fl1CIDENCE 'FECTS ON A STATIC HCVCPAFT by
A.J. 1exaxerX
This work was carried out by Dr Alexander while at Ship Division
as a consultant. The model and rig used were built by Ship Division
for research under contract to the Ministry of Aviation.
SU1ARY
Experiments have been carried out on a static, circular, annular jet Hovercraft model in order to measure the changes in lift and/or power requirements associated with a change of incidence relative to the
surface beneath the craft. The fan power needed to maintain a given
lift at constant hover height increases with incidence, the power
incremant being roughly proportional to incidence,.
The overall lift force measured on a strain gauge balance was much
less than the lift obtained by integrating lower surface static pressures but this fference was accounted for mainly by the dilrerence in rate.
of moment flux between the inlet to the model ar
the jet exit.
The remainder was due to jet induced suctiais (negative lift) acting on tire model lower surface outboard of the jet.The balance measured drag force was approximately 1C greater than
the component of the balance lift, Lift sin a. This difference was due
(i) to peripheral changes in jet thrust with the model at incidence producing a small horizontal (effective) drag component and (ii) to peripheral changes in tlre jet induced suctions on the lower surface outboard tire jet contributing to the drag.
Pitching moments were calculated from the lower-surface static
pressures and show that the model is stable for t/h = -1/3 and unsthble
for /h = 1/6 in the range of incidence 00 - 6°
1. Introduction
The development of Hovercraft has proceeded with great rapidity
since the basic ideas were first conceived some ten years ago and today large and sophisticated Hovercraft are available commercially. Fundamental research in this new field, however, has not kept pace with development and it is unfortunately true tlBt comparatively little
is known about the aerodynamics and. hydrodynamics of these vehicles. Hovereraft research was first undertaken in Ship Division, NPL in 1961 and a series ' tests was carried out with a circular, annular
jet model, ref.1, to provide much-needed information on the mechanism of wave-resistance. The presmit static rig tests were designed to
provide further information on the behaviour or hovercraft when they
are not trimmed parallel to the surface underneath.
Modifications were made to the lower surface of the model, see
section 2, to enable static pressures to be measured and the lower surface contribution to lift and pitching-moment to be calculated Overall lift and drag were measured separately on a strain-gauge
balance.
In ref..1 it is shown that with a Hovercraft it is not possible to escape wave drag penalties by trimming the craft level and. that extra
lift power may be necessary to maintain hoverheight at forward speed.
over water, at least up to hump speed.
In this condition extra jet
thrust is necessary to maintain equilibrium in the cushion. An estimate of the extra power required can be obtained from static tests at
incidence over a ground board and this was the main object nf the investigation. In addition, the overall lift and drag were
divided into their various components and some assessment of static
stability obtained using the pitching moments calculated from the lower surface pressure distribution.
From the limited amount of information available it appears tlm.t
single annular jet machines are stable at low hoverheights for a small
range nf incidence beyond which the downgoing jet splits and a cross-flow
is set up which rcsults in instability.
The cross-flow can be preventedand the instability cured by either a twin-jet system as in S.R.N,I, or
by longitudinal and transverse jets which compartment the cushion.
More recently the tendency has been to use flexible stability curtains
or to dispense with compartmentation altogether for the very low
effective hoverheights obtained with fledble sidewalls However, since
these systems are the result of development work rather than research and do not necessarily represent the optimum it is possible that a better
understanding of the mechanism of instability would produce a more
efficient system.
2 Apparatus
The basic model is a circular, single annular jet model or 3
ft.
diameter and is adequately described in ref. 1. As originally designed
it was only possible to measure the integrated pressure on the bottom surface and this was not adequate for the proposed tests over a fixed
ground. It was desirable to be able to measure static pressures at all points on the bottom of the model andfor this purpose a false bottom was designed which consisted or a circular sheet of 0.080 in. dural, 35j- in. diameter, having a row of 19 static pressure holes along one
diameter. Since it was difficult to move either the model or the false bottom and it was desirable to obtain a complete lower surface pressure distribution with the model at various angles of incidence, a false
floor was also constructed which iabled the effective incidence of the model to be changed by tilting the floor and the complete pressure distribution obtained by rotating the floor relative to the fixed line of pressure holes.
