[Binoy]

Hello everyone,

I am new to CFX and I am modelling a submarine slope where the soil is clayey soil. The soil is modeled as a non-Newtonian fluid and I wish to vary the undrained shear strength of clay soil.
The dynamic viscosity is expressed as: Undrained shear strength/ Strain rate
Undrained shear strength= 30000 [Pa]
Strain Rate = if (clay. Shear Strain Rate< 0.001[s^-1], 0.001[s^-1], clay. Shear Strain Rate)

(1) When I use shear strength= 30000 Pa, it works. But when I use shear strength greater than this (35000 Pa, 40000 Pa, 50000 Pa) then it gives an error.
'Run Fluid Flow CFX 001': No data source for Workspace, cannot monitor Solver data.
The error message is:
| ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Floating point exception: Overflow
ERROR #001100279 has occurred in subroutine ErrAction. |
| Message: |
| Stopped in routine FPX: C_FPX_HANDLER
An error has occurred in cfx5solve: |
| |
| The ANSYS CFX solver exited with return code 1. No results file |
| has been created.

(2) And another problem is:
When I use a constant shear strength of 30000 Pa, then theoretically Pressure (vertical stress) not supposed to change at different time step. Because shear strength is constant and I do not apply any external load. Also, I use no slip wall in all boundary. But here it changed at first 5 or 6 time step then it becomes more or less constant.
Pressure is expressed as:
downPress = (denClay-denWater) *(g)*VF1*(Yc-y)
Where, VF1 = if(t>0.1[s], clay.Volume Fraction, downVFClay) and
Yc= slope height

I am sorry for my grammatical and spelling mistakes.

Hopefully, someone can help me.

Best regards,
Binoy

[ghorrocks]

Your viscosity model is highly numerically unstable. Having step functions in the viscosity is always going to cause huge convergence issues. I recommend you look at the built-in non-Newtonian models and select one of them which is close to your application and adjust the model constants so that it gives a similar behaviour to your idealised step function model.

[Binoy]

Hi Glenn,
Thanks for your answer.

You recommended me to use built-in non-Newtonian models. But in my modeling I need to vary the strength by using sensitivity i.e., strength will reduce with strain (strength degradation). And as far I know the built-in models only applicable for constant shear strength. What then can I do?

Also please give me suggestion about the second problem.

Thanks in advance.
Binoy

[ghorrocks]

If you need a variable shear strength then I suggest you replace the step function with a continuous function and blur the transition over a physically realistic space. An example of doing this using the tanh function is in the CFX documentation for blurring free surface model initial conditions - Modelling Guide, section 7.18.4.3

Can you post an image of what you are modelling and the output file?

[Binoy]

Hi Glenn,

I am sorry. I can't understand. Can you please explain again about the blurring function? And where I have to use this function? In Strain rate?

Here I have attached an image and output file for constant Shear strength of 30 kPa.

[ghorrocks]

Your function Mu = Su/Strain Rate and Strain Rate is capped at a minimum value of 0.001 is a discontinuous function. Things will improve significantly if you can replace this with a continuous function which smooths the transition from the 0.001 line to the SSR = clay.SSR line. I will leave it up to your imagination to dream up a function here, but my suggestion of a tanh function does not appear suitable.

[Binoy]

Thanks Glenn for your suggestion. I will try to do this.

[Binoy]

Hello Glenn,

Please give me suggestion about the continuous function which I can use here for the smooth transition from the 0.001 line to the SSR = clay.SSR line. I tried this but I couldn't do it.

Thanks in advance.
Binoy

[ghorrocks]

There are lots of ways.

Simple ones like define a linear function (call it s) which is 0.0 below 0.001, 1.0 above 0.002 and linearly varies between 0.001 and 0.002.

Then if you set mu = s*(Su/Strain Rate) + (s-1)*0.001 then you have a function which is still a little discontinuous, but much better than your function.

The function s can be made continuous by making it a tanh function. Then it will be continuous for all values. But I will let you work out how to configure a tanh function for this case.