rAP
CEF
Lab.
.Technische Hoscc,.o
Deift
VIRTUAL MASS AND MOMENT
OF
INERTIA OF PROPELLERS
(Determination of entraine4 water effects, for use inconnèxion with torsional and axial vibration calculations)
By PROFESSOR L. C. BURRILL, M.Sc.,
Ph.D., Fellow,
and W. ROBSON, B.Sc.
23rd March, 1962
SYNoPsIS.The paper describes the results of tests carried out inthe Research Laboratories of the Department of Naval Architecture and Shipbuilding at King s College, Newcastle upon Tyne, to determine the effect of entrained water on the mass moment of inertia and virtual mass of propellers oscillating both torsionally and axially in water These tests were carried out with a total of forty nine I6in model propellers, the chOracteristics of which were varied systematically, and the
resulls, which proved to be extremely consistent, reveal the effects of such important
variables as blade-area. ratio, pitch ratio, number of blades, etc., on the entrqined
mass effects.
From the experimental results, a series of simple empirical expressions have been derived for first estimation purposes, and finally a more detailed method of
calcula-tion is proposed which allows the entrained mass effects to be estimated directly
from the completed propeller drawing, and thus takes into account all the significant
variables, including the actual blade shape and pitch variation finally adopted, 1. Introduction
has for a long time been felt that our knowledge of the entrained
water effect on the mass moment of inertia
of propellers oscillating inthe torsional modes of vibration of main line
shafting, is very sketchyand incomplete.
In this country, it is usual to allow for this
effect bymaking a percentage addition
the calculated mass moment of inertia
of the metal propeller, varying from some 25 per cent to 30 per cent, but
it has been realized that this takes no proper account of the shape of
the blades, or the size of the boss.
In 1950, some preliminary
experi-mental work was undertaken on this subject with a limited range of
small propellers which were vibrated in torsion, in air
and in water and
the results, published in Ref. 1, showed that the percentages to be added
to the propeller mass moment of inertia varied considerably, and were
influenced by both the pitch angle of the blades
and by the amount of
blade surface area.
More recently, in connexion with a large passenger liner, it was decided to carry out a similar test with the l6in.diameter manganese bronze model of the 4-blãded propeller, as fitted to the ship, which had been made for the usual
cavitation tunnel experiments.
This test showed that the percentage addition to be madeinthis instance was
as high as 46 per cent, orabout half asmuch again as the normal allowance, the
pitch ratio m this instance being 1 17 and the
blade area ratio 73 This suggested the need for a systematic series of tests to determine, if possible, the326 BURRILL & ROBSON VIRTUAL MASS AND MOMENT OF INERTIA OF PROPELLERS
effect of pitch ratio, bládè-area ratio, number of blades and other design features, on the proper allowances to be made for entrained water when carrying out the usual torsional frequency calculations, and, the test with the l6in. model propeller having proved to be successful, the present work was put in hand. As the work
proceeded, and the results obtained were found to plot in a very consistent
fashion; it was decided to extend the investigation to determine the axial entrained mass in addition to the entrained inertia effect in torsion, using the same experi-mental arrangement. This, at first, led to only a partial success, as the long thin
rod used in the torsional tests was found to oscillate in other than purely axial
modes; but this difficulty was overcome by using a much stiffer and shorter rod
in conjunction with a horizontal spring, and the results, as described in the paper, were found to be entirely successful. Various methods of plotting the experimental results were tried, and some of these are recorded in the several diagrams given in the paper. Finally, a method of calculation was devised whereby both the entrained inertia in torsion, and the added mass in axial translation may be estimated directly from a detailed drawing of the propeller.
2. Test Procedure
The method used for determining the polar mass moment of inertia of the
seveial propellers, in air and in water, was similar to that employed in the earlier tests described in Ref. 1, but owing to the smaller size of the model propellers, which were uniformly 1 6in. in diameter, several refliements were introduced in
order to secure the necessary accuracy. In the first place, the propellers were
suspended on a thinner rod, which was m m diameter and 14 5ft long and
the irieasureinents were calibrated by swinging three steel discs of known momentOf inertia. Secondly, the frequency measurements were obtained by making an
extensive record of the oscillations on a paper strip by means of a capacitor
transducer and an accurate secondsclock, the propeller being displaced slightly
man angular direction and then allowed to oscillate freely in torsion, the lower end of the rod, which projected some 3m, below the propeller boss, being
sUpported in a socket to prevent sideways motion.
Changes in capacitance proportiOnal to the angular displacements, occurred when a horizontal centre plate, rigidly attached to the oscifiatisig rod, moved cyclically between two fixed plates attached to the stationary test-rig. The
changes of plate area and hence capacitance altered the frequency of a pre-set
oscillator, the output of which was converted from frequency to voltage variation by a lOw pass filter device operating at the centre of the cut-off frequency. In
this position, the slope of the response curve enables changes of input frequency
to re-appear as changes in output voltage, which, after suitable amplification, operate the armature coil of the pen-motor and the attached writing stylus. A
second pen, activated by the seconds clock, provides a time trace for calculation
of fEequency, and hence 1/f2. M the initial displacements lay well within the
elastic limit of the steel rod, the frequency of the recorded oscillations remained
independent of amplitude, throughout the test records.
