Hydrodynamic Design of Planing Hulls By Daniel Savitsky! The elemental hydrodynamic choracteristics of given which describe the lift, drag, wetted area, cond empirical planing equations cre center of pressure, and porpoising stabil speed, trim angle, deadrise included to demonstrate the application of FUNDAMENTAL research on the hydrodynamics of planing surfaces has been actively pursued in both this country and abroad for well over 40 years. ‘The original ‘impetus for this planing research was primarily motivated by the hydrodynamic design requirements of water- based aircraft and to a somewhat lesser extent by the development of planing boats. In recent years, how- ever, the research emphasis has been on planing forms with application to planing boats and hydrofoil craft, " Associate Profesor, Hood of Applied Mechanics Group, David- ‘on Laboratory, Stevets Instituie of Technology, Hobekea, NF Prosented atthe Januaty 1964 Meeting of the New York Metro- poltsn Section of Te Soctery ov NavaisAncireors ano Masts * Numbers in brackets designate References at end of paper Nomer Cy = ttictionarag coefficient = Dy cos seceloratio ‘S23 tue ‘ongle, and loading, lote simple computationc! procedures to predict th trim, drott, and porpoising stability of prismatic pl prismatic planing surfaces are discussed ty limits of planing surfaces es o function of These results are combined to formu the horsepower requirements, running loning hulls, Ilustrotive examples ore the computational procedures, Some of the earliest experimental studies on pris- matic planing surfaces were made by Baker [I]? in 1910 but the first’ comprehensive experiments which received wide attention were those of Sottorf [2]. ‘These were followed by investigations of Shoemaker [3], Sambraus [4], Sedov (5], and Locke [6]. The efforts of these researchers resulted in a large accumulation of test data describing the hydrodynamio characteristics of constant. deadrise prismatic planing surfaces operating at fixed fizim, fixed mean wetted length, and constant speed ‘To make these data suitable for practical use it was de- sirable to establish empirical equations which would ex- press the relations between the many planing variables and the hydrodynamic lift, diag, pitehing moment, and wetted area. Under sponsorship of the Office of Naval fnclature, SE 2 due to gravity, Fm distance between 7 and CG (reasured normal to 7, f¢ Bf Vib? a Z wetted chin length, ft 1 ~ propeller threst Cay lift coelfcient, zero deadrise, = elted be lanath i on = inclination of thrust line relative evan ee emacoroniraee | quan ye Cig = Hit Soetiien, dendine suracs amicforge wactor with el (ineasured normal to A te vas * nogaeaeafed one Meal, Fe difference between wetted keel “8 fzontal velocity of plaking a and chine lengths, = (Le Cx, = dynamié component of lit oeti- mma rao ove Batam ofa gl been tee eon ooo anal lang euface fr, 3) tea 4a = cifrencebutween kel nd chine ry = busyant component of Ut coeti- = angle of endnse' of lading roe lnglis meted by level Sater es gs ote surf, Oy = ditaaee of center of premure 4 = 1h SME ey wean wotted length, fe = (Ea + Ghydredyeamiofores) measured} = (dnernatis Wiseorty of fuid, 14/ ray ee feo specito weight of water, pel = ; 9 = pas density of water, / {gle between spray 0% line and Sx peediosint = V4) 2, = Rydeostatiolit competent’ 1b feel “line miaaured in pane Re = Reynolds number, = Vidb/r. parallel to keel deg X= mean wettedlength-beam ratio also a 7 ~ trimangleof planing area, deg atts + PRageonponeny I 1o6 T longiuinal sane eee of ea companea ravity from transom (mess Y= mea, poi loth team ni rosisiance ‘component due to Bel tong ety ranwom wurbod watersurtace Te 4 = vertical depth of trailing edge of angle between the kesl and spray a Ting at ket blow Tovel water ‘edge measured in plane of bot- vere sirius fe 8, .b = beam of planing surface, ft N= component of resistance force ‘otal wetted spray aren, sq ft 7 Dy = fictional’ dragrores ‘componoat tong bottom face; I = cor ain (Gneasur OcTosER, 1964 Reprinted from MARINE TECHN reaal to bottom, Ib distance between “D, end CG distance of eanter of gravity above ‘eel Tino, measured sormal to etnoraal to By), Keel ft n (OLOGY, Vo. 1, No 1, pp. 71-95wave mise LEVEL waren, ovat SURFACE Fig. 1 Wave rise on a ft planing surface Research, US. Navy, the Davidson Laboratory of Stevens Institute of Technology, in 1947, undertook a theoretical study and empirical-data anelysis of the phenomenon of planing. This study produced 16 tech- Aical reports (listed in the Appendix), which consider planing-surface lift, drag, wetted area, pressure distribu- tions, impact forces, wake shape, spray formation, dynanie stability, and parallel planing surfaces. Where possible the ONR sponsored research utilized existing planing data and theoretical results but in many areas additional experimental results and new theoretical analysis were provided by the Davidson Laboratory. In 1949, Korvin-Kroukovsky and Savitsky [7] pub- lished a summary report on the then completed studies of planing lift, drag, and wetted area and, in 1950, Murray [8] utilized these results in developing a compu: tational procedure for predicting planing performance In 1954, Savitsky and Neidinger [9], continuing the ONR study, developed an extensive set of empirical planing equations which increased the range of applica- bility to parametric planing variables well beyond those developed in (7) ‘The purpose of the present paper is to utilize the re- sults of the studies of [9] to describe the elemental hydrodynamic characteristics of prismatic planing sur- faces and then to combino these results to formulate simple computational procedures to predict the horse- power requirements and porpoising stability of pris- matic planing hulls, Some of the material of (9] is re- peatod in this paper since [2] had a limited distribution and is currently out of print Hydrodynamics of Prism: A knowledge of the elemental hydeodynamic character- istics of simple planing surfaces is necessary prior to undertaking the design of specific geometric planing boats. In this section of the paper attention will be given to the development of equations for wetted area, lift, drag, center of pressure and stability limits of hard= chine prismatic surfaces in terms of deadvise angle, trim. angle, and forward speed. ‘The prismatic planing sur- face is assumed to have constant deadvise, constant beam and a constant running trim for the entive wetted planing