Components of Hull Resistance ive. D2 sp Mp—Ma=p ff wu +udcay G5) -ia-h This rate of change of momentum can be equated to the forces on the fiuid in the control box and, neglecting friction on the walls, these are R (hull resistance), Fa (pressure force on plane A) and Fs (pressure force on plane B). Therefore, Mg — Ma=-R+Fa-Fp Bemoulli’s equation can be used to derive expressions for the pressures at A and B, hence, for forces Fy and Fx Pal H=445 p 2 where APis the loss of pressure in the boundary layer and AP/p is the corresponding loss in total head. If atmospheric pressure is taken as zero, then ahead of the model, P4 = 0 on the free surface where z = 0 and the constant term is H = }U?. Hence, Py AP gat UtuP te tw] t ect (3.6) AP 1 Pr=-p {ecs + > t 3 [2U" bu pv? + wl (3.7) On the upstream control plane, br o 42 0 ran ff Padzdy--ee f fe deay— yee [ eo} Spgbh®. (3.8) -ba-h -bamh On the downstream control plane, using Equation (3.7), bn op AP 1 ; i= f J Padzay =-p f J {ecw + gta we +7 +wifec dy bh =b/2-A ° LP bn op ino ayes | Wie f faracdy-2 f fev rw+ w]dz dy. “ia iach aback (3.9) ‘The resistance force R = Fx — Fs — (Mz ~ Ma) Substituting for the various terms from Equations (3.5), (3.8) and (3.9), 1 a2 Wo R= hive f bayd yf fo surauyded}o ff ardcay. (6.10) “a bac “ba In this equation, the first two terms may be broadly associated with wave pattern drag, although the perturbation velocities v, w and u, which are due mainly to wave orbital velocities, are also due partly to induced velocities arising from the viscous shear in the boundary layer. The third term in the equation is due to viscous drag. The use of the first two terms in the analysis of wave pattern measurements, and in formulating a wave resistance theory, are described in Chapters 7 and 9. 1920 Ship Resistance and Propulsion [Link] Identification of Induced Drag A ficticious velocity component u’ may be defined by the following equation: PESO tu +e ul + gen 3? G1) where w’is the equivalent velocity component required for no head loss in the bound- ary layer. This equation can be compared with Equation (3.6). u’ can be calculated from Ap since, by comparing Equations (3.6) and (3.11), yo(U+uy = tp(U +uy + dp, then, a oD te 1 2ay 4), 2 R= joe | darrze f fo abe —w?)de dy hath + ff {avs zoe—we)} deay @12) ‘The integrand of the last term (Ap + }p(w? — u?)} is different from zero only inside the wake region for which Ap + 0. In order to separate induced drag from wave resis- tance, the velocity components (u7, vy, w7) of the wave orbit motion can be intro- duced. [The components (u, v, w) include both wave orbit and induced velocities] The velocity components (u, vy, w1) can be calculated by measuring the free-surface wave pattern, and applying linearised potential theory. It should be noted that, from measurements of wave elevation ¢ and perturba tion velocities u, v, w over plane B, the wave resistance could be determined. How- ever, measurements of subsurface velocities are difficult to make, so linearised poten- tial theory is used, in effect, to deduce these velocities from the more conveniently measured surface wave pattern £. This is discussed in Chapter 7 and Appendix A2. Recent developments in PIV techniques would allow subsurface velocities to be measured; see Chapter 7. Substituting (uy, vy, wz) into the last Equation (3.12) for R, R=Ry+Rvy +R, where Rw is the wave pattern resistance bn a 1 1 2 Rw = 508 f tdy +50 f (i+ uf -ui)dedy, (G.13) oa “ban Ryis the total viscous resistance 1 2 Ry ff {r+ zou - » dzdy (3.14) and Ry is the induced resistance 1 ee Ri= 5 f [-d + w? — wh — uw’? + uj) dz dy. (3.15) aban For normal ship forms, R is expected to be small.Components of Hull Resistance 2 3.1.3 Systems of Coefficients Used in Ship Powering Two principal forms of presentation of resistance data are in current use. These are the ITTC form of coefficients, which were based mainly on those already in use in the aeronautical field, and the Froude coefficients [3.13] [Link] ITTC Coefficients The resistance coefficient is G.16) where Rr = total resistance force, S = wetted surface area of hull, V = ship speed, Cr = total resistance coefficient or, for various components, Cr = frictional resistance coefficient (flat plate), Cy = total viscous drag coefficient, including form allowance, and Cw = wave resistance coefficient. ‘The speed parameter is Froude number Vv vel and the hull form parameters include the wetted area coefficient Fr and the slenderness coefficient where V = immersed volume, L [Link] Froude Coefficients A basic design requirement is to have the least power for given displacement A and speed V. Froude chose a resistance coefficient representing a ‘resistance per ton dis- placement’ (ic. R/A). He used a circular notation to represent non-dimensional coef- ficients, describing them as®,©.