Quantcast
Channel: Solving, solving\ - Latest Discussions - COMSOL Forums
Viewing all 3988 articles
Browse latest View live
↧

Coupling magnetic field (no current) with solid state BH curve

July 4, 2012, 6:24 am
$
0
0
Hi All,

I want to model magnetostriction phenomena. I found an example (nonlinear_magnetostriction.mph) where the magnetic field is created by a coil (current density J0) and which uses Magnetic Field (mf) physics equations. In my study I use a permanent magnet, and so I have to use Magnetic Field No Current (mfnc) equations to solve my problem.

Solving the magnetic part alone is not a problem, when I use a BH curve of my magnetostrictive material or when I use only a constant permeability value µr.

I add the solid mechanic physics equation (linked to the magnetic field), only for my magnetostrictive material.
I am perfectly able to find a solution when I use a constant µr value to define “magnetically” my magnetostrictive materials.

But when I change the magnetic definition of my material to a BH curve I have the following error:

Failed to evaluate variable.
- Variable: mod1.mfnc.Hz
- Geometry: 1
- Domain: 1
Failed to evaluate variable.
- Variable: mod1.mfnc.mfc3.minput_magneticfieldz
- Defined as: mod1.mfnc.Hz
Failed to evaluate variable.
- Variable: mod1.mfnc.normB
- Defined as: (mod1.mat4.BHCurve.BH(((unit_m_cf*sqrt(((mod1.mfnc.mfc3.minput_magneticfieldz^2)+((mod1.mfnc.mfc3.minput_magneticfieldy^2)+(mod1.mfnc.mfc3.minput_magneticfieldx^2)))))/unit_A_cf))*unit_T_cf)
Failed to evaluate expression.
- Expression: subst(mod1.mfnc.normB,mod1.mfnc.HX,(unit_A_cf/unit_m_cf))
Failed to evaluate expression.
- Expression: nojac(((unit_m_cf*(subst(mod1.mfnc.normB,mod1.mfnc.HX,(unit_A_cf/unit_m_cf))-mod1.mfnc.normB))/unit_A_cf))
Failed to evaluate variable.
- Variable: mod1.mfnc.normDiff
- Defined as: nojac(((unit_m_cf*(subst(mod1.mfnc.normB,mod1.mfnc.HX,(unit_A_cf/unit_m_cf))-mod1.mfnc.normB))/unit_A_cf))
Failed to evaluate expression.
- Expression: (mod1.mfnc.normDiff*mod1.mfnc.HZ)

The error is concerning the variables: mod1.mfnc.Hx, mod1.mfnc.Hy or mod1.mfnc.Hz with no rules.

I have already done some tests:
- Even if the solid mechanics has no linked to the previous magnetic calculation, I have the error.
- If I add a heat transfer instead of the solid mechanics, I am able to find the magnetic solution.

I am using Comsol 4.2a.
The BH curve that I use is the inverse of the HB curve from the magnetostrictive material used in the nonlinear magnetostrictive example.

What could I change to find a solution with a BH curve?
↧
Search

Failed to find a solution. The relative error is greater than the relative tolerance.

August 22, 2014, 10:48 am
$
0
0
Hi all,
I'm trying to simulate a Piezoelectric transformer using the piezoelectric devices physics.
My structure is quite simple:
it's a thin circular membrane of piezoelectric material thick 1um (AlN), with upper concentric electrodes for primary and secondary (inner circle is the primary while the secondary electrode sorrounds the primary electrode).
Under the AlN membrane there's silicon thick 4um
I run the eingenfrequency study step in order to find resonant frequencies, but when I run frequency domain study step, in a small range of the chosen resonance, I get the error: "Failed to find a solution. The relative error is greater than the relative tolerance".
As the problem has a cylindrical simmetry, this model is set in 2D-Axysimmetric

- I have to say that I get this error when the frequency is extremely close to the resonant frequency.
- I have inserted damping and losses for all materials, thus getting a finite quality factor (so the problem is not related to undamped response)
- I ve tried an extremely fine mesh
- I ve tried to change solver parameters (although i m not an expert in solvers) and this seems to change a little the frequency where I get the error (since the Q is around 4000, I've set a 100Hz frequency step), but at the end I always get the error.
- I ve read all forum discussions about this error, but nothing seems to help me.
- If I change the radius of the structure from about 500um to 50um, I do not get the error, furthermore the simulation is faster!

How can I resolve this error?

Thanks in advance!

Antonio
↧
↧

Use constraints with multiple modules

August 7, 2015, 5:41 pm
$
0
0
Hi all,

I am working on a problem where I use several different modules, and I want to add another constraints to the model. Say in my model component 1 I want to solve for u(t), d(t) and phi(t), each with one module. Now I want to use one equation (ux*phix+uy*phiy)*step1(d-dm)=0 as a constraint that is to be satisfied globally, how should I implement it? I tried adding "Global constraint" but it seems not working. Any help is appreciated.