area, Variations from these conditions will be discussed in the section on design procedure. Only hard-chine planing forms are cousidered in this paper since, at present there is a scarcity of basic planing data on round-bottom forms, Planing Surfaces 72 PRESSURE biSTaISUTION LEVEL wareR. SURFCE Fig, 2 Typical pressuse distribution on flat planing susface ‘The planing coefficients and symbols used in the sub- sequent analysis are based on Froude's law of similitude and are the same as those used in the analysis of water- based aireraft and hydroskis, Each symbol is specifi- cally defined in the section on nomenclature. Tt will be noted that the beam is the prime nondimensionaliing dimension rather than the length of the boat which is usually considered by the naval architect. The justification for this is that for planing hulls, the wetted Tength of the boat varies with trim, loading, and speed while the wetted beam is essentially constant. More- over, it is possible to change the overall length of a plan- ing boat without changing its hydrodynamic character isties at high speed. Shape of Wetted Area of Planing Surfaces A separate analysis is given of the shape of the wetted area for flat-bottom and deadvise planing surfaces Wave Rise for Flot Plening Surfoces In the ase of planing surfaces with no deadrise (fat~ bottom planing surfaces), water rises in front of the sur- face, thereby causing the running wetted length 1 t0 be larger than the length defined by the undisturbed water- level intersection with the bottom &, Fig. 1. Wagner {10] had made a mathematical study of the flow at the leading edge of a planing surface of infinite length and found that the rising water surface, mentioned in the fore- going, blends into a thin sheet of water flowing forward along the planing surface. ‘This sheet is the source of spray in a planing surface and the rogion of its origin has been designated by Wagner as the “spray-root” region, Fig. 2 shows the spray root and the pressure distribu- tion resulting from it. The term wetted area, as used in this paper, designates that portion of the wetted area over which water pressure is exerted and excludes the forward thrown spray sheet, ‘The wetted aren used in this sense is often designated in the literature as the pressure area” and geometrically, inchdes all the wetted bottom area, aft of a line drawn normal to the planing bottom and tangent to the curve of the spray root. ‘This line is clearly discernible from underwater photographs, As scen in Pig. 2, the stagnation pressure is developed MARINE TECHNOLOGYEL WATER ‘SURFACE TIT TTT TT WETTED LENGTH-BEAM RATIO BASED ON LEVEL WATER SURFACE, a, eS 160,-0.30 x8 (0<,<1) E (ls asa) iB ™ MAb tat ° 1 2 3 4 WETTED LENGTH-BEAM RATIO BASED ON WAVE RISE, 4 Fig. 3 Waverrise variation for fat planing surfaces at a short distance aft of the epray-root line, At very small values of trim angle the stagnation line and spray root line are nearly coincident. As the trim angle in- breases, the stagnation line moves farther aft of the spray- root line, Fiat-plate, wetted-length data from all available sources are shown plotted in the form of \ versus Ay in Fig. 3. Here ) represents the running mean wetted Jength-beam ratio (I/b) and dy represents the ealm-water length-beam ratio obtained from the relation bx = d/b sin, where d is the depth of the trailing edge of the OcroseR, 1964 planing surface below the level water surface during a planing run. Tt is seen, from Fig. 3, that, for the range of test parameters considered, the wave tise on a flat- bottom planing surface is only a function of the running wetted length. The mean curve fitted through the test data is defined by the following empirical equations: 160-030. tan Fane ® MARINE TECHNOLOGY‘The wave rise in the spray-r00t area is accounted for by the following consideration, Wagner computed the wave rise for a two-dimensional wedge penetrating. © uid surface vertically, and found that the actual wetted vrdth ofthe wedge was 4/2 times the wetted width de- fined by the ealmevvater intersection with the bottom, ‘The motion of a deadsise planing suviace can be repre: sented aga two-dimensional problem by tonsidering the water fow between two vertical planes normal to the plane of symmetry of the planing surface, ‘To an ob Eerver located between these tivo planes, the passage of the prismatic Vee planing surface will appear identieal to the vertical immersion of a wedge. This being the caso, the +/2 waverise factor computed by Wagne is applicable, and the difference between actual wetted keel length and chine length for a prismati planing se face is given by 2 tang lant h-k @ It is seen that this length is a factor 2/ times the cor- responding length defined by the level-water intersection with the Vee planing surface. A plot of this relationship is given in Fig. 6. Since the wetted keel length can be defined in terms of the draft of the aft end of the keel as Ly = d/sine w then the mean wetted length-beam ratio, A, which de- fines the pressure area is given as Rig) men Experimental evidence indicates that (3) is applicable for all deadrise and trim combinations when the speed coefficient is greater than C, = 20. This indicates a full development of the spray-root and water pile-up as predicted by Wagner. For deadrise surfaces of 10, deg or less, (3) continues to be applicable at C, = 1.0 For the 20-deg deadrise surface, at C, = 1.0 andr < 4°, experimental values of Ly — L, are larger than those predicted by (3), indicating a partial breakdown of the spray-root formation. Experimental evidence for 30- deg deadrise surfaces showed similar effects except that, at C, = 1.0, the spray-root formation breaks down when + < 6°. It appears that, for C, = 1.0, the spray- root formation will begin to break down when, for a given deadrise, the trim is reduced to a value such that the theoretical value of J, — L, is approximately equal to 1.68. ‘This quantity (Z, — 1,)/b ean actually be con- sidered to be a measure of the angle (y) between the spray-root line and the keel line measured in a plane along the keel. Hence, it may be generalized that the spray-root formation at C, = 1.0 will begin to break- down when the theoretical value of + is less than ap- proximately 17 deg for a given trim-deagrise combina tion. It is easily shown that y = tan? (Panr/2 tang) A series of photographs illustrating the breakdown in the spray-root line is given in Fig. 7 where bottom arcas are shown for a 20-dog-dencirise