Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Modeling a spring and damper system

Indranil Goswami
I am having trouble modeling a simple 2D spring mass damper system. The geometry comprises the spring at the upper end anchored (fixed) attached to a square mass which in turn is attached to a damper at the bottom of the mass which is also anchored. Any help on modeling both the spring and damper would be appreciated. Thanks

11 Replies Last Post Jan 26, 2016, 2:52 AM EST
Posted: 8 years ago Aug 25, 2009, 7:02 PM EDT
Do you care about stresses on the spring/damper? Or are you just looking for the displacements and speeds and the like?

If it's only displacements you're interested in, it might be much easier to model the spring/damper combo as equations in a transient analysis:

1.- Model a square with the characteristics you're looking for.
2.- add a restriction limiting to zero movement in the x axis.
3.- Under integration coupling variables, add two point variables, choosing the spring contact point: One for displacement (expression "v") and speed (expression "vt").
4.- Add a spring force -k*displacement where your spring connects to the piece.
5.- Add the damper force -c*speed where your damper connects to the piece.
6.- Add some sort of excitation force (a time-stepped function) F*(t<te) or whatever jolt to destabilize your system.

That should give you an answer. For reference, I'm uploading a very simple model of a .5m x .5m x .5m block. I added all the forces in the subdomain section, but you can play around. The model's settings can be changed under the constants menu, where k, c, F, m and te can be modified.

You can plot speed and displacement using the variables damp and disp, respectively.
Do you care about stresses on the spring/damper? Or are you just looking for the displacements and speeds and the like? If it's only displacements you're interested in, it might be much easier to model the spring/damper combo as equations in a transient analysis: 1.- Model a square with the characteristics you're looking for. 2.- add a restriction limiting to zero movement in the x axis. 3.- Under integration coupling variables, add two point variables, choosing the spring contact point: One for displacement (expression "v") and speed (expression "vt"). 4.- Add a spring force -k*displacement where your spring connects to the piece. 5.- Add the damper force -c*speed where your damper connects to the piece. 6.- Add some sort of excitation force (a time-stepped function) F*(t


Posted: 8 years ago Aug 26, 2009, 6:58 AM EDT
Hi

here you have a different model, in fact not that different, but I usually apply the forces to points, or I distribute them over boundaries. The approach is the same, but I get somewhat different results than Mario's, probably because I'm not sure he has correctly checked all his dimensions in his formulas, in 2D one often gets confused with the subjacent "thickness".

That is why I prefer to work as far as possible with calculated values (through boundary integrals) and I use the application mode density rho_... instead of mass for the gravitational field forces.

This is the nice way of COMSOL you might add some physics with just some formulas here and there, not that easy to do in other FEM programmes.

One thing tough If you start to add masses on points, check carefully that these are taken into account if you use the eigenmode analysis, you might well need to multiply them by -(jomega)^2 or a lambda^2 depending on how you define the masses and the application mode

Good luck
Ivar
Hi here you have a different model, in fact not that different, but I usually apply the forces to points, or I distribute them over boundaries. The approach is the same, but I get somewhat different results than Mario's, probably because I'm not sure he has correctly checked all his dimensions in his formulas, in 2D one often gets confused with the subjacent "thickness". That is why I prefer to work as far as possible with calculated values (through boundary integrals) and I use the application mode density rho_... instead of mass for the gravitational field forces. This is the nice way of COMSOL you might add some physics with just some formulas here and there, not that easy to do in other FEM programmes. One thing tough If you start to add masses on points, check carefully that these are taken into account if you use the eigenmode analysis, you might well need to multiply them by -(jomega)^2 or a lambda^2 depending on how you define the masses and the application mode Good luck Ivar


Dobromir Filip
Posted: 7 years ago Jun 17, 2010, 4:23 PM EDT
Hi Mario,

I am trying to simulate two concentric springs that share the same deflection axis (connected at both ends) in COMSOL. Would you happen to know where I can find tutorial for a spring simulation in COMSOL. Thank you
Hi Mario, I am trying to simulate two concentric springs that share the same deflection axis (connected at both ends) in COMSOL. Would you happen to know where I can find tutorial for a spring simulation in COMSOL. Thank you

Giorgio Calanni
Posted: 7 years ago Sep 17, 2010, 3:06 PM EDT
Ivar,




And why do you have two blocks? and not 1 only? I am a bit confused.



why nothing happens when you set Gravity = 0 ? Is that the external load you are applying? Correct? [Nevermind, I found the formula you used]
Ivar, And why do you have two blocks? and not 1 only? I am a bit confused. why nothing happens when you set Gravity = 0 ? Is that the external load you are applying? Correct? [Nevermind, I found the formula you used]

Posted: 5 years ago May 25, 2012, 4:31 AM EDT

Do you care about stresses on the spring/damper? Or are you just looking for the displacements and speeds and the like?

