Quantcast
Channel: Latest Discussions - COMSOL Forums
Viewing all 26323 articles
Browse latest View live

Build mesh (help)

$
0
0
Hello good, a doubt about meshes, to have if someone can help me. I have a doubt. I have created a model with three domains. When I create a mesh and select user controled mesh, it automatically creates a size for each domain, a free tetrahedral, corner refinement and boundary layers.

This procedure would be correct or it is better to create a free tetrahedral for each domain with its size.

I would appreciate if you could guide me on the correct procedure to create a mesh.

Greetings.

Number of particles in the space charge limited emission

$
0
0
Hi,

What is the meaning of the paramater "number of particles per release N" in the space charge limited emission node?(Comsol 5.2)

From Comsol's particle tracing module user guide:

Number of particles:
Enter the Number of particles per release N (dimensionless). The default value
is 1 ×10^3.

How i can use to model the space charge limited emission? How is related to the current emitted from a simple planar chatode?

Thanks

Attempt to evaluate nonintegral power of negative number.

$
0
0
Hi, I am new to comsol when creating a model with a default mesh the model works, but to check the independence of the mesh I want to simulate the model with different sizes of mesh, to introduce a mesh finer I get the following error:

Attempt to evaluate nonintegral power of negative number.
- Function: ^
Failed to evaluate variable.
- Variable: comp1.ht.hf1.fLow
- Defined as: (((0.3387*comp1.ht.hf1.kExtFluid)*(comp1.ht.hf1.Pr^0.33333333333333331))*(comp1.ht.hf1.Re^0.5))/(comp1.ht.hf1.L*((1+((0.046800000000000001/comp1.ht.hf1.Pr)^0.66666666666666663))^0.25))
Failed to evaluate expression.
- Expression: nojac(2*((comp1.ht.hf1.fLow*comp1.ht.hf1.hsExprLow)+(comp1.ht.hf1.fHigh*comp1.ht.hf1.hsExprHigh)))
Failed to evaluate variable.
- Variable: comp1.ht.hf1.h_expr
- Defined as: nojac(2*((comp1.ht.hf1.fLow*comp1.ht.hf1.hsExprLow)+(comp1.ht.hf1.fHigh*comp1.ht.hf1.hsExprHigh)))
Failed to evaluate variable.
- Variable: comp1.ht.hf1.h
- Defined as: comp1.ht.hf1.h_expr
Failed to evaluate variable.
- Variable: comp1.ht.hf1.q0
- Defined as: comp1.ht.hf1.h*(comp1.ht.hf1.Text-comp1.ht.hf1.Tvar)
Failed to evaluate expression.
- Expression: (comp1.ht.hf1.q0*comp1.ht.d)*dvol
Failed to evaluate Jacobian of expression.
- Expression: (comp1.ht.hf1.q0*test(comp1.ht.hf1.Tvar)*comp1.ht.d)*(dvol)


How can I know what the error is?

Error while using mphinterp function

$
0
0
Dear All,

please help me regarding mphinterp function.....

while using,

coordinate=[29970,10080]'
'' electric_field=mphinterp(model,'(es.Ey*es.ny)','coord',coordinate)''
in comsol 5.0 with matlab, i am always encountered with the error message

Error using mphinterp
mphinterp : Java exception occurred:
Exception:
com.comsol.util.exceptions.FlException: Failed to evaluate
expression
(rethrown as com.comsol.util.exceptions.FlException)
(rethrown as com.comsol.util.exceptions.FlException)
Messages:
Failed to evaluate expression

Failed to evaluate expression

Cannot evaluate expression.