©, @ and. The speed parameters include (4x was introduced for wave speed considerations). The hull form parameters include the wetted area coefficient22 Ship Resistance and Propulsion and the slenderness coefficient @= ve (now generally referred to as the length-displacement ratio). The resistance coeffi- cient is @- WOE (3.17) A®’ i.e. resistance per ton R/A (R = resistance force and A = displacement force, same units) ‘The 1000 was introduced to create a more convenient value for ©. The subscript for each component is as for Cr, Cr, etc.,.e.©z, ©x. The presentation R/A was also used by Taylor in the United States in his original presentation of the Taylor standard series [3.14] Some other useful relationships include the following: in imperial units, 427.1PE 5 av G18) (Pe = HP, A = tons, V = knots), and in metric units, (3.19) (Pr =kW, A = tonnes, V = knots, and using 1 knot = 0.5144 m/s) ‘The following relationships between the ITTC and Froude coefficients allow conversion between the two presentations: Cs = S/(V.L)"?. ®=siv", @=L/9"", Cs =@/" and Cr = (8 x 7/1000) x ©. It should be noted that for fixed A and V, the smallest implies least Pz, Equation (3.19), but since S can change for a fixed A, this is not true for Cr, Equation (3.16). Itis for such reasons that other forms of presentation have been proposed. These have been summarised by Lackenby [3.15] and Telfer [3.16]. Lackenby and Telfer argue that, from a design point of view, resistance per unit of displacement (i.e. R/A), plotted to a suitable base, should be used in order to rank alternative hull shapes correctly. In this case, taking R/A as the criterion of performance, and using only displacement and length as characteristics of ship size, the permissible systems of presentation are R Vv ) sanps 2) on Tz 2®) RL Q) Ayan R Vv “, (3) 5 oneither 75 or FrComponents of Hull Resistance Consistent units must be used in all cases in order to preserve the non- dimensional nature of the coefficients. For skin friction resistance, Lackenby points out that the most useful character- istic of ship size to be included is wetted surface area, which is one of the primary variables affecting this resistance. A coefficient of the form Cr = Rp/l/2 pSV? is then acceptable. In which case, the other components of resistance also have to be in this form. This is the form of the ITTC presentation, described earlier, which has been in common use for many years. It is also important to be able to understand and apply the circular coefficient notation, because many useful data have been published in©® — Fr format, such as the BSRA series of resistance tests discussed in Chapter 10. Doust [3.17] uses the resistance coefficient RL/AV? in his regression analysis of trawler resistance data. Sabit [3.18] also uses this coefficient in his regression analysis of the BSRA series resistance data, which is discussed in Chapter 10. It is useful to note that this coefficient is related to as follows RL L . Ayi = 2.4938 © x [ss] 3.20) 3.1.4 Measurement of Model Total Resistance [Link] Displacement Ships The model resistance to motion is measured in a test tank, also termed a towing tank. The first tank to be used solely for such tests was established by William Froude in 1871 [3.19]. The model is attached to a moving carriage and towed down the tank at a set constant speed (V), and the model resistance (R) is measured, Figure 3.8. The towing force will normally be in line with the propeller shafting in order to minimise unwanted trim moments during a run. The model is normally free to trim and to rise/sink vertically and the amount of sinkage and trim during a run is mea- sured. Typical resistance test measurements, as described in ITTC (2002) [3.20], are Carriage Dynamometer Model oF Water level Figure 3.8, Schematic layout of model towing test. 2324 Ship Resistance and Propulsion Environmental Carriage Hull model ronmental conditions ¥ ¥ ¥ Speed Resistance