The value -of the entrained inettia (4), reckoned as a percentage of (Ia), the
moment of inertia of the propeller in air, may be calculated from the experimental
frequency data by means of the formula:
Percentage (4)
(fwl_fiil)
< 100 = (4/li,) X 100 %The numerical value of (4), however, cannot be determined unless the value of
(I,) is known.
The percentage figure derived from this simple formula includes a residual
frequency error, since fa (and fw) are the frequencies of oscillation of (propeller
plus rod and fairings), and are related to (l + 4) say, and (Ii, + I + 4), where 4 is the inertia of the light rod and fittings. In the numerator, the residual inertial0 cance1 out, but 4 persists in the dénominator
BIJRRILL & RoBsoN: vIRTUAL MASS AND MOMENT OF INERTIA OFPROPELLERS 327
denomihatorproduceS no significant error, but with the smaller model propellers
the residual value of I, cannot be ignored For this reason and also to deter-mine the numerical value of 1,, for the propellerin air the apparatus. win-
calF-brated by ineans of three discs of known momentof inertia.
The frequency of oscillation was measured for (Ia)
(I + 13) (I + I + 1)
and (I + 4 + 4 + 4) where 11 4 and 13 were the known
moments of inertiaof the discs and a calibration curve was then drawn as shown m Fig 1 which shows the actual calibration curve for torsional oscillation where (1 If') is
plotted to a base of moment of inertia in (lb.ft.2) units. The value of 4 is given
by the intercept on the "x" axir.
The example given in this diagram indicatçs the manner in which the thêasured
entrained inertia was derived, by interpolation -on theabove curve, i.e.,
(I + 1)
(4 + Ip) = (1 -
Je (lb.ft.2 units)The smaller numerical values of 4 derived from the percentage formula may be observed by notmg the displacement of the dotted percentage value line from the actual calibration line shown in full.
The complete apparatus is shown on the right-hand side of Fig 16. The depth of immersion was limited to about l2in., beyond which depth no sigthficant changes in frequency were observed.
To determine the entrained mass in the axial direction, the propeller was rigidly attached at mid-span to a simply supported steel beam by means of a
short stiff vertical rod, of length 1 8in., sufficient to permit immersion to a depth
of l2in. The apparatus was first calibrated by the addition of knownweights to the free end of the vertical rod The calibration curve for axial vibration (Fig 2) shows (1/f2) plotted to a base of mass in lb weight The intercept on the "x" axis again indicates the residual mass of the beam plus fairihgs. The
accuracy of the apparatus is exhibited by the fact that the values of (1 /f2), for all the propellers tested, lie precisely along the imtial calibration hne when tested in air.
The example given on the corresponding calibration curve, Fig. 2, shows the
method of determining the entrained mass We from W in
the immersedcondition and W,, in air, i.e.
(W0+ W)(W0+ W)(W W)= Welb.
The recording of the vertical frequencies of vibration in air and in water was carried out by means of a condenser device similar to that used in the torsional oscillation tests. For this application, the centre plate was attached to the beam at mid-span and oscillated between two fixed plates attached to the stationary test-rig.
The apparatus used for axial vibration of thepropellers is shown on the left-hand side of Fig. 16.
A complete lit of the propellers tested and of thecorresponding measured values of Wp, Ip, We and Ic is given in Table 1, together with the calculated values obtained by applying the proposed new method of estimating these quantitiCs which is described in section (4) of this paper, and illustrated in Table 2.
As a matter of general interest, the values of We and 'e are also expresed in
terms of percentages of the cOn esponding Wp amid Ip figures.
3. Experimental Results
Torsional
The first use which was made of the experimentally derived values was to plot these against the major parameters (i.e. pitch ratio and blade-area ratio, etc.), to
find out whether these exhibited any consistent trends, or laws of variation,
328 BIJRI1.IIL & ROBSON: VIRTUAL MASS AND MOMENT OF INERTIA OF PROPELLERS
exaxñple; shows the measured values of Je the entrained inertia determined for the' KCA 3bladed series of Admiralty type propellers, plotted to a base of b.a.r., the blade-area ratiO. This shows that the measured values for each pitch ratio were readily separated, and that the variation was substantially linear with
blade-area ratio for each pitch zistio. Fig. 4 shows the same results plotted to a base of pitch ratio (P/D). This diagram shows that the law of variation with pitch ratio was also substantially linear for each blade-area ratio. This suggested that it might be useful to plot the results to a combined parametet (ba.r. x PJD)
and this is shown in Fig. 5. It was then found, that all the results could be represented with reasonable accuracy by a single straight line, which could be used as a basis for a simple empirical formula, for thst éstimatiön purposes.
Strictly speaking, this fOrmula, which is
Entrained inertia (Ic) = 1 37 (b.a.r. X P/D)-030 lb.ft.2'
can only be applied to propellers conforming to the basic 3-bladed design, but the diagram also includes the results obtained for the KCD 4-bladed series of merchant-ship type propellers, and several results for 5-bladed and 6-bladed screws, which, although not sulficient for a very complete analysis to be made for such propellers, would suggest the following additional formul for first
estimation purposes:
lb. ft.2
4 (four blades) = lO9 (b.a.r. x P/D) - O23
le (five blades) = O98 (b.a.r. x F/D) - O21
'e (six blades)= 090 (b.a.r. x P/D) - O2O
It will be seen that the entrained inertia dftnihishes with increase in the number of blades it also increases steadily with increase in either the blade area ratio or pitch ratio.
