Latest Discussions - COMSOL Forums Most recent forum discussions Fri, 24 May 2019 18:31:17 +0000 COMSOL Forum: Latest Discussions Particle Agglomeration in a flow <p>Hi, I am new to this software. I want to simulate particle agglomeration in a flow. Which module of COMSOL should I use for my analysis? Thanks</p> Fri, 24 May 2019 18:31:17 +0000 4.2019-05-24 18:31:17.236082 Mesh error (Intersecting edges) <p>Recently i have been working on muffler design so during pressure acoustics analysis this mesh error occured. I read previous thread about interseting edges error and tried to solve by that but it didnt resolve.[][] above are drive links for the screenshot of the error occured during meshing and model with mesh errors.(Becz of size issue i uploaded them on drive.Sorry for inconvenience) Anyone help !!</p> Fri, 24 May 2019 15:11:13 +0000 4.2019-05-24 15:11:13.236081 use result from a study to calculate a variable and use this as input for another physics module <p>Dear COMSOL users,</p> <p>I'm a student and attempting a RF system in combination with a DC magnetic field.</p> <p>First, I simulated the magnetic field of my model in COMSOL. Now in my next study (frequency domain, RF physics) I need to use the previously calculated magnetic field to calculate the permittivity matrix: I need to calculate the local cyclotronfrequency, which depends on the local magnetic field (mfnc.Bz).</p> <p>Now my question is: How do I use my result from the local magnetic field study to calculate all local cyclotron frequency and use this as input for my permittivity matrix in the RF module?</p> <p>Thanks in advance.</p> <p>Kind regards, Rik</p> Fri, 24 May 2019 09:51:03 +0000 4.2019-05-24 09:51:03.236071 Error message - error log attached <p>Hi, I'm unsure as to how to solve this error message. The system has plenty of storage, so unsure as how to solve the problem. The programme keeps crashing, and files still says they're locked despite not being open.</p> Fri, 24 May 2019 09:45:06 +0000 4.2019-05-24 09:45:06.236061 How to export intensity (gop.I) data in ray optics ? <p>Hi, I'm simulating a model quite similar to linear wave retarder in the application libraries in the ray optics. I can successfully trace the ray, but somehow I cannot export the intensity (gop.I) data. It just noticed me an error like the screenshot below.</p> <p>So how can i fix this error and export the data i want ? thx</p> Fri, 24 May 2019 04:52:14 +0000 4.2019-05-24 04:52:14.236053 Useful power/absorbed power <p>Hi, I am using COMSOL version 4.2.a. I work on induction heat treatement. I'd like to know how to find the general power absorbed by the specimen/piece of work. Thank you!</p> Thu, 23 May 2019 23:27:46 +0000 4.2019-05-23 23:27:46.236012 Coupling RF Module with Semiconductor Module <p>Hello all,</p> <p>I am using COMSOL 4.4. I am simulating the effect of an incident electric field wave over a semiconductor substrate. I know the semiconductor and the RF module are not coupled through the multiphysics node, but I wanted to simulate the same model with the RF module and use the results in the semiconductor module. I was wondering if the General Extrusion was a way to go in this application by coupling two components (one using the RF module and one using the semiconductor module). However, I am not sure if this is a suitable application for the General Extrusion.</p> <p>Jose Santos.</p> Thu, 23 May 2019 22:03:23 +0000 4.2019-05-23 22:03:23.236002 Resonance Frequency Shift due to Magnetic Force Gradient <p>Hello,</p> <p>I am finding some problems in simulating a Magnetic Force Microscope (MFM). I have a magnetized cantilever which interacts with a non uniform magnetic field; the gradient of the magnetic field acts on the magnetized cantilever with a force, the gradient of the force affects the stiffness of the beam changing its resonance frequency.</p> <p>The geometry is very simple: a permanent magnet, a magnetized cantilever and air surrounding everything. I am using the Magnetic Field No Currents interface in order to find the field gradient produced by a permanent magnet and in order to evaluate the magnetic force on the magnetized beam. The magnetic force is used as a Body Load of the cantilever in the Solid Mechanic interface. I have also applied a Load to the cantilever to simulate the harmonic forcing term. Basically, the magnetic force gradient seen by the oscillating cantilever changes its stifness; therefore, I need the Body Load (due to the magnetic force) to change with the displacement of the beam.</p> <p>I have initially set up this study:</p> <p>STEP 1 - STATIONARY: to evaluate the the field and the force; STEP 2 - FREQUENCY DOMAIN: to calculate the effect of the force on the frequency spectrum of the cantilever oscillations. Since I should observe a shift in the resonance frequency of the cantilever, this step does a sweep on the frequencies.</p> <p>In the result I plot the RMS displacement of the beam with respect to the frequency sweep.