Isothermal sloshing in a circular tank (CCP-WSI Blind Test Series 5)

Isothermal sloshing in a circular tank (CCP-WSI Blind Test Series 5) image

The “CCP-WSI Blind Test Series 5: isothermal sloshing in a circular tank” investigates physical modelling of a horizontal cylinder partially filled with water. Two separate experimental campaigns have been conducted exciting the tank in two different directions. The experiments are 1) forced harmonic horizontal excitation, corresponding to the x-direction of the global co-ordinate system defined figure 1a. 2) forced harmonic vertical excitation, corresponding to the y direction of the global co-ordinate system defined figure 1.

The blind test is therefore split into two parts:

  • 1) Forced harmonic horizontal excitation.  
  • 2) Forced harmonic vertical excitation.  

In each experiment data is acquired through a high-speed camera. Through the acquired images the displacement of the tank, the global free surface shape, and the fluid centre of gravity (COG) is extracted by means of an image processing algorithm (IPA).

Contributors

  • Stuart Colville
  • Scott Brown
  • Ed Ransley
  • Yeaw Chu Lee
Contact person
Stuart Colville

Description

The “CCP-WSI Blind Test Series 5: isothermal sloshing in a circular tank” investigates physical modelling of a horizontal cylinder partially filled with water. Two separate experimental campaigns have been conducted exciting the tank in two different directions. The experiments are 1) forced harmonic horizontal excitation, corresponding to the x-direction of the global co-ordinate system defined figure 1a. 2) forced harmonic vertical excitation, corresponding to the y direction of the global co-ordinate system defined figure 1.

The blind test is therefore split into two parts:

  • 1) Forced harmonic horizontal excitation.  
  • 2) Forced harmonic vertical excitation.  

In each experiment data is acquired through a high-speed camera. Through the acquired images the displacement of the tank, the global free surface shape, and the fluid centre of gravity (COG) is extracted by means of an image processing algorithm (IPA).

Experimental Set-up

The physical experiments were completed in the COAST Laboratory at the University of Plymouth, UK. A horizontal cylinder geometry was manufactured from cast acrylic. A cast acrylic sheet was bored-out using a lathe - to create a circle with a diameter of D=0.4m. Two gaskets were clamped, with a series of bolts, between the bored-out piece and the two end caps to ensure a watertight design. Figure 1b shows the final manufactured tank. The global co-ordinate system is defined with the x, y and z axis as displayed in figure 1a. The origin (0,0,0) of the co-ordinate axis is indicated at the centre of the tank. The diameter of the tank is 0.4m and the depth of the tank (z-axis) is 0.04m.

Figure 1: a) Circular tank layout showing co-ordinate system (z axis into the page) b) Image of manufactured sloshing tank. 

Figure 1a shows the coordinate system and parameters of the experiment. Each experiment has three parameters which can be varied, the fill level (h), the forcing amplitude (A) and the forcing frequency (ω). Figures 2 and 3 show the experimental rig for the horizontal excitation setup. In both forced excitation setups (vertical and horizontal) the sloshing tank (a) is mounted onto four bearings (b). Two parallel rails (C) are inserted through two pairs of ball bearings. The rails are built onto the metal frame (e). The sloshing tank is mounted to an electromechanical actuator (d) which provides a sinusoidal input acceleration. The actuator is also housed on the metal frame (e).

Figure 2: Horizontal sloshing excitation experimental setup (reproduced from Colville et al. 2023). 

Experimental Test Program

Isothermal conditions are assumed for all test cases. The temperature of the fluid is assumed to be at room temperature (20°C) and atmospheric pressure (1.013bar). Thus, the liquid density (water) is assumed to be 998kg/m3 and the gas density (air) is assumed to be 1.204kg/m3. The sinusoidal acceleration is generated using an electromechanical actuator in both cases. The fill level (h) is presented in dimensionless form through the tank diameter (h/D). The amplitude (A) of the sinusoidal input acceleration is expressed in millimetres (mm). The frequency of the sinusoidal wave is expressed in hertz (Hz).

  • 1) Forced harmonic horizontal excitation: In the horizontal cases the acceleration input starts at the maxima of the sinusoid. The actuator initially ‘pushes’ the tank (negative x direction corresponding to the co-ordinate system).
  • 2) Forced harmonic vertical excitation. In the vertical case the acceleration input also starts at the maxima of the sinusoid. The actuator initially ‘pulls’ the tank (positive y direction referring to the co-ordinate system).
Table 1: Properties of the forced harmonic horizontal excitation tests
CCP-WSI test ID Fill level (h/D) Amplitude (A) Frequency
(mm) (Hz)
1HIBT5 0.3 0.5 1.3
2HIBT5 0.3 2.0 1.3
3HIBT5 0.3 2.0 1.18
4HIBT5 0.7 0.5 1.45
Table 2: Properties of the forced harmonic vertical excitation tests
CCP-WSI test ID Fill level (h/D) Amplitude (A) Frequency
(mm) (Hz)
1VIBT5 0.3 2.0 3.790
2VIBT5 0.5 4.0 3.882
3VIBT5 0.5 8.0 3.882

Assessment Criteria

For test case 1HIBT5 and 2HIBT5 the CCP-WSI requests left and right sidewall free surface y co-ordinates from the beginning of excitation to a time of 50s. For test cases 3HIBT5 and 4HIBT5 the left and right sidewall free surface co-ordinates commencing after 80 cycles for a period of 15s are requested.

For test case 1VIBT5 the CCP-WSI requests left and right sidewall and free surface centre y-co-ordinates from the beginning of the excitation to a time of 60s. For test case 2VIBT5 the CCP-WSI requests left and right sidewall and free surface centre y-co-ordinates from the beginning of the excitation to a time of 30s. For Test case 3VIBT5 left and right sidewall and free surface centre y-co-ordinates from the beginning of the excitation to a time of 15s.

Please note that the maximum resolution of the camera in the physical modelling (i.e 1 pixel) is 0.46mm. In post-processing of the physical data, the sidewall is assumed to be 5 pixels (2.3mm) inside the wall since the signal contains more noise right at the edge.

Relevant References

Colville, S, Ransley, E, Gambioli, F, Lee, YC & Greaves, D 2023, 'Fluid Response of Sloshing in a Horizontal Cylinder Due to Horizontal Excitation', Paper presented at the The 33rd International Ocean and Polar Engineering Conference, Ottawa, Canada, June 2023.