Latest Discussions - COMSOL Forums Most recent forum discussions Thu, 17 Oct 2019 03:19:03 +0000 COMSOL Forum: Latest Discussions Export boundary condition with boundary node in mesh <p>I have found the mesh result could be exported from the export options, but the output does not show the boundary condition with boundary node, it only contains the element numbers and corresponding node numbers in elements. Is there any way to export the boundary condition with boundary nodes?Thanks.</p> Thu, 17 Oct 2019 03:19:03 +0000 4.2019-10-17 03:19:03.244533 Voltage/current Source with harmonics <p>Hello,</p> <p>I wan t to apply a voltage or current source which is is square wave and it not sinusoidal purely. How can I define it as bellow: V(t)=V_ac<em>sin(2</em>pi<em>f</em>t)+ (V_ac/3)sin(2<em>pi</em>3f<em>t)+ (V_ac/5)sin(2</em>pi<em>5f</em>t)</p> <p>Besides, how I can calculate losses in E-core transformer?</p> Thu, 17 Oct 2019 00:36:57 +0000 4.2019-10-17 00:36:57.244532 Checking for modules in COMSOL 5.2a <p>Hi,</p> <p>I am very new in unsing COMSOL 5.2a. Could you please inform how to check for existing modules in a COMSOL 5.2a inside a Cluster. Thank you.</p> <p>Regards, Koushik</p> Wed, 16 Oct 2019 17:28:26 +0000 4.2019-10-16 17:28:26.244522 Parameter Values of Damage Integral Analysis in Bio Heat Transfer <p>In Damage Integral Ananlysis of Bio Heat Transfer Module the by default values of the parameters are as fowllowing as A=7.39E39 and dE=2.577E5. I am very much interested to know is this value used in litereature by other researchers, as I need a reference for research article.</p> Wed, 16 Oct 2019 14:54:54 +0000 4.2019-10-16 14:54:54.244521 Multiphysics coupling <p>hello everyone,</p> <p>i'm tyring to model an isothermal flow of fluid through a pipe. the heated pipe will heat the fluid when it move along the pipe. my question is, in the multiphysics section, should i choose fluid coupling and temperature coupling? can anyone explain what temperature and fluid coupling are for?</p> <p>thank you</p> Wed, 16 Oct 2019 14:47:14 +0000 4.2019-10-16 14:47:14.244511 How to construct particle tracing and analyze particle motion in the coupling field of electrostatic field and fluid <p>In the coupling field formed by electrostatic field and fluid, only fluid flowing particle and charged particle physical field can be simulated separately at present. How to find a particle tracking scheme in the coupling field</p> Wed, 16 Oct 2019 14:38:35 +0000 4.2019-10-16 14:38:35.244503 Failed to find consistent initial values. Attempt to evaluate nonintegral power of negative number. <p>Hello comsol community,</p> <p>I am in the process of coupling Electrostatics, Heat Transfer in Fluids, Transport of Diluted Species to Two Phase Laminar Flow (Phase Field) and having some difficulty. The simulation scenario is a water droplet evaporating on a surface under the electric field between two electrodes. I expect to see deformation of the water droplet and a water concentration profile as it evaporates. However, I was faced with this problem:</p> <p>Failed to find consistent initial values. Attempt to evaluate nonintegral power of negative number. Function: ^ Failed to evaluate variable. Variable:, Defined as: ((1+(((^0.5)<em>(( v/ a)^0.25)))^2)/(((1+( a/ v))^0.5)</em>2.8284271247461902909) Failed to evaluate variable. Variable: moist, Defined as: ((<em></em><em></em> Failed to evaluate variable. Variable:, Defined as: moist Failed to evaluate variable. Variable: effrr, Defined as: Failed to evaluate variable. Variable: T, Defined as: (( eff<em>mod1.Tt)-( effrr</em>mod1.Trr))-( effzz<em>mod1.Tzz) Failed to evaluate expression. Expression: abs( T) Failed to evaluate expression. Expression: nojac(max(((((0.5</em>abs( T))<em> T)</em>sqrt( T))/(( T+((h spatial^2)<em>sqrt(eps)))</em>( eff)))-((( T^2)<em> T)/(eps+(( eff</em> T)/,0)<em>(1.4142135623730951455^(1-shapeorder(mod1.T)))) Failed to evaluate variable. Variable: T Failed to evaluate temporary symbolic derivative variable. Variable:${test8} Failed to evaluate expression. Expression: d((2</em><em>pi</em>r)<em>(dvol spatial),{test8}) Failed to evaluate Jacobian of expression. Expression: (2</em><em>pi</em>r)*(dvol spatial) Last time step is not converged.</p> <p>Why the viscosity may be negative? I don't think this problem is caused by my setting of physics model conditions, so I have tried adjusting the setting of mesh many times, but it dosen't work. Did anyone also encounter this kind of problem? How do you manage it? I hope to learn something from someone.</p> <p>Best regards, Sunny PS: In the picture "model", the semicircle represents the water droplet, and the rest field represents air.</p> Wed, 16 Oct 2019 14:11:29 +0000 4.2019-10-16 14:11:29.244493 nanopore <p>I would like to model nanopore sequencing. I probably have a record that holds a small of about 2nm. This plate separates two chambers. These chambers are connected through the hole. There is an ionic solution in both chambers. There is an eletrode in the two chambers where a voltage is applied, so there is now a current from the negative chamber to the positive chamber. A small particle now passes through it, which blocks a large part of the current and therefore reduces the current being measured. In my model I have already tried something, but I do not know how to proceed.