/SCDSE/2014/ENU/?url=/cloudhelp/2014/ENU/SimCFD/files/GUID-D39DB8C7-05C5
It is typicaly used in electronic cooling models where we need to add radiative effects of nearby parts without modelling them. There is no need to use it here. Typically we would just apply a temperature to the outer walls of the air domain.
For accurate heat transfer I suggest you refine the solid mesh - we need a mid-node throughout all solid parts (2 elements through their thickness).
You also have the ground floating - I suggest you model the air domain within CAD rather than in CFD as then it can touch the ground. I would also suggest you make it the same width as the ground plane.
Once this is in place, you may also wish to add a film coefficient on the underside of the ground - as heat would be conducted out through here also.
I hope this helps!
I really appreciate your assistance in this matter.
Could I please ask what the process is to refine the mesh to create intermediate nodes at the middle of a volume, to develop multiple elements as you suggest?
Thank you in advance.
I would do this for each solid separately unless they have the same thickness (and therfore require the same mesh size).
Move to meshing
Select a volume
Press 'uniform' which chooses the smallest element size currently in the part and applies it to the entire solid.
Pull the slider to the left until you can see a mid-node through the thickness (use an edge to see this)
Press Apply to make is permenant
Press Spread Changes to spread this out into the adjacent volumes
Does that make sense?
Thanks for your advice Jon, ill test out the multiple element system and see if it works. Is that really all that would have been making my results only increase? My results should be coming out in an oscillating curve, a little like a sine curve, as the building heats up during the day and cools during the evening.
I should also note that I'm now using a temperature boundary condition (transient piecewise data set).
At this late stage, what would be the implications of leaving the ground and air environment geometry as they are rather than rebuilding them in Revit?
Grateful for your advice,
Honestly I am not sure - you would need to ensure that you proerly captured that small air-gap with a really good mesh. I would imagine the runtime for this would be longer than making the change in Revit. Plus you might not be properly capturing the physics, who knows what air temps you would have travelling beneath the ground - not entirely realistic.
Also then, if you run into issues with running this, we would suggest to close that gap - do it in advance and we rule it out as a future problem.
Again, thank you for your assistance, it is most valuable.
I have attempted to construct the model using my CAD program (Autodesk Revit), however, upon transferring the model from Revit into Simulation CFD, i need to be able to create a surface on the northern (positive y axis) side to fill the window space, and then make the inside of the building a separate air volume. I can successfully make the window surface using the void fill tool, however it appears that bringing in an external environment from Revit leads to the recognition of the inside of the building as a volume already, and upon creating the surface in the northern most end of the building, I can no longer apply a material (air) to the inside of the building...as though there were a vacuum inside the building. Please see my dropbox .cfz file which illustrates this occurance here.
Note that my interest here is to have an internal air volume and an external air volume, separated by a completely transparent surface (hypothetical window)
Secondly, I have attempted to develop the mid point nodes and view multiple elements through the thickness of the model, however all dragging the slider to the left appears to do is increase the nodes on the edge of the model components. Please see the .cfz file for this illustration here. https://www.dropbox.com/s/zqgkane5lw9245a/Insulate
(note that this is the file for the ammended model, with temperature boundary conditions and correct settings, however the ground still 'floats'.
I really appreciate any and all the assistance you can provide in these matters,
I suggest you model the window in Revit - just leave the internal hollow and we will create an air volume for you when read in from Revit.
The meshing you are seeing is as expected - we only show the preview on the edges - although you can right click on the part and preview the mesh to see it on a surface.
Once you run the model the mesh will be created and then you can view it through the volumes on a cut-plane.