SEMCAD Tutorial

Michael Burdumy
[email protected]

January 29, 2009

1 Two Metallic Spheres

The goal of this exercise is to become familiar with the low frequency solver
of SEMCAD. We’ll create two metallic spheres, assign different potentials
to them and let SEMCAD calculate the electric field. Be aware that this is
the first version of this tutorial, so any comments would be appreciated.

1.1 Creating the Model

1. Set the model unit to cm

2. Create a sphere by using the Sphere icon in the modeling toolbar. Set
the radius to 1cm.

3. Create a second sphere. Set the radius again to 1cm. Set the translation
to x = 17cm. You can also change the color of this sphere.

1.2 EM-Simulation
1.2.1 Simulation Settings

Next we have to set up the simulation settings.

1. Click on the EM-Simulations tab. There will already be a default

simulation entered.

1
1.2 EM-Simulation 1 TWO METALLIC SPHERES

Figure 1: Two metallic spheres

2. Change the solver in Settings → Solver to Low Frequency Solver.

3. Make sure the Solver Type is set to Electro Static. This is the default
setting.

1.2.2 Solid Regions

Now we have to change the material of the spheres and assign the potential.

1. Click on Solid Regions

2. Choose Sphere1 and select PEC/Metal as Type. Set the Potential

Amplitude to 1V.

3. Repeat the same procedure for Sphere2, this time choosing −1V as
the amplitude.

Note that the spheres count as sources when using the low frequency
solver, so we don’t have to set up any other sources.

2
1.2 EM-Simulation 1 TWO METALLIC SPHERES

Figure 2: Simulation Settings

Figure 3: Solid Regions

3
1.2 EM-Simulation 1 TWO METALLIC SPHERES

1.2.3 Boundaries

Figure 4: Boundaries

1. Now switch to Boundaries. We want to use Dirichlet.

2. Select X low and choose Dirichlet from the drop down box. Right
click on X low and copy and paste it to all other boundaries.

We don’t need to set up the sensors as SEMCAD implements an Overall

Field Sensor by default.

1.2.4 Grid

Now we have two metallic spheres, each with a different potential. Next
we have to set up the grid that SEMCAD uses to break down its calculations
to finite parts.

1. Click on Grid, select Sphere 1 and set the Mode to Geometrical. This
will tell SEMCAD to better refine the grid on the object. Select Sphere
1 and check Gang Axes. Set the Curvature Resolution to 0.5.

2. Repeat for Sphere 2.

4
1.3 Extracting and Interpreting the Results
1 TWO METALLIC SPHERES

3. Right click on Voxels and select Make Voxels. The Grid will now be
calculated.

Figure 5: Grid Settings

4. Start the simulation by clicking on the green start button in the toolbar.

1.3 Extracting and Interpreting the Results

Once the simulation has finished, you can view the results in a lot of
different ways. Here we will provide one example on how to view those
results.

1.3.1 Extracting

1. Click on Results in the EM Tab, choose Overall Sensors and E(x, y, z)

and select Slice Field View.

2. Click Next on the appearing screens.

3. Switch to the Viewers tab (if SEMCAD didn’t do so automatically).

Here you will be able to manipulate at which coordinate the model
is sliced and a lot of other options that can’t be explained in detail
here, but should be self-explaining.

5
1.3 Extracting and Interpreting the Results
1 TWO METALLIC SPHERES

Figure 6: Extracting Results

4. If you right click on Simulation 1 in the viewer tab you get several
options to view the simulation data differently, e.g. Surface Field
View.

Figure 7: Viewing the Results

1.3.2 Interpreting the results

You are free to try different settings for the amplitude, the size of the
spheres and the distance. Note that in order to change parameters like the

6
1.3 Extracting and Interpreting the Results
1 TWO METALLIC SPHERES

distance done in the modeling tab you have to delete any results and grid
already calculated.

7
 2 CAPACITOR

2 Capacitor
In this exercise we will create a model of a parallel-plate capacitor and
calculate its charge with the help of SEMCAD.

2.1 Creating the model

1. Create a new SEMCAD project.

2. Set the Model Unit to cm.

3. Create Brick 1 using the Brick icon in the modeling toolbar, use the
origin as the first corner, set the second corner to (15,15,0).

