Rodolfo Gonçalves | The University of Tokyo (original) (raw)

Papers by Rodolfo Gonçalves

Recent studies suggest that the vortex-induced motions (VIM) of a semi-submersible found in model... more Recent studies suggest that the vortex-induced motions (VIM) of a semi-submersible found in model tests over-predicts the response in the field, which may lead to an over conservative design of the mooring and riser system. Within the Vor-tex Induced Motion Joint Industry Project (VIM JIP), run by MARIN and University of São Paulo (USP), possible reasons for this over-prediction are investigated using model tests and CFD [1–6]. A model test campaign was carried out at MARIN to test different candidates that might explain the observed differences. The results obtained with an air bearing setup regarding damping, mass ratio, draft variations and sinusoidal tow velocities , have been published elsewhere [6]. The present publication focuses on the influence of waves on the VIM response. The model was a generic bare hull semi submersible with four rounded square columns at scale 56.5. A simplified mooring system consisting of four springs was designed to match the desired natural period in the sway and yaw direction. The model was towed at different velocities, corresponding to the range of reduced velocities where the highest response is expected. A VIM tow test campaign was carried out in calm water as a benchmark. The model was then tested at 7 different wave-current conditions , and the results are compared with the benchmark case. The results suggest that two factors are important for VIM response in waves: the wave height, and the relation between wave and current direction. Comparing to calm water condi-*

The VIM-vortex-induced motion of multi-columns platforms, SS-semi-submersible, TLP-tension leg pl... more The VIM-vortex-induced motion of multi-columns platforms, SS-semi-submersible, TLP-tension leg platforms and FOWT-floating offshore wind turbines require advances in the study of flow around fixed cylinders. On this matter, is important to understand and predict the influence of flow around fixed structures, in order to understand the vortex detachments and the source for oscillatory lift and drag forces. The main purposes are to carry out V&V-verification and validation to numerical results produced by a CFD code for flow around single fixed cylinders. This code was employed for two different geometries, circular and square, for fixed cylinders with low aspect ratio L/D = 1.5. The V&V was carried out by varying time steps, and mesh sizes and the results were methodically compared with experimental results. Satisfactory results were obtained in this comparison, and valuable knowledge on the flow behavior was obtained to be applied for engineering designs.

Experiments were carried out in a towing tank regarding the flow around a stationary array of fou... more Experiments were carried out in a towing tank regarding the flow around a stationary array of four cylinders with low aspect ratio, L/D = 1.5, piercing the water free surface. One isolated cylinder and one array of 4 cylinders were tested; this number of cylinders corresponds to the number of columns of a typical multi-column offshore floater system, such as semi-submersibles and TLP platforms. Three different geometries of cylinder sections were studied: circular, square and diamond. The aims are to better understand the influence of the relative position of the cylinders on the flow around this type of structure; to create a reliable database for CFD studies and afterward to use this research in VIM calculations for free multi-columns systems. Two parameters were investigated for a distance between column centers, S/L = 3: current incidence angles 0, 22.5 and 45 degrees and column section geometries. The Reynolds number range covers from 50,000 up to 250,000. Forces were measured in each column by using a three degree-of-freedom load cell. Results of lift and drag forces are presented for each column separately and for the whole system. This paper is a piece of an extensive experimental study on flow around an array of multi-cylinders.

Experiments were carried out in a towing tank regarding the flow around a stationary array of cir... more Experiments were carried out in a towing tank regarding the flow around a stationary array of circular cylinders with low aspect ratio, H/D=1.5, piercing the water free surface. One isolated cylinder and one array of 4 cylinders were tested; this number of cylinders corresponds to the number of the columns of the typical multi-column offshore floater system, as semi-submersibles and TLP platforms. The aims are to create a reliable database for CFD studies and to better understand the influence of the relative position of the cylinders on the flow around this type of structure and afterward to use this research in VIM calculations for free multi-columns systems. Two parameters were investigated: distance between column centers, S/D = 2, 3 and 4; and current incidence angle, 0, 22.5 and 45 degrees. The Reynolds number range covers 50,000 up to 250,000. Forces were measured in each column using a three degree-of-freedom load cell. Results of lift and drag forces are presented for each column separately and for the whole system.