Temperature measurement, || dynamometer || measurement ‘rasurement: tachometeriprobe ‘devices yermometer y y ¥ Resistance / Tank water Model speed external temperature tow force y Signal conditioning and data acquisition | Data analysis Figure 3.9. Resistance test measurements. shown in Figure 3.9, An example of a model undergoing a resistance test is shown in Figure 3.10. Model speed is measured either from the carriage wheel speed (speed over ground), from the time taken for the carriage to travel over a known distance (speed over ground) or by a Pit6t-static tube or water speed meter attached to the carriage ahead of the model (speed through water). If the model speed remains constant dur- ing the course of a run and the tank water docs not develop any significant drift during the test programme, then all of these methods are equally satisfactory. Figure 3.10. Model resistance test. Photograph courtesy of WUMTIA and Dubois Naval Architects Ltd,Components of Hull Resistance The total hydrodynamic resistance to motion (R) is measured by a dynamometer. The dynamometer may be mechanical, using a spring balance and counterbalance weights, or electromechanical, where the displacement of flexures is measured by a linear voltmeter or inductor, or where the flexures have strain gauges attached. In all cases, a calibration procedure of measured output against an applied calibration force will take place before and after the experiments. Two-component (resistance and sideforce) or three-component (resistance, sideforce and yaw moment [torque]) dynamometers may be used in the case of yawed tests for assessing manoeuvring performance and for the testing of yacht models. Model test tanks vary in size from about 60 m x 3.7 m x 2 m water depth up to 300 m x 12 m x 3 m water depth, with carriage speeds ranging from 3 mys to 15 mys. A ‘beach’ is normally incorporated at the end of the tank (and sometimes down the sides) in order to absorb the waves created by the model and to minimise wave reflections back down the tank. This also helps to minimise the settling time between runs. Circulating water channels are also employed for resistance and other tests. In this case, in Figure 3.8, the model is static and the water is circulated at speed V. Such channels often have glass side and bottom windows, allowing good flow visualisation studies, They are also useful when multiple measurements need to be cartied out, such as hull surface pressure, skin friction or PIV measurements (see Chapter 7). Circulating water channels need a lot of power to circulate the water, compared say with a wind tunnel, given the density of water is about 1000 times that of air. For many years, models were made from paraffin wax. Current materials used for models include wood, high-density closed-cell foam and fibre reinforced plas- tic (FRP). The models, or plugs for plastic models, will normally be shaped using a multiple-axis cutting machine, Each model material has its merits, depending on pro- ducibility, accuracy, weight and cost. Model size may vary from about 16 m in a 60-m tank up to 9m in a 300-m tank. ‘The flow over the fore end of the model may be laminar, whilst turbulent flow would be expected on the full-scale ship. Turbulence stimulators will normally be incorporated near the fore end of the model in order to stimulate turbulent flow. Turbulence stimulators may be in the form of sand strips, trip wires or trip studs located about 5% alt of the fore end of the model. Trip studs will typically be of 3 mm diameter and 2.5 mm height and spaced at 25 mm intervals. Trip wires will typically be of 0.90 mm diameter. For further details see [3.20], [3.21], [3.22] and [3.23]. Corrections for the parasitic drag of the turbulence stimulators will normally be carried out; a detailed investigation of such corrections is contained in Appendix A of [3.24]. One common correction is to assume that the deficit in resistance due to laminar flow ahead of the trip wire or studs balances the additional parasitic drag of the wire or studs. In order to minimise scale effect, the model size should be as large as possi- ble without incurring significant interference (blockage) effects from the walls and tank floor. A typical assumption is that the model cross-sectional area should not be more than 0.5% of the tank cross-sectional area. Blockage speed corrections may be applied if