These empirical formula have been derived directly from the experimental
values of I for the model propellers which were all 1 6in.. in diameter, so that in
order to obtain the corresponding values for full-size propellers it is necessary to multiply the results by the appropriate scale factor (Did)5, where D is the
diameter of the full-size propeller and d is the diameter of the model propeller. Furthermore, since the results were obtained in fresh water, it will be necessary
to apply a further correction for density, i.e.
I D'\5
Entrained inertia (full-size) = 4 (formula) x x 1O24 (lb.ft.2)
Axial
Fig. 6 shows the ineasured values of We the entrained axial mass for the KCA
3-bladed series, plotted to a base of b.a.r., and Fig. 7 shows the same values plotted to a base of P/D.'
It will be seen that while the values of We increase linearly with h.a.r., they
decrease steadily with increase in, pitch ratio, and it was therefore found
impos-sible to obtain a single line plotting by combining these two parameters. In view of this, a new parameter Cos2O was tried in place of F/D, where is the
pitch angle at two-thirds radius. This is 'shown in Fig. 8, from which it will be
seen that the values of W increase linearly with increase in this parameter. Finally, Fig. 9 .shows the same results plotted to a base of the combined
para-meter (b.a.r. x Cos282). It will be seen that all the results now lie substantially along a single line, which can be represented by
We = axial entrained mass
347,(b.a.r. x Cos2O) - 4-2 lb.
(3 blades)an approximate formula which can therefore be used for first estimation
purposes.The results for 4-bladed, 5-bladed and 6-bladed propellers are not sulTIcient to allow a similar analysis 'to be made, but as it is not expected. that this should
BURRILL & R0BS0N: VUiTIJAL MASS AND MOMENT OFINERTIA OF PROPELLERS 329
indicate any majOr differences from the 3-bladed analysis, the following
appr9x-irnate formul are- suggested for 4bladed, 5-bladed and 6-bladed screws,
respectively, based on the limited, data available.
(4-blades) We = 347 (b.a.r -x cos2
- 67 lb.
(5-blades) We
341 b.a.r. x Cos2 0
- 83 lb..
(6-blades) We347 (b.a.r. x Cos2 0) - 96 lb.
As it is not anticipated that 5-bladed and 6-bladedpropellers will- be adopted
over a wide range of blade-area ratio or pitch ratio, and the propellers tested are
fairly representative of possible applications it is considered that the above
expressions should be quite adequate for present requirements.
These empirinal expressiOns have also been derived, directly from the
experi-mental values of We for the model propellers, and
therefore for full-size propellers the values will be obtained fromAxial entrained mass (full-size) - We:(formula)
X (
j___3)x l024 lb.
In order to complete the'analysis, Fig. 10 has been prepared showing the variation of 1e'wjth respect to the çomplèrnentary function sin2 0a for torsional motion.- 4. Cakidatioñ Method
Torsional
Having obtained the above data from experiments covering quite a wide range of variables, it was thought desirable to explore the possibility of establishing a method of calculation which would enable the designer to-estimate the entrained main and inertia for a particular propeller directly from the drawing, as is done; for example, in estimating the propeller weight and mass moment of inertia. As
a starting point, use was made of the knowledge gained in dealing with the similar problem of estimating the entrained mass of a. ship vibrating as a
free-free beam. . . .
Basically this is dealt with by using the simple expression forthe entrained
mass of an infinitely long cylinder of circular cross section oscillating sideways
or at right angles to its axis in which case the entrained mass per unit length
is given by
Entrained mass = iT/4 b2 x p per unit length where b beam (or diameter of the cylinder)
p density of the fluid
-and for ship calculations this expression is extended to entrained mass =
C;J. 'x n/4. b2 x p per'unit length, where C is a constant which allows for
the 'shkPe and proportions of the-actual cross-section a 'any point of the length,
and J'is a further
onstant which takes account of the finite length. The factor - is inti:odi.iced becaus,e the ship is floating at a freesurface.-In the case of a propeller oscillating m torsion, the problem is to some extent simplified by the fact that the blades are essentially flat plates, and the influence
of thickness and section shape is therefore probably quite small,
but, on the
other hand the sections are not moving in a direction perpendicular to the chord, the effect of interference due to overlapping of the-blades is unknown, andthere is no theoretical treatment which would assist in the assessment of the
three-dimensiOnal constant J. Nevertheless, as a first approach, it was decided
to make a preliminary attempt to resolve the problem by assuming that the
amount of the entrained volume of fluid at each radius. could be.represeited by drawing a circle round the extremities of the sections as projected on to a plane
at right-angles to the motion the area of this section being given
by iT/4330 BURRILL & ROBSON: VIRTUAL MASS AND MOMENTOF INERTIA OF PROPELLERS
effect, of the finite blades this basic area at each radius was then reduced by applying a factor K which- varied elliptically from the axis to the tip,, and the
mass moment of inertia of the volume so described was calculated in the usual
way. This tentative procedure was found to give results which agreed very closely with the experimental figures for propellers with moderate pitch ratios and low blade-area ratios, but failed to predict those for the propellers -having
high pitch ratios and very wide blades, for which the calculated values were too
high. This suggested that, apart from the influence of blade interference, there was probably an aspect ratio or (blade-length/blade-width) effect, which was not
included in the calculations.
