# Buoyancy problem

Hello, I wrote a few weeks ago but I think I wasn't able to properly explain my problem.

I apologize, I'll try to explain it better.

I have been using DualSPHysics for a month and when simulating bodies entering a fluid I get an error with the densities: bodies with greater density than the fluid float.

This image corresponds to the before and after the entry of a sphere with a density of 1500 into a fluid with a density of 1000 (water):

*There are walls on every side except above

I decreased the distance between particles, the process was better seen but it kept floating.

I increased the density, passing 2000 it sinks.

All parameters and constants are the ones used by the designsphysics model for water. Artificial viscosity with alfa 0.01, SPH and none density diffusion term:

I also tried other geometries such as cylinders and I get the same result: after the density of 2200 it sinks.

a) I would like to know why this is happening. Why do bodies with densities greater than the fluid float?

b) What is the influence of dp on buoyancy? Empty spaces influence the calculation of density?

Thank you for your time

• I also separated the walls thinking that they could influence but I got the same result despite having more fluid.

Any help would be welcome

• Hi,

From my experience, it has to do with the gap of the boundary formulation (https://dual.sphysics.org/faq/#Q_13). My interpretation is that it changes the apparent volume of your sphere and thus its density.

Try to take it into account by reducing the radius of your sphere (gap = 1.5 sqrt(3*dp*dp)) and imposing directly the right mass to the sphere rather than the density.

Not an expert advice, just a tinkerer trick...

Best regards

• I've tried this before, and the same thing happens. My suggestion is that after you have calculated the corresponding mass yourself, do the simulation calculation by giving the mass instead of the density. As for how the mass is calculated, it is your actual density value multiplied by the volume or area of the model in the simulation.

• Hello,

Does the gap of the container interfere with the gap of the sphere ? If it's the case then, that could mess things up also. For the container, you could use mDBC.

Regards

• Hello, when you did the simulation using mass did it work? Because I tried but still no changes

• Hello, I also tried with another boundaries where the separation between walls and the sphere es bigger. It was the same result

I tried with mass and bathing changed. I honestly don't know how to scale this difference

• Hi WhyIsThisAsked, did you ever get to the bottom of this? I'm having the same problem.