Latest Discussions - COMSOL Forums

Time Dependent Study Piezoelectric Element

Fri, 22 Sep 2017 22:48:04 +0000

Hello,

Im a new user to COMSOL and as a result I think I am stuck on a pretty basic part of the program.

I want to create a time dependent study for a piezoelectric element where the electric potential of each sector changes with time, see attached image. However, when I run the study, my electric fields seems not to change. Any ideas why this might be happening? Could it be because I have defined an initial voltage instead of one that varies with time ?

What Comsol module would be best for modeling the force from solar radiation pressure on a solar sail?

Fri, 22 Sep 2017 19:48:35 +0000

The equation for the force from the reflection and absorption of photons from the sail on a flat square sail is:

The is the incident solar radiation direction, is the normal vector from the sail, is the diffuse reflection coefficient, is the specular reflection coefficient, is the absorption coefficient, is the solar radiation pressure, is the area of the sail. The 's come from the absorption of the solar radiation, and they are the surface emissivity, with the subscripts b and f referring to the back and front side of the sail. The comes from the diffuse reflection of the solar radiation, and is the non-lambertian coefficient, with the subscripts b and f referring to the front and back side of the sail.

To start, I am just trying to find the force vectors on a flat square sail, but eventually I would also like to be able to calculate torque as well. I would also like to be able to use a non-flat sail and vary the angle of incidence.

How to open models before version 3.5a, The model is posted below. Please can I get help

Fri, 22 Sep 2017 18:50:27 +0000

The below post is related to an archived discussion

[start here]

Using "withsol" inside of study to extract solutions

Fri, 22 Sep 2017 17:56:42 +0000

I've been struggling to implement the "withsol" operator in my study. I'm trying to model the change in eigenfrequency due to thermal diffusion.

However, since the eigenfrequency solver can't take a time series as input, I need to extract discrete solutions from my Time Dependent thermal diffusion model to feed into the Eigenfrequency solver. "Withsol" seems like the best tool for this job, but I'm unsure how to include it into my study to get the solutions at the time steps I am interested in. Has anyone had success with something like this?

Frequency response of a multibody model

Fri, 22 Sep 2017 12:07:40 +0000

Hi there,

I've made a multibody model - actually it is a double pendulum, with a spring attached between the two 'free' ends These two free ends are constrained to motion in one direction only; excited by a harmonic predescribed displacement of one of the ends. I would like to get the frequency response of the opposite free end, when the other is excited with this predescribed displacement (and not a force).

I'm not familiar with the theory of multibody dynamics, and therefor not sure about if what I want to achieve actually can be achieved. Specifically I've defined a Rigid Connector to the excitation point, specifying the harmonic displacement - and right-clicking the rigid connector node, choose Harmonic Pertubation and specifying the maximum displacement. This works fine in a time-dependent analysis. When doing a frequency response analysis though, the displacement of the excited point in the structure is greater than I've specified. Also, the opposite free end of the double pendulum exhibits greater displacements than physically possible, if the pendulum arms are rigid links.

Is the approach I am applying the way to go, or should I do something else? And why do my structure displace more than physically possible?

Thanks in advance.

Best regards David

Error; how to interpret.

Fri, 22 Sep 2017 07:00:08 +0000

Hi there,

I've set up a multibody-model with contact, though when I open the model file, I get the following message:

Invalid property error: - Property: segcomp 'Selection' is an array of strings Variable component strings

My model runs fine as far as I can see, but I'm not comfortable with getting errors I don't know what means. Does anyone know what the error message refers to, or - in general - where / how to figure it out?

Thanks in advance!

Besr regards David

Defining an inhomogeneous material as a function of the electric field in a previous simulation

Fri, 22 Sep 2017 02:12:31 +0000

I'm simulating a structure with a 3rd order nonlinearity in the permittivity. In my first simulation, I've assumed that the nonlinear structure has a permittivity dependence on the local electric field. As a sanity check, in a second simulation, I want to define this structure to be linear and have an inhomogeneous permittivity which is a function of the local electric field in my first simulation. Any ideas how I can define an inhomogeneous material which has a local permittivity equal to a function of my first simulation?

