Quantcast
Channel: Latest Discussions - COMSOL Forums
Viewing all 26527 articles
Browse latest View live
↧

PML error

April 11, 2013, 4:28 am
$
0
0
Hello

I am attempting to use a PML in the acoustics-piezoelectric interaction frequency domain module. I choose cartesian coordinates type as I am working with parallelpiped geometry. When I attempt to compute I get a "could not find any cartesian regions" error.

I have seen a thread wherin it was suggested the the units used must be meters and I have ensured this to no avail.

Any suggestions? Thank you
↧
Search

Waveguide Verification time domain (TE10)

April 11, 2013, 5:52 am
$
0
0
Hello!

I am testing COMSOL with a simple 2D model of a waveguide for the TE10 mode (physics: transient electromagnetic waves, study: time dependent), i am using the scattering boundary condition for the entry and exit of the wave.

When the waveguide depth is a multiple of the wavelength, the solution converges to the analytical solution after a few cycles, but when it's not, no matter how small the element size or the time step are, it will converge to a solution with an error in amplitude (in same cases very high, the amplitude I am imposing is E0= 1 V/m, and somethimes it converges to an amplitude of E0=1.4 V/m). I have tested for smaller and longer waveguides but the conclusion was always the same. Anyone has an explanation for this?

I have attached four images, two with the evolution of the amplitude in a point in the middle of the waveguide (one where the wavelength is a multiple of the waveguide's depth, and another where it isn't), and two with the respective electric fields.
↧
↧

about heater simulation in the room

April 11, 2013, 6:31 am
$
0
0
Hello!

I am use comsol 4.3a , and use the non-isothermal flow physics to simulate the heater in the room.
And the room has a door , and I want to set the door is open, and it is connect to the enviroment. So the heat and air can go out through the opened door , also the air from enviroment can supply the room by the opened door.

So I select the open boundary to the door (I don not know it is correct or not correct),
1.if the heater is remove, the simulation can sucessful, but when I see the velocity field , the result the opened door have a very high velocity inlet to the room, and it is not resonable.


2.When the room has heater,and the simulation will fail , and the result is not convergence.

So anyone know how to set the open boundary for open door?



↧

Thin liquid film with thermal gradient in contact with air (tpf, ht, ALE, AMR)

April 11, 2013, 8:27 am
$
0
0
Dear all,

I am modelling a thin film liquid in a rectangular channel (2mm), as well as the air on top of it. At the initial condition everything is within room temprature (Tc) then one side of the heater reaches a higher temprature (Th).
Therefore I am using the laminar two phase flow integrated with heat transfer in fluids for this problem.
Now I want to model the thermocapillary effect at liquid interface with air. I expect to see the deflection and change in height of the liquid and finally attain the liquid surface velocity.

I have attempted to formulate this problem in 2D by using the incompressible NS and adaptive multigrid meshing since the zone of interest is mainly the interface. Here are the setps:

1) Thermal insulation at the bottom wall of the thin film liquid.
3) IC: all sides are at Tc
2) BC: right side is at Tc while left side is at Th with deltaT higher ( 25°C).
4) Outflow (the air also consider as another 2mm layer) so the upper boundry is considered as outflow
----------------------
5) Wall1: is the lef and right sides of the channel which are considered as wet walls with pi/2 contact angle.
6) Initial values for both fluids U=V=0, p=0.
7) Initial interface: a line for the surface in between air and thin film liquid.
8) Wall2: top and bottom wall no-slip condition.
9) Pressure point constraints two top corners
10) Volume force to consider the effect of gravitional force. (rho.g.dt.alpha)

Here is the problem, by solving this I get the velocities but not the one which I want at the interface, it takes a very long time to solve (converge) even with a extremly coarse mesh and medium level of tolerance, therefore I thought of ALE application mode which is in progress. But using initial interface is the best way to consider the surface between liquid and air? And how to insert the change of height with respect to temprature in the problem? (dh/dy ~ dT/dx).
Do I need to consider another body force in air since the marangoni effect happens in the air too? (conductivity of liquid is very small)

Thank you in advance for sharing your opinions with me.
↧

Postprocessing-visualization of symmetric profiles

April 11, 2013, 8:43 am
$
0
0
I solved the thermal stresses in a plate using two faces of symmetry, saving computational cost. I wonder if COMSOL allows me to mirror the two symmetric faces and visualize the deformation in the full plate rather than in the quarter that I simulated.


Thanks
↧
↧

Varying the order of diffraction in Plasmonic Wire Grating

April 11, 2013, 10:56 am
$
0
0
Hi,
I am simulating in Comsol 4.2.I started simulating the PLASMONIC WIRE GRATING given in the model library.the simulation there is done by varying the angle of incidence.(in the parametric sweep angle of incidence is varied).

I want to modify this a little bit.So instead of angle of incidence i want to vary m(the diffraction order)

i have made the necessary changes .

