Flow Induced Vibration

Abstract

A central goal in the thermal design of tubular heat exchangers is to utilize the available shellside pressure loss to maximize the shellside film coefficient. The most effective means of accomplishing this objective is to arrange the shellside fluid to flow across the tube bank. Single segmental baffles (Fig. 1.4.5) provide the nearest practical alternative to a pure cross flow arrangement. Due to its ability to actuate maximum heat transfer rate for available pressure loss in a minimum amount of space, the single segmental baffle design has become a standard construction detail. Early texts on heat exchanger rating methods [16.1.l–2] acknowledged the dominance of the single segmental baffle design. It was not until the ’60’s, when widespread failure of tubes began to occur [16.1.3], that another consequence of shellside cross flow — flow induced vibration — came to the fore. In retrospect, relatively high fouling factors used in the petrochemical, chemical, and process industries had indirectly prevented uncovering of the vibration problem. High fouling factors specified by the user blunted the heat exchanger designer’s motivation to maximize the shellside coefficient by maximizing cross flow velocities. Lower fouling factors, coupled with greater flow rates used in the power industry, removed this disincentive. Moreover, the specified pressure loss in power plant heat exchangers also tended to be greater. Conventional wisdom held, correctly so, that higher cross flow velocity (in the absence of erosion related problems) produce better operating heat exchangers by reducing the deposition rate of crud and debris on tube surfaces. Designers have since learned that ignoring the effect of increased cross flow velocity on tube vibration failure can be an expensive mistake. It is indeed true that sporadic failure of tubes in operating heat exchangers occurred even before the ’60’s. However, the significance of flow induced vibration on heat exchanger reliability did not attract sufficient attention until tube failures became commonplace, particularly in the power industry. Finally, it should be added that parallel flow has also been reported to be responsible for some tube failures. However, even in ostensibly parallel flow induced failures, cross flow components of the turbulent stream have been held responsible by some investigators [16.1.4]. From a practical designer’s point of view, the design focus should rightfully rest on the cross flow component of the flow velocity.