Question about equation customization in viscoelastic model.

Thu, 21 Sep 2017 23:54:54 +0000

Hi everyone, I am now doing a simulation on a spheroid viscoelastic model. With given displacement on the boundary layer. I would like to explore the inner stress distribution in the sphere. However, the model i want to use is slightly different from the Kelvin-Voigt model which was given in COMSOL. I would like to customize the equation in order to substitute one of the spring module with a dashpot module. Is this feasible in COMSOL? Best, Mingshi

Contact Geometries

Thu, 21 Sep 2017 16:02:49 +0000

I am trying to simulate displacement of a neural probe inside of brain tissue using a 2D axisymmetric model, however, I have some basic questions about how to form these geometries. Right now, I have the two geometries built, but I am unsure if they are truly two separate entities, or if they are just divided by a line. When I create form assembly (which I believe I have to do since I will be applying a downward displacement to the top of the "probe"), the contact pair is not created at the interface between the probe and the block, so it seems as if COMSOL does not recognize them as two domains. How should I proceed? I previously tried "cutting out" a section of the brain block first using a Difference, and then building the "probe" where the difference was, but it seems like this was excessive. Any suggestions would be appreciated!

Thanks, Allison

Electric current and link with Magnetic and Electric fields module

Thu, 21 Sep 2017 14:19:29 +0000

Hi Everyone,

I would like to run a model of 3 conducting layer between them there is insulation. There are a connection between the 3 layers at one pooint that allows the currents flow. I have done this by making an electric current module that i define the current that flows from the upper layer to the lower layer.

Afterwards, i link the current densities that are generated by the electric current module with the magnetic and electric fields module. Then i run both modules.

I noticed some points. 1- There are some currents that flow into the insulating parts ( Not expected) 2- The electromagnetic losses that are generated from both modules are different. ( I run stationnary model)

Can you give me more details how to tackle this problem.

Ahmed Hemeida

What to do with this error " Failed to find consistent initial values. Undefined value found.Undefined value found in the stiffness matrix"?

Thu, 21 Sep 2017 13:47:11 +0000

Hi guys,

I'm trying to simulate 3D model of Vibrating beam in fluid flow which has a 2d counterpart in Comsol application gallery. I've met this kind of frequent error like shown below. I checked the boundary conditions serveral times and it's not a complicated one, I didn't find the problem. I also scaled down the dimension of the geometry and refined the mesh, it still turned out the same error. I f anyone has some idea about how to deal with the error, I'll be very grateful. Best Regards

Failed to find consistent values. Undefined value found. Undefined value found in the stiffness matrix.

For mesh-case there are 54 equations giving NaN/Inf in the matrix rows for the variable x.

For mesh-case 1 there are 54 degrees of freedowm giving NaN/Inf in the matrix columns for the variable x.

For mesh-case 1 there are 54 degrees of freedowm giving NaN/Inf in the matrix columns for the variable y.

For mesh-case 1 there are 54 degrees of freedowm giving NaN/Inf in the matrix columns for the variable z.

Last time step is not converged.

last time step did not converge when deleting the outlet or outflow in physical fields

Thu, 21 Sep 2017 08:35:55 +0000

I am simulating the process of ice melting. I used the physical fields of laminar two-phase flow(phase field) and heat transfer in fluids. First time the computation gave a reasonable result with the outlet boundary added. But then I thought no ice or water came out in the process. So I deleted the outlet boundary and then an error appeared: last time step did not converge. Has anyone ever seen this problem?ps: after I adjusted the model and added the gravity, the results seemed very irrational——the ice decreased at first and then increased greatly.

Error: dependent variables are not coupled to all degrees of freedoms in the model

Wed, 20 Sep 2017 23:35:53 +0000

Hello!