In global variables i have taken the transmission angle beta_m=asin((m*lam0)/(d*nb)) which comes from m*lam0=d(nb*sin(beta_m)-na*sin(alpha)) here alpha=0 for normal incidence
na=R.I of air(upper domain of simulation)
nb=R.I of glass(lower domain of simulation)

The propagation vectors which i have defined are :

ka=emw.k0
kb=nb*emw.k0
kbmx=kb*sin(beta_m)
kbmy=kb*cos(beta_m)

The geometries and material properties have been defined as usual

In port1 i have taken the upper boundary ,wave excitation on and Eo=exp(-i*ka*x) and beta=abs(ka)

In port2 i have taken the lower boundary ,the Eo=-cos(beta_m)*exp(-i*kbmx*x)
and Eo= sin(beta_m)*exp(-i*kbmx*x) and beta=abs(kbmy)

Then i have added the periodic boundary condition as Floquet condition

In the parametric sweep i have varied m(-2,0.1,2)

But this gives me erroneous results!!
probably i am unable to vary /parametric sweep m.

Kindly help I want to find the transmiision for various m
↧

Evaluate parameter value in MATLAB

November 11, 2010, 3:15 am
$
0
0
Hi,

If I have a parameter, e.g.:

>> model.param.set('outer_z', '((volume*formfactor^2)/(2*pi))^(1/3)');

How can I in the easiest way evaluate the value of this parameter (like the 'Value'-column in the GUI)?

I get an error message when trying to use:
>> mphglobal(model,'outer_z')
"com.comsol.util.exceptions.FlException: Unknown dataset feature"

Regards,
Johan
↧

Fluid-Porous Slip boundary conditon

August 19, 2012, 11:29 pm
$
0
0
Hello,

I focus now on the flow and solute transportation simulation in the free flow and porous media. There are traditionally one singe formulation, typically Brinkman and two-domain coupling with Navier-Stokes/Stokes and Darcy's or Brinkman's. For the former unified one I guess as exerted in fp (free flow and porous media), there is no need to specify the boundary conditon along the interface between the two medias, which provides continuous velocity profile across the transition zone. However, I would also like to try the coupling with specification of a slip jump velocity across the interface. There is some literature in comsol conference proceedings. Could anyone explain to me in detail how to impose this jump boundary condition? What is the difference between the choices of 'slip conditon' and ''slip velocity'', could ''slip velocity'' be prescribed as experssion like the Beaver&Joseph boundary dU/dy=α/√k(Uf-Ud), or necessarily assign a value? What is the difference with the maxwell boundary conditon? Could anyone explain to me more in detail or find me some reference?

Best

Yuexia
↧
Search

Unable to create mesh : Error Free Tetrahedral 1

February 25, 2013, 12:31 pm
$
0
0
Hi everybody,

I am having this error when I try to mesh a 3D structure :

The following features have problems :
Features : Free Tetrahedral 1


I don't find a way to have more information about this error.
My geometry is really complex, with many points/edges. I cannot make it simpler.
I still have the same error when I change the element size to be 'Extremely coarse'.

Does someone have an idea to fix this error ?

Thanks.
↧
↧

Effect of prestress on eigenfrequency

March 20, 2013, 1:50 am
$
0
0
Hello,

I am trying to model the effect of the prestress on eigenfrequency of a vibrating string (clamped at both end with a prestress)
The problem is that I am using 3.4 here. Would you please let me know how to load a static analysis result into a eigenfrequency analysis?
What should I do to introduce stress to the material?
Should I use matlab to load the prestress?
Is there a simple way to do that?
Is there simply a way to do it?

I am not able to find good exemples corresponding to that problem with my COMSOL version (3.4).

What should I do to introduce stress into the material?

Thanks.

↧

Evaluation of Volumteric displacement

April 8, 2013, 1:06 am
$
0
0
Hi all,

I am modelling a diaphragm deformed under pressure loading in an axisymetric model.

I was wondeing if there is a way to evaluate volumetric displacemnt of the diaphrgam due to displacement.

Currently I have considered taking the avegage vertical diaplacemnt of the diaphragm across its length and then multiplying by area to give an effective displacement.

Is there a direct way of evaluating this in COMSOL?

Possibly by evaluating the intergral of area beneath the diaphrgam and an arbitary line?

Thanks,
Matt
↧

Linking a Porous Media Flow to a Transport of Concentrated Species

April 8, 2013, 6:29 am
$
0
0
Hello,

I have created a chamber that has sugar water running through a compartment and carbon cloth modeled by a free and porous media flow physics. I also want to create a transport of a concentrated species to model where the sugar is in the solution as it flows through this chamber. My questions are as follows:

Is it possible to link a transport of concentrated species physics to a porous media flow physics while using incomplete geometries? I don't care where the sugar end up in the carbon cloth, so I have deselected it for the geometries, and have been receiving negative feedback from the software.

How do you handle a physics when certain parts of the geometry have different characteristics? For example, the carbon cloth would have a different diffusion coefficient in comparison to the purely fluid filled chamber, however I cannot find where I can mathematically represent this.

Thanks in advance for any help.

