CST "wavelength refinement feature"

Hello everyone,
I am simulating a wideband printed slot antenna with substrate height of 1.5 mm. I have a concrete block on top and an epoxy box at bottom attached to it. The problem i face is that if i simulate this antenna in the specified band of 100-800 MHz i get a mesh size of almost 9000 cells. The simulation is aborted by an error of "reducing the stability factor". I did the mesh refinement by using Wavelength refinement and the simulation ran fine and i got results. I dont what could have been the results but they look pretty good for me .
Now if i do another way of mesh refining that is by Adaptive mesh i cant get it through unless i reduce the stability factor to 0.5 or less from the original value of 1.
I tried to simulate the same antenna in HFSS but it takes very long to converge almost 20 passes and i dont get the same results as i get in CST using the wavelength refinement.
What does wavelength refinement actually does? How much you can rely on these results as far as the reality is concerned?
Soon I have to get it manufacture to see the results but want to be sure that i am doing right simulation.
Will be grateful for ur suggestions and desparately need them .
Regards
Safwat

Hello,
wavelength refinement makes the mesh denser in areas with Er>1. If i'm not mistaken, this feature applies the "lines per wavevlength (LPW)" setting in the dielectric slab with the dielectric constant taken into consideration. For example, if your setting is 10 LPWs in an air brick (Er=1), then for a dielectric brick of Er=4 you'll have twice as many mesh cells for each direction - Meaning, that for a 3D brick the total mesh cells count will increase by a factor of 8.
BTW, for a more stable result I would try the energy based adaptive meshing.
P.

If by using wavelength refinement i get good results and if i run adaptive meshing and it never converges what thing to rely on ? ... using wavelength i don run adaptive meshing....

It is quite hard to tell without actually checking the model myself. In cases like this I usually start with a simple model, make sure that it is running in a satisfactory manner, and only then I move on to a more complicated model.

how can i send u the design ?.. i tried with HFSS and the results are completely different on the passimistic side :(. And i m unable to converge the design in CST and it gets too slow ... i m using CST2006B ... will really appreciate ur help in this regard.

Hi safwatirteza,
Have you got a cst file for your simulation? If yes could you please upload it here, so can we check it.
Regards
Praful

i am attachin what i v been tryin .. u can see everythin here .. i v to model the antenna with concrete .. u can remove the epoxy material if u want to .. but in order to get a converged result my pc dies :(... the process is too slow n i m unable to tune the antenna ... n i dont know what can be the mesh settings so as to get the accurate results .. HFSS also takes long time for this model n secondly the results are completely different from CST .. i dont know why .. i wud be grateful for ur suggestions regards.

I'll try it today and let you know the outcome.. [:)]