I am trying to run the following model, but keep getting an error:

"The following dependent variables are not coupled to all degrees of freedoms in the model: comp1.cPAHx, comp1.cPAx. - Domain: 1"

Are there typical reasons for this type of error?

Thank you in advance for any help you can provide!

Lawrence

Modeling interface movement during evaporation of a single water droplet

Wed, 20 Sep 2017 23:33:44 +0000

Hello,

I am trying to model the evaporation from a single water droplet (initially at 50 degrees celsius) evaporating into stagnant air (initially at 20 degrees celsius). I want to have movement of interface due to mass loss resulting from evaporation and I do not understand to accomplish this. Could someone please help me with this.

I have three physics nodes (a) Laminar flow with level set for droplet spreading (b) Heat transfer in air and water phases (c) transport of diluted species for transport of water vapor from air-liquid interface into dry air.

Error in simple contact model

Wed, 20 Sep 2017 19:12:49 +0000

Hello,

I am trying to model a simple tool-sample kind contact problem. It is a simplification of what we are trying to do, but it does not work even in this simple case.

The model consist in a block of a linear elastic material (the idea is to add plasticity in the future), against a rigid wedge. The block moves towards the wedge making the penetration going higher. I've tried with a time-dependent solver, and using parametric sweeps with different types of stationary solvers, but the model always ends up with a convergence problem.

I have also tried different contact methods (augmented lagragian and penalty) with different factors. Nothing seems to work. Could you give any recommendation?

Turbulent flow in a concentric annulus with heat transfer and magnetic field

Wed, 20 Sep 2017 10:38:41 +0000

Hi all,

I am currently modeling the magnetocaloric effect inside a horizontal concentric annulus to model a magnetic refrigeration system. However, after some investigation I saw that the flow is pretty turbulent and shows oscillatory behavior. Therefore, I want to change my current physics (laminar flow, heat transfer) to turbulent flow.

I saw in Comsol 5.3 that there are several models to choose from. Does anyone have any idea which one might provide the best combination of accuracy and robustness regarding my system?

Many thanks!

Evthimios

Mass flow rate in heat transfer model

Wed, 20 Sep 2017 09:16:40 +0000

Hello,

I'm currently busy with a model for scalp cooling. This means the head is cooled with a cold fluid that flows through a cap wich is placed over the head. I know this cooling fluid gets warmer as a result of the biological heat transfer. But this heat will be taken away because of the mass flow system of the cooling cap.

My problem is that I don't know how to model this mass flow. If someone could explain me how to make a mass flow inlet with a certain temperature and volume and a mass flow outlet with the same volume but warmer because of the biological tissue heat transfer I would be very gratefull.

Sincerly,

Max Graduate student applied physics

How can i give complex function as load?

Wed, 20 Sep 2017 05:53:30 +0000

I am doing a plate vibration problem there the boundary phase load is f=F0e^(iwt). I tried to create this wv1 of sine function and wv2 of cosine function. In the boundary phase i just gave the load as *wv1 + iwv2 *. Then i triend to solve my problem with time dependent study type, Then it is showing error, that complex number not accepting?? Please help me?

FUNCTIONS

Wed, 20 Sep 2017 04:31:16 +0000

Hi everyone. I need your help please.

How do I get a Time dependent load in COMSOL?

I tried using a step function but it gives me a very sharp transition zone as shown in my image.

My main problems are:

Obtaining a constant load for t seconds (e.g -10 MPa for time 0<=t<=40 sec)
After 40 sec my stress profile should be rising steadily to a maximum where it stays constant again.
My transition zone is too sharp.

Thank you for your support

Wear Simulation

Tue, 19 Sep 2017 23:24:22 +0000

Hello Everyone,

My name is Sai Krishna, I am trying to do wear simulation since many days. I am unable to figure out a couple of things. Can anyone please kindly help me.

1) I am unable to figure out how gave gap calculation thing and 2) where did we apply archard's Law for wear simulation.

This is the main step of my thesis. Can you please kindly take some time to explain me. I request for your help.

Thanking you,

With regards,

Sai Krishna Enabothula