COMSOL 6.2 - Feature Specific Variables and Accessing the Solution at Different Phases

Feature Specific Variables and Accessing the Solution at Different Phases

When time periodic problems are solved, the solution contains variables that may have different values at different phase values.

In postprocessing, it is possible to use the built-in phase variable that can be specified as an input in a dataset. Variables in the physics interface that depend on the phase value typically end with the suffix _tpph. For example, the magnetic flux density norm, as a function of phase, can be accessed by writing mmtp.normB_tpph (assuming that the physics interface tag is mmtp).

It is also possible to access the value of a variable at a particular time by using the suffix _tpn, where n is an integer between 1 and the number of time frames nTP. For example, the magnetic flux density norm at phase zero is mmtp.normB_tp0 and the same variable at the phase π/2 is mmtp.normB_tp4 (assuming also that nTP = 16).

Phase variables exist, for example, for the following quantities:

•

Current density and electric field.

•

Magnetic flux density, magnetic field, and out-of-plane magnetic vector potential.

•

Feature specific postprocessing quantities such as voltages and force densities.

•

Lumped and integral quantities such as magnetic coenergy or axial torque.

Since forces and torques are specific to the node they originate from, some extra details must be taken into account for their variable names:

•

The total axial torque is added for each node, so the variable contains the tag of that node. For example, mmtp.rcon1.Tax_tpph.

•

The node distinguishes forces on different objects by an extra suffix that is specified as an input in the node. For example, the normal outward Maxwell stress tensor as a function of phase in the X direction becomes mmtp.nTX_1_tpph (given that the input is 1).

When phase-dependent variables are defined, a number of related variables become available by substituting the _tpph suffix. In Table 5-7, available variables and operators are shown, using the X-component of the magnetic flux density as an example.

Table 5-7: Possible suffixes and operators for variables in the Magnetic Machinery, Rotating, Time periodic interface.
Suffix	Description	Example	Operator
_tpph	Function of phase	mmtp.BX_tpph	mmtp.BX(phase)
_tp0	Time frame 1	mmtp.BX_tp0	mmtp.BX(0)
_tp1	Time frame 2	mmtp.BX_tp1	mmtp.BX(2*pi/mmtp.nTP)
_tpavg	Time periodic average	mmtp.BX_tpavg	mmtp.BX_fft(0)
_tprms	Time periodic root mean square	mmtp.BX_tprms	n.a
_tpdev	Time periodic standard deviation	mmtp.BX_tpdev	n.a
_fft1	Fourier component 1	mmtp.BX_fft1	mmtp.BX_fft(1)
_fft2	Fourier component 2	mmtp.BX_fft2	mmtp.BX_fft(2)

Axial torque and force density variables from the and features are also equipped with versions giving the maximum and minimum values over a time period.

In Table 5-7, the last column shows the definition of operators that can be used for accessing the variables at a given phase, instead of the phase value specified in the dataset. These operators are available both during model setup and in postprocessing, where they can be found in the section under in plot menus. It is important to note that the phase variable has a specific value (typically 0) in the physics. For accessing the phase at each time frame, the variable <phystag>.phaseTP can be used. However, this is not a true variable and cannot be plotted or included in other variable definitions. It may be used in the following inputs:

•

for and .

•

for .

•

for and .

•

for , , , , and .

As an example, given that the physics interface tag is mmtp, it is possible to apply a current density with a DC bias of 1 A/mm2 and a zero-phase harmonic ripple with an amplitude of 0.5 A/mm2, by entering the expression 1[A/mm^2]+0.5[A/mm^2]*cos(mmtp.phaseTP) in the feature.