surface planing at a ocToweR, 1964 trim angle of 4 deg and at five values of C,. The ealeu- lated angle 7 = 17°. It is seen that, at C, = 2.01, 3.02 and 4.00, the spray-root line is one continuous line and the value of ZL, — Ze. is in agreement with that. com- puted by equation (3). At C, = 10, the leading edge of the wetted ares is now defined by a broken line made up of two straight segments. ‘The forward segment is the usual spray-root line formation and makes an angle of approximately 17 deg with the Keel. ‘The after sez ‘ment of the leading-edge line makes an angle with the Keel which would correspond to the calm-water inter section with the bottom. At G, = 0.6, the same phenomenon is in evidence except that the length of the spray-root portion of the line is reduced. Wetted-Spray Area of Deadrse Planing Surfaces ‘The total wetted bottom ares of a planing surface is actually divided into two regions. One is aft of the spray-root line, commonly referred to as the: pressure area and the other is forward of the spray-root line, re- ferred to as the spray area. The prescure area, which has been defined in the preceding sections of this paper, is the load-carrying area of the planing bottom. ‘The forward spray area contributes to the total drag butis not considered to support any portion of the load. ‘The flow directions in both wetted areas have been determined by underwater photographs of tufts such as shown by Pierson and Leshnover in Figs. 4 and 5 of reference [13]. An enlarged sketch of the flow diree- tions on a deadrise surface is given in Fig. 8 of this paper. It is found that the flow in the pressuré-aeel is predominantly aft with some transverse Row along the chines. The flow along the spray-root line is primarily along the direction of the stagnation line. In the spray wetted area the directions of the fluid flow are such that the space angle between the oncoming fluid particles and the stagnation line is equal to the angle between the direction of the spray jets and the stagnation line; i.e, any line of motion in the spray area is nearly a refleotion about the stagnation line of the incident velocity direc tion. Since the pressure in the spray area is nearly at- mospheric, then, by Bernoulli, the spray velocity ean be assumed to be equal to the planing speed. Equations defining the spray direction in terms of trim and deadrise angle are given by Pierson and Leshnover [13]. ‘The actual spray area extends from the spray- root line forward to the spray edge. The angle © be- tween the keel and spray edge measured in the plane of the bottom is Ath tan ahs @ where: 4 |sinte( — 9K) + K? tantr{(1/sin'g) — sintr)}* cosr + K tanr sinr ., - Ktane n= 7TAN 9 = TANG cos @ ze TANS Tawas Fate FLOW DIRECTIONS ‘SPRAY LINES a SPRay E0sE VIEW OF BOTTOM ON PLANE PARALLEL TO KEEL Flow direction slong planing prism and extent of spray ares and £(,_ Stent @.cos 8 _ tang sin’ Ba F(— Sar pee — wap se ‘The total spray area, both sides, projected on a plane along the keel line is given by [ 4 tome 2 Getanr ~ Tiana cond, In making visual observations of the wetted chine length during a planing run, itis important to distinguish be- fbveen the spray-root intersection and the spray-edge intersection with the chine, Fig. 9 illustrates the two intersection points. Tt is seen that the spray edge is slways ahead of the spray-root intersection with the bottom. o Lift of Planing Surfoces The following discussions will first develop the lift equations for fiat planing surfaces and then show how these results are modified to account for finite deadrise 7a Lift of Flot Ploning Surfeces ‘The lift on a planing surface (at fixed draft and trim) can be attributed to two separate effects; ie., one is the dynamic reaction of the fluid against the moving surface, ‘and the second is the so-called buoyant contribution to lift whieh is associated with the static pressures cor- responding to a given draft. and hull trim, In effect, the buoyant contribution represents the influence of gravity. At very low-speed coefficients, the buoyant it component predominates. As speeds are increased, the dynamic-lift effects begin to develop. At first the dynamic effects tend to decrease the load which a given prismatic surface can support and then, as the speed is further increased, the load on a given’ surface will in- crease. At very’ high-speed coefficients the dynamic contribution to lift predominates and the statie-pressure effects can be neglected. The formulation of an empiri- cal planing lift equation was based on a combination of the dynamic and statie efects, It will be recalled that the fiuid-flow directions over the pressure area of a planing surface were a combination of longitudinal flow and some transverse flow across both chine lines. Prom aerodynamie theory it is known that MARINE TECHNOLOGY9) Characteristic features of vee-bottom planing surf Dechine; E—whisker spray edge; G—spray-root region wedge; Botransom) Ck lifting surfaces of high aspect ratio (small A) have a pre- dominantly longitudinal (chordwise) flow and that the lift is directly proportional to x. For surfaces of very small span and infinite length, ie, » = ©, the flow is in a transverse direction and lift is proportional to 7? Henee for a normal low aspect-ratio planing surface, the lift cam be expressed in the form Cy = Art Bet ® For the range of A-values applicable to planing surfaces, the second term takes the form of a small correction to the first term and it is found that equation (8) can be approximated by using r to the 1.1 power. Hence 0, C) @ Sottort’s analysis of high-speed planing data, where the hydrostatic term is negligible, showed that, for a given ‘rim angle, the dynamic component of the lift coefficient varied as \”*, Hence we ean consider this component to be of the form: Cag = ental where cis a constant to be determined. The hydrostatic component of lift for a flat plate of beam, b, mnean wetted length-beam ratio, 4, and angle of trim + ean be written as follows: Jn = § gb" — 0.30)? tanr ay Dividing both sides by 391 and assuming that ( — 0,30)? ean be replaced by KA" where D and n are con- stants to be determined, results in (a0) ‘ocToseR, 1964 Acmodel of planiog Erefection of speay (2) the difference between tanr and + is neglected Cz, can be written Dy c. Combining equations (10) and (13) gives a form of an empirical equation for the lift coefficient of a planing surface, ie, y= ota (eqn DR Coes (w a) As with any empirical equation there are several ways ‘to formulate the equation for planing lift. ‘The form of relation given in (14) has the advantage of readily illustrating the effect of the prime variables on planing lift and also is easily applied in design of planing hull, ‘The constants C, D, and n are evaluated by applying the foregoing formula to the large collection of planing data contained in the existing literature. ‘The mechan= ics of this evaluation are described in [9]. As a result of this analysis the empirical planing lift equation for a zero deadrise surface takes the following final form: = +[oorzo. + Cy = Ee (a3) ay (15) where + isin degrees, ‘This empirical equation is applieable for 0.60 < Oy < 13.00; 2° < 7 < 15%; and a < 4. 