Daniel Guggenheim Airship lnsti tute
Akron , Vhio
TABLE vF C i.JliTE1TS
Pago
Introduction
1 Object
1 Test Appo.ratue
11) Gust Apparatus
2 b) Airship .• odel lllld Recordinc Apparatus
4 Teot Procedure
4 Re•ulte
5 Dieouseion
6 Table
7 Table II
0 Table III
9 C onolusions
Figure•
Daniel Guggenheim Airship Institute
Akron , Ohio
MEASUREMF.JlT OF BE@ING MOMENTS IN THE HULL OF AN AIRSHIP
MODEL DURING ITS PASSAGE THROUGH AN ARTIFICIAL GUST
Item 5), Contract Nos - 47286
Introduction
Following is the final r eport on measurements of the bending moments at
specified c r oss sections in the hull of an airship model dur ing its passage
through an artificial gus t . These tests were carried out dur ing February
and !larch , 1938, at the Daniel Guggenheim Airship Institute, Akron, Ohio.
The object of these tests was etated in the contract as follows: "Utilizing
the model (Item 2) modified to permit flexibility in the model "'t sections (a)
appr oximately at center of buoyancy and (b) at appr oximate l y .15 the ai r ship's
length behind the center of buoyancy , and utilizing t he special gust chamber
equipment to be provided by the contractor , taste shal l be made to determine
bending moments occurring in the hull of an airship dur ing passage through
gusts . The co.nnection of the three sections of the model shall be through
instruments 11hich will r ecord local bending moments in the model . The test
series shall include the ·same pitch angles and rudder settine;s as fo r Item 3)
giving a t ota l of 32 teet seriee. The report covering this item shall include
all pertiment data ."
Test Appe.ratus
a) Gust Apparatus
The gust apparatus at t he Daniel Guggenheim Airship Institute has been described
in the r eport submitted on Item 4a) of this contract. Most of the
tests here reported were made using the "sha rp11 gust . In this case the width
of the transition zone bet ween the ext r eme edge of the gust (zero vertical
velocity) and maximum gust speed is a pproximately 50 cm. The extreme ohord
of the 1/75 scale Mark II fins from the tip of the leading edge to the trailing
edge of the elevators is 49 om. The transition zone of the "sharp" gus t
is therefore approximately equa l to the fin ohord length . Several check runs
were made ueing the 11diffuee11 gust , in which the transition zone is now aesumed
to be approximately one meter.
Fol lowing the test runs , R. survey was me.de of the gust velocity in the model
path in order to c ~1 eck a previou:: survey. Results of t hese later measur ements
e.re given in Figi:. 1 and 2, which show the variation in gust velocity across
the jet for the nf!ha r p11 and "diffuse" gusts respectivel y . The fi rst or lowest
gust speed (approximatel y 4 . 3 io/s ) was used in t hese teste.
-2-
JJanrer-uuggen m E~r•!Up Ine1'n>lf1'"'•~---Akron,
Ohio
b) Airship llodel and Reoording Apparatus.
The general oonetruotion of the airship model used has been desoribed in the
report on Item 2) of this oontraot. Briefly, it is a l/75 acale model of the
U. S. Airfihip Akron 1 made in three sections. The forward section is composed
of a thin copper shell closed by a pressure bulkhead. This shell forms the
hull of the airship from the nose to the point a/.R • . 40. (,/ • tota l length
of model and a ... dietance of a given cross section from the nose.) The center
section includes a oonneoticn for attarJhing it rigidly to the whirling arm,
the recording apparatus and the surface of the hull from a/.R • . 40 to a/# = .60.
The rear section, like the front, ie a copper shell cloeed by a preiniure
bulkhead. This seotion finiehes out the model hull .
For the tests here reported, the front and rear sections were hinged to the
rigid center section about horizontal transverse axes . Longitudinal wires
.03011 in diameter under a certain initial tension connected the top and bottom
of the bulkheads of these eeotione with the rigid center section (see Fig . 3) .
