ABSTRACT

Conductor pipe group sections have been tested In steady flow, harmonically oscillating flow and In oscillating flow with current to determine the drag, lift and Inertia coefficients The purpose of the tests was to investigate shielding effects associated with dense pipe groups

The test program was carried out In two steps Steady flow tests were performed In the cavltatlon tunnel at Marlntek A total of 12 different pipe configurations, involving from 1 to 12 pipes, were tested at 7 different flow directions The pipes had a surface roughness of k/D = 2–3 10−3 A Reynolds number range of up to Re = 5 1 103 was covered, which for the chosen surface roughness was sufficient to reach the transcritical flow regime It was found that the average drag coefficient for pipe groups was reduced compared with CD for a single pipe, up to 70 % reduction was found A general trend rn the results is that the average drag coefficient for the pipe groups varies less with Re than it does for a single pipe

Oscillatory flow tests were performed In a towing tank A single cylinder, two different 2x4 arrays and a 3x4 array were tested The surface roughness was the same as In the steady flow experiments The Keulegan-Carpenter number was systematically varied between 60 and 130 The Reynolds number, based on oscillating velocity amplitude, was In the range Re = 20–50 105 The effect of relative current velocity was examined Results are presented for drag, lift and Inertia coefficients A similar trend of reduced CD with increasing number of pipes was found as In the steady flow tests Furthermore, the drag coefficient was found to decrease with increasing relative current velocity when keeping t e other parameters fixed

INTRODUCTION

In the design of conductor pipe groups the hydrodynamic loading due to current and waves Is a governing parameter The force coefficients to be used In load estimation are often difficult to determine, since they depend on many different parameters

The drag coefficient of one single cylinder In steady current IS generally a function of the Reynolds number, the surface roughness and the turbulence level In the flow T his dependence has been extensively tested In experiments and is described In the Literature, l g Sarpkaya and Isaacson (1981)

For a group of pipes the total drag force depends on additional parameters which are specific for the group geometry (I e number of rows and columns), spacing-to-diameter ratio, and incident flow angle relative to the group In general this dependence is far less known Some results have been published, especially for two pipes (see Zdravkovich(1 987)) For pile group configurations with a central tube with a number of small diameter outer tubes, data have been published by Loken et al (1979), Ottsen-Hansen et al (1979) Demirbilek(1988)

In the present tests, pile group configurations wlth tubes of equal diameter have been considered

This content is only available via PDF.
You can access this article if you purchase or spend a download.