Experimental data of the influence of the number of rows of a tube bank on the FIV

Published: 31 August 2021| Version 1 | DOI: 10.17632/cxp38zvhzk.1
Contributors:
,
,

Description

Data result from careful measurements with accelerometers in an aerodynamic channel to study the phenomenon of fluid-elastic instability in a square arrangement rows of cylinders and tube bank with all fixed tubes and with one cylinder free to vibrate. The Reynolds number for the tested cases ranged between 11850 and 64104, based on one cylinder diameter and the kinematic viscosity at the environmental conditions of the laboratory. The gap velocity is obtained applying mass conservation considering the reference velocity and area upstream the test section and the free area with the row of cylinders. The test section is assembled with smooth cylinders with 25mm diameter and a p/D-ratio of 1.26. One of the cylinders has one degree of freedom. It is free-to-vibrate in the transversal direction, being attached to two blades that allow the variation of length, to change the stiffness. The positions of the cylinders were numbered and the cylinder free to vibrate was assembled in positions 3 (central cylinder of first row), 8 (central cylinder of second row), 13 (central cylinder of Third row), or 18 (central cylinder of fourth row). The assembly of the cylinder free to vibrate is the same used in Neumeister et al. (2021, a-b). A total of 27 cases were conducted, starting with a single cylinder (reference case), than rows of cylinders, changing the number of rows, the stiffness, and mass ratio. The tested cases present the nomenclature indicating if the case is a single-cylinder (SC), if it is one row (1R), two rows (2R), three rows (3R), four rows (4R) or five rows (5R), the position of the cylinder free to vibrate (P3, P8, P13, P18) and the first natural frequency. The acceleration of the cylinder free to vibrate is the monitored variable. The measurements were executed with accelerometers ADXL 335 with an acquisition frequency of 1 kHz with an anti-aliasing low pass filter at 0.3 kHz. Data acquisition was executed employing an A/D board 16bits NI USB-9162. The acceleration results are presented through the values obtained during the data acquisition for each case tested. Uncertainty values are about 7 % for the acceleration of the cylinder; for the velocity, between 7 and 8 %; for the frequency, uncertainty is about 9 %, leading to a value of 10.5 % for the reduced velocity. REFERENCES Neumeister, R.F., Petry, A.P., Möller, 2021-a, Analysis of the coherence between FIV acceleration and flow velocity in a square array tube bank with P/D = 1.26. Nuclear Engineering and Design, v. 373, p. 111016, DOI: 10.1016/j.nucengdes.2020.111016 Neumeister, R.F., Petry, A.P., Möller, 2021-b, Experimental Flow-Induced Vibration Analysis of the Crossflow Past a Single Cylinder and Pairs of Cylinders in Tandem and Side-By-Side. Journal of Pressure Vessel Technology-Transactions of the ASME, v. 143, p. 031402, 2021, DOI: 10.1115/1.4048101

Files

Steps to reproduce

The tested cases present the nomenclature indicating if the case is a single-cylinder (SC), if it is one row (1R), two rows (2R), three rows (3R), four rows (4R) or five rows (5R), the position of the cylinder free to vibrate (P3, P8, P13, P18) and the first natural frequency. Data with time [s] and acceleration [m/s²] are ordered by rows. A total number of 155 datasets are in .txt format to be read by almost all software. Each case is identified by the nomenclature for each case, the natural frequency (fn), the reduced velocity (VR) and the gap velocity (GV) in [m/s]. Decimal points have been replaced by hyphens to avoid some conflict in the analysis software and different operating systems. An example is: Case2R01_VR18_fn15-63Hz_GV_7-1ms: Case 01, with 2 rows, reduced velocity 18, first natural frequency 15.63 Hz, Gap velocity 7.1 m/s.

Categories

Fluid Dynamics, Mechanical Vibration

License