If it's only displacements you're interested in, it might be much easier to model the spring/damper combo as equations in a transient analysis:

1.- Model a square with the characteristics you're looking for.
2.- add a restriction limiting to zero movement in the x axis.
3.- Under integration coupling variables, add two point variables, choosing the spring contact point: One for displacement (expression "v") and speed (expression "vt").
4.- Add a spring force -k*displacement where your spring connects to the piece.
5.- Add the damper force -c*speed where your damper connects to the piece.
6.- Add some sort of excitation force (a time-stepped function) F*(t<te) or whatever jolt to destabilize your system.

That should give you an answer. For reference, I'm uploading a very simple model of a .5m x .5m x .5m block. I added all the forces in the subdomain section, but you can play around. The model's settings can be changed under the constants menu, where k, c, F, m and te can be modified.

You can plot speed and displacement using the variables damp and disp, respectively.


Hello Mario Elenes,
i'm trying to simulate a spring/damper in 3d in order to get information about the displacements and speeds. I tried to tranfer your 2d-Model into 3d, but it doesn't work. The stresses are zero und over time there is no displacement shown in the graphic interface.
Any idea on what i'm doing wrong or how can i modell a spring/damper system in 3d?

[QUOTE] Do you care about stresses on the spring/damper? Or are you just looking for the displacements and speeds and the like? If it's only displacements you're interested in, it might be much easier to model the spring/damper combo as equations in a transient analysis: 1.- Model a square with the characteristics you're looking for. 2.- add a restriction limiting to zero movement in the x axis. 3.- Under integration coupling variables, add two point variables, choosing the spring contact point: One for displacement (expression "v") and speed (expression "vt"). 4.- Add a spring force -k*displacement where your spring connects to the piece. 5.- Add the damper force -c*speed where your damper connects to the piece. 6.- Add some sort of excitation force (a time-stepped function) F*(t

Posted: 5 years ago Apr 21, 2013, 4:29 PM EDT
Hi

I am trying to model the 3 element Hill Muscle Model in COMSOL fo muscle fibers.. I am a bit confused on how to use the PDE coefficient form in order to accomplish this. I would like to have a single step function response for the force and would like to study the displacement for the muscle fibers.

Best,
Shoaib
Hi I am trying to model the 3 element Hill Muscle Model in COMSOL fo muscle fibers.. I am a bit confused on how to use the PDE coefficient form in order to accomplish this. I would like to have a single step function response for the force and would like to study the displacement for the muscle fibers. Best, Shoaib

Posted: 4 years ago Apr 22, 2013, 2:50 AM EDT
Hi

if you have the "solid" module, make a simple model with two adjacent domains (a rectangle in 2D with a layer at half way) and add a thin elastic layer in between the two blocks, then fix one of the external boundaries (not adjacent to your thin elastic layer, add some useful boundary load on the opposite side.
NB: start to give it a zero stiffness in the direction tangential to the boundary and add a roller condition on the side of the block and study first only the effects in the normal directions

Then check the doc/help, try out the different combinations of spring and damper settings of this special BC, and particularly look at the underlying equations (turn ON equation view under "options - preferences - show"

Then study how COMSOL treats these equations, this would give you the solution for how to set up your own PDE if you prefer this way

I have attached a simple 4.3a model, as well as the same for a spring foundation (perhaps simpler to catch.) Once you understand the difference in modal response of these two and how it is implemented, you can easily model any kind of non linear force displacement system ;)

--
Good luck
Ivar
Hi if you have the "solid" module, make a simple model with two adjacent domains (a rectangle in 2D with a layer at half way) and add a thin elastic layer in between the two blocks, then fix one of the external boundaries (not adjacent to your thin elastic layer, add some useful boundary load on the opposite side. NB: start to give it a zero stiffness in the direction tangential to the boundary and add a roller condition on the side of the block and study first only the effects in the normal directions Then check the doc/help, try out the different combinations of spring and damper settings of this special BC, and particularly look at the underlying equations (turn ON equation view under "options - preferences - show" Then study how COMSOL treats these equations, this would give you the solution for how to set up your own PDE if you prefer this way I have attached a simple 4.3a model, as well as the same for a spring foundation (perhaps simpler to catch.) Once you understand the difference in modal response of these two and how it is implemented, you can easily model any kind of non linear force displacement system ;) -- Good luck Ivar