Figure 8: Parallel-Plate Capacitor

4. Right click on Brick 1, select copy and then paste the first brick.

5. Change the name of the copied brick to Brick 2. Set Translation Z to

3cm.

8
2.2 EM-Simulations 2 CAPACITOR

2.2 EM-Simulations
1. Switch from the Model tab to the EM-Simulations tab to enter the
simulation settings.

2. Delete the default simulation, then right click and choose New Sim-
ulation → Low Frequency to create a new low frequency simulation.

3. The default settings of the new simulation are Low Frequency Solver
with Solver Type set to Electro Static solver.

4. In Solid Regions select Brick 1, change the Type to PEC/Metal and

set the Potential Amplitude to 1V to assign a potential to the plate.

Figure 9: Solid Regions

5. Select Brick 2, change the Type to PEC/Metal and set the Potential
Amplitude to −1V.

6. It is not necessary to apply any settings to the sources as the Solid

regions act as such.

7. The Overall Field Sensor is used to record the field distribution. This
sensor is always used by default.

9
2.2 EM-Simulations 2 CAPACITOR

8. The Boundaries are automatically configured to Dirichlet.

9. Click on Grid to generate a default grid.

10. In Global Settings change the Padding Low and High for the Z-Axis
to 0.7m. The Padding adds extra space outside of the model so that
the Boundary conditions can be met.

11. Select Brick 1, check Gang Axes and set the Refine on Lower and
Upper Boundary to 0.6. Apply the same settings to Brick 2.

12. The resulting grid should have 0.964467 MCells.

Figure 10: Grid Settings

10
2.2 EM-Simulations 2 CAPACITOR

13. Make Voxels.

14. Start the simulation.

15. View the results.

16. Copy Simulation 1, rename it to Simulation 3 and insert a dielectricum

between the two plates. This can be done by changing the epsilon of
the background. Click on File→Materials..., copy Water(distilled)
and paste it on the Background.

17. Make Voxels.

18. Start the simulation.

19. View the results.

20. Now create a new project.Create the same parallel plate capacitor as
before, but this time set the distance between the plates to 3mm.

21. Calculate a fine grid.

22. Start the simulation and view the results.

Figure 11: Viewing the results

11
2.3 Extracting and Interpreting the Results 2 CAPACITOR

2.3 Extracting and Interpreting the Results

Calculate the theoretical values of Q for all settings ( 0 = 8.8542 · 10−12 F/m).
Given that |D| = Q/A for a parallel plate capacitor, calculate the value
of Q by using SEMCAD’s numeric results for D. Extract D(x, y, z, f 0).
You can extract the results of SEMCAD by using a provided script on
the website. The script is called QDA.py. Open it by clicking on SEMCAD
X Script → Open Script and press F5. Why are the results of the numerical
and theoretical calculations different?
Calculate the theoretical value of E = U/d. Compare this with the
maximum of the numerically calculated E.

12
 3 PLANE WAVE

3 Plane wave
This exercise will deal with a plane wave. SEMCAD provides easy ways
to model plane waves in 3D and analyzing those models for different
frequencies.

3.1 Modeling and Simulating

1. Choose the Plane Wave Source icon. Enter the points (-150, -150,
-150) and (150, 150,150) for the First and Second Corner respectively.

2. Click Done: Plane Wave Source 1 is added to the model.

3. Set the Frequency to 3000 MHz for the harmonic simulation.

4. Select the Plane Wave Source 1: the E, H, and k orientation and the
signal Scope is shown. For this simulation the k vector of the Plane
Wave will be oriented in positive Z-direction, so it is not necessary to
change phi, theta or psi.

5. Make Voxels and Simulate.

6. Save the project under a different name, excluding the results.

7. Insert a dielectricum, a cube of a size smaller than the cube you’ve

created in the beginning. As with the plate capacitor, get the param-
eters of distilled water from the material list and paste those settings
to the small cube.

8. Make Voxels and Simulate. You’ll have to change the grid size by
changing the scale factors. Be sure to stay under 1 Mio. Voxels.

9. Extract the E-Field and look at the results of both simulations.

3.2 Questions
How would you explain the differences in the 2 simulations?

13
3.2 Questions 3 PLANE WAVE

Figure 12: Plane Wave

Figure 13: Plane Wave with dielectricum inserted

14
 4 TELEPHONE AND HEAD

4 Telephone and Head

In this exercise we will create a model of a phone, import a 3D model of
a head, assign parameters to the materials and extract results from this
simulation setting.

