Latest Discussions - COMSOL Forums Most recent forum discussions Thu, 25 Apr 2019 17:24:14 +0000 COMSOL Forum: Latest Discussions Comsol 5.2 not starting on Debian <p>Hello there!</p> <p>I'm desperately trying to get Comsol 5.2 running on Endless 3.5.8 (Linux/Debian) but it won't.</p> <p>Can anyone help me out here? Thanks in advance, Kevin</p> Thu, 25 Apr 2019 17:24:14 +0000 4.2019-04-25 17:24:14.232871 simple test case not converging. <p>Hi all,</p> <p>I have a simple setup to understand the logic of COMSOL and FE simulation:</p> <p>linear elasticity, strutural mechanics interface, time-dependent study type.</p> <p>a block of aluminum, give a "prescribed displacement" to the whole domain through expression "1*step1(t/1[s])".</p> <p>The step function is given to introduce a gradual increment of displacement.</p> <p>what I find <strong>mysterious</strong> is that:</p> <p>when the step function possesses an <strong>non-zero displacement at t=0</strong>. this simple model converges.</p> <p>However, when the step function starts from <strong>zero displacement at t=0</strong>. this model does not converge.</p> <p>what is the reason?</p> Thu, 25 Apr 2019 14:38:57 +0000 4.2019-04-25 14:38:57.232861 Thermally and Hydrodynamically Fully Developed Flow <p>Hi, I am trying to model a thermally and hydrodynamically fully developed flow (laminar flow) in a channel. I am using the "Fully developed flow" boundary condition for the "Inlet" (hydrodynamically fully developed), but I don't know which boundary condition is suitable in "Heat Transfer in Fluids" to model the thermally fully developed flow???</p> Thu, 25 Apr 2019 13:08:56 +0000 4.2019-04-25 13:08:56.232853 how to couple two shell face? rotation component? <p>I know that I can define the prescribed displacement to make the two boundaries coupled in solid mechanics module. But how can I do this in shell module, specifically the rotation component?</p> Thu, 25 Apr 2019 11:26:32 +0000 4.2019-04-25 11:26:32.232851 what's the meaning/definition of initial magnetic susceptibility <p>hello! I am studying the ''non linear_magnetostriction'' model in COMSOL. I'm confused about the meaning of "initial" in "initial magnetic susceptibility", whcih is a parameter of magnetostrictive material. Does it mean the magnetic susceptibility will change over time? But the study type is stationary, which means varibles do not change over time, so does initial magnetic susceptibility equal to magnetic susceptibility? thanks!</p> <p>Regards Yi</p> Thu, 25 Apr 2019 07:41:04 +0000 4.2019-04-25 07:41:04.232843 How to choose a particular freq? <p>I get 4 eigenfrequencies by the eigenfrequency study and study step, and now I want to caculator the width of the 1st and 2nd eigenfrequency(for example: freq(2)-freq(1)),but I dont know the correct syntax.</p> Thu, 25 Apr 2019 02:41:22 +0000 4.2019-04-25 02:41:22.232833 Java: Invalid cd root dir error <p>Hi I'm trying to install Comsol on a Ubuntu 18.04 machine. However when I run:</p> <p>./setup in the installation folder the terminal prints:</p> <p><code>./setup: line 756: [: /home/joey/Documents/2019-4/Install: binary operator expected Extracting temporary COMSOL installer components in /tmp...</code></p> <p>Following that a Java screen pop up which reads: <code>java.lang.RuntimeException: Invalid cd root dir: /home/joey/Documents/2019-4/Install</code></p> <p>I followed and already searched to web for similair stories. I am running Java 1.8.0_191</p> <p>Any clue what this could be and how to resolve it?</p> Wed, 24 Apr 2019 20:12:59 +0000 4.2019-04-24 20:12:59.232821 Flux boundary condition <p>I recently noticed the flux boundary condition in the new version is different from the old versions. For the flux boundary <strong>n</strong>.