With this arrangement any foroe applied to either the front or rear seotion
of t he model giv~rise to a bending moment about the hinge axis which is
opposed by the elastio forces set up in the longitudinal wires. Sinoe the
deflection in the wires is proportional to the bending moment, the moment was
detennined by recording the elonr~ ation of the wires.
The recording apparatus coneieted essentially of two f!mall diamond points
scratching on a polished glass oylinder. This cylinder, l" in diameter by 1 ~ 11
long, was located in the rigid center section with its axis on the axis of
the mode l. A synchronous clook motor turned the cylinder at a constant speed
of l rev. every 3 min . The two diamond points were carried on light aluminum
tubes extendine; diagonally across the rigid center section of the model. One
tube was connected to the front section at the top of the bulkhead close to
the longitudinal wire connection, while the seoond was connected to the r ear
section at the bottom of the bulkhead. Fl•t spring flexure joints were uaed
at these connections. Each tube wae constrained to move in a direction
parallel to the axis of the model, not parallel to its own axis . This was
accomplished by two pairs of levers connecting each tube to the rigid center
section through flexure joints. Each pair of' levers: formed the eidee of an
ieoceles triangle, with the vertex attached to the tube . The base, formed
by the rigid center section , lay in a transverse plane . The vertex of such
a triangular frame could only move in a direction perpendicular to its base,
or parallel to the longitudinal axis of the model. The aluminum tube wae~
therefore.,__ compelled to move in this same manner, the extent of motion being
determined by the deflection of the bulkhead to which it was connected . The
diamond points accordingly moved along an element of the glass cylinder and
ae the cylinder turned recorded the deflections of tho longitudinal wires.
Theee scratches were magnified 250 times linear scale and photographed . The
bending moments were computed from the curves so obtained.
The construction of the model and the recording apparatus is shown in Fig. 3 ,
a close up photograph of the oenter section of the model with the cover removed.
The oonnection by which the model is attached to the whirling arm has
also been removed to ehow in detail the aluminum tubes which carry the diamond
points . The vertical wires attached near the oenter of these tubes were
found neces sary to prevent transverse vibrations from building up in the tubes.
-------lJElnre-i- mJgg· nmn: :i--r- ·n:r n- -,;r ,;:ui::;e•----Akl"
on, Ohio
-3-
The synohronous clock motor and the glass recording cylinder are not shown
in this photog ra ph. The oylinder was located with its axis horizontal direotly
behind the brackets carr ying the diamond points.
Fig. 3 also ohows olearly the nearer two of the four ball bearing joints oonneoting
the front and rear shells with the center section. It will be observed
that the beariDg housings are attached to the end eeotions by means of 3"
aluminum channels which lie horizontally across the bulkheads. •
The apparatus ao described thus far did not prove entirel y satisfactory in
trial runs due to vibrations appearing in the records. Certain of these vibrations
appea red to have a frequency proportiona l to the speed of the whirling
arm. Theee were aecribed to the pinion and large bevel gear used in the
whirling arm drive . To reduce the amplitude of these disturbing frequencies,
a dynamic vibration absorber wae devised, shown in Fig . 4. This oonsit!ted of
a s ection of 2" channel approximately 2 ft. long hung parallel to the whirling
arm and about 411 below it. The inner ond was carried by a shaft turning in
ball bearings while the outer end was suspended from a pair of stiff' s pr ings.
The weight was clamped to-the channel and was so chosen that the suspended
system had a natural frequency of vibration equal to that of the gear frequency
at the teet speed. The natural frequency of the dynamic absorber could be
changed by shiftin~ the weight along the arm. This dynamic absorber improved
the records to a considerabl e extent, but another t roublesome frequency still
was present . Oil dash-pot dampers were installed but had no appreciable effect.
Finally the conclusion was reached that these vibrations were due to weakness
in the mounting of the ba ll-bearing houeings to which the f ront and rear sections
were hinged . To confirm this tests were made in which heavy re inforc ing
angle~ vrere clamped from the rigid center eection aorose the ball-bearing
housing to the aluminum channels on the bulkheads. This made the ball-bearing
hinge joints quite rigid both with respect to rotation and transverse deflection.
