How to implement a charge-discharge cycling BC with constraints in Coefficient Form PDE module?

April 16, 2020, 7:26 pm

≫ Next: Transport of Dilute Species (tds)

≪ Previous: Can you give me some advice about Phonon crystal plate wave?

$

0

0

Hi,

I am trying to build a lithium ion battery model using Coefficient Form PDE equations in Mathematics module, instead of using the in-built Lithium Ion Battery module.

For this, I want to implement a charge-discharge cycling boundary condition which involves 3 steps.

Step 1: Charge at a constant current (I_cell) upto a maximum cell cut-off voltage (V_max). Step 2: Then, charge at a constant voltage (V_max) until the cell current drops to a small value (I_min). Step 3: Next, discharge at constant current (-I_cell) until the cell voltage drops to a lower cut-off value (V_min). Repeat this sequence of steps for a given number of cycles.

It can easily be done using the in-built charge-discharge cycling boundary condition in the lithium ion battery module. However, I would like to implement the same BC in my model which is based on Mathematics module.

Can someone suggest me the approaches to acheive this?

Thanks in advance! Ganesh

---

Transport of Dilute Species (tds)

April 16, 2020, 11:23 pm

≫ Next: femtosecond laser ablation

≪ Previous: How to implement a charge-discharge cycling BC with constraints in Coefficient Form PDE module?

$

0

0

Hi there How can I modify the governing equation in the transport of dilute species to add an extra mass flux term induced by the temperature gradient?

J=D∇c + DT c∇T

where DT is the user-defined thermal diffusion coefficient.

femtosecond laser ablation

April 17, 2020, 12:39 am

≫ Next: Using previously computed solution as input for new simulation.

≪ Previous: Transport of Dilute Species (tds)

$

0

0

Hello... I want to do femtosecond laser ablation based on TTM(two temperature model). Frist laser gives the energy to the electron and that electron heated up. Now, this electron gives its energy to ion grid and relax quickly. I want to build this model but I don't know how to couple electron temperature and ion grid temperature  If someone knows about it please tell me. I want to see the temperature rise in the electron and iron grid. 

Using previously computed solution as input for new simulation.

April 17, 2020, 1:11 am

≫ Next: Cluster Batch Job Error

≪ Previous: femtosecond laser ablation

$

0

0

Hello,

I am doing a stationnary study of thermal strain using the solid mechanics module. I want to simulate thermal strain in a system with different material that have been deposited at different temperatures one by one.

For that I would like to do a simulation with the first two material, then use this solution as input for the three-material case, solve it, and so on.

Is that feasible? How can I use a previous solution as inital condition? The 'initial values' condition only allows to put initial constant displacement.

Thanks for your help

Cluster Batch Job Error

April 17, 2020, 5:01 am

≫ Next: Use 2D-Axisimetric component (not only geometry) inside a 3D component

≪ Previous: Using previously computed solution as input for new simulation.

$

0

0

Hi. I've recently gotten the cluster sweep running through a SLURM cluster Architecture from the COMSOL GUI. I can submit my jobs just fine, monitor, detach and attach jobs, and collect results at the end of the run. I can then view my results, save my file, close my file and re-open with success.

However, after a period of several hours, I re-open the saved models and get the following error when trying to view my results: Old model cannot be reconstructed. Do Clear Solution to expunge. Failed to load file for interpolation function.

I have not changed anything in terms of the cluster output generated status files, save files or anything else! Just a close... wait several hours... and re-open to this error message. Any ideas?

Best James

Use 2D-Axisimetric component (not only geometry) inside a 3D component

April 17, 2020, 5:02 am

≫ Next: Slider value.

≪ Previous: Cluster Batch Job Error

$

0

0

Hello, I am designing an array of speakers to analyze them within an enclosure with certain obstacles and to analyze the pressure field in the enclosure. The speakers are based on this example (https://www.comsol.com/model/lumped-loudspeaker-driver-using-a-lumped-mechanical-system-55901) and the enclosure would house an array of 15-20 speakers.

Is there a way to use the entire 2D-axisymmetric component within the enclosure? I need to be able to position the speakers at specific positions within the enclosure. Or some way to enter the speaker's calculated parameters as initial conditions of the enclosure (using the speaker's exterior field (sphere pressure source) )?