7900s > 004 ey10 eo 003 ~ i 307 [a . $0 PB 20 o 730 ee Y | oor 4:7 Woor »”* +0058 7/2) | ° 20 30 40 » Fig. 10. Life coeficienr of «Gat planing surface; = 0° For convenienee in use, equation (15) is plotted in Fig. 10 in the form Ci/7* versus ) for a wide range of Cevalues. Examining this plot at a fixed value of itis clear that the buoyant contribution to lift is signif cant up to speed coeflicients as high as approximately 10. At, > 10, the dynamic lift is predominant. and the lift coefficient is then independent of speed. In fact, for C, > 10.0 the flat-plate lift coefficient ean be simply expressed as Cz, = 0,0120 )'/r!1 To illustrate the loss in lift experienced by a planing surface at very low speeds (C, < 1.0), Fig. 12 preseats 80 8 comparison between the resultant lift and that cor responding to the purely static lift (buoyaney) for a given draft and trim of the planing surface. A form of load coefficient is plotted against speed coefficient for three wetted lengths at three trim angles. ‘The solid curves are the planing loads as predicted by (15) and are seen to vary with C,. The dotted curves are the buoyant. loads computed by (11). This hypothetical load is inde- pendent of @.. ‘The comparison between the planing oad and calculated buoyant load is limited to C, > 0.60 since this is the range of applicability of (15). As C, MARINE TECHNOLOGYee os - 7 03 - L ce of 009 ° Prijit Or 02 05 a4 OS fe Cue 008 ; oor at I | c,+¢,,-0.0088 6g 152 0 08 7 0.05, 23, * ee oot oos 0a | 201 c | | Msrrtititlilititil, G02 0a 08 G08 G0 Figs, Lift coeficlat of dead plang surface approaches zero, it is naturally expected that the calew- lated load should approach the buoyant load, It is interesting to note from Fig. 12 that in the range 0.60 < C, < 1.00, the motion of the planing surface reduces the lift below the value which would be expected on a purely displacement basis, This effect is somewhat similar to the sinkage experienced by displacement vessels at low speeds. At C, ~ 1.0, the total planing load is approximately equal to the hypothetical buoyant load At C, > 10 the positive dynamic reaction of the fluid on the planing bottom increases rapidly as the speed increases, ocrosEr, 1964 Liff of Oeadrise Planing Surfaces For a given trim and mean wetted length-beam ratio, the effect of increasing the deadrise angle is to reduce the planing lift. This lift reduction is caused primarily from a reduction in the stagnation pressure at the lead- ing edge of the wetted area, Tt will be recalled from the discussion of wetted areas that the angle between the stagnation line and Keel is given by the equation y = tan~* (lan 1/2 tan 8). When 6 = 0 the stagnation line is normal to the keel and normal to the free-stream velocity so that full stagnation pressure }épV? is de- veloped, For increasing values of 8, the angle y de aoa v2.2 Br TOTAL PLANING Loans asi2rgb= 7" [o.o120x’*c,"+ 0.0055 x 3 2 ~ EQUIVALENT DISPLACEMENT LOAD=a/1/2 £9 b'= ( 4-0.30) TANT . 200 a 0.30 060 120 Cy ce cy Fig. 12. Planing Jond versus calculated displacement load for a at planing surface at vasious velocity coeficients creases 0 that full stagnation pressures are no longer developed; hence the planing lift is reduced. In effect. then, the presence of deadrisé causes the stagnation line to be “swept” aft and leads to a lift reduction not unlike that on a swept-back wing, ‘To formulate an empirical equation for the planing lift, of a deadrise surface, the lift coefficient af a Vee surface Was compared with that of a flat plate at identical values of x 2, and C,. It was found in [7] and (9) that the lift of a deadrise surface ean be represented by the fol- owing equation: Cag = Coy — 0.00858 C228" (a6) where Cig = lift coefficient for a deadtise surface 8 = deadtrise angle, deg 2 Cz, = lift coefficient of a flat plate operating at the same 7A,and C, as deadrise suriace For convenience in use, equation (16) is plotted in Fig, 11. Drog of Ploning Surfaces ‘The total hydrodynamic drag of a planing surface is composed of pressure drag developed by pressures acting normal to the inclined bottom and viscous drag acting ‘tangential to the bottom in both the pressure area and spray area. If there is side wetting then, of course, this additional component of viscous drag must be added to the hydrodynamic drag acting on the bottom of the plan- ing surface. For the present analysis it will be assumed that there is no side wetting of the hull For a frictionless fluid, the tangential forve is zero. Hence for a trim angle r, 8 load a, and a foree N normal MARINE TECHNOLOGYto the bottom the resistance component D, due to pressure forees is shown in Fig. 13 to be Dy = dtanr an ‘When the viscous drag D,acting tangential to the bottom is added, the total drag, D, is shown in Fig. 13 to be D, cosr as) D = Atanr + ‘The friction component D, is shown in [9] to be com- puted by the following equation: where Vit) 2 cos8 * C= Schoenhorr [14] turbulent friction coefficient V; = average bottom velocity ‘The average bottom velocity (V1) is less than the forward planing velocity (7) owing to the fact that the planing bottom pressure is larger than the free-stream pressure. Sottorf, Parkinson (15] and Locke [16] have presented data and analytical expressions for defining the average bottom velocity at very high-speed coeffi- cients where the buoyant contribution to lftis negli Savitsky and Ross (17] developed an expression for the mean bottom velocity which is applicable over a speed range from C, = 1.0 to C, = 13.0. This development was based on the following considerations: Taking first, the ease of a zero deadrise hull, the dynamie con- ‘tribution to planing lift is given by the first term in (15) tobe (a9) Dy= Cog = 0.0120" 20) ‘The dynamic load on the bottom is