Trial runs showed that this bracing increased the d1"turbing frequency
and msterially reduced its amplitude. Surprisingly , however , the deflections
in the gust passage• were very nea rly equal to those obtained with the ballbearing
hingee f ree . Thie indicated that either the aluminum oroes channel
was twieting under the bending moment or, more probably , that the bulkhead
it.eelf was bending . Since strengthening these parts of the model would have
required considerable time , the behavior of the model with tho reinforcing
bare clamped ac ross the hinge joint• was further investigated . By applying
forcee at different points along both front and rear shells , it wa~ found
that the defleotions of the diamond points wer e propor tional to the bending
moments about the planes of the bu l kheads . The final tests were ~ therefore,..
run with the reinforcing bars still clamped ao r oss tho hinge joints . Check
calibrations fol lowing the teat runs were in good agreement with the initia l
calibrations . The center of moments for the front section under these conditions
was located at a/..i 2 .363 . For the rear section the center of moments
was at a/,£ ~ .616.
Figure 5 is a photograph from the front of the mode l at the least angle of
the gust with respect to the model . This indicates how near the inner fin
would be to the wake of the whirling 1 ~rm shaft and the dynamio vibration
absorber if the mode l were set at -32 pitch and was traveling at 15 ni/s.
Figure 6 is a photograph taken from below the model showing the point of
attachment to the whirling arm.
-4-
Daniel Guggenheim Air•hip ln1titute
Akron, Ohio
Test Prooedure
The body foroe tests had all been run at 12 m/s model velocity ueing both tho
sharp and the diffuse gueto. .As stated in the object, the bending moment
tests wer~ to be run at the same pitch angles and rudder settinge as the body
force tests. To obtain the same relative angles in the fully developed guet,
the horiiontal speed should necessarily be the same. However, einoe it was
found possible to obtain fairly satisfactory bending moment recorde at 15 u(e
it -..a• decided to run teste at both 12 and 15 m/s uaine, only the sharp gust.
Several check run.e: were made wi th the diffuee gust to determine whether the
change in the •h<>rpness of the gust affected the bending moments aoutely.
Teets were accordingly run at -3~ 0 , o0
, -+-3~ 0 and T 1° model gitch with and
without fins, the tests with fine including o0 , ·~o 0 and -10 elevator settings.
The sharp gust was ueed and model veloci:!;ies were 12 and 15 m/e .
To determine whether the whirling arm and the dynamio vibration absorber interfered
appreoiably with the flow over tho inner fin •everal cheok run• were
made with it removed . The tests were all made with no vertioal fins installed.
The procedure in making test runs was as follows: the model was firet set
to the desired pitching angle and the elevators were set to the required angle .
The whirling arm and the guet wore then started simultaneously. The guot
thue had an interval to build up while the model was brought up to the •teady
test velocity. When steady test conditions had been attained the ewitoh controlling
the synohronous olook motor driving the glass oylinder was turned on.
The model was allowed to make from 8 to 12 guet paeeages during each run .
Shortly before and direotly after the model had pe.esed through the guet the
olook motor switch was opened· for a short interval. This procedure gave two
dietinot marks in both reoord• by which they could be synohronized. Following
the run the glass cylinder wa~ examined under the mioroeoope to make sure
that the diamond points had been correctly adjusted to give a fine distinct
eoratch. The model was then eet to a different pitoh angle or elevator setting
and the pr ooess repeated.
Results
Results are gi -von in terms of
rear sections, C!llf and Cmr •
bending moment ooeffioienta for the front and
Theee are defined by the following equation•:
Cmr • --~a~·f_ _
5'/2 v2.Vol
and
"1f and Mr = bflnding moments exerted by the front and rear sections
respectively, the moment centers being at a/1 .... 383
and a/_g = . 616.
g = density of air:
v = resultant velocity of airship model relative to the
fully developed gust.
Vol = total volume of the airship model.
