Latest Discussions - COMSOL Forums https://www.comsol.com/forum/ Most recent forum discussions Wed, 21 Feb 2018 17:46:06 +0000 COMSOL Forum: Latest Discussions https://www.comsol.com/shared/images/logos/comsol_logo.gif https://www.comsol.com/forum/ Peristaltic axial tube- Walls of tube not closing https://www.comsol.com/forum/thread/166511/peristaltic-axial-tube-walls-of-tube-not-closing?last=2018-02-21T17:46:06Z <p>Hi,</p> <p>I am currently trying to build a peristaltic tube in Comsol. The model consists of an outter wall which is axial rotated to form a tube and water flows through the tube. Boundary loads are then applied to the tube to simulate acutuators acting on the tube. Picture shown below. These two loads then move up the tube and (hopefully) generate presure head and then top. I based my model on standard the peristaltic pump model that you can find on the comsol website and my model uses all the same physics.</p> <p>The issue is that i want the walls of the tube to close to the centre when the boundary force is applied. I can get the wall to reach about halfway by increasing the load, however if the load is increased passed a certain point then the model spits out errors and doesnt solve. It says that it may have reached a singularity and that it has not converged.</p> <p>Does anyone have any insight into how i could get the walls of the tube to close fully?</p> <p>Much appreciated,</p> <p>Rory</p> Wed, 21 Feb 2018 17:46:06 +0000 4.2018-02-21 17:46:06.166511 Successfully modelling gravel. https://www.comsol.com/forum/thread/166501/successfully-modelling-gravel?last=2018-02-21T13:13:51Z <p>Hi all comsol wizards!</p> <p>I'm trying trying to model gravel as isolation in a floor between apartments. As for now i have modelled the gravel isolation as a block in comsol 5.3a and if i now assign the block any certain material, it will have a stiffness matrix and take up bending stress and in other words behave as a solid.</p> <p>How do i prevent this and create a non-solid object? Is it correct to define a new linear elastic material in the solid mechanics field and reduce its stiffness parameters to zero?</p> <p>Thank you in advance!</p> <p>Best regards, Mark</p> Wed, 21 Feb 2018 13:13:51 +0000 4.2018-02-21 13:13:51.166501 1D line plot in 2D axisymmetric geometry https://www.comsol.com/forum/thread/166491/1d-line-plot-in-2d-axisymmetric-geometry?last=2018-02-21T13:16:18Z <p>Hi!</p> <p>I have a model with a 2D axisymmetric geometry and now, for the postprocessing, I need a 1D plot line that goes from -r to r, instead of just 0 to r.</p> <p>I used the 2D revolution, then "Cut Line" to define the line where I want to evaluate my results, and when I plot it it's ok, as long as I use "Arc Length" for my x-axis. When I change the expression to "r", it only plots my results from 0 to r. I tried to define a range for my x-axis using "range(-x,step,x)", but COMSOL doesn't accept that range as an expression.</p> <p>I'm attaching a couple of figures so it's easier to understand what I mean. In both figures I used exactly the same data set (cut line), only changed the expression for x.</p> <p>Can anyone help me with this?</p> <p>Thanks!</p> <p>Clarissa</p> Wed, 21 Feb 2018 11:16:31 +0000 4.2018-02-21 11:16:31.166491 error: Curl operator only supported for vector components in the differentiation system. - Expression: ∇×Eb - Variable: curlEb - Provider: Background Field https://www.comsol.com/forum/thread/166481/error-curl-operator-only-supported-for-vector-components-in-the-differentiation-?last=2018-02-21T09:25:33Z <p>I am using EM-frequency,</p> <p>I got this error:</p> <p>Error in multiphysics compilation. Curl operator only supported for vector components in the differentiation system. - Expression: ∇×Eb - Variable: curlEb - Provider: Background Field What possibly could the reason be?