This compromise solution enabled the work to be carried. out in the
short time available but bad two main disadvantages: (i) no pressure plotting holes were available on the false floor and. although the
velocities and pressures could be measured on the model it would be
difficult to describe flow patterns from data obtained on one boundary only and (ii) since the pressure plotting holes were fixed relative to the model any serious variations in slot width or jet thrust could affect the results although from the available evidence ref,2 and
section 3 it seems that this effect is likely to be
all,
The traversing pitot-static, ref,1, was used to measure total head
and static pressure in the jet,
All pressures were measured on a36 tube tilting M.T.M. using alcohol as the gauging liqui&
3, Experimental results
The experimental work was divided into three
parts:-(i) Measurement of lower surface static pressures for the following cases: h
ij in.
a = o0i0s20302f0]PC ,1,1- in.water and = 135°,180°,225° h 3 in, a = 0°,2°,4°,6°
and. 270°.
(2) Measurement of overall lift and drag on the balance with the model
tilted for the following cases:
h = 1 in. a = 00,10,20,30,20,50 _)
h = 3 in. a = 00,10,20,30,2f0,50,603 Pc0
*11 inwater.
() Measurement of jet total head. and. static pressure for the following cases:
h = ij in.
a = 0°, p 4-5° pc0 = in,water.h 3 in. a = 0°, = 0,45,135,180,225,315°, pe0 = -,1,1- in.water, h = 3 in. a = 30,60, p = 0,180°, pc0 = in. water,
Values of lift and pitching-moment were calculated from the lower
surface pressure distributions for the cases in (i) above. These results are plotted against incidence in fig. Ia with the lift given directly
and the pitching moments converted to C,P, shifts,
Maximum lift was obtained at zero incidence in these tests with a steady fall-off in lift as incidence increased. The fact that the incidence range is negative is purely one of convention and has no particular significance. Maximum lift reduction for h 1
in. at
.-4° was about 1O3 and far h = 3 in, a. -6°
was about I.
The upper graph, fig.lb, shows the C.P. shift for the cases
h = 1- in. and h
= 3 in, Readings taken at three blowing pressuresfor each case fall reasonably close to the mean lines. The model is
stable at h = 1- in. thraighout the incidence range and is unstable at
h 3 in, The apparent C.P. shifts at a. = 0° are thought to be due to errors in setting the grcund plate, which was slightly dish-shaped,
but are not likely to affect the overall results.
In fig, 2 the stability of the NPL circular model is compared with
experimental data from ref 4 The only other circular model results
available had vertical jets and large clearances where the configuration is unstable but tests on an elongated shape with semi-circular ends, pitching about its major axis and. having the jet inclined inwards at 45°,
show moderate agreement.
Following the procedure adopted in ref .2, the valueof pc0 was
maintained constant as incidence was increased, The horse-power required
for each condition was calculated from average values of total head. and
static pressure in the duct. Actual model values are plotted in fig3a and show veiy little change so that constant pc0 corresponds
approximately to constant fan power. This is in broad agreement with
ref.1, which predicts a lift reduction as incidence is increased while
fan-power and hover height remain constant
Simple momentum and energy considerations lead to the relation
I . 2
Curtain area (PC)
i + x
H.P.
--
x x1100 where ii is an efficiency.
If for the incidence case we take
pressure lift
=---
we have plan area Pc H.P. I1+x
K7ressure lift'\2
1100 r(Curtain area) plan area
5
Values cf K are plotted in fig.3b using the results from figs,la and 3a and are compared with the theoretical values. At a.
O
the efficiency is about 60% but this includes losses in the model. There is some scatter in the points and the best straight line has been drawn through
them, The sl.oDe af tI lines has been plotted against Wt in
fig,3c.
For perfonnance calculations a suggested formula is
H.P. I
i+xf
ÔKI f'2ressure lift\2 11.00 x2 L aa
--
(Curtain Area)plan area
where is obtained from fig.3c. ôa
One result obtained in the NPL Tank tests with the model at 0' trim,
h = 1- in, velocity 5.5 ft/sec PC 675 in (3.5mq ft) gave the
mean angle of the water surface as 40 and the corresponding power increment
as 0.017 H.P.
An estimate from the formula alone gives the power increment as. 0,0302 H.P. Allowing for the 60% efficiency this gives 0.018 H,P, which
is surprisingly good,
Fig..4 shows the overall lift as measured by the strain gauge balance.