Figure 14: Head and phone model

4.1 Modeling the Case of the Phone

1. Create a Group by clicking on the Group icon and rename this group
phone.

2. Create two points p1(20, -16, -130) and p2(-20, 0, 10) by clicking on
the point icon.

3. Click on the brick icon and select p1 and p2 to build the brick. Rename
the brick Case.

15
4.2 Modeling of the Monopole Antenna 4 TELEPHONE AND HEAD

Figure 15: 2 Points

Figure 16: Phone Case

4.2 Modeling of the Monopole Antenna

A 1mm gap is left between the phone case and the antenna where an
edge source can be placed. The antenna is to be modeled as a 79mm long

16
4.3 Adding the Source and the Sensors 4 TELEPHONE AND HEAD

cylinder with a radius of 1.25 mm.

1. Create three points p3(-12, -8, 11), p4(-12, -6.75, 11) and p5(-12, -8,
90), and click on the Arc icon to create a circle. Select p3 as the center
and p4 as the start and end point on the circle.

2. Click on the Extrude icon and select the previously created circle (Arc
1) as the profile to extrude. Select the points p3 and p5 to create the
cylindrically swept body. Rename it Antenna.

3. Create the point p6(0, 0, 0) and rename it Speaker Point. This point
will be used later.

4.3 Adding the Source and the Sensors

1. Create the point p7(-12, -8, 10), which is the intersection of the axis of
the antenna with the top of the case.

2. Create the Edge Source by clicking on the corresponding icon and

select p3 and p7 as the termination points.

Figure 17: Complete Phone with Edge Source

17
4.4 Importing the SAM phantom model 4 TELEPHONE AND HEAD

Note that it is not necessary to add any sensors, as the necessary sensors
are added by default, the Sensor of Edge Source, Overall Field Sensor,
and Far Field Sensor.

4.4 Importing the SAM phantom model

1. Select the root (Model) in the modeling tree, generate a new modeling
group for the phantom and activate it.

2. Import the SAM Phantom model via the File → Import Model →
ACIS SAT File menu command. The zipfile containing the phantom
model (SAM Head.zip) is located in the directory in which the tutorial
projects are installed. If there are no tutorials installed on your ma-
chine download the model from http://www.semcad.com/sup down-
loads.html. This model contains the three predefined reference points
at the openings of the left and right auditory canals and at the mouth
of the phantom.

Figure 18: Head Model

18
4.4 Importing the SAM phantom model 4 TELEPHONE AND HEAD

4.4.1 Moving and Rotating the head

You need to move the phantom near the generic phone such that the left
ear reference point of the phantom coincides with the Speaker Point of the
phone. In order to do so, select the modeling group which contains the
phantom. Selecting a modeling group instead of a single item will apply
any operation, such as moving, to all items contained in that group. Click
on Movein the modeling toolbar and then on the left ear reference point of
the phantom to designate it as the starting point of the translation vector.
The end point will be the origin by default, where the speaker of the phone
is located. You can complete the move operation by clicking on the Done
button.

Figure 19: Translation of head phantom

In order to move the phantom and the phone into the desired positions,
the head will be rotated instead of the phone. This is to avoid staircasing
errors at the case and at the antenna of the phone, since in this way they
will remain aligned with the axes of the coordinate system.
1. Select the model group which contains the phantom and click on
Rotate in order to rotate it. In the Rotate Selected Parts dialog

19
4.5 EM-Simulations 4 TELEPHONE AND HEAD

window you are required to select two points for the rotation axis.
Here, you can either choose the two reference points at the phantom’s
ears, or, alternatively, enter two arbitrary points on the y-axis (see
Figure 20). Finally, enter an angle of -61◦ in the dialog window and
click on Done to finalize the operation.

2. Rotate the phantom around the z-axis (e.g. (0, 0, 0) and (0, 0, 100)) by
an angle of −3.9◦ . Proceed in the same way as in the previous step.
Be sure that the profile of the phantom’s ear is aligned with the case
of the phone.

