Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.698571
Title: Interfering of two tandem cylinders undergoing VIV : mathematical modelling and experiments
Author: Armin, Milad
ISNI:       0000 0004 5991 7526
Awarding Body: University of Strathclyde
Current Institution: University of Strathclyde
Date of Award: 2016
Availability of Full Text:
Access through EThOS:
Access through Institution:
Abstract:
Behaviour of two cylinders in tandem subjected to a stream of steady uniform flow is investigated. This research includes experimental and mathematical simulation so that a better understanding of the interaction between two bodies undergoing vortex induced vibration (VIV) can be achieved while one is submerged in the wake of the other. Amplitude and frequency of oscillation are observed for both cylinders when they are placed at different distances from each other. Positioning cylinders at various spacings can help to shed some light on the interaction mechanism of the upstream wake and trailing cylinder. An experimental investigation was carried out in sub-critical Reynolds number and low mass-damping to simulate the conditions in which offshore structures are deployed. Initially, two identical cylinders are placed in-line with the stream at various spacings. The response of leading cylinder is observed to be similar to that of an isolated cylinder experiencing VIV at all spacings. On the other hand, trailing cylinder response is observed to increase with flow velocity at small and medium spacings. Moreover,as the spacing grows large downstream response becomes more similar to that of the leading cylinder. Motion trajectory of trailing cylinder is significantly influenced by the leading body, and does not follow the typical figure of eight observed for an isolated cylinder at all velocities. Frequency power spectrum of obtained time histories reveals that two sources of excitation exist for trailing cylinder. Corresponding motion to each excitation source is determined using Fast Fourier Transform. The second set of experiment was conducted using similar cylinders with different natural frequencies to observe how it influences the interaction between two cylinders. It was observed that behaviour of trailing cylinder alters in comparison with initial set up. However, two sources of excitation are still detected in this set-up. Mathematical simulation is pursued by modelling the oscillating cylinders with a simple mass-spring-damper system. Furthermore, the force exerted to cylinders by wake is simulated by wake oscillators which can capture the self-exciting and self-limiting nature of VIV phenomenon. Two equations are coupled together by assuming that wake force is proportional to cylinder acceleration. Then, the system of equations is solved analytically,and results are compared to those obtained by a SimuLink model of the system. SimuLink model is solved by numerical RungeKutta method. It was observed that model is successful in simulating leading cylinder vibration amplitude,while it, initially, fails drastically to predict the oscillation amplitude of trailing cylinder due to buffeting vortices. Two terms were added to accommodate the effect of upstream wake on the trailing cylinder, to modify the force in the equation of motion which is respectively proportional to acceleration and velocity of the cylinder. Moreover, acceleration term is determined by fitting a linear function of the variable to the difference between upstream and downstream wake force obtaining from the experimental investigation results. Additionally, the damping term is determined by optimization of variance between simulation and experiment results. Such an observation can confirm that upstream turbulent wake has a significant influence on added mass coeffcient of the trailing cylinder which both are observed to be dependent on upstream Strouhal number. Overall, the agreement between mathematical model and experimental results is evaluated for both cylinders. Model error for trailing cylinder is calculated between 20% to 30% in cross-flow direction. This error is lower than that of the leading cylinder which is a well established method in literature for simulating VIV of an isolated cylinder.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.698571  DOI: Not available
Share: