Latest Discussions - COMSOL Forums Most recent forum discussions Fri, 19 Oct 2018 11:06:36 +0000 COMSOL Forum: Latest Discussions Electromechanics simulation <p>Hi</p> <p>I am working on electromechanical simulation of fixed-fixed beam. Shall I use same moving mesh method as per application available for electrostatically actuated cantilever to fixed-fixed beam?</p> <p>In this module when parametric sweep of pull-in voltage for 2 micro meter deflection is applied, actual deflection shows 0.67 micro meter. Why 2 micrometer deflection not shown for applied bias? Is it correct way of interpreting pull-in voltage?</p> <p>pls suggest...</p> Fri, 19 Oct 2018 11:06:36 +0000 4.2018-10-19 11:06:36.210561 How to couple my MATLAB solver to the COMSOL's Segreated-step solver <p>Suppose I have a multiphysics coupling model, heat transfer and electrical field. It is a time-dependent problem. At each time step, I suppose to use the segreated scheme to deal with the coupling between Heat transfer and Electric field. I build the heat transfer model in comsol with comsol's heat transfer (HT) interface. But, at each time step, I want to solve the electric field using my own MATLAB code. Under the node "Time-Dependent Solver-->Segreated", I add a "segreated step" to solve HT problem. The question is: how to couple my MATLAB code for the electric field to the "segreated step" for solving HT problem?</p> <p>Thanks in advance!</p> Fri, 19 Oct 2018 03:38:42 +0000 4.2018-10-19 03:38:42.210533 Squeezing tube peristaltic pump: sliding roller along lenght of the tube <p>Good evening, I'm trying to simulate peristaltic pump working with fluid structure interaction module in 2D geometry. I have obtained the first deformation promoting the squeeze of the tube with prescribed displacement of the domain as boindary condition along x. Now with this solution I would like to promote deformation with prescribed displacement along y and so along the lenght of the tube but I have a strange deformation of the tube. It's like there is no sliding of the roller on the tube. geometry is shown in following. How it is possible to solve this problem? Thank you in advance for your support. I attached a gif animation that shows the problem without the first deformation but it is the same with that.</p> Thu, 18 Oct 2018 19:02:42 +0000 4.2018-10-18 19:02:42.210521 how to remove planes interactions if I made many extrudes above each other . <p>how to remove planes interactions if I made many extrudes above each other even I make the whole model one unit but still intersections planes between diffrent extrudes surfaces ?</p> Thu, 18 Oct 2018 17:43:20 +0000 4.2018-10-18 17:43:20.210511 Moving the Position of Upper Boundary in a 2D Model <p>Hi everyone,</p> <p>I am trying to create a 2D heat transfer model in which the position of my upper boundary changes with time. Let's say from t = 0 to 300 days , the upper boundary lies at y = 5 m. After 300 days, the upper boundary shifts from y = 5 m to y = 5 + 5m as another layer is placed on top of the existing layer. So, from t = 300 to 600 days, it stays at y = 10 m and so on, for other consecutive layers which are placed on top. Therefore, my upper boundary is moving up along the y axis with time. One way to do such an analysis is to do divide the placement of different layers into separate studies and conduct an analysis such that the results of the previous simulation are carried forward to the next and so on so forth. But if I want to do the entire analysis in a single study, what are my options?</p> <p>Thanks, Simran</p> Thu, 18 Oct 2018 16:11:21 +0000 4.2018-10-18 16:11:21.210502 Understanding the log file generated by 4 nodes with cluster computing <p>Hi all,</p> <p>I am trying to read the log file generated while using cluster computing on 4 nodes. I am using distributed memory approch. Moreover, the domain decomposition solver has been used in the study. The log file suggests that job is "Runing in distributed mode using 4 nodes.". I have attached my log file here. However, the log is provided only for 3 nodes. Also, the log for all nodes look exactly the same. Is every node performing the same job?</p> <p>Please let me know if you find out the mistake that I am making.</p> <p>Thank you.</p> <p>Regards, Prajakta Nakate, Graduate student, Delft University of Technology, Netherlands</p> Thu, 18 Oct 2018 16:02:53 +0000 4.2018-10-18 16:02:53.210501 EM harvester with moving coil. <p>Hello,</p> <p>I'am newby to COMSOL. i'm working on an EM harvester with a printed coil on a vibratng beam and a permanent magnet. I have some technical questions for the design.</p> <p>1- is it possible to design a printed coil ? i was able to draw a 3D geometry with a liitle high as if it was printed and used numeric coil but it needs too much memory.