Quantcast
Channel: Latest Discussions - COMSOL Forums
Viewing all 686 articles
Browse latest View live

Energy balance with variable density

$
0
0
Hello everyone,

I am interested in finding the temperature in a system that changes the density in time and I would like to formulate the energy balance using this variable density.

What I have in mind is to use

Cp * d( rho T)/dt = Q

but I find it pretty tricky to implement it in COMSOL, because of the derivative. Does anyone have any experience with this?

Nomenclature: Cp is the specific heat, rho is the density (rho = f(t)), T is the temperature and Q is the heat source in W/m3, which can be any number (just for reference)]

Rotary woofer simulation

$
0
0
Hello.

I am trying to simulate this recently developed rotary woofer:
www.rotarywoofer.com/
www.jpark.us/pubs/JASA_125(4)_RotaryWoofer_2009.pdf

My main goals are to find out air flux and energy flux rates going through it, as a function of rotation velocity and signal frequency.

Is that possible?
I reckon simulating an axial fan is probably the first step to take, but I get a bit lost in comsol even for that.
Any tips or directions?

Thanks

Modify continuity on internal bounderies

$
0
0
Hi everyone,

I have some doubts about how to modify continuity condition on internal boundaries
and I would ask for some help.
I want to solve hydrodynamics on a unit cell of the Taylor flow (known also as
slug flow or bubble-train flow) in a circular millichannel. This unit cell is composed by a
bubble and two liquid half-slugs. I consider that bubble have a fixed shape and
remains immobile.
So, what I want is to solve Navier-Stokes equations into the two subdomains (bubble and liquid)
considering the gas-liquid interface as an internal boundary but establishing a "No
penetration" condition as done by the "slip boundary condition".

Thanks in advance,
Freddy

About check the fluid setting

$
0
0
Hello when i calculate the fluid model where and How could i know the Reynolds number and Stokes number?

Boundary condition

$
0
0
How to impose perfect contact (equal displcaement) between internal boundaries of different surface (resin-mold interface).

Defining a variable limited only for the boundary

$
0
0
Can we define a dependent variable for a physics which is only valid for particular boundary and not for the whole domain?

cone-jet mode in electrostatic field

$
0
0
Hi,

Here is my problem.
I want to simulate a electrohydrodynamic (EHD) phenomenone , a nozzle with full water (or electrolyte) in a electrostatic field, the the water is polarized and deformed the Taylor cone,then comes the cone-jet mode. My models include the level set -laminar flow and electrostatic,but still cann't get the cone-jet result.Is there anybody also working on a similar model like mine?Did you get the cone-jet results?

Thank you.

Form assembly or Form union

$
0
0
Dear All,

I encountered a problem when I models fluid flow through conduit surrounded by solids at two sides. The models involves laminar flow module and structure module. According to the description of reference manual, When the contact features are included in structure mechanics, the "form assembly" in geometry is recommended. So I choose the "form assembly" box. However, when I imposed no-slip BC at the interface between solid and liquid domains, and run it, the error occurs.
Undefined variable.
- Variable: comp1.spf.U
- Geometry: geom1
- Domain: 2

Form the attached image, we can see that domain 2 is solid zone, no fluid, whereas the error didn't detect fluid velocity in the soild zone. It seems weird. I try many times to resolve it but failed.

So I am wondering whether the "form assembly" in my model is the criminal. But considering the contact feature in my model (circled region in red line), I am not sure for it.

Does anyone give me some hints or suggestion on that? I will appreciate your assistance.

Warning: SOR line did not find any lines.

$
0
0
Hi everyone,
I am running a laminar two-phase flow (level set) simulation with a 3D geometry. I had previously simulated the same geometry in 2D with the same initial values and boundary conditions and have no problem reaching a solution. However, when I try to run the simulation with a 3D geometry, the warning "SOR line did not find any lines" keeps coming up and eventually I get error saying
Nonlinear solver did not converge.
Divergence of the linear iterations.
Time : 0
Last time step is not converged.

Has anyone encountered this warning before and know what to do with it? Thanks a lot in advance!

Level set method dealing with two phase flow with extraordinarily high density ratio (>100,000)?

$
0
0
Dear all,

I am working on the simulation of droplet formation and transfer in GMAW (gas metal arc welding) using COMSOL.

A schematic view of GMAW system was shown in attached figure.

A constant current is applied to the electrode through the contact tube. Argon gas is provided from the gas nozzle as shielding gas.An arc plasma is struck between the electrode and workpiece. The steel electrode will be melted at the tip due to the high temperature arc. Droplets are formed at the molten electrode tip, and then detached and transferred to the workpiece.