There is some scatter in the results and mean curves have been drawn, The tendency for lift to fall as incidence is increased is clearly shown and follows the trend obtained with the lower surface pressure lifts, fig,..la. The absolute values of lift, however, are much lower than those of fig.la. The flagged symbols were obtained at +3° and agree reascnably well with
those at 3° except for the case h = 1- in, Pco = 1 in Possible
scatter appears to be about ±1 lb. The lift as measured by the balance is made up of four parts:
LBalance =Llower surface +
0cos45°-1L
' lower surfaceDuct) (
(between jets) (jet) 'outboard af jets)
Duct rate of momentum flux was calculated using average pressure measurements obtained from the duct pitot rake for two positions at right angles, Jet momentum flux was calculated from pitot and static traverses of the jet at
six positions 3 = O 45°, 135°, 180°, 225°, 315°. It was not possible to traverse at = 90° and 270° because the static tubes were brought out at these points. The negative lift due to suctions on the lower surface outboard of the jets caused by jet entrainment was measured approximately with a small static tube placed on the surface, Maximum down load was
1 lb for pc0 1. in.
--In fig,5 - cos 450 (change of vertical jet thrust (p + p12)A from inlet to outlet through model) is plotted against undersurface pressure lift for a. = 00 and h = 1- in. and 3 in, Also shown is the negative lift LR due to suctions on lower surface outboard
of jet.
In fig.6 the difference between lower surface pressure lift and
overall,liftis plotted against undersurface pressure lift and. cpared
with (M1 - M cos 45° + LR) from fig.5. Forh = 1- in. the lift is
completely accounted far, within the experimental error, but for h = 3 in.. there is a discrepancy of I- 2 lbs.
The reason for this error is not clear but it should be noted that in fig.4. the balance readings at a. = 00 are not always consistent with readings at largerincidence, This mayaccount for the discrepancy. In general it would appear that the difference between overall and lower surface lift has been accounted far satisfactorily.
The balance measurements at incidence also give the drag and. this
is plotted in fig.7
as lift x sina against measured drag. Although there is a fair amount of scatter most of the points lie near the meanexperimental line which shcws a drag approximately 1( higher than
lift sin a..
Part of this difference is accounted, for by the change in jet thrust due to differential change in height as the model tilts, Traverses of the jet at 1 = 0° and 180° (Bow and stern) were made forh = 3 in, a. 3°, P0 = - in, I in, 1- in. and the additional drag
force due to the jet out-of-balance is' plotted against total drag in fig.8., This accounts for 40% of the difference between lift sin a and measured drag.
The forces due to induced sucticris on the lower surface outboard of the jet also have a horizontal component and, due to peripheral variations in the jet when the model is tilted, produce an additional drag force, This was measured approximately, as for the lift, with a surface static tube and. produces a drag equal to the jet out-of-balance
drag.
Thus af a difference of 0.32 lbs between lift sin a. and drag at
a. = 6°, Pco
l- in,
0.26 lbs is accounted for, The remaining 0.06 lbsrepresents only of the total drag.
Measured total drag = measured total lift sin a. + Jet out-of-balance drag + drag due to jet induced surface suctions,
4-. Conclusions
Very little data are available on the effect of incidence on
Hovercra.ft and appear to be concerned mainly vith overall effects. The
purpose of these tests was to investigate the problem in a little more detail, as it relates to a circular, single anruiaa' jet model. The conclusions drawn are thought to be valid fairly generally although greater accuracy of some of the results is desirable. A two-dimensional rig is better suited to this fundamental work and would provide mare accurate results and a better comparison with theory, bearing in mind
7
The results can be divided into two parts:
Power requirements
As incidence is increased a Hovercraft will either lose height if fan power is kept constant or alternatively will need extra power to maintain height, this being a direct result of a change in the pressure
equilibrium of the cushion. This theoretical result by Hogben, ref.1,
is confirmed and the non-dimensional power required is shown in fig.3b. The extra power required increases both with incidence and height.
From an analysis of the results a correction is suggested which enables the effect of incidence to be takn into account in the standard formula for calculating lift power. Comparison with one result obtained in tank tests with this model shows good agreement.
Breakdown of measured overall lift and drag into components
In this particular model the air is supplied from an external source which is separated from the model by a small air gap. The thrust of the air entering the model produces, effectively, a negative lift making the measured overall lift much less than the lower surface cushion plus
annular jet lift.
