Latest Discussions - COMSOL Forums Most recent forum discussions Sat, 20 Apr 2019 10:26:47 +0000 COMSOL Forum: Latest Discussions <p>** Momen Trim Keto** On his or her third and fifth meal have a light-Weight Loss Tips carbohydrate like rice or new potatoes. Each protein should be portioned out to four ounces for and also two ounces for adult women. The veggies (especially greens) you may have a little really if you like. ** **</p> Sat, 20 Apr 2019 10:26:47 +0000 4.2019-04-20 10:26:47.232571 Trying to build Trimorph <p>Hello,</p> <p>I am trying to buuild trimorph but in bimorpph it is not giving me any solution. As i am new to comsol so kind of stuck in this situation. All i just to want to calculate the resonance frequency an total displacement. Can anyone help me regarding this? Thanks</p> Sat, 20 Apr 2019 06:39:11 +0000 4.2019-04-20 06:39:11.232562 Creating an Air Geometry inside a structure <p>Hi! I'd like to model air between the comb fingers of my structure. So far I have tried making a sphere, assigning it the property of air, and then placing my structure inside it, and then use the difference operator to subtract the structure from the air. This gives me the geometry of the air, but I no longer have the structure that I wished to model. Is there any way to essentially build a shape for the air placed perfectly between the gaps in my structure, and build the structure itself?</p> <p>See file attached. Thanks!</p> Sat, 20 Apr 2019 02:33:44 +0000 4.2019-04-20 02:33:44.232552 Problems of Integration of magnetic field and electric field computation in time dependent studies <p>Hi there,</p> <p>There are two problems that we can not address when simulations. Could anyone please help check and figure ways out?</p> <p>Question #1: How to integrate electric and magnetic field computation in time dependent studies.</p> <p>Please find a simple winding 2D model attached. We used magnetic and elctric field physic combined with eletric circuit in simulations. The winding is constructed with four serial coils, and excited by a pulse voltage source. At voltage rise font, there are supposed to have ringing on the voltage across each coils, which was already observed from our experimental results. This ringing is induced by the interaction of electric field and magnetic field. However, in simulations, it is apparent that electric field was not solved, not to mention the interaction between two fields.</p> <p>To represent the effect of electric field, we manually added some capacitors between the conductor and the stator. Then we did see the ringing at voltage rise front. In the attached file, capacitors are set to be 0. If you change them to 10 pF ~ 20 pF, you will see the ringing that we are talking about.</p> <p>However, this is just a temporary solution, and we are looking for a full-wave simulation approach. Is there a way to integrate the electric field, magnetic field and their interaction directly in one FE simulation only? We will really appreciate it if you can help to figure it out.</p> <p>Question #2 : 2D model depth setting</p> <p>We are having problems with the 2D model depth setting, because we could not find a portal to set this parameter. Is there any guidance for this?</p> <p>Best, Kessa</p> Sat, 20 Apr 2019 01:30:37 +0000 4.2019-04-20 01:30:37.232542 Novel Solar Concentration System <p>Hello Everyone</p> <p>I'm doing my thesis paper , the title of my thesis is optimization of novel solar concentration system . I'm a new user for comsol , and actually I need a strong support for working on it. I have to design the system , and study the optimum orientation and the direcrtion of the ray's which wil provide us the maximum power. So, from where do I have to start and how can I build the system and study it ??</p> <p>Appreciate your support</p> Fri, 19 Apr 2019 15:47:41 +0000 4.2019-04-19 15:47:41.232511 Long setup time before initial time step for study for model with large number of probes <p>I have a model with over 2400 probes, and it takes about 4 hours before it starts taking the initial time step for the study. I created a simpler model with varying number of probes and logged the time it took before it started the initial time step. I found that the number probes seems to scale as O(n^3), which means at around 1800 probes, the simple model takes over an hour to start up. Is there a particular reason why probe setup is an O(n^3) operation?