</p> Wed, 21 Feb 2018 07:00:03 +0000 4.2018-02-21 07:00:03.166481 Stokes Equation https://www.comsol.com/forum/thread/166471/stokes-equation?last=2018-02-21T06:07:51Z <p>I want to simulate my problem using Stokes equation and convection diffusion equation. I want the Stokes equation to not have any transient term. But the creeping flow module has time dependency, as I am doing a time-dependent study and the convection diffusion equation has a transient term, which is required to see the species evolution. Both the modules are coupled, so I cannot do a separate study, i.e., stationary for Stokes and transient for convection diffusion equation.</p> <p>A way around this problem would be modeling the Stokes equation using PDE module. But I am unable to create a Stokes equation, grad(p) = mu * laplacian(u), using the PDE module.</p> <p>Any kind of suggestion would be appreciated.</p> Wed, 21 Feb 2018 06:03:34 +0000 4.2018-02-21 06:03:34.166471 PDE is coupled with ODE https://www.comsol.com/forum/thread/166462/pde-is-coupled-with-ode?last=2018-02-21T04:44:03Z <p>How can i solve a system of equation in which a PDE is coupled with ODE (as attachment file)?? this is a mass transfer with reaction by considering boundary interface. I am a bigginer in comsol, so i would appreciate if anybody can help me.</p> <p>Thanks in advance,</p> <p>Best regards, Mojtaba</p> Wed, 21 Feb 2018 04:44:03 +0000 4.2018-02-21 04:44:03.166462 Sample Model of a Single-Sheet Tester https://www.comsol.com/forum/thread/166441/sample-model-of-a-single-sheet-tester?last=2018-02-20T22:47:50Z <p>Hi,</p> <p>I am trying to study a single sheet tester and need a sample model of one. Any help is greatly appreciated.</p> Tue, 20 Feb 2018 22:47:50 +0000 4.2018-02-20 22:47:50.166441 "Unexpected unit of input" error for material property defined by piecewise function https://www.comsol.com/forum/thread/166432/unexpected-unit-of-input-error-for-material-property-defined-by-piecewise-functi?last=2018-02-21T09:37:45Z <p>Hello,</p> <p>I am using Comsol 4.2. defined a temperature-dependent heat capacity using a piecwise function. The argument is T and I defined the units of the argument and function as K and J/(kg*K), respectively. I named the function Cp_eff. When I enter Cp_eff(T) under the Value column for the Heat capactiy property, the entered value turns orange with the warning "Unexpected unit of input". The function unit is correct, so I don't know what the problem could be.</p> <p>Thanks.</p> Tue, 20 Feb 2018 21:16:18 +0000 4.2018-02-20 21:16:18.166432 Export comsol data file to tecplot https://www.comsol.com/forum/thread/166422/export-comsol-data-file-to-tecplot?last=2018-02-21T07:45:12Z <p>Is there any way to export COMSOL surface and contour plot to export in tecplot?</p> Tue, 20 Feb 2018 17:24:05 +0000 4.2018-02-20 17:24:05.166422 3D Coil with Assembly https://www.comsol.com/forum/thread/166412/3d-coil-with-assembly?last=2018-02-20T15:07:48Z <p>Hello,</p> <p>I am currently working on a 3-D, time-dependent model that uses the magnetic fields interface. This model uses the Coil feature to excite a 3-D conductive structure with a specified voltage waveform. This part of the model works as expected.</p> <p>I am now trying to add some additional functionality which allows part of the "coil" to move with respect to the rest of the conductive structure. I have implemented the moving mesh feature as well as some global ODEs to describe the movement. In addition, I've switched the geometry to an assembly and defined an identity boundary pair between the moving and stationary structures. I then defined a continuity boundary condition within the magnetic fields interface for this identity pair.</p> <p>Now, the coil functionality is not working as before. Even with the moving mesh and ODEs removed, the coil feature appears to not function properly using an assembly with identity pairs. The model will solve, but no current flows through the "moving" portion of the coil.</p> <p>Is there something simple I have missed somewhere that will allow the coil feature to function properly with an assembly? Alternatively, is there a better way to excite my "coil" aside from the coil feature? Unfortunately, I am unable to upload my .mph file due to some sensitive information within, but I would appreciate any advice you can offer.</p> <p>Thanks!