Learning Forum

[jonatassilvajunior]

Hi, I am a novice user in  ANSYS. I am studying mechanical of the fracture computational  and I would like of a help. Someone knows like inserting a difusion effect of hydrogen in solids materials?

Explaining my study:

I have a vessel with hydrogen pressured, in a point has a crack. I inserted a semieliptical crack (actually, I would like add a embeebed crack, but I don't know how do). In this moment, I would like inserting in the plate of vessel the effect of migration of hydrogen. 

Someone has a tutorial? (video classes would be great, but anything goes).

[Michele De Giosa]

Have you a solution for this?

[Rob]

On the fluids side, we assume there's no flow into the solids. However, with a diffusion coefficient and knowledge of the maths it should be possible to code up a scalar UDF in Fluent. Does the hydrogen just alter the strength related material properties or does it also enhance hydrogen diffusion once the solid has been "wetted"? 

You'll need to post in the Mechanical section for cracks etc. 

[Michele De Giosa]

Hi Rob. 

A question. I'm trying to study hyd. diffusion in a superalloys.
So i'm using also Multiscale.Sim for doing a steps like: model microstructure (I chose the polycrystalline model), homogenization, macrostructure and localization (last for seeing the behaviour of hyd. in grain boundaries). 
Now, after these steps.. Is this set of steps sufficient, and then simulate the diffusion with Fluent or Transient Thermal, or do you recommend some other step that I am missing? Thx for help.

[Rob]

In Fluent we can look at hydrogen flow/pressure on the solid surface. If it's a fairly static system with on hydrogen we can probably assume values. CFD is not suitable for hydrogen penetration into the crystal structure: you're into molecular flow at that scale and not continuum physics. 

I'll ask a colleague to have a look: my structural knowledge is limited as it was assumed we could get a Mechanical Engineer to do all of that for us when I was at Uni. 

[Michele De Giosa]

Hi Rob,

thanks for your time and for your suggestion. I have values for diffusivity, intra and intergranular, for simulating this behaviour. So, my only issue is to create a geoemtry, as explained above.
I'll wait your collegue for others suggestions.. Have a good day :)

[Michele De Giosa]

Are there some news? 

[Rob]

Apologies, I asked and then didn't reply. As mentioned the CFD codes can't do the flow, but could be modified to allow diffusion based on theoretical/empirical correlations. We're fairly sure Mechanical can't do the diffusion, but it may be possible to script something in much the same way as we would in Fluent. Either way you're going to need to add most of the models yourself. 

[Michele De Giosa]

Ok thanks Rob, now i can continue my work on the path I had set out on.

[Sheldon Imaoka]

Diffusion can be solved in Mechanical APDL - please see the "Coupled Field Analysis Guide" under the Mechanical APDL Help documentation.  There are different types of coupling available, such as structural-diffusion, meaning that the concentration can cause 'diffusion strain'.

I don't know if in your application hydrogen concentration causes any noticeable diffusion strain, but I just mention that such coupling capabilities are available, but you can also solve diffusion problems directly, too.  However, you could post in Structures forum as Rob mentioned for more details/help.