[rohanjha]

Greetings of the day,

I have been running a 6DOF simulation in REEF3D for a rigid cylinder via a .STL file. The cylinder diameter (D) is 3.5 cm, and the length is 24.5 cm. It has been placed in a domain of dimensions [40D x 20D x 10D]. Surface 1 is inflow, surface 4 is outflow, and surfaces 2,3, 5 and 6 are walls. The cylinder is oriented with its axis along the Y direction. It has been given freedom only in the heave direction (Z-axis). The cylinder for reduced mass (m*) = 2 keeps falling out of the domain. I have attached my ctrl and control files. The problem persists with the cylinder being mounted on a spring or m* being increased but does not occur for m* < 1.

control.txt

C 11 1
C 12 21
C 13 21
C 14 2
C 15 21
C 16 21

B 1 0.025
B 10 0 1.4 0 0.35 0 0.7

M 10 4

ctrl.txt

B 10 1
B 50 0.0001
B 60 1
B 61 1

D 10 4
D 30 1
F 30 3
F 40 3
F 50 2
F 60 0.7

I 10 1

N 10 14
N 11 11
N 40 6
N 41 30.0
N 47 0.3

M 10 4

P 10 1
P 20 10

T 10 2

W 11 0.8016 0 0
W 22 -9.81

X 10 1
X 11 0 0 1 0 0 0
X 13 2
X 21 1998.4.0
X 41 2.1
X 180 1
X 182 0.35 0.175 0.35
X 183 0.35 0.175 0.35 90 0 0

Regards,

Rohan

[Felix_Sp]

Hello,

I think your cylinder is just falling through the bottom since your x21 value is greater than the water density (I think there is one dot too much in your post) and the 6DOF algorithm of REEF3D is not yet capable of interacting with solid boundaries. Therefore, for the solid object, the boundary is not present, and it just keeps accelerating until infinity. Your observation of m* behaviours strengthens my beliefs (if m* is the relative mass in your post).

Cheers

[rohanjha]

Originally Posted by Felix_Sp

Hello,

I think your cylinder is just falling through the bottom since your x21 value is greater than the water density (I think there is one dot too much in your post) and the 6DOF algorithm of REEF3D is not yet capable of interacting with solid boundaries. Therefore, for the solid object, the boundary is not present, and it just keeps accelerating until infinity. Your observation of m* behaviours strengthens my beliefs (if m* is the relative mass in your post).

Cheers

Thanks for the response, Felix. I was thinking the same thing; I faced this issue because I was trying to induce VIV, and for a random trial case, the cylinder just fell through the domain.

[valgrinda]

Hi Rohan,

there has been an issue with certain geometries regarding normal vector alignment, thus resulting in wrong signs when performing the force integration on the floating body. The cylinder was one of the effected geometry types. This has been fixed in the most recent master 23.01. Maybe also try the release candidate.

[rohanjha]

Originally Posted by valgrinda

Hi Rohan,

there has been an issue with certain geometries regarding normal vector alignment, thus resulting in wrong signs when performing the force integration on the floating body. The cylinder was one of the effected geometry types. This has been fixed in the most recent master 23.01. Maybe also try the release candidate.

Thank you for the response Hans. I have been using version 23.01 and continue to face these issues. The body falls through even for very large values of the spring coefficient.

[valgrinda]

Hi Rohan,

there were a couple of things that needed fixing, so I have just attached the control files. Generally, the mesh was also way too fine. You are also better off with running it just in 2D. Also very important, since you seem to go for VIV: the mesh size needs to be fine enough to allow for vortex shedding and RANS turbulence modeling suppresses the shedding (so I turned it off).

In the end though, it appears, that the density of floating body should be slightly lower than that of the water, so that you get some tension onto the spring.

[valgrinda]

Btw, I have increased the spring stiffness. If you reduce it that back to your value, the motion amplitude will increase accordingly.