</p> Fri, 19 Apr 2019 13:49:19 +0000 4.2019-04-19 13:49:19.232502 how to couple two different physics <p>Hello.</p> <p>I'm trying to simulate and get result about how EHD Inks are changed by electric currents. I already downloaded example(inkjet.mph) from this repository( Using this example, I added electric currents and did compute how inks were getting changed under each voltage while inks were flowing. but the thing is the results were out seperately such as one of fluid model and one of electric currents model. So I tried to couple "Levelset, flow fluid" and "electric currents". but whenever I checked Levelset or flow fluid with electric currents, it doesn't show available physic in "mutiphysics coupling" button in physic tab. how can I simulate what I want?</p> <p>plz help me out.</p> <p>thx in advance.</p> Fri, 19 Apr 2019 08:19:27 +0000 4.2019-04-19 08:19:27.232493 2D model of a Phosphoric Acid Fuel Cell <p>I'm trying to make a 2D model of a Phosphoric Acid Fuel Cell. For the physics part I'm using electroanalysis by adding cyclic voltammetry for the study. I've used the same parameters as the one in COMSOL application gallery (see attachment). I would like to know if i'm defining the components correctly in the electroanalysis or not? (Check the PDF file please). And i would like to know if there's anything i need to add or fix?</p> <p>Thank you</p> Fri, 19 Apr 2019 07:00:47 +0000 4.2019-04-19 07:00:47.232492 Modeling heat transfer and fluid flow in a greenhouse using evaporative cooling and fans <p>I am currently on a capstone project in which we are designing a greenhouse specifically for Hawaiian or tropcial climates. We currently have a proposed design but I am having trouble modeling it correctly on COMSOL. The pad system, or evaporative cooling, causes temperature change on the inflow of air, and that air is outputed through exhaust fans, or the outet. There are also fans situated in the greenhouse to induce flow.</p> <p>I did tutorials such as pipe flow, evaporative flow, and basic heat transfer. I am just wondering if there are certain tutorials or suggestions to make it easier for a fairly new COMSOL user.</p> <p>Thank you!</p> Fri, 19 Apr 2019 06:22:19 +0000 4.2019-04-19 06:22:19.232482 combine solutions (stitch solutions) <p>The below post is related to an <a href="//">archived discussion</a></p> <hr /> <p>Combine solutions / stitch solutions / join solutions</p> <p><a href="http://"></a></p> Thu, 18 Apr 2019 19:43:21 +0000 4.2019-04-18 19:43:21.232462 Memory usage increase with looped LiveLink for MATLAB <p>Question: How can I eliminate the LiveLink for MATLAB server from storing "things" in my computer memory when I run it?</p> <p>Situation: I am working on simulating li-ion battery at different conditions using For-loop function in MATLAB. Since I am running many iterations, my computer slowly run out of memory before I can finish all the possible conditions. These different conditions need to be run in a single run for optimization purposes, thus the option of separating them into smaller chuck cannot apply here.</p> <p>Why can't the LiveLink for MATLAB runs like the normal COMSOL software, where with each new run the memory usage is reset.</p> <p>Thing I have tried: I have look through the User Guide for LiveLink for MATLAB and here are what I found: 1. We can remove all the model objects by using the following command: ModelUtil.clear 2. We can disable the model history by using: model.hist.disable Even after I tried these two commands, the memory usage for the LiveLink server still increase everytime I run the model. And the memory does not reset itself. I have to close the program and rerun it to reset the memory.</p> <p>Please let me know if you have any suggestions.</p> <p>Thank you!</p> Thu, 18 Apr 2019 17:23:46 +0000 4.2019-04-18 17:23:46.232452 Add piezo-resistive coefficient in doped channel for MOSFET <p>Hi,</p> <p>I am simulating the Ids current in a MOSFET. I want to see the Ids current change on application of mechanical stress. For that, I wanted to add the π-44 piezo-resistive coefficient of the doped channel region...