Sec. 56.5 



OBSERVED SHIP-RESISTANCE DATA 



301 



more than 2.00 m or 6.56 ft in length; some of 

 them were as short as 1.80 m, or 5.91 ft. There 

 are no data to indicate the water temperatures at 

 which these models were run nor what steps were 

 taken, if any, to insure turbulent flow throughout. 

 Copies of these books, for those who wish to 

 use the data, are to be found in: 



(a) SNAME Headquarters in New York 



(b) Bureau of Ships Technical Library, U. S. 

 Navy Department 



(c) Bureau of Ships Preliminary Design Section 



(d) TMB Library. 



56.5 Gertler Reworking of Taylor Standard 

 Series Data of 1954, with Contours of Cr . When 

 the American Towing Tank Conference decided 

 in 1947 to adopt the Schoenherr friction-resist- 

 ance formulation it was realized that this pro- 

 cedure would predict effective powers not directly 

 comparable to those calculated from the original 

 TSS contours mentioned and illustrated in Sec. 

 56.3. 



The differences in the calculated effective 

 powers result from two causes: 



(a) The differences between the friction resist- 

 ances obtained from the Schoenherr formulation 

 and those from the old EMB 20-ft friction-plank 

 results in the model range 



(b) The differences between the friction resist- 

 ances obtained from the Schoenherr formulation 

 in the ship range and from the Tideman data used 

 by D. W. Taylor [Schoenherr, K. E., SNAME, 

 1932, p. 285; PNA, 1939, Vol. II, Table 9, p. 114]. 



Item (a) is reflected as a difference in residuary 

 resistance and thus requires that a lengthy cor- 

 rection to made to D. W. Taylor's Rr/A con- 

 tours to render them comparable to modern data. 

 Item (b) merely requires a substitution of the 

 Schoenherr formulation with the appropriate 

 roughness allowance to correspond to the Tide- 

 man data in the ship-prediction procedure. 



In view of this situation, plus the fact that 

 water temperatures and turbulence stimulation 

 were not taken into account in the original TSS 

 testing, it was decided at the David Taylor Model 

 Basin to reanalyze the original test data on the 

 Taylor Standard Series models. In the reanalysis, 

 the methods and procedures employed were 

 essentially the same as those currently used at 

 Carderock. A total-resistance coefficient for the 

 model was computed, from which an ATTC 

 1947 or Schoenherr friction-resistance coefficient 



was subtracted to give the residuary-resistance 

 coefficient Cr . 



The method of correcting for the effects of 

 transition flow is based on the assumption that 

 at low Froude numbers the specific residuary- 

 resistance coefficient Cr = Rr/(0.5pSV^) is 

 practically constant, as indicated in the left 

 portion of Fig. 7.H. The original data showed that 

 in general Cr decreased with decreasing speed so 

 long as wavemaking resistance was important. 

 There was then a short range of speed for which 

 Cr remained constant, after which, as the speed 

 was still further reduced, the coefficient again 

 began to decrease. This latter decrease was 

 attributed to transition flow, and was ignored. 

 The practically constant value of the coefficient 

 Cr is therefore used for all lower Taylor quotients 

 T„ and Froude numbers, down to values of 

 T, = 0.5, F„ = 0.149. 



Although this procedure is not rigorous, a 

 number of recent tests of TMB 20-ft models 

 which were towed with and without a turbulence- 

 stimulating device indicate that in general tur- 

 bulence stimulation results in no resistance change 

 for models which experience only minor transition 

 effects and those only at the lowest speeds. Good 

 agreement was attained in most of these cases 

 between the residuary-resistance coefficient curves 

 from the unstimulated experiments, faired in the 

 manner described, and those resulting from the 

 tests with the turbulence-stimulating device. 

 This seems to be especially true with forms having 

 the TSS type of bow. Two 20-ft models of the 

 Taylor parent form, having (longitudinal) pris- 

 matic coefficients Cp of 0.613 and 0.746, were 

 tested at the Taylor Model Basin in 1951. In 

 both cases it was found that turbulence stimula- 

 tion Avas required only at low speeds. The pro- 

 cedure described in the foregoing gave reasonable 

 agreement with the curves in the turbulent range 

 [Todd, F. H., and Forest, F. X., SNAME, 1951, 

 p. 678]. 



Corrections for restricted-channel effects in the 

 Washington Basin were made by using the form- 

 ulas given in TMB Report 460 entitled "Tests of 

 a Model in Restricted Channels," by L. Land- 

 weber, dated May 1939, with the appropriate 

 model and basin dimensions. This correction was 

 in most cases small; even for the fullest model of 

 the series it amounted to a decrease in resistance 

 of only 2 per cent. 



The results of the re-analysis of the Taylor 

 Standard Series data, carried out by M. Gertler 



