0
Review Article

Adaptive Composite Marine Propulsors and Turbines: Progress and Challenges

[+] Author and Article Information
Yin Lu Young

Professor
Department of Naval Architecture and
Marine Engineering,
University of Michigan,
Ann Arbor, MI 48109
e-mail: ylyoung@umich.edu

Michael R. Motley

Assistant Professor
Department of Civil and
Environmental Engineering,
University of Washington,
Seattle, WA 98195
e-mail: mrmotley@u.washington.edu

Ramona Barber

Department of Civil and
Environmental Engineering,
University of Washington,
Seattle, WA 98195
e-mail: rbbarber@u.washington.edu

Eun Jung Chae

Department of Naval Architecture and
Marine Engineering,
University of Michigan,
Ann Arbor, MI 48109
e-mail: imchae@umich.edu

Nitin Garg

Department of Naval Architecture and
Marine Engineering,
University of Michigan,
Ann Arbor, MI 48109
e-mail: gargn@umich.edu

1Corresponding author.

Manuscript received January 26, 2016; final manuscript received August 22, 2016; published online October 3, 2016. Assoc. Editor: Xiaodong Li.

Appl. Mech. Rev 68(6), 060803 (Oct 03, 2016) (34 pages) Paper No: AMR-16-1010; doi: 10.1115/1.4034659 History: Received January 26, 2016; Revised August 22, 2016

In this paper, the advantages, state-of-the-art, and current challenges in the field of adaptive composite marine propulsors and turbines are reviewed. Adaptive composites are used in numerous marine technologies, including propulsive devices and control surfaces for marine vessels, offshore platforms, unmanned surface and underwater vehicles, and renewable energy harvesting devices. In the past, most marine propulsors and turbines have been designed as rigid bodies, simplifying the design and analysis process; however, this can lead to significant performance decay when operating in off-design conditions or in spatially or temporally varying flows. With recent advances in computational modeling, materials research, and manufacturing, it is possible to take advantage of the flexibility and anisotropic properties of composites to enable passive morphing capabilities to delay cavitation and improve overall energy efficiency, agility, and dynamic stability. Moreover, active materials can be embedded inside composites to enable energy harvesting, in situ health and condition monitoring, mitigation and control of flow-induced vibrations, and further enhancements of system performance. However, care is needed in the design and testing of adaptive composite marine propulsors and turbines to account for the inherent load-dependent deformations and to avoid potential material failures and hydroelastic instabilities (resonance, parametric excitations, divergence, flutter, buffeting, etc.). Here, we provide a summary of recent progress in the modeling, design, and optimization of adaptive composite marine propulsors and turbines, followed by a discussion of current challenges and future research directions.

Copyright © 2016 by ASME
Your Session has timed out. Please sign back in to continue.

References

Figures

Grahic Jump Location
Fig. 1

Predicted performance for an adaptive turbine (left) and propeller (right) considering changing TSR or J, respectively, as a function of inflow velocity or rotational speed only. The results are based on data from Refs. [32,40].

Grahic Jump Location
Fig. 5

BEM representation of the no-load and loaded geometry of composite propellers 5471 (no-twist) and 5479 (self-twist) at the design condition of J = 0.66 at n = 909 rpm in uniform flow. (Reprinted with permission from Young [31]. Copyright 2008 by Elsevier.)

Grahic Jump Location
Fig. 7

Predicted undeformed and deformed geometry (magnified by ten times) of one of the blades for propellers 5471 (no-twist) and 5479 (self-twist) at the design condition of J = 0.66 and n = 909 rpm in uniform inflow obtained using the coupled 3D BEM–FEM solver presented in Ref. [31]. The results are based on data presented in Refs. [30] and [31].

Grahic Jump Location
Fig. 8

Comparisons of the predicted (using the coupled 3D BEM–FEM solver presented in Ref. [31]) and measured (by NSWCCD, as presented in Ref. [177]) open water thrust coefficient (KT) and propeller efficiency (η) for propellers 5471 (no-twist) and 5479 (self-twist) from the first propeller set on the left, and propellers 5474 (no-twist) and 5475 (self-twist) from the second propeller set on the right. The torque coefficients (KQ) are not shown since it can be calculated based on KT and η. (The left figure is reprinted with permission from Young [31]. Copyright 2008 by Elsevier. The right figure is based on data presented in Refs. [179] and [180]).

Grahic Jump Location
Fig. 2

Lift (CL), drag (CD), and pitching moment (CM) coefficients for a range of angle of attacks (α) for type II SS versus CFRP00 hydrofoils on the left, and SS versus CFRP30 hydrofoils on the right, for 0.2×106≤Re≤1.0×106. The test range only included conditions where the total hydrodynamic lift force is less than 1 kN to avoid material failure. Hence, the maximum angle of attack was limited for the higher Re cases. The dependence of the hydrodynamic load coefficients on Re indicate load-dependent hydroelastic response. The results are based on data presented in Ref. [171].