necessary. Typical blockage corrections include those proposed by Hughes [3.25] and Scott [3.26], [3.27]. The correction proposed by Hughes (in its approximate 2526 Ship Resistance and Propulsion form) is as follows AV ‘Vo T-m—Fr (3.21) where AV/is the correction to speed, Fr, is the depth Froude number. Fr, vgh where /r is the tank water depth and m is the mean blockage ay © Ap Ly Ar’ where Lyy and V are model length and displacement, Aris the tank section area, and ais the model section area, ‘The temperature of the test water will be measured in the course of the experi- ments and appropriate corrections made to the resistance data. Viscosity values for water are a function of water temperature. As a result, Re and, hence, Cr vary with water temperature in a manner which can be calculated from published viscosity val- ues (see Tables ALI and Al.2 in Appendix A1). Standard practice is to correct all model results to, and predict ship performance for, 15°C (59°F), ic. Cras) = Crcry) + Cras) — Cec): (3.22) The ITTC recommended procedure for the standard resistance test is described in ITTC 2002 [3.20]. Uncertainty analysis of the results should take place, involving the accuracy of the model and the measurements of resistance and speed. A back- ground to uncertainty analysis is given in [3.28], and recommended procedures are described in [3.20] After the various corrections are made (c.g. to speed and temperature), the model resistance test results will normally be presented in terms of the total resis- tance coefficient Cr (= R/4pSV*) against Froude number Fr, where S is the static wetted area of the model. Extrapolation of the model results to full scale is described in Chapter 4. m [Link] High-Speed Craft and Sailing Vessels. High-speed craft and sailing yachts develop changes in running attitude when under way. Compared with a conventional displacement hull, this leads to a number of extra topics and measurements to be considered in the course of a resistance test. The changes and measurements required are reviewed in [3.29]. Because of the higher speeds involved, such craft may also be subject to shallow water effects and correc- tions may be required, as discussed in Chapters 5 and 6. The typical requirements for testing high-speed craft, compared with displacement hulls, arc outlined as follows: (i) Semi-displacement craft High-speed semi-displacement craft develop changes in running trim and wetted sur- face area when under way. The semi-displacement craft is normally tested free to heave (vertical motion) and trim, and the heave/sinkage and trim are measured dur- ing the course of a test run. Running wetted surface area may be measured during aComponents of Hull Resistance A Tow force I v R_—JIP® Model free to trim and heave Sr -—r t Figure 3.11. Trim compensation for offset tow line. T > eecied rusts run, for example, by noting the wave profile against a grid on the hull (or by a photo- graph) and applying the new girths (up to the wave profile) to the body plan. There are conflicting opinions as to whether static or running wetted area should be used in the analysis. Appendix B in [3.24] examines this problem in some detail and con- cludes that, for examination of the physics, the running wetted area should be used, whilst for practical powering purposes, the use of the static wetted area is satisfactory. It can be noted that, for this reason, standard series test data for semi-displacement craft, such as those for the NPL Series and Series 64, are presented in terms of static wetted area. Since the high-speed semi-displacement craft is sensitive to trim, the position and direction of the tow force has to be considered carefully. The tow force should be located at the longitudinal centre of gravity (LCG) and in the line of the expected thrust line, otherwise erroneous trim changes can occur. If, for practical reasons, the tow force is not in line with the required thrust line, then a compensating moment can be applied, shown schematically in Figure 3.11. If the tow line is offset from the thrust. by a distance x, then a compensating moment (1 x x2) can be applied, where (w X x2) = (R x x1). This process leads to an effective shift in the LCG. w will normally be part of the (movable) ballast in the model, and the lever x7 can be changed as necessary to allow for the change in R with change in speed. Such corrections