The 3D speaker over-calculates, but 2D-axisymmetric is very fast, so I want to calculate the 2D speakers within the 3D enclosure. Is there such a possibility or do I have to settle for everything in 3D (with work planes)?

Thanks

Slider value.

April 17, 2020, 5:54 am

≫ Next: Failed to find consistent initial values

≪ Previous: Use 2D-Axisimetric component (not only geometry) inside a 3D component

$

0

0

I have defined a slider in an application (in the Application Builder). I want it to control the time paramater (t) for which some plots will be shown. I have defined a method with a few lines like

Integer ind;
ind = app.form("form1").formObject("slider1").getValue();

The method getValue() is not valid (I wrote just "from the heart", I'm looking for the correct one).

I only need to GET the value of the slider (the step, in my case there are 20 in total) and use it to set the time value and update the plots. Thanks in advance.

Failed to find consistent initial values

April 17, 2020, 7:10 am

≫ Next: what does the error "para_sweep_2 is not monotone mean in comsol and how to solve this ?

≪ Previous: Slider value.

$

0

0

I am trying to run a simulation of a DC magnetic coil. The simulation runs with no issues as a stationary problem and gives a sensible result. I would like to run the analysis as a transient problem*. I get a message saying:

Failed to find consistent initial values Undefined value found NaN or Inf found when solving linear system Last time step is not converged.

I am using the same initial values as the stationary solution which works. I have checked the documentation and forum posts. There is a suggestion to run a stationary analysis and then initialise the transient analysis with the stationary result. I have set this up and get the exact same error message.

I have tried using a tiny dt. I have tried with an enormous dt. I have tried changing the prescribed initial conditions. Same error

I guess there is a condition or setting I need to apply to get this to work in transient mode but I am at a loss to work out what this could be - I am a fairly new COMSOL user.

Any advice would be welcomed.

• I understand there is no reason to analyse a DC magnetic field in a transient manner but I wish to couple the analysis with other transient physics problems later and want to test/validate the magnetic fields independently first.

---

what does the error "para_sweep_2 is not monotone mean in comsol and how to solve this ?

April 17, 2020, 8:54 am

≫ Next: Functions vs Variables

≪ Previous: Failed to find consistent initial values

$

0

0

what does the error "para_sweep_2 is not monotone mean in comsol and how to solve this ?

Functions vs Variables

April 17, 2020, 2:05 pm

≫ Next: MMA algorithm

≪ Previous: what does the error "para_sweep_2 is not monotone mean in comsol and how to solve this ?

$

0

0

Many calculation tasks can be handled either with a variable or a function. Are the calculations handled differently in the code? Are there advantages of one over the other, or are they just different ways of organization?

For example, a model might have a source term R that depends on many things. One could define R as R: myVariable and in a variables node: myVariable = thing1 * thing2^2 / sqrt(thing3)

Alternatively, one could use a function: R: myFunction(thing1, thing2, thing3) and then define the analytical function myFunction(a,b,c) = a * b^2 / sqrt(c)

Once the model is run, are these approaches different in efficiency or stability?

MMA algorithm

April 18, 2020, 3:35 am

≫ Next: Spatiotemporal varying relative permittivity based non-reciprocal transmission

≪ Previous: Functions vs Variables

$

0

0

Why does the MMA algorithm take the integral form for solving the objective / constraint function of the topology optimization problem? What is the theoretical basis behind it?

Spatiotemporal varying relative permittivity based non-reciprocal transmission

April 18, 2020, 7:01 am

≫ Next: Simulating a toroidal B-field inside a magnetic focusing horn

≪ Previous: MMA algorithm

$

0

0

I am trying to reproduce the paper: Complete optical isolation through indirect interband photonic transitions

The non-reciprocal transmission they have demonstrated in fdtd by their in house simulation code. I want to reproduce the same results by spatiotemporal varying the relative permittivity in comsol multiphysics. I am not getting how to do/implement it. Any helpful contructive suggestion is highly appreciated.