In the light of this result, a second attempt was made by applying the
three-dimensional inertia coefficient factors k' obtained by Lamb' (Ref. 2), ; 155, in-connexion with the calculation of the entrained inertia for prolate ellipsoids of various proportions, rotating about a minor diameter. These factors k' give the relationship between the moment of inertia of a volume of fluid described
by circles at each radius corrdsponding to the elliptical distribution, and the final inertia of the actual entrained fluid volume, and are therefore directly applicable to the basic volume used in the previous estimate.
-The new approach did, in fact, indicate an important aspect-ratio effect (in terms o.a/b where a = major axis radius, and b = minor axis radius of the ellipsoid), but the application of these k' factors led to the unexpected result that
while the calculated values again agreed closely for the moderate pitch ratios
and low blade-area ratios, those for the higher pitch ratios and blade areas now
worked out below the experimental values. This indicated clearly that the correction made for aspect ratio was too great for the wider propellers, and an examination was therefore made of the actual projections of the blades on to a transverse plane, in relation to those of the several ellipsoids, in an attempt to
find for a given propeller blade shape an equivalent ellipsoid which would have the same entrained inertia. Although this procedure was not successful, it did
draw attention to two factors which have an important bearing on the failure
of the ellipsoid analogy. The first of these is that while the projection of the
blades on to a transverse plane is approximately elliptical for the moderate blade area ratios, it is distinctly wide-tipped for higher blade-area ratios. The second
is that for a given diameter (2 a) there is a maximum value of 4 which can be achieved as the axis diameter (2 b) is increased, and that thereafter the value of 4 diminishes to zero when 2a = 2b, as the ellipsoid then becomes a sphere.
This is illustrated in Fig. 11 where the experimental values of 4 for the 3-bladed
series are plotted to a base of b, the equivalent mean value of the fore and aft
projection of the blade widths, extended to the axis.
As a further extension of this procedure, the values for the 4 derived from the
above mentioned work of H. Lamb have been calculated for the special case when the dimension c at right angles to b approaches zero, and the ellipsOids therefore become flat elliptical plates set at right angles to the direction of
motion. Itwillbe seen that this explains the failure of the previous analogy, as
the calculated values now lie within the region of the experimental results, but
that this method of approach still does not fully explain the effects of blade-area ratio and the influence of pitch ratio, as revealed by the experimental results.
As a result of this work, it was decided to approach the problem in a different
way, and to endeavour to derive the appropriate K factors from an analysis of
the actual measured results This was done by plottmg the values of a derived
K factor represented by the ratio
measured- value of I
-K(fl
- calculated value of 4
for the propellers of the 3-bladed series against various parameters.
Theresult of this examination is illustrated in Fig. 12 i,vhen it was found that the
majOr variable was undàubtedly the blade-area ratio, as the value of K1 for each
pitch ratio was sensibly constant over the middle part of the range, and only
-BURR1LL & RoBsoN: VIRTUALMASS AND MOMENT OEINERTIA OFPRQPELLERS 331 It was therefore decided to re-calculate all the valuesof 4 for the full range of propellers using a single line of K5 plotted to a base of b a r as shown in Fig 13
These new calculated values were then plotted against the residual parameter pitch ratio and the resultant errors. were so small that it wasP considered un
necessary to apply a further correction for pitch ratio. The final calculation method is illustrated in Table 2,, for the unity pitch ratio, 65 b.ar. propeller.
Axial Entrained Mass
The above described investigations were all undertaken in
relation to the
entrained mertia values, at a time when the axial entrained massvalues were unknown and a supreme test of the final method arrived at as a result of thevarious speculations, was to calculate the entrained axial mass values in precisely
the same manner, but substituting the value of (chord x cosine 0) in place of (chord x sine 0) at each radius This was, in fact done m one day, and the results when plotted in the form of
- measured value of axial enttained mass
K axiaj)
-
calculated valuewere found immediately to fall in line
Having regard to the fact that the
quantities 4 (lb. x in.2) and We (lb.) are of a different order, this was considered
to be a sufficient proof that the basic method of calculation was at least sound in principle.
It was not expected that the values of K(i,tja) and K(iM)
would necessafily be equal, but it will be seen that they follow similar line or
trend with b.a.r., and in fact the rãtiô
K(jacrtja)= 1 O8, was found to be
constant over the full range of b.a.r. values tested.The same procedure was later applied to the 4-bladedpropellers of the KCD series and the resultmg values of KA and K1 are shown in Fig 14 It will be
seen that these are slightly higher than the 3-bladed values, but that they
exhibit the same trend with b a r
and can, in fact be related to these by a
simple ratio, throughout the range of the tests. Similar curves for the 5-bladed
propellers are shown in Fig. 15, together with the correspondingvalues for the
single 6-bladed propeller tested. For calculation purposes, the values of K1 and ((A are listed at 02 intervals of b.a.r. in Table 3-for 3,4, and 5-bladed propellers
respectively.
5. Further Discussion of Experimental Results
It wifi be obvious from Table 1 that the method of estimating the entrained
water inertia by adding a fixed percentage to the propeller mass moment of inertia is totally inadequate excepting only for verysmall variations m b a r and pitch ratio from a given basic propeller. It so happens that there were three variations
of the KCA 110 propeller hiOh' were identical in all respects apart from the blade thicknesses, m that with respect to the normal propeller of 045 blade
thickness fractiOn, one model was deliberately made thinner,and another made
thicker, the corresponding blade-thiOkness fractions being, respectively, O3O
and 060.