Experiments regarding flow-induced vibration on floating squared section cylinders with low aspec... more Experiments regarding flow-induced vibration on floating squared section cylinders with low aspect ratio were carried out in an ocean basin with rotating-arm apparatus. The floating squared section cylinders were elastically supported by a set of linear springs to provide low structural damping to the system. Three different aspect ratios were tested, namely í µí°¿/í µí°· = 1.0, 2.0 and 3.0, and two different incidence angles, namely 0 and 45 degrees. The aims were to understanding the flow-induced vibration around single columns of multi-column platforms, such as semi-submersible and TLP. VIV on circular cylinders were also carried out to compare the results. The range of Reynolds number covered was 2,000 < Re < 27,000. The in-line and transverse amplitude results showed to be higher for 45-degree incidence compared with 0-degree, but the maximum amplitudes for squared section cylinders were lower compared with the circular ones. The double frequency in the in-line motion was not verified as in circular cylinders. The yaw amplitudes cannot be neglected for squared section cylinders, maximum yaw amplitudes around 10 degrees were observed for reduced velocities up to 15. Keywords: flow-induced vibration (FIV), vortex-induced vibration (VIV), vortex-induced motion (VIM), low aspect ratio, floating squared section cylinder

The present study provides original results of an experimental approach to understand the VIV phe... more The present study provides original results of an experimental approach to understand the VIV phenomenon on long flexible cylinders launched in catenary configuration. The tests were carried out in a towing tank by means of a movable floor attached to the carriage car. Three groups of tests were considered, comprising the incidence of uniforme current profiles in catenary arrangements transverse and longitudinal to the inflow, the second one tested in two different directions. The catenary models were built by means of silicone tubes fulfilled with steel microspheres leading to achieve a mass ratio of m * = 3.72. Displacements along the models were measured by means of an underwater image tracking system and a large number of passive targets along the lines. A Galerkin's modal decomposition scheme is applied to obtain the modal-amplitude time histories and then the root-mean-square amplitudes as functions of the modal reduced velocities. Curiously, the modal results are quite characteristic of simpler systems with low m * , particularly in terms of the synchronization behavior observed in VIV reponses of rigid cylinders elastically supported. The modal decomposition procedure , therefore, is presented as a great technique for improvement of the investigations concerning the VIV of long flexible lines, revealing interesting results not yet observed experimentally in a so accurate way.

The vortex-induced motions (VIM) of offshore platforms stand as an intriguing and challenging eng... more The vortex-induced motions (VIM) of offshore platforms stand as an intriguing and challenging engineering problem, drawing attention from industry, universities and research institutes. Field observations, model tests and calculations have extensively showed that the complex fluid-structure interaction can result in appreciable motions and increased fatigue of mooring and risers. It is thus a very relevant issue from the engineering standpoint. A large volume of experimental research has been carried out, mainly to verify designs and characterize the occurrence of VIM. Conversely, the numerical investigations applying CFD tools have shown to be a more flexible approach enabling better understanding of the physics at play due to the possibility of investigating the effects of different parameters upon the vor-tex induced motions of floating platforms. Moreover, the CFD * Currently, also at Argonáutica Engineering & Research. † Currently, at Federal University of Santa Catarina (UFSC). calculations enable investigation of the full-scale behavior of the platforms under VIM, a very controversial issue presently. Bearing upon these issues, the VIM Joint Industry Project aims at increasing physical insight of this phenomenon by means of investigating the influence of geometric design variations, flow conditions and scale effects with the objective of improving practical knowledge that can be applied in the design stage of floating platforms. In this paper, we present some of the CFD studies, results and observations carried out within the JIP, regarding the VIM of a semi-submersible with circular columns in 0 and 45 degrees and over a wide range of reduced velocities. It is confirmed that the 0 degree incidence results in larger motions than the 45 degrees-incidence case, in contrast to the VIM behavior of a semi-submersible with square columns. The tests campaign carried out at the University of São Paulo for the same platform agree with these results. Within the lock-in range, the frequency synchronization of the lift forces on columns and pontoons cause