</p> <p>2- to use a permanent magnet i need all the structure to be housed in an air environement for boundaries condition. when i have a vibrating beam, is it possible to use a moving mesh or do i need to use fluid-structure-interaction module ? in this case must i keep the solid mechanics module or only fsi ?</p> <p>Thanks</p> Thu, 18 Oct 2018 12:16:58 +0000 4.2018-10-18 12:16:58.210491 WBK Tunneling Model and Thin Insulator Gate <p>Hello,</p> <p>I am using COMSOL 5.4 with the semiconductor module and I'm trying to simulate a structure that is similar to this one <a href=""></a> , but with a Thin Insulator Gate on top of it (Boundary 6 in the model). I applied a voltage of 1 V on this gate.</p> <p>My problem is that as soon as I add this gate, the simulation doesn't work anymore and gives me the following error:</p> <p>Undefined value found. - Detail: Undefined value found in the equation residual vector. There are 23 degrees of freedom giving NaN/Inf in the vector for the variable comp1.Efn_. at coordinates: (4.08452e-08,0), (4.04603e-08,1.18994e-08), (4.2633e-08,2.95524e-09), (4.44208e-08,5.91047e-09), (4.63485e-08,2.95524e-09), ... There are 23 degrees of freedom giving NaN/Inf in the vector for the variable comp1.Efp_. at coordinates: (4.08452e-08,0), (4.04603e-08,1.18994e-08), (4.2633e-08,2.95524e-09), (4.44208e-08,5.91047e-09), (4.63485e-08,2.95524e-09), ... There are 23 degrees of freedom giving NaN/Inf in the vector for the variable comp1.V. at coordinates: (4.08452e-08,0), (4.04603e-08,1.18994e-08), (4.2633e-08,2.95524e-09), (4.44208e-08,5.91047e-09), (4.63485e-08,2.95524e-09), ...</p> <p>Please help me, I can't figure out what I did wrong!</p> <p>Cheers, Simon</p> Thu, 18 Oct 2018 11:32:12 +0000 4.2018-10-18 11:32:12.210481 Discrete phase model <p>I want to simulate spray cooling which needs thermal interaction between two the flow of discrete phase and continuous phase flow, and also vaporization of discrete phase, I was wondering if it is doable with COMSOL or not?</p> <p>P.S: FLUENT has the capability of simulating discrete phase flow with interaction between thermal and fluid flow madule! But I couldn't find anything in COMSOL!</p> <p>Any feedback will be appreciated.</p> <p>Thanks.</p> Thu, 18 Oct 2018 08:31:59 +0000 4.2018-10-18 08:31:59.210472 How to solve the error in using multiple solution number for data evaluation? <p>Dear all,</p> <p>I tested the example for using mapmatrix in the LiveLink™ for MATLAB® User’s Guide. The matlab code is listed as follows</p> <p><em>model = mphopen('modeltutorialllmatlab'); std ='std1'); time = std.feature.create('time', 'Transient'); time.set('tlist', 'range(0,1,25)');; str = mphmatrix(model,'sol1','out',{'K'},'initmethod','sol','initsol','sol1','solnum',15);</em></p> <p>However, if another solution number is specified, an error will be found.</p> <p>str = mphmatrix(model,'sol1','out',{'K'},'initmethod','sol','initsol','sol1','solnum',11);</p> <p>The error imformation is:</p> <p>Error using mphmatrix Cannot set the property 'solnum' to '11'. Allowed values are: 'auto' 'first' 'last' 'interp' 'manual' '1'</p> <p>How to solve this problem? Thanks.</p> Thu, 18 Oct 2018 01:17:53 +0000 4.2018-10-18 01:17:53.210463 impedance measurement in comsol <p>The below post is related to an <a href="//">archived discussion</a> hello..i have made a microfluidic channel in comsol and want to simulate it and measure the impedance with varying amount of it possible in comsol multiphysics 5.2??</p> Wed, 17 Oct 2018 20:04:16 +0000 4.2018-10-17 20:04:16.210441 How can I determine the heat dissipated in a pin fin model created in COMSOL? <p>So I have a problem that looks like the attached picture.</p> <p>I have modeled the problem in COMSOL and obtained isothermal contours as well as a 3D figure displaying the temperature distribution.</p> <p>But how can I calculate the value of the heat dissipated (i.e. q). Any help is appreciated.</p> Wed, 17 Oct 2018 16:02:15 +0000 4.2018-10-17 16:02:15.210432 storing one variable only <p>Dear all, I have a model with 3 variables of T, P and v. what is the best way of only storing T in solutions to save on size? do I need to do it via batch mode? I dont have COMSOL with matlab so probably matlab is not an option.</p> <p>many thanks, Regards, Asal</p> Wed, 17 Oct 2018 10:12:59 +0000 4.2018-10-17 10:12:59.210421 Extract point values from line and reference them in expressions <p>Hi, it feels like a very simple question that I have, but I cannot figure out how to do it.