<strong>j</strong>=j_0 is <strong>j</strong> in the domain always considered as the diffusive <strong>j</strong>=-D(grad c) in the newer version? If I enable convection in the domain, does this the flux automatically switch to <strong>j</strong>=-D(grad c)+<strong>u</strong>c or I will have to manually adjust for the convective flux (<strong>u</strong>c) in terms of expression of j_0?</p> Wed, 24 Apr 2019 17:15:54 +0000 4.2019-04-24 17:15:54.232812 Adding a Mass Flux to Slope Stability Model <p>So I am trying to add a mass flux to a model with richards EQN to a model, but it doesnt add the way it should be adding. How do i add a mass flux to a model</p> Wed, 24 Apr 2019 16:33:36 +0000 4.2019-04-24 16:33:36.232802 Phase field with mass transfer <p>Hi !</p> <p>I'd like to use a phase field model (meaning a potential like Cahn-Hilliard with interface energy) but with mass transfer between the phases. Unfortunately, the Phase field Physics doesn't allow directly to do so, even now one can divide the nonconservative convection term by phipfx and phipfy, but still, the potential is imposed and sometimes I'd like to change it a little bit. (Another point is that in the ternary phase field there is apparently no possibility to implement a non-conservative flux, which is annoying if you want to introduce another field.)</p> <p>I tried to do it with the Mixture Model Physics, and with the Stabilized Convection Diffusion Equation but nothing gave me a satisfying result. I'm not satisfied with the sizes of the interfaces, and also the fact the the mesh is too apparent in the patterns. I tried to implement an adaptative mesh but it's very very slow and until now it has never really converge.</p> <p>Have you ever tried to do something similar ? If not would you have an idea how to do it ?</p> <p>Thank you in advance</p> <p>Joseph</p> Wed, 24 Apr 2019 12:49:20 +0000 4.2019-04-24 12:49:20.232791 Movable boundary condition while using stationary study <p>Dear all,</p> <p>I want to model a process in steady state modling (like etching a sacrificial layer with a constant rate) but the boundary condition (BC) is required to be changed for every steps (imagine a beam underneath of which has be released step by step). This could be possible by applying new BC in each step, but as there might be thousands of them, it is a really time consuming process. I was wondering if there is a way to change the BC automatically for each step. In transient study I know that is is possible by defining a time dependent BC, but for steady state as there is no time, I think it would not be not possible to use such a method. would you tell me what to do to address this issue?</p> <p>Thank you so much</p> Wed, 24 Apr 2019 12:09:23 +0000 4.2019-04-24 12:09:23.232771 Phase Change in Two-Phase Flow <p>Dear all,</p> <p>I am trying to solve a two-phase flow problem, where one of the fluids is a gas and the other a polymer melt. Isothermal computation works out of the box without any severe problems.</p> <p>However, I want to make a non-isothermal simulation, where heat is transfered from both fluids to the surroundings (Dirichlet condition with constant temperature for the beginning). Now as maybe known, polymers tend to phase transitions at their glass transition or melt temperature and essentially solidify, when they are cooled down to the respective temperature.</p> <p>I wondered, how I can implement this in two phase flow with a level set method. So straight forward I added "Heat Transfer in Solids and Fluids" and added the polymer as a phase change material. Started the computation, which ran without a problem and got some results. However, it appears there is no coupling of the temperature fields at all and subsequently no solidification of the polymer melt.