-Payton G
↧

Coupling B/W two solid stress-strain modules

April 8, 2013, 8:00 pm
$
0
0
Hi all,

I am trying to model a 3D structural mechanics problem with two separate solid stress-strain modules coupled together at the interfacial boundary through equal displacement boundary condition (i.e. Rx, Ry and Rz for one module is equal to u, v and w of other module) where Ri represents applied displacements in x, y and z directions while u, v and w represents the displacements (the solid stress-strain module always solves for u,v and w in each domain along with the stresses and strains). I think the constraint on the interfacing boundary in this way means that at the interface the displacements must be same. However I cannot acheive convergence at all and the simulations always diverges. I am using default solver settings and unstructured tetrahedral elements mesh. Can some one guide me what should I do.

Regards
↧
↧

General Projection Bulk Temperature Example & Question

April 11, 2013, 4:30 pm
$
0
0
Dear all,

The following model may prove useful to somebody (other than myself) who wishes to calculate the average of some scalar value as a function of axial distance. In this case it's the bulk fluid temperature.

The model was made on COMSOL 4.2.0.228 with basic 'classkit' licence.

It does this:

- Calculates the surface integral of temperature in the xy plane.

- Displays a line graph of the average xy temperature variations in the axial z direction using a general projection.

Also, it has fully developed inlet flow specified with boundary PDE solution.

The general projection method used relies on the symmetry of the geometry, it takes a slice right down the middle of the pipe to calculate the average temperature at different axial positions.


THE PROBLEM:

If I cut the pipe in half the technique used to calculate the bulk average temperature would no longer be valid.

Does anyone have any advice for how average bulk temperature could be calculated for a semi-circular geometry?


Thanks,

Adam Parrott
↧

Typical fluid-structure interaction model

April 11, 2013, 7:51 pm
$
0
0
Hi

I am new and struggling to start a simple model.

I have been playing with blood vessel model from the model library.

I would like to look at the vascular behavior when blood vessel is impacted by a piston.

It is going to be dynamic analysis and I want control the velocity of piston and depth of impact.

My question is that how I can make piston move with certain velocity and control the displacement of piston.

And also how can I solve the problem with dynamic analysis?

Any help will be appreciated.

Thanks
↧
Search

How to apply harmonic and random vibration's load?

April 11, 2013, 8:45 pm
$
0
0
Hello, I want to know how to apply harmonic and random vibration 's load in comsol v4.2,
my previous operation is that first, apply a boundary load, then give a range of frequency, am I right?
or please give me a detailed guide, many thanks.
↧

Induced charge on a surface by a charged object at different relative positions

April 12, 2013, 12:02 am
$
0
0
Hello,

I am new in COMSOL and I was wondering if anyone could five me any idea. I want to simulate the evolution of the induced charge on a surface as the relative position to a charged object varies, this variation must be parametrized by the user and not due to the electrostatic effects. My first idea was to parametrize the position of one object as a function of time but i couldn't. Is possible to do it in any other way? Thank you very much in advance!

Gerard
↧
↧

Electric potential boundary on thin structure

April 12, 2013, 12:38 am
$
0
0
Hi!

My model contains a thin film with a thickness of 30 nm. I have applied different potentials and grounds on the thin boundaries of this film without any problems, but now I'm trying to apply these two boundary conditions on one of the flat sides of the film and the resulting potential doesn't turn out the way it should.

Is there a way to get a proper result by using these bc's on the flat side or do I have to stick with my old solution (i.e. potentials on the thin boundaries)? I made a quick model to illustrate my problem. thick.png shows what I would like to get (here I've increased the thickness to 30 µm) and thin.png shows the unwanted result due to the thinness of the geometry.

Thanks.
↧

Transport of concentrated species, Membrane gas separation

April 12, 2013, 3:07 am
$
0
0
I've attached simplified model describing transport of the one component through membrane. I've made it simple to make you catch a problem easily. In my advanced model I'm using Laminar Flow physics and 2D axissymmetry as well.
The problem is with mass balance in this scheme - see Line Graphs. As a result it gives mass/mole fractions higher than 1 or lower than 0. Likely I don't get some sense of using Transport of concentrated species interface.
How can I make all of the components passing the membrane? As I'm choosing 3 components, I can define flux only for 2.
What happens on the feed side when part of the component is leaving by the wall? I think the pressure should decrease along the wall (according to decreasing of component partial pressure and friction influence on fluid flow), also mass flux decreases. Does Comsol "knows" that?
Are the Laminar flow and Transport of concentrated species interfaces sufficient to create property model in this case?

I wish you know the answers ;-)
I'm beginner - I'm sure you've guessed.

Anna
↧

Circumferential direction

April 12, 2013, 4:58 am
$
0
0
Hi

My equations are in default cartesian(x,y,z) coordinates. But my geometry is circular and I want to see the temperature distribution only in circumferential direction. It would have been better if I could only consider heat transfer along the circumference. My heat source term is expressed in x,y,z coordinates. Any help would be appreciated in this regard.

Thanks.
↧
Viewing all 26527 articles
Browse latest View live
Search