3. Now, bring the phantom into ”touch position”. Select two points on
the x-axis (e.g. (0, 0, 0) and (100, 0, 0)) and rotate the phantom by −4◦ .

Figure 20: Rotation of head phantom

4.5 EM-Simulations
1. Keep the Excitation Mode as Harmonic.

20
4.5 EM-Simulations 4 TELEPHONE AND HEAD

2. Keep the Frequency of the simulation at 900MHz and the Simulation

Time at 10 periods.

3. Select the case of the phone and change the Type in the Materials
window to PEC.

4. Select the antenna of the phone and set it to PEC.

5. Select the SAM Liquid, and set the Rel. Permittivity to 41.5 and the
Electrical Conductivity to 0.97S/m.

6. Select the SAM Shell, and set the Rel. Permittivity to 3.7. Repeat for
the Reference Points by either manually entering the value, or using
the Copy command on the SAM Shell and then the Paste command
on the Reference Points.

7. Set the Absorption Strength of the X low boundary to Low.

8. Right-click on the X low boundary, choose Copy, then multi-select

the remaining boundaries and Paste the settings.

9. Select Grid: the grid is automatically calculated. To obtain an accu-

rate representation of e.g. the impedance it is necessary to refine the
grid slightly.

10. Select the Antenna either by selecting it directly in the Grid Regions
List or using the Picker and clicking on it in the 3D modeling window.
For the local grid settings set the Refinement on Lower and Upper
Boundary to 0.06 for the Z-Axis.

11. Select the Case, check Gang Axes and set the Refinement on Lower
and Upper Boundary to 0.15.

12. Select the SAM Liquid either by selecting it directly in the Grid
Regions List or using the Picker and clicking on it in the 3D modeling
window. Change the Mode to Bounding Box. This will introduce
Baselines, resulting in a finer grid in the region the maximum SAR

21
4.5 EM-Simulations 4 TELEPHONE AND HEAD

(a) Antenna (b) Case

Figure 21: Grid Settings for phone

is expected in, giving a more accurate result. The total grid should
now have 0.84048 MCells.

Figure 22: Grid Settings Head Phantom

13. Copy and Paste Harmonic Simulation 1.

14. Rename the copied simulation Harmonic Simulation 2.

15. Change the frequency to 2000MHz.

16. Click on Grid. There will be too many voxels now, the light version
of SEMCAD can only handle 1Mio. Voxels. So change the Overall
Scale Factor to 2. This should result in a voxel count slightly below 1
Mio..

17. Make Voxels for both simulations.

22
4.6 Extracting and Interpreting Results 4 TELEPHONE AND HEAD

18. Click on Run Batch and choose both simulations.

19. Lean back and wait for the simulation to end.

4.6 Extracting and Interpreting Results

4.6.1 Defining SAR

The ”Specific absorption rate” (SAR) is a measure of the rate at which

radio frequency (RF) energy is absorbed by the body when exposed to
radio-frequency electromagnetic field. It is defined as the power absorbed
per mass of tissue and has units of watts per kilogram. SAR is usually
averaged either over the whole body, or over a small sample volume
(typically 1g or 10g of tissue). The value cited is then the maximum level
measured in the body part studied over the stated volume or mass. It can
be calculated from the electric field within the tissue as:

σ|E|2
SAR = (1)
2ρ

In Europe, the limit for a cell phone is 2 W/kg, averaged over a volume
of 10 grams of tissue.1

4.6.2 Calculating SAR

In order to calculate the average SAR follow these steps.

1. Start with Harmonic Simulation 1.

2. Right click on the Spatial Peak SAR[IEEE-1529] icon of the Overall

Field sensor results in the project tree and select the Slice Field Viewer.

3. The SAR distribution can be averaged over an arbitrary mass accord-

ing to IEEE 1529, set it to 10g.

4. Select the solids to be considered for the SAR computation. Select

SAM Liquid. Click Next to start the computation.
1
Source: Wikipedia

23
4.6 Extracting and Interpreting Results 4 TELEPHONE AND HEAD

Figure 23: Spatial Peak SAR

5. Choose view details. This will show you the Averaged SAR based
on IEEE-1529. The value should be 0.0139817 W/kg

6. Now switch to Harmonic Simulation 2 and extract the results in the

same way. Compare the Average SAR. How would you explain the
results?

24