To represent the GMAW system in COMSOL, the physical fields I used include Electric Currents, Magnetic Fields, Heat Transfer in Fluids, and Laminar Two-Phase Flow, and I have tried both levet set method and Phase field method.

In the fluid flow field, the two phases considered in this model are liquid steel and argon gas, and the density ratio of these two phase is about 7200/0.05=144000.

Although the process in COMSOL is working, the results are quite different from others' results in references using VOF method.

So I am wondering if the level set or phase field is capable to handle two phase flow with extraordinary high density ratio. And if not, how can we incorporate VOF method in COMSOL?

I would be very appreciated if anyone could help me figure this out. Thank you.

How to add a mass source term to NS equation

$
0
0
Hello everyone

I have a problem in adding a mass source term to the NS equation. I use it to account the mass loss due to vaporization during laser melting process. Currently I use the interface moving speed based on mass flux in the moving mesh module. But I'm not sure if it works properly.

Previously in V3, I can easily add the source term to the NS equation in the equation system.
But in the current V5 or V4 COMSOL, I cannot find the equation system, or any method to add a weak contribution to the continuity equation.

Do you have the solution? Thank you very much for your help.

serval questions of heat transfer in my project

$
0
0
Hello
As the figure shows, I am using the fllowing models
(Heat Transfer>Conjugate Heat Transfer>Laminar Flow
Stationary situation
Surface-to-Surface Radiation), and try to find the temperature distribution of the air and the spheres.
I simplified the original object into two spheres,since its shape was very complex. The latter sphere has a boundary heat source.
I have some questions to solve:
1.Since the object was simplified, so the real convective area was changed, i need to modify the convective heat transfer coefficient, but I can't find the modify interface in comsol.
2.The Surface-to-Surface Radiation was considered, and the model is periodic, so I need to define the periodic boundaries as perfect mirrors (the default insulated conditions are just fine to the convective heat transfer and conduction heat transfer), but I can't find the interface in comsol,too.
3. Since the original object was simplified into two spheres, so the heat conduction was incorrect,because the spheres are not contacted.I want to find the ht.gardT between the spheres and use the Fourier's law of heat conduction to define extra heat source on spheres, is it right?
thank you

Gas mixture flow simulation

$
0
0
Hi everybody,

Gas 1 is flowing in my geometry (a round pipe) and I want to inject gas 2 into the flow.

My idea was to assign both gases to the same geometry (using pipe flow module twice) and then use the transport of concentrated species module to adjust the molar fraction of gases in a way to tell COMSOL that initially there was just gas 1 and then an x molar fraction of gas 2 is injected into the flow.

My question is: Is this a right way to deal with the problem and how to tell COMSOL to assign different materials (gas 1 or gas 2) to the same domain dependent on the physics?

Thanks.

Matrix has zero on diagonal

$
0
0
I have to solve a lid driven cavity (laminar flow) + melting problem (heat transfer in fluids and solids with moving mesh). I was able to successfully implement this in 2D.

I have now shifted to a 3D model, still solving the same physics. I get an distinct ( i haven't encountered this previously) error message which I’m unable to understand :
Failed to find a solution for the initial parameter.
Matrix has zero on diagonal for the SSOR update submatrix in the Vanka algorithm.
Returned solution is not converged.

In the Comsol reference manual, explanation for matrix equation for vanka algorithm is attached in the picture

Is the diagonal element (Sj) = 0? This S corresponds to the average values of the velocity vectors?
My initial condition is set to zero.

I have also attached my model file.

Kindly help

Non-isothermal pipe flow (nipfl)

$
0
0
Gooday,

I'm using non-isothermal pipe flow model and heat transfer in solids physics in order to simulate a bore hole heat exchanger (BHE). the problem consist of a tube modeling by nipfl and a ground around the tube modeling as a solid heat transfer. The program seems to simulate the problem in a reasonable way but when the thermal resistance (R) is calculated, the obtained R by Comsol is higher almost 30 % compared with R obtained by formula or 2D model. I guess since this module simulate the pipe flow as a 1D problem and there is a gap between the pipe and solid which in graphics shows as a solid part, it adds the value of thermal resistance. Can anyone help or has a suggestion?

In attachment you can find the file.

kind regards,
Amin

Discrepancy between particle trajectory and streamline?

$
0
0
Hello,

I'm trying to teach myself how to use the particle tracing module, so I'm modeling a simple T-junction microfluidic channel. I get an unexpected result, however, when I trace the particles. Since it is a microchannel, I expect the particles to follow streamlines. However, they do not (see attached). I've tried both the Newtonian and massless configurations. This is a 2D, incompressible model, assuming a shallow channel of 30 um. Could anyone lend any insight into what's happening or which results I should trust?

