Learning Forum

[Aswin]

Hi,

I am trying to perform the viscoelastic analysis of asphalt pavement. I inputted the data for prony shear relaxation and william Landel ferry shift function. I got these data from the literature. I also performed the curve fit. I have also defined the isotropic elastic property of the materials. But when I solve the result, there is no difference between isotropic elastic and viscoelastic solution. Meaning I get same result when I remove the viscoelastic parameter. I am not sure if I need to do anything in the Model  like update the analysis setting or anything else once I define the viscoelastic properties in the Engineering data. 

 

Thanks,

Aswin Karki

[Aswin]

I am new to ANSY and need to finish my college project by end of the week. I will appreciate any suggestion or comments.

[peteroznewman]

Did you assign the material to the body?  That is a common novice user mistake.

[Aswin]

Yes, materials has been assigned to the body. It solves perfectly for the inelastic case but the results doesnot change anything when I solve for viscoelasticity case.

[peteroznewman]

Read this post. It describes fitting the Prony series to your data.

[Aswin]

Thanks Peter. I had read this earlier  and had  fitted my curve. The issue may be the STUDENT VERSION the software that I am using. I have attached the model for the reference. I would really appreciate if you can take a look at the attached model.

Thanks.

[peteroznewman]

Aswin,

You have used a Mohr-Coulomb material model in your analysis. That material model defines yielding when the shear stress on any plane in the material reaches this criterion: 

However, you have set the value of cohesion at zero.  Please define a non-zero cohesion shear strength.  Why are you even including a Mohr-Coulomb model in the first place?  I thought you were interested in the Viscoelastic behavior of the asphalt.

You should also slice the geometry through the center of the bodies and set up a 1/4 symmetry model so that the bottom can have a Y=0 constraint instead of a X,Y,Z = 0 fixed constraint which is over constraining that bottom face.

If this is a stack of soil layers, shouldn't there be a gravity load to provide the down force on the mass to create a pressure as a function of depth? And shouldn't the cut faces have a pressure that increases from zero at the surface to larger values as a function of depth to balance the pressure created by the gravity force?

You should use Shared Topology in the Geometry editor, SpaceClaim.  Then you can delete the Contact definitions between the layers.

Your model issues have nothing to do with using a Student License, which only becomes an issue when you have too many nodes or elements in your model and that will be clearly indicated in the error message when the solver refuses to run.

[ali.ismail]

Hello, I'm using a student version of Ansys workbench and I have the problem of "The solver engine was unable to converge on a solution" while using Mohr-Coulomb to define a soil with very low cohesion. I will be very grateful if you can tell me what to do to solve that asap.

I can send you also screens of the boundary conditions and mesh despite I don't think they are the problem 

[Aswin]

Thank you Peter for your detailed response. I am trying to model the viscoelastic behavior of asphalt surface and stress/strain dependent nonlinear behavior of granular base and subbase. The pavement have 5 different soil layers: asphalt, base, subbase, compacted subgrade and subgrade(native soil). This is why I created separate discussion topics" one for viscoelastic and another for nonlinear. I can continue to discuss both on this one if you prefer. 

I agree with your definition of Mohr column. The reason for assigning the c as 0 is that the base and subbase are granular materials which are cohesionless and only have internal friction. The mohr column defines the nonlinear behavior of the base and sub base layer upto the residual internal friction of phi/2.

About the boundary condition, the assumption is that the soil layer at a certain depth is assumed fixed with no impact due to traffic load. I ran KENPAVE software for the linear elastic behavior ( no visco elastic and nonlinear case: meaning no mohr column and prony series parameters of soil layers) which is specially designed for pavement analysis by inputting parameters as ANSYS model  in order to validate the ANSYS model and all the values for stress, strain and deflections including the deflection basin  are very much comparable. I used other literature where they have used the same parameter. I will recreate the model as per your suggestion and compare the difference. 

I agree with your effect of soil own weight and lateral pressure. The reason it is not included in the model is I am only analysing the effect of truck/traffic load on the pavement. The load from the tyre of the truck is defined by the contact stress using a circular pressure load as you can see in the model.  I would have included the vertical and lateral soil load effect if I was analysing the foundation or retaining wall.

I recreated the model with 2 simpler models: model 1 with only one layer of soil(asphalt layer)  and model 2 with only base layer . I ran two cases: linear elastic and viscoelastic on model 1 and two cases: linear elastic and non linear analysis on model 2 but I am not getting any different values. I did these using the static structural and transient structural analysis method but I got same results. 

At this moment, I  feel like I am missing something in the model menu. 

 

Thank you.

 

 

[Aswin]

One another thing I noticed on the prony series table I created is, the cure fit gets removed everytime i close the workbench and I will have to input it again when I open it next time. Not sure if this is common or just me.

Thanks.

[Rahulreddy96]

Hi,

I'm a beginner in this tool and performing an analysis of flexible pavements, for the modelling of asphalt concrete layer what is the importance of giving the shear data and Prony series tables. What is the source for the data?