will also be applied as necessary to inclined shaft/thrust lines. (ii) Planing craft A planing craft will normally be run free to heave and trim, Such craft incur sig- nificant changes in trim with speed, and the position and direction of the tow line is important. Like the semi-displacement craft, compensating moments may have to be applied, Figure 3.11. A friction moment correction may also be applied to allow for the difference in friction coefficients between model and ship (model is too large). If Rp» is the model frictional resistance corresponding to Cry and Rrns is the model frictional resistance corresponding to Cy, then, assuming the friction drag acts at half draught, a counterbalance moment can be used to counteract the force (Rim — Ryn). This correction is analogous to the skin friction correction in the model self- propulsion experiment, Chapter 8. Care has to be taken with the location of turbulence stimulation on planing craft, where the wetted length varies with speed. An alternative is to use struts or wires in the water upstream of the model [3.29]. a728 Ship Resistance and Propulsion Air resistance can be significant in high-speed model tests and corrections to the resistance data may be necessary. The actual air speed under the carriage should be measured with the model removed. Some tanks include the superstructure and then make suitable corrections based on airflow speed and suitable drag coefficients. The air resistance can also cause trimming moments, which should be corrected by an effective shift of LCG. Planing craft incur significant changes in wetted surface area with change in speed. Accurate measurement of the running wetted area and estimation of the fric- tional resistance is fundamental to the data analysis and extrapolation process. Meth- ‘ods of measuring the running wetted surface area include noting the position of the fore end of the wetted area on the centreline and at the side chines, using underwa- ter photography, or using a clear bottom on the model. An alternative is to apply the running draught and trim to the hydrostatic information on wetted area, although this approach tends not to be very accurate. The spray and spray root at the leading edge of the wetted area can lead to difficulties in differentiating between spray and the solid water in contact with the hull. For high-speed craft, appendage drag normally represents a larger proportion of total resistance than for conventional displacement hulls. If high-speed craft are tested without appendages, then the estimated trim moments caused by the appendages should be compensated by an effective change in LCG. Captive tests on planing craft have been employed. For fully captive tests, the model is fixed in heave and trim whilst, for partially captive tests, the model is tested free to heave over a range of fixed trims [3.29]. Heave and trim moment are mea- sured, together with lift and drag. Required values will be obtained through interpo- lation of the test data in the postanalysis process. Renilson [3.30] provides a useful review of the problems associated with mea- suring the hydrodynamic performance of high-speed cratt. (Gii) Sailing craft A yacht model will normally be tested in a semi-captive arrangement, where it is free to heave and trim, but fixed in heel and yaw. A special dynamometer is required that is capable of measuring resistance, sideforce, heave, trim, roll moment and yaw moment. Measurement of the moments allows the centre of lateral resistance (CLR) to be determined. A typical test programme entails a matrix of tests covering a range of heel and yaw angles over a range of speeds. Small negative angles of heel and yaw will also be tested to check for any asymmetry. Typical test procedures, dynamome- ters and model requirements are described by Claughton et al. [3.8] In the case of a yacht, the position and line of the tow force is particularly impor- tant because the actual position, when under sail, is at the centre of effort of the sails. This leads to trim and heel moments and a downward component of force. As the model tow fitting will be at or in the model, these moments and force will need to be compensated for during the model test. An alternative approach that has been used is to apply the model tow force at the estimated position of the cen- tue of effort of the sails, with the model set at