</p> <p>Let's say I have a plate that is heated on one side, so a temperature profile is developing across the plate. I now how to plot the temperature profile along a line (e.g. the center line) via a "Cut Line" data set. But how can I print the values of temperature that are shown in the plot to a table? I'm thinking of something like "print the temperature for 100 evenly distributes points along the cut line". There has to be a more sophisticated approach than to create 100 cut points and 100 derived values by hand.</p> <p>Also, i want to call those table cells in expressions, subsequently. For a start, I'd be happy to know how to reference a table cell or a derived value in an expression in general.</p> <p>Thanks in advance! -Jan</p> Wed, 17 Oct 2018 08:38:45 +0000 4.2018-10-17 08:38:45.210401 Electric field parallel to electron current density <p>Hi,</p> <p>I am trying to make an electro-thermal GaN-HEMT device model on Comsol using the Semiconductor and Heat Transfer modules.</p> <p>I have tried to incorporate the Caughey-Thomas mobility model to evaluate field dependent mobility at high drain voltages (Vds). However, I could not succeed in predicting the experimental data accurately. Therefore, I have decided to manually input a field-dependent mobility expression (known as Canali model). For that purpose, I need to make "Electric field parallel to electron current density" an input to my user defined expression.</p> <p>When I activate the Caughey-Thomas mobility model (mmct), this particular parameter is automatically calculated by Comsol. Its variable name is: comp1.semi.smm1.mmct1.Epn. In the Caughey-Thomas Mobility tutorial, it is possible to plot the variable as it is readily evaluated.</p> <p>However, when mmct is not used/activated, this parameter is not recognized by Comsol.</p> <p>How can I evaluate this parameter in terms of the existing variables in the Semiconductor module? I have been focusing on writing a summation of dot products of J and unit vectors and taking the squareroot of that expression. However, I could not yet find the correct parameters.</p> <p>I appreciate if you could share your experience or ideas on this.</p> <p>Thanks, Bugra</p> Wed, 17 Oct 2018 08:28:28 +0000 4.2018-10-17 08:28:28.210393 Questions regarding defining a geometry with variables <p>I recently have met a question when using COMSOL to simulate the deformation of an elastomer membrane. The elastomer membrane I am using is not a flat membrane but has a inhomogeneous surface profile. I want to optimize the inhomogeneous profile (which is a polynominal equation with unknown coefficient variables) to reach the desired surface profile upon a certain deformation. So I may need to define a geometry with coefficient variables in building the model. I am not sure whether I can do this in COMSOL, but I found a sample file (optimization of a beam) and see that it uses deformed geometry with prescribed mesh deformation. I have no idea how it defines that prescribed deformation. Can someone give any suggestions on how to solve my problem? Very appreciated.</p> Wed, 17 Oct 2018 04:55:31 +0000 4.2018-10-17 04:55:31.210392 CO2 injection <p>Hi everyone,</p> <p>I am currently doing my final year research project on ‘A modelling study of petro physical properties changing during CO2 injection into carbonate reservoirs’. In my first part of the research I am validating experimental result. I having problem finding modelling equation related to this topic.Please do assist me in this matter if you are able to. Your assistance would be greatly appreciated. Thank you</p> Wed, 17 Oct 2018 03:58:12 +0000 4.2018-10-17 03:58:12.210381 Ball-screw <p>Hi everyone,</p> <p>Is anybody here know how can i model a ball-screw with nut?</p> <p>Thanks</p> Wed, 17 Oct 2018 01:37:57 +0000 4.2018-10-17 01:37:57.210372 UvsI Coil and Required Resistor in Parallel in Electric Circuit Node <p>Hello COMSOL team and users:</p> <p>I am using an AC current source to excite a UvsI coil in the Electric Circuit node and I see that I have to connect a resistor in parallel with the current source and with the UvsI coil and I would like to know what is the purpose of this resistor? In the mmf node I specify my multi-turn coil's settings and set the values for the coil wire's cross section area, coil number of turns, the coil wire conductivity, and my the length of my 2D model is specified, so I figured that the resistance of my coil is already built in my model so what is the resistor in parallel with my AC current source and UvsI coil doing and representing?</p> <p>Thank you!</p> <p>David Giglio</p> Wed, 17 Oct 2018 00:29:18 +0000 4.2018-10-17 00:29:18.210362 EHD Cone Jet <p>The below post is related to an <a href="//">archived discussion</a></p> <hr /> <p>Hi,</p> <p>Did you get any reasonable results for the EHD-cone jet simulation. If so, can you share the procedure?</p> <p>Regards, Tawsif</p> Wed, 17 Oct 2018 00:28:09 +0000 4.2018-10-17 00:28:09.210352