Recent studies suggest that the vortex-induced motions (VIM) of a semi-submersible found in model... more Recent studies suggest that the vortex-induced motions (VIM) of a semi-submersible found in model tests over-predicts the response in the field, leading to an excessively conservative design of the mooring and riser system. Within the Vortex Induced Motion Joint Industry Project (VIM JIP), run by MARIN and University of São Paulo (USP), possible reasons for this over-prediction are investigated with model tests and CFD. The results presented herein are part of the model tests carried at MARIN, where damping, mass ratio, draft variations and sinusoidal tow velocities were put to the test. The model was a generic bare hull semi submersible with four rounded square columns at scale 56.5. The experimental setup allowed only 3DoF, with heave, roll and pitch restricted by an air bearing plate. A classical VIM tow test campaign was carried out as a benchmark, against which different variations are compared. A newly developed active damping system was used to apply linear damping and virtual mass to the model. This system has been verified with dedicated tests, and its performance is discussed. The tested mass ratios show minor impact on VIM response. Different damping ratios were tested for different drafts. It is found that adding reasonable levels of linear damping (1) significantly reduces VIM amplitude, from 25% reduction up to VIM suppression, (2) narrows the reduced velocities at which VIM lock in occurs and (3) shifts the response peak to higher reduced velocities. All these effects are present in the field measurements published by Ma et al. (2013). The presented results suggest that damping is a strong candidate to explain the differences between model test VIM and field observations. With the active damping system, future VIM model test campaigns can include the damping coming from risers and mooring, assuming that a reasonable estimation of the damping level is available.

Experiments regarding flow-induced vibration on floating rounded squared section cylinders with l... more Experiments regarding flow-induced vibration on floating rounded squared section cylinders with low aspect ratio were carried out in an ocean basin equipped with a rotating-arm apparatus. Floating squared section cylinders with rounded edges and aspect ratios of L/D = 2.0 were elastically supported by a set of linear springs in order to provide low structural damping to the system. Two different incidence angles were tested, namely 0 and 45 degrees. The Reynolds numbers covered the range from 2, 000 to 30, 000. The aim was to understand the flow-induced vibrations around single columns, gathering information for further understanding the causes for the Vortex-Induced Motions in semi-submersible and TLP platforms. Experiments on circular and squared sections cylinders (without rounded edges) were also carried out to compare the results with the rounded square section cylinders (with rounded edges). The amplitude results for in-line, transverse and yaw amplitude for 0-degree models showed to be higher for squared section cylinders compared to those for the rounded square section cylinders. No significant difference between the 45-degree models was observed. The results of ratio between frequency of motion in the transverse direction and natural frequency in still water confirmed the vortex-induced vibration behavior for the squared and rounded square section cylinders for 45-degree incidence; and also the galloping characteristics for 0-degree incidence cases. The rounded effect on the square section cylinders showed to be

Experiments regarding free-end effects on vortex-induced vibration (VIV) of floating circular cyl... more Experiments regarding free-end effects on vortex-induced vibration (VIV) of floating circular cylinders with low aspect ratio were carried out in a towing tank. Four cylinders with low aspect of ratio, L/D = 2 (Length / Diameter) were tested with different free end corner shape types, namely by the relation between chamfer rounding radius (r) divided by the radius of cylinder (R) (r/R = 0.0, 0.25, 0.5 and 1.0). For the initial case, r/R = 0.0 represents flat tip and r/R = 1.0 the hemispherical tip. The aims were to understand the effect of different free-end types on VIV behavior of cylinders. The floating circular cylinders, i.e. unit mass ratio m* = 1(structural mass/displaced fluid mass) were elastically supported by a set of linear springs to provide low structural damping on the system and allow six degrees of freedom. The range of Reynolds number covered 3, 000 ≤ Re ≤ 20, 000. To conclude, cylinder with r/R=0.25, shows lower amplitudes in transverse direction. The same occurs for the cylinder r/R = 0, but for amplitudes of vibration in in-line direction. Behaviors of the vibration frequencies in in-line and transverse direction don't have significantly differences. Regarding force coefficient, flat tip cylinder (r/R=0) presents higher values compared to the others however, for the lift coefficient, results converge in similar values for the same velocities that were observed higher transverse amplitudes. The visualization experiments show an expressive reduction of the recirculation bubble for r/R = 0.5 model compared with the flat tip, can therefore justify the lower values for this model obtained in draft amplitudes and drag coefficient compared with the flat tip model.