Aa = 4 9V%%0.0120\""r!4) ey ‘The average dynamic pressure is A _ 00012074 Vp ; Pe= ere (22) 2A oor Applying Bemnoulli’s equation between the free-stream conditions and the average pressure and velocity condi- tions on the bottom of the planing surface: =v (i 2)" ° substituting (22) into (23) gives v= 7 (1 — 90 The average bottom velocity for specific deadrise angles is computed in an analogous manner using the lift co- efficient for deadrise surfaces given by (16). The ratios, ¥4/V have been computed for four deadzise angles and the results are plotted in Fig, 14 in @ convenient form for use by the designer. Tt will be noted that the wetted area used in (19) is octoser, 1964 TANT O)FRICTIONLESS FLUIO Taner. diviscous FLUID Fig. 13 Drag components on a planing surface the bottom pressure area, Xb, In previously published Davidson Laboratory reporis [9, 17] consideration was given to the friction drag developed by the spray area, ‘equation (7), ahead of the spray-roat line, ‘The analysis of {0] and [17], which were based on certain assumptions as to the spray thickness and the frietion drag coefficient in the spray area, resulted in a simple formulation for an additional increment, AXb%, in wetted areas to be added to the pressure area Ab* These results were based mainly on data obtained at planing trim angles greater than 4 deg. Recent studies at the Davidson Laboratory have indicated that at trim angles less than 4 deg (usual for planing boats) the spray thickness is considerably Jess than had been assumed previously. In fact, the spray sheet appears to be much thinner than the dis- placement thickness of a normal boundary layer at the same Reynolds number. Hence, until this effect is more fully studied, it is recommended that at trim angles less ‘than 4 deg, the area used for computing the viscous drag be Ab% Vor larger trim angles the results of [9] and [17] should be used. In summary then, the hydrodynamic drag of a planing surface is given by the following equation: _oV eC att 2 0088 cose D = Atanr + (25) cywhere Vj is plotted in Fig. 12, and Cy is the Schoenherr turbulent-friction coefficient. “The Reynolds number is defined, Re = Vi Ni/e, where v is the kinematic viscosity. Dragclift Ratio of Planing Surfaces Prior to computing the drag-lift ratio of planing sur- faces it would be advantageous to examine the typical variations in drag.lift ratio as a function of speed, wetted Jength, and trim angle, Tor this purpose the’ experi- mental data for a 9-in, beam, 20 deg deadise surface (aiven in reference [9]) are plotted in Fig. 19 for sepa. rate values of tzim angle. The abscisea for these plots is a form of speed coefficient. based on wetted length, defined as C,/N's, which will be recognized as being the well-known Froude number ¥/(gi)". Other forms of Froude number representation could have been used (eg, based on load), but the ratio C,/x"* is used. since it is identical in form and equal to 0.296 times the speed. length ratio. It is emphasized that the drag-lift ratios siven in Fig. 15 apply only to the 9-in. test model and are not to be directly applied to full-scale boats, The plots are given merely to indicate typical variations in the drag-lift ratio of planing surfaces. It is seen from Fig. 15 that the ratio D/A plotted against C./x"* generally collapse onto a single curve for each test trim over the test ranges of \ and C,.. Tt is also seen that, up to a ratio of C4/t"* = 1 there is a very. rapid increase in the ratio D/A for all test trims. At +> 2° and at C,/n" > 1, the ratio D/A is nearly con stant for any combination of speed and wetted length For + = 2°, the curve of D/A appears to approach « ‘constant value for ratios of C,/n" > 2 ‘The above variations of D/A can be associated with observed changes of the flow conditions around the planing curface, It was found that, at C, > 2.0 there was a clean separation of the fluid from the chines and the transom. “Further, at C, < 1.00 the degree of flow separation from the transom was, at a given trim angle, 8 function of the wetted length, the shorter the wetted length, the greater the flow separation. With increasing degree of flow separation from the transom, the drag force is increased and hence the ratio D/a is increased until complete flow separation has occurred along the chines and transom. If planing is defined to exist when the fluid breaks away from the transom and chines, theo, using Fig. 15, the inception of planing can be defined to occur when CX = 1 for r > 4° and at C/N!" = 2 for r= 2°, In essence then, planing occurs when the drag-ift ratio ata given trim angle is essentially constant. Other defi- nitions of planing ean be found in the literature. For example, Locke [6] defines the inception of planing to eceur when, at a given 9 and z, the load carried by the planing surface vaties as the square of the speed. ‘This implies that the buoyant component of the lift is neg. ligible. In both definitions only the bottom of the planing surface is wetted. The use of the ratio C,/\i" = 1 defines the point at which this phenomenon first occurs 34 An exact definition of the inception of planing is, of course, not important. The foregoing criterion appear to be a convenient guide in classifying boats. Tt is clear from Fig. 15 that when a boat does start to “plane” it hhas the largest resistance for a fixed trim angle. The resistance decreases sharply when the ratio Cy/a"™ is reduced to values less than 1.0, From equation (25) the drag-lift ratio of a planing surface can be ealeulated as follows: D eV eCab? oe 24 coss cose ‘Multiplying and dividing the second term of the right- hand side by V? and substituting Cz for 24/p V%* results (28) Vi\t D (Fea en a> 07 + er cosr cod In the foregoing expression C, = Cz, if 8 = 0 and Cx = Cia if 6 #0. The ratio V1/V is given in Fig. 14. ‘The friction coefficient C, is a function of Reynolds number \which in turn increases with inereasing size of the planing boat. Since, as shown by Schoenherr, the turbulent friction coefficient decreases with increasing Reynolds number, the ratio D/A will decrease slightly with in- creasing boat size for a given combination of d, r, 8, and Equation (27) has been used to compute the ratio D/4 for 0°, 10? and 20° deadrise surfaces at trim angles of 2°, 4°, 6° and 8°, Mean wetted length-beam ratios, A, were varied from 1 to 4, and speed coefficients up to C, = 10 were used in various combinations. ‘The com- utations were made for a beara, b, of § ft and 10 it. As