I want to make a subject about the hygro-thermo-mechanical-chemical coupling of concrete in early-age (from casted to 28 days). The material is concrete whose porosity, density and saturation are changed. And the hydration reaction of concrete in early-age couples with moisture and temperature. At the same time, external force influences the porosity of concrete. Witch modules should I select? Thank you !
↧
hydro-thermo-mechanical-chemical coupling, which modules should I select?
↧
how to model a moving water droplet on hydrophobic surface in comsol
hello all,
i am new to comsol environment. i am working on a problem to simulate the moving of a water droplet on hydrophobic surface but do not know from where to start. can any one suggest me an effective tutorial and how to proceed.
thanks
i am new to comsol environment. i am working on a problem to simulate the moving of a water droplet on hydrophobic surface but do not know from where to start. can any one suggest me an effective tutorial and how to proceed.
thanks
↧
↧
Calculating cubic roots with conditions
Hello everyone
I am using Comsol 3.5a and I am trying to perform a calculation but unfortunately it seems not to work. I am using Comsol just to solve differential equations and , hence, I am using the mathematical tool that is incorporated. My domain is a square with variables x and z and side equal to 10 dimensionless units. What I want is to calculate the square root of (x^2-y^2) but only for x>y. In all other cases I want the outcome to be 0. I tryied to use absolute value (using the abs(x^2-y^2)) but it seems that in that way I also create solutions for x<y , as well.
Also I am trying to use a condition like: sqrt (x^2-y^2)*(x>y) but also is not working.
Do you have any idea what should I do?
Thank you very much
Yuri
I am using Comsol 3.5a and I am trying to perform a calculation but unfortunately it seems not to work. I am using Comsol just to solve differential equations and , hence, I am using the mathematical tool that is incorporated. My domain is a square with variables x and z and side equal to 10 dimensionless units. What I want is to calculate the square root of (x^2-y^2) but only for x>y. In all other cases I want the outcome to be 0. I tryied to use absolute value (using the abs(x^2-y^2)) but it seems that in that way I also create solutions for x<y , as well.
Also I am trying to use a condition like: sqrt (x^2-y^2)*(x>y) but also is not working.
Do you have any idea what should I do?
Thank you very much
Yuri
↧
optimization
Dear Sir
I bend a cantilever beam using 3 piezoelectric patch in hyperbolic shape. But desired profile is not meet. By changing the voltage of piezoelectric I want to match the curvature of the cantilever beam with desired hyperbolic profile. For this propose we calculate the summation of square of error at every node along length ( x=0 to L ). This is my objective function and I try to minimize the objective function. Error is define as difference of deflection of cantilever beam (z direction displacement) and actual deflection (z_actual = -a + a*((b^2+x^2)/b^2)^0.5 ) at particular location of x (node). My question is :-
1. How to write objective function i.e z direction displacement ( deflection of every node from x=0 to L). How I define variable that represent z direction displacement.
2. How define x in objective function.
Best regards
Sumit
Junior Research Fellow (Research Scholar),
Synchrotrons Utilization Section,
Homi Bhabha National Institute - Raja Ramanna Centre for Advanced Technology, Indore,
Department of Atomic Energy,
Government of India.
Mobile: +91-9170927292
Email: sumit102790@nitp.ac.in
I bend a cantilever beam using 3 piezoelectric patch in hyperbolic shape. But desired profile is not meet. By changing the voltage of piezoelectric I want to match the curvature of the cantilever beam with desired hyperbolic profile. For this propose we calculate the summation of square of error at every node along length ( x=0 to L ). This is my objective function and I try to minimize the objective function. Error is define as difference of deflection of cantilever beam (z direction displacement) and actual deflection (z_actual = -a + a*((b^2+x^2)/b^2)^0.5 ) at particular location of x (node). My question is :-
1. How to write objective function i.e z direction displacement ( deflection of every node from x=0 to L). How I define variable that represent z direction displacement.
2. How define x in objective function.
Best regards
Sumit
Junior Research Fellow (Research Scholar),
Synchrotrons Utilization Section,
Homi Bhabha National Institute - Raja Ramanna Centre for Advanced Technology, Indore,
Department of Atomic Energy,
Government of India.
Mobile: +91-9170927292
Email: sumit102790@nitp.ac.in
↧
Stokes flow (calculation speed)
Hi everyone,
If we try to simulate water flow on a microfluidic channel with the same model (geometry, mesh, physic) and compare only the change from laminar flow to stokes flow (neglect inertial forces) shouldn't the convergence be reached faster in the stokes flow case?
Thank you everyone
If we try to simulate water flow on a microfluidic channel with the same model (geometry, mesh, physic) and compare only the change from laminar flow to stokes flow (neglect inertial forces) shouldn't the convergence be reached faster in the stokes flow case?
