Learning Forum

[2108397]

Hi,
I am getting this warning while calculating the SAR in HFSS "[warning] Average SAR calculation with invalid tissues, please check your SAR settings."

I have edited the density of the tissues using the ITIS database (https://itis.swiss/virtual-population/tissue-properties/database/database-summary/). The screenshot is attached for the SAR settings of the simulation.

Kindly guide me to resolve this warning.

Secondly, the SAR distribution is not uniform around the antenna structure, also shown in screenshot 2. I have tried different human models with 2mm, 4mm, and high precision models, but all are the same. Kindly guide me. Is it because of the antenna structure?

 

[Irina G.]

I can not reproduce the error, in general, material should be able to dissipate the power for SAR to be reported. I suggest checking the options:

1 - Mass density in material definition (not in SAR setting) is reasonable. Note that in Material definition it's in kg/m^3

2- in SAR setting, check that Mass of Tissue and voxel size is reasonable for the design scale.

SAR dittribution, even  Average SAR, is not supposed to be uniform. I dontt see anything wrong with the picture. The outline of the structure is probably defined by the geometry and material properties. if you want to make Avarage SAR more uniform, increase Mass of Tissue. To speed up the report, I suggest switching to Gridless definition from IEEE. 

 

[2108397]

Dear Irina,

Thank you for your reply.

Can you please recommend to me why I have to use gridless, I know that the IEEE follows the standards which are used as a standard for the compliance of the SAR value. 

Please see the attached screenshot:

1. I have already inserted the density of the material using ITIS database.
2. I am using 10 grams of tissues for averaging the over 10 grams mass of the tissues (SAR_10g).
3. I am using a voxel size of 1mm, but the resolution of the human body model is 2mm. 

Please explain to me why I have to move to Gridless as it does not follow IEEE standards for averaging the SAR values. Secondly, does the voxel size I am using 1mm voxel size for averaging with the 2mm resolution of the model is correct? Thank you

[Irina G.]

Hi, 

I suggest you look at the on-line Help entry titled "Modifying SAR Settings" for general description. Below are my answers / comments:

• it's not "necessary" to use "Gridless" definition; but it's fast and it doesn’t do additional operation of creating the voxels (cubes) from the HFSS mesh. Note that  HFSS mesh is tetrahedral based and conformal to geometry. The Average SAR plot is smoother using Gridless option, use it as a quick check of a baseline.
• Q: does the voxel size I am using 1mm voxel size for averaging with the 2mm resolution of the model is correct? A: there is no right or wrong setup. Note that the field and SAR values are reported based on mesh inside the object, not the model resolution. However, it's reasonable that the voxel size is smaller than the model resolution. In general, the smaller the voxel size, the smoother the plot will be (it'll take more time). Also, it's better if many voxels fit into the volume defined by "Mass of Tissue" setup. SAR_10g with 1mm voxel looks reasonable. Another consideration is the size of the actual solved mesh compared to the size of the voxel. You can seed mesh at the area close to the antenna - SAR plot should become smoother.

• I was able to reproduce the original message: it appears when Object list is not used and averaging process adding different objects with materials not suitable for SAR reporting.

If you can, file an official ticket to discuss further details. 

[2108397]

Thank you for your reply. I am still struggling to understand why SAR distribution is disruptive for IEEE 62704-4 Algorithm, whereas for gridless, it is smooth. Even in CST software, it is smoother. 

I have reviewed all HFSS, helping material, and IEEE standard for FEM. But, still not understanding it. I have increased the mesh of the model by assigning it inside the tissues --> select humanaverage tissue --> assign mesh --> inside --> length based -> max element 10000. It is still the same.

 

 

[Irina G.]

Hi,

if "gridless" option is smooth and IEEE is choppy, it's the mesh /voxels mismatch. 

Are you eligible for official support? if yes, we can share the model. Visualize the Mesh on the tissue object. 

To have a smooth picture, the mesh size shuld be less than the voxel size. To restrick the mesh size, use the first option of the screen,

"Set max element length", start with the size of voxel. Don't make the voxel size too small compared to the total size of the tissue object, or  it will take a while. 

On the bigger prospective: why can not you use gridless SAR? It'll give you the same averaging based on "Mass of Tissue" without the pain of converting the geometry conformal mesh to box-like voxels. CST results are smoother probably because they use brick-like mesh to begin with. Note that the stats of the mesh built by HFSS is available in the results.

 

 

 

[2108397]

Dear Irina,

Thank you for your reply.

I am eligible for official support from Qfinsoft Ansys partners in South Africa. But, sometimes, they respond late. Please let me know how I can share my model with you.

 

Please verify for me that the "Gridless" averaging method can be used for certification of SAR, as the only difference is it does not consider voxel size. Thank you

[Irina G.]

Please, go through the official route. Channel partners can request help from the core Ansys team when necessary.

If certification is specifically based on voxel based IEEE definition, it's necessary to use it. Denser HFSS mesh should smooth IEEE based SAR plot.