heal objects+hfss
Mesh generation has failed. Specific details may be located in 3D Modeler> .....
Simulation completed with execution error on server: Local Machine.....
I do not quite understand what's mesh generation, when i refer to the logs, I found something about misaligned entities... I try using the "healing" option there but can't align the faces...
any suggestion so I could run the waveport simulation smoothly?
I got the same error when I ran it.
The design check says your design is OK.
Questions -
1 - how come you do not have any radiation boundary?
2 - Did you draw the structure, or you imported it. You would only need to heal it if you imported the structure. It seems like you drew it.
I looked through your errors. There are numerous material types. On the exterior level, your design might look perfect with no intersections. In the log that lists the misalignment, you need to zoom in, and make sure that the two face do not intersect.
If you have the help file,
Go to technical notes -> Geometrical Objects -> Healing and Meshing.
I hope that helps. Let me know if you are still stuck.
I
hi peleda,
i draw the structure myself, and i have tried the 'healing' option but there's a error message saying that it can't heal the model.
i'm sorry but not sure what's a radiation boundary?
i will try to look into the model again for intersection faces. i come back again if i have further problems.
thanks so much for the reply. :D
when you run an HFSS simulation, fields get radiated from the source. You have to define a boundary in which the solver will solve for the fields.
This in fact affects the accuracy of your solution.
For a structure like yours, you would create a rectangular boundary. generally the boundary tends to be at least 1/4 of a wavelength in length from the edge of your object.
If radiates from the top, which I am guessing, you do not know, you would create a rectangular box that is (1/2 by 1/2 by 1/2 )* wavelength.
Then click the box, and set the boundary to radiation boundary. Use the center frequency to determine the wavelength.
How come you did not use variables in your design? The benefit of variables is that when you alter the length of something, the other related structures alter their own length in unison. In that way, you do not have to edit every single dimension.
Thanks..
I have one more question, how to set the solution frequency for the analysis setup in order to get the correct graph that we want? Is there any method to determine the solution frequency need to be set?
Thank you.
The solution frequency depends on the application. For example, if you are designing a wireless system for home use, the frequency used for this purpose is 2.45GHz, hence your solution frequency will be 2.45GHz.
So for your case, you mentioned this is for saw packaging. I do not know what that is. But if you are concerned about noise from the electrical surrounding, then a solution frequency of 60Hz (??? not sure what the power line frequency is for the US) will suffice.
If you have a frequency band in mind, for example you are wondering what happens between 1GHz and 2GHz, then
- for fast sweep, you use the midpoint (1.5GHz) as the solution frequency
- for discrete sweep, you use the upper boundary (2GHz) as the solution frequency
You only need a radiation boundary, if you expect your system to radiate fields. The dimensions of system seemed small (in the mm range). So I am guessing your frequency range for seeable results should be in the GHz range.
Hi, the frequency I use is 433.93 MHz, so is it i just set this frequency as the solution frequency in the HFSS simulation setting under SOLUTION SETUP?
Then, how do I set the START and STOP FREQUENCY and STEP SIZE under the EDIT SWEEP for the FAST SWEEP TYPE i use? When I change the START and STOP FREQUENCY, i get the different result in the graph of the report. So I would like to ask, is there any method to set the START and STOP FREQUENCY? I just use the try-and-error method to get the result i want, but this spend a lot of time and it's hard to get the correct result in the graph.
Thanks for your advice..
Yes, you set 433.93MHz as the solution frequency.
To do a frequency sweep, right click on the solution setup you just made, and click "add sweep".
Change the selection to Fast sweep.
Edit the start and stop frequency to your choice.
There is also a drop down list, I think you want to choose linear step, and you input your frequency in the step frequency, or you use linear count, and you input how many frequency points you want in between that range.
After simulation, you can either view a function with frequency as the independent variable,
or each frequency you view its result.
i hope this helps...
hi,
thanks for all the help i got. but...
i still won't able to run the simulation. still getting the same error of "mesh generation has failed". I got some suggestion that i should not run the "healing" after I have drawn the model before the simulation, but even I re-draw the model and try simulating it, it seems the same "mesh" error is still there...
I try to bear in mine not to intersect the models, but I'm not sure whether I got the 'intersect' meaning right. Does your 'intersect' means that two faces shouldn't touch it other, or i should have overlap it into the model???
confused...
A proper licensed HFSS shouldn't have this problem unless (I speculate here) you are using a cracked version having a damaged mesh3d_ng.exe
actually, i tried running the file myself.
The HFSS I used is definitely not cracked.
The structure is somewhat complicated. So easy to make drawing errors. Given that it was drawn by hand, it should not need healing. But i am still taking a look @ it.
nope, i am using the original software.
anyway i figured out the problems, that i have some surface drawn on really small measurements ~nanometers, which i change to the ~um range which now my model runs smoothly.
I am trying to get my desired simulation results now. thanks again for all the help i've got. especially to peleda.