Thank you everyone
↧
↧
Added mass force
Hi
Can comsol model added mass force(virtual mass force) in particle tracing module or no?
Thanks alot
Can comsol model added mass force(virtual mass force) in particle tracing module or no?
Thanks alot
↧
Adding geometry between studies
Hi
Is it possible to add geometry between the studies and use the result of a previous study to solve the next one?
I am going to simulate this situation (simple heat transfer in solids 2D):
1. One block of steel cooling from 1000 K by natural convection for 1000 s from the top surface (the other three sides are isolated).
2. After this time, another block of steel is placed on its top, and they cool for another 1000 s together from the top surface of the top block (so the heat flux BC is transferred from the top of the lower block to the top surface of the top block).
I really appreciate if you guys do the simulation and send me the file, or at least guide me here please.
Is it possible to add geometry between the studies and use the result of a previous study to solve the next one?
I am going to simulate this situation (simple heat transfer in solids 2D):
1. One block of steel cooling from 1000 K by natural convection for 1000 s from the top surface (the other three sides are isolated).
2. After this time, another block of steel is placed on its top, and they cool for another 1000 s together from the top surface of the top block (so the heat flux BC is transferred from the top of the lower block to the top surface of the top block).
I really appreciate if you guys do the simulation and send me the file, or at least guide me here please.
↧
Elastic perfectly plastic contact
Hi,
I have two touching geometries and I used a thin elastic layer to model the contact. I determined the spring constant of the thin elastic layer as a function of the spring displacement (uspring) to have a elastic perfectly plastic contact to have slipping displacements. The input values of this function changes along the length of the contact. The stiffness expression is "(abs(solid2.uspring3_tel1)<(1[mm]))*100[MPa]*x". During running my parametric study, at a certain value of the parameter that the spring displacement exceeds the 1[mm] limit and the slipping starts I get error.
I want to know what is the expression for spring constant (like "uspring" for it's displacement) to plot it along the contact length and see what happened. I think maybe after that one point reaches my displacement limit the spring stiffness becomes zero at all of the points of the contact (not only at the "x value" I expected and so my model do not converge. I would be happy to know what kind of expression would be better to model this type of contact to prevent what I think was happening.
Thanks in advance
I have two touching geometries and I used a thin elastic layer to model the contact. I determined the spring constant of the thin elastic layer as a function of the spring displacement (uspring) to have a elastic perfectly plastic contact to have slipping displacements. The input values of this function changes along the length of the contact. The stiffness expression is "(abs(solid2.uspring3_tel1)<(1[mm]))*100[MPa]*x". During running my parametric study, at a certain value of the parameter that the spring displacement exceeds the 1[mm] limit and the slipping starts I get error.
I want to know what is the expression for spring constant (like "uspring" for it's displacement) to plot it along the contact length and see what happened. I think maybe after that one point reaches my displacement limit the spring stiffness becomes zero at all of the points of the contact (not only at the "x value" I expected and so my model do not converge. I would be happy to know what kind of expression would be better to model this type of contact to prevent what I think was happening.
Thanks in advance
↧
The meaning of imaginary part of refractive index in COMSOL
What's the meaning of imaginary part of n in COMSOL and the convention? What does positive or negative of imaginary part stand for? Gain, losses, scattering, reflection or what else?
For example, my work is about simlution of microresonator. And the extinction coefficient of water is 1.4680e-8. So what should I set the imaginary part, -1.4680e-8 or 1.4680e-8?
Thank you so much.
For example, my work is about simlution of microresonator. And the extinction coefficient of water is 1.4680e-8. So what should I set the imaginary part, -1.4680e-8 or 1.4680e-8?
Thank you so much.
↧
↧
How to use COMSOL batch to solve many files repeatedly without GUI?
Hi,
When I open an .mph file with a given geometry and settings, then click "Solve", COMSOL creates the mesh then solves the problem.
I would like to do this in batch mode (i.e., without GUI), however, I am not successful.
I think I should use something like "comsol batch -input myfile.mph" and that should suffice, however, COMSOL quits saying that the file doesn't contain a mesh. The COMSOL help files state that the .mph file's geometry must be meshed before using the above to solve the project.
Can anyone explain how may I *simply* tell COMSOL to open file after file after file, *create a mesh* and solve each?
Thank you
When I open an .mph file with a given geometry and settings, then click "Solve", COMSOL creates the mesh then solves the problem.
I would like to do this in batch mode (i.e., without GUI), however, I am not successful.
I think I should use something like "comsol batch -input myfile.mph" and that should suffice, however, COMSOL quits saying that the file doesn't contain a mesh. The COMSOL help files state that the .mph file's geometry must be meshed before using the above to solve the project.
