Quantcast
Channel: Latest Discussions - COMSOL Forums
Viewing all 26447 articles
Browse latest View live
↧

Interpolation of geometry according to chosen element type

September 11, 2017, 6:03 am
$
0
0
Hi,

I am new to Comsol and have a basic question.

When modeling a curved geometry (solid mechanics) meshed with a coarse mesh with linear interpolation, the curved parts gets edgy due to the limited number of elements. To fix this, I would like to choose a higher order of interpolation, i.e. cubic serendipity or lagrange elements. I do this under 'Discretization' under the Solid Mechanics node.
However, it seems that I can only achieve this for the displacement field alone, which leaves me with an edgy geometry, with only the displacements interpolated. I would like the intermediate nodes of the elements to be fitted to the geometry according to the shape functions defined by the element. How do I achieve this?


David
↧
Search

Having problem with the heat transfer part of my model

September 11, 2017, 6:25 am
$
0
0
Hi all,

I have been dealing with a problem in model attached. Basically, this model is to simulate water evaporation, located in a container, across a porous media.
The problem is with the heat transfer part of the model. In my experiments with the setup which I am modeling, I observed constant temperature of water during the drying process.
According to the experiments, water temperature decreases and reaches thermal equilibrium with surrounding air which results in constant temperature of water. While, in the model I have not been able to simulate this observation for the water inside the container.

The model is somehow large. I could not upload it to this post. But, if you are interested in looking at the model, I can send it to you via email which enables me to share a big file.

How can I solve this issue?

Babak
↧
↧

Potential errors with Surface Maximum node

September 11, 2017, 2:50 pm
$
0
0
I'm curious about how maximum value measurements are found in FEM, in particular: results using the Surface Maximum node. I know that issues of infinite stress can arise at corners in structural modeling and I wonder if I am experiencing something similar in a heat transfer model.

I am modeling a half channel of a heat exchanger. I have noticed something odd while performing a mesh sensitivity analysis. Sparing you all of the details, one of the measurements I am interested in is the maximum temperature on the channel wall. To find this I am using a Surface Maximum node. As I refine the mesh initially (from 10,000 elements to 20,000, to 30,000...) the max values and my other measurements do not change (<0.5% difference). If I continue to refine the mesh (60,000+ elements) the max temperature begins to rise and shows no signs of leveling off. All other measurements I'm taking remain the same as they did with a coarser mesh. I have found the same issue with maximum heat flux. Is this some type of numerical error, or artifact of FEM? Or is it actually physically meaningful?

Thank you for the help,
Bryan
↧

how to simulate shear wave propagation in a transverse isotropic viscoelastic medium?

September 11, 2017, 11:53 pm
$
0
0
Hello,
I simulated according to the tutorial "Focused Ultrasound Induced Heating in Tissue
Phantom" and analyzed the acoustic pressure and sound pressure level. but now, I want to involve the assumption of the acoustic wave propagation( linear) and
also the amplitude of shear waves in the tissue domain in my previous simulation.
please suggest me.

Thanks
Najme
↧

Initial values for Added Geometry in Time Dependent Loop Simulation

September 12, 2017, 12:23 am
$
0
0
Hi,

In my time dependent problem my geometry is devided into two parts A1 and A2. The transient solution in carried out in two phase.

Phase 1: Domain A1 is considered and A2 is ignored and results are calculated for time "0" to "t1".
Phase 2: Domain A1 and A2 is considered and Initial values for domain A1 is used from Phase-1 solution.


My query is how can I initialize the initial values for domain A2. By default it is getting "0".
Thanks

Mridul
↧
↧

test

September 12, 2017, 6:26 am
↧

Modification of variables in parametric sweep

September 6, 2017, 7:35 pm
$
0
0
Hi community, I'm trying to get solved my model with LiveLink.
Is it possible when study runs with parametric sweep and
in each parameter(voltage in my model), change the variable?

I've done this before.

Before runs my model I already set up variable(x1 as "a")

and After runs my model.sol('sol1')
I'd reset up my variable x1 with other value(b) at certain parameter(voltage)

and runs it again.
But when I looked it up in comsol, x1 has only "b" in every parameters.
I meant it has to be different value in different parameter.

simplifying it.
V=0~1V , x1=a;
V=1.1~2, x1=b;
like this.

But What I got is
V=0~1v, x1=b;
V=1.1~2, x1=b;

Any suggestions would you give?

Best regards,
Min
↧

test

September 12, 2017, 7:20 am
↧
Search

It’s here! Welcome to the new user forum.

September 12, 2017, 9:53 am
$
0
0
The COMSOL discussion forum has a new design and we hope that you find it intuitive and easy to use. Along with the new look, we have focused on improved usability, search tools and tagging, as well as a mobile-friendly interface.

While we have tested the system extensively, if you find any problems or bugs, please e-mail us at <web@comsol.com> so we can address the issue promptly. In addition to any fixes, we hope to continue to roll out improvements and new features over the coming months.

Enjoy the new forum!

↧
↧

parameter estimation with surface reaction module

September 12, 2017, 10:05 am
$
0
0
I am trying to estimate some parameters (Kon, Koff) for a surface reaction in a microchannel. The surface reaction is happening along a small part of the boundary in a 2D space. i have the surface reaction module set up to monitor the surface concentration of the formed species (cs_AB).