</p> Tue, 20 Feb 2018 15:07:33 +0000 4.2018-02-20 15:07:33.166412 Splitting into several domains https://www.comsol.com/forum/thread/166411/splitting-into-several-domains?last=2018-02-20T14:54:17Z <p>The below post is related to an <a href="//www.comsol.com//forum/thread/68161/how-to-split-into-different-domains-when-importing-cad-files-by-other-softwares">archived discussion</a></p> <hr /> <p>Hello,</p> <p>I have a rectangular geometry with several channels, which water is supposed to flow through. Other than the channels, the rest of the geometry should be Aluminium. The geometry is imported from SolidWorks as a STEP file.</p> <p>I am trying to investigate the heat transferred and the flow of the water but I encounter the problem when the geometry is imported, it becomes one whole domain, thus creating problem when I'm trying to assign water as material inside the channels.</p> <p>My question is, how can I separate the channels into a different domain? I tried 'split' operation but did not seem to work.</p> Tue, 20 Feb 2018 14:46:52 +0000 4.2018-02-20 14:46:52.166411 IntVolume Parameter in MatLab https://www.comsol.com/forum/thread/166402/intvolume-parameter-in-matlab?last=2018-02-20T14:52:34Z <p>Is it possible to add the IntVolume expression parameters with a variables outside of the model? I want to calculate the coupling coefficient using a data from two different runs. I assigned the first data to a variable and tried to use it as a parameter for IntVolume function together with a second data but I am not sure about the syntax.</p> <p>Ez1=mpheval(model,'emw.Ez');</p> <p>model.result.numerical.create('int2', 'IntVolume'); model.result.numerical('int2').set('expr', 'emw.Ez'*Ez1);</p> Tue, 20 Feb 2018 14:52:34 +0000 4.2018-02-20 14:52:34.166402 Infinite element domain parameters https://www.comsol.com/forum/thread/166401/infinite-element-domain-parameters?last=2018-02-20T10:59:26Z <p>Does anyone know the meaning of two parameters, Physical Width and Pole Distance defined for infinite element domain? Or if possible, can anyone provide a reference?</p> <p>And the region for the infinite element as shown in the figure is generated automatically or I need to create a circle myself? I use axissymmetric 2D and version 4.3</p> <p><img src="http://" alt="https://imgur.com/a/WoYJj" /> Thanks!</p> Tue, 20 Feb 2018 10:32:07 +0000 4.2018-02-20 10:32:07.166401 memory management https://www.comsol.com/forum/thread/166391/memory-management?last=2018-02-20T09:19:15Z <p>Hello everybody, i am using comsol 5.3 and i have sumile a lot of dof ~ 1000 million and i have out of memory error is there a possibility of stoker derictly on hard desque without using RAM? thank you. N.B. I use a computer core i5 2.38 Ghz 16Gb RAM</p> Tue, 20 Feb 2018 09:18:52 +0000 4.2018-02-20 09:18:52.166391 undefined variable comp.2.ht.rhoref, coupling between heat transfer and laminar flow after upgrade to 5.3a https://www.comsol.com/forum/thread/166373/undefined-variable-comp-2-ht-rhoref-coupling-between-heat-transfer-and-laminar-f?last=2018-02-20T05:31:11Z <p>Hi everyone,</p> <p>I m modelling a coupling between heat tranfer and laminar flaw using a nitf multiphysics coupling. My model runs welll in 5.3 but when i use 5.3a an erro msg comes out as if the solver can not find the reference density though it is set to be taken from ht interface.</p> <p>error: undefined variable. -variable: comp1.ht.rhoref -Geometry: 11 Failed to evaluae the Jacobian of expression : dvol(((-comp1.ht.Ceff)comp1.ht.timeDerivative )Comp1.ht.</p> <p>Can anyone give me a quick hint?</p> <p>Regards,</p> Tue, 20 Feb 2018 05:31:11 +0000 4.2018-02-20 05:31:11.166373 Magnetostrictive energy harvester using villari effect https://www.comsol.com/forum/thread/166363/magnetostrictive-energy-harvester-using-villari-effect?last=2018-02-20T02:55:53Z <p>I am going to design a magnetostrictive energy harvester using villari effect. Where if we applied a compressive and tensile stress on a magnetostrictive material with the present of parmanent magnet bias field and if a coil is wound around the magnetostrictive material a e.m.f is generated across the coil. can anybody say how can i apply a continuous compressive and tensile stress on the trip of the harvester..