expected, for r = 4°, the D/A ratio was essentially con- stant when C,/d% > 1.0. For r = 2°, the D/A ratio was essentially constant when C,/x" > 20. The re sults of this computation are given in Fig. 16 to illustrate the effect of trim, deadrise, and size af boat on the drag- lift ratio, “Each computed point represents the average of five different combinations of C/A", On the aver- ‘age, there was approximately a 5 percent spread in the computed values for any trim-deadrise combination For more exact values of D/A it is recommended that detailed evaluations of equation (27) be carried out for specific cases It is evident from Fig. 16 that for any given deadrise, there is an optimum trim angle for lowest ratios of D/A. ‘Small decreases in trim angle below the optimum cause large increases in resistance. Small increases in trim angle above the optimum result in moderate increases in resistance. Increasing deadrise angle inoveases the resistance for a given trim angle. For a deadrise of 0°, the lowest resistance that ean be expected is approxi mately 12 percent of the load at a trim angle of approxi- mately 45°. Tt will be noted that the optimum trim angle increases slightly with inereasing deadrise angle, ‘The effect of increasing the size of the boat beam from 5 to 10 ftis to reduce the D/A ratios by nearly 4 percent. MARINE TECHNOLOGYV, AVERAGE BOTTOM VELOCITY V FORWARD PLANING VELOCITY ~ 100 yiver/1- 2981207 x? cost 2 ogof © | 5 8:0" Eb / oso LIA. Tete 18% 210° buittiiit 100 | : aso 7.00) 200 x 300 [ra S| Ce =e [12 ag fF B=30° Ciptitisirtipritiiiy Woo 200300 4 Fig. 14 Magnitude of average botton velocity for a pleniog surface Included in Fig. 16 is a plot of tanr which is the pressure component of the total drag. ‘The difference between tan and the curves D/A represents the drag component due to viscous (friction) drag. It is seen that at low trim angles the total drag is predominantly friction drag while at Ligh trim angles itis predominantly pressure drag. At r = 4° the total drag for # = 0 is nearly one half pressure drag and one half friction drag, The foregoing trends in resistance variation with trim and deadrise have been shown by many experimenters in eross plots of their specific test data. Fig, 16 presents the results of computations and includes a recognition of the fact that D/A ratios for a given trim angle, are essentially independent of various combinations of C, and. d providing that C,/x"" > 2 for r = 2°, and C,/x"*"> 1 for r > 4° Center of Pressure of Planing Surfaces It has been shown in [9] that the resultant center of pressure of planing surfaces ean be fairly accurately evaluated by separate considerations of the buoyant and dynamic force components of the lift. The center of pressure of the dynamic component is taken to be at 75 percent of the mean wetted length forward of the tran. som, while the center of pressure of the buoyant force is assumed to be 33 percent forward of the transom ‘These distances are, of course, approximations butt are octoser, 1964 acceptable in the empirical development of this paper. Adding the moments taken about the transom for each of the two components of the total lond and then dividing by the total load gives an expression for the distance of the center of pressure forward of the transom. By using ‘the values of the buoyant and dynamie force components given in (13), the center of pressure, Cy, is found to be a distance forward of the transom equal to 1 (28) Gs 1 SS + 2.39 where C, is the ratio of the longitudinal distance from the transom to the center of pressure divided by the ‘mean wetted length, ‘A comparison between (28) and actual test data is given in Fig. 17 of reference [9]. Excellent agreement exists between the formula and data, It is seen that Cy is essentially independent of trim angle and/or dead- rise angle. A working plot of equation (28) is given in Fig, 17 of this paper. When the wetted length and speed coefficient are known, the value of C, ean be quickly determined from this chart, Porpoising Stability Limits Porpoising is defined as the combined oscillations of @ boat in pitch and in heave, of sustained or increasing as030 020 ITsI0" a 10 030 Tis 020 Da 10 60 re20 040 ova 020 10 20 Man 3040 10 30 ig. 15. Variation of deagilife ratio with speed coeficient amplitude, occurring while planing on smooth water. It is peculiar to high-speed planing hulls and will lead ‘to structural damage when the motions become so severe that the hull is thrown entirely out of the water. It may also result in diving (tripping over the bow) when ‘the low trim angles, reached in the lower part of the por- poising cycle cause the bow to dig in. ‘This longitudinal instability has been responsible for many serious boating accidents, and at one time, was considered to be a rather mysterious unknown phenomenon. With the constantly increasing speed of modern planing boats, porpoising is becoming e major problem in planing-boat design. Designers of water-based aircraft were faced with the problem of porpoising instability early in 1930. Perring and Glauert [18] in Bngland developed a theory of por- oising instability in 1933. The practical application of this theory to seaplane design problems was not suc- cessfitl since the theory reqnired an acenrate knowledge a6 of certain hydrodynamic derivatives which could only be obtained experimentally. In fact, the experimental determination of these derivatives ‘were more time- consuming and more involved than @ direct measure of the actual porpoising limits. In 1942, Sottorf [19], in Germany, conducted a systematic model study on the stability limits of a series of float designs suitable for Mloat sesplanes. Sottorf’s experimental work showed that porpoising limits for seaplane floats could be easily predicted in terms of the basic planing coefficients Cs, Cz, and +. In the United States, Davidson, and Locke [20], Benson [21], Parkinson [22] also conducted systematic experimental studies of porpoising limits for water-based aircraft and also showed that the inception of porpoising could be predicted in terms of the basic planing coefficients, With the water-based aireraft experience as a guide, Day and Haag [23] in 1952 undertook a systematic series MARINE TECHNOLOGY020) os oe 008 DRAG-LIFTRATIO, 0/8 004 PRESSURE DRAG 0 20. 