Posted: 4 years ago Apr 23, 2013, 11:50 PM EDT
Hi Ivar,

I have tried modeling the 3 element Hill Model for muscle fibers (the image is attached below). I am using a simple geometry of a muscle fiber as a cylinder. I was able to set damping and spring constants to the geometry, but I am unable to ensure that this is actually a spring and a damper in parallel. For the series spring I was thinking of adding a thin elastic layer to represent this. I am trying to show correlation between a fixed force and displacement over a period of time. I have attached an image of the equation used for the Force, although I plan to set the force to a constant 500N. I am unsure about how to represent this COMSOL. The displacement seen in my model seems incorrect. Please Advise any input.

Best,
Shoaib Ahmed
Hi Ivar, I have tried modeling the 3 element Hill Model for muscle fibers (the image is attached below). I am using a simple geometry of a muscle fiber as a cylinder. I was able to set damping and spring constants to the geometry, but I am unable to ensure that this is actually a spring and a damper in parallel. For the series spring I was thinking of adding a thin elastic layer to represent this. I am trying to show correlation between a fixed force and displacement over a period of time. I have attached an image of the equation used for the Force, although I plan to set the force to a constant 500N. I am unsure about how to represent this COMSOL. The displacement seen in my model seems incorrect. Please Advise any input. Best, Shoaib Ahmed


Posted: 4 years ago Apr 24, 2013, 5:39 AM EDT
Hi

spring and damper to a fixed reference is the "Spring" BC, thin layer physics really only appear between two domains. In you case the spring and damper is also distributed in the domain, so its whatever lienar or non linear structural physics to which you ad damping via a sub node

--
Good luck
Ivar
Hi spring and damper to a fixed reference is the "Spring" BC, thin layer physics really only appear between two domains. In you case the spring and damper is also distributed in the domain, so its whatever lienar or non linear structural physics to which you ad damping via a sub node -- Good luck Ivar

Chien-Hung Lin
Posted: 4 years ago Jun 5, 2013, 10:01 AM EDT
Hi Ivar

I am interesting in your "solid" module. But I can't open 4.3a model, the files of ThinElasticLayerTst1.mph and SpringFoundationTsts1.mph could transfer to 3.4 model.

Best,
Chien-Hung Lin
Hi Ivar I am interesting in your "solid" module. But I can't open 4.3a model, the files of ThinElasticLayerTst1.mph and SpringFoundationTsts1.mph could transfer to 3.4 model. Best, Chien-Hung Lin

Maizuar Maizuar
Posted: 2 years ago Jan 26, 2016, 2:52 AM EST
Dear Ivar

I would like to model a prestressed concrete bridge beam. The beam is supported by elastomeric bearings (Laminated Rubber Bearing LRB, 350 x 280 mm) . Since the stiffness of elastomeric bearings have been specified in bridge design code as follow:

Calculated compressive stiffness at zero shear = 168000 kN/m
Mean shear stiffness = 810 kN/m
Calculated rotational stiffness = 307 kNm/rad
Shear deflection capacity = 42 mm

So, I want to know whether it is correct to model the elastomeric bearings as "SPRING FOUNDATION" or THIN ELASTIC LAYER? If yes, how do I assign the above parameters.

Many thank for your advice and assistance. I also attached the file for you to see.


Kind regards


Maizuar
Dear Ivar I would like to model a prestressed concrete bridge beam. The beam is supported by elastomeric bearings (Laminated Rubber Bearing LRB, 350 x 280 mm) . Since the stiffness of elastomeric bearings have been specified in bridge design code as follow: Calculated compressive stiffness at zero shear = 168000 kN/m Mean shear stiffness = 810 kN/m Calculated rotational stiffness = 307 kNm/rad Shear deflection capacity = 42 mm So, I want to know whether it is correct to model the elastomeric bearings as "SPRING FOUNDATION" or THIN ELASTIC LAYER? If yes, how do I assign the above parameters. Many thank for your advice and assistance. I also attached the file for you to see. Kind regards Maizuar

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.