Undefined variable.
- Variable: comp1.es.ny
- Geometry: geom1
- Domain: 2

Failed to evaluate expression.
- Expression: real((comp1.es.Ey*comp1.es.ny))

Failed to evaluate expression.
- Expression: real(comp1.es.Ey*comp1.es.ny)

Error in Plea_Linepractice60micW_1micro_final29_2dModel (line
239)
electric_field=mphinterp(model,'(es.Ey*es.ny)','coord',coordinate)

Simulation of semiconductor electric potential

$
0
0
Hi all,
Now I am simulating of electric potential. Simply I will solve Poisson equation in different medium. Support I have two medium and they are connected. So there is a interior boundary condition at the junction. There are charges in this junction so I have the Neumann boundary condition like this: D1-D2=ro. I think this is a discontinuous boundary condition. In my model, the space charge density is a function of electric potential V, which means this is a nonlinear Poisson equation. I select the general PDEs to solve this model, but the problem is that how to set the discontinuous boundary condition in the junction: D1-D2=ro.
I look for the USR ´S GUIDE, there are some words on PAGE 477 describe this, but the problem is that I really do not know how to set this discontinuous conditions.
Also in ACDC MODUL >electrostatics, I can set this discontinuous boundary condition, but the other problem is that I can only set space charge density as a constant, but can not set it as a function of electric potential V
I am using COMSOL 4.0
Thanks,
Yan

Convergence problem in RF module

$
0
0
I am doing some plasmonics simulation using silver nano particles in comsol. For silver particle convergence is either very very slow or there is no convergence in most cases for some specific wavelengths. The same model works fine for gold nano particles. Kindly help.

Elliptical cylinder

$
0
0
I am completely new to comsol, so please forgive me for posting a problem I suspect have a simple solution. I have tried to search for a solution myself, unsuccessfully.

In a microfluidic problem I need to use a pipe with an elliptical cross section. I have attempted to use a parametric surface and then convert it to a solid, but I get a warning that the output is empty.

Any help or tips would be much appreciated.

Particle Release Initial Velocity from File in Particle Tracing Module

$
0
0
I have a list of particle velocities in a text file that resulted from a Monte Carlo simulation. I'd like to use these values as input for initial particle velocities in a particle tracing simulation. What is the best way to go about this?

Helix with managable ends

$
0
0
Hi there. Im fairly new to Comsol. I am trying to model an evaporation champer of an absorption chiller. I descided to model the geometry in Comsol, since i have a load of import problems from Creo 2.0(study version). My problem at the moment(im sure i will have more problems) is, that the cooling cirkut cosist of two spirals made of cobber. Now i have used the Helix to create the spirals, the problem is now, how to i control where the end stops, so that i can stop it 180 degrees off of the inlet, and hopefully make a sweep to return to the flange where the pipe/spiral/helix starts. Hope someone can help me.
The pictures show how i would like the model to look like. The picture is from creo.
Kristian

Heat flux of a wheel break

$
0
0
Hey guys,

I am working on a project for university, and we want to do a research on how much heat is produced by breaking a bike in 6 seconds.

For this purpose we want to simulate a rotary motion. At the beging we have a velocity of v0. The breaking acceleration remains constant. Now we have the problem, that we don't know which Physics to use for this study.

Attached is a function of the angle and the anglevelocity depending on time.
The breaking pad is now producing heat through friction, which is converted in thermal energy.
It will be pleasant, if we could rotate the pad in the simulation.

I would be more than happy if I get any helps.

Strain and Compression plot

$
0
0
Hi Comsol Community,

i had a question to a structural mechanic simulation of a beam. I have a beam, which is fixed on the one side and has a boundary load at the other end. The load is 100N in the negative z-direction. Now i have simulated it and wanted to know the strain of the beam. I have plot the first principal strain and a arrow volume. But actually one side of the beam should be strained and the other compressed. I don't get any negative values in the color bar (compression). How do i get the compression?
A picture is attached.

Romol

Particles release inlet setting

$
0
0
Hi,

It is possible to set the number of particles per release from a boundary(inlet) dependent on the electric field (es.normE) present on the boundary itself?

Setting a time-dependent study, the elecrical field present on the chatode(inlet) could be variable for the possibility of a space-charge effect due to the emission of particles.

Smoothing a pulse made with Analytic function

$
0
0
Hello,

I have made a sine pulse with an analytic function.
My step function goes from 0 to 1at the location of 2e-4. My analytic function is "100*(sin(2*pi*10^6*t)-step1(t[1/s])*sin(2*pi*10^6*t))" and the lower and upper limits are 0 to 0.001.

