Water entry of object

waradevivek

i am trying to simulate water entry of cylidrical object. I have mentioned relative weight of object as 2.4. As relative weight is more than 1, my body should sink, but it is coming out of the fluid after some time. Please suggest me where i am going wrong.

Alex

The relativeweight is used to define the mass of the individual particles of the floating

If you are creating particles only in the faces, perhaps the total mass of the object leads to a weight which is not higher than the buoyancy

Try to create the cylinder full of particles, so that, particles also inside

To avoid problems you should define the mass of the cylinder once this is created with the proper geo (only faces or full as your cylinder should be)

Regards

waradevivek

Dear Alex,
Thanks for info,
I have some further queries.
how to create cylinder full of particles , I am importing STL file of object.
Also i just wants to know if it is possible to create more particles only on floating object so as to capture the physics more accurately. Now it is creating particles which are equivalent to "dp". Can we assign different dp for different objects.

Regards,
VIvek

Alex

of course, you have to fill the cylinder with the same "mk" of the faces of the cylinder
you can use filling algorithms such as fillbox where you have to define the seed point and the limits of the box

examples in: page 45 of XML_v4.0_GUIDE

http://dual.sphysics.org/index.php/faq/#Q_27

zilonglee

@waradevivek
Right now it seems that we can't set a different "dp" for floating objects, and the same "dp" for fluid and floating objects causes some problems. Especially for a 3D case the floating objects is complex and the whole computational domain is large compared to the floating objects. We may need more particles to perfectly represent the geometry of the floating objects then we set a small "dp", but a small "dp" will simultaneously let the particles of the computational domain increase exponentially, that's a nightmare. Hope the developer of DualSPHysics can take some considerations of this question.

waradevivek

Dear Alex,
can we use multi gpu for version 4.2 ? since number of particles are increasing exponentially as i am decreasing dp. I have machine with 8 gpus "Tesla V100-SXM2-16GB".

sam

I agree with @zilonglee. A constant value of dp creates an unnecessary amount of fluid particles if you need to refine any boundary object, which is the case in most practical applications. Unfortunately, the developers will release the variable particle resolution option in v5.0 together with the Multi-GPU implementation. Hope they consider releasing atleast one of these features (either variable particle resolution or Multi-GPU) in their next version.

waradevivek

thanks.. @zilonglee and @sam

waradevivek

Dear Alex,
While solving the Problem which i have mentioned above, cylindrical body is now sinking in the fluid, but we are not able to get the desired sink rate. Sink rate according to analytical calculations is found to be 11.5 m/s, but what we are getting from SPH is around 2 m/s.
@Alex please guide us where we are going wrong.

Alex

check the mass of the object and be sure is the correct mass
the best is to define the mass in the XML

waradevivek

Dear Alex,
Thanks..
I am specifying "massbody" for floating object.
Also .xml file generated by Gencase is also showing correct masspart. (I.e. if i multiply masspart and no of floating particles, i am getting mass equal to floating object.)

Alex

If you want, please send to us the files to run your case and we will take a look when we can

Regards