HFSS - is this the most sensible way to speed up a calculation?
It it taking ages. It has been running 3 hours, has used 17 GB RAM, and is still on only the 2nd adaptive pass. At this rate I can imagine it will take several days at least. (The machine has far more than 17 GB RAM, so its not swapping to disk). Despite my other questions about setting up for best use of parallel processors, it is clear this part of the simulation is using all 8 cores, as the load average is close to 8. Here's the output from "top" which Linux users will know. For Windows users, consider the task manager would show the CPU usage is very high.
top - 17:15:18 up 22:07, 3 users, load average: 7.63, 5.46, 4.37
Tasks: 209 total, 1 running, 208 sleeping, 0 stopped, 0 zombie
Cpu(s): 81.8%us, 0.7%sy, 0.0%ni, 17.5%id, 0.0%wa, 0.0%hi, 0.0%si, 0.0%st
Mem: 44927468k total, 17408008k used, 27519460k free, 157816k buffers
Swap: 35291128k total, 0k used, 35291128k free, 5860432k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
17544 deborah 17 0 17.7g 10g 15m S 659.9 24.5 41:56.29 hf3d
10630 deborah 16 0 12764 1184 820 S 0.7 0.0 1:46.66 top
1 root 15 0 10372 696 584 S 0.0 0.0 0:01.91 init
1) To get the radiation pattern one needs to use a small mesh on the radiation box. That need to be lambda/6 or smaller at the highest frequency, so length of a tetrahedral on the radiation boundary needs to be 10 mm or less for accuracy at 5 GHz.
2) To get accurate results at the low-frequency, (200 MHz), the radiation boundary needs to be a distance of at least 750 mm (lambda/4) away from the antenna
So I end up with a large volume of airbox to ensure accuracy at the low frequency, and a fine mesh on the radiation boundary to ensure accuracy at the high frequency. But neither of these are simultaneously necessary. The airbox needs to be large for the low frequency, but the mesh does not need to be very fine on the radiation boundary. At the high frequency, the mesh needs to be fine, but the airbox does not need to be large.
So I'm thinking I should split this problem into a 2 (or more) different simulations, rather than 1. Perhaps if I split it into 2, use one to cover the range 200 MHz to 1 GHz, and a second to cover the range 1 GHz to 5 GHz. Then using a small mesh and small airbox for the 1-5 GHz simulation, and a larger airbox and larger mesh for the 200 MHz to 1 GHz simulation.
Am I missing a trick here? Is this a sensible approach?
Deborah.
This actually finished in far less time than I expected. The delta S was below 0.02 after only two adaptive pases, but I usually force there to be two consequtive converged passes. So three were done. From there, the total simulation time was about 6 hours, rather than the days I was expecting. But still, I would like to find if there is a quicker way of doing what I wanted.
Can you please explain a bit how can we control the meshing to make the simulation quick. As in recently i came to know how meshing affects the design simulation. But i need to know how to use it to our own benefit.
please help.
Thanks and regards.
Ashish
Ashish... start a new thread, hijacking a thread is annoying and likely not going to result in productive answers for you.
Deborah,
As this is an electrically small antenna, there are several things to be aware of:
1) Since Gain will be very low, radiation box can be much closer than lambda/4 (try it and you will see ;))
2) As this is an electrically small antenna, you are likely matching it to a circuit network. As such, the antenna impedance is likely very inductive and converging on delta S may not result in the accuracy you need. If the antenna is looking at a short, then one pass with mag S = 0.990 and next pass at 0.998 will look converged as delta S is 0.008. However, you are so far from a match that this is useless info. You can/should converge on the inductance of the antenna or at least the imaginary part of Z_input for electrically small antennas
3) Feel free to post your project and I will take a look
Have Fun
Hi tallface65.
I was aware that the distance to the airbox could be reduced in areas where an antenna was "weakly radiating" - that information came from an Ansys engineer. But I never asked for any guidelines on what was "close", and what was "weak". I believe the main effect on having the airbox too close is inaccurate far field patterns and hence gain - it does not change S11 too much. I should try varying the airbox size as you say, to see the effect for myself.
The excessive run time of the simulation, which prompted me to start this thread, ultimately gave really stupid results, which I put on another thread.
https://www.edaboard.com/thread261690.html
where the .hfss file is attached, though https://www.edaboard.com/attachments...olinear-v5.zip is a more direct link. It would be helpful to know what's wrong with that .hfss file, as not only was it taking ages, but the results are silly.
I would add, the "antenna" in the .hfss file is supposed to represent a bit of coax where most of the braid is removed to make it radiate, like a leaky feeder. But my aim was to make a phased array antenna with high gain, not a leaky feeder, so there would be sections like in the .hfss file, and other sections which are just normal coax.
The silly results could be due to at least one of:
- The convergence criteria was not suitable. I had not given very much thought to this, but the fact ΔS converged after only two adaptive passes did make me suspicious. Your comments above saying one should converge on the inductance of the antenna, not ΔS, is making me think that was the problem.
- I'm not totally convinced I have got the waveports set up properly. I'm sure you can tell that quickly.
- A post by fred23 here https://www.edaboard.com/thread261475.html#post1119141 has convinced me my antenna should not radiate at all! . As such, I have scrapped this design for this exact antenna, though I am working on a modification of it.
Deboah.
PS,
if you have a minute, could you please take a look at
https://www.edaboard.com/thread262046.html
I'd really like to solve that problem, as it would certainly help me in many ways, including seeing the effect of varying the airbox size if I swept the size of the airbox as a parameter.
Hi Deborah,
I notice u want to study the antenna performance in a large freq range, 200 MHz - 5 GHz. I suggest u to split the freq range into several sub-ranges (two sub-ranges may not be enough). According to my experience (especially with HFSS v11), a singe simulation for this large freq range is not only troublesome to set up (the airbox size vs mesh size dilemma), but also tends to deliver wrong results, even use discrete freq sweep.
Please be aware that HFSS 11 is nearly 5 years old. The current version of HFSS has several new features and solvers to simplify the setup for UWB antennas, including output variable convergence and dependent mesh convergence. Also, with the new convergence features of the interpolating sweep, the sweeps no longer need to be setup in sub ranges :)
Have Fun
Here is a link to another software (ADF) which claims much faster simulation for larger antennas / arrays. I am awaiting the evaluation version, thus very minimal info with me though the software looks promising !
http://www.idscompany.it/page.php?f=92&id_div=7