The simulation works when tested with some arbitrary values. Now I want to estimate the real parameters from my experimental data. So i set up my reference data which contains the concentration along the reaction surface, and the time.

i have 2 questions:

1. the surface concentration is stored in the variable cs_AB , however when I enter this variable in the parameter estimation module (in the expression field for value), it gives the error that the expression 'cs_AB' is not known while solving.
also entering 'comp1.cs_AB' gives this error.although in the parameter estimation equations view, it recognizes the expression and automatically labels it as a 'surface concentration'. is the surface reaction module even compatible with the parameter estimation?

2. my reference data actually contains the integral of the surface concentration along the boundary. is there a way to integrate the found surface concentrate cs_AB along my boundary, and then use this value to fit my parameters to the reference data?

↧

[Help] Comsol takes too much temporary space while computing

September 12, 2017, 10:40 am
$
0
0
So I was just running a simple microfluidic tube on Comsol 5.1. Comsol was able to handle it with no problem for the first couples run (giving 4-6 hours) per run (case studies). However, what I observed a couple days ago was that COMSOL required or took up to some 600 GB to compute and run into the "out of internal memory" problem. My the CPU I'm using comes with 500 GB, so I think it would be impossible to have the files that large. And that 600GB is just temporary and disappear if I close or stop the program. Has anybody bumped into this problem please let me know. Would extra internal hard drive space help me in this case ? Thank you !
↧

Challenge in meshing model

September 12, 2017, 5:45 pm
$
0
0
I am having a challenge in meshing my model. Please, find the attached model and help me. Thank you.
↧

compound parabolic concentrator

September 12, 2017, 11:02 pm
$
0
0

I'm working on a project, which i am making a study about the compound parabolic concentrator modelling.i have try to do it using comsol rayoptics module but i cant create the specified curvature regarding to it. I don't know how to model the cpc in ray optics.if there is any tutorial or suported materials please reply.
Pasan
↧
↧

problems on modeling magnetoelectric material

June 18, 2014, 5:10 am
$
0
0
Hi there ,I am recently modeling magnetoelectric material, but I have some problem with it ,which is the stress was not right coupled between magnetostrictive material and piezoelectric one. there do is stress ,but no diplacement . I referred to the example of nonlinear_magnetostriction_actuator_sensor. while the difference is that the magnetic field in my model is provided with a permanent magnet. so my question is what's the different between MF(no current) and MF on my situation. moreover, how to define the boundary conditions if I want it to transfer stress, or strain? thank you very much !
↧

velocity profile

September 9, 2017, 10:22 am
$
0
0
greeting
I'm using the slip velocity for the wall condition and I got the following velocity profile
can any one explain why this decrease in velocity near the wall, please.
↧
Search

Plastic dissipation calculation: solid.Wp

September 13, 2017, 12:37 am
$
0
0
Hello everyone,

I am running a time dependent model of a crack growing and propagating in an elasto-plastic medium (solid mechanic module) and i am analyzing its energy balance evolution in function of time.

I was curious to know how the plastic dissipation energy density (solid.Wp) is calculated by Comsol. I looked into the equation thanks to the option "equation view" but i didn´t find anything. Does anyone know how this variable is calculated?

Thank you very much,

Rémi
↧

Velocity of deformable droplet

September 13, 2017, 2:05 am
$
0
0
Hello,
I'm building a unsteady computational model of 'Motion and deformation of droplet through minichannel using Level set method'. I ant to plot a graph between droplet velocity and the time scale. How can I do it when droplet is itself deforming and interface is changing with time.?
Kindly help me.

Regards,
Akshay Tiwari
↧
↧

Import complex geological geometry

September 13, 2017, 3:45 am
$
0
0
Hi all,
I am trying to import geological layers from Petrel software to COMSOL. Importing the grid directly is not possible as the formats of Petrel (*.egrid, *fgrid, *gsg, ASCII) are not recognized from COMSOL and vice-versa.
So my approach is to do this only with the defining surfaces and the mesh the whole domain again, this time in COMSOL. I am able to export contact surfaces in the form of points (X,Y,Z txt file) and import this into COMSOL. These are large areas of about 13km x 5km and the original mesh is 50m x 50m.
However, I am not sure how to proceed. How can I convert these imported points to a "solid" surface and assign the domain between them as geological space to be populated with properties such as porosity, permeability etc?
Any tips or guidance would be greatly appreciated
↧

Simultaneous Boundary Conditions

September 13, 2017, 4:04 am
$
0
0
Hello,
I am making an exercise to reproduce a model with a set of PDE using math model in COMSOL, the model include a set of boundary conditions like this:
y=0 ---> Ψ = 0
and
y=0 ---> ∂Ψ/∂y = - S ∂θ/∂x
When I introduce the second one overwrite the first one,
How can I introduce this simultaneous conditions?,
or simply COMSOL can not manage this?



↧

Convergence issue

September 13, 2017, 5:02 am
$
0
0
Hello,

I am working on a 3D simulation for an h shaped fluidic channel. There are two different concentration of solution flowing through the two channels at different flowrates. I would like to look at the ocncnetration profile at the interconnect between the two channels. This works well at 2D, however, on 3D, it gives me convergence error. I have checked for various values of torences, but I cannot seem to solve the proble. Please let me know if you have an explnation or solution to this problem.

Thanks in Advance.

Sahana
↧
Viewing all 26447 articles
Browse latest View live
Search