Which EM analysis software is best?
ANSOFT?CST?ANSYS?ZELAND?...
We plan to buy a EM analysis software.
But we don't know which one is best.
Please give me some advices.
Thanks!
It's depend on your application but in my opinion CST is better than Ansoft.Both of ANsoft and CST is accurate but Ansoft HFSS needs huge memory to simulation.
Hi!
Also depends on what kinds of EM problems are you going to solve..
Please specify more!!
Rgz,
eirp
HI Msh ,
I don't think Cst is better than HFSS .We have compared a few results from HFSS ,CST with measurement(espcially on antenna's reflected coefficient and pattern ) ,the HFss's is much accord with measurement ,the CST's is far away :(
I think that HFSS is better when simulating a mid-narrow frequency bandwidth ,Maybe the Cst is adapt to Wideband according it's FDTD method .
CaoCao
HFSS needs huge memory installed, 1GB is minimum IMHO!!
CST is much more efficient. New CST-MWS 5 will be better yet.
I have mostly obtained pretty good agreement with CST and measurement, anyway manual mesh adapting is necessary for some structures.
smilezhi,
specify for what purposes you need EM simulator.
Best u can do -> measure your kind of structure(s), tell us the dimensions etc. and we will try to simulate. Everybody with his favourite software. You can decide viewing the results :)
Regards,
eirp
i have tried both Hf$$ and IE3d in microstrip patch antenna design
i think iE3d is more effecient ,and it much faster than HF$$
IE3d is 2.5D software,it based on MoM
the HFSS is 3D software,it's accuracy is more approaching measurement results .
Accuracy is first and speed is second :)
Smilezhi
erip had brought forward a good idea to decide which EM software your need ,you can tell us some parameter about you want to test ,erip can use CSt and I can use the HFSS .:)
CaoCao
Hi, CaoCao:
1. IE3D is no longer a 2.5D EM simulator. It has been able to solve 3D metallic structures. The IE3D 10.1 just released can also handle finite dielectrics and it is a true 3D feature.
2. I think we should not just say 3D solvers are more accurate. You may want to check a paper by Pozar in comparing the Agilent HFSS and other MOM simulators on modeling patch antennas. The MOM simulators are more accurate and much faster. I think accuracy and efficiency are related to what kinds of problems you are solving and how you use hte software. Thanks!
Dear jian,
will be appreciated if you can upload mentioned Pozar's paper here please.
BTW, we have measured some vertical dielectric stubs. I'm planning to compare measured data+MWS+IE3D 10.1...
Best regards,
eirp
It is on the IEEE AP-Symposium 2000 or 2001? I am not sure whether there is a detail comparison on the digest. However, he has detailed comparison on the talk.
I also think that CST is better for wideband applications. But for narrowband and resonance structures HFSS is more perfect.
many things has changed since pozar's publication... but the comparision and actually its way is quite intersting.
I can say and honestly i can say that ie3d is really good choice even if you consider cst or ansoft products.
Guys, lemme ask once more that we don't enter same endless topic as before. We did it many times and sometimes it went far too hot.
To me a company that is considering procuring 3D EM solvers must have ppl in there able to sort which one is better for their particular needs. The question as posed here is simplistic, devoid of any practicality and am not sure is even ethically correct. Same is to ask which car is best - same nonsesne - like ODE or PDE with no boundary conditions. Can be anything. It is why, I kindly ask the mods here to require no more questions of that kind. We have delved into the nuts and bolts of nearly every solver with pros and cons and see no reason to start it over again.
We all know that 3D solvers are rather pricey and I know no real purchase endorsed without the chief engineer signature. It is hard to believe company ready to expend $25.000+ based on forum like ours. We all deal with different topics and each and every one holds an opinion - which is what makes elektroda folks so precious commodity. We can share, argue, agree etc and still be very useful to each other.
One other question I was about to aks many times our contributors. Guys, many times there are questions arising by someone not reading the manuals. When we address these questions let just point the page and the title of the manual. Else, we need too much space to say something that was said in the manual.
It is much more instructive and effective if we just point our colleagues to a more exhaustive sourse than trying to encompass the answer in a few lines. Like the question - what is the difference between EM and circuit solver. Many times the web sites of the respective vendors provide excellent information and a good link can do better job than our short posts here.
I will appreciate your opinion on the subject matter. My purpose is that we strealine the topics and improve the quality rather than increasing the volume. I wish that we propose accuracy tests of at least few structures and then providing comparative data.
Thank you in advance.
Cheng
I know CST and HFSS. HFSS doesn't have a good drawing tool. I used to draw my models in autocad and put in HFSS. CST however has a very good interface. I like this feature. you have the control of your model. you can parameterize, you can go back to history etc. It really makes my job easier.
In terms of accuracy, HFSS and CST (depending on solver ) don't have significant difference.
I prefer HFSS. Numerical and experimental results agree very well. Take a look at some actual example:
http://www.emtalk.com/forums/viewtopic.php?t=7
Also Ansoft's Designer is very good especiallly version 3.0 and higher. It is MoM based.
Hi,
I agree with cheng. At the end of the day you and you colleagues have to work with the programs. It therefore make a lot of sense to check the programs yourself before you buy it. I believe most vendors offer a trail period of their product. This also gives you the possibility to "test" the technical support of these vendors because I guess you certainly will have some questions concerning your particular structure which are not answered by the Handbooks.. I would look for the usual suspects in this forum and I then would request a trail license.
Best regards,
F.