Welcome to EDAboard.com

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.

Anybody can help me out of this? CST suite studio model cannot be simulated

Status
Not open for further replies.

Jason Hou

Newbie level 6
Joined
Sep 26, 2012
Messages
13
Helped
1
Reputation
2
Reaction score
1
Trophy points
1,283
Location
Singapore
Activity points
1,369
Alright, here's the thing. I have made a patch antenna and set the material of it properly, before I simulate, I set the background material to 'Normal', but the simulation aborts everytime and shows 'The calculation domain is completely filled with metal(PEC), please check the background material and layer properties'. What should I do? Can anybody help me out of this?:?:
 

thylacine1975

Full Member level 6
Joined
Dec 7, 2010
Messages
337
Helped
227
Reputation
448
Reaction score
218
Trophy points
1,323
Location
Canberra, Australia
Activity points
5,127
How are you feeding your structure? Sometimes CST whinges (with cryptic, irrelevant errors) if the excitation ports are too close or touching the calculation boundaries. In particular, make sure your boundaries are set to "open" (i.e. PML) for antenna simulation. If you inadvertently have a "electric" boundary set, CST adds an (invisible) wall of PEC at the edge of the calculation space which can wreak havoc with waveguides/coaxial feedlines which otherwise *appear* open.

Also, click the mesh view button (and/or take a peek at the number of meshcells displayed at lower right) to make sure the meshing appears sensible and it hasn't simply quantized your entire structure into a PEC block!

If that doesn't help - post your file and I'm sure someone will be able to set you straight :)
 

majidmad

Junior Member level 2
Joined
Jan 20, 2010
Messages
20
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Location
usa
Activity points
1,420
How are you feeding your structure? Sometimes CST whinges (with cryptic, irrelevant errors) if the excitation ports are too close or touching the calculation boundaries. In particular, make sure your boundaries are set to "open" (i.e. PML) for antenna simulation. If you inadvertently have a "electric" boundary set, CST adds an (invisible) wall of PEC at the edge of the calculation space which can wreak havoc with waveguides/coaxial feedlines which otherwise *appear* open.

Also, click the mesh view button (and/or take a peek at the number of meshcells displayed at lower right) to make sure the meshing appears sensible and it hasn't simply quantized your entire structure into a PEC block!

If that doesn't help - post your file and I'm sure someone will be able to set you straight :)

I have the same problem. I don't understand what you mean by your second paragraph. I selected Mesh view button. at the lower right corner it mentioned 8580 cells. So what does that mean? how can I recognize my problem by this number !
thank you
 

thylacine1975

Full Member level 6
Joined
Dec 7, 2010
Messages
337
Helped
227
Reputation
448
Reaction score
218
Trophy points
1,323
Location
Canberra, Australia
Activity points
5,127
Hiya majidmad,

The number of mesh cells indicate the number of discrete calculation domains that your model has been "meshed" (or sliced/cut up) into. The physical object being modelled is assumed to be homogenous within any given mesh cell, i.e. there is *no* variation in the physical material within the mesh cell. This means that the mesh must be fine enough to capture all practical physical details/characteristics of your model. This is what the mesh view allows you to see - where the meshing algorithm has decided to place the calculation domains with respect to the physical model. This allows you to check that it hasn't decided to lump the (for example) entire cross-section of a piece of coax into a single cell and thus erroneously model a structure with conductors and dielectric as one uniform block of stuff.

Deriving a sensible mesh is a non-trivial exercise and the penalty for being conservative is bloated computation times. While the CST included algorithms are pretty good, they can't easily differentiate between important features (e.g. transmission lines) and unimportant detail (such as redundant fasteners) and hence need a little help sometimes. Meshing theory/algorithms have undergone radical development since Yee's original uniform cubic mesh in 1966.

The number of mesh cells can be used to give you a quick feel for how sensible a given mesh might be. If you assume cubic mesh cells (which the CST ones aren't - but for our purposes this assumption is OK) then the cube root of this number gives you the approximate number of slices in x, y & z the solver will take through your model. For example, for the 8580 cells you mention, this gives ~20 slices in x, y & z. That is, your model has been quantized into only 20 discrete segments across it's width, height and depth. Whether this is sufficient depends upon the size of your model, the detail it contains and your frequencies of interest... Despite not knowing anything about your simulation, 8580 cells still feels a little low... by ~an order of magnitude. In my post #16 in this thread https://www.edaboard.com/threads/260093/ I describe how to increase it - hopefully that'll work for you!
 
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top