Latest Discussions - COMSOL Forums Most recent forum discussions Sat, 16 Dec 2017 16:52:23 +0000 COMSOL Forum: Latest Discussions heat transfer coefficient <p>Hello guys,</p> <p>I'm having some problems trying to work on a quite simple model, that you can see attached. 3 of the walls (half tube, and the 2 pieces of carter) have 3 different temperatures and there is air in the middle. I would like to know the exchange of energy between the 3 elements due to convection and I'm using a stationary solver. The strange thing is that I was able to have a solution for a lot of initial conditions but sometimes, changing the temperature, the model doesn't work anymore giving the following error: Undefined value found in the equation residual vector. There are 260 degrees of freedom giving NaN/Inf in the vector for the variable comp1.T. at coordinates: (0.11073,-0.054257), (0.110392,-0.0503973), (0.110246,-0.0612593), (0.110176,-0.0575849), (0.109958,-0.053865), ... There are 447 degrees of freedom giving NaN/Inf in the vector for the variable comp1.u. at coordinates: (0,-0.062), (-5.08159e-020,-0.0612062), (-1.11795e-019,-0.0602536), (0.00382759,-0.062), (0.11073,-0.054257), ... There are 447 degrees of freedom giving NaN/Inf in the vector for the variable comp1.v. at coordinates: (0,-0.062), (-5.08159e-020,-0.0612062), (-1.11795e-019,-0.0602536), (0.00382759,-0.062), (0.11073,-0.054257), ...</p> <p>. How could I solve a problem like this?</p> <p>ps: for this type of study is better to use the gravity or to define a gravity force?ps: for this type of study is better to use the gravity or to define a gravity force?</p> Sat, 16 Dec 2017 16:52:23 +0000 4.2017-12-16 16:52:23.162091 Enhanced light-harvesting capability for silicon single-nanowire solar cells coupled with metallic cavity <p>I want to simulate the enhanced light-harvesting capability for silicon single-nanowire solar cells coupled with metallic cavity with the TE, TM and non polarisation incidences. But I don't know how I do it because I am a beginner in comsol, someone can help me to do this simulation</p> Sat, 16 Dec 2017 09:56:33 +0000 4.2017-12-16 09:56:33.162081 Solidification <p>Hi, all,</p> <p>Anyone has a solidification model or similar example here? Especially the KKS model of binary alloy.</p> <p>Thanks, Xing</p> Sat, 16 Dec 2017 09:18:23 +0000 4.2017-12-16 09:18:23.162072 Heat and moisture transport in comsol? <p>I would like to medelize heat and moisture transfer from building ground level through unsaturated soil, with time dependent study . What physics should i use for it ?</p> Fri, 15 Dec 2017 23:22:50 +0000 4.2017-12-15 23:22:50.162071 Novice Question - Partitioning Domains in 2-D <p>Hello all,</p> <p>I'm working on a Comsol design of a device through which fluid will flow. I used SolidWorks to design the device, and imported the relevant face as a .DXF to obtain the 2-dimensional cross section that I want to model flow through. The overall design is thus: <img src="" alt="Design" /></p> <p>I'd like to keep this domain: <img src="" alt="Domain to be kept" /></p> <p>And subtract these domains: <img src="" alt="Domains to be subtracted" /></p> <p>I've used the "partition domains" function to identify/highlight the relevant portions, but I seem to be unable to use the "difference" tool to delete the domains I want gone. I need to use the edges of the to-be-subtracted domains as walls, and the vertical lines connecting the innermost rings as the outlets.</p> <p>Any help is greatly appreciated, please let me know if more clarification is needed.</p> Fri, 15 Dec 2017 19:26:37 +0000 4.2017-12-15 19:26:37.162052 Initial particle velocity changes in time. <p>Hello, I am tracing electrons in the electric field (2D) and I want to investigate changes of their velocity vectors while moving. So, I want to compare actual angles (atan2(vy,vx)) with initial ones. To do this I need initial particle velocities cpt.v0xinl1 and cpt.v0yinl1 - but these initial velocities change with time. Even in case of no electric field, when the tracks are stright lines and cpt.vy is a constant value (which is expected), the cpt.v0xinl1 changes with time. How can initial value change in time? How can I obtain initial velocity vectors for each particle? Best regards, Jacek.