Jiangzhi
↧

Convergence problem in time dependent solver

September 26, 2011, 12:58 pm
$
0
0
Hi,

I'm getting this error when running a time dependent simulation of laminar flow in a micro channel. What I want to do is change the input flow on one arm of a y-channel and follow the response in time. Anyway it's this error:

Repeated error test failures. May have reached a singularity.
Time : 0
Last time step is not converged.

Now I've seen people with similar problems where it is due to too large a change in the time dependent parameter but I don't see how this is the case here. I am simulating for t=0 with the stationary solver first which is giving the correct result and at t=0 for the time dependent solver the boundary conditions should be exactly the same. I've tried several different time steps and different arrangements for the step function I'm using to change the flow. If it failed as the flow changed I'd think there was a point to the error but this doesn't make sense to me. Anyone have any ideas?
↧

Failed to find consistent initial values, Dense suspension simulation

August 10, 2015, 3:08 am
$
0
0
Hi,

Im new to comsol and i want to simulate shear induced migration of particles in a suspension at low reynolds number. I tried to implement it in the mixture model laminar flow but the model had some extra terms which I cant seem to get out of the model equations.

In 2d, I have two concentric circles, in which the inner one moves, I have pics of my model attatched

Using creeping flow model, (spf)
rho*du/dt= del.[-pI+mu( del u + (del u)^T)] +F

Here i express the viscosity mu equal to a variable I defined as
eta_f=((1-(phi/phid_max))^(-1.82))*eta_f0

Where phid_max and eta_f0 are parameters = 0.68 and 4.95.

phi I define via a coefficient for pde:
d phi/dt+ del.(-(D1*phi*spf.sr+D2*spf.sr*phi*phi*d(eta_f,phi)/eta_f)- alpha*phi) +u. del phi=0

alpha= D1*phi*phi*d(spf.sr,x) .........in x
D1*phi*phi*d(spf.sr,y).....in y

For the creeping flow I have initial values u,v =0 ,p=1, outer circle as wall, inner circle as inlet,
u0= ay in x
-ax in y

for the pde I have no flux on all walls and initial value 0.5.

when i run it, it always says last time step not converged. Inconsistent initial values. should I make any changes in the solver to make it converge. I also tried to remove viscosity from the variables and tried to implement it as a pde with only f as the function and all others zero (where it did run once but Couldnt run it again after tweaking it a bit, cant seem too get back the earlier run)

Do i have to add any weak constraint or change any discretization or anything?? I am using the fully coupled solver, ill also try the segregated solver now.

I would be very grateful on any help on this to converge. I have pics attatched of my model.

Thank you
Nikhil



↧
↧

Results of fixed Boundaries

August 11, 2015, 10:48 am
$
0
0
Hi all,

I've got a Problem. I try to create a phononic crystal lens using pressure acoustics (frequency domain). Even at clamped edeges I recieve acceleration values unequal to zero.

How can I explain that?

↧

1D axisymmetric equation based modelling basics

August 11, 2015, 10:58 am
$
0
0
I am pretty new to comsol. I need to model a 1d axisymmetric problem of fluid flow at the moment and develop the model further. I cant find the fluid flow physics for 1D. so i decided to try equation based modelling.

Here are the eqautions i modelled using a pde

It is a simple couette flow model between to concentric cylindersof radii 0.0238 and 0.0064m




I am always getting last time step not converged.
Do you know what boundary conditions i should use for pressure.

Thank you
↧

Zero flux boundary conditions issue with Poisson equation

August 11, 2015, 12:32 pm
$
0
0
Good morning,

I am using version 4.3b. One of the steps in my project requires me to solve a Poisson equation with zero flux conditions at the boundaries. My geometry is a simple 2-by-2 square (domain: [-1,1]x[-1,1]). I have 2 questions regarding a tolerance error that keeps popping up.

1). When I have a source term as simple as: f=cos(2*pi*x), it says:

Failed to find a solution.
The relative error (25) is greater than the relative tolerance.
Returned solution is not converged.
- Feature: Stationary Solver 1(sol1/s1)
- Error: Failed to find a solution.

But if I replace just 1 of the side walls with a Dirichlet condition, I get the analytic solution.

I have verified the consistency condition for Poisson's equation with Neumann conditions (integral of f over the region is equal to zero, which is the same as the integral of zero (flux) over the boundary). I don't understand why COMSOL can't handle the zero flux condition.

