Warning 4600

Nooshin

Hello Administrator,

I faced this warning in my simulation in Feko as I wanted to calculate SAR.

WARNING 4600: Spatial-peak SAR computation and output as a function of the medium or label is not possible for the MoM with special Green's function

Would you please help me to solve this warning?

Regards,

Nooshin

Torben Voigt

Nooshin said:

Hi Torben,

Thank you very much for your help. I defined the dielectric above the structure finite along z-direction as you see in the attached figure. However, I still face the same error. What should I do? Moreover, when I request for near-field distribution on different plane above my structure, in POSTFEKO I can see the SAR distribution on those planes too without any SAR request in CADFEKO. what is the definition of this calculated SAR? What are the differences between this SAR calculation and the ones in SAR request window (volume-average SAR and spatial-peak SAR)?

Regards,

Nooshin

Hi @Nooshin ,

You can get plenty of information in the internet about SAR. It can be calculated by knowing the E-field and the mass density of a medium:

SAR = (conductivity * E^2) / (2 * mass density)

In your case at 3 GHz for material "Gland": SAR = (1.818*self^2)/(2*1041)

Therefore you can output it directly as nearfield value in POSTFEKO. The "Volume average SAR" is calculated by Feko in the same way and then output as a single value. For the spatial-peak SAR Feko searches an area and finds the position of the cube. This usually takes a very long time.

Best regards,
Torben

Torben Voigt

Hi @Nooshin ,

Could you attach your model (.cfx) here? Which version do you use?

Best regards,

Torben

Torben Voigt

Hi @Nooshin ,

I just confirmed: The SAR can't be computed inside the infinite dielectric bit it is accurately calculated in any finite dielectric. In your case this would bean that you'd have to model a finite dielectric above your structure to calculate the SAR inside this region.

Hope this helps!

Best regards,
Torben

Nooshin

Torben Voigt_20420 said:

Hi @Nooshin ,

I just confirmed: The SAR can't be computed inside the infinite dielectric bit it is accurately calculated in any finite dielectric. In your case this would bean that you'd have to model a finite dielectric above your structure to calculate the SAR inside this region.

Hope this helps!

Best regards,
Torben

Hi Torben,

Thank you very much for your help. I defined the dielectric above the structure finite along z-direction as you see in the attached figure. However, I still face the same error. What should I do? Moreover, when I request for near-field distribution on different plane above my structure, in POSTFEKO I can see the SAR distribution on those planes too without any SAR request in CADFEKO. what is the definition of this calculated SAR? What are the differences between this SAR calculation and the ones in SAR request window (volume-average SAR and spatial-peak SAR)?

Regards,

Nooshin

Torben Voigt

Nooshin said:

Hi Torben,

Thank you very much for your help. I defined the dielectric above the structure finite along z-direction as you see in the attached figure. However, I still face the same error. What should I do? Moreover, when I request for near-field distribution on different plane above my structure, in POSTFEKO I can see the SAR distribution on those planes too without any SAR request in CADFEKO. what is the definition of this calculated SAR? What are the differences between this SAR calculation and the ones in SAR request window (volume-average SAR and spatial-peak SAR)?

Regards,

Nooshin

Hi @Nooshin ,

Sorry for the misunderstanding! What I meant by "finite dielectric above your structure" was a dielectric cuboid for example. The warning says that the spatial-peak SAR (1g or 10 g cube) cannot be calculated inside the inifinite layers (infinite in x and y, planar Greens functions).

Of course, the SAR can just be calculated from the near field anywhere in your model. But then you do not have the position of the 1g or 10g cube. Is it important to you to find this cube? If you are simply interested in the SAR in a certain area, then you can simply create a near field request and display the SAR.

Something like this:

Note, the dielectric cube should have a very small distnace to the first infinite dielectric layer, else you'll get 

ERROR 53143: A dielectric triangle may not be located on an interface between two layers of the planar Green's function definition

Hope this helps!

Best regards,
Torben

 

Nooshin

Torben Voigt_20420 said:

Hi @Nooshin ,

Sorry for the misunderstanding! What I meant by "finite dielectric above your structure" was a dielectric cuboid for example. The warning says that the spatial-peak SAR (1g or 10 g cube) cannot be calculated inside the inifinite layers (infinite in x and y, planar Greens functions).