</p> Fri, 15 Dec 2017 17:14:45 +0000 4.2017-12-15 17:14:45.162051 Now available: COMSOL Multiphysics® 5.3a – Features innovative modeling tools, faster solvers, and more <p>Hello,</p> <p>I'm pleased to announce that COMSOL Multiphysics® 5.3a is now available for download. For details, visit: <a href=""></a></p> <p>Release highlights:</p> <ul> <li>Hybrid Boundary Element-Finite Element Method for acoustics and magnetostatics</li> <li>New modeling and visualization tools, faster solvers</li> <li>Thermodynamics properties library, new Turbulence model, moisture flow multiphysics coupling</li> <li>Revolutionary method for plasma simulations</li> <li>New shape memory alloy material models, 3D solid-beam connections</li> <li>Faster rendering in simulation Apps, new publishing and administration features</li> </ul> <p>Enjoy!<br /> Valerio</p> Fri, 15 Dec 2017 13:40:46 +0000 4.2017-12-15 13:40:46.162022 Distributed current source <p>Dear all, I am trying to import a current density file and use it as a current source. I couldn't find any place to add distributed current source and also I couldn't find any documentation about it. Can someone please tell me where to do so? Thanks in advance, Gulsah</p> Fri, 15 Dec 2017 13:02:40 +0000 4.2017-12-15 13:02:40.162011 Inducing a "diffusion"-like blur to smoothen sharp data peaks <p>Hi there, I'm currently working on a 2D-axisym. model of a droplet on a wall. The droplet consists of a certain species which starts building depositions on the wall, where the concentration of that species on the droplet-wall-interface is >50%.</p> <p>The problem is that the meshpoints on the wall, where the species concentration is >50% are not very well distributed (only 1 / 2 meshpoints do have concentration >50%) and so there are really high, rather "unphysical" peaks of the concentration. So depositions are also build badly distributed.</p> <p>At the moment to solve this issue I've induced a normal distribution (gaussian) to the concentration to get better distributed depositions. But this step of course is also "unphysical".</p> <h3>So my question is:</h3> <p>Is there a possibility to blur the >50% concentration peaks, so that it "diffuses" to the adjacent meshpoints as well? (As it would do when I solved a diffusion-equation for the problem itself) The blur should only affect the deposition building, but not backfire on the concentration itself.</p> <p>Thank you very much in advance! I searched in the COMSOL community for approaches but I couldn't find any.</p> Fri, 15 Dec 2017 12:35:29 +0000 4.2017-12-15 12:35:29.161991 How to couple two components comp1 and comp2 with the time-dependent coupling in COMSOL ? <p>I need to couple two components comp1 and comp2 in logiciel Comsol. The purpose is to use the final time-step Temperature T1 of comp1 as the initial value T0 of comp2. What I did, I assigned T0 = comp1.genext1(comp1.T1) for the target domain. Is there anyone encountered the similar problem. I found some cases in the model library of logiciel COMSOL. However, they all solve stationary issues. How to deal with the time-dependent coupling?</p> Fri, 15 Dec 2017 10:32:08 +0000 4.2017-12-15 10:32:08.161981 Shaped metal deposition <p>hi,</p> <p>i want a simulate a shaped metal deposition which a shaped layer of metal is laid on a base substrate succesively for 6 layers ..can someone plzz help me out??</p> Fri, 15 Dec 2017 09:10:25 +0000 4.2017-12-15 09:10:25.161971 Updating Mesh for Topology Optimization <p>Hello!</p> <p>I am trying to implement an algorithm using LiveLink for MATLAB in order to perform a topology optimization. In short, I want to "disable" elements that verify a certain condition (ex: Von Mises stress &lt; XXX MPa). Is there any way to, in the mesh generation process, disable a specific element? If so, how do I update the model in order to use the new mesh?</p> <p>Thank you in advance!</p> Thu, 14 Dec 2017 18:29:38 +0000 4.2017-12-14 18:29:38.161911 Thrust Measurement <p>Hi there,</p> <p>Anyone have idea on how do I go about in measuring Thrust vs Voltage under fluid flow?</p> <p>Thanks alot in advance.