Quantcast
Channel: Latest Discussions - COMSOL Forums
Viewing all 27024 articles
Browse latest View live
↧

Electromagnetic heating in a 2D axissymmetric AC model

April 22, 2020, 3:52 am
$
0
0

Hello,

I am currently working on a project to simulate the electric field behaviour at different temperatures. The application is shown in the attached image. I already simulated it in DC with the temperature dependence as 2D axissymmetric model. Therefore I used the "electric currents" and "heat transfer in solids" physic coupled with "electromagnetic heating". As study step I used "stationary". Now I want to do the same for AC. I tried using the "frequency-stationary", but I always get an error while computing.(First few rows of the error code are at the end of the post) I couldn't figure out why the solver gets that error. Hence I looked for a different way of solving it and changed "electric currents" to "magnetic and electric fields" because I saw that in an example. However, in this new setup I can't apply the voltage on the electrode or the ground.

I am pretty sure the simulation should work with my first attempt, does anyone know what I did wrong or what I can/should change?

Thank you!

Error Code: - Feature: Stationary Solver 1 (sol1/s1) Attempt to evaluate negative power of zero. - Function: ^ Failed to evaluate temporary symbolic derivative variable. - Variable: comp1.ec.normE@VDN${realdot@3} - Defined as: ((realdot(comp1.ec.Er,comp1.ec.Er)+realdot(comp1.ec.Ez,comp1.ec.Ez))^(-0.5))0.5 ...

↧
Search

FSI vs Acoustic-stucture interaction

April 22, 2020, 10:11 am
$
0
0

I try to understand the difference between FSI and acoustic-structure interaction (including thermoviscous acoustics and pressure acoustics).

After exploring some blogs articles and examples, I have an impression that FSI is more "complete" since it starts from the lower level-- the properties of the solid and the fluid, with moving mesh in general. So it can handle all kinds of fluid structure interaction problems, including the ones in which the sturcture vibrates in fairly high frequency (~kHz). As long as the time step is small enough (1/nth of a period of the vibration), FSI time depent solver should be able to find all the behavior of this system, including dampling. The only practical problem is that it can be expensive and be even inpractical for very high frequency vibration, and must be carefully setup when dealing with fast changing conditions, like shocks.

Thermoviscous acoustics and structure coupling, howerer, is a weaker solution compared to the FSI one, as it doesn't have moving mesh. So, it can't handle motions or vibrations with large amplitude. But it is more efficient than FSI for vibration scenarios, as long as the vibration amplitude is not large enough. This case condisers both added mass due to the fliud to the structure and the thermal damping due to the fliud.

The pressure acoustic and structure coupling is further simplied thermoviscous acoustics structure coupling, because it simply drops the thermal damping part. For macrosopic systems, that parts is in genearl insignificant.

If the above is in general correct, I had a laminar flow--solid mechanics coupling problem. And I tried to find the eigenfrequency of the coupled system using eigenfrequency solver, but it seems the result is not much different from the same system without the laminar flow part. Did I choose the wrong coupling (FSI)? or did I choose the wrong, laminar flow?

↧
↧

definiing an elastic modulus that depends upon location in axisymmetric problem

April 22, 2020, 2:30 pm
$
0
0

If you check the 1st attached image, it shows a multilayered portion of a circular "shell". In the middle layer (shaded in blue), I want to define a gradient in modulus such that the modulus decreases as one approaches the outside layer in a direction normal to the layer (I could call this the "radial" direction). I want to define a formula such that at each point in the blue layer, the modulus is computed in terms of the coordinates of the point as well as the values of the modulus values of the outer and inner layers, plus the thickness of the outer layer. I have done such a thing successfully in a simpler non-axisymmetric problem (see 2nd attached image of an example rod with a gradient in modulus defined in the middle; note equation for modulus in gradient region), but I am unsure how to translate this approach to the axisymmetric problem. I think my question boils down to this: When entering the formula for the modulus into "Materials", do I define it in terms of (r,z) coordinates or (x,y,z) coordinates? Thanks.

↧

ntfluxInt - kinks in plot

April 22, 2020, 4:57 pm
$
0
0

I want to simulate a block of aluminum that has a certain temperature and radiates to an empty space. So I set up a COMSOL simulation:

  • Physics: heat transfer in solids, time dependent

  • I created a simple block

  • added a material (Aluminum 6063-T83)

  • set the temperatures (delta is 10K)

  • turned on surface-to-surface radiation

  • all surfaces of the block are diffuse surface with a given epsilon of 0.5

  • mesh is physics controlled (coarser)

When I do so I get reasonable results (I verified them by hand), but I noticed some kinks in the temperature plot. I investigated that they result from even bigger kinks in the ntfluxInt global value (see graph). I already varied the mesh sizes, time steps, material, and several other parameters, but none of that removes the kinks.

So my question is: where do these kinks come from? Are they simulation errors? Is there any way to get rid of them?

