Session: 01-08-01: Passive, Semi-Active, and Active Noise and Vibration Control

Paper Number: 166417

Investigation of Clamp Numbers and Positions to Mitigate Flow-Induced Vibration in High-Speed Superheated Steam Flow Through a Pipe Elbow 

Fluid transmission systems carrying high-pressure fluid often experience breakdown, excessive vibration, and even fatigue failure, particularly at critical locations. Although external excitations play a role in pipe dynamics, recent findings have pointed to internal sources such as Forced-Induced Vibration (FIV). FIV is defined as the reciprocating motion of a body caused by force imparted by a flowing fluid over the body. If the flowing fluid is turbulent inside pipelines, the amplitude of the vibrations of a pipe is significant in both longitudinal and transverse directions, which leads to damage to the pipe through fatigue failure and affects the pipe’s supports. Thus, understanding the interplay between fluid and structure, especially at critical locations like bends, tees, and valves, is necessary. Researchers have been using the Fluid-Structure Interaction (FSI) method for the abovementioned issues over the years. The ideology behind the FSI method is that flowing fluid generates pressure fluctuation, causing the structure to deform, affecting the flow regime, and implying a further change in structure dynamics. However, literatures investigating the dynamics of pipes carrying low-speed fluid through the FSI method often limit the fluid velocity to a mere 8 m/s. Limited works have been done on the effect of superheated steam flow on pipe dynamics. However, the operating conditions for these industrial pipelines were often limited to 500 KPa to 1 MPa only, demanding further insight into superheated steam pressure crossing 10 MPa. Lastly, to achieve clamp position optimization for a wide range of fluid properties, simulating the effects of superheated steam onto the pipe and how it affects clamp position is necessary. Thus, the objectives of this research were to investigate the fluid pressure distribution phenomena of superheated steam in an industrial pipeline section using CFD analysis and analyze the transient vibrations through structural analysis by mapping the pressure distribution. Finally, the optimal clamp numbers and positions were determined from the Fast Fourier Transform (FFT) of acceleration and deformation data for varying mass flow rates. Therefore, we chose a 90-degree bent section of a pipe with 2 m downstream and 4 m upstream sections from the bend from an industrial pipeline network, which is well-suited for extreme boiler operating conditions. A simplified version of an industrial-grade clamp was also designed to keep computational demand low. We chose a superheated steam of 12.58 MPa inlet pressure for the high-speed fluid for three different mass flow rates: 180, 200, and 220 kg/s, based on boiler load factor and efficiency. The idea behind choosing three different mass flow rates spanning a 10% threshold was that the flow rate would never have a constant flow throughout the practical application. RANS method has been used in ANSYS® Fluent for the fluid domain due to its low computation time for multiple structure domains. The Reynolds Stress Model (RSM) was selected among different RANS-based turbulence models as it also showcases shear-stress correlations.  The inflation layer was kept at 1.15, and an O-grid mesh with hex-dominant elements was selected. For the structure domain, the critical sections were finely meshed. For the boundary condition, the inlet section was kept fixed for the direct connection with the boiler, and movement of the outlet section was restricted to the transverse section. The clamp connection was considered frictionless, and the outer surface of the clamp was kept fixed. We validated our fluid and transient structural model by comparing velocity and pressure distribution contours and acceleration data with established results for low-speed oil before moving on to the simulation stage for current research work on superheated steam. Results for superheated steam accounted for maximum pressure ranging between18-28 MPa and maximum velocity ranging between 60-74 m/s for three different mass flow rates at pipe bend. The acceleration and deformations were recorded from 10-16 mm/s² and 0.4 to 0.7 mm, respectively. As for the investigation of optimal clamp position, we fixated 5 clamp positions at the inlet and 10 at the outlet with 0.36 m regular intervals. Both the inlet and outlet sections were simulated for total acceleration by placing one clamp per simulation at one of the strategic clamp positions. After selecting the clamp positions for which total acceleration was least at the first natural frequency, simulations were performed again, keeping the clamps at the best possible locations for three different mass flow rates. For the acceleration, the reduction was recorded at 31% on average for a single clamp at 0.4 m away from the inlet towards the bend and 33.7% for an additional clamp at 0.360 m away from the outlet towards the bend, implying an insignificant effect upon introducing a second clamp. On the other hand, a single clamp at the exact location on the inlet section only reduces deformation by 14% on average for different mass flow rates; however, adding another clamp at the optimal outlet section provides a 35.5% average reduction in displacement, which is more than twice. Our study on how high-pressure and high-speed fluid affect pipe dynamics applies to various sectors with high-pressure fluid transmission systems, including but not limited to nuclear plants, oil piping, and even space exploration.

Presenting Author: Tanvir Hossain Islamic University of Technology

Presenting Author Biography: Tanvir Hossain is an undergraduate researcher in the Mechanical Engineering program of the Mechanical and Production Engineering Deptertment (MPE) at Islamic University of Technology (IUT), Bangladesh, under the supervision of Dr. Md. Zahid Hossain. His research interest lies in the intersection of Fluid-Structure Interaction, Flow-Induced Vibration, Reinforcement Learning and Control.

Authors:

Hasin Ahmad Zafir Islamic University of Technology
Tanvir Hossain Islamic University of Technology
Arafat Ahmed Islamic University of Technology
Chowdhury Sadid Alam Islamic University of Technology
Ahmed Abuhatira Waha Oil Company
Md. Zahid Hossain Islamic University of Technology