These coefficients are plotted as functions of the position of the nose of
the mode l with respect t> the extreme edge of the vertical gust. The position
y,---S
-6-
Daniel Guge;enhe im Ai r1h1p ln1ti tut e
Akron, Ohio
of the noee i• given by the dimensionleH ratio e/L where o • dietanoe of the
nose from the entrance edge of the guet and ,t • model length.
A aoale ie also plotted on the ordinatet of the curvee showing the value of
the coefficients when the following definitions are uoed:
Cmr • Mr and r2 v8 Vol"}
, where i ie tho total length of the
mode l .
During the !'light of tba model through the undilturbed part of the ohannel,
the bending momenta will aeaume constant values depending on the pitching
anr;le and the elevator setting. Aa the model entere the gust, the momente
ohani;e. In evaluating the reoords the oon1tant value in the undisturbed
ohannel wae u1ed as a base line from wHoh the deflections in the guet were
:reaeured. Aooordincly, the coeff'ioiente are labelled Ll Cmf and LICmr
The following curves are given to ehow the reeulta obtained with the aharp
guat:
l"~b· 7a , sh~wing ~~mr tor o0 and
7b, '9"7°and
8a, o0 and
Sb, T7° and
9&, Lj~mr ,, 00 and
9b, T7o and
lOa, o0 and
lOb, T7° and
model
"
pi~oh, 12 ~s m~el velo~ity.
• 12
15 It.
I 15 II
1 12 "
12 u
• 15 tt
15 II
liig. lla i;ives curves of <'.!Cmf !'or o0 model pitch with the diffuse guet, the
model velocitiee being 12 and 15 r.i/s. Fig. llb r;ives similar ourvee for
llGmr. The curves of LlCmr, Figuroa 9a to lOb inclusive and aleo Figure llb
givo !"esul te both for the bare hull and !'or the hull with !'ins, elevator
settings beinr, Tl0°, o0 and -100. Figures oa and lOa show in addition the
reeults obtained with the inner !'in removed.
Disouesion
The .tlCmr ~urvee in general show tho samo characterietics, riling ta a maximum
value ns the section comes in-co the gust completely, then falling any 1lii:;htly
and rising again to a aeoond maximum just before the aeotion starts to leave
the gust. The variation in <ICmr aoroee the fully developed aeotion of the
gust is undoubtedly due to tho variation in the gu1t velooity, as the local
tranoveree !'oroe on the hull is proportional to the gu@t velooity. In Fig. 12
tho c:oan ourvo of LlCmr for o0 model pi toh and 12 ,,Ya test speed io shown by
the solid line. The points i;ivon tfere derived by takin,; the value of LlCmr
at the oenter of the guet and, asouming different positions of the model in 111 e
e;ust , multiplying it by the ratio of the avere.i;e i;ust velocity over the front
aeotion to the gust velocit;t at the oentor of the jet. The point• thus determined
are eeen to lie very close to the ex;orlmental ourve.
-6-
Daniel Guggenheim Aire hip Inoti tute
Akron, Ohio
A oomparieon of the value1 of .1 Cmr here measured has been made with ooeffioiente
computed from wind tunnel mea1urementa of the presaure distribution
over the hull of a l/40 aoale model of the U. S. a irohip Akron r;iven in
llACA Report 11"43, by ~·reeman. The bending moment at any point along the axil
of the model is equal to the awnmation from the noee 'kl the moment oenter of
the local tranever1e force times it1 dietanoe from the chosen center of moments.
plus the awmna ti on over the same eurfaoe of the local longitudinal
forces timee their nonnal distances from the moment oenter.
Theoretioally the numerical Talue of 6 Cmr should be the eame for all the
model pi tohing angleo, ae the ohange in pitching angle is oonatant and the
transverse foroes are oc. to the pitch angle. The averae.e of the valuea of
ncmf taken from the ourvee for all four pitch angles ie .195 for the 12 m/•
te•t epeed and .170 for the 15 m/e test speed. Theae values oorreepond to
pitohing angle• of 19°17' and 15030' reopeotively, using the average guo t
velocity of 4.3 m/s. Computatione from wind tunnel pressure measurementa at
20° and 15° model pitoh give re•pectively Cm values of .215 and ,161 .