Concomitant Vortex-Induced Vibrations (VIV) and axial motion prescribed to the top of a vertical ... more Concomitant Vortex-Induced Vibrations (VIV) and axial motion prescribed to the top of a vertical and flexible cylinder were experimentally investigated in a towing tank facility. The technological motivation for such experimental campaign is the risers dynamics scenario. Vertical motions are applied to the top of such structures, causing geometric stiffness modulations. Such modulation cause modulations in the natural frequencies which, in turn, may affect VIV phenomenon. Both statistical and modal analyses were carried out, showing that the f t : f N,1 = 2 : 1 top-motion ex-citation enhances the crosswise vibrations for all free-stream velocities tested. Focus was put on the analysis of the modal-amplitude time-histories corresponding to the first-vibration mode. It was also observed that for large reduced velocities, the Fourier spectra obtained from these time-histories are broadbanded. Still considering the latter range of free-stream velocity, the presence of top-motion enhances the characteristic oscillation amplitude if compared to the non top-motion excited case.

An experimental investigation of Vortex-Induced Vibration (VIV) on a long and vertical flexible c... more An experimental investigation of Vortex-Induced Vibration (VIV) on a long and vertical flexible cylinder is presented and some results are analyzed. Optical instru-mentation allows to directly measure the cartesian coordinates of 43 targets placed along the model. At each instant , modal decomposition based on Galerkin's scheme is applied and, then, modal-amplitude time-histories are obtained. The modal-amplitude time-histories are analyzed , allowing to identify similarities in the modal response at the first and the second natural modes lock-in. Jumps and phase-shifts transitions are also obtained from the analysis of the modal-amplitude time-histories. Finally, it is also found that modal response amplitudes, plotted as functions of the corresponding, reasonable collapses onto the same curve, providing an experimental evidence that the modal response are the same for different excited modes.

Experiments regarding the Vortex-Induced Motions (VIM) phenomenon of deep-draft Semi-Submersible ... more Experiments regarding the Vortex-Induced Motions (VIM) phenomenon of deep-draft Semi-Submersible (SS) platforms were carried out in a towing tank in order to investigate the effects of two types of columns, i.e., rounded square sections and circular columns. Two different models at scale of 1:100 were built based on the typical geometry of a SS with four columns and each model with the same arrangement of four pontoons in closed configuration. The models were elastically mounted by means of a set of linear springs, which provide low structural damping. The range of towing velocities in the reduced velocity was 4 up to 25 covered Reynolds numbers from 7,000 to 80,000. Three angle of incidences for each model were tested, namely 0, 22.5 and 45 degrees related to the current incidence. Different VIM response was observed due to the column design. The maximum amplitude in the transverse direction was practically the same for both column designs, but occurring at different incidence angles, i.e., 0-degree for the SS with circular columns and 45 degrees for the one with rounded square columns. In terms of the yaw motions, the highest amplitudes were observed for the SS with circular columns for 22.5 and 45-degrees incidence, particularly around the reduced velocity based on yaw natural period equal to 8. On the other hand, yaw amplitudes increased continuously for the SS with rounded square columns, with no substantial difference related to the heading. These results confirmed that the heading effects on the VIM of SS strongly depend on the column design. The results presented herein are part of a more complete campaign of experiments carried out within the VIM JIP which is managed by MARIN and USP. Within this campaign, the effect of roughness on VIM response was also investigated using the same models. Those results can be found in a complementary paper in this conference.