a predetermined yaw angle. This is a more elegant approach, although it does require more complex model arrangements (3.8)Components of Hull Resistance Figure 3.12, Pressure variations around a body. 3.1.5 Transverse Wave Interference [Link] Waves When a submerged body travels through a fluid, pressure variations are created around the body, Figure 3.12. Near a free surface, the pressure variations manifest themselves by changes in the fluid level, creating waves, Figure 3.13. With a body moving through a stationary fluid, the waves travel at the same speed as the body. [Link] Kelvin Wave Pattern The Kelvin wave is a mathematical form of the wave system created by a travelling pressure point source at the free surface, see Chapter 7. The wave system formed is made up of transverse waves and divergent waves, Figure 3.14, The heights of the divergent cusps diminish at a slower rate than transverse waves, and the divergent waves are more predominant towards the rear. The wave system travels at a speed according to the gravity-wave speed relationship, see Appendix AL8, = /F 3) Ve VaR 6.23) and with wavelength of the transverse waves, Qn? = G24) & The ship wave system is determined mainly by the peaks of high and low pres- sure, or pressure points, that occur in the pressure distribution around the hull. The ‘Stern wave system Bow wave system Figure 3.18. Ship waves.30 Ship Resistance and Propulsion Moving pressure source = 1g Ayp=2aVig | Figure 3.14. Kelvin wave pattern. overall ship wave system may be considered as being created by a number of travel- ling pressure points, with the Kelvin wave pattern being a reasonable representation of the actual ship wave system. The ship waves will not follow the Kelvin pattern exactly due to the non-linearities in the waves and viscous effects (e.g. stern wave damping) not allowed for in the Kelvin wave theory. This is discussed further in Chapter [Link] should also be noted that the foregoing analysis is strictly for deep water. Shallow water has significant effects on the wave pattern and these are discussed in Chapter 6. [Link] Wave System Interference ‘The pressure peaks around the hull create a wave system with a crest (trough for negative pressure peak) situated some distance behind the point of high pressure. ‘The wave system for a ship is built up of four components, Figure 3.15. (1) Bow wave system, with a high pressure at entry, starting with a crest. (2) Forward shoulder system, due to low pressure between forward and amidships. Bow wave Forward shoulder wave - Ater shoulder wave ao Stern wave Figure 3.15. Components of wave system.Components of Hull Resistance Bow'{source) Stetn (sink) — 24 — Figure 3.16. Bow and stern wave systems. (3) After shoulder system, due to low pressure at amidships and higher pressure at stern. (4) Stern wave system, due to rising pressure gradient and decreasing velocity, start ing with a trough. ‘Components (1) and (4) will occur at ‘fixed’ places, whilst the magnitude and position of (2) and (3) will depend on the form and the position of the shoulders. Components (2) and (3) are usually associated with fuller forms with hard shoulders. The amplitudes of the newly formed stern system are superimposed on those of the bow system. Waves move forward with the speed of the ship, and phasing will alter with speed, Figure 3.16. For example, as speed is increased, the length of the bow wave will increase (Equation (3.24)) until it coincides and interferes with the stern wave. Asa result of these wave interference effects, ship resistance curves exhibit an oscillatory nature with change in speed, that is, they exhibit humps and hollows, Figure 3.17. Humps will occur when the crests (or troughs) coincide (reinforcement), and hollows occur when a trough coincides with a crest (cancellation). [Link] Speeds for Humps and Hollows The analysis is based on the bow and stern transverse wave systems, Figure 3.16, where Lps is the pressure source separation. Now Lys = k x Lgp, where kis typically 0.80-0.95, depending on the fullness of the vessel. The lengths 2 of the individual waves are given by Equation (3.24). Itis clear from Figure 3.16 that when the bow wavelength A is equal to Lps, the crest from the bow wave is on the stern trough and wave cancellation will occur. At lower speeds there will be several wavelengths within Lps and the following analysis Ship resistance ‘Speed Figure 3.17. Humps and hollows in ship resistance curve. 