Can anyone explain how may I *simply* tell COMSOL to open file after file after file, *create a mesh* and solve each?
Thank you
↧
Error: no 'smooth' property with 'mphinterp'
As specified in the reference guide, I could modify the 'smooth' property to force the continuity of the results from 'mphinterp' on element edges. But when I tried to do the modification, matlab told me there was no property called 'smooth' with 'mphinterp'.... Is there perhaps another name for this property?
And if my purpose was to smooth the extrapolation (for points outside the intended domain) results, does anyone know which properties help?
And if my purpose was to smooth the extrapolation (for points outside the intended domain) results, does anyone know which properties help?
↧
COMSOL 5.3 extract the electric field with mphinterp
Hi everyone,
I have a problem with the command mphinterp, since in the version 5.1 my code was working fine, while when i switched to the version 5.3, it is not working anymore.
I am using an adaptive mesh refinement as follows:
model.sol('sol1').feature('e1').create('ad1', 'Adaption');
And to extract the electric field (x component for example) of the modes of my fiber:
Ex_data = mphinterp(model, 'emw.Ex', 'coord',p, 'solnum',[Mode_num],'dataset','dset2');
When I run my simulation with the version 5.3, I have the following error:
The following feature has encountered a problem:
- Feature: Eigenvalue Solver 1 (sol1/e1)
Cannot use adaptation without referring to a study.
I am using the same code I was using with the version 5.1. I am wondering if something is changed between the two versions.
Many thanks for any answer.
I have a problem with the command mphinterp, since in the version 5.1 my code was working fine, while when i switched to the version 5.3, it is not working anymore.
I am using an adaptive mesh refinement as follows:
model.sol('sol1').feature('e1').create('ad1', 'Adaption');
And to extract the electric field (x component for example) of the modes of my fiber:
Ex_data = mphinterp(model, 'emw.Ex', 'coord',p, 'solnum',[Mode_num],'dataset','dset2');
When I run my simulation with the version 5.3, I have the following error:
The following feature has encountered a problem:
- Feature: Eigenvalue Solver 1 (sol1/e1)
Cannot use adaptation without referring to a study.
I am using the same code I was using with the version 5.1. I am wondering if something is changed between the two versions.
Many thanks for any answer.
↧
Domain number Assignment
Dear All,
I wrote a code for my comsol multiphysics problem using comsol livelink matlab. The code is working and producing desired geometry but there is problem regarding domain number as comsol assign domain number itself .
Is there any way to assign domain number of one's need while working in comsol livelink matlab .
Please suggest some solution.
-----------------------
Asha Sharma
I wrote a code for my comsol multiphysics problem using comsol livelink matlab. The code is working and producing desired geometry but there is problem regarding domain number as comsol assign domain number itself .
Is there any way to assign domain number of one's need while working in comsol livelink matlab .
Please suggest some solution.
-----------------------
Asha Sharma
↧
↧
I don't know what wrong with this model. Can someone helpme please
I try to use spiral antenna for RF heating but a program tell me that
Undefined variable.
- Variable: comp1.emw.etaRef
- Geometry: geom1
- Boundary: 34
- Feature: Time-Dependent Solver 1 (sol1/t1)
I am new for this program so i don't know how to solve this problem
Thank for all comment who try to help I really appreciated
drive.google.com/open?id=0B6QEE2-TG9t3ckhWLVFXeEZOVUE
Undefined variable.
- Variable: comp1.emw.etaRef
- Geometry: geom1
- Boundary: 34
- Feature: Time-Dependent Solver 1 (sol1/t1)
I am new for this program so i don't know how to solve this problem
Thank for all comment who try to help I really appreciated
drive.google.com/open?id=0B6QEE2-TG9t3ckhWLVFXeEZOVUE
↧
Draw Cut Line
Good morning,
I use Cut Lines on a specified plane within the "Data Set" section, in order to establish a measurement path for post-processing.
The goal is to compute the electric field on a specified zone.
I have two questions :
- to verify the position of the cut line (after having inserted the parameters for the start/end of the line...) I usually plot it on the geometry. This operation ("draw a line") requires several minutes to be completed. Why is this the case ? I have a powerful Workstation, so it is not a problem of computational power.
- Is there a way to define the resolution of the "Cut Line" ? How many point/unit of length are computed ?
- Is there a way to define different shapes within the dataset section (arc ?) In some occasions, you want to know the electric field along a curved line (which might not have a "simple" definition for its analytic expression...)
Thanks,
Daniele
I use Cut Lines on a specified plane within the "Data Set" section, in order to establish a measurement path for post-processing.
The goal is to compute the electric field on a specified zone.