Tests were therefore carried out with these three propellers and theresults are tabulated
below:=-Basic Propeller KCA 110 Blade-area ratio 80 Blades 3
Pitch ratio 1-0 Blade thickness fraction 045
Propeller Wp Ip
4
% additiOn- Calculated 4KCA 110 (light) 1381
l083
-765 70-6% -746KCA 110 (normal) l7l2
1685
-785466%.
746KCA llO(heavy) 20-12 2-004 -761 38-0% 746
Propeller Wt (Wp) We % addition Calculated W
KCAIIO(light) 13-81 18-00 130-3% 17-78
KCA11O(normal) l7-12
.i798
1Q5.0.%1778
KCA11O(heavy)2Ol2..
1.7-70 88-0% 17-78332 BVRR]LL & ROBSON: VIRTUAL MASS AND MOMENT OF INERTIA OF PROPELLERS
It will be seen that the measured values of 'e and We are quite independent of the propeller mass moment of inertia and weight, and are governed only by the geometrical characteristics of the blades. It is also clear that the thickness of
the blade sections 'is of no importance in determining the entrained water mass. For the 4-bladed KCD series there were also thtee variations of the unity pitch
ratio, 587 b.a.r. propeller, in that, relative to the basic propeller, one had wide-tipped blades with narrow root widths and another had narrow-wide-tipped blades
with wide roots, all other characteristics remaining the same. This provided.the
possibility of makmg a check on the adequacy of the proposed calculation procedure to reflect correctly the effect of blade shape n the entrained inertia
and mass values..
The results obtained with Propeller KCD 4 (wide tip) KCD 4 (normal) KCD 4 (narrow tip) Propeller 'Wp We
KD 4 (wi4e tip)
.' 19, 38 '978
KCD 4 (parent)l938
.,947
KCD4 (narrow tip)2038
9.39and it will be seen that, the measured values are completely calculated figures. '
Finally, it was decided to determine the effect of removing two opposite
blades from One of the four-bladed propellers, and the results are given below:
'Basic Propeller: KD 2
B.a.r. = 50 P/D = P846
Blades: 4 Propeller IJ) KCD 2 (4-blade) I o99 KcD 2 (2-blade)559
Propeller Wp Kcp 2 (4-blade) .l388
KCD 2 (2-blade) 9 12these propellers are given below:
'P
4
l429
.393l387
385
l440
' 381 Calculated 'e' 39l 381 .377 alcula1ed We 9.79 9.57 9 50 in accord with the4
Galculated 4 C'orresponding K, -'226 ' ''235 ' -875il8
'118
108
We Calculated 4 c'orresponding K4 7-83 7-88. -938 4-23 4-23. 1-00It will be seen that in this special case of two opposite blades the factors K4 and Kj-approach to unity,so that the initial assumption as to the natifre of the entramed volume being descnbed by drawmg a circle round the extremities of
the sections as projected on to a plane at right angles to the motion is reasonably
correct; and no further correction is required.
-' ' 6. Final Comments and Suggestions
It is' not suggested that the results given by this series of, tests should be regarded as a complete and final resolution of the prOblem of estimating the
entrained mass effects to be included in the usual calculations for the torsional
and axial criticals of main hne shaftmg
It is, however thought that they
provide a much more fleAible and logical means 6f estimating these effects than the present method 'of making a percentage allowance, based on the calculated weight and mass moment of inertia of the metal propeller. The final adequacyof- the proposed method of calculation, for full-size propellers, can, only be
assessed after it has been applied, in practice to a wide range of actual
installa-tions, but the consistency of the results Obtained on the model scale gives
sufficient confidence 'in the method to suggest that it should at least be tried Out
in practice. . . .
Th specimen calculation given in 'Table 2 has been' drawn up in terms of the standard radii used in defining the m'ödèl propellers, but can readily be adapted
BURRiLL & ROBSON VIRTUAL MASS AND MOMENT OP INERTIA OF PROPELLERS 333
to suit other methods of definingthe blade sections of the full-size. propellers (i.e., Ift. intervals, etc.) The same calculãtiön form can,. in fact, readily be extended to include the calculation of weight and polar moment of inertia for the metal propeller (Wy) and (I,,) by inserting the appropriate areas of the
bladesections at each radius and using the same Simpsons multipliers to find the corresponding volume and inertia functions, down to the crown of boss. The volume and polar moment of mertia of the boss can then be calculated iepãrately
and added to that for the blades, thus giving the radius of gyration and mass
moment of inertia of the completed propeller Experience has shown that the final figures thus obtained agree within one per cent with the results of full scale tests obtained by swinging large propellers in air, using the usual tri-filar method of suspension.
During the course of the investigation attention was directed to two paper (Ref 2 and Ref 3) which had been written on this subject in Athenca and in
particular to theeiiipiricalexpreSSiOris
-.
