commercial vessels with box-shaped. superstructures may be made by rounding the comers, leading to reductions in drag. It is found that the rounding of sharp comers can be beneficial, particularly for box - shaped bluff bodies, Hoemner [27] and Hucho (28) However, a rounding of at least r/B3=0.05 (where r is the rounding radius and Bs is the breadth of the superstructure) is necessary before there is a significant impact on the drag. At and above this rounding, decreases in air drag of the order of 15% — 20% can be achieved for rectangular box shapes, although it is ly such decreases can be achieved with shapes which are already fairly streamlined. Its noted that this procedure would conflict with design for production, and the use of “box type’ superstructure modules. Investigations by Molland and Barbeau [26] con the superstucture drag of large fast ferries indicated a reduction in air drag coefficient (based on frontal area) from about 0.8 for a relatively bluff fore end down to 0.5 for a well streamlined fore end, a decrease in air drag of about 38%. If this change were applied to the ferry in Table 2, this would lead to a decreas power of about 1.5%. 4.5 PROPULSIVE EFFICIENCY 4.5(a)_Propulsive efficiency ‘The components of quasi propulsive coefficient (75) may be written To = Th. Th Tk © where 1p is the open water efficiency of the propeller, ‘my_is the hull efficiency and 1 is the relative rotative efficiency. ‘me takes account of the differences between the propeller in the open water condition and when behind the hull, and lies typically between 0.98 to 1.02. ‘mq, takes account of the interaction between the hull and propeller and is defined as: ony © where ¢ is the thrust deduction factor and wr the wake fraction. mi lies typically between 1.0 and 1.25 for displacement ships. The formula indicates how changes in thrust deduction ({) due, for example, to the presence of a rudder or other device will influence overall propeller efficiency. Similarly, the influence of wake fraction (Wy) can be seen and quantified. tp is the open water efficiency of the propeller and will depend on the propeller parameters and operating conditions. 4.5(0) Individual components of propeller open water efficiency: For a fixed set of propeller parameters, 9 can be considered as being made up of: = te tm % o | Printed by DOME e147 R215-01 where 1, is the ideal (or axial) efficiency, n, accounts for losses due to fluid rotation induced by the propeller and ry accounts for losses due to blade friction drag, Dyne [29], [30], Molland et al [31]. Bear Figure 3: Propeller blade element diagram An investigation has been carried out to determine the likely values of these three components of efficiency. Blade element-momentum theory was used, [31], (32] and, based on the blade element diagram, Figure 3, it can be shown that: 1 7" Tray i) 1,= (-@) ” = tang 10) ™ an(O+7) a where a and a’ are the axial and rotational inflow factors, derived from momentum considerations and corrected for finite number of blades using Goldstein correction factors (31]. ‘The investigation used a propeller with a pitch ratio P/D = 1.0, BAR = 0.700 and 4 blades for a range of J values, hence thrust loading, Cy,as shown in Table 4. ‘Table 4: Range of thrust loadings investigated J | c 2 ‘04 0.35 10.32 0.45 5.34 0.55 2.94 0.65 1.64 0.75 0.86 0.85 0.38 0.90 0.20 ‘The thrust loading coefficient Cis defined, and related to Kz/F, as follows: o-—t— - 48 ay 0592 yat For a fixed pitch, decrease in J Wend oan inrease in thrust loading Ky/’, or Cr. The results of the investigation are shown in Figure 4. Itcan be noted that the ro curve closely replicates the02 | oo 01 02 02 04 08 04 07 00 00 10 oper vance coement 3 Figure 4: Components of propeller efficiency level of that for a Wageningen B4.70_ propeller Working from a low J (high thrust loading Cx) to higher J (lower thrust loading) it is seen that the rotational losses decrease, the frictional losses increase and there is a significant decrease inthe axial losses. At a typical design condition of say J = 0.75 (Cy = 0.86) it is seen that the losses are typically 60% axial, 10% rotational and 30% frictional. At a lower J, higher thrust loading, of J = 0.35 (Cr = 10,32) the losses are typically 80% axial, 15% rotational and 5% friction Thi of ne, Equation (7) indicates where pote @ me _ ‘Sa using the minimum blade thickness consistent with strength considerations and minimum blade area consistent with cavitation requirements. Reduction in friction can also be achieved by decreasing the local inflow velocity (by reducing revolutions) with an appropriate increase in diameter. Friction loss may also be reduced by reducing revolutions and increasing pitch, although this will lead to increased rotational losses which may have to be compensated by stators downstream, Surface finish influences friction, particularly during operation. Indications of the Potential savings were derived using the propeller blade element-momentum theory with J'= 0.75, P/D = 1.0 and BAR = 0.700. It was found that a 10% change in blade thickness led to about 1% - 2% change in 1, @ decrease in BAR from 0.700 to 0,600 led to an increase in 7 of about 2% and an increase in drag coefficient of 20% due to say roughness and fouling led to a decrease in mp 0f 2% - 4%. Gi) Recovery of rotational losses can be achieved in various ways. The most effective way is to ‘employ contra-rotating propellers, where much of the flow rotation loss can be removed. In the absence of such propellers the most common way is by the use of ‘pre-and post-swirl stators. The rudder (downstream of ‘a propeller) acts as a post-swirl stator, but also blocks rad by Ose ameP 8825-1 the flow, resulting in pressure (axial) losses which can negate much of the rotational savings. Pre-swil (upstream) stators can entail fins, while post swirl downstream can entail fins attached to the rudder. ‘Asymmetric stems can be employed which put pre- rotation into the propeller inflow. Levels of savings in power for these various devices are discussed in ion 4,6, It should be noted that pre-and post-switl 8 and asymmetric stems ean increase the resistance, hhence reducing some of the effective savings in propeller effcioney. nz Axial losses are by far the largest, Figure 4, Theory and practice indicate that an increase in ameter with commensurate changes in P/D and rpm il lead to improvements in axial efficiency (together with some improvement in frictional efficiency) Maximising diameter is therefore of fundamental importance. Accelerating ducts may be used to advantage in conditions of high thrust loading, although duct friction will tend to remove any savings at ‘moderate or low thrust loadings. An upstream semi duct has been employed to improve the axial efficiency by directing part of the frictional wake to inside the propeller diameter. study was carried ‘whose particulars fn load draught. The ‘where operation is less than the design load case. This can create a significant power penalty. For example, when this range of diameters is applied to the 145m vessel in Table 1, then the propeller diameter ‘would be between 6.9m and 7.9m. If this were transformed into propeller efficiency improvements, then the order of increases are shown in Figure 5, which include changes in wr and ¢ (hence 7x ) with change in diameter. With the propeller efficiency 1p, Equation (5), improving from 0.726 to 0.765, there is 1, pprovement of some 5%. To be able to incorporate Proper cio 2 0700 107 08 09 70.74 72,79 747578 77 78 79 80 Proper lamer) Figure 5: Change in propeller efficiency mo ‘with change in propeller diameter