</p> <p>Clearly this approach was a failure because, as I have understood, STEP 1 evaluates the force without considering the displacement of the beam, then in STEP 2 the applied magnetic force is always the same, i.e. the cantilever does not see any gradient of the force.</p> <p>My question is: how can I set up the simulation so that I have a Body Load due to the magnetic force which is updated with the displacement of the beam and, in parallel, a Body Load to implement the harmonic forcing term which drives the oscillation of the beam at different frequencies?</p> Thu, 23 May 2019 16:04:24 +0000 4.2019-05-23 16:04:24.236001 Why does Comsol plots probe results in "Iteration Number" instead of time step? <p>Hi all,</p> <p>I have a time dependeent study where I am plotting a probe in time. However, sometimes the plot's X axis is in ITERATION NUMBER, isntead of TIME. Also, the probe table experiences the same thing.</p> <p>I am using PARDISO as the Direct Solver.</p> <p>Can you help me revert the ploting in time?</p> <p>Kidn regards, Ion</p> Thu, 23 May 2019 15:33:10 +0000 4.2019-05-23 15:33:10.235991 how can I find variables in prefixed equations <p>Hello.</p> <p>I want to change prefixed equations and use variables to other equations. I have seen some of the articles say not be able to change it. it also says if I really wanted to change it, change equation's variables in equation view. So I tried, but I couldn't find what variables are what. for example, variable "F" in flow fluid model and variable "V" in electric current model. there are tons of variables in equation views and declared variables are not as same as "F" or "V" like that. Please let me know how to find this kind of variables in equation view. looking forward to your answers.</p> <p>Thx in advance.</p> Thu, 23 May 2019 13:22:02 +0000 4.2019-05-23 13:22:02.235983 Query regarding the Sweep function in COMSOL <p>Sir/Madam, I'm just starting to learn about the software. I was trying to design a parametric solid 3D curve. I used parametric equations to draw the actual curve(which can be seen in picture 1 which is not sweeped; represented by a line) then use a work plane and a square face to perform the 'Sweep' function inbuilt in COMSOL. I have attached two pictures out of which 'parametric geometries_2' is obtained by checking the box 'Reverse Direction' (an option under the spine curve) and the picture 'parametric geometries_1' contains a distorted figure with warning. Did the warning happen due to the choice of 'Faces to Sweep' ? And why the use of reverse direction solved the problem?</p> <p>Thank You in advance.</p> Thu, 23 May 2019 09:22:59 +0000 4.2019-05-23 09:22:59.235951 3D RC circuit <p>We are interested in modeling a 3D RC electrical circuit similar to the one presented in the attached figure (from; maybe something in the order of 10 x 10 x 10 RCs.</p> <p>Each node (i,j,k) represents an RC circuit. All adjacent nodes in the x, y and z directions are connected electrically to each other. An RC system away from the edges or the faces shall be connected to its neighbors through 6 connections. Facial RC have 5 connections, Segment connections have 4 connections and edge RC have 3 connections. This topology can be viewed as a 3D expansion of 2D series-parallel association of RC circuits.</p> <p>We have tried many combinations using EC, Cir and Electrostatics modules in Comsol 5.0, however, we are not funding any difference from ideal capacitors.</p> Thu, 23 May 2019 05:37:03 +0000 4.2019-05-23 05:37:03.235931 How can I plot the deformed boundary? <p>Hello,</p> <p>I kinda have a simple question. Maybe an interface issue. (I'm using 4.3b ver.) Whenever I perform a mechanical (mostly about deformation) simulation, the defromed boundary of the geometry is not clear. Especially when a composite is imbeded into another composite, it is hard to see the interface. Is there any method to keep the boundary plot not only for the initial state but also after the deformation??</p> <p>Thank you.</p> Thu, 23 May 2019 03:47:13 +0000 4.2019-05-23 03:47:13.235923 Cap faces feature without CAD <p>The below post is related to an <a href="//">archived discussion</a></p> <hr /> <p>[start here] I've made parts directly in COMSOL (not CAD) but I need to identify the volume inside an enclosed space a domain. It appears that the "Cap faces" feature would accomplish this but since I built it directly in COMSOL it doesn't appear available. How can I accomplish creating this domain without CAD?</p> <p>Thank you Osa</p> Wed, 22 May 2019 22:39:07 +0000 4.2019-05-22 22:39:07.235922 Calculate R squared in Comsol <p>Hi,</p> <p>If I have the measured and the computed curves in the same plot (see attached file), how can I calculate R^2 in Comsol? Thanks.