Yade

How to get the induced Stress/strain in DSR test?

Asked by behzad

Hi guys,

I'm currently working on simulation of Dynamic Shear Rheometer testing. You can check the test configuration here:
http://www.pavementinteractive.org/article/dynamic-shear-rheometer/

This is a test to measure the viscoelastic properties of materials. Parameters like Complex Shear modulus and phase angle are measured in this test.
In this test, the material is like a disk which is sandwiched between two plates. The upper plate has a sinusoidal rotation.

An oscillatory shear is applied to the material by the upper plate and the induced stress is measured. The complex shear modulus is calculated as:

G*=Stress_amplitude/Strain_amplitude.

The question is; in the numerical simulation, how do we calculate the induced shear stress?

Can we use getStress() function and then get the xy components?

Or we calculate the shear force on the upper_plate?

If you have any experience in such a model, I appreciate if you can share your ideas.

Thanks,
Behzad

Question information

Language:
English Edit question
Status:
Solved
For:
Yade Edit question
Assignee:
No assignee Edit question
Solved by:
Jan Stránský
Solved:
Last query:
Last reply:
Revision history for this message
Jan Stránský (honzik) said : 		#1

Hi Behzad,

> Can we use getStress() function and then get the xy components?
>

I would say yes, with correct volume argument. Alternatively you can
compare it with the average of bodyStressTensors,

>
> Or we calculate the shear force on the upper_plate?
>

this is another way to check the results. In static case (and considering
correct volume/area), all three methods should lead to comparable results.
In the dynamic case it would depend on how much it is dynamic :-)

cheers
Jan

PS: have you tried your model in some simple scenario, like uniaxial
tension? It might happen, that on interaction level the model would behave
according to Burger's model, but macroscopically it could be different..

Revision history for this message
behzad (behzad-majidi) said : 		#2

Hi Jan,

1) How can we get the sigma xy in the sample?

2) "have you tried your model in some simple scenario, like uniaxial
tension? It might happen, that on interaction level the model would behave
according to Burger's model, but macroscopically it could be different..."

Yes, I tried and compared the microscopic and macroscopic behaviors. Both were dynamic by the way.
In the microscopic level (the response of two contacting spheres) is what expected by theory of Burger's model. But, the many-spheres system is strange. I was thinking perhaps there's a time-step problem

Revision history for this message
Best Jan Stránský (honzik) said : 		#3

>
>
>
> 1) How can we get the sigma xy in the sample?
>

e.g.

s = getStress(volumeOfTheDisc)
sxy = s[0,1]

>
> 2) "have you tried your model in some simple scenario, like uniaxial
> tension? It might happen, that on interaction level the model would behave
> according to Burger's model, but macroscopically it could be different..."
>
> Yes, I tried and compared the microscopic and macroscopic behaviors. Both
> were dynamic by the way.
> In the microscopic level (the response of two contacting spheres) is what
> expected by theory of Burger's model. But, the many-spheres system is
> strange. I was thinking perhaps there's a time-step problem
>

then try to decrease the time step. What I meant is the interplay between
normal and shear components. Even in the easy case of perfectly linear
elastic behavior you would get "strange" stiffness depending non-trivially
on normal and shear interaction stiffness. If you have additional dampers,
I would expect the macroscopic damping to behave similarly..

Jan

Revision history for this message
behzad (behzad-majidi) said : 		#4

Thanks Jan Stránský, that solved my question.