Adjusting theae computed valueo to oorreeponli with l 90 l7 1 and 15030', the
oomparil!!on i1:~iven in Table I.
TABLt I
ritoh angle Cmr. pressure di1tribution Cmf• whirling arm 0/o Diff.
19°17' .207 .196 -so Lo
15°30' .166 .170 •2.4°/o
In li'ib• 13 a oomparison is given between the experirtental curves obtaiiied for
o0 model pitch at 12 and 15 m/• test speeds and a theoretical ourve computed
from the equation for the transverse foroe distribution along an airship hull .
t • 'SV Vn cos 2r ~ • where
dx
r e local tranaver1e foroe at the oroee eeotion x.
S = density of air.
V = Yelooity component along model axis.
Vn e n n perpendicular to model axis.
-r • angle between model axis and tangent to hull at cross eeotion x.
£ = rate of change of oroas-aeotional area at x.
dx
If the moment center is taken L units from the nose and x ie used to designate
the distance from a given croes section to the moment center. the bending
moment is then given by ;-.L
2
dS
Bending i.oment • J f' V Vn ooa r _ x dx ,
K=" dx
ainoe (~ Vn ooo"r- ~)x ii the contribution of the local traneverse foroe to
dx
the total bending moment.
v-7
Daniel Guggenheim Airship Institute
Akron, Ohio
- 7-
The simplified i;uat nlooity dirtribution ahovm at the top of Fig . 13 was used
in makin<, theu oomputation1. Tho nose of the model wno auumed at different
posi ti one and the local traneveree foroee were computed tor varioue etation1
alonl the model. From these valuee the local oontributione to the moment were
fOl.llld and the latter inteirated over the front section gave a value for the
bending moment for the particular location of the model, The theoretical curvH
so obtained give inaximum d Cmf valuee of , 218 and , 182 for 12 and 15 ,./1 reapectivel:,"
These nre 12 and 7 o/o hii;hor than the nerai;e llvmf vdues from the
experimental curves. Howe..,er, thi.e ."1ethod does not take into account the effect
of the longitudinal components on the bending moment which oppose the moment due
to the transverse foroee. The general oha racterietio• of tho theoretical curves
aro quite Eimilar to those obtained experimentally,
The ourvee for the rear moment coefficient are not quite at oontistent aa those
for the front. However, the general characteristic• are well defined, differing
Nrkedly from those exhibited by the front moment coefficient, The moat
important oharaoteriatic aeeme to be the more gradual building up to the maximum
value . Thia will be commented upon more fully in a later paragraph.
ltoment ooeffioiente for the rear eeotion without tail !urfa.oee can be oomputed
from pr cesure d11tribution measuremente in the wind tunnel in the same rr.anner
aa for the front section . A oompa.riaon of the experimental valuee with thore
coipputed from wind tunnel meaaurementa ia given in Table II.
Change in
Pitch Angle
19°47'
TA!IU. II
6. CDl.r, oomputed from
Pre1eure di1t.
-.041
-.027
Ll Cm ,
guat ieata
-.030
-.032
The experimental valuee r;iven are the average of the maximum values for the
different model pitch angle• , Al thou(,h the 0/ o difference ie qui to high , con•
iderine; the low numerical value of tho coefficient it ie felt that the check
is sufficiently ood,
ln Fii; . 14 the experimental curves for the rear seotion without fins for
12 ,./e model speed are compared with a theoretical curve oomputed in the eame
manner as the theoretical ourve given for Cmr • In this caEe the agreement ia
not at all good. Thie is to be expected, due to the oo-called wake forces
developed nt the tail . However, the theoretical curve i• helpful for furthe r
analysis.