</p> <p>I wonder wether I did something wrong or this method just does not work.</p> <p>Also, maybe there is a different approach to solving this. I also thought about directly implementing this into the material properties, but still working on how to do it properly. Just setting the viscosity to some value won't work I guess.</p> <p>Does anybody of you has experience with this kind of model and can help me out with advice?</p> <p>Thanks and best regards, André</p> Wed, 24 Apr 2019 10:32:41 +0000 4.2019-04-24 10:32:41.232761 Randomize any surface thickness <p>Hi everyone! I would like to know how I can make rough curve surfaces. To clarify: I started generating a random surface, as explained <a href="">here</a>, but I want to fit it to a curve surface, such as a sphere or, better, to a part of it. How can I do that? Thanks</p> Wed, 24 Apr 2019 09:40:34 +0000 4.2019-04-24 09:40:34.232751 Corona discharge in COMSOL <p>Hello guys, I just started to learn COMSOL, and I found a weird thing that makes me confuse a lot. I was trying to simulate the plating on the cathode and checking the corona discharge. First, I made a cone structure and the current density seems reasonable.(Figure 1)</p> <p>However, when I increase the height of surrounding, the concentrated current density on the tip just disappear and change to the edges. (Figure 2)</p> <p>I think it might change but not like this. I used the axisymmetric model in this case, and it remains the same if I change to 2D or 3D models.</p> Wed, 24 Apr 2019 05:48:43 +0000 4.2019-04-24 05:48:43.232742 Scaling of heat transfer coefficient with Thickness in 2D heat transfer/fluid flow model <p>Hi,</p> <p>I'm modeling heat transfer of a simple cylinder in convective flow in 2d, with a defined volumetric heat source (General Source in W/m^3).</p> <p>If I add a constant heat flux to the outside of the cylinder, and change the heat transfer thickness dz, the temperature of the cylinder is unchanged, as I would expect (since the heat source is volumetric).</p> <p>However, if I add CFD turbulent flow and do a multiphysics solution, the temperature changes dramatically when dz is changed (ie, the effective heat transfer coefficient has changed).</p> <p>Anyone have a clue what may be going on here?</p> Tue, 23 Apr 2019 23:33:55 +0000 4.2019-04-23 23:33:55.232722 Using output of a study as input in a new study. <p>Hello,</p> <p>I am trying to run a time dependent CFD problem with a variable flow rate, and obtain the reaction forces on the wall of the lumen. Then I would like to apply these reaction forces in a new time dependent solid mechanics problem to obtain the stressed inside this material. I can't however figure out how to properly apply the time dependent load on the wall in the second study. Any help would be greatly appreciated.</p> <p>I am using COMSOL version 5.4.</p> <p>Kind regards, Max</p> Tue, 23 Apr 2019 10:42:48 +0000 4.2019-04-23 10:42:48.232711 Names of Boundaries <p>How can the boundaries that there are in COMSOL be named according to their names in the CAD geometry imported from SolidWorks?</p> <p>×Names do not need to be "string", being "number" is enough.</p> Tue, 23 Apr 2019 08:24:51 +0000 4.2019-04-23 08:24:51.232702 Far field profile of grating <p>Hello,</p> <p>I am trying to simulate the far field of a grating. In my simulation I am using periodic boundy conditions and a port exitation. When I am cacluting the far feild profile, it seem that it did not take into account that my stracture is perioidic.</p> <p>Is there any way of taking into account finat\infanat mumbers of periods in the Far fiel simulation?</p> <p>Thank you very much!!!