Experiments on the Vortex-Induced Motions (VIM) acting on deep-draft Semi-Submersible (SS) platfo... more Experiments on the Vortex-Induced Motions (VIM) acting on deep-draft Semi-Submersible (SS) platforms were carried out in a towing tank in order to investigate the roughness effects on the surface of two different 1:100-scale models. Both models were built based on the typical SS geometry with four columns and each model with the same arrangement of four pontoons in closed configuration. The models were elastically supported by means of a set of linear springs, which also provided low structural damping, and then submitted to a range of towing velocities in the reduced velocity range from 4 up to 25 covering the Reynolds numbers from 7,000 up to 80,000. Three different levels of roughness were tested: 1) bare models; 2) intermediate level of roughness and 3) high level of roughness. For the SS with rounded square columns, no significant differences were observed in terms of the VIM response For the SS platform with circular columns, some difference on the transverse and yaw motions were observed. These results seem to confirm two important aspects: the important role played by the separation point near the rounded corners of the rounded-square columns and the roughness effect on the boundary layers of the circular columns at the Reynolds numbers just below the critical regime. The results presented herein are part of a more complete campaign of experiments carried out within the VIM JIP which is run by MARIN and USP. Also within this campaign, the effect of column design on VIM response was investigated using the same models. Those results can be found in a complementary paper in this conference.

Experiments with yawed cylinders were carried out at the recirculating water channel facility of ... more Experiments with yawed cylinders were carried out at the recirculating water channel facility of NDF-Fluids and Dynamics Research Group, University of São Paulo, aiming to study the effects of the inclination in the force coefficients and in the vortex-shedding frequency. The cylinders were inclined 0, 10, 20, 30 and 45 degrees. Forces and Strouhal number were evaluated by using the Independence Principle (IP), which states a dependence of the flow characteristics only on the component of velocity normal to the cylinder axis. The results showed dependence on the inclination angle, even if the IP is considered. Differences were also observed if the cylinder was inclined in upstream or downstream directions.

Experiments regarding vortex-induced vibration on floating circular cylinders with low aspect rat... more Experiments regarding vortex-induced vibration on floating circular cylinders with low aspect ratio were carried out in a recirculation water channel. The floating circular cylinders were elastic supported by a set of linear springs to provide low structural damping on the system. Eight different aspect ratios were tested, namely 0.2, 0.3, 0.4, 0.5, 0.75, 1.0, 1.5 and 2.0. These aspect ratios were selected to cover the aspect ratio range of the main offshore circular platforms, such as spar and monocolumn. The aims were understanding the VIM of such platforms; due to this, the cylinders were floating, or. The range of Reynolds number covered 2,800 < Re < 55,400. The amplitude results showed a decrease in amplitude with decreasing aspect ratio in both directions. The frequency results confirm a different behavior for cylinders with ; in these cases, the cylinder free-end effects were predominant. The resonant behaviour was no longer observed for. The decrease in Strouhal number with decreasing aspect ratio is also verified. All the results presented here complement the work presented previously for stationary circular cylinder with low aspect ratio presented by Gonçalves et al. (2013), Experimental Study on Flow around Circular Cylinders with Low Aspect Ratio, OMAE2013-10454.

Experiments with a flexible and submerged cylinder were carried out to investigate fundamental as... more Experiments with a flexible and submerged cylinder were carried out to investigate fundamental aspects of risers dynamics subjected to harmonic excitation at the top. The flexible model was designed aiming a high level of dynamic similarity with a real riser. Vertical motion, with amplitude of 1% of the unstretched length, was imposed with a device driven by a servomotor. Four values of the ratio between the exciting frequencies and the first eigenfrequency were investigated, namely f t : f N,1 = 1 : 3; 1 : 1; 2 : 1 and 3 : 1. Cartesian coordinates of 43 monitored points positioned all along the span were experimentally acquired by using an optical tracking system. A simple Galerkin's scheme applied for modal decomposition, combined with standard Mathieu chart analysis, led to the idenfica-tion of parametric resonances. A curious finding of this paper is that the Mathieu instability may simultaneously occur in more than one mode, leading to interesting dynamic behaviors, also revealed through standard power spectra analysis and displace