O2lpD(WR)2fl
e -
[1 + (p/d)2] (03 F MW1)
-..()
Ic= O02 W .p2
...(2)
which had been suggested. An attempt was made to interpret the results
obtained in the present series of tests by applying these formuhi, but it was found that although they had a linlited application to the propellers of small b.a.r. andlow pitch ratio they failed to predict themeasured results for the wider blades
and higher pitch-ratios. For example,
One important point did, however, emerge from this examination, and that is
that the authors of Ref. 3 showed by their analysisthat the apprOpriate value
of W,, to be Used in the axial vibration calculation was not that obtained from
the tests with rotational restraint, as described in this paper, but an equivalent entrained mass W' given by
We
We'
=
W lb...(3)
+\2)
i,
fp\2
Wewherep = pitch in feet, and the term (j x
\.71J1p
is a coupling factor to allow for a propeller inlongitudinal (i.e., axial) motion, but unrestrained by the shaft in rotation. The writers of the present paper are unable to comment with authority on this matter, as they have no extensive
experience of the results of axial vibration calculationsin relation to measured values obtained from actual ships, but having studied the analysis presented in
Ref 3 it would appear that this correction is fullyjustified
Itis hoped that
marine engineers who have had practical experience of the results of axial
vibration calculations will be able to comment on this point. Propeller
Ref.No. P.R. B.A.R.
Exp.W
Formula
w
Exp.Ie Formula -Ic 1 -6050
11-21 11-89 190 -152 2 -6065
16-04 17-42 -256 -223 4 80 -50 10-44986
-288225
5 -80 -65l534
14-44427
329
28 1-60 951748
il56
l85
1-05 32 2-00 95 15-28822
2-33117
334 BIJRRILL & ROBSON; VIRTUAL MASS AND MOMENT OF ThJERTIA OF PROPELLERS
A further matter of interest is the relationship vetween 4 and We which is
implicit in equation (2) above, which may be re-writtenas
= OO2p2
This has been examined in relation to the measured values obtained in the
present series of tests, and has been found to be substantially correct, although
the values of the constant vary from about OO23 to 0O25 and a more suitable
mean value appears to be about 0024.
Acknowledgments
The authors wish to acknowledge their indebtedness to Mr. R. H. Curry, B.Sc., for his valuable assistance in connexion with the development of the calculation method described in the paper and the preparation of the various diagrams, and
also to Mr. R. Greenwell who carried out the tests in the
laboratory. Dr. R. L. Townsin is also to be thanked for his exploratory workon the ellipsoid analogy described in Section 4 of the paper.BURRILL & RoBsoN: VIRTUAL MASS AND MOMENT 01' INERTIA OF PROPELLERS 335
REFERENCES
BURRILL, Prof. L. C. and BoGGIs, A. G.," Electronics in NavalArchitecture,"
N.E.C.Inst., 67, 1951.
KANE, J. R. and MCGOLDRICK, R. T., ' Longitudinal Vibration of Marine
Propulsion-Shafting Systems," S.N.A.M.E., 57, 1949.
LEWIS, F. M. and AUSLAENDER, J.," Virtual Inertia of Propellers."
Detailed drawings of the KCA and KCD propellers are to be found in the
following:-KCD Series: Buiuw.i, Professor L. C. and EMERSON, A., "Propeller Cavita-tion: Some Observations from l6in. Propeller Tests in the New King's College Cavitation Tunnel," N.E.C. Inst., 70, 1954. KCA Series: GAWN, Dr. R. W. L. and BuRIuLL, Professor L. C., "Effect of
Cavitation on the Performance of a Series of l6in. Model
TABLE IList of Propellers Tested w Pro eller Number P/D B.A.R. Number Of Blades We! hr wp. lb.
Measured Values Calculated Values Percentage Values
Entrained Mass We..lb. Moment oJI,zeriia lb. ft.a.
jp.
-Entrained Inertia 'e lb.ft:2 -Entrained Mass W. lb. Entrained Inertia 'e lb:,(t.2 (W 1We0, P /0 (1 /1 0I '60 '50 3 1I'19 I1'21
983
. 9Ø
.1094. -'-164 1002I93
2 -60 65 3 14'56
l604
1'290 -256 1599 '240 I10'2198
.3
'60 80 3 18062074
161I
'2992059.
-311
114-8 .185
4 .50 31I06
10-44 .983..
2881020
.27-3 94'4 29'3 5 8065
3143l
15'34i'3d4
427
14.92 399 107.2 -327 6 '-80 80 3 1825 19'75' 1.604.. 562
1922
'5171082
350
7 '80 '95 3 22'OO 24'75 1955 - '7O5'2458
'660
112'5360
8 -80110
3 2650- 29'90- 2'282 . -832 .29-74 795 112'8 . 36'4 9 I'OO 50 3 I0'94946
960 .357 9.45 -395865'
372
10 i'OO65
3 1475 13"64 1'300 -'581 -1382- ;577 92'S 44'7 111(L)100
3i381
1800
1"083 765 17'78- '.746 130'-3706
12(P)i00
80 3 1712 17'98. 1'685 . 785 1778.-. '746 105'0466
,13(H) 1.00 . '-80 3 20 12 17'70 2'004 - '761 17.78:.746
.88'O380
14 1.00' . "95 - 3 22'12 22'88 1'893 ,94-32275
954I034
49'-8 15 1.00110
3 -27'122668
2391 -F095 2753- ,1.'14898'4
45.7 16120
'50 3 't1'12828
'976 51-3 8;68 :52! 74.5526
17 1'20 '65 3 14'56 12'32 1'-267 -'765 1269 '763-84'6
604
18 1'20 '80 3 1756 16'62 1-584 ,'993 -16'34 '986946
62'7 19120
.95 32100
,20'30 1"901 - F22'9 20'90 1-259.967
65'i
20 1'20 1'lO 3 25'38 25'OO 2'293 1'4952529.