3132 Ship Resistance and Propulsion Table 3.1. Speeds for humps and hollows Humps A/ Les 2 213 2s 27 Fr 0535 0309 ©0239 0.202 Hollows A/Les 1 12 18 a Fr 0378 0.268 0219 0.189 indicates the speeds when favourable interference (hollows) or unfavourable inter- ference (humps) are likely to occur. Hollows: Bow crests coincide with stern troughs approximately at Les ‘Humps: Bow crests coincide with stern crests approximately at: a 222 Ips 3°5'7 ‘The speeds for humps and hollows may be related to the ship speed as follows Ship Froude number and and, from Equation (3.24), the wavelength 2a pr ghts and (3.25) If say k = 0.90, then speeds for humps and hollows would occur as shown in Table 3.1. The values of speed in Table 3.1 are based only on the transverse waves. ‘The diverging waves can change these speeds a little, leading to higher values for humps and hollows. At the design stage itis desirable to choose the design ship length and/or speed to avoid resistance curve humps. This would, in particular, include avoiding the humps at or around Froude numbers of 0.30, the so-called prismatic hump and 0.50, the so-called main hump. Alternatively, a redistribution of displacement volume may be possible. For example, with a decrease in midship area (Cy) and an increase in prismatic coefficient (Cp), displacement is moved nearer the ends, with changes in the position of the pressure sources and, hence, wave interference. Furthermore, a decrease in bow wavemaking may be achieved by careful design of the entrance and sectional area curve (SAC) slope, or by the use of a bulbous bow for wave cancella- tion, These aspects are discussed further under hull form design in Chapter 14. ‘The wave characteristics described in this section are strictly for deep water. The effects of shallow water are discussed in Chapter 6.Components of Hull Resistance 3.1.6 Dimensional Analysis and Scaling [Link] Dimensional Analysis In addition to the physical approach to the separation of total resistance into identi- fiable components, the methods of dimensional analysis may also be applied [3.31]. [3.32]. Physical variables and their dimensions are as follows: Hull resistance: Rr (force), Hull speed v (velocity), Hull size: L Fluid density: p (mass/unit volume), Fluid viscosity: M (stress/rate of strain), Acceleration due to gravity: g (acceleration). According to the methods of dimensional analysis, the relationship between the quantities may be expressed as (Rr. VL, p, i, 8,0) = 0 or Rr = SVL, pH & 01), (3.26) where a; are non-dimensional parameters of hull shape. If geometrically similar models are considered, the method of dimensional anal- ysis yields , &) ] . (3.27) where v = 4, and Rr _ y 3, = F 3. aya, = FIRE], (3.28) or say Cr = f(Re, Fr), where Cr= (3.29) where S and So L? for geometrically similar hulls (geosims), and 0.5 in the denominator does not disturb the non-dimensionality, where Re= ee is the Reynolds number, and is the Froude number. Ya Hence, for complete dynamic similarity between two hull sizes (c.g. model and ship), three conditions must apply. 1. Shape parameters o; must be the same. This implies that geometric similarity is required34 Ship Resistance and Propulsion 2. Reynolds numbers must be the same 3. Froude numbers must be the same ‘The last two conditions imply: VL a Li/gib: Lay! gob ry Sib. ay 8242 or UL v2 sili _ wl} vy vB Ly" [z] for constant g. Hence, complete similarity cannot be obtained without a large change in g or fluid viscosity v,and the ship scaling problem basically arises because complete similarity is not possible between model and ship. ‘The scaling problem can be simplified to some extent by keeping Re or Fr con- a (3.30) vy stant, For example, with constant Re, Vili _ Valo uo ie Yi Lavy _ Lr Va Tim Ty if the uid medium is water in both cases. If the model scale is say ¥, fs = 25, then = 25, ie. the model speed is 25 times faster than the ship, which is impractical with constant Fr, Vv: Ve vali Vele ie Yi_ flv Va Vin 5 for 1/25th scale, then the model speed is 1/Sth that of the ship, which is a practical solution. Hence, in practice, scaling between model and ship is carried out at constant Fr. At constant Fr, Rey _ Vila Rey vy L Hence with a 1/25th scale, #4 = 7 the model. 3p, 2 Ly Vila 23? and Re for the ship is much larger than Re for