Heat Transfer - Latest Discussions - COMSOL Forums

Modelling a moving, breathing animal

Thu, 17 May 2012 14:14:34 +0000

Hello,

I am rather new to COMSOL and so far I have not been able to pinpoint how to model an animal moving inside of a closed box. I have tried to use parametric sweeps but I don't think that is quite what I am looking. Basically, I need an animal (modelled as a solid ellipsoid) to move around the bottom of the box and then view the rather delicate pressure perturbation incurred by this movement.

The next step of the project would be to simulate the animal breathing and view the pressure & velocity changes in the box as the animal breaths in and out.

Any suggestions as to how to start? i.e. stick to the parametric sweeps and model the breaths as an interior wall with periodic flow?

Thanks,

Arman

Convergence issues for transient freezing (apparent heat capacity phase change)

Tue, 15 May 2012 17:49:38 +0000

Hello,

I am fairly new to COMSOL and am trying to solve a transient (over 600 seconds) freezing (phase change) problem in a gel (similar thermal properties to water). I am using temperature-dependent thermal conductivity and density (both implemented as cubic-smoothed lookup functions, f(T) ). I am simulating the phase-change with the apparent heat capacity method (also implemented as a cubic-smoothed lookup function, f(T) ).

Across the model, my temperatures vary from 140K to 300K, with fairly high temperature gradients near the cooling interface (order of 40K/mm).

I have a 2D geometry with phase change occurring over <1K, that solves without issue. However, I have a slightly more complicated 3D geometry that I am having problems converging. I can get the solution to converge, treating the phase change as a triangular function over >30K, but the solver gets hung up for any more aggressive phase change.

I would like to reduce the temperature duration over which my phase change is implemented (less than 10K). I am assuming a triangular “pulse” will provide the most gradual change for specific heat, but is this an incorrect assumption and will another function provide less of an issue for convergence? Is there any other settings I can look at to help with convergence? I am using the default settings for the transient solver.

Any assistance you can offer would be greatly appreciated.

Thanks!
Michael

defining a variable for each element

Tue, 15 May 2012 08:47:01 +0000

hi ,
i would like to know how can i define a variable (which i am using in calculating other properties of the material) such that it is solved for each element separately ?
Right now what COMSOL does is solve for a time step and use that value to calculate material properties for all elements together , whereas i would like COMSOL to solve that variable for all elements differently (much like temperature in heat transfer) in each time step.

- Hirendra

Thermal Dispersion in a Poreus Medium Module Problem

Thu, 10 May 2012 11:33:25 +0000

Hi,

I am working on simulation an Aquifer Thermal Energy Storage (ATES) [1] system. In this I am trying to use thermal dispersion, which is part of the heat transfer module. However I cant get it to work. I have tried the manual and the internet, but I cant figure out what the problem is and let alone solve it. Maybe one of you guys can see what the problem is, and how I would be able to solve it?

I've included the error log and the model, which is a 4.0a model.

Greetings Nick

[1] http://www.iftech.co.uk/index.php?option=com_content&task=view&id=21&Itemid=33