The problem is when I zoom in the graph does not look like sine wave. Please see the attached image.
If I plot the same function with the frequency of 10KHz instead of 1MHz, it looks fine.
It seems that there is a problem with the graph resolution. Does anyone have any idea how to solve this?

Thanks,
Hamed

Modelling simple diffusion in water in a closed system

$
0
0
I am trying to characterize concentration gradients that form in a closed system as a function of time. I have a 2D simulation where in I have defined Initial values of concentration and used no flux conditions at the boundary. I am unable to think of any more boundary conditions since I am trying to solve for time -varying fick's diffusion equation. I have an initial condition that was given by initial values (1st order in time) and the no-fluxes at boundaries (2nd order in space in x and y-direction). Can anyone tell me if I am missing anything as the results I am getting are a little weird. I get linear gradients through my ROI which kind of doesn't make sense as the diffusion should slow down with time and it should follow more of an e^-x kind of curve (which I have successfully tested experimentally). Any insights would be really helpful.

PS: I tried attaching my model for further reference but I got an error message - 'file size error' (I am guessing my model file is too big to be uploaded). I can share my model with you via google dropbox if you think you need to take a look at the model!

particle counter not giving accurate results using .alpha

$
0
0
I have a particle counter set as a cylinder and I am setting the boundaries of the counter as a wall and having the condition set to freeze. I then emit 10 particles and all 10 hit the cylinder and freeze there as expected. However when I I use cpt.pcnt1.alpha and set time selection to last with integral using summation I get a value of 0.55 or something that is not quite right. In reality I expect 1.00. I have attached a screen shot as well as my .mph file. Any help is appreciated.

a Rigid Sphere travelling inside Fluides

$
0
0
Hello Everyone!
I want to simulate a rigid sphere inside flowing fluid. Physical Specifications followed:
1. This rigid sphere DO NOT DEFORM
2. The rigid sphere and fluid is fully coupled
2.1 fluid loading on the rigid sphere in order to make it move inside fluid
2.2 the velocity is transmitted to the fluid on the surface of rigid sphere
3. As the rigid sphere moves the mesh also moves

Which models should I adopt? I have tried
1. CFD model + Structure Model + ALE + FSI Fixed Geometry combinations. (Do not work as rigid domain is not accepted by solvers)
2. FSI model (The solid deforms, Not Acceptable)

Is there a method in COMSOL to simulate that? I am open to any advices, Thanks in advance!
Sincerely
Xiaoqiu HE

not converged in spite of the low error value

$
0
0

Hi,
I'm modeling a reactor with turbulent fluid flow and heat transfer and got really confused. The Nonlinear
solver reaches the error value lower than 1e-10 however the solver dose not stop the calculation. Could any one assist me with this problem?

Movement of a particle due to dielectrophoresis

$
0
0
Hello all,

I'm new to COMSOL and I'm trying to build a simple model to show how a particle would move in the presence of dielectrophoresis. I'm also trying to show the regions where there will be a positive and negative DEP.

Could someone please guide me in the right direction. Are there any tutorials or models available for this?
How do I start?

Random function independent of mesh

$
0
0
I'm running a time-dependent simulation using the general form PDE interface. I have to assign random numbers for my initial condition. Since the random number generator in COMSOL takes the spatial coordinates as input, the distribution of the initial condition is very mesh sensitive. Is there anyway that I can make the initial condition mesh independent? I was thinking about generating a uniform distribution of random numbers independent of mesh and then some how do the interpolation and find nodal values corresponding with that distribution. Is this possible? How?

Measurement of magnetic field density in the air gap of electrical machine

$
0
0
Hi all,

Can any one help me with the measurement of radial magnetic field density in the air gap of an electrical machine along a circle at each point with respect to time.

I tried to follow this step,
Data set --> More data sets--> average.---> selected boundary of the circle where I want to calculate magnetic field density

Now I go to

Derived values --> Point evaluation ----> data sets-->Average1---expression -->sqrt(rmm.Bx^2+rmm.By^2)


But I am not able to get the magnetic field density at each point along the circle.

Can any one help me?

With best regards,
Gaurav

Viewing all 26323 articles
Browse latest View live