Viscous Drag Force around a Floating Body

[valbooxd]

Hi,

I am new to Autodesk CFD and currently starting my Final Year Project in University. I am hoping someone could advice me if Autodesk CFD is able to calculate the viscous drag found around a floating cylinder and how i should go about doing it.

Thanks so much in advance!

[Amal.C]

Hi, 

Welcome to Autodesk Community 🙂 

CFD is able to compute the drag force. 

please check this out. 

please let me know if you have further questions. 

thanks, 

Amal

[valbooxd]

Hi Amal, 

Just to double confirm, the solution attached is for the calculation of drag coefficient on vehicles in air, however in my case i would like to analyse the drag of a floating body in water. Would the steps be the same?

Thank you so much!

[Jon.Wilde]

Hey,

Sort of... although I guess you will have a portion of this in water and a portion in air?

Are you looking to essentially run a free surface analysis looking at what happens when the water reaches the body?

Note that we would only be considering the forces from the water here - the air would not be considered in a free surface analysis.

Might you be able to run this as a 2D analysis? 

Thanks,

Jon

[valbooxd]

Hi Jon, 

Yes, i will have a small portion of the model in air and will be creating a free surface for the analysis however i will only be looking at how the water affects the model underwater and more specifically the viscous drag resulting from the heaving motion of the body in water. And if i do run this as a 2D analysis, how to i go about doing it?

Thank you!

[Jon.Wilde]

Hey,

It sounds like you can run 2D! Definitely do this if you can, SO much simpler and faster 🙂

Are you also looking at motion though? Will it be prescribed motion or flow driven?

You simply need to draw the model on the x-y plane in CAD, volumes become surfaces and surfaces from the 3D model become edges in the 2D. 

Check out the attached file for what can be achieved.

Hope that helps,

Jon

[Jon.Wilde]

Hey!

How did this work out in the end? Did you need anything further or is it looking good?

Thanks,

Jon

[valbooxd]

Hi Jon,

I am currently testing a single cylinder within CFD, however i will need to simulate three cylinders connected in a triangle format, therefore i think i would have to simulate that in 3D.

I do have another question though, while trying to make a wave and using the boundary condition harmonic function, i could only set the input wave. So how do i set an output to allow the waves to leave the volume i created around the model? Is there any other boundary conditions i would have to set before starting the simulation?

Thank you!

[Jon.Wilde]

Sounds like you will need 3D, yes.

Maybe a 1/2 symmetry model?

For the boundary conditions, check out the mini guide here.

[valbooxd]

Hi Jon,

I was trying out the models over the past few weeks but i could not get any drag force being exerted onto the model. I was wondering if any of my boundary conditions are wrong or there is something else i have left out.  

Thank you!

[Jon.Wilde]

Hi,

What do you see when you look at the results? Might you be able to share?

The setup looks like it might be sensible - does the flow look OK when running?

What does the mesh look like from inlet to outlet and over the unit? - Please show the mesh on a cut plane 🙂

Thanks,

Jon

[valbooxd]

Hi Jon,

Thanks so much for the quick reply!

For now, i used a velocity of 0.068m/s as an input into the scaled down model which you can see in Fig 3, however since i used pressure 0 as an outlet condition, there seems to be a backflow of fluid which you can see in the video i attached. Is there any way i can get rid of it? I am quite worried it my affect the results of my model as i run more iterations.

Currently, the meshing i am using is quite coarse for the flow under the model and i have plans to further refine it once i could get an output force from the model.

I used the wall calculator method in this example (https://sustainabilityworkshop.autodesk.com/products/calculating-coefficient-drag-urbee-example) to see if the force exerted on my model can be shown. However, i got 0 newtons. 

Thank you!

[Jon.Wilde]

Why not run a 1/2 symmetry model here? Saves on mesh - or leaves you scope for more.

Could I suggest trying a better mesh now also? This can actually run faster with a VOF analysis than slower 🙂

You can also pull flow out of the outlet with these models as there are P=0's around to prevent over constraining.

It might help to start the entire fluid volume with an initial velocity - saves waiting for it to ramp up.

Hope that helps,

Jon

[valbooxd]

Hi Jon,

I tried pulling flow out through the outlet (around 1/3 of my original velocity) and assigned an initial velocity to my model. I plotted a point right in front of my model and i realized my simulation did not start at 0 seconds and the velocity dropped drastically to around -18m/sec soon after starting the simulation. I used the wall calculator to see my results and there was not any forces in the x direction so i could not compute the drag force.

Do you have any idea why i would experience this drastic drop in velocity, no force on the model and the simulation time starting at 1.4 seconds??

Thank you!

[Jon.Wilde]

It could be that CFD is filling up the space above the fluid right? I would think it is some sort of wave that needs time to stabilise.