Attempt to evaluate negative power of zero.
- Function: ^
Failed to evaluate temporary symbolic derivative variable.
- Variable: mod1.ht2.k_dispxx@VDN$mod1.py
- Defined as: (((1/max(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))),eps))*((mod1.ht2.rho*mod1.ht2.Cp)*(((((1/(mod1.dl.g*mod1.dl.rho))*(-mod1.dl.Kyy))))*(mod1.ht2.uy*2))))+((((((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))^-0.5)*0.5)*(((((1/(mod1.dl.g*mod1.dl.rho))*(-mod1.dl.Kyy))))*(mod1.ht2.uy*2)))*(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2)))>eps))*((-((mod1.ht2.rho*mod1.ht2.Cp)*(((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))))/(max(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))),eps)^2))))
Failed to evaluate temporary symbolic derivative variable.
- Variable: mod1.ht2.k_eqxx@VDN$mod1.py
- Defined as: ((((1/max(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))),eps))*((mod1.ht2.rho*mod1.ht2.Cp)*(((((1/(mod1.dl.g*mod1.dl.rho))*(-mod1.dl.Kyy))))*(mod1.ht2.uy*2))))+((((((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))^-0.5)*0.5)*(((((1/(mod1.dl.g*mod1.dl.rho))*(-mod1.dl.Kyy))))*(mod1.ht2.uy*2)))*(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2)))>eps))*((-((mod1.ht2.rho*mod1.ht2.Cp)*(((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))))/(max(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))),eps)^2)))))
Failed to evaluate temporary symbolic derivative variable.
- Variable: mod1.ht2.k_effxx@VDN$mod1.py
- Defined as: (((((1/max(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))),eps))*((mod1.ht2.rho*mod1.ht2.Cp)*(((((1/(mod1.dl.g*mod1.dl.rho))*(-mod1.dl.Kyy))))*(mod1.ht2.uy*2))))+((((((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))^-0.5)*0.5)*(((((1/(mod1.dl.g*mod1.dl.rho))*(-mod1.dl.Kyy))))*(mod1.ht2.uy*2)))*(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2)))>eps))*((-((mod1.ht2.rho*mod1.ht2.Cp)*(((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))))/(max(sqrt((((mod1.ht2.ux^2)+(mod1.ht2.uy^2))+(mod1.ht2.uz^2))),eps)^2))))))
Failed to evaluate expression.
- Expression: d(d((-mod1.ht2.k_effxx*mod1.T2x*test(mod1.T2x)-mod1.ht2.k_effyy*mod1.T2y*test(mod1.T2y))*(dvol),{test@3}),mod1.py)

Heat distribution of an infrared focused laser beam

Thu, 10 May 2012 04:21:38 +0000

Hi all,

I am looking to model the localized heating of a focused infrared beam. In specific I am looking to model a 1064nm optical trap, given the absorption of the trapping medium and the power of the beam.

The assumption I make is that all absorption events are dissipated through heat which is likely reasonable at these frequencies. I would also like to make this a transient model in the future, however I am suffering from fundamental problems.

I model the intensity profile of the beam, which seems to work out well. However when I couple the absorption (attenuation) coefficient as the heat induced I am encountered with non-physical temperatures.

Am i missing something simple or is my assumption just fundamentally wrong?

Thanks in advance
Jakrey Myers

Temperature Sensor Model

Tue, 08 May 2012 21:11:51 +0000

Hello,

I am designing a temperature sensor using an interdigitated capacitor. Using the electrostatics modules, I was successfully able to design the sensor with the starting capacitance value I wanted. However, regarding changing capacitance with temperature is where I run into problems. Thus far, I have been using the joule heating and thermal expansion module. Can anyone tell me if this is the correct module for my goal?

I have had some measure of success using it. However I find that when I remove the surrounding rectangle with air associated with it in the positive z direction and silicon in the negative z direction, the study completes, but with that rectangle, I get an error of singular matrix/solution does not converge and cannot complete the study. Why is this?

Also, without the surrounding rectangle, and without the proper dielectric in place, it seems that I cannot calculate capacitance correctly, and the values are all very small, approaching zero.

I have attached my model.

I would greatly appreciate help understanding this, thank you.

problem with conditional material properties

Tue, 08 May 2012 10:50:02 +0000

Hi,

I am trying to give a conditional electrical conductivity as a material property. It will be something like when the voltage becomes 0 then the conductivity should be 1 other than that it will change with the voltage. I am using a rectangular pulse as a pulsed voltage. So after the pulse the conductivity should be 1. I am using v4.2a. will be very glad if someone replies.