Cmr values for tha rear eeotion with fin! were al10 computed from wind tunnel
preaeure di•tribution meumrement1. The experimenta.-1 data available: for o0
elevator setting covered only the range from 3° to 20° 111odel pitch while the
maximum pitch angle developed in theee test• wee 26cl7'. It ie known that u
the model a11umea hirh pi tohint anglee the pressure force! over the rear 1ection
of the hull increase more rapidly than the fin forces. Accordingly, in
order to estimate from wind tunnel teste the pa.rt of the moment cooffioient
due to the hull foroes in steady flit;ht at the hi@,heat pi tchinr; anglea a ourve
wae plotted ehowine; the moment coefficient due to the hull foroea for 3°, 6°,
-8-
llaniel uuggenheim Ai rehip Ineti tut e
Akron, Ohio
9°, 12°, 15° and 20° pitching angle. Thie our ve was then extrapolated to give
values of Cmr for the entire range of pitch angles oove r od in these teete .
The contribution of the fin forces to the rear moment coefficient waa computed
from normal force ooefficients and center of preesure data. for the ll.B.rk II fins
given reEpeotivel y in Fig. 28 and Table VIII of NACA Report #604. The no,.,,.l
force ooeffioiente are there given for the fin area exclusive of the elevator
e:o the forces on the elevators wera necleoted. Table III givee the values obtained
experimentally for comparison with tha computed ooeffioients. Thi~ tab l e also
ehowo separately the contributions of the hull and fine to the computed coefficients.
TAELC: III
Pitoh An~le A Cm
r
Initial Final Computed on basis of wind tunnel Gust 0/o Diff .
preeeure distribution meaeuremente Teet
Hull Fine Total
00 19°17' . 034 .129 .163 .168 ' 3
3°30' 22°47' . 049 .135 .184 . 208 •13
70 26°17' , 067 .142 .209 . 230 •10
00 15°30' . 020 .101 .121 ....130 '7 .4
3°30 ' 19° . 034 .108 .142 .180 •27
70 22°30• . 049 . 112 .161 ,190 Tl8
l'he agreement is quite good in all but two instances, the g~et test values being
higher in all ca a ea. The elevator forcee which were neglected in the computation
would account for part of this discrepancy ea that the check may be
ooneidered aatiEfaotory.
Returning now to the general shape of the ~ Cmr curves, Fig. 15a ehows .6Cmr
for o0 pitch, 12 n/s model •peed with fine at o0 elevator •etting. Compared
with it ie a curve showing how the contribution of the fin forces oan be assumed
to vary. ThiE shape of the latter curve is computed from the .6Cn curve1 obtained
in the fin force teat• under Item 4), the final value being that given
in Table III. A travel of approximately 1, 75 m. ( . 56 hull lengths or 3.5 fin
chord lengths) from the moment the leading edge of the fin enters the gu•t
transl ti on zon·e ig required to build up these force a . The differenc t between
the experimental LI Cmr curve and the curve Ehowing the contribution of the fin
forces is plotted in F'ig . 15b and reoreeente the development of the body fo r ces.
The development of ll Cmr is seen to be quite similar in the main to the deyelopment
of the fin forces. It 'ff'llY be noted , however, that the initial pElrt of
the curve iE due solely to hull forces 1 so that part of the slow rear moment
development is apparently due to the delayed building up of the frictional wake.
In view of the limited test accuracy it ie hardly po•oible to analyze the
development of the hull forcee in any detail. as relatively small error e in
either the A Cmr or the fin force curve could give riee to eerioue error in the
hull moment curve , whioh ie obtained by ::"ubtraction of two large values.
In this connection it may be pointed out that although the maximum vs l uas of
6 Cmr agree very well with those computed from wind tunnel pre1eure distribution
measurements, the ehape of the DCmr curves could hardl y be pred-ic ted f r om the
-9-
wind tunnel reaul ta. If, for example, the forces on the fine be aseumed to
reach one half the final value in one ohord length, increaeing exponentially
thereafter to approximately 96°/o of the maximum val ue in five ohord lengths
in general accord with the Wagner law, the curve shown by the broken line in
Fig. 15 results . Also, if the moment developed by the hull is assumed to
develop in accordance with steady wind tunnel preaeure distribution measurements
applied to tho seotion in the gust, the broken ourve shown in Fig . l6b results .
