Calculation

中文/English

模擬計算 (Calculation)

(Calculate Time)

(Time-Step) : This means the total calculate time step, and carefully, the time step need to achieve the steady state. The simulation flow chart describes below.

Courant S :

In FDTD, the time dt need to less than a condition, or the simulation system will be unstable. please refer to any the FDTD book.

FDTD:

2D: dt=S/cc/sqrt(1/dx^2+1/dz^2);

3D: dt=S/cc/sqrt(1/dx^2+1/dy^2+1/dz^2);

SFM :

2D:dt=S/cc/sqrt(1/dx^2+1/dz^2)*(1-Sinq);

3D:dt=S/cc/sqrt(1/dx^2+1/dy^2+1/dz^2)*(1-Sinq);

CompactFDTD:

2D: dt=S/cc/sqrt(1/dx^2+(beta/2)^2);

3D: dt=S/cc/sqrt(1/dx^2+1/dy^2+(beta/2)^2);

Total calculation time = time-step * dt

The figures at below show "arrive steady-state (left)" vs "not arrive steady-state (right)"

Spectrum : Pulse Wave

the calculation time of simulation system arrive steady-state

the calculation time of simulation system not arrive steady-state

This is a scattering/absorption spectrum of a metal cylinder, as the calculation time not longer enough, the Fourier integral time not enough, then will obtain the unstable result.

How long does the calculation time step is enough? Ans : usually take the ten times of diagonal length of the simulation system (10~20* diagonal length). For 2D example, if the total grid numbers X=100, Z=100, then the calculation time need larger than 2000 time step. If the simulation system include a high refractive index material, then the calculation time need to set longer.

Set Stop Condition

If the calculation time step default set at 500, the time step increments set 1000. As the system calculate time step arrive at 500, the program will check if the inner electric field maximum less than the set value. If no, continue calculate til 1500 time step and then check the inner electric field. Until the inner electric field less than set value or calculate time step arrive the "Maxium Time Setp".

The calculation will stop while the stop condition satisfy one of following

1. Time Step arrive "Max. Time Step"

2. The maximum Electric field value inside the system lower than "Max Cut-OFF filed"

linear to dB ==> 10*log10(linear)

0.1 = -10 dB,

0.01 = -20 dB

0.001= -30 dB

(Attention! this stop condition only work on 脈波(Pulse)

For example :

if set the stop condition like this

Then can get the spetrum at n=300, n=1300, n=2300 .........

The calculation will stop while the stop condition satisfy one of following

1. Time Step arrive "Max. Time Step"

2. The maximum Electric field value inside the system lower than "Max Cut-OFF filed"

(Output Files)

□ Output Steady State - Fields )(Poynting DFT) (Discrete Fourier Transform)

Pulse

TFSF-RCS

TFSF-RCS-Substrate

RCS - radar cross section calculate for Scattering/Absorption Spectrum

Extinction=Scattering + Absorption

★If only polarization 1 incident, then the output file

Result_RCS_Absorption.csv

Result_RCS_Absorption_Imag.csv

Result_RCS_Scattering.csv

Result_RCS_Scattering_Imag.csv

Result_RCS_CrossSection.csv

Result_RCS_CrossSection_Imag.csv

In "Result_RCS_CrossSection.csv", the Scattering is the sum of the six direction of "Result_RCS_Scattering.csv"。

In "Result_RCS_CrossSection.csv", the absorption is the sum of the six direction of "Result_RCS_Absorption.csv"

In "Result_RCS_CrossSection.csv", the extinction is the sum of the six direction of "Scattering+Absorption"

Imag : means the imaginary part.