</p> <p>How do I change the property of just that region? And should adding this region property help to see variations in Ids with stress?</p> <p>Thank you very much..</p> Thu, 18 Apr 2019 16:32:46 +0000 4.2019-04-18 16:32:46.232442 Total enthalpy calculation <p>Hello I´m simulating a system where I want to know the inlet specific entalphy, the fluid is water so Cp=4186 J/kg*K. With the boundary probe Comsol gives me the values I´m looking for but I want to be sure they are correct. I know the temperature of the inflow is 298.15 K and the result of the total enthalpy is 101.51 J/kg. For the entaplhy I used this expression "ht2.H0" which Comsol says it´s "Total enthalpy". Can someone tell me how does Comsol calculate values with this expression? I thought that, in first approximation, dh=cp * dt but the results are not the same. Thank you for your help</p> Thu, 18 Apr 2019 15:07:45 +0000 4.2019-04-18 15:07:45.232431 Implementing custom boundary heat source with local x coordinate <p>I want to insert a custom boundary heat source which depends on position x along an edge.</p> <p>I have two questions on how to implement this:</p> <p>How can I create a local x coordinate along this edge? Something like xedge_1</p> <p>How to insert this local x coordinate inside my function, is it as easy as y(x) = f(xedge_1)?</p> <p>For reference purposes, I included the function, dependent on x and also the specific boundary. A general description on the above two steps would suffice.</p> <p>Kind regards,</p> <p>Marvin</p> Thu, 18 Apr 2019 13:16:46 +0000 4.2019-04-18 13:16:46.232411 Problem with Interpolation (Comsol 5.3a) <p>Dear all,</p> <p>this simulation model interpolates a data from a file which plots refractive index and voltage. It is able to import the file but when I start to run the model, it fails. I think there is a problem with defining the units/function in the interpolation section.</p> <p>The error says: Circular variable dependency detected. - Variable: comp1.mat5.rfi.n_iso - Geometry: geom1 - Domains: 3, 5</p> <p>Can anyone help me figure out the problem?</p> <p>Kind regards, Grace</p> Wed, 17 Apr 2019 12:23:02 +0000 4.2019-04-17 12:23:02.232341 Unblocking in Mesh <p>Q: How can i unblocked some areas in mesh?</p> Wed, 17 Apr 2019 05:33:58 +0000 4.2019-04-17 05:33:58.232323 How do I edit equation in COMSOL Multiphysics? <p>I am using COMSOL Multiphysics 4.3a. I want to simplify the Helmholtz equation in "Pressure acoustics". How can I make some variable (Monopole and Dipole source item etc) value zero to simplify it? I am attaching a picture here of the equation which I want to simplify. Thanks in advance.</p> Wed, 17 Apr 2019 01:35:39 +0000 4.2019-04-17 01:35:39.232312 The size of PZT-5H transducer <p>Hi everyone,</p> <p>I have been wondering if the size of PZT-5H transducer in my ultrasound modelling will affect the wave propagation or ultrasound behaviours. If yes, is there any reasons causing that to happens?</p> <p>Thank you.</p> Tue, 16 Apr 2019 16:39:26 +0000 4.2019-04-16 16:39:26.232311 Temperature Distribution in a Vacuum Flask with a Matlab problem <p>hello,guys, I'm learning from a case study,but I meet a problem,when i click Compute,it shows a error:Multiphysics compilation error.Can not start Matlab,- Library: csmatlab.dll could you teach me how can i figure it out.</p> Tue, 16 Apr 2019 12:58:19 +0000 4.2019-04-16 12:58:19.232263 Surface plots in a log scale <p>Hi there</p> <p>I am generating a 2D surface plot of the velocity magnitude. I have small regions with very high velocity magnitude compared with the remainder of the plot. This means I have small regions of red and the remainder of the plot is blue so the velocity gradient within the blue portion of the plot cannot be visualised very clearly. I tried plotting log(spf.U) and this helped with the visualisation but then the colour legend was not easy to interpret. I would like to have the colour legend labelled with the original velocity values, not the log values. Is this possible?</p> <p>Thanks</p> Tue, 16 Apr 2019 10:52:54 +0000 4.2019-04-16 10:52:54.232261