2). When I try something even simpler with source term: f=0, I get the same error if my initial guess is u=2, but I get a valid solution when my initial guess is u=0. With zero flux on the walls, both u=2 and u=0 everywhere should be valid solutions. What is the reasoning for this error?

Thanks in advance,
-Peter
↧
Search

evaluating of percentiles

August 12, 2015, 5:35 am
$
0
0
Hi,

I'm looking for a java code (not matlab) or a possibility to evaluate percentiles of a 1D, 2D or 3D distribution.

I calculate the time step of each element using the CFL number during simulation. For a max. possible timestep during simulation (gen. alpha, time step manual) I want to use a 10% quantile.

Is there any good idea?

Best regards
Michael
↧
↧

Singuarity error and mesh size

August 13, 2015, 1:50 am
$
0
0
Hi
I have a pde based model where


It is a 1d problem which simplifies it to an extent. when I try to run it at coarser mesh sizes I get an error saying may have reached a singularity. When the mesh I use is made finer the singularity error occurs at a lower value of time.

I have been reading on trying to reduce the singularity error and I think that it might be because of gamma which is a derivative of u (du/dr). There are sharp point in the u vs r graph, so I think this might be the problem.

Can anybody help me in solving this problem. Is it because of the meshing. So how do I get rid of the singularity error by proper meshing of the 1d problem. I know that the singularity always occurs at r-around 0.5. How can I increase the meshing there.

Thank you for taking the time to help me out.
↧

Convergence problems for magnetic circuit

August 13, 2015, 11:14 am
$
0
0
I am quite new in using Comsol and currently facing convergence problems in the following situation:

I want to solve the magnetic field in a magnetic circuit where I put a coil of a rectangular cross section modelled by addition of an External Current Density node. Properties of magnetic material are given by HB curve. I think everything is properly defined because for common magnetic materials the solution is converged without any significant problems and looks as expected. However, the convergence problems arise when the HB curve is quite steep for higher B. The solution hardly converges for quite rough mesh and for denser mesh which I need to use for further calculations it does not converge at all.

I was trying to solve the problem for lower excitation current density, where the solution converges, and use the converged solution as initial values in following study for higher excitation. However, it seems that such an approach does not help and I am looking for another hints how to overcome such problems. Has anybody faced anything like this?
↧

How to plot quantum efficiency

August 14, 2015, 5:14 pm
$
0
0
Hi all,
I'm a new COMSOL 4.4 user. For my master study I'm trying to learn RF module applications in COMSOL. My topic is quantum well infrared detectors with prism and pyramid geometry. I have completed the draw the geometry of the prism but I couldn't understand the definition of quantum efficiency graph.
The paper that I have studied is:
www.researchgate.net/publicati...m_Well_Infrared_Photodetectors
I'm trying to plot the Fig 2 which is wavelength versus quantum efficiency
I couldn't understand that is there any prepared expression for quantum effiency or should I write a formula to plot the graph?
↧

Two phase darcy law

August 18, 2015, 5:45 am
$
0
0
Hi
I am using transient two phase darcy law for resin impregnation through a porous material. The porous medium (rectangular block) of certain porosity and permeability contains cylindrical porous tubes of a different porosity and permeability. The purpose is to study flow front variation in the inter-cylinder and within the cylindrical bodies simultaneously.
The entry of resin occurs at prescribed pressure and saturation while at the outlet vacuum is applied. All other surfaces of the rectangular block have no flux condition. The surfaces of the cylindrical tubes have NO FLUX condition as NOT APPLICABLE. No other boundary condition is specified for these tube surfaces.
The simulation runs into singularity error only after 0.06 seconds. I have attached the simulation file here for further review. Please revert with any possible solutions to this issue.
↧
↧

Capillary diffusion and relative permeability

August 18, 2015, 5:53 am
$
0
0
Hi
I am using transient two phase darcy law for resin impregnation through a porous material. The porous medium of certain porosity and permeability is being impregnated by resin as it displaces air.
The entry of resin occurs at prescribed pressure and saturation while at the outlet vacuum is applied. All other surfaces have no flux condition. The simulation does not converge and runs into singularity errors.
Different results are obtained when fluid 1 is resin and fluid 2 is air compared to fluid 1 being air and fluid 2 being resin. The second case works relatively better.
1) What is the role of capillary diffusion term and relative permeability ? In our case these values are default.
2) How do you select values of capillary diffusion and relative permeability ?
3) Is there any mechanism of selecting which material (resin or air) will be fluid 1 ?

Please revert with any possible solutions to this issue.
↧

Running big simulation- time step

August 18, 2015, 6:06 am
$
0
0
Hi,

I have long data (hourly time step for three years). If I want to run this amount of data,then it takes long time to solve the problem. Is there any way to reduce time calculation in COMSOL?