I used this but this is not converging,

if(v>0,(2.205*(10e-3*0.05)/(v*1*pi*0.65)),1)

Thanks ,
Arif

Please help me

Tue, 08 May 2012 09:44:52 +0000

Hi to all,
Please help me.I am doing heat transfer simulation in laser tube forming field. My distribution is guassian. i have used following conditions to beam rotate 180 degree on over the tube(top surface). I have moved the heat source by using cylindrical co-ordinates. I have used only GUI method.
My model radius 10[mm]
To Circular movement,
Y=R*cos(omega*t)
Z=R*cos(omega*t)
X=L/2
Analytic function is,
exp(-(r/rb)^2)
boundary Heat source is
(0.865*Ab*P/(pi*rb*rb))*an1(sys2.r[1/m])
My question is,
I want to increase the number of laser passes (only 180 deg)
Now I have done only one pass on the tube.
I am interesting to move the laser heat source on tube at many times (GUI method) (2, 3, 4 etc.)
I tried many times. But I couldn’t do it.
Please help me anyone…
Thanking you
Jay

non uniform heat flux

Mon, 07 May 2012 18:00:40 +0000

Hello all,

I'm trying to model simple 2D conduction model in a slab. What I want to do is use a non-uniform heat flux for a boundary condition. I already have the heat transfer coefficient distribution along that boundary, which is a gaussian type distribution. Instead of using a gaussian distribution I would prefer to actually call a file that will assign the values of the heat transfer coefficient along the boundary. Do you know if that is doable?

Thanks,
Anastassis

non isothermal heat transfer in room

Mon, 07 May 2012 14:42:34 +0000

Hello,

I am trying to model a simple room 4 m x 3 m, 2.90 m high with:
- open door (to large volume of air) - outdoor = open boundary, 20°C
- open window (to large volume of air) - indoor = open boundary, 20°C
- a heat source in the room = gas furnace, 100 m3/hr of fluid, temp 1000°C = inlet, laminar inflow 100 m^3/hr, temperature 1000°C
- gravity (-g_const*nitf.rho), defined as volume force

Physics is nitf, non isothermal transfer in fluid.

My problem is that it never converges although I reduced and improved a lot mesh size.
If I change the "open boundary" condition with "outlet", it can converge but setting a temperature at the outlet is not realist.

Does air density variation with temperature is automatic in 4.2 o do I have to introduce a formula to take this in account?

Thanks for any help you can provide,

Pedro

electro-thermal analysis of micro-heeater

Mon, 07 May 2012 04:07:32 +0000

sir,when i tried to simulate it it is showing some errror, and i could not find the error .can u please help me to solve this structre..

Mass balance problem when using user defined density.

Sun, 06 May 2012 22:12:06 +0000

I used user defined density in a Darcy's Law model because the temperature is very low (35 K for hydrogen) but there is mass balance problem (more than 10%) while when I used ideal gas as the mass balance is ok.
I have checked the density I defined there is no problem for this part actually the density is obtained from NIST/REFPROP through MATLAB.

Fan in Comsol

Sun, 06 May 2012 00:48:36 +0000

Is it possible to make a fan that spins by itself to produce an air flow? I know the other option is to just have an inlet somewhere with a fluid flow rate, and get the fan to spin because of the flow rate, but i am trying to get the fan to spin, and initiate the flow rate. My goal is to model the cooling system of a gaming system, so i want to be able to have the fan have different speeds to simulate when the system is just turned on and in standby and also when it is operating at maximum capacity for a long period of time.

Error in parametric sweep simulation

Fri, 04 May 2012 17:58:55 +0000

Hi,

I'm trying to run a joule heating simulation on a wire that is on a sapphire substrate. The wire has a hot spot in the middle and current is flowing threw it. My idea is to determine how long takes the system to thermally relax. I'm running a parametric sweep changing the thick of the wire, but when I compute the solution I get this error message:

Error in sweep: The following feature has encoutered a problem:

The following feature has encoutered a problem:
Extrude requires distances to be distinct.
Error in multiphysics compilation.
-Feature: Compile equations: Time dependent (sol1/st1)
-Error:Extrude requires distances to be distinct.

I've checked the busbar example in the tutorials and I can't find any differences between the example and my simulation.

I'm attaching the file I'm using

Thanks

Emanuel

Conjugate, non-isothermal u-pipe, time-dependent study

Tue, 01 May 2012 19:04:23 +0000

I wanted to simulate the heat transfer of fluid in a u-pipe. However, I'm really confused about which module and model should I use?