Simulating a toroidal B-field inside a magnetic focusing horn

April 18, 2020, 7:46 am

≫ Next: Modeling the transition from undeveloped flow to fully developed flow on a pipe

≪ Previous: Spatiotemporal varying relative permittivity based non-reciprocal transmission

$

0

0

Dear COMSOLers,

I am trying to study the magnetic field map produced inside a magentic horn (geometry photos and simplefied model attached). The current is flowing in the r-z-plane rather than azimutally (i.e. in phi), hence the field inside is a toroidal field along the phi-plane. I am defining the physics conditions through "Magnetic Field" module. Under "coil -> Geometry Analysis" I define the current input port. However, I am facing the follwoing problems: 1- When I add the surrounding Air domain, I am not able to define the current output port anymore, why? 2- I am getting the follwoing error message when try to solve my model "Numeric coil Coil 1 (coil1) not solved for. Solve it in a Coil Geometry Analysis step.", what dose it mean? 3- What I am missing in the model to make it running?

Regards, Tamer

Modeling the transition from undeveloped flow to fully developed flow on a pipe

April 18, 2020, 10:02 am

≫ Next: free licence for students ??

≪ Previous: Simulating a toroidal B-field inside a magnetic focusing horn

$

0

0

First of All, I appreciate you for taking the time to read my question.

I need some help finding the right path to start modeling the transition from undeveloped flow to fully developed flow on a pipe (in laminar flow region). For example, I want to model the flow from the beginning when the tube is empty until it is filled with the liquid. I presume this model has to be a time-dependent model, but I do not know which modules I have to implement other than laminar flow module. Also, I do not know if I had to use a "moving" mesh, with which I have zero experience. What should I try first on COMSOL?

free licence for students ??

April 18, 2020, 5:31 pm

≫ Next: Comsol 5.5 cannot be launched after installation

≪ Previous: Modeling the transition from undeveloped flow to fully developed flow on a pipe

$

0

0

does comsol provide a free student licence or free trail to their software ? is there an online software of comsole multiphasics ?

---

Comsol 5.5 cannot be launched after installation

April 18, 2020, 11:27 pm

≫ Next: Evaluate Impedance of a coil

≪ Previous: free licence for students ??

$

0

0

After i install it, there is no react when I click it. Only the update is working.

Evaluate Impedance of a coil

April 19, 2020, 1:43 am

≫ Next: Ac electric field

≪ Previous: Comsol 5.5 cannot be launched after installation

$

0

0

Hello All, I am trying to include impedance calculations in my model.

To keep it simple, let's consider my model to be AC/DC analysis with a single conductor coil. To realize it I followed the example of ecore_transformer, so I did the coil geometry analysis and then I did the analysis in time-dependent (frequency is 50 Hz). I managed to trace arrows of the density of current and the induction magnetic.

Now I'd like to evaluate the impedance and inductance of the coil. I know that I should use a global evaluation under derived values, but there aren't these choices in default calculations for coil parameters. I can evaluate only other parameters like current, voltage, and resistance in DC.

Could you tell me why I can't find other default calculation? Thank you very much for your help.

Ac electric field

April 19, 2020, 9:11 am

≫ Next: Piecewise constant function over a boundary

≪ Previous: Evaluate Impedance of a coil

$

0

0

Hi! I cannot understand how to apply an AC voltage to an electrode.

Piecewise constant function over a boundary

April 19, 2020, 2:15 pm

≫ Next: Variable time step for pulse features

≪ Previous: Ac electric field

$

0

0

Hello,

I have a 2D axisymmetrc model and I'm trying to define an input pressure on the edge as a piecewise constant function of the r-coordinate.

Here's what I want to have for each segment: P=p_0*exp(1i*phase_j), for r_j < r < r_(j+1) where p_0, phase_j, r_j and r_(j+1) are input parameters.

I have tried several approches, but there is a jump in the input pressure at the segment boundaries after meshing. I'm wondering what is the best way to do this?

Thank you in advance for your help.

Variable time step for pulse features

April 20, 2020, 12:53 am

≫ Next: Modeling no separation in Comsol

≪ Previous: Piecewise constant function over a boundary

$

0

0

Dear All,

I am trying to model a dc discharge in nitrogen with plasma module.

The time dependent solver works fine to a point with time steps of 10^-8s. However, at a cetain instance in time, there is a very sharp rise (or pulse) in one of the paramteres (current). At this point the solver gets upset and starts diverging. I am guessing that I need finer time steps at this point.

Is it possible to use variable time steps so that I can 'program' to select 'finer time intervals' when the sharp rise in current occurs and use regular time steps when the parameters are mostly uniform. How?