Thanks,
Alisha

Parachute CFD simulation

$
0
0
Hello,

I am trying to perform a CFD flow field simulation of a 3D parachute to find its lift and drag. The geometry model is imported from a CAD software as surfaces only (no thickness). I attached one picture of how it looks like. In order to make the simulation, I created the computational domain (box) big enough to neglect the boundary effects (I saw in a post from an airfoil SST model that a recommended value of 100 chord-lengths toward all sides is enough). For the physics, I am using the SST turbulence model with inlet (velocity input), outlet (pressure output) and symmetry (box lateral sides) boundary conditions. I used an interior wall boundary condition on the surfaces of the parachute with no slip condition. For the mesh, I have tetahedrals for the domains and prisms for the boundary layer over the parachute surfaces. After running the simulation, I visualized clearly the vortices created in the interior of the parachute. However, when calculating the surface integrals to determine lift and drag (I saw in a comsol blog to use spf.T_stress to calculate lift and drag at the correspondent axis direction), I am getting nonsense values (my drag is on the opposite direction resulting in a propulsive force). For sanity checks, I was also trying to see how the pressure differs on the parachute surfaces from the inside and outside but I only see the same values on both sides.
My questions would be:
1. Is this approach valid for this type of scenario?
2. How would it be possible to find the pressure difference on the inside and outside surfaces of the parachute? (I am not sure if this has something to do on how the interior wall boundary condition works). From my understanding, I thought this internal wall boundary condition would act as a baffle (very thin surface) for the parachute surfaces (which will therefore have both inside and outside surfaces). Please correct me if I am wrong on this.

I would really appreciate some help with this issues. Thanks.

Problem in 1D EHD simulation

$
0
0
Dear all,

I have built a 1D EHD in COMSOL. But it does not work. There always exists an error. Can anybody give me any suggestion? Please help me! My *.mph file is in the attachment. The error is as follows

The following feature has encountered a problem:
Failed to find a solution.
Singular matrix.
There are 249511 void equations (empty rows in matrix) for the variable comp1.P.
at coordinates: (2.5949,0), (2.5948,0), (2.59482,0), (2.59484,0), (2.59486,0), ...
There are 1 degrees of freedom that do not occur in any equation (empty columns in matrix) for the variable comp1.H0.
at coordinates: (0,0), ...
Returned solution is not converged.
- Feature: Stationary Solver 1 (sol1/s1)

Thanks a lot!

Comsol not following ideal gas law or conservation of mass

$
0
0
Hello,
I am currently working on modeling a high temperature hydrogen heat exchanger. I decided to start as simple as possible and just model the heat transfer into an axi-symmetric tube of hydrogen surrounded by a solid shell.

I am running into two problems. First, COMSOL has decided that the conservation of mass is a suggestion at best and need not be followed. As a check on the results I was getting, I checked the mass flow rate per unit area. I found that the exit mass flow rate is about 25% higher than the inlet mass flow rate. Thus something is obviously wrong with this.

Second, although I have experimented with treating the hydrogen as both a fluid and an ideal gas, at the moment I am letting it be an ideal gas. However, COMSOL doesn't seem the ideal gas law any more than it did the law of conservation of mass. When the pressure and temperature at the inlet are used in combination with the specific gas value of hydrogen (4124 j/kg), a rearranged ideal gas equation rho=P/(R*T) says that the density should be about 1.4 kg/m^3. COMSOL however, returns a value 17 thousand times less than that.

Does anyone have any thoughts or ideas on these problems?
Dan

Inlet pressure boundary condition not allowing COMSOL to converge

$
0
0
Hi,

In my simulation (file attached here), I'm modeling a slight variation of the 2D axi-symmetric 'stirrer in a cylinder' situation, which is on COMSOL model library.
In my variation, there is applied pressure on the top of the cylinder in which the stirrer rotates. Below the cylinder-stirrer setup is a highly impermeable layer (modeled by Brinkmann equation) through which the fluid has to pass.
This is a simulation of a lab setup, so I know that the applied pressure at the top is usually 4 bar. The outlet below the membrane is open to atmosphere (so the fluid has to pass though the cylinder, through the, membrane and then the open outlet) .

However the pressure boundary condition at inlet does not allow comsol to converge.

I've seen this happen even in a simple pipe flow situation where beyond a certain max pressure, COMSOL can't converge.

Please provide me suggestions!
Viewing all 686 articles
Browse latest View live