Combining these two quaeisteady "theoretical" curves gives the curve in Fig . 15a
labelled "theoretical". The difference between thi,e curve and the e.xperimenta~
one is both apparent and important. The slower development of the exper imenta l
~Cmr curve shows that the rear bending moment loads are not applied ae suddenl y
as assumption of quaeisteady foroee f rom wind tunnel teat11 would indi oate .
Results of the two teat• made with the inner fin r emoved are given in Figur es 9a
and lOa ~~th the other &Cmr ourves for o0 model pitoh, These tests are of questionable
value, In both cases the hull forcee contribute at leut 200/0 of the
total rear moment coefficient . Removing the inner fin would certainly ohange
the nature of the flow over the tail •ection and might have a greater effect than
the interference of the arm. In view of the ol oee ohecks between the eXperimental
and the computed coefficients no attempt was made to use these run• for fu rther
interpretation.
The tests with the diffuse gust are in all respeots esoentially simil ar to those
with the sharp gust , The resulting ourves build up and fall awo.y somewhat more
gradually than do the ourvee obtained with the eharp guet. This is, of course,
entirely to be expected .
The results are believed a.oourate within 5°/o for the values of £1Cmt and 100/o
for the va luea of .:1Cmr
Conclusions
From the resul te: obtained in theae tests it may be concluded that the aerodynamic
bending moment cauoed by the fore body of an air•hip hull can be computed with
satisfactory acouraoy from steady flight wind tunnel pressure di11tribution
measurements . However, these teats indicate that the bending momente in the
rear section of the hull develop mor e slowl y than wind tunnel tests in steady
conditions ·,.ould prediot, Accordingly, bending moment• oomputed on the basis
of wind tunnel tests would seem to furnish a conservative basie for design.
y~/ r.l
1.
2.
3.
5 .
6.
7a.
7b.
Sa.
Sb.
9n.
9b.
lOa.
lOb.
lla.
llb.
12.
13.
1~.
"'~L~VU.IV ... l•.l ""' ll•Jr'
Akron, Ohio
FIGURES
Velocity dietribution acroao sharp gu1t.
Volooity diatr ibution aorose diffuse guat.
Photograph •hawing re ordiri,, apparatua .
Photograph or model on whirlinc an:i 1howini; dynamic vibration
n~,uo rber and braoinc at end of whirlini; arm.
l'hotoi;raph taken from front of model at leaet resultant guat angle.
Photocrap~ from beneath model •how1n£ point of attachment, to whirlint
arm. ·
Curvee or l\Cmr for 0 and •St model pitch, 12 nV• model speed.
'7 -3~ 12 nV•
0 ... 3n • 15 nV•
•7 -~~ 15 nV•
" LlC"'r " 0 T3~ 12 nV•
•7 -3~ • 12 nV•
0 ..-s~ • 15 nils
•7 -3~ . 15 nV•
• urvea of t>Cmr ror o 0 pitoh, diffu•o guet , at 12 m/e apd 15 nl•
model speed. ·
Curve. of LICmr for oo pitch, dit'fuae cust , at 12 m/s and 15 nil•
tlodel speed.
Coz::parieon or L>Cmr ror o 0 pitch, 12 m/a model epeed with point•
computed from gust velooi ty dietribution.
Cooparilon of L>Cmr curves with with curves computed from theoretioal
treneverse roroe distribution.
C o:nparison or 4C111r curves w1 th curves computed from theoretical trans•
verse force distribution.
15a, Comparison of LICmr curve with curve ohowing contribution of fin forces
and with theoretical curve baeed on wind tunnel measurements.
16b. Comparhon of hull moment with theoretical curve ooeed on wind tunnel
J:Mt&.l!uremente.
v-//
t
r ~-----:-·
I
- I
+
,-.-r - .. -,
a .
b .
c.
d .
