Complex antenna designs cannot be simulated without using electromagnetic solvers. HFSS, CST and WIPL-D are compared for accuracy and simulation time. Pro and cons of these softwares are stated based on the simulation results.
Complex antenna designs cannot be simulated without using electromagnetic solvers. HFSS, CST and WIPL-D are compared for accuracy and simulation time. Pro and cons of these softwares are stated based on the simulation results.
Complex antenna designs cannot be simulated without using electromagnetic solvers. HFSS, CST and WIPL-D are compared for accuracy and simulation time. Pro and cons of these softwares are stated based on the simulation results.
Advances in Computational Mathematics and its Applications (ACMA) 203
Vol. 1, No. 4, 2012, ISSN 2167-6356
Copyright World Science Publisher, United States www.worldsciencepublisher.org
Simulation Comparison between HFSS, CST and WIPL-D for Design of Dipole, Horn and Parabolic Reflector Antenna 1 Fahad Shamshad Muhammad Amin
Department of Electrical Engineering Satellite Development and Research Centre Institute of Space Technology, Islamabad, Pakistan Lahore, Pakistan
E-mail: 1 fahad.shamshad3@gmail.com
Abstract-- Antenna designs are becoming increasingly complex with recent advancement in the communication systems. These days it is very important to sort out the software which is best suited to our required antenna design. Complex antenna structures cannot be simulated without using these softwares, known as electromagnetic solvers. The important criteria of simulations are accuracy of results, time and how electrically large structures can be simulated. In this paper three famous electromagnetic solvers High Frequency Structure Simulator (HFSS), CST Microwave studio (Transient solver) and Wires Plates and Dielectrics (WIPL-D) are compared for accuracy and simulation time with simulations of dipole, Horn and parabolic reflector antenna. In the end pro and cons of these softwares are stated based on the simulation results. Keywords-- HFSS; CST; WIPL-D; Dipole; Horn; Parabolic reflector
1. Introduction
Several real world electromagnetic problems are not analytically solvable like scattering, radiation and transmission of energy via waveguides because of complexity of structures. Therefore use of computational electromagnetic field (CEM) [1] is inevitable. It models the interaction of electric and magnetic fields with physical objects and their environment to find the numerical approximation of Maxwells equations. Electromagnetic solvers are result of advancement in this field. These are specialized programs or softwares that solve the subset of Maxwells equations [2] directly. Three different electromagnetic solvers with different computational electromagnetic techniques are compared in this paper. Each technique has its advantages and drawbacks, the knowledge of which helps to get best result out of these softwares. Softwares that are compared here are HFSS stands for high frequency structure simulator and based on finite element method (FEM). [3] CST stands for computer simulation technology and based on finite domain time difference method (FDTD).(Only transient solver for CST will be considered here).[4] WIPL-D stands for wires plates and dielectrics and based on method of moments (MOM).[5] Central to all computational electromagnetic methods is the idea of discretizing (first calculated in small area by gridding and then combine these to produce overall result) some unknown electromagnetic property which is Electric field for FEM Electric and magnetic field for FDTD Surface current for MOM All simulations in this paper are performed on HP xw 8400 workstation Intel(R), Xenon(R) CPU, E5345@2.33 GHz and 3GB of RAM. [6][7][8]
2. Dipole antenna comparison in HFSS and WIPL-D Dipole has been simulated in HFFS and WIPL-D. Results and time taken by these solvers to simulate the Fahad Shamshad & Muhammad Amin, ACMA, Vol. 1, No. 4, pp. 203-207, 2012 204
structure has been obtained from respective simulation software. Gain and impedance plot of dipole in WIPL-D and HFSS are given in Figure 1 . Parameters of dipole are Frequency =300 MHz Transmission line length = /4 Dipole length = /2 Radius of wire used =0.1mm Frequency sweep =80 - 440 MHz Step size =20 MHz In HFSS dipole antenna is simulated inside the rectangular shape Radiation boundary. All parameters in both softwares are set to default except those who has been mentioned.
