Pierre Beynet

Full Scale VIV Response Measurements of a Drill Pipe in Gulf of Mexico Loop Currents

In September 2004 BP retrieved a Gulf of Mexico drilling riser due to excessive loop currents. Whilst waiting for the currents to diminish an impromptu full scale riser Vortex Induced Vibrations (VIV) test was carried out using the drilling riser standalone monitoring system. A 6-5/8 inch drill pipe was instrumented and deployed under the drilling vessel to a depth of 1,000ft. The drilling vessel drift was varied relative to the loop current to produce a range of current profiles and surface current speeds. Current loading along the length of the drill pipe was measured using the on-board acoustic doppler current profiler (ADCP). The paper describes the test set-up and measured VIV response. It includes observations of cross flow VIV, in-line VIV and additional response at higher frequencies not currently predicted with industry VIV analysis tools.

Lock-In, Transient and Chaotic Response in Riser VIV

We show using experimental data on a model riser that lock-in of long flexible risers placed in sheared or uniform cross-flows is a much richer phenomenon than lock-in of flexibly-mounted rigid cylinders under similar conditions. In particular, we find that the frequency content of the riser response may be either narrow-banded around a single dominant frequency (Type I response) or distributed along a relatively broad range of frequencies (Type II response). Distinct transition from Type I to Type II response, and vice versa, can occur several times within a single experimental record. Type I responses reveal features of a quasi-periodic oscillation, often accompanied by large 3rd harmonic components in the acceleration and strain signals, increased correlation length, stable riser trajectories, and monochromatic traveling or standing waves. Type II responses, on the other hand, are characterized by features of chaotic oscillation with small or negligible 3rd harmonic components in the acceleration and strain signals, reduced correlation length, and a continuous spectrum. We study how the fatigue damage differs in the two types of riser response.Copyright

Traveling Wave Response in Full-Scale Drilling Riser VIV Measurements

The dominance of traveling wave VIV response is observed in full-scale measured drilling riser data for the first time. This paper presents the novel methods developed to identify the presence of traveling versus standing wave riser structural response in the full-scale data and the observations. This paper reports on some of the work conducted under the most recently completed phase of the DeepStar JIP (Phase 9). The paper uses four different novel methods to identify the presence of traveling versus standing wave structural response in data obtained from a full-scale Gulf of Mexico drilling riser during a loop current event. The techniques are: 1) Observation of RMS accelerations between synchronized accelerometers; 2) Observation of filtered displacements between synchronized accelerometers (using a new postprocessing synchronization technique); 3) Displacement versus angular rate phase diagrams; and 4) Derivation of upward/downward curvature components via an algorithm proposed by two of the co-authors. High level conclusions are drawn about the structural response types. Recommendations for future instrumentation campaigns are made.Copyright

Distributed Wake-Body Resonance of a Long Flexible Cylinder in Shear Flow

The fluid-structure interaction mechanisms involved in the development of narrowband and broadband vortex-induced vibrations of long flexible structures placed in non-uniform currents are investigated by means of direct numerical simulation. We consider a tensioned beam of aspect ratio 200, free to move in both the in-line and cross-flow directions, and immersed in a sheared flow at Reynolds number 330. Both narrowband and broadband multi-frequency vibrations may develop, depending on the velocity profile of the sheared oncoming current.Narrowband vibrations occur when lock-in, i.e. the synchronization between vortex shedding and structure oscillations, is limited to a single location along the span, within the high current velocity region; thus, well-defined lock-in versus non-lock-in regions are noted along the span. In contrast, we show that broadband responses, where both high and low structural wavelengths are excited, are characterized by several isolated regions of lock-in, distributed along the length. The phenomenon of distributed lock-in impacts the synchronization of the in-line and cross-flow vibrations, and the properties of the fluid-structure energy transfer, as function of time and space.Copyright

Reliability-Based Factors of Safety for VIV Fatigue Using Field Measurements

The development of a Vortex Induced Vibration (VIV) fatigue Factor of Safety (FoS) consistent with state-of-the-art industry design practice is cast within the coherent framework of reliability analysis. The proposed methodology consists of the following steps: (i) define the failure criteria or limit-state function, (ii) set-up a deterministic analysis model, (iii) characterize the uncertainties involved in the problem, (iv) propagate the uncertainties through the deterministic model and assess the probability of failure due to VIV fatigue and (v) calculate the FoS required to achieve a given failure probability. The proposed methodology is demonstrated by determining the FoS associated with using state-of-the-art VIV prediction models to attain varying reliability levels (probabilities of failure) in a hypothetical design scenario. Prediction uncertainty is based herein on measured flow and response data for several full-scale drilling risers working in the field. Results indicate that depending on the reliability level required of a particular design, different FoS than those that currently appear in guidance may be appropriate. Results also indicate the sensitivity of the FoS to the riser and prevailing current type; analysis program and input parameters; and accumulation of conservatism in aggregate vs. single-event damage predictions.Copyright

Re-Evaluation of VIV Riser Fatigue Damage

The paper describes a new characterization of the properties of the vortex-induced vibrations (VIV) of marine risers, which emerges from processing field and experimental data. We show that two currently employed assumptions: (a) that VIV is a statistically steady-state response containing one or several frequencies, and (b) that VIV consists of alternating dominant modes (mode-sharing), are inadequate. Instead, we find that the response either contains strong traveling wave components accompanied by high force harmonics; or consists of a chaotic wandering among several traveling and standing waves, associated with a wide-band spectrum; both types of response require careful consideration for correct fatigue evaluation.Copyright