Grahic Jump Location
Fig. 3

Nondimensionalized plot of the lift (LS3/EIb)versus bending deformation (h/b) on the left and moment (MS/GJ) versus twisting deformation (θ) on the right for the SS, Al, CFRP00, and CFRP30 hydrofoils for 0.2×106≤Re≤1.0×106 and α≤10 deg. The results are based on data presented in Ref. [171].

Grahic Jump Location
Fig. 4

Left: Photograph of the model-scale 0.61 m (24 in.) diameter no-twist composite propeller 5474. Middle: Illustration of the wake screen used to generate the four-cycle wake and blade tip deformation measurement locations in the cavitation tunnel test. Right: Axial inflow velocity distribution at the propeller plane generated by the four-cycle wake in the cavitation tunnel. Details of the experimental setup and results can be found in Ref. [177].

Grahic Jump Location
Fig. 6

Comparisons of the predicted (using the coupled 3D BEM–FEM solver presented in Ref. [31]) and measured (by NSWCCD, as presented in Ref. [177]) open water thrust coefficient (KT), torque coefficient (KQ), and efficiency (η) for propellers 5471 (no-twist) and 5479 (self-twist) in uniform inflow at two different rotational velocities (n). (Reprinted with permission from Young [31]. Copyright 2008 by Elsevier.)

Grahic Jump Location
Fig. 9

Predicted in-air (dry) versus in-water (wet) modal frequencies, and mode shapes of propellers 5471 (no-twist) and 5479 (self-twist). The no-load geometry is shown in dashed lines in the lower figures, while the deformed modal geometry is shown in solid lines. Note that the measured first modal frequencies in water are shown in open symbols on the top plot. (Reprinted with permission from Young [31]. Copyright 2008 by Elsevier.)

Grahic Jump Location
Fig. 10

BEM representation of the no-load and loaded geometry of composite propellers 5474 (no-twist) and 5475 (self-twist) at the design condition of J = 0.66 and n = 780 rpm in uniform inflow. The results are based on data presented in Refs. [179,180].

Grahic Jump Location
Fig. 11

Predicted undeformed and deformed geometry (magnified by ten times) of one of the blades for propellers 5474 (no-twist) and 5475 (self-twist) at the design condition of J = 0.66 and n = 780 rpm in uniform inflow obtained using the coupled 3D BEM–FEM solver presented in Ref. [31]. The results are based on data presented in Refs. [179,180].

Grahic Jump Location
Fig. 12

Comparison of the measured performance curves (left) and cavitation bucket(right) of the model-scale no-twist composite propeller 5474 and the self-twist composite propeller 5475 at U = 11 ft/s at varying J values in both uniform inflow andfour-cycle wake conditions. The results are based on data presented in Refs.[177,179,180].

Grahic Jump Location
Fig. 13

LES predictions of contours of fluctuating pressure and total velocity vectors on a lateral plane cut (left), pressure contours on the face side (middle), and back side (right) of a five-bladed, zero skew, 0.3048 m diameter aluminum propeller 4381 in a low-amplitude event (top) and high-amplitude event (bottom) at J = −1. The results are based on data presented in Ref. [113].

Grahic Jump Location
Fig. 14

FEM predictions of contours of bending (σ22) and shear (σ12) stress on the face and back side of a five-bladed, zero skew, 0.3048 m diameter aluminum propeller 4381 during a low-amplitude (Rev = 290) and high-amplitude (Rev = 300) event at J = −1. The results are based on data presented in Ref. [113].

Grahic Jump Location
Fig. 16

Static/dynamic divergence and flutter velocity and flutter frequency as a function of the relative mass ratio for a NACA16-010 foil in air–freon mixture with experimental data from Ref. [225] (top row), and for a NACA16-012 hydrofoil in water with experimental data from Ref. [120] (bottom row). The elastic axis (EA) is located at 0.39c for a NACA16-010 foil and 0.25c for a NACA16-012 hydrofoil. (Reprinted with permission from Chae et al. [61]. Copyright 2013 by AIP Publishing.)

Grahic Jump Location
Fig. 15

In-water to in-air natural bending and twisting frequency ratios (f¯h=fh*/fh and f¯θ=fθ*/fθ) and fluid bending and twisting damping coefficients (ζf,h and ζf,θ) with varying Uθ¯ and μ of a cantilevered, rectangular, homogeneous NACA0015 hydrofoil. The results are based on data presented in Ref. [60].