Figure 1(a). Dipole model in HFSS Figure 1(d). Dipole model in WIPL-D
Figure 1(b). 3D gain plot of dipole in HFSS Figure 1(e). 3D gain plot of dipole in WIPL-D
Figure 1(c). Impedance plot of dipole in HFSS Figure 1(f). Impedance plot of dipole on WIL-D
2.1 Observation HFSS is not suitable for wire like structures. Its gain value 2.77 dB is not close to the theoretical value . Fahad Shamshad & Muhammad Amin, ACMA, Vol. 1, No. 4, pp. 203-207, 2012 205
Moreover HFSS took 26 minutes to complete this simulation. The gain as calculated by the WIPL-D is 2.28 dB which is very close to theoretical value and almost the same as gain calculated by the Necwin plus software for the same structure. For this simulation WIPL-D took 31.20 seconds. Thus for wire like structures WIPL-D is about 50 times faster than HFSS. 3.1 DISH ANTENNA COMPARISON IN CST AND WIPL-D Dish is not suitable to design in HFFS due to its large electrical length. So Dish antenna has been designed in CST software with circular horn as a feed and results are measured . Then same dish antenna has been exported to WIPL-D Pro CAD software to compare the time taken by both software's to run the simulation. Simulated results are shown in Figure 2. . Dish parameters are Diameter =600mm Focal point =300mm Thickness =0.01mm Material =Perfect Electric Conductor Taper angle =53.1 o
Horn parameters are Upper radius of cone =30.88 mm Lower radius of cone =15 mm Waveguide length =31.25 mm Waveguide radius =15 mm Total length of horn =43.25 mm Edge taper =11 dB Material =Perfect Electric Conductor Thickness of horn sheet =0.001 mm Excited by =Wave port (In CST) Dipole (In WIPL-D)
Figure 2(a). Dish antenna model in CST Figure 2(c). Dish antenna model in CST
Fig.2(d) 3D gain plot of dish in CST Fig.2(d) 3D gain plot of dish in CST
Fahad Shamshad & Muhammad Amin, ACMA, Vol. 1, No. 4, pp. 203-207, 2012 206
Fig.2(e) 3D gain plot of dish in CST Fig.2(e) 3D gain plot of dish in WIPL-D
3.1. OBSERVATION
Above simulation was run in CST using transient solver and default mesh properties. CST took 1 hour 36 minutes to complete the above simulation while WIPL-D took 9 min 35 sec for the same task. Thus WIPL-D is about 10 times faster than CST for dish antenna simulations. Results are also very close to that of CST results. However CST is more versatile giving us much more options as compared to WIPL-D. For example CST give the phase centre of Horn antenna used in this simulation while WIPL- D have no such option. WIPL-D is best in terms of time and wire-plate like structures (like dipole and helix) while CST is more versatile and give us more options to see in depth of our results.
4. Horn antenna comparison in HFSS and CST Horn antenna has been compared in HFSS and CST and time taken by both the solvers to complete the above simulation has been noted Figure 3. Horn parameters are Upper radius of cone =30.88 mm Lower radius of cone =15 mm Waveguide length =31.25 mm Waveguide radius =15 mm Total length of horn =43.25 mm Edge taper =11 dB Material =Perfect Electric Conductor Thickness of horn sheet =0.1 mm Excited by =waveport
Figure 3(a). Model of horn antenna in CST Figure 3(b). Model of horn antenna in HFSS
Fahad Shamshad & Muhammad Amin, ACMA, Vol. 1, No. 4, pp. 203-207, 2012 207
Figure 3(c). 3D gain plot of horn in CST Figure 3(e). 3D gain plot of horn in HFSS
Figure 3(d). 2D gain plot of horn in CST Figure 3(f). 2D gain plot of horn in CST 4.1 Observation CST took about 4 minutes to complete the above simulation with default mesh properties while HFSS took 40 seconds for the same task. Results are almost same. So HFSS is better for narrow band frequency problems as horn is operated on a single frequency. 5. Conclusion Each software has its strong and weak points which one must know to get the best results out of these. After these simulations we conclude that WIPL-D is best suitable to electrically large structures like parabolic dish. CST transient solver is suitable for wide band antenna simulations and electrically large structures but take lot of memory and time. HFSS is not suitable for electrically large and wire like structures but best for narrow bands problems. References [1] Elliot P , The applied computational electromagnetic society, Antenna and propagation Magazine , IEEE, Volume 33, Issue 1, 3 rd August 2002, pp..18-19.
[2] Mathew N.O.Sadiku , Elements of Electromagnetics, Edition 4, Oxford University Press , Nov 2 , 2010.
[3] www.ansoft.com/products/hf/hfss/
[4] www.cst.com/
[5] www.wipl-d.com/
[6] Abdul Basit, Zain-ul-Aabidin Lodhi, Farhan Zafar, Waqar Aziz, Design Analysis of /4 Monopole VHF Ground Plane Antenna, , Advances in Electrical Engineering Systems (AEES), Vol.1, No. 3, 2012, pp. 146-151.
[7] Nabeel Arshad, Muhammad Ali J amal, Dur E Tabish, Saqib Saleem, Effect of Wireless Channel Parameters on Performance of Turbo Codes, Advances in Electrical Engineering Systems (AEES), Vol.1, No. 3, 2012, pp. 129-134.
[8] Fahad Shamshad, Usman J aved, Saqib Saleem, Qamar-ul-Islam, Physical Layer Aspects of 3GPPs Long TermEvolution (LTE), Advances in Computer Science and its Applications (ACSA),Vol. 2, No. 1, 2012, pp. 287-2