I’m sorry, I’m new with Prepomax/Calculix, I don’t understand where to set your command line
I modified the source code to add the integral value (classical trapezoidal formula)
I get a close value, but I’m pretty sure the calculation is not performed with the normal flux cause I don’t calculate the dot value, I can’t see how to retrieve the flux vectors cause GetNodalValue give the scalar value and not an array…
Moreover, the Integral must have a correct unit which is not defined in Prepomax
Is there a manual for developpers ?
You can use the built-in Keyword Editor to add new keywords like this:
*SECTION PRINT, SURFACE=surface_name, NAME=section_print_name
FLUX
For such advanced postprocessing, it might be easier to use ParaView where you have multiple filters to perform calculations on the results, including e.g. Python expressions.
- yes freecad/paraview maybe a good alternative
- I’m definitely not an expert of calculix, I tried to add a built-in keyboard attached to a surface section (Edge_T1), but I don’t know where to add that keyword in the calculix script, moreover I can’t see any “print” in the monitor dialog
radiation3.pmx (1.5 MB)
FreeCAD won’t help much here but ParaView is very good for advanced postprocessing.
You defined it correctly. But since this option is not yet supported in PrePoMax, you can find the result only in the .dat file written by the solver in the same location as the .frd file with contour plots. The path is shown in the Result list (top left part of the toolbar).
1 Like
Ok, thanks @FEAnalyst
-
I can see the file “Analysis-1.dat” containing all the fluxes at the integration points and also at the end the total heat flux on the surface_name. Is it possible to explicitly name the file as “Edge-T1.dat” for instance ?
-
The theoretical heat flux is 11.10^3 W/m, but I get a total heat flux = 11.10^7, I don’t know the unit, I guess it’s mW/m ? How to know ?
Thks.
PS : I need to read the calculix manual… 
The .dat file generated by CalculiX always has the same name as the input file name which in turn is named after the analysis name in PrePoMax. The same .dat file stores all requested tabular outputs.
CalculiX doesn’t state the units anywhere and they depend on the unit system used for inputs. Here it’s SI(mm) so heat flux should be in mW.
1 Like
make sense !
The unit is SI(mm) so the heat flux is mW/(mm²) and the integral mW/mm which is same as W/m, am I wrong ?
No, there is no manual for developers.
PrePoMax uses only consistent unit systems. So all input values and all result values are written in consistent units. In all files. But in a 2D simulation, I think, the heat flux is also computed on the surface, so it should have a unit of Power/Area. It is not a Power/Lenght unit.
Section Print total surface flux provides overal heat leaving or entering the body throught that surface. It integrates the flux across the surface [mW/mm2 * mm2]. Units should be [mW] in your case.
It match the Energy balance.
| Entering to the system | ||||||
|---|---|---|---|---|---|---|
| 2000 | W/m2 | 3 | m2 | 6000 | W | (Vertical Surface) |
| 1000 | W/m2 | 5 | m2 | 5000 | W | (Tilted Surface) |
| 11000 | W |
You could request SECTION PRINT for any of the other two surfaces to check
Just as a note, keyword editing is not necessary if you direct request RFL (or just use history output for the edge and select Totals = only)
And if you divide RFL by the thickness (1000mm) you have 11000 W/m
1 Like
Is there a way to read the Cavity sink temperature that ccx has determined?
As the surface temperature is not uniform there is no way to extract it from the equation known q.
Okay, this post is more addressed to the bug section but relies on this topic so i decided to post it here.
I went back to my “unitless” .pmx file and noticed a strange problem with the Field output results for RFL, they are completely wrong:
BUT when i request History output for RFL in addition, these AND the Field outputs for RFL are suddenly okay:
I don’t know if it’s a problem on the prepomax or calculix side.
And i think there is still something wrong with the units by importing material properties from the library. As i created the file first, i added copper material from my material library where i stored all properties in mm,ton,s,°C and forgot to check, if the units fit to the SI system (m,kg,s,K). Most values didn’t change, but the Specific heat did. Now I think that for the unitless system no values should have changed, correct?
Another thing is, when i try to create History output in my unitless file and request node coordinates, i get the error no abbreviation is specified for length unit… So coordinate output actually can’t be used without specified units. Is there the chance that it can be changed in the future or will it remain a limitation?
But even when i do the same in my other .pmx file with units, the error disapears but the tables are empty (maybe coordinate output is in general not possible for heat transfer analysis?):
I used Prepomax v2.1.7 and ccx 2.21
radiation_no_units.pmx (105.4 KB)
radiation_units.pmx (85.6 KB)
Regards and thanks @Matej
1 Like
Thanks for the files. I will look into it to see what is going on. Hopefully, I can recreate all the problems.
Thank you, I have to add that the problem also occurs with the other file where i used the mm/ton/s/°C unit system:
I can confirm this. But is seems the problem is CalculiX. If you look at the .frd file the values are the same as shown in PrePoMax. I think it would be good to notify Guido about it (@FEAnalyst).
You are right. There is a bug when converting values with units to unitless system of units. I will fix the problem. The question is, what whalues are expected for the material properites when unitless system is used. Probably values entered when creating the material. The problem with this approach is that now each material property remembers the units it was entered in. So user can use arbitrary combination of units when creating the material. Once this materil is used in uniless system the values enered will not be converted, only the units will be removed.
A bug. I have fixed it.
A feature 
The coordinate output was based on results field DISP. If no field was present, no coordinates could be computed. I have fixed it.
Are you planning to integrate the temperature field.? mmmh. 
That could be a solution to find/check for the sink temperature.
I think that’s okay. Conversely it would mean that, if one creates materials in the unitless system and save them to the library, the values wouldn’t be either converted into another unit system (that would be illogical). In that case one could create a separate database (one with/one without units).
Personally, I don’t know any other software that offers the option of working with or without units in order to be able to make a comparison.
1 Like
Sorry I can’t help you finding answers to your question. I just wanted to plot the temperature as line graph as Matej suggested here:
https://prepomax.discourse.group/t/temperature-graph-on-the-surface/1901/6?u=gunnar
in addition to coordinates, direct arc length of the selected edge would be nice to have as option…
1 Like
Yes, materials entered in a unitless system would not be converted when used in a model with units.