Learning Forum

[a.moussa]

Dear All,

I am trying to model a pile embedded in a soil. The applied loads are static concentrated loads at the pile cap. Could anyone provide me with the general modeling procedure for creating the soil model in Ansys. Thank you in advance for your kind help and cooperation.  

[peteroznewman]

Hello Moussa,

I am not an expert in soil modeling, but I did find a few articles on the Internet (1) (2).

What results do you want from your model?

Are the static loads on the pile cap vertically downward?

Are you modeling a single pile or a group of piles connected to a cap?

Are the piles supported at the bottom on bedrock or only soil?

Are you going to have different layers of soil properties?

Is the cap above the soil or in the soil?

Are you on a Student license or a full license? The Student license has a limit of < 32,000 nodes+elements which will make it difficult to build a continuum model, but you can easily build a beam model with nonlinear springs to represent the soil, however, that was in paper (1) that was studying earthquake ground motion causing inertial side loads on the structure.

The procedure if you use beam models is to create a beam for the pile with the correct cross-section and material properties and attach springs. The procedure if you use a continuum model is you create a block and a pile, subtract the pile from the block, share topology and assign materials to the pile and soil.

If you go with a continuum model, ANSYS provides a list of Geomechanical material models, some of which are appropriate for soil modeling.

 

 

 

[Rana Nasser]

Hi,

I'm using ansys 15.0 and I can't find the "Geomechanical" material models in the engineering data toolbox, any help?

this is the engineering data toolbox in my version

[peteroznewman]

It looks like there are a few variations of Drucker-Prager, which has been used to model soil in the literature.

[Rana Nasser]

 but I tried to use the linear drucker - Prager model it gives me the ability to add " yeild stress at zeto pressure and slope in degrees" only, the other models in the "brittle/granular" material models didn't give me what I need also. I need to add the bearing capacity, the unit weight of the soil, c, and angle of friction!, so what is the material model can I use?

 

[Rana Nasser]

Their is an other question, I will use this material in a transient analysis simulation and I didn't fined the " brittle/granular" material models in the engineering data of the transient structural, so I added separated engineering data from the tool box to define the soil and linked it with the engineering data of the transient structural. Is there any negative effect of this?!

 

[peteroznewman]

Engineering Data is configured as a Toolbox. When you create a new material name, it has no properties. To do Static Structural (without Inertial loads), you only need Elasticity, so you add Isotropic Elasticity and provide Young's Modulus and Poisson's Ratio.  Now you want to do Transient Structural, you have to add Density, which you could use for the unit weight of the soil. Now you want to do Elastic-Plastic behavior, you add Bilinear Plasticity and provide the Yield Strength and Tangent Modulus. If there are properties in a material that are not needed by the analysis, it doesn't hurt if they are defined, they are just not used.

ANSYS keeps track of which properties are compatible and deactivates an old one if a new, incompatible property is added. For example, if you add Orthotropic Elasticity to a material that already has Isotropic Elasticity, the Isotropic Elasticity will me marked with a red x. You could delete that property.

ANSYS keeps track of which properties are needed for an analysis, so if you have finished a Static Structural analysis with a material that has no density, and drop a Transient Structural on that Model item, you will get a ? in the Setup cell because of the missing material property. If you try to solve, it will give you a message saying the density property is missing.

 

[Rana Nasser]

Peter I can't find enough words to express how I'm thankful for your continuous help!!  

I want to make sure that I got you, you mean that there is no problem in adding the a separated engineering data tool box to add material properties which not supported in the transient structural engineering data tool box because ansys will deal with the material properties which are needed in the analysis and ignore the others and if there are two opposite properties it will alarm me and stop solving.

** actually I have used the density property to represent the soil unit weight comes from the test by dividing gama by 10 and adding the result in the density cell, but I thought that if there are material models prepared in ansys to represent soil material it will include explicit gama.      

[peteroznewman]

Rana, you are welcome!

You understand correctly, you can't put incompatible material models together since Engineering Data will disable one before you even start the solver. Once you have built up a good soil model, you can save that material definition and reuse it in many models.

I don't know anything about soil models or what gama is and why you would divide by 10, but density is simply mass/volume, in the right units.

Here is what Drucker-Prager looks like in ANSYS 18.2...

[Nayef Al Haddid]

Hello everyone, I'm a newbie to ANSYS and currently using it on my dissertation to run static structural analysis on a 2-span unreinforced masonry arch bridge (using a macro-modelling approach), I've been trying to produce a soil material to model the fill inside the bridge, I've used a mohr-coulomb model (see picture) and frictional contact between the fill and the spandrel walls and arche barrels of the bridge. Just wanted to get some insight if there's a better way of doing this or if I am going about it the wrong way???

 

Thanks

 

[Rana Nasser]

Hi Nayef,

I think you are on the right way, but I have used orthotropic elasticity instead of the isotropic elasticity to give me the ability to input the tension and compression capacity in the horizontal directions different from the vertical direction.

 

I wish this can help you!

[Rana Nasser]

Hi everyone! 

Is there any way to add friction angle and cohesion to a material model in ansys workbench ver. 15 or the only way to do this in this version is using APDL? 

*By the way the load that subjected to the soil will not cause soil failure so I think I don't need to use "Drucker - prager" failure model.  

There is another request, If anyone have papers or any reference in modeling soil in ansys "workbench" please share it!

Regards!!

[yasser1987]

Dear All

 

i want to model a soil under  tripod bucket offshore foundation  and getting the displacement and rotation by applying  wave load , the soil should be modeled with mohr Colomb theory and it is sandy clay, but unfortunately i do not know how to do it , could you please someone help and guide me, i mean if there is any tutorial or sample project could please introduce to me 

actually i do not know which steps must be followed on ansys workbench19.1 in order to simulate the soil,i?step by step approach 

many thanks  

[peteroznewman]

Yasser's post here is a duplicate of this discussion that includes a diagram.

[shiv]

Hi everyone! 

Is there any way to add friction angle and cohesion to a material model in ansys workbench ver. 15 or the only way to do this in this version is using APDL? 

*By the way the load that subjected to the soil will not cause soil failure so I think I don't need to use "Drucker - prager" failure model.  

There is another request, If anyone have papers or any reference in modeling soil in ansys "workbench" please share it!

Regards!!

 

I think the best possible soil model that is appropriate in your case will be Mohr-Coulomb soil model.

If you want to access Mohr Coulomb model then only possible way is to call in the material model using APDL Macros in ANSYS 15.0.

Where you can add the its friction angle, cohesion and density as unit weight. And while solving treat the problem as Non-Linear with enough iterations and substeps. 

[shiv]

Adding to my previous comment, use SOLID65(Considering the model you are trying to execute is brittle in nature) element for the model with Mohr-Coulomb material property for easy convergence and better accuracy. 

[ali.ismail]

Does anyone know how to solve a model using mohr-coulomb material for soil with a very low cohesion without getting the problem of "The solver engine was unable to converge on a solution"? 

I will be very grateful if anyone tells me what to do asap