• i,
Figure 3 . Photograph showing recording apparatus .
Fronr section of model.
Rear section of -model.
Ball bearing hinge joints .
Cross channels supporting hir:ee joints.
2 , 3 4 . Longi tudi~al wires supporting
front and rear sections .
f 2 • Diagonal aluminum tubes fastened to
f r ont and rear sections respectivel y .
g2 . Brackets carr ying diamond points . The
motion of tte diamond points is indicated
by the a r rov1.
h , h . Triangil la r lever assemb l ies supporting
aluminum tubes .
i , i • Fr amewo r k fo r suppor ting sheet metal
cover of center section .
j . Light ve r tical wi r es to prevent transverse
vibrations in a lurninum tubes .
hotoo••pb or odel on wti1r11n arm •bowing 4ynalll1c
•ib1'8tio" ab1orber and braotn at and or wh!rllnr; &l'll.
Figure 5 . Photograph taken from front of model at least resultant gust angle .
Hgure 6. l'hotograph fro benffth mod.el •howinr; pol..nt of attacr.ment
to whirli aria.
i'
.". •....
C>
~ c
..!,>. .,.".,..
IP'
• .I. .. C· ., :.r ..
..... If c.: .....
"' H ;
.". 1t
.".
' I
I ,,,.. t
f .
t I •
f :
i
1·
j
.J'
·.·l .• I,~
LLJ'
v- If
+
1
I
I t-
I
I
.1 -~rt+·
{.)."··=' !''
TC- ~ t ~- ~
'
t
j,
I r
1-l j
t
f
r
I
AlBSHIP lNSTITUTll
.A.DON, Omo
i' t.. VRY.;:i$'. ~ ... t1 !i"M,r I' •
•il'~.;.lt ,¥ Mc/2 P1r<rH
I~ t1r'J31 ¥rlv-_.: t n z:>
~.
I .
: t H-: ·. t·. L -~ l ~ - -
t
I
~,;j' . {•
.__ ....,...~ T- ...... ....,...
1
t.
t
~·I
J
I
I -r + -
...,..._....,.k~
i' h I 1.
"Ci, f I
~
J ,,
'
1
1 l
' "'1 .•,i • .. ~ 1 '
1 ...,
...
r i·.
I '
.,
I I
I .t -+b .. ~ + .,
I r
,.--,,
t
Lor
( u
L- '---"-~-
l
.l
'
' l
'
I LI :...
'-
T
<!) ".. /.
>;.
I
~ii
'-1\0'
r"' l
l't
ltT I l
- .....
/•O
""'H.u!.L GUGGl!1NHEIM
Al38RIP INBTITU'rE
, AKRON! 0.i:HO
,/,;::
~I" L "., ::i
l"ll- .':!,
I
.,.. (..,. ,- J.-
r'
"1 ~ m f°'R Ll.C!H,- ~
6 "l'"li ».::LL fl.z,_ ii ~ L.Z
"'N J I"'/' ~ I). L -;- iE.e'l>
I l. <'Al' {i .,r
1
I
!
1(;'}
l
!"'a.1 !(]1.1.rr
#_'\Sl.H·
- T
J 1 I ' I
_J___ _ =-"'- L _1~
1
r-- --,- T
[' t '
~
·I t
~ t "I ...,I '
J ,,,...[},,;.,1:.r
t:"l'i-t.f-
1.:: il'/<if
[--.1r;_
AIRSHIP STITUTll
AKRON, OHIO
L
.;j t - j
+- ·~
I J
I T + t' r -" r
-~ ;l_t...;'1-'fl! f"'&
ri:'fi,,_j.j:,' ~
4€.L S./'l!i"'.1.
-4VJ'.f'
_,1-1~\o _
-~
-~~
//"' z_f
I
~-
~~~~~~~~
t -fl•lt,
f _ ~·2o
I
I
t-'
.l
I - ' "
t •
!
+
I
4 r ~
FIJV.S I WIT/f
L ~.
I t
f
"I
f
l
j
j
I