.
Auhurn Universi ty Libraries
IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII
3 1 706 025 84997 2
File D S2.332 / 23 McCOOK FIELD REPORT, SERIA L No . 2093
AIR SERVICE TION CIRCULAR
PUBLISHED BY THE CHIEF OF AIR SERVICE, WASHINGTON, D. C.
Vol. V May I, 1923 No. 443
THE STRENGTH OF WING RIBS,
( AIRPLANE