Of course, the SAR can just be calculated from the near field anywhere in your model. But then you do not have the position of the 1g or 10g cube. Is it important to you to find this cube? If you are simply interested in the SAR in a certain area, then you can simply create a near field request and display the SAR.

Something like this:

Note, the dielectric cube should have a very small distnace to the first infinite dielectric layer, else you'll get 

ERROR 53143: A dielectric triangle may not be located on an interface between two layers of the planar Green's function definition

Hope this helps!

Best regards,
Torben

 

Hi Torben

Thank you for your reply. if I want to define 1g or 10g cube, should I consider a specific dimension for it? How about the position of this cube above the structure? moreover, if I want to calculate SAR by requesting near field in a certain area, for example on a plane above the structure, what will this SAR refer to? what is the definition of this calculated SAR? What are the differences between this SAR calculation and the ones in SAR request window (volume-average SAR and spatial-peak SAR)? I have to compare this simulated SAR value with standard values. I am not sure which standard value I have to consider since this SAR is obtained from near E-field distribution on a x-y plane. 

Additionally, when I request for near-field, I face warning 121: Coarse segmentation for triangular patches in feko, whereas I use fine auto mesh in my design.

Sorry for lots of questions.

thanks for your consideration an time in advance.

Regards,

Nooshin

Torben Voigt

Nooshin said:

Hi Torben,

Thank you very much for your help. I defined the dielectric above the structure finite along z-direction as you see in the attached figure. However, I still face the same error. What should I do? Moreover, when I request for near-field distribution on different plane above my structure, in POSTFEKO I can see the SAR distribution on those planes too without any SAR request in CADFEKO. what is the definition of this calculated SAR? What are the differences between this SAR calculation and the ones in SAR request window (volume-average SAR and spatial-peak SAR)?

Regards,

Nooshin

Hi @Nooshin ,

You can get plenty of information in the internet about SAR. It can be calculated by knowing the E-field and the mass density of a medium:

SAR = (conductivity * E^2) / (2 * mass density)

In your case at 3 GHz for material "Gland": SAR = (1.818*self^2)/(2*1041)

Therefore you can output it directly as nearfield value in POSTFEKO. The "Volume average SAR" is calculated by Feko in the same way and then output as a single value. For the spatial-peak SAR Feko searches an area and finds the position of the cube. This usually takes a very long time.

Best regards,
Torben

Nooshin

Torben Voigt_20420 said:

Hi @Nooshin ,

Sorry for the misunderstanding! What I meant by "finite dielectric above your structure" was a dielectric cuboid for example. The warning says that the spatial-peak SAR (1g or 10 g cube) cannot be calculated inside the inifinite layers (infinite in x and y, planar Greens functions).

Of course, the SAR can just be calculated from the near field anywhere in your model. But then you do not have the position of the 1g or 10g cube. Is it important to you to find this cube? If you are simply interested in the SAR in a certain area, then you can simply create a near field request and display the SAR.

Something like this:

Note, the dielectric cube should have a very small distnace to the first infinite dielectric layer, else you'll get 

ERROR 53143: A dielectric triangle may not be located on an interface between two layers of the planar Green's function definition

Hope this helps!

Best regards,
Torben

 

Hello,

I followed your suggestion about using a finite cube above my structure, but I still face that warning: "WARNING 4600: Spatial-peak SAR computation and output as a function of the medium or label is not possible for the MoM with special Green's function"

Is my calculated spatial-peak SAR reliable with this warning?

Every guidance from you is appreciated!

Regards,

Nooshin

Torben Voigt

Nooshin said:

Hello,

I followed your suggestion about using a finite cube above my structure, but I still face that warning: "WARNING 4600: Spatial-peak SAR computation and output as a function of the medium or label is not possible for the MoM with special Green's function"

Is my calculated spatial-peak SAR reliable with this warning?

Every guidance from you is appreciated!

Regards,

Nooshin

Hi @Nooshin ,

The warning seems to occur whenever a SAR request and planar Greens functions are present in a model. As long as you don't relate the SAR request to any of the mediums used in the planar greens functions, you should be fine. 

In this example I have an infinite layer (Dielectric1) and I request the Spatial-peak SAR of a 1g cube in the cube above the infinite layer (Dielectric2). The warning is given, but the SAR in the cube is correct.

Best regards,
Torben