I have two questions :
- to verify the position of the cut line (after having inserted the parameters for the start/end of the line...) I usually plot it on the geometry. This operation ("draw a line") requires several minutes to be completed. Why is this the case ? I have a powerful Workstation, so it is not a problem of computational power.
- Is there a way to define the resolution of the "Cut Line" ? How many point/unit of length are computed ?
- Is there a way to define different shapes within the dataset section (arc ?) In some occasions, you want to know the electric field along a curved line (which might not have a "simple" definition for its analytic expression...)
Thanks,
Daniele
↧
Error with Matlab function ....
Hello all,
I am calling a function "myfunction" from matlab into comsol desktop. While writing "myfunction" in the function section, arguments in the argument section and specifying the arguments limits i receive the error message "error using myfunction, too many output argumnets" when i click to plot.
Could anyone help what is wrong...
Regards.
I am calling a function "myfunction" from matlab into comsol desktop. While writing "myfunction" in the function section, arguments in the argument section and specifying the arguments limits i receive the error message "error using myfunction, too many output argumnets" when i click to plot.
Could anyone help what is wrong...
Regards.
↧
Bug with "Axis" item within "Views" ?
Hello,
within the item "2D plot group" I use different Views to get standardized images of my postprocessing results.
The target is to get a series of PNG files to export once the simulation is finished.
Each View (which is called by the corresponding 2D plot) contains a definition for the Axis.
Problem :
When I close the model, and re-open it (or when I store a model under a different name...) and then I plot the results again, the "Axis" settings are not taken in account anymore.
I have to manually "add" a constant value (like 0.0001) in order to "force" COMSOL to take in account of them, and give to the graphic window (and to the exported file as well...) the correct size.
Is this a knwon bug ?
How to solve this problem ?
Regards,
Daniele
within the item "2D plot group" I use different Views to get standardized images of my postprocessing results.
The target is to get a series of PNG files to export once the simulation is finished.
Each View (which is called by the corresponding 2D plot) contains a definition for the Axis.
Problem :
When I close the model, and re-open it (or when I store a model under a different name...) and then I plot the results again, the "Axis" settings are not taken in account anymore.
I have to manually "add" a constant value (like 0.0001) in order to "force" COMSOL to take in account of them, and give to the graphic window (and to the exported file as well...) the correct size.
Is this a knwon bug ?
How to solve this problem ?
Regards,
Daniele
↧
↧
ask for a question
hello, when I run the comsol 5.2a for computation a case. the error occurs in the following:
Failed to find a solution for the initial parameter.
Out of memory during LU factorization.
Returned solution is not converged.
- Feature: Stationary Solver 1 (sol1/s1)
I don't know how to resolve this question. i hope someone could help me. thanks.
but if the case use comsol 5.3 that is ok.
Failed to find a solution for the initial parameter.
Out of memory during LU factorization.
Returned solution is not converged.
- Feature: Stationary Solver 1 (sol1/s1)
I don't know how to resolve this question. i hope someone could help me. thanks.
but if the case use comsol 5.3 that is ok.
↧
Google Cloud computing with Comsol through Matlab
Hi everyone,
I need to run Comsol on the cloud and for specific reasons I am restricted to Google Cloud.
I am already using Matlab on the cloud.
My question is :
If I run Comsol through Matlab, which runs on the cloud with Techila, will I have the performance of the cloud machine or my laptop? In other words, when I call the Comsol kernel in Matlab, is the calculation happening in Matlab on the cloud or does the comsol instance calculate locally ?
Best,
Konstantinos
I need to run Comsol on the cloud and for specific reasons I am restricted to Google Cloud.
I am already using Matlab on the cloud.
My question is :
If I run Comsol through Matlab, which runs on the cloud with Techila, will I have the performance of the cloud machine or my laptop? In other words, when I call the Comsol kernel in Matlab, is the calculation happening in Matlab on the cloud or does the comsol instance calculate locally ?
Best,
Konstantinos
↧
Using the laminar flow module without setting inlet and outlet
Hello
I want to use laminar flow module, but I do not want to set inlet and outlet for my system. In the other words, I want to set the boundary condition for velocity on some of walls in my microsystem and then study the surface velocity field in my system by solving Navier-Stokes equation which is in this module.
Although, I set velocity of inlet to zero or use the pressure option in the inlet section and outlet too, but my results has been affected by the position of inlet and inlet I sat and when I disabled these option, I came up with error about the convergence problems…
Is there any solution not to use these options and solve Navier-Stokes equation? Or is there any other modules to study the Navier stokes equation without these options..?
In my real work, the solution is in the closed chamber and there is no inlet and outlet there.
I will be very grateful if anybody can help me
Many thanks in advance for your time and consideration
↧