Grahic Jump Location
Fig. 17

Probability distribution of equivalent unidirectional fiber angles as a result of Monte Carlo analysis of random variations in material stiffness properties and laminate fiber angles. (Reprinted with permission from Young and Motley [282]. Copyright 2011 by International Symposiums on Marine Propulsors.)

Grahic Jump Location
Fig. 18

Normalized tip deflection (left) and change in tip pitch angle (right) as a function of propeller advance speed for an adaptive composite propeller with θeq=3.0 deg, 5.0 deg, and 7.0 deg. (Reprinted with permission from Young and Motley [282]. Copyright 2011 by International Symposiums on Marine Propulsors.)

Grahic Jump Location
Fig. 19

Delivered thrust as a function of propeller advance speed for an adaptive composite propeller blades with θeq=3.0 deg, 5.0 deg, and 7.0 deg. (Reprinted with permission from Young and Motley [282]. Copyright 2011 by International Symposiums on Marine Propulsors.)

Grahic Jump Location
Fig. 20

Plot depicting how the number of function evaluations scales with the number of design variables for the multidimensional Rosenbrock function. While the gradient-free methods (NSGA2 and ALPSO) scale poorly, the gradient-based methods (SNOPT and SLSQP) scale linearly with the number of design variables. The results are based on data presented in Ref. [323].

Grahic Jump Location
Fig. 21

Figure depicting the single-point optimization results for tapered NACA 0009 hydrofoil at CL=0.75. Results show the difference in optimization with 18, 48, 200, and 720 FFD control points for a simple, unswept, hydrofoil at Re=1.0×106 and M = 0.05. Note that the total number of design variables includes FFD control points and three spanwise twist design variables (defined at the root, the midspan, and the tip of the foil) in each case. The optimization problem specified was to minimize CD for a given CL of 0.75 and cavitation number of 1.6. (A) The difference in the CD values was found to be very small, with the maximum difference in CD values just less than 0.7%. (B) The optimization results converged to similar geometries (particularly in terms of the twist and camber distribution), but with significant differences in the sectional geometry and pressure profile near the root for the case with 18 FFD control points and 720 FFD control points. Black horizontal line represents the constraint on cavitation number. The optimized solution is practically the same for 200 and 720 FFD control points, except the region very close to the root section. (Reprinted with permission from Garg et al. [322]. Copyright 2015 by Society of Naval Architects and Marine Engineers (SNAME).)

Grahic Jump Location
Fig. 22

Performance of the original NACA 0009 hydrofoil compared to the multipoint hydrostructural hydrofoil optimized with cavitation, maximum stress, and manufacturing constraints at Re=1.0×106 and V = 12.4 m/s. The multipoint optimized geometry shows an increase in overall efficiency of 8.5% compared to the original NACA 0009 hydrofoil and an increase in cavitation speed by 45% for an assumed submergence depth of 1 m, while meeting the stress and manufacturing constraints. The results are presented in three rows. First row presents results for CL=0.65 (design point); on the left-hand side, Cp (pressure coefficient) contour plots on the suction side are displayed for NACA 0009 hydrofoil and the multipoint optimized foil. Gray contour region along the leading edge of the tapered NACA 0009 foil displays the area with −Cp≥σ. A reduction of 8.3% in CD can be noticed for the multipoint optimized foil, over the original NACA 0009 hydrofoil. On the right-hand side, the figure shows the sectional −Cp plots and the geometry profile of the foil at two sections along the span of the hydrofoil. The dark (blue) solid line represents the NACA 0009 hydrofoil and the light (green) solid line corresponds to the multipoint hydrostructural optimized foil with manufacturing constraint. The horizontal line represents the constraint on cavitation number (when −Cp=σ=1.6). As noted, there is significant decrease in maximum −Cp from the NACA 0009 to the optimized hydrofoil. Difference in the sectional shape as fraction of the span (z/s) between the original and optimized foil is also shown in the bottom of each subplot. Second row presents results for CL=0.75 (stress critical point); on the left-hand side, 1.1σv/σs,f (stress constraint) contour plots on the suction side are displayed for NACA 0009 hydrofoil and the multipoint optimized foil. The light (yellow) colored region near the root of the tapered NACA 0009 foil displays the area where the stress constraint was not satisfied. Third row presents results for CL=−0.15 (cavitation critical on pressure side); on the left-hand side, Cp (pressure coefficient) contour plots on the pressure side are displayed for NACA 0009 hydrofoil and the multipoint optimized foil. A minor increase of 1.5% in CD can be noticed for the multipoint optimized foil, over the original NACA 0009 hydrofoil. The results are based on data presented in Ref. [329].

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In