Learning Forum

[Dong Ha KIM]

Hello,

I am a novice for the Ansys Fluent and I am trying to simulate the steel ball (diameter of 0.05m) dropping experiment using DPM (Discrete Phase Model). The steel ball is dropping 0.385m above the water pool of 1.5m deep. 2 phases of water (primary phae) and air (secondary phase), and a discrete phase of steel ball are defined in the multiphase DPM model. The DPM penetration length result from the 22R1 version shows that the steel ball bounces back  when it drops about 0.3 m ( near the water pool surface). Then it jumps up about 0.2 m and then starts going down again (refer to the attached file). The tank consists of upper part (for the air) and lower part (for the water pool), and there is nothing between the two parts. The mesh type is a cube and its size is 0.1m. My questions are: 1) Is the DPM model appropriate for the simulation of steel ball dropping experiment under this  condition? 2) why does the steel ball bounce back? (When I drop the ball 0.2m above the pool, there is no bounce back and the steel ball reaches the tank bottom gradually) 3)  I read from the manual that the discrete phase reacts with the primary phase only. When I model the free surface problem mentioned above with the primary phase of water and the secondary phase of air, which phase reacts with the discrete phase (steel ball) while the steel balls drops? Thanks.

[C N]

Hello,

For this particular application, The Discrete phase modelling is not applicable. The VOF model approach can be used. For further understanding refer this link 22.3. Setting Up the VOF Model (ansys.com).Because in DPM approach the ball will be considered as a point mass without any volume so it will not displace anything .

Thanks

[Dong Ha KIM]

Thanks for the quick response to my post. As you mentioned, DPM has several limitations including your comments. By the way, In this simulation I am interested in the behavior of steel ball instead of water  pool. Though the VOF method is a good way to  simulate the multiphase flow, it gives a numerical diffusion to follow the steel ball location. Hence I am trying to use the DPM model in this case. By any chance, is there any reason why the steel ball bounces back near the water pool surface?  I expect the steel ball experiences the gravity force as well as the drag when it drops down from the air to the water pool, keeping going down to the water pool though the water drag slows down the ball.   Is the Macroscopic Particle Model (MPM) included in the Fluent Addon Modules applicable to this case?  Thanks in advance.

[C N]

Hello,

The physics behind the steel ball bouncing back on the water surface initially and then sinking into the water at a greater height compared to a smaller height is due to the law of conservation of energy because at a greater height you have more potential energy stored, so when you drop the ball from a greater height the potential energy gets converted to kinetic energy and the ball gains more velocity which is more than sufficient to overcome the shear resistance of the water surface so it bounces back initally and then later sinks into the fluid. At a smaller height there is less potential energy stored so less velocity is obtained during the conversion of potential energy to kinetic energy which is not sufficient to overcome the shear resistance of the water surface since velocity is less there is more pressure drag is experienced and it sinks into the fluid. For a more detailed in depth understanding Particle Velocity, Particle Force balance and Particle torque balance can be analysed in which one of the most important force is the drag force which depends on particle relaxaton time and particle relative reynolds number.But in your case Qualitatively speaking even at both the heights the steel ball will sink inside the water. Yes Macroscopic particle modelling can be used.

Thanks

[Dong Ha KIM]

Thanks again. If Mr. Natraj doesn't mind, other Ansys engineer's comment will be helpful to understand the bounce back phenomena. Also if I use the Macroscopic Particle model in this simulation, are the similar results (bouncing back) expected? Thanks.

[Prashanth]

Hello Dong,

Here's my two cents on your model. I would expect steel balls dropped over a hard surface to reproduce the same results. The bouncing also looks unphysical even for a hard surface though, but your's a water surface. Note that the balls that you are dropping are point masses that don't displace any water volume. 

You might have to reconsider your model setup even if you are not interested in water pool splash. Instead of DPM or MPM, try using VOF + 6DOF with overset.

[Dong Ha KIM]

Thank you so much. I guess your comments are worthy more than 2 cents.