</p> Tue, 23 Apr 2019 00:18:47 +0000 4.2019-04-23 00:18:47.232692 how to plot the specific value vs time? (heat front vs time) <p>Hi there,</p> <p>Could you please tell me how to plot the curve of the specific value vs time? Such as the 1D transient heat transfer case (Tinitial=10 deg, Ttop=10 deg, Tbot=-10 deg), if I want to show the position of T = 0 deg vs time, how can I do? I attached the model file. Many thanks!</p> <p>The depth of 0 degree heat front vs time.</p> <p>Xi</p> Mon, 22 Apr 2019 23:22:03 +0000 4.2019-04-22 23:22:03.232682 How to Apply Convection Boundary Condition for Concentrations in Transport of Diluted Species Model <p>Hello,</p> <p>I am trying to apply a diffusive boundary layer as my upper bounary condition in my reaction transport model. My upper boundary should include an external concentration and a gradient.</p> <p><img class="latexImg" src="data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAASkAAAArCAMAAAAT+I4PAAAABGdBTUEAALGPC/xhBQAAAAFzUkdCAK7OHOkAAAAgY0hSTQAAeiYAAICEAAD6AAAAgOgAAHUwAADqYAAAOpgAABdwnLpRPAAAADlQTFRF////AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA////AuJdXgAAABF0Uk5TAIi7M5mqEWbMIkTdVe530ci0He6bAAAAAWJLR0QAiAUdSAAAAAlwSFlzAAAAeAAAAHgAnfVaYAAABRpJREFUaN7tWum6qygQpAFB0DDj+7/sAA3KYiIq92b5pn6E5BxppUIXvYSQ//E3AJSGd4wPnIMQ0MOsBIQU715gP8DoB8GpcgPwLkwRvUj7qmBi715gN1DPjBg5fpwX3cWsXJQfYXr3ArthmYkjalg/9zHLA0PC761fgF6cksBj1ZMgW/NNu1PYo2Tp483vhQKYvUypZDlIEaMXbUbT61Zablr6BDAqiH44jiCIygZ9c31Rpqz3ff+eEg+n3V7Ax0p27zIVZcqeEN+vU9z5nfYCXnvIXaZWmeJWBwVo+Oa4Cv3OM5IwZXC4ydQqU8LdxUZqYnz3cq+Deb/DaGpceZFBXm4ytcoUTAJPB/qdEahWMTywdMkkSFAmXnGPqShT8mEJMp6p7xR273gPS5dZBJu3CJ1Fou4yhTIlYHRbC76XKeZ9Q3INjM6YzNBh1hq2cPMWUzAs1GbHlKNYFUzJjn740pYR+dhmMUvtUcaJcLeJWZ6SJk347p59KWbPVAwWdNy312oX2axoi6m9SwXPxzZkqf14nKr0ZEo40ZpiJDrEx79Su8hnBVsz6JnnXAm/Qh1M6zO3SFN71ZDes3/6MWV3tIK4pTju3Gu1i2IW2vLLYXkUInGJNPBHVesd8tTemOPr2b8dmSJKx0cNcdW12kUxK9jCTZn7CUdG50jsCf9LU3vWoHA9vS+FREeoaxctKGYFW/4gJzQjYwpb+EHycQfIhqQDfpmnU/s/xdQw4+OUtYsWlLPQFmauhG4PrAEWDv5UpMG16XMX9zaBz+GYO53atzHF9Ia2s3/0j1zXLlpQzkJbGpmC9IFlVGJu8rGGT0q020e4KU+n9m1MQYomu6hK46WScTkLbUWmUr3j0RXjUyVPJ/Jv11PoX/CSLLUP93kG2s7UEzaeQK+rO5YAsbNby1k5UymNUab2mFLJlzsjP4ytdOWpfQv+lE5VTD1xi2I9u7PQlvKNgFSn7J+im+4wVQCiNeYpylL7jkyZq95X1S5aUM4K/obKS4ftwlWmGpgaBLKlcpnyqX1HplJFbwsgJ3SVrHZhA+9B4zZ6yVtZ8UBbZMB4KtmcTqbQiULmuY47C52kzYkEULSdpfYdmToPPNnz2gW4cr5kriz6MiYuKx4hStBuGyjLoprCP2w2zvB/sSr2ojrGqJXQIEpFal8BaE1LJ6Zme+9s4xv8lNYu/KI0ipB6uTOLikewRQwXylGhHsEDJZ8DZw2Rp4s2mqsbUJduO+V9di25W6p4q612gRdwLwziwIezisdmy4SfP6yFkeA7TdlMSxocsVNl65T31Q/Bay7cVpfAnTScKiXVtsp1xAoDf6U6ZiDNWOqMopP3DZXwqPq5mJUG485lGM51rktbrGCOQT7u40Q1Vi/1N9mJKV0ftrJj/6+wVX4tXOTjPloKBiRpuAsX4Zm1jPSUqVO9O2Gm2v8+uzr8DGvDXRhXuBDzwg6YOtO7Y1zpH2ghO2wNd8e7C0OOulhnenfClaLHE3r5wdga7m7tj0RPkSk2g5ZZ4Hymd+czeviBn7qQrOHueluJP3umrEvqkUxsS1fFmd6dZx5+Yk+lDXe7XVL5QaZElVqeYAp/HEIbivafj7ThjpmmKHXKl1W3YlHRu3sJ5rsCjyvVzc/D2nDXC/Olq/VA9UwxSSyX4CKI6H1Z7+4A7jJzru/5sVgb7np0b/QWnvgzznBjc838oEt7dwdwBn+EqKTh7t5k6enTQGDr3R3jxKV/A/8BOjAnj6B0iB8AAAAldEVYdGRhdGU6Y3JlYXRlADIwMTktMDQtMjVUMTE6MTM6NTItMDQ6MDB/e2oLAAAAJXRFWHRkYXRlOm1vZGlmeQAyMDE5LTA0LTI1VDExOjEzOjUyLTA0OjAwDibStwAAACJ0RVh0cHM6SGlSZXNCb3VuZGluZ0JveAAxNzh4MjYrMjE2KzYyONWd57wAAAAndEVYdHBzOkxldmVsAEFkb2JlRm9udC0xLjA6IENNU1kxMCAwMDMuMDAyCncbARAAAABJdEVYdHBzOlNwb3RDb2xvci0wAC9kZXYvc2htL3pmMi1jYWNoZS9lNWUwZTlkZTVmYTJkNDMwYjNhNjdmYjAxNTcyOGU4OC5kdmkgLW/woaGrAAAARXRFWHRwczpTcG90Q29sb3ItMQAvZGV2L3NobS96ZjItY2FjaGUvZTVlMGU5ZGU1ZmEyZDQzMGIzYTY3ZmIwMTU3MjhlODgucHPzA5O2AAAAAElFTkSuQmCC" /></p> <p>The closest approach for my boundary condition is in this heat transfer model: <a href="http://"></a></p> <p>However I don't see a simillar node when dealing with concentrations in a transort of diluted species physics.</p> <p>Any suggestions?</p> <p>Many thanks, Yaniv</p> Mon, 22 Apr 2019 21:33:46 +0000 4.2019-04-22 21:33:46.232671