★ If the incident ☑ polarization 2 checked, then the output file are

Result_RCS_Scattering_RCP_Imag.csv

Result_RCS_Scattering_RCP.csv

Result_RCS_Scattering_p1_Imag.csv

Result_RCS_Scattering_p1.csv

Result_RCS_Scattering_p2_Imag.csv

Result_RCS_Scattering_p2.csv

Result_RCS_Scattering_LCP_Imag.csv

Result_RCS_Scattering_LCP.csv

Result_RCS_CrossSection_RCP_Imag.csv

Result_RCS_CrossSection_RCP.csv

Result_RCS_CrossSection_p1_Imag.csv

Result_RCS_CrossSection_p1.csv

Result_RCS_CrossSection_p2_Imag.csv

Result_RCS_CrossSection_p2.csv

Result_RCS_CrossSection_LCP_Imag.csv

Result_RCS_CrossSection_LCP.csv

Result_RCS_Absorption_RCP_Imag.csv

Result_RCS_Absorption_RCP.csv

Result_RCS_Absorption_p1_Imag.csv

Result_RCS_Absorption_p1.csv

Result_RCS_Absorption_p2_Imag.csv

Result_RCS_Absorption_p2.csv

Result_RCS_Absorption_LCP_Imag.csv

Result_RCS_Absorption_LCP.csv

LCP (Left Circular Polarization)

RCP (Right Circular Polarization)

p1 (Polarization 1)

p2 (Polarization 2)

Result_RCS_Absorption_Imag.csv = Result_RCS_Absorption_p1_Imag.csv

Result_RCS_Absorption.csv = Result_RCS_Absorption_p1.csv

Result_RCS_Scattering.csv = Result_RCS_Scattering_p1.csv

Result_RCS_Scattering_Imag.csv = Result_RCS_Scattering_p1_Imag.csv

Result_RCS_CrossSection.csv = Result_RCS_CrossSection_p1.csv

Result_RCS_CrossSection_Imag.csv = Result_RCS_CrossSection_p1_Imag.csv

TFSF-Line

SoftSource

HardSource

SoftSource(Load)

HardSource(Load)

Reflection, Transmission file have Poynting integral of -x,+x,-y,+y,-z,+z six directions

In TRSpectrum are the sum of Transmission+Reflection (T+R) 及 Transmission, Reflection value

Imag : means imaginary part.

★If only polarization 1 incident, then the output file

Result_Reflection.csv

Result_Reflection_Imag.csv

Result_Transmission.csv

Result_Transmission_Imag.csv

Result_TRSpectrum.csv

Result_TRSpectrum_Imag.csv

In "Result_TRSpectrum.csv", the Transmissionis the sum of the six direction of "Result_Transmission.csv"

In "Result_TRSpectrum.csv", the Reflection is the sum of the six direction of "Result_Reflection.csv

In "Result_TRSpectrum.csv", the T+R is the sum of the six direction of "Transmission+Reflection"

★ If the incident ☑ polarization 2 checked, then the output file are

Result_TRSpectrum_RCP_Imag.csv

Result_TRSpectrum_RCP.csv

Result_TRSpectrum_p2_Imag.csv

Result_TRSpectrum_p2.csv

Result_TRSpectrum_p1_Imag.csv

Result_TRSpectrum_p1.csv

Result_TRSpectrum_LCP_Imag.csv

Result_TRSpectrum_LCP.csv

Result_Transmission_RCP_Imag.csv

Result_Transmission_RCP.csv

Result_Transmission_p2_Imag.csv

Result_Transmission_p2.csv

Result_Transmission_p1_Imag.csv

Result_Transmission_p1.csv

Result_Transmission_LCP_Imag.csv

Result_Transmission_LCP.csv

Result_Reflection_RCP_Imag.csv

Result_Reflection_RCP.csv

Result_Reflection_p2_Imag.csv

Result_Reflection_p2.csv

Result_Reflection_p1_Imag.csv

Result_Reflection_p1.csv

Result_Reflection_LCP_Imag.csv

Result_Reflection_LCP.csv

LCP (Left Circular Polarization)

RCP (Right Circular Polarization)

p1 (Polarization 1)

p2 (Polarization 2)

CW(Continue Wave)

