Defining a stiffness of a spring damper element through a curve, any help?

Altair Forum User

Hello there,

 

I need some advice for defining the stiffness of a spring damper element with a curve. My task is to model a 1 mass oscillator with a non-linear stiffness. But first of all I want to understand how to define a stiffness through a curve. So I tried to model a linear stiffness through the curve and verify it with the regular linear simulation.

 

My questions are:

1. Can anybody tell me what I have done wrong with my curve definition?
2. I defined a sinusodial force with +/- 10kN but the mass is just swinging in the negative area?
3. Can anybody specify what the independent variable is? What do I have to put there when my curve is defined as force over displacement?

 

I have attached my files.

 

I appreciate every help /emoticons/default_smile.png' srcset='/emoticons/smile@2x.png 2x' title=':)' width='20'>

 

 

 

 

Altair Forum User

Hi @Peter99

 

Thank you for sharing the file. 

 

The results are very strange. Can you let me the units used in your model and also for the curve?

Praful

Altair Forum User said:

I have attached my files.

Hi @Peter99 - Somehow, I am not able to see the files that you have attached (if at all they are attached)

 

Speaking in general - When the spring damper is linear, one provides a constant value as stiffness which is the slope of the Force - Deflection curve. When defining it as a non-linear entity, now the spring damper actually works as a action reaction force (single component). So the curve should represent Force deflection curve. If representing a linear stiffness through the curve, this would be a linearly increasing curve with a constant slope.

The independent variable should  represent the deflection of the spring. By default, MV sets the independent variable as {sd_0.DM}, where sd_0 is the variable name of the spring.

This results in the function DM(i, j) which tracks the magnitude of the distance of the 2 end markers of the spring. Note that this function does not give you the exact  deflection in the spring. To get that one has to subtract the initial length of the spring. (This is assuming that the force-deflection curve represents the deflection and not the force v/s spring length)

 

Attached here is an example model that should help you understand how this works. You may chose to turn off the damper part by setting it to Linear = 0.0

 

Regards

Praful

Unable to find an attachment - read this blog

Altair Forum User

Hey,

 

sorry for the late response.

 

Altair Forum User said:

Somehow, I am not able to see the files that you have attached (if at all they are attached)

 

Something went wrong I uploaded it again. (1mass-oscillator)

 

Altair Forum User said:

Attached here is an example model that should help you understand how this works. You may chose to turn off the damper part by setting it to Linear = 0.0

 

Thank you for your example it helped a lot.

 

Altair Forum User said:

This results in the function DM(i, j) which tracks the magnitude of the distance of the 2 end markers of the spring. Note that this function does not give you the exact  deflection in the spring. To get that one has to subtract the initial length of the spring.

 

This advice also helped a lot.

 

Altair Forum User said:

When the spring damper is linear, one provides a constant value as stiffness which is the slope of the Force - Deflection curve.

 

I think I got this right.The right picture desrcibes the slope of the force-deflection curve. 

 

 

Altair Forum User said:

When defining it as a non-linear entity, now the spring damper actually works as a action reaction force (single component). So the curve should represent Force deflection curve.

 

I didn't understand this part for the non-linear case. How do I need to define my curve for the non-linear case? Same as the linear case?

 

I have attached my current files. (Nicht_linear_Steifigkeit3.rar)

 

Thank you 

Unable to find an attachment - read this blog