</p> Tue, 20 Feb 2018 02:55:53 +0000 4.2018-02-20 02:55:53.166363 getting error for failed to find a solution: singular matrix https://www.comsol.com/forum/thread/166361/getting-error-for-failed-to-find-a-solution-singular-matrix?last=2018-02-19T14:56:37Z <p>hii, I am getting error 1 void equations for mesh : case 0 : empty rows in matrix in COMSOL multiphysics 5.2 during computation time. Can anyone help me for resolving this error. I am using EM waves frequency domain. According to previous discussion I have changed the conductivity from default 0 to 0.1. Still I am getting same error given below.</p> <p>**Failed to find a solution. Singular matrix.</p> <p><strong>For mesh-case 0 there are 1 void equations (empty rows in matrix) for the variable comp1.emw.S2x. at coordinates: (0,0,0), ... and similarly for the degrees of freedom (empty columns in matrix). Returned solution is not converged.</strong> **</p> Mon, 19 Feb 2018 14:56:37 +0000 4.2018-02-19 14:56:37.166361 Modelling 'Helix' Turbine- Meshing Problems https://www.comsol.com/forum/thread/166351/modelling-helix-turbine-meshing-problems?last=2018-02-19T15:13:37Z <p>Hi All,</p> <p>I'm trying to model an alternative type of turbine, which is a version of vertical axis turbine that is deployed horizontally in a micro hydro electric power setting.</p> <p>I'm relatively new to COMSOL, but I think I've created an accurate geometry model for the turbine. However, when I try to mesh the geometry, I'm getting numerous errors around edges being to small to compared to the element size. I've tried to mesh on the smallest size, but I still have the issue.</p> <p>I believe the issue comes from the fact that the blades slightly overlap their casing (Turbine rotates with casing, not on a central axis), however this is key for providing a flush finish. I attempted to use a union to remove the interior boundries, between the overlapping blades and the casing. This works at one end of the turbine, however, at the opposite end the interior boundries aren't removed and I believe this is causing issues???</p> <p>I've also tried changing the mesh manually, but I'm still getting the same issues.</p> <p>Any meshing tips would be greatly appreciated!</p> <p>Also my final goal is to calculate the torque of the turbine, and my current calculate is to perform a surface intergration with the expression:</p> <p>p<em>(ny</em>x-(y+0.2)*nx)</p> <p>I'm also just using the turbulent flow physics package. If anyone can let me know if im on the right lines, or if I should be using a different physics package to find the torque from the flow please let me know!</p> <p>Sorry for the questions!</p> <p>Tom</p> Mon, 19 Feb 2018 14:43:16 +0000 4.2018-02-19 14:43:16.166351 reducing comsol file size https://www.comsol.com/forum/thread/166331/reducing-comsol-file-size?last=2018-02-19T11:52:03Z <p>Hi all, I am trying to reduce the size of my comsol file. I need to define several cut planes and then tell comsol only save the T distribution over these planes (x,y,z,T). is this possible? to define a explicit surface, i can only see the boudaries of the model can be selected as surface and does not allow me to add my cutplanes as surfaces of interest.</p> <p>Regards, Asal</p> Mon, 19 Feb 2018 11:52:03 +0000 4.2018-02-19 11:52:03.166331 Export 1D Line Graph to spreadsheet - how to prevent sorting? https://www.comsol.com/forum/thread/166321/export-1d-line-graph-to-spreadsheet-how-to-prevent-sorting?last=2018-02-19T10:56:05Z <p>I created a 1D Line Graph along a custom selection of lines (boundaries). The plot looks as intended (see lineplot_correct--.png).</p> <p>However, I can't export this plot! I do Export / Plot / Line Graph and the resulting spreadsheet is sorted by the x value (see lineplotafterexport--.png), which is a problem when the curve is going "backwards". If I check Advanced / Sort or not doesn't make any difference.</p> <p>Thanks in advance for any suggestions!</p> Mon, 19 Feb 2018 10:39:38 +0000 4.2018-02-19 10:39:38.166321