</p> <p>Best regards, Zhidong</p> Wed, 22 May 2019 20:57:26 +0000 4.2019-05-22 20:57:26.235921 Pulsed RF ICP <p>Hello. I am wondering about COMSOL Multiphysics can solve Pulsed ICP problem. I know there is Pulsed CCP example in application library. But, I couldn't find any information about Pulsed ICP. Who has some ideas about Pulsed ICP ?</p> Wed, 22 May 2019 14:50:34 +0000 4.2019-05-22 14:50:34.235891 Contraction of a cell on a gel <p>Hi!</p> <p>I created a 2D axisymmetrical model to examine the deformation of a (stem) cell on a (polyacrylamide) gel after contraction.</p> <p>In order to do so, I should define a prestress over a part of the cell (the cytoplasm). However, since the cytoplasm is made up of a hyperelastic Neo-Hookean material, it is not possible to define a prestress -I used "Initial stress"- in a way it is possible for a linear elastic material (also, when applying linear elastic materials, the model does only converge for a positive prestress, so radially outward, instead of radially inward as in a contraction).</p> <p>What could cause the model to not converge for a radially inward prestress?</p> <p>And is there any other way to define a prestress for a Hyperelastic Material, to imitate contraction?</p> <p>Thank you in advance!</p> Wed, 22 May 2019 14:25:05 +0000 4.2019-05-22 14:25:05.235881 Different S11 values for the same frequency <p>Hi all,</p> <p>I simulate a pin-fed patch antenna at 3.5 GHz. When I set Frequency to <strong>3.5&#91;GHz]</strong>, The S11 value is -28.474 dB. But if I enter <strong>range(3.3&#91;GHz],0.1&#91;GHz],3.7&#91;GHz])</strong> phrase into the frequencies text box. I get the following results, which is -23.569 dB at 3.5 GHz.</p> <pre><code>3.3000 -3.9214 3.4000 -8.0754 3.5000 -23.569 3.6000 -10.408 3.7000 -5.3938 </code></pre> <p>Why are two S11 values different at the same frequency (3.5 GHz) even I don't change anything else?</p> <p>Thanks.</p> Wed, 22 May 2019 09:50:48 +0000 4.2019-05-22 09:50:48.235751 Computing transmittance and reflectance of glass for different wavelengths <p>Hello everyone, I'm a beginner user of COMSOL, and I'm stuck on a little problem. I'm using the "<em>Electromagnetic Waves, Frequency Domain</em>" interface to model an incident wave on a smooth surface, in order to calculate the transmittance and the reflectance of an <strong>N-BK7 glass</strong>. I followed the instructions of the "Plasmonic Wire Grating" tutorial, but my purpose is to compute the aforementioned quantities for different wavelengths. For this reason, I don't want the <strong>refractive index</strong> of my material to be constant: it must vary with the wavelength. Therefore, I created an analytic expression for the real part of the refractive index (which I named "<em>n_NBK7</em>"), as well as an interpolation function for the imaginary one (which I named "<em>k_NBK7</em>"). As concerns the air above my sample, I did the same thing with the real component (naming the function "<em>n_air</em>"), while I set the extinction coefficient to 0. I took all these relations from this <a href="">link</a>. My questions are the following: <strong>1)</strong> In the study settings, I can make the wavelength sweep within the range of interest: how can I tell the software that the wavelength must be the <em>input variable</em> for my interpolation/analytic functions (i.e.: <em>n_NBK7</em>, <em>k_NBK7</em> and <em>n_air</em>)? <strong>2)</strong> In the <strong>Port boundary conditions</strong>, the ewfd interface needs to know the refractive index of the air (Port 1) and of the N-BK7 glass (Port 2), in order to compute the diffraction orders. Provided that I don't have a constant value for them (they depend on the wavelength), I want to <em>call</em> here the functions <em>n_air</em> and <em>n_NBK7</em>, respectively: is this feasible? Which is the correct syntax to do it? <strong>3)</strong> If I understood correctly, the software computes the number of <strong>diffraction orders</strong> as a function of three parameters: the angle of incidence, the refractive index of the material (in my case, the N-BK7 glass) and the wavelength itself. This means that, during the sweep study that I'd like to run, the software should be able to adjust the number of Ports in order to compute the different diffraction orders: I mean, without requiring me to click manually on the "<em>Compute Diffraction Orders</em>" button (provided that it is disabled while the solver is running!). Am I reasoning right? Is there a way to automate the computation of diffraction orders?</p> <p>Any help is very much appreciated! Thank you</p> <p><strong>PS</strong>: I'm going to attach the .mph file of my model, it could be useful to better understand my problem</p> Wed, 22 May 2019 08:50:15 +0000 4.2019-05-22 08:50:15.235741 How to Define Poisson Boltzmann equation in Cylindrical coordinates <p>Hi, I want to define poisson Boltzmann equation in cylindrical coordinates in comsol, can some one please help me. The equation is dsqrU/dRsqr + (1/R)*dU/dR + dsqr U/dZsqr = (kh)^2 * sinh(U)</p> Tue, 21 May 2019 16:18:50 +0000 4.2019-05-21 16:18:50.235721