Learning Forum

[iainb]

Hi all, I have started a simulation in which I am testing the erosion rate of certain materials. The set up of this simulation is fairly basic with a sphere, to represent a raindrop, modelled above a plate to represent the target material. The sphere has the material property of water pre-defined in Ansys and I have inputted a shock EOS. I am using the pre-defined Epoxy Res material in the Ansys library for the plate material. The model is setup with a Lagrangian/Euler approach with the Sphere modelled with Euler and the plate modelled with Lagrangian. The reason I chose this was because of the expected high deformation with the raindrop upon impact. The simulation was set up with an initial velocity of 140m/s and a fixed support on the underside of the plate. The Euler domain is set to the sphere only.

The simulation runs fine but upon output for stress, strain and deformation there is 0 for all. When I play the simulation what I see is the sphere disappearing almost instantly. I am wondering if anyone has any insight as to why this is happening? Initially, I believe that the Euler domain is causing the sphere to disappear by stopping the sphere from falling. All of the faces of the domain are set to flow out. I haven't used this software for a while and appreciate any help you can give me.

Thanks

Iain

Edit: I have just run the simulation with the domain over all bodies and gotten some results. The thing is I'm attempting to replicate results so the domain needs to be over just the sphere.

[peteroznewman]

Hi Iain 

A quick Google search on "how fast do raindrops fall" says the terminal velocity is about 10 m/s, and you know the Internet is never wrong ; )   I'm curious why you chose 140 m/s.

It sounds like you may be using Explicit Dynamics, but Fluent also has the features you describe. Please confirm which solver and which release of ANSYS you are using.

Regards

Peter

 

[iainb]

Hi Peter

Thanks for replying. Firstly I suppose I should mention that this is a basic setup to solve raindrops impacting a wind turbine blade which can go as fast as 140 m/s. The simulation is about material response so no point in modelling for the whole blade

I am using Explicit Dynamics. I am using Ansys Student 18.2. I am hoping to have access to the full version of Ansys soon but I'm not sure it's the student version that's restricting me and more something in my settings.

Regards

Iain

[peteroznewman]

Hi Iain,

In Analysis Settings, scroll down to  Erosion Controls, did you set No to On Geometric Strain Limit?
In Output Controls, did you set a high Number of points for output like 2000 instead of the default 20?

 

If these don't help, you can create an archive file and attach the .wbpz file to your reply.

Regards

Peter

 

[iainb]

Hi Peter

I will try your suggestions. The simulation is running now, I will let you know how it goes.

Regards

Iain

[iainb]

Hi Peter

It didn't work I'm afraid. Attached is the archived file. When I simulated the motion it looked like the sphere is impacting upon the euler domain lower face.

Thanks,

Iain

[peteroznewman]

Hi Iain,

Why did you choose to give velocity to the drop and hold the plate fixed?

Why not let the drop have zero velocity and no boundary condition, and have the plate have the 140 m/s velocity to move through the space occupied by the drop?

What is the diameter of sphere you are using? When raindrops are falling, air pressure causes them to flatten out from a spherical shape.

A face travelling at 140 m/s will cover 4 mm in about 3e-5 seconds.  What is the End Time for your simulation?

Regards

Peter

 

[iainb]

Hi Peter,

As I said I am mimicking prior work in an attempt to recreate a proven model that I can then use to adapt to different parameters. I fixed the plate and dropped the sphere as I that is what the previous model contained. 

I am using a 3mm diameter sphere. I realise that the raindrop may flatten out. This is ideal conditions. I have applied standard earth gravity to the model so maybe the analysis will account for that?

The simulation is running for 2e-3 I believe

Regards

Iain

[peteroznewman]

Hi Iain,

It takes 2e-5 seconds for an object moving at 140 m/s to travel 3 mm.

If a simulation runs for 2e-3 seconds and you request 2000 frames, each frame is 1e-6 seconds, so the splat will cover 20 frames and the next 1980 frames is what happens next. That seems a bit of a waste of time.

I recommend you use an end time of 2e-4 so you see the effect of 10 drop diameters of time elapse and the splat will cover 200 frames and the next 1800 frames is what happens in the next 9 drop diameters. That is a bit more balanced, but you could also cut the end time in half again and get 400 frames of the splat.

Regards

Peter

[iainb]

Hi Peter

I understand that and will fix it, thank you.

What about the Euler domain infringing on the sphere though. That is what's causing the issue. Every time I use it surrounding the sphere the domain causes the sphere to splat. Any help on this?

Thanks,

Iain

[peteroznewman]

Hi Iain,

I have little experience using an Euler domain in Explicit Dynamics, but I think you have to have a large block of air surrounding the spherical fluid.  Most Fluent models are Euler domain.

What happens if you leave the drop of water as Lagrangian?

I found a site that will sell you an ANSYS model of a bottle drop with fluid inside for $13 to their PayPal account. I'm not recommending that, just pointing it out.

https://www.youtube.com/watch?v=-nqN-31-xt0

 

Regards,

Peter

 

[iainb]

Hi Peter

I figured out the problem. Was pretty simple in hindsight. Because the Euler body could only flow in a Euler Environment, I had to simply extend the domain further past the impact point. Thanks for all your help.

Best Regards

Iain