How to use the dummy objects in Ansoft tools?
You use dummy objects to create better meshing in that region to get better results. It is very useful!
Suppose you have a trace on a PCB board (could be antenna, whatever) and would like to visualize the EM field radiated by that piece of metal.
Most likely you would be more interested into the field close to the metal and therefore would like the mesh to be more dense in that region. So you could put a dummy object, like cylinder or small box, around the trace. This way, instead of having one area between the trace and the model boundary box, you had 2 regions.
1. Between the surface of the trace and the surface of the dummy object. This should have a very dense mesh, as it is the area of the interest
2. Between the dummy object and the boundary box. This area has a coarse mesh(small number of elements), but you still have a results.
Please note that both regions have the same physical properties, so Dummy object is there just for the purpose of tweeeking the mesh. As far as the electrical properties of the model are concerned, they are the same as if it were not there.
Hope this helps!
:D
Well, the info definitely helps. but is this dummy object a model or a non model object? If it is a model object and I have this dummy object inside another object (even though it is of the same material), does it not change the model then?
Also, does the size of the dummy matter?
The whole point of "dummy objects" is to cause the mesh to concentrate in areas where you have a particular interest.
Non model objects will not participate in the meshing. All the usual rules about objects apply.
HI, meantares
dummy object should be model objects and they won't change the simulations results.
Hi,
As I was told by Ansoft representative a long time ago when version 8 was out, the concept of dummy object was introduced to help the mesher by applying human knowledge and experience to come to the optimum mesh for the particular problem type with the less effort (passes) and waist of dense mesh points where it is not needed.
It is easier to explain this on a example - microstrip analysis. The large objects are initially evenlly meshed in the whole of their volume whole microstrip substrate. If left on its own , the program mesher will in later passes, based on the calculated energy stored in a single mesh tetrahedron, refine the mesh so that finely mesh areas correspond to large field/energy variation volumes. Finally, after some number of passes, the mesher will concentrate the dense points underneeth the microstrip track as this is the area of most rapid field variations, and the rest of the substrate will be much coarsely meshed.
However, although we don't know the solution to our problem and have to use EM sumulator to do calculations, we do know how the field of microstrip line looks like so we can be smart and place a dummy object, approximately 5 widths of the microstrip line in total symmetrically around the center of the microstrip line (2.5w to the left and 2.5w to the right) and dummy object should have the same height as microstrip substrate and use the same material. The object is much smaller than the whole microstrip substrate and will be initially meshed much finely that the rest of the substrate. So, by applying this method, the initial mesh will be much closer to the optimum one than in the "free running" mesh case.
However, HFSS mesher has been considerably improved since v8 and for v11 that is currently available, it is not needed to use dummy elements (as I was again told by the representative)
flyhigh