</p> Wed, 16 Oct 2019 13:27:35 +0000 4.2019-10-16 13:27:35.244491 Laminar Flow from a cylinder into a block? <p>If I create only one block and create an inlet on one side and an outlet on the other I am able to simulate a laminar flow through that block. But as soon as I create a cylinder and a block, my flow only flows through the cylinder but no longer into the block. What do I have to change? How do I "open" the cylinder so that the flow can enter into the block? Thanks in advance!</p> Wed, 16 Oct 2019 08:54:50 +0000 4.2019-10-16 08:54:50.244471 Simulation of photonic crystal waveguides to find band structure by COMSOL <p>Hi everybody,</p> <p>I am trying to simulate a photonic waveguide using COMSOL. The geometry is more or less as mentioned in</p> <p>The point is that when I plot the dispersion curves they are all identical!!!!!!!!!!</p> <p>Any idea how to simulate and get the correct band structure?</p> <p>Best, Maryam</p> Wed, 16 Oct 2019 06:57:14 +0000 4.2019-10-16 06:57:14.244461 Rotational mechanics in microwave oven plate <p>Hi all i am trying to rotate plate ina microwave oven to study the process as well as as an exercise for roatational mechanics. while doing so ,i had some issues after importing the cad file. after importing the cad file the identity pair ;source and destination seems to be same and couldnt figure out why comsol kept like that. I know that the air gap needs to have a boundary ( identity pair) in order for body to move but i couldnt figure out why the identity pair remained the same. It will be helpful if someone could let me know to solve this roatational issue.</p> <p>Thanks in advance</p> Wed, 16 Oct 2019 06:27:02 +0000 4.2019-10-16 06:27:02.244451 Buoyancy--simulating contamination of a buildings ventilation air intake by its exhaust air <p>Hi all, I am trying to model a cross section of vertically aligned ventilation ports on the ouside of a building. I am using a 2D domain to represent this outside space where an inlet delivers warm air (a buildings exhaust air), and an outlet pulls air (a buildings fresh air intake) through an insulated wall (the wall of the building). Also in the model I have left two walls as open boundaries (since this space is open to the environment). At some point I would like to parameterize the geometry of the ports to optimize them against contamination (particle tracing?)... but I am not there yet. I am new to COMSOL and could use some help in just figuring out how to set up the file I attached, the error I recieve is a failure to converge. And beyond that, I am not sure how to set the temperature of the inlet, nor how to set up temperature dependent buoyancy. If anyone could give me feedback on the file, or share tips on achieving convergence that would be great--I am new to comsol so no suggestion is too simple! Thank you, R.</p> Wed, 16 Oct 2019 00:28:22 +0000 4.2019-10-16 00:28:22.244442 2D grid of points and curve fitting according to an equation <p>Hello Everyone, I need to fit a curve according to an equation, between the selected points in the 2D grid of points. I am making an app. in comsol where there has to be a grid of points, when I select few points on the grid, It should form a smooth curve fitting in the given equation. I went through the COMSOL programming guide, It says I have to</p> <p>choose the option <strong>Blank Model</strong> in the Model Wizard. To create a new method, go to the <strong>Developer tab</strong> in the <strong>Model Builder</strong> and click the <strong>New Method</strong> button in the ribbon. In the Application Builder, paste the code into the new method.</p> <p>But then, I am not finding the <strong>developer tab</strong> in the model builder!</p> <p>Can someone help me fit a curve according to an equation, between the selected points in the 2D grid of points.</p> <p>Thank You,</p> <p>Regards/Kishor</p> Tue, 15 Oct 2019 20:12:32 +0000 4.2019-10-15 20:12:32.244431 Simulating the Hall Effect <p>I have done a search on this forum and all the threads I have found are from years ago.</p> <p>Is there a simple way to simulate the Hall Effect? It really doesn't seem like it would be that complicated.</p> <p>All of the AC/DC modules seem to <em>partially</em> have what I need. Magnetic Fields (the one with currents) lets me define a static, uniform magnetic field, but no real way to just apply a voltage between two ends of a conductor.</p> <p>I've seen people suggest calculating my own conductivity matrix which takes into account the magnetic effects, but I also read that you can't get a charge pileup reading from it, so it's really pointless.</p> <p>The module Magnetic and Electric fields doesn't have a way to define a static background magnetic field. You can't select between "full field" and "reduced field". I can make a 3D block, voltage across X axis, magnetic field along Z, trying to measure charge pileup in Y-axis. I can set the Y-axis walls to have a magnetic field pointing in Z. This magnetic field influences the current across X. This I have verified. So why can't I just put a static uniform field in the model?</p> <p>Thank you.