There is also a small negative lift produced by the entrainment effect of the annular jet on the downward facing surfaces outboard of the jet.Measured lift = lower surface pressure lift + direct lift from
annular jet - inlet thrust fr an external source - jet induced negative lift from surfaces
outside annular jet.
Agreement is very good at h = 1 in, with discrepancies of up to
io at h = 3 in. probably due to inaccuracies in the strain gauge balance
readings.
With the model at incidence the model experiences a drag force due
to the tilting of the lift vector (L sin a.).
Measured values of the dragwere some 1C greater than L sin a. and were mostly accounted far by slight changes in jet thrust as one side of the model neared the ground plus a corresponding change in the jet indiced suctions outboard of the jet.
Measured drag Measured. lift x sin a. + small peripheral change
in jet strength + associated peripheral change in jet induced suctions outboard of jet.
5. Acknowledgments
The work described in this report was carried out in the period
20th July - 14th August 1964 while the aithor was a Vacation Consultant in Ship Division. Acknowledgment is made to the DSIR for the opportunity
to undertake this research and to the Ministry of Aviation for permission to use the model and rig, In addition the author wishes to express his appreciation for all the help received in particular to Dr. N. Hogben
and. Mr. J.T. Everest for many helpful discussions and to Mr. EJ. Neville
and Mr. H.R. Sayer for their assistance with the experiment work and
6. References
Author Title
HOBEN, N. n Investigation of Hovercraft Wavemaking. Ship Hydro. Lab. NPL report.
HOGBEN, N, Static rig pressure and airflow measurements on a circular Hovercraft model.
Ship Divn Tech.Mem. No.21. Nov. 1962.
EVERT, JT. Unpublished, note on pitching moment measurements.
Princeton Symposium on Grcund Effect Phenomena
LIST OF SYMBOLS
h height of model centre line above ground board
p0 cushion pressure
reference cushion pressure at centre of model
t jet thickness
t
x (1+cos e)
K
(Curtain :::; (re33.l1ft)2
LR jet induced negative lift on lower surfaces outboard of jet
rate of momentum flux entering model
rate of momentum flux leaving model
jet efficiency
e jet angle, irsvard from vertical
p air density
60
I I0
1-50 III +h=i
c.=li
x hi pf
i3' Ico*l
® h3'
c.'1
h=1-
co2
(I,h=31
VARATON OF LOWER SURFACE PRESSURE LIFT
AND CORRESPONDING CP SHIFT WITH INCIDENCE
FIG. L 30
20-N.B. FALS( ZERO 10 I IC
I
a-0
8 6 4 2*
NWAR FROM VER11CALSYMBOL' REF hID
e *
MODEL SHAPE+
O04?
450 CIRCULAR)0 083 45'
CIRCULAR j1 N.P.L X 4,P94 O°IS0'
CIRCULARA
0.30
00 CIRCULAR4,p43 05O
4501 1IIJ
o
o 10
JH
EQUIVALN DIAM LUSED FOR hID-4
-2 0 2 4 6
O'2.-0
0.003 Lu U' 2 a-0.00015 000010 O0005
h-3
hi
o,0oI-0 -2 3 o 4COMPARISON OF THEORETICAL AND MEASURED
POWER REQUIREMENTS
FiG 3C
3 5 6
VARIATION OF WITH
-40
w U2
60 50 20 10 qi +hm1
h 3 1 'Ch=1
h= 3°
h I0
h 3_
12 i2r
2 2FIG4
FLAGCD VALUES OBTAII'JED
4'
ATo(+3°
6 2 10 -,
J
w Li.-j
4 U) w 02
0O0'
I. hDUE -to SUCO
ON L0NER UFT SURFACE OUTBOARD OF
JT
-
1D 45°+ L) 0 10 20 30 40 50UNDERSURFACE PRESSURE UPT (tb)
60
FIG6
60 50
20 30 40
3.5
30
2 S 2.0 1.0 0 FLAGGED VALUES OBTAIN.O AT O(=+ FIG7 FIG. 8. 05 1.5 2.0 25 30 3.S MEASURED DRAG (Ib')VARIATION OF DRAG WITH LiFT
20
S 30 3.5DRAG (Ib)