★If only polarization 1 incident, then the output file

Result_FieldPattern.csv

★ If the incident ☑ polarization 2 checked, then the output file are

Result_FieldPattern.csv

Result_FieldPattern_Imag

□E,H,V,I (FFT : Fast Fourier Transform)

if checked, will output the file

Data_ProbeFields.csv

the simulation result record in the file

Result_ProbedFields.csv

□(Output Temp.- Fields)

output file

Result_Temp_Field_T_p1_0001.csv

Result_Temp_Field_T_p1_0002.csv

Result_Temp_Field_T_p1_0003.csv ...

if checked, the output loop condition is

for n=1:TotalStep

if（n>=start & n<=End）

if(n/Spare=.....0 be divided with no remainder)

output the temporary electric/magnetic field

end

★ If the incident ☑ polarization 2 checked, then the output file are

Result_Temp_Field_T_p1_0001.csv

Result_Temp_Field_T_p1_0002.csv

Result_Temp_Field_T_p1_0003.csv ...

Result_Temp_Field_T_p2_0001.csv

Result_Temp_Field_T_p2_0002.csv

Result_Temp_Field_T_p2_0003.csv ...

Using the【暫態分析(Temp Analysis)】to analysis the temporary electric/magnetic field.

□歸一化 Normalized (Normalized the Poynting integral to 1, recommend use for TFSF-Line incident. (will normalize the transmission + reflection = incident =1 )

Calculation

中文/English

模擬計算 (Calculation)

(Calculate Time)

(Time-Step) : This means the total calculate time step, and carefully, the time step need to achieve the steady state. The simulation flow chart describes below.

Courant S :

In FDTD, the time dt need to less than a condition, or the simulation system will be unstable. please refer to any the FDTD book.

FDTD:

2D: dt=S/cc/sqrt(1/dx^2+1/dz^2);

3D: dt=S/cc/sqrt(1/dx^2+1/dy^2+1/dz^2);

SFM :

2D:dt=S/cc/sqrt(1/dx^2+1/dz^2)*(1-Sinq);

3D:dt=S/cc/sqrt(1/dx^2+1/dy^2+1/dz^2)*(1-Sinq);

CompactFDTD:

2D: dt=S/cc/sqrt(1/dx^2+(beta/2)^2);

3D: dt=S/cc/sqrt(1/dx^2+1/dy^2+(beta/2)^2);

Total calculation time = time-step * dt

The figures at below show "arrive steady-state (left)" vs "not arrive steady-state (right)"

Spectrum : Pulse Wave

This is a scattering/absorption spectrum of a metal cylinder, as the calculation time not longer enough, the Fourier integral time not enough, then will obtain the unstable result.

Set Stop Condition

The calculation will stop while the stop condition satisfy one of following

1. Time Step arrive "Max. Time Step"

2. The maximum Electric field value inside the system lower than "Max Cut-OFF filed"

linear to dB ==> 10*log10(linear)

0.1 = -10 dB,

0.01 = -20 dB

0.001= -30 dB

(Attention! this stop condition only work on 脈波(Pulse)

For example :

if set the stop condition like this

Then can get the spetrum at n=300, n=1300, n=2300 .........

The calculation will stop while the stop condition satisfy one of following

1. Time Step arrive "Max. Time Step"

2. The maximum Electric field value inside the system lower than "Max Cut-OFF filed"

(Output Files)

□ Output Steady State - Fields )(Poynting DFT) (Discrete Fourier Transform)

Pulse

TFSF-RCS

TFSF-RCS-Substrate

TFSF-Line

SoftSource

HardSource

SoftSource(Load)

HardSource(Load)

CW(Continue Wave)

Transmission + Reflection = 1 (set the incident equal 1)

Parallel - Message Passing Interface (MPI)

Message Passing Interface (MPI)

Need to be divided with no remainder!!!

(example : grid number)

X:100

X=2 , 100/2 ..0 (correct)

X=3 , 100/3 ..1 (error)

X=4 , 100/4 ..0 (correct)

X=5 , 100/5 ..0 (correct)

x=6 , 100/6 ..4 (error)

計算(Cal.) - Calculate : start simulation。

停止(Stop) :stop simulation