The flow around free-surface piercing, low aspect-ratio circular cylinder is investigated by mean... more The flow around free-surface piercing, low aspect-ratio circular cylinder is investigated by means of unsteady Reynolds averaged Navier-Stokes (URANS) calculations together with verification procedures and comparison with small-scale experimental and Particle Image Velocimetry results. A two-phase interface capturing model is used to handle the free-surface flow, together with k-ω SST turbulence model. We investigate physical and modeling aspects of this problem in order to gain more knowledge about the interaction of free-surface and free-end effects so that this mechanism is better understood and taken into account when modeling the problem in engineering-applied situations , such as the vortex induced motion of spars, tension-leg platforms and semi-submersibles. The case herein presented is a captive, low aspect-ratio * Visiting MARIN as invited researcher. cylinder (L/D = 2.0) with flow velocity corresponding to Reynolds and Froude numbers (both based on diameter) of Re = 4.3 × 10 4 and Fn D = 0.31, respectively. We will show that appreciable free-surface effects are perceived on the flow, but with dominance of free-end effects, at least in terms of forces. Furthermore, we investigate different boundary conditions that would represent this free-surface problem to show that the separation of viscous and free-surface effects is not valid in this instance. Therefore, the interaction between viscous and free-surface effects is also tangentially investigated. In order to support our conclusions, we will show forces with uncertainty estimation and field variables obtained with different modeling strategies, unveiling physical and numerical aspects of this problem .

Experiments were carried out in a recirculating water channel regarding the flow around stationar... more Experiments were carried out in a recirculating water channel regarding the flow around stationary circular cylinders with low aspect ratio piercing the water free surface. Eight different aspect ratios were tested, namely , 0.2, 0.3, 0.5, 0.75, 1.0, 1.5 and 2.0; this range corresponds to aspect ratio related to circular offshore systems, such as spar and monocolumn platforms. Force was measured using a six degree-of-freedom load cell and Strouhal number is inferred through the transverse force fluctuation frequency. The range of Reynolds number covers 10,000 < Re < 50,000. PIV measurements were performed in some aspect ratio cases, namely 0.3, 0.5, 1.0 and 2.0 for Reynolds number equal to 43,000. The results showed a decrease in drag force coefficients with decreasing aspect ratio, as well as a decrease in Strouhal number with decreasing aspect ratio. The PIV showed the existence of an arch-type vortex originated in the cylinder free end.

Vortex Self-Induced Vibrations (VSIV) of a reduced scale model of a catenary riser are experiment... more Vortex Self-Induced Vibrations (VSIV) of a reduced scale model of a catenary riser are experimentally investigated. The riser model dynamics was assessed with a submerged optical motion capture system and significant VSIV were revealed as result of oscillatory vertical motion imposed to the top. Such a behavior recovers similar ones reported in the technical literature by other authors and resembles previous fundamental studies, by Sumer and Fredsoe, with rigid cylinders forced to oscillate in a plane and elastically mounted in the transversal direction. The present experiments are preliminary and pertain to a much more comprehensive experimental set, within a research project aimed at studying the nonlinear dynamic behavior of risers, through experimentally validated analytical and numerical, nonlinear reduced-order models.

Recent studies suggest that the vortex-induced motions (VIM) of a semi-submersible found in model... more Recent studies suggest that the vortex-induced motions (VIM) of a semi-submersible found in model tests over-predicts the response in the field, leading to an excessively conservative design of the mooring and riser system. Within the Vortex Induced Motion Joint Industry Project (VIM JIP), run by MARIN and University of São Paulo (USP), possible reasons for this over-prediction are investigated with model tests and CFD. The results presented herein are part of the model tests carried at MARIN, where damping, mass ratio, draft variations and sinusoidal tow velocities were put to the test. The model was a generic bare hull semi submersible with four rounded square columns at scale 56.5. The experimental setup allowed only 3DoF, with heave, roll and pitch restricted by an air bearing plate. A classical VIM tow test campaign was carried out as a benchmark, against which different variations are compared. A newly developed active damping system was used to apply linear damping and virtual mass to the model. This system has been verified with dedicated tests, and its performance is discussed. The tested mass ratios show minor impact on VIM response. Different damping ratios were tested for different drafts. It is found that adding reasonable levels of linear damping (1) significantly reduces VIM amplitude, from 25% reduction up to VIM suppression, (2) narrows the reduced velocities at which VIM lock in occurs and (3) shifts the response peak to higher reduced velocities. All these effects are present in the field measurements published by Ma et al. (2013). The presented results suggest that damping is a strong candidate to explain the differences between model test VIM and field observations. With the active damping system, future VIM model test campaigns can include the damping coming from risers and mooring, assuming that a reasonable estimation of the damping level is available.