1'520 - 98'S 65"2 21 - 1'20 1'25 3 27'122870
2622
17412950
1'7-75 105'8 66'4 22 23I'40
' 65 3 14'19 1141 1'264 P893i162:
-951 .. 80'4706
140
80 3 17'19 14'81 1'547 - '1231 '1496 1'232 - 86'2 79-5 -24 1'40 '95 3 2-1.75 18"95 1'915 1'555 19'15 1.574 87'I.811
TABLE 1List of Propellers Tested (continued) Propeller Number ,
'"
.'"
NumberDi.4es . Weight Wp lbMeasured Values Calculated Values Percentage Values -Entrained, Mass We. lb. Moment of Inertia lb. 11.2 Jp Entraüed Inertia le lb.ft.2 Entrained Mass W. lb. Entraized inertia
Ie lbfe.2
-(WIW
0/ 0 (Je/Ip)0/ 0 26 27 28 29 30 31 32 33 34 p35(P) 36(N) 38 39 40 41 42 4.3 44 45 46 47 48 49160
160
60160
200
200
200
60 983 983983
120
140
1'60 983 983 983 846846
965117
1181 .73 65 80 95110
50
65
95587
587 587 587 587 P587 587587
587x587
x587
50
25588
73 802.4
3 3 3 3 3 3 3 4 '4 4 '4 4 4 4 4 3 3 4 4 2 5 4 . 5 6 14381819
2156
2650
1131 15122262
2006
.1.956 19382038
19381962
1862
18752100
1819
1'638''
1318912
19252325
2250
1.225i022
1381 .17482050
609
878
1414
1164
1022 .947939
978 '
868
758
685
10.60 1'41602
783
423
863
12'75.1310
9.95 1285 1591 1917 2351 .9731317
1994
1405 .1366
1:387 1.440 1429 14011350
i386
1'3831219
1149
1099
.559-1405
i'598
1679 1.256 1081 14231856
2239
963 1342.2328
.197 297 385 381 393: 499 P594 .738517
.334263
-226 118 '326 .743 752 . 2251062
1367
1748
2i16
604
813
1454 1137 1051957
950
979
.851
7.73696
1014
681
622
788
423
8'oS1264
1311995
1125 1'457 1''8622242
10O9 F4742438
P168 P276 381 , 377 391 510,627
738450
-298254,
P235 P118 ' '330 741 763 225 71-1 75.9811
774
531
.581656
58Q522
489
461
.,507:442
407
365
505
407
368
564
464
441
-,548
58'2 81'2 841894
961
952
99O 1O19 1:16.8 14O2!7
28'4264
274.
35:.6 44'Q532
374
274
228
206
211
23'2 46'5 44'8 17'9K.C.A. 410 PJD
10
B.A.R. = 0.65.
From Fig 13 (orTable 3)
K, =0599andK4
O646Entrainedinertiaje =1
R2 xx N x , x K1 x
jb.fj.2=
x 64 x
! x625
>x 381164 x0599
Calculated
le = 0577 lb. ft.2
Measured: value I = O58O lb. ft.2
Axial entrained mass We
=
N X X L4 ' K4 X lbWe =
>< 3IT x625
x 375 9367 x 0646x1
Calculated
We = i382ib.
Measured value
We = 1364 lb.
TABLE 2Specimen Calculation for
4 and W
Number of Blades
- 3 = N
/R Pitch iii.Tzo,
=
00 Chord in in. (Ozord x Sin 01 (Giord XSifl 0)2 / R2 S M. Function (C'hordx Cos 0) x C'os 8)2(Chord Function
o 250
0375
O5OO 0 625 Ø75Q0875
1 000 -l6in l6in. l6in. l6in l6in. 16in l6in 1 273108489
06365
0 509304244
03638
0 318351 51mm
40°-20min. 320-29min.27° 0mm
23° 0mm.
20° Omia.