</p> Tue, 15 Oct 2019 19:28:46 +0000 4.2019-10-15 19:28:46.244422 Residual stress <p>Hello all,</p> <p>I am trying to simulate the behaviour of a structure that goes through thermal stress. The temperature is applied through a time dependent expression and the temperature goes from 293K to 1436K. The study done is time dependent.</p> <p>My structure is a thin film with a gage section made of silicon nitride over a silicon substrate. From the measurements of my devices, I know there is a tensile residual stressof 390MPa on the thin film, however, whenever I use the "initial stress feature" and input a 390MPa as the value when I check the stress on my structures at time 0s I get values of about 2.7E9. There are no other loads applied to the device. The only added boundary condtion comes from constraining the 4 points at the corners of the device. I do notunderstand why my stress levels get bumped so high. I check the initial conditions of my model and it does show the whole film prestressed to 390MPa</p> <p>I have read that the "initial stress" function is kind of confusing as well, so I tried to apply a statitc study that brought my temperature down from 950K to 293K, which generated a stress of 390MPa in my results and then use that result as my input for a second study where i try to do what i mentioned above. However, the same thing happens when I put my simulation at 0s my stress then becomes really low to about 2.7MPa. I have tried several approaches and combinations of options but none of them give me the result I believe I should be getting.</p> <p>I am attaching my .mph file any help is appreciated</p> Tue, 15 Oct 2019 18:48:45 +0000 4.2019-10-15 18:48:45.244412 Spiral flat Coil Simulation <p>Hi there, I am trying to simulate a printed spiral coil and I am using the Coil sub-branch in the magnetic field physics. I would like to see the magnetic flux generated aroud the coil. In the stationary study an error is generated: "Numeric coil Coil 1 not solved for. Solve it in a Coil Geometry Analysis step." Does anyone have any idea? Thank you</p> Tue, 15 Oct 2019 17:45:54 +0000 4.2019-10-15 17:45:54.244402 Flow and heat transfer through packed bed of spheres <p>Hello, The study is time dependent, it's 2D axisymmetric, and the physics used are laminar flow and heat transfer in fluids with heat transfer in solids node added to it. It is not converging and the time-dependent solver (BDF) found in the progress section found just below the graphics window is not showing any progress at all. It says unsymmetric matrix found. It stops the simulation and shows errors. Please help.</p> Tue, 15 Oct 2019 15:10:34 +0000 4.2019-10-15 15:10:34.244401 Extracting results of specified coordinates <p>Hi, I am doing simulation and got a results of 2D plot as shown below. I want to extract the output of specific coordinates for example coordinate of (0,-1100), (1,-1100), ... (45, -400) , etc.. Do you know how to extract this information automatically? I know how to use the evaluation features but it is manual and not automatic.</p> <p>Thank you.</p> Tue, 15 Oct 2019 12:33:37 +0000 4.2019-10-15 12:33:37.244391 Modeling of soil-wall interface (Earth retaining wall) <p>Hi, I have been modelling a diaphragm wall with COMSOL. During the modelling I have faced problems to model the interface soil-wall.</p> <p>It seems that to define an internal boundary condition, one needs to define “pairs”. At the beginning, we tried identity ones, but due to the nature of the problem I decided to take contact pairs instead, to allow new contacts and detachments. For this purpose, I have unsuccessfully tried:</p> <p>1) Prescribed displacement coupled with a general extrusion to map only the normal displacement from one domain (the soil) to the other (the wall) (simulating frictionless with identity pairs)</p> <p>2) Contact boundary condition (modelling with contact pairs)</p> <p>Entually, I want to control the parameters of this interface (say from frictionless to a Mohr–Coulomb failure criteria) similar to what all specialist geomechanics software does.</p> <p>Thank you for your help</p> Tue, 15 Oct 2019 05:46:19 +0000 4.2019-10-15 05:46:19.244382 Is it possible to scale the horizontal and vertical coordinates differently? <p>I am studying some microfluidic flow problem. When I'm doing the analytical work, I can often simplify the problem under the assumption that one spatial coordinate is much smaller than the other one (say H/L&lt;&lt;1 where H is typical vertical scale, and L is typical horizontal scale). Now I want to study the problem without any approximation in COMSOL using the Laminar two-phase flow module , but it is still true that L is quite larger than H. The problem is that the geometry of the problem becomes very wide (width is usually more than 10x the height) while the actual variation rate in the horizontal direction looks quite small compared to the vertical direction, and I have a feeling that I'm using a lot more mesh elements (and time) without actually improving the simulation. Is it possible to scale the two spatial coordinates differently to make the size and typical variation in both dimenions comparable? (say x->x/10um, y->y/0.5um)</p> Mon, 14 Oct 2019 22:17:52 +0000 4.2019-10-14 22:17:52.244362