search

tmolteno / necpp

NEC2++ is a C++ rewrite of the Numerical Electromagnetics Code (NEC-2) with many new features like automatic error detection when you specify the structure incorrectly and much faster execution. Nec2++ can analyse radiating as well as scattering properties of structures. The simulation engine in Nec2++ is compiled into a library for easy integration into automatic antenna design systems or GUI tools. Examples are included for using Nec2++ from C/C++, Ruby and Python.
http://elec.otago.ac.nz/w/index.php/Necpp
GNU General Public License v2.0
239 stars 66 forks source link

RCS computation of dihedral corner reflector #53

Closed vineethajoy closed 6 years ago

vineethajoy commented 6 years ago

I was trying to compute the monostatic RCS of a dihedral corner reflector(Plate length=0.3m, Plate thickness=0.005m, f=9.4GHz) using NEC. But somehow I am not getting correct results. The output I am getting is given below and the NEC input file is attached herewith. plot

PEC-Dihedral-Tab-Delimiter-Tag-no-modify.txt

Any suggestion would be really helpful.

ytakeyasu commented 6 years ago

As NEC2++ output Runtime Error for your model: GEOMETRY DATA ERROR -- WIRE #1 (TAG ID #1) INTERSECTS WIRE #4 (TAG ID #124) I tried the model using 4nec2 & NEC2DXS on Windows and got the attached result. I suppose the result is what you expected. It seems NEC2++ and/or NEC2 is not the problem. pec-dihedral-tab-delimiter-tag-no-modify

vineethajoy commented 6 years ago

Thank you for your response. I tried the same in FEKO and I am getting similar results there too. But the correct plot is as given below: correct plot

I think there is some fundamental problem with the modelling. But I am not able to figure it out.

ytakeyasu commented 6 years ago

I'll suggest you to review your description of "EX" and "RP" command in accordance with NEC2's definition. The issue may be the direction of incident plane wave.

ytakeyasu commented 6 years ago

Are you referring to this article ?

http://ijamt.com/article_html.php?did=3188&issueno=0 INTERNATIONAL JOURNAL OF ADVANCES IN MICROWAVE TECHNOLOGY - Volume 2, Issue 3, August 2017, Pages: 110-116 Hybrid MoM-CG RCS Estimation of Dihedral Corner Reflector

Your model is not the Fig.2, but the Fig.3 which is the cross section of the Fig.2.

vineethajoy commented 6 years ago

Exactly. The model I dealt with is the cross-section in the XY plane i.e. Fig. 3 .
The results from comsol were perfectly matching with the results from an MoM code we wrote by our own and that was experimentally verified too. I did try various changes in the EX card like different values for theta, phi and eta. ETA= 0/180 gave no output whereas ETA=90/270 gave similar results. I also tried plotting in various planes (XY,YZ and XZ) by modifying RP card. But still no luck.

The multiple bounce effect around the corner is not coming into picture.

vineethajoy commented 6 years ago

As the thickness of the strip is very small compared to wavelength, I think the solvers are ultimately treating them as wires instead of strips. I tried a Dihedral two-plates problem (3D problem) and I am getting correct results. The plots are attached herewith. Problem dimensions: Problem Satement: 100cm*100cm f=600MHz Segment length=0.04 Segment radius=0.002

Ref: Fig.3.33 (P.no: 204) of the book "Complex electromagnetic problems and numerical simulation approaches" by Levent Sevgi. monostatic rcs bistatic rcs

ytakeyasu commented 6 years ago

The plot in my first post was "bi-static" RCS, but vineethajoy's expected plot was "mono-static" RCS. To get mono-static RCS from NEC2++'s bi-static RCS, 360 degree scattering pattern to be generated for each value of Phi and we have to extract the mono-static RCS value corresponding to the specific value of Phi amongst the 360 degree values.