Thanks for your help

↧

How can I customize piezoelectric materials correctly？？

April 22, 2020, 7:03 pm
$
0
0

Hello，everyone， Genius and wisdom. Recently, piezoelectric materials have been customized, but the non-convergence and inconsistency of initial value are occured in the simulation model. Then I modified the properties of piezoelectric materials in the comsol material library（Elastic matrix Coupling matrix），these problems will also occur. The number of decimal places of the parameters in the material library is consistent， see picture. How can I customize materials correctly.

↧
↧

Help with the model of wicking in a paper strip

April 22, 2020, 9:59 pm
$
0
0

Hi,

I was trying to use the model provided on Comsol website to work on a time dependent wicking problem in a paper strip where the width is a not constant value (the shape is a 6cm-width rectangular wicking pad with a 1cm-width leg below).

Here's the link I found from comsol：

https://www.comsol.com/model/wicking-in-a-paper-strip-79061

I used same boundary conditions as the above link and I think my mesh is OK, but there's always an error after running for a while. For some unknown reason, the s value ( the effective saturation) of air (or maybe water) becomes negative.

Except for the geometry, all I changed were the parameters like properties of air and water. Anyone can give me any idea about which part goes wrong and why s becomes negative?

Any help is appreciated. Thanks a lot.

Jing

↧

Help with BC and contact pairs

April 23, 2020, 12:12 am
$
0
0

Hello!

I'm new in Comsol and trying to create a simplified track model in 2D, where I'm applying load on the top (of the sleeper). The model has 5 layers: sleeper, ballast, asphalt, sub-ballast, and subgrade. On the top and bottom of the asphalt layer, I want to put a contact pair with friction applied. But it is only possible to compute when the side of the layers have fixed boundary conditions. But there should not be any fixed BC for the first 4 layers, as there is nothing like that in reality (nothing on the sides of the structure), except for the bottom layer. When I try to compute without fixed BC on the top 4 layers, it doesn't compute (getting errors) because of incorrect BC. Do you have any ideas on how to fix this, what kind of BC (that aren't fixed) can I apply, so it will work in my case? The link to the model (Comsol 5.5) is: https://www.dropbox.com/s/d9p5sxextfik61w/NEW_MODEL.mph?dl=0 Thank you in advance!

Kind regards, Aleksandra

↧

Material properties

April 23, 2020, 3:37 am
$
0
0

How to apply materials properties for composite block?

↧
Search

Help with mesh building for a simple 2D geometry

April 23, 2020, 11:28 am
$
0
0

Hi,

I was working on a model that couples the physics module Phase transport in porous media and Darcy's Law together. In Comsol Application Gallery, there's an example ' Wicking in a paper strip' where the width is constant. I was trying to study the 2D effect of this process, so I changed the geometry. But there's always error occurs, I'm wondering if it's the problem with my mesh.

I've attached two screenshots, one is the example from Comsol (paper strip), the other one is the mesh I built. Could anyone take a look and let me know whether this mesh looks appropriate?

Thanks a lot.

Jing

↧
↧

How to calculate/export left eigenvectors within the eigenvalue solver?

April 23, 2020, 1:49 pm
$
0
0

Dear COMSOL users/employees,

after quite some research, I still haven't found a way to calculate and export the left eigenvector to Matlab when performing an eigenvalue study in COMSOL.

Some background on the issue: Currently I am trying to perform reduced order model simulations in time domain for acoustic wave propagation. The governing equations are implemented using the Weak Form PDE Physics in COMSOL 4.4 (newer releases are available but rarely used). Therefore, in a first step I execute an eigenvalue study and extract the system matrices, the eigenvalues and the (right) eigenvectors via the COMSOL LiveLink for Matlab using the functions mphmatrix and mphgetu. The model order reduction is then performed in a second step in Matlab. However, for the model order reduction the right AND left eigenvectors are required but only the right eigenvector can be extracted in a straightforward manner. So far, transposing the system matrices in Matlab and returning them to COMSOL via mphinputmatrix did the trick to get the left eigenvector. For systems with large numbers of degrees of freedom, however, the function mphinputmatrix fails due to limitations of the Java Heap Memory (OutOfMemoryError: Java heap space) in Matlab, although the Java heap space was already increased to a maximum value of 8GB. As the manipulation of the load vector is more convenient to do in Matlab, the reduced order model simulations are preferred to also being performed in Matlab.

Long story short: Is there any possibility to extract the left eigenvector calculated in COMSOL to Matlab? Do you have any other suggestions to solve that issue?

Any help is kindly appreciated.

Best regards, Gerrit

↧

How do you couple the modules “Transport of Diluted Species and Transport of Concentrated Species”?

April 23, 2020, 4:49 pm
$
0
0

Hello, Thank you so much for reading my post.