40 60 TRIM ANGLE, DEGREES BEAM: TOTAL DRAG [viscous brag - —TANr, PRESSURE DRAG 80 0 20 40 60 80 TRIM ANGLE, DEGREES Fig. 16 Variation of drag.if catio for prismatic planing surfaces of tests of constant dendrise prismatic planing surfaces to determine porpoising limits for planing-hull forms ‘The purpose of their study was to provide the boat de- signer with useful data on the inception of porpoising in terms of the boat trim, speed, weight, and deadrise. ‘The results of the research by Day and Haag are pre- sented in this paper in a graphical form which can be used by the designer of planing boats. These results are constantly used by the Davidson Laboratory 8 a guide in estimating the porpoising limits of planing hulls Briefly the results of the porpoising study showed that for a given dendrise angle, there was a specific relation- ship between trim angle, r, and lift coefficient, C, which defined the inception of porpoising. These relations are shown graphically in Fig, 18 for 0°, 10° and 20° dead- rise prismatic planing surfaces. The combinations of rand C; whieh fall below the limit, eurves indicate stable operation while those above the line indicate the existence of porpoising. It is seen that, as the lift coefficient is decreased, indi- cating @ lightly loaded hull and/or a high planing speed, the trim limit for stability is deerensed. Further, the effect of increasing deadrise is to inerease the trim angle before the inception of porpoising. In any case, if a boat is porpoising at a given speed and load, the rule is to lower the trim angle to avoid porpoising. ‘The lower trim angle can be achieved in several ways. One method is to move the longitudinal center of gravity forward, If this caunot be done and if the boat dimensions are fixed, the addition of a small transverse wedge across the bottom at the transom will lower the running trim at only a small cost in added resistance, Tk may be of interest: at this point to compare the trim ocroser, 1964 requirements to avoid porpoising with the trim angle which results in minimum resistanee, Tt was shown in Fig. 16 that a trim angle of approximately 4° to 5° re- sulted in minimum draglft ratio, ‘The porpoising limits in Tig. 18 require a trim angle as low as 1° to 2° to achieve stable operation of a high-speed boat. Hence, because of porpoising considerations it is necessary to operate the boat at an unfavorably low trim angle where the resistance is high. Increasing the hull deadrise alleviates this situation since as shown in Fig. 18 the trim. angle required to avoid porpoising increases with inereas- ing deadrise angle, Henee, increasing the deadrise will enable a planing surface to operate at trim angles more closely approaching those required for minimum drag- lift ratios: Methods for computing the running trim angle for planing surfaces will be discussed in a subse~ quent section of this paper. It will be noted that the porpoising limits are not de- pendent upon the piteh moment of inertia of the boat. Experimental studies by Locke [24] wherein the moment of inertia was increased and decreased by significant amounts showed a negligible effect on the porpoising inception boundary. What was observed was a change in frequoney of oscillatory motion; increasing frequency for small values of pitch inertia and lower frequency for large inertias, Method for Evaluating Performance of Prismatic Planing Forms ‘The preceding sections of this paper have presented the results of elemental studies of the fundamentals of planing and have summarized the results in terms of equations and design charts, To be of tse to the de- 7igner, it is imporlant that these data be combined to formulate simple computational procedures to predict © the horsepower requirements and porpoising stability of prismatic planing hulls. This section of the paper pre- sents a method for computing the running trim, wetted length, resistance, power requirements and stability of a tiven planing hull over wide epeed ranges and for arbi- trary locations and inclinations of the propeller shaft line relative to the center of gravity of the hull In 1950, Murray [8] presented a computational pro- cedure for predicting resistance which was based on the clemental planing data available at that time, reference [7], No consideration: was given to the effect of pro- peller thrust on the hull lift and pitching moment and, since porpoising information was not, at that time avail: able, porpoising stability limits were not defined. The new planing equations presented in this paper (based on {9)) are applicable for much lower speed coeficients than those used in Murray's paper and, in addition, the new expression for center of pressure is much simpler in form than that used by Murray. DuCane [25] pre- sents & computational procedure which is based on the early planing equations and which is essentially similar to that presented by Murray, In 1959, Clement and Pope (26) presented a series of graphs for predicting the resistance of planing hulls at high speeds. The lift and moment equations used by these authors were those developed by Shuford [27] and are applicable only at C, > 10 where the buoyant forces are negligible. Most planing surfaces operate at lower speed coefficients wherein the buoyant contribution to lift is important. In 1963, Koelbel [28] used the new Davidson Laboratory planing’ relations, reference [9], to develop a simple graphical procedure for predicting the powering require- ments of planing hulls when the effect of propeller thrust on lift and pitching moment is neglected and when it can be assumed that the viseous component of diag passes through the center of gravity. The relative sim= plicity of Koebel's design charts are so attractive that they are included in this paper. ‘There are in the literature test results on related series of planing boats which provide excellent design informa~ tion on families of specie hull designs. Davidson and Suarez [20] present the results for EMB Series 50, a family of planing boats designed by DTMB. Clement and Blount [30] have developed a new hull series desig- nated TMB Series 62 and their results are presented ia [80]. These series data can be used to predict the per- formance of projected new designs which are similar in geometry, loading, and operating conditions to those ‘bull forms investigated in the series Performance Prediction Methods—Anclysis In the present paper the object is to utilize basie plan- ing equations to formulate methods for predicting the performance of a prismatic planing hull whether or not it be a member of a tested series. The computational method involves the determination of the