Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.
My opinion would be use of CST MWS. It is amazing regarding modelling of objects and ease of use. It is pretty accurate too. I don't know if it is more accurate than HFSS (that has been a constant source for discussion on this forum).
When I simulate this antenna, there is a warning like "some critical cells have been expanded to full size ... ", as a result, parts of the radiation patch or feeding strip hase been shorted to ground. Though I tried to change the mesh method, the warning exsists and the problem cannot been solved, WHY and HOW to solve this problem?
When I simulate this antenna, there is a warning like "some critical cells have been expanded to full size ... ", as a result, parts of the radiation fix or feeding strip hase been shorted to ground. Though I tried to change the mesh method, the warning exsists and the problem cannot been solved, WHY and HOW to solve this problem?
"critical cells" warning means that conformal fit algorhitm wasn't be able to treat curved surface withing the cell. In result, given cell is now modelled as a "brick", like when classical staircase fdtd is used. And yes, it can make unwanted short connection.
Turn on the mesh view and check for blue bricks - they are the criticall cells and try to improve mesh locally there.
Follow on to eirp's comment. Create an air object (preferably a box) at the location near to the expanded object and name it "dummy". Then alter Dx, Dy, Dz accordingly so that they are not equal zero, at the same time the number should not be rediculously small (result in a very fine mesh). Just enough to avoid the critical cell error.
Normally, when dealing with curved structure, I would prefer HFSS. However, in most antenna problems when a wide frequency spectrum and multiband characteristics exist, I will go for a time domain method like MWS. I think in your case, you should carry on with the use of CST MWS. The critical cell problem is the only major hassle, but becoming hands-on in the dummy object creation near critical cells, you should eventually breeze through the problem. Perhaps later on you may consider comparing your result with HFSS.
Thanks eirp and sassyboy.
Yes, I want to compare the CST results with HFSS resullts, unfortunately, I find it is very difficult to simulate my antenna(size of abou 1*3 lambda) on my PC due to the too many cells and very low computing speed.
red-sun, I have just a small thought. If you are really struggling with the simulation using either HFSS or MWS. I would recommend you try IMST Empire. Ask the company for a 30 days full trial. Now the difference this makes for you is quite dramatic in terms of speed or total mesh volume. Because this uses the near-field to far-field transformation to obtain radiation pattern. Therefore the bounding box needs only Lamba/16 away from the nearest radiation edge. Therefore the total mesh is reduced significantly compare to HFSS and MWS.
The only problem though is that this is a time-domain method and uses the Yee mesh cell. So, a dense mesh is still necessary along the conformed/curved surfaces for the best approximations.
In summary, all numerical methods require some sort of compromise for this particular task.
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.