Initially what I wanted to do is to choose the heat transfer in porous media modulue, however, I found difficulties of solving the problem of directing the fluid from vertical pipe into u-shaped corner by defining a velocity. Based on my shallow experience with Comsol so far, it seems that this modulus only allows the fixed velocity vector to be defined at the input stage, and I found it difficult to resolve for the model.

I have also tried to use the conjugate heat transfer as well as non-isothermal flow... then I started to get more confused because my analysis has to be carried in time-depenent study. Would anyone give me some advices?

Thanks for your advice in advance!

calling variables in COMSOL

Sun, 08 Apr 2012 00:38:20 +0000

Hi,

I am creating a heat transfer model. I have imported a file as a parameter under Global Definitions, but COMSOL automatically labels the first column t and the second column f(t). In reality, the first column is actually T, temperature.

I am mainly confused because I extract data for the properties for thermal conductivity and Cp. I want certain values to only be used during specific temperature intervals. For example, I have imported files called, Cp_material1, Cp_material2, Cp_material3, k_material1, k_material2, and k_material3. Each of these files have the first column referring to temperature (not time) and the second column referring to the material property. When the temperature is below 200K I want the model to be using properties for material1, when the temperature is between 200-500K I want the model to be using properties for material2, and when the temperature is above 500K I want the model to use properties of material3.

I currently have something like this for k: k_material1(t<200)+k_material2(t>=500)+k_material3(200<=t<500).

What is COMSOL referring to as t and T? Should the expression above be using T instead of t?

Thanks

Dielectric heating between parallel plate electrodes

Fri, 09 Mar 2012 23:25:06 +0000

Hello all, I'm fairly new to Comsol. What I'm trying to do is model the heating of a dielectric in between two parallel plate electrodes at a frequency of 27.12 MHz. I would appreciate any suggestion as to which physics to use. Thanks in advance.

Extruding to 3d

Wed, 29 Jun 2011 18:00:15 +0000

Hi,
2 questions:

1) I am creating a ring in 2d, and it seems perfectly circular, but when I extrude it to 3d, it suddenly gets slightly jagged sides and becomes a sort of polygon. I tried to fillet it in 2d and then extrude, but it makes no difference, the 3d still has edges. Anyone know how to fix this?

2) Is there any way to put a 2d face on a 3d object? For example, on a cylinder, I would like to put a small circle on its top face to model a laser's heat deposition. I think this would let me impose heat transfer conditions on the boundary. Other suggestions are also welcome.
Thanks

error 6250 with Piecewise Analytic Functions

Thu, 23 Dec 2010 07:37:25 +0000

Hello,

I am using COMSOL 3.5 with the general heat transfers model.

I would like to impose an energy flow to a sub-domain, and this energy flow is function of time. Thus, I have written the polynomial equations of the flow in "options, functions, Piecewise Analytic Functions". The argument is t (time). The flux is cut into several equations during time.

But when I try to run the temporal model I have an error 6250 : Failed to evaluate variable jacobian. And the problem comes from the variable defined in “functions”

I have the same error with every equation I try, even the simplest (like f(t)=0 for 0<t<end). But when I write the same equation directly in the sub-domain, I have no error.

I guess the polynomial equations are correctly written because I can have a view of them when I press on “graph”.

As I try the model with simple equations, I don’t think the extrapolation or smoothing can help.

Where does my problem could come from?

Sincerely,
Anne-Maude Bazin

Determine Arrenhius damage integral for a solution

Wed, 12 Aug 2009 18:05:07 +0000

I am trying to apply an Arrenhius damage integral DURING the solution of an electrosurgical problem using the Bioheat Transfer Module along with the Conductive Media DC module (T and V are the dependent variables) . The integral is of the form:

Adi = int(A*exp(-E/(R*T),t)

where Adi is the Arrenhius damage integral, int stands for integral and A, E, and R are constants. T is the nodal tissue temperature and t is time.

Is there a way to apply this into the Global Equations by taking the derivative of Adi? Thus making the equation:

Adit = A*exp(-E/(R*T)

I tried this, but received an error that T could not be solved for. Do I need to do something in order for Global Equations to access my dependent variables?

Thanks for any help...I am also interested in any other way that anybody can think of!