17 39mm 4 38593
725
8 26879
8'25 -3 4443838
3893
3 7503434
2822
-0 7414 2O714 37889 5 4932 66332. 6O972 -2 1 2 1 2 0 3707 4 1428 37889 10 9864 66332 12.1,944 2 7064520
61I5
7 360 80917753
7 3224 2O4304 3:73932 54 169665464
60'1090 3 6612 408608 37-3932 108 3392 654643 i'202'1'80= 381164
3759367BURRIL.L & RoBsoN: VIRTUAL MASS AND MOMENT OF INERTIA OFPROPELLERS 339
TABLE 3K, and KA Factors for 3, 4 and 5bladed Propellers
B.A.R. K, KA 3 Blades 4 Blades 5 Blades 3 Blades 4 Blades 5 Blades 0-50 0696 0875 0752 0938 2 0-683 0-840 0737 0-905 4 0670 0811 0899 0723 0878 0-977 6 0-657 0788 0-873 0700 0853 0950 8 0-644 0-768 0-850 0-695 0-832 0926
060
0-631 0750 0830 0681 0-814 0905 2 0618 0.734 0-812 0667 0798 0-887 4 0605 0720 0797 0653 0783 0-870 6 0592 0708 0-783 0-639 0-770 0-855 8 0579 0695 0-769 0-625 0757 0840 070 0567 0685 0757 0612 0-745 0826 2 0555 0674 0745 0-600 0733 0813 4 0544 0663 0734 0588 0-721 0801 6 0-533 0723 0576 0789 8 0522 0-712 0564 O777080
0-511 0701 0552 0765 2 0500 0-690 0540 0753 4 0490 0680 0529 0-742 6 0-480 0518 8 0-471 0-508090
o462 0498 2 0453 0488 4 0-444 0479 60435
0470 8 0-427 0461 1-00 0419 0-452 2 0-412 0444 4 0-403 - 0436 6 0396 0428 8 0-389 04201iO
0-382 04130+ 04 03 0 02 U 'U 02C 2
0
Lil 015 a-010 005340 BURRILL& ROBSON: VIRTUAL MASS AND MOMENT OF INERTIA OF PROPELLERS
,
F /,/
.. -x 1, BY :AL,eRATIo 0f /- !f'BJLk__
I 4-41
'f .020H
.0 25 3 35 +0 MOMENT OF INERTIA IN L FTZ0020 0016 0012 000 000 kU
r
C -4r
> z C C z - C z -4 > C '1 C -1'lr
C, ._!PQ1!U041 N'Il..
-CALlbRA1IOH ,411614TN'II._
CALIBRATIOkI JEIIfl -W.IM1ER:EPr O -3CAIJ. I W N'!l W1 4-4 7geIb.-W..31.I I-Fol N'U.
1 + B 16 20 24-PROPELLER WEIGHT IN LB.
342 BURRILL & RoBsoN: VIRTUAL MASS ANDMOMNTOF INERTIA OF PROPELLERS LU
z
LUz
z
LUBLADE REA RATIO.
Fig. 3Measured Valuesof 4to Bzse
of
B.AR. (3-bladed Popellers)06 04 10 12
PITCH RATIO.
Fig. 4Measured Values of 4 to Base ofPitch Ratio (3-bladed Propellers)
----.0 .A
d4L4
05 06 07 O8 09 .10 24 1:20 -I6 LU 12 LU I.- 08 LU04
16 lB30
BURRILL & ROBSON: VIRTUAL MASS AND MOMENT OF INERTIA OF PROPELLERS 343
S 24 0 II 'i-I , 3øLAE RLAE5. 0+ o.s a 16 BAR. x P..
Fig. 5Meàsurëd Values of 4 to Base of
(B.A.R x P/D)
to
05
06. 07 0503
BLADE AREA RATIO:
30 25 2 20 LJ
z
z
-J< 10
>c 5344 BURRILL& ROBSON: VIRTUAL MASS AND MOMENT OF INERTIA OF PROPELLERS
30
25
5
PITCH RATIO.
Fig. 7Measured Values of W to Base P/D
06 07 05
VALUES OF CO5Gz.
Fir. 8Measured Va/ties of Wç to Ba.e cos2o ,
09
- 06 08 tO lZ 14
t6
- IA En--30
25
S
BURRILL, & R0BSON:. VIRTUAL MASS AND MOMENT OF INERTIA OFPROPELLERS 345
02 04 Ob b
BAR. % COS29a
Fig. 9Measured Values of W to Base B.A . R.
Cos2 0213
20
01 02 03 0
VALLJESOF 51N29
Fig. 10Measured Values of
4
to Base5j,12 0 2/3
I,-'I
s4 I
It.-
SBl.ADE -J4ADES.fiNs
s/f...-o'
r#A
..-a w
z
w2
z
346 AURRWL &Ronsoi:vijftuAL MASS AND MO?kENTOF INERTIA OF PROPELLERS
2-16 pa 08 0+ 1 b iN JNCI4Z5.
Fig. lIC'omparison oJ Measured- 'e Valus with Theoretical Values
- ---- for Prolate Ellisojd
/
-- //
- - '-a b rtho,. C &AbIu ANGLES T b. CONWrANT Aii 1ADiU AT QuilTI,
-
1/
--f
-..-- __i-:
07
06
05 ow LU 0 002
0IBURRILL & RoBsoN: vIRTUAL.'MASS AND MOMENT OF INERTIA OF PROPELLERS 347
050 8A.R.!065 S 080 O B.AR .O95,
Il0,
S 06 08 (0 I2. I PITCH RATIO. measured va/tieof 4 Fig. 12--Derivation ofFactor K,calculated va/tieof 4
348 BURRJLL & ROBSON VIRTUAL MASS AND MOMENT OF INERTIA OF PROPELLERS
05 06 07 08 -
09
10 IIBLADE AREA RATIO.
P
BLJRRILL & RoBsoN: VIRT1JAL MASS AND MOMENT OF INERTIA OF PROPELLERS 349.
09
06
05
Fig. 14Curves
of
K1 and KA to Base B.A .R. (4-bladed Propellers)Fig. 15Curves of K1 and KA to Base B.A .R. (5-bladed Propellers)
09
N
k
k 06 050 055 0 6005
070 075BLADE AREA RATIO.
K,-GBLAES.
K-6 BLADES.
055 0-60 065 070 075 080
BLADE AREA RATIO. 08
-o
07 08 2 < 07(a) Axial Vibration (b) Torsional Vibration