running trim and resistanee whieh will provide for equilibrium eondi- ae tions of the hull at a given running speed, load, and center of gravity location. The accompanying sketch shows the forces and moments acting on a planing hull. = | where 1 = propeller thrust, Ib Ag = weight of boat, Ib D, = viscous component of drag, (assumed as acting parallel to keel line, midway between keel and chine lines), Ib z= trim angle of keel, deg LCG = longitudinal distance of conter of gravity from ‘transom, measured along keel, ft CG = center of gravity inclination of thrust line relative to keel, deg resultant of pressure forees acting normal to bottom, Ib distance between D, and CG (measured normal to D,), ft distance ‘between T and CG (measured normal toshaft line), ft distance between N and CG (measured normal toN), ft 8 = desdrise angle, deg beam, ft wetted Keel length, ft Z, = wetted chine length (from transom to spray root intersection with chine), ft aning speed, fps @ = draft of Keel at transom, ft Por Vertical Equilibrium of Forces: y= N cosr + TP sin(r + «) —D,sinr (29) Por Horizontal Equilibrium of Forces: Teos(r +) = D,cosr+Nsinr (30) For Equilibrium of Pitehing Moments: Ne+D,a- 1 =0 1) For a given boat design the quantities 4,, a, b, « LCG, J, and @ are specified. ‘The unknowns in the foregoing equations of equilibrium are evaluated by a solution of these simultaneous equations together with the planing formulas for lift, drag, and center of pressure, An analytical solution of these equations is extremely tedious and cumbersome and hence a numerical computational MARINE TECHNOLOGYTable 1 Computational Procedure Hy- drodynamie Performance of ‘Prismatic Planing Hull (General Case) EQUUBRUM TRA Ce ‘Trim at which (30) = 0 Assume ling 42 interpaation-between nd = 3" es 149,960, + 836,800 Horizontal Drog Force 3 D = 424 — oss4 ~ san) 5 23° +a ane IvEN ‘& *60,000.8 eni39FT Less2a0rr 1050FT voss20 Fr eat b =14 FTIAVERAGE) B +10" (averace) V_*40KNOTS (67.5 FT/SEC) PLANING COEFFICIENTS sequiReo POWER REQUIREMENT PORPOISING UNIT POWER REQUIREWENT Ve40 KNOTS D = 9605 > Cy *VAab= 40.x1.69//32RRT4 + 3.18 Eevee gt fp V88-60,00070.97%67 s%e14'-0.069 pup = PXV | 9005 079 suis, # 550 550 ef ‘auton Mean Wetted Lensi-Bsom Ratio yank Soares = <2 Tee 3 oma f eu [ais fo Nem 885 — (885 — 2.60) 75 = 3.20 rs Figen tt] aes] 5g Jt Wetted Keel Length fig/t Tt (aay 0397, 025k 0185 ne Bian Run rise 19 {3.85 | 9.60 | 108 Mt a tane ", Figure | 67.0] 66.6 6.2 ens me Wp |5 8) a ake x ETI a] en 8 + 2 tan Schosnterr | 00124 | cor» | 00190. ae ae MREgkaar’ | eso. | coor [com Soerar a, erat (hs (| —o0ait_} ooo | oat tan ERG 13a °, ma eT ke | sie Dra ot Keel reno a8 a> Tales = 860% tn 28° oo bE ag] ae aa ann in tg | sa Fergcing Sony on : 36 Cigh2) = : sis a on Ties $38 O.0a451 0.186 a oy + 05) [9h Ory From Fig. 18, porpoising will occur if ~ 1450; hence, resent pani Dou salable Se s a a5 5 5s : (ee 8 7 . eae 7 wife ie sinfe a) rig 1392 Taster inte 2a 961] 230) (23) eae 5.46 10.70 sie —ece | os an = Pra 4 (8) ease “332,000 | 4s, 000} wea ‘aye a ae A 9 ee) a0] a) + (a) | ees =k alo | 36,600. | a, 0 ‘octoneR, 1964 cyTable 2 Computational Procedure Hydrodynamic Performance of Prismatic Plane ing Hull (Case When all Forces Pass Through CG) GIVEN: REQUIRED: ‘A =60,000 LB POWER REQUIREMENT Los: 290 FT PORPOISING STABILITY 20 FT 14 FT (AVERAGE) lo (AVERAGE) 40 KNOTS: POWER REQUIREMENT a=: fee =O Y= 40 KNOTS (67.5 FT/SEC) PLANING COEFFICIENTS: Cy *V/V gb 40x1.69//3225T4 «3.18 Uae eat © gt 4/ 3p V*8"+60,000/0.975.67.5%14810.069 Row want ‘ ive 1 ct Figure 085 2 7/5 tare 2.07 3 a Figure 19 3.5 oly * cyan Figure 9 Os: 5 oT Oe) 2a 6 cs Ee eo 2.239 a tent 03 8 rea 2,340 9 | nb? (3)b2 6; io vA Figure me | 66.9 1 te Wade 3.22 x 108 2 co Schoenbere 00177 3 eh ‘ATTE,Stendara ‘0008 i eae (ia) + (i) | —0eei7 VRE (cy + a ce) 15 Pe era eos atime | cee 16 oe/eoet Sa 7 ° + G7 | s010 18 | eu 0-5 ¥/550 1100 19 fig 185 a z.porpoisieg | Flgare 18] &a3° Joout 1s seasie procedure is recommended. To simplify the computa- Teose = Asine + D, (2) tional procedure the equilibrium equations are rearranged as follows: Substituting (82) into (29) and assuming that cos « = 1 Tt can be shown that vesults in biel MARINE TECHNOLOGY100 Sse0sGe Res 521 OF /22 42.38 _£ 080 4 ~ a MT : —— 1 & oso cE oer ee : a ae rd & o40 by 5 & Ee LEVEL WATER Ao AsL/b NeRESULTANT OF NORMAL 4 [e BOTTOM PRESSURES ULL uti PH ° 2 3 4 5 6 7 8 9 VELOCITY COEFFICIENT, Cy+W//EB Fig. 17. Ceoter of pressure of planing surfaces A= Neosr + Asinr sin(r + 6) (83) equal to zero and e = 0. This is the condition analyzed that by Murray, Clement, and Koelbel in their respective — computational procedures. The moment equation (31) All = sine sin(s + )] is hence satisfied since a, f, and ¢ are equal to zero, Tt a] G4) is, of course, implicitly’ specified that xC,b — LCG. Substituting (82) and (34) into (31) aft sins + ele _ jsinr cos? + Daf) = 0 (35) When +, ¢, and D, satisly equation (35) the planing hull in equilibrium and the resistance, power, and stability are then easily evaluated. Case When Thrust Avia is Parallel to Keel In many boat designs the shaft axis is nearly parallel to the keel line. If it is assumed that « = 0, equation (85) simplifies to Ale cosr —bsinr] ++ DAa—f)=0 (36) Case When Thrust Axis ond Viscous Force Coincide ond Pass Through Center of Gravity This case is the simplest to evaluate sinee, to achieve equilibrium in pitch, the hydrodynamic pressure force ‘must pass through the center of gravity. Tt is assumed in this condition that the distanecs @ and f and e are octoser 1964 Hence combining (29) and (30) = afesr L rob = L6G ‘These two equations will satisfy the conditions of equilib- ium for the ease when @ 0. There are many practical planing-boat designs wherein’ these conditions are very nearly applicable. 7) Performance Prediction Methods—Computational Procedures ‘The computational technique for the general ease is developed in the form of tabulations which ean be eom- pleted as a routine procedure. By setting e = 0 the computations can be made applicable to Case 2; by setting a = f= ¢ = ¢ = Oand ACjb = LCG, the com- putations can be made applicable to Case 3. For the relatively simple Case 3, the detailed computations can be replaced by a design’ nomogram, General Cose Itis assumed that the hull geometry and loading condi- tions are known and that the trim angle, wetted length, power requirement, and meastire of porpoising stability fre required over a range of design speeds, Specifically a