</p> Thu, 14 Dec 2017 05:12:57 +0000 4.2017-12-14 05:12:57.161883 Can CFD model be used to mix two miscible fluids? <p>Hi I just wanted to know if it's possible to mix two miscible fluids like Ethanol and Water with a high velocity injection? I just don't seem to be able to figure it out since I have been getting convergence and singularity errors all the time.<br /> Another question I have is if it is possible to mix the same two fluids like Ethanol and Water with orbital motion at a high speed. I seem to get the same errors. I have uploaded the two files. Pls any help is appreciated.<br /> Thanks</p> Thu, 14 Dec 2017 01:37:58 +0000 4.2017-12-14 01:37:58.161882 How to upload pictures? <p>I want to upload some picture in my topic. When I pressed the icon for inserting pictures the following strings pops up. Does it mean that we need to first upload the pictures on some website and then add a link to it? "i""[]""(http://)"</p> Wed, 13 Dec 2017 17:04:09 +0000 4.2017-12-13 17:04:09.161881 Regarding to the effective elastic property of a double-layered perforation <p>Attached please find ther doc describing the modeling problem and the two reference papers. Thanks!</p> Wed, 13 Dec 2017 16:46:53 +0000 4.2017-12-13 16:46:53.161871 Thermal Contact Resistance in 3D modeling <p>Hi,</p> <p>I'm trying to define thermal contact resistance in a 3D model. One way to do it is to define surface roughness and contact pressure. Are there any other way to force the thermal contact resistance to a value? I'm trying to do reduce the thermal conductivity of one side to increase the thermal contact resistance which is not ideal.</p> <p>Bests, Aria</p> Wed, 13 Dec 2017 16:26:27 +0000 4.2017-12-13 16:26:27.161862 Boundaries and domains <p>I am new to COMSOL and trying to model a waveguide for MW application. For simiplicty, I used a rectangluar box as surface for the waveguide, however, I only can select it as boundries is there a way to define as domin. Also for the boundaries is there a way to select the interior face or for materials or both will be the same becuase I only see the outter surface is being highlighted ?</p> Wed, 13 Dec 2017 12:42:36 +0000 4.2017-12-13 12:42:36.161841 Extracting matrix with multiple components in MATLAB <p>Hi,</p> <p>I've built a model with 2 components. Each of them consists of the same geometry with different meshes.</p> <p>And I ran it with one study consisting of two steps. Step1 solves with the comp1 and, at step2 comp2 is solved with an information from step1.</p> <p>What I supposed to do is extracting matrices for each mesh of both components. So I've wrote this script: "mat = mphmatrix(model, ...)" And I got a 1042 by 1042 matrix, but... wait There are only 242 DOF for comp1 and 18 for comp2?!? I'm working on 2D solid mechanics, so, interested in u and v of each components. And I know there are more, such as displacement in z direction or something... Though it is, the total number of DOF is only 687. (Another way to see if it's right is to check the length of "U = model.sol('sol1').getU" or "U = mphgetu(model).")</p> <p>Definitely, I can find the DOF numbers for each variables for comp1 and 2 by using "mphxmeshinfo." But the extracted system matrix does not match at all!!</p> <p>can anyone help?</p> Wed, 13 Dec 2017 12:14:52 +0000 4.2017-12-13 12:14:52.161833 Matlab with COMSOL - unable to link <p>Hello, I have just started to learn COMSOL and I would like to integrate MATLAB into my workflow as I have lots of experience with it. I followed my documentation for Livelink, COMSOL version 5.3m MATLAB 2017a. The following are the commands I used in a sequence: mphlaunch model=ModelUtil.create('Model'); model.param.set('L','9[cm]','Length'); geom1=model.geom.create('geom1',3); wp1=geom1.feature.create('wp1','WorkPlane'); wp1.set('quickplane','xz'); r1=wp1.geom.feature.create('r1','Rectangle'); r1.set('size',{'L','0.1'});</p> <p>at This point I expected to see some update in the COMSOL interface, but alas there is none!</p> <p>I tried this twice, with a complete restart of the system in between. Not sure exactly what is going wrong, so please give me some pointers!</p> <p>Thanks</p> <p>Yameng</p> Wed, 13 Dec 2017 10:14:53 +0000 4.2017-12-13 10:14:53.161831