Is it possible to divide the simulation to sub-simulation and then run together?

Thanks in advance for your answer.
Sepideh
↧
Search

Detail: Undefined value found in the equation residual vector

August 18, 2015, 6:29 am
$
0
0
Hello Experts,

Am trying to solve a packed bed reactor problem am facing this error. Could anyone help me out please?

- Detail: Undefined value found in the equation residual vector.
There are 3 degrees of freedom giving NaN/Inf in the vector for the variable mod1.c_T2.
at coordinates: (0.163432,-0.451967), (0.226592,-0.454277), (0.218286,-0.5), ...

I have attached my file below. Thanks a million in advance
↧

Magnetic Field of a Multi Layer solenoid

June 18, 2015, 3:19 pm
$
0
0
Hello Everyone,

I have no experience in using COMSOl.

My professor wants me to run an example project in this software and I have a few specific questions.
The problem statement is:
I have a multi layer coil. Each layer has 10 wounds and there are 5 layers. The coil is wound on a ceramic rod. I need to give AC current as an input and plot the magnetic field distribution around the coil. ( more importantly along the axis of the rod).
How I proceeded
I designed the coil in SOLIDWORKS and imported the file into COMSOL. I designed an cubical test area around the coil (infinite element domain). I specified material for coil as copper and for the test area as air.
What should be the selection in Ampere's law?
I included coil ends and the sidesof test area cube in magnetic insulation.
Then I chose Single turn coil and gave boundary feed (How can I input AC current?) to one end and ground to another.
I chose a fine mesh size and meshed the coil using free tetrahedral. For the test area I chose free triangular.
I input 10 kHz as the frequency.
I don't see any fancy plots coming up. All it shows is 0 and faint green
Can someone tell me where I am going wrong? How do I proceed further?

Thanks in advance,
Rahul Korpu

↧
↧

Cantilever beam fracture

July 31, 2012, 6:38 pm
$
0
0
Hi Everybody

I have a question regarding analysis of deflection of the cantilever beam.
As it is clear, If I apply a small force on tip of the cantilever it will deflect. However in simulation I don't know how I can evaluate if the deflection (or applied force) is small enough to avoid fracture (breaking) of the beam. [beam dimensions are as follow: Length 500 to 100 um , width 30 to 50 um, thickness 0.7 um] and the force is in range of 100 pN
Can anybody give me some idea?


-I went through some mechanical books and they said the maximum stress that a material can tolerate before fracture is equal to yield stress (or fracture strength), but i don't know in my simulation I should compare which parameter with yield stress.
↧

Limiting current from Rotating Disk Electrode - Simulation doesn't match theory

August 19, 2015, 6:37 pm
$
0
0
I am having an problem getting a value for the current at an electrode, that is being rotated in solution, that would be expected from theory.

The Levich equation ( en.wikipedia.org/wiki/Levich_equation ) models the diffusion and solution flow conditions around a Rotating disk electrode (RDE). The Levich equation gives the height of the sigmoidal wave observed in rotating disk voltammetry, or the limiting current. The limiting current can be approached, for example, by increasing the electric potential or decreasing the rate of mass transfer to the electrode. The RDE reduces the rate of mass transfer through induced flux and is equivalent to a silent solution at steady state.

I began with model library. I found the microdisk volatmmetry model. This model shows the limiting current as the applied potential is changed. I added swirl flow and got rid of the infinite boundary domain because the laminar flow physics did not allow me to select it. I am integrating the local current density along the electrode and revolving around the z axis (elan.iloc_er1*r*pi*2). this does not match the expected solution from the Levich equation. Also the current applied will result in the maximum current for this system.

What am I doing wrong? I've been at this for a while and have had no luck at all.
↧

Limiting current from Rotating Disk Electrode - Simulation doesn't match theory

August 19, 2015, 6:37 pm
$
0
0
I am having an problem getting a value for the current at an electrode, that is being rotated in solution, that would be expected from theory.

The Levich equation ( en.wikipedia.org/wiki/Levich_equation ) models the diffusion and solution flow conditions around a Rotating disk electrode (RDE). The Levich equation gives the height of the sigmoidal wave observed in rotating disk voltammetry, or the limiting current. The limiting current can be approached, for example, by increasing the electric potential or decreasing the rate of mass transfer to the electrode. The RDE reduces the rate of mass transfer through induced flux and is equivalent to a silent solution at steady state.

I began with model library. I found the microdisk volatmmetry model. This model shows the limiting current as the applied potential is changed.
↧
Viewing all 3988 articles
Browse latest View live
Search
<script src="https://jsc.adskeeper.com/r/s/rssing.com.1596347.js" async> </script>