I have to domains that are separated for a thin region that will serve as a semi-permeable membrane. However, I want to specify in one of the domains "“Transport of Diluted Species" while in the other domain I want to specify "Transport of Concentrated Species.” I want the diffusion through the semi-permeable region of the chemical species to depent on both mechanisms (modules). In one region the chemical species are at diluted concentration while the ohter region the chemical specias are at saturated concentrations.

Thank you,

↧

Frequency domain modal analysis for piezoelectric device coupled with electrical circuit

April 23, 2020, 5:45 pm
$
0
0

Hi all,

What I am trying to do is to build a piezoelectric device with electricalstatic and solid mechanics interface and I coupled it with an electrical circuit interface. In the study I want to do a frequency domain modal analysis. But it seems to be stuck in the eigenfrequency study because it just takes too long and no progressing.

In the eigenfrequency step there is a orange triangle beside the electrical circuit step, just wonder what is wrong in there?

↧

Problem with lower coil excitation frequency in Magnetic Field mf module

April 23, 2020, 9:45 pm
$
0
0

Dear All,

I'm working on the eddy current sensor simulation with the different target material. According to my model, I cannot apply the coil excitation frequency lower than 20 kHz, for example, 1,000 Hz. The results do not converge.

Please advise how to adjust the simulation model to solve this problem.

Many thanks,

Siripong

↧
↧

Failed to find a solution. Maximum number of Newton iterations reached for Laminar Flow generated by a Microswimmer

April 23, 2020, 10:07 pm
$
0
0

Hi All,

Thank you so much for taking the time to read this. I am currently encountering some issues with finding a solution for my simulation. For some background, I have a 2D simulation of an acoustically actuated microswimmer in a tank filled with water that generates 2nd order steady streaming. The microswimmer is actuated by a piezoelectric plate underneath at frequency f0 via Thermoviscous Acoustic-Structure Boundary, and the resulting acoustic velocity is found via Frequency Domain. Then these results are linked and used in a Stationary study for Laminar flow to obtain the 2nd order steady streaming caused by the oscillation of the microswimmer tail. While these simulations work/converge for microswimmer eigenfrequencies corresponding to modes up to the 5th order, the stationary study starts failing to converge at f0 = 24.104kHz, which is the 6th order eigenfrequency of the microswimmer. I am very curious on why this is so and would greatly appreciate any help. I have also attached my simulation. Thanks!

↧

How to solve the errors in the comsol log file?

April 24, 2020, 1:35 am
$
0
0

Hi, everyone. I used the " Plane wave expansion" approach to set the tightly focused gaussian beam, but when I increased the wave vector count to 100, an error occurred and the log file showed "java.lang.StackOverflowError". I have no idea about this, can you help me ? Best wishes Sincerely, Miao Peng

↧
Search

who know the ''ec.intWe'' means?

April 24, 2020, 1:55 am
$
0
0

In AC/DC，in Electric Currents，in Frequency Domain，in Global Evaluation，what is the exactly expression of ''ec.intWe''?

↧

shear stress of ZnO and Cu thin films

April 24, 2020, 3:28 am
$
0
0

can anyone please post the shear stress phenomenon when ZnO and Cu materials used as thin film platform? please respond asap.. thank you.

↧
↧

Modelling Fluorescence Enhancement Near A Gold Nano-Particle

April 24, 2020, 5:02 am
$
0
0

Hi ! Can anyone help me in modelling fluorescence enhancement near a gold nano-particle. I've tried a few times but couldn't succeed. Has anyone done it before?

Please let me know if anyone has any clue regarding this.

Thanks, Priyadarshi

↧

Double-layer charging on an electrode with an ion-exchange membrane.

April 24, 2020, 8:40 am
$
0
0

Hello,

Disclaimer: I am new to COMSOL, this is the first thing I am trying to model with the software.

I am trying to model an electrode assembly to see what would happen when I place an ion-exchange membrane (lets say a cation-exchange membrane like Nafion), close to an electrode surface, and then apply a voltage, without any electrochemical reactions taking place. Only trying to charge the surface. See the attached picture, which should hopefully be self explanatory.

Physically, what I am imagining should happen is that cations should be pulled through from the potential field, and move to the surface, causing a low concentration at the Nafion membrane, and continue to do this until the cationic charge balancing the electrode potential, and the rest of the volume fills up to balance the bulk concentraion of cation. But maybe I am completely wrong... so I want to model it!

Attached is my attempt so far, but it doesnt work, or make any sense, so hoping I can get some help!

↧

Add mass to end of cantilever beam

April 24, 2020, 11:58 am
$
0
0

Hi, I want to get the natural frequency of a cantilever beam with one end fixed and the other end with a mass. I tried both adding a mass at the end of the beam, and applying a point load. Both approaches give me lower eigenfrequencies (1-2 order of magnitude away from the numerical result). I saw on the forum from previous years that adding point load or boundary load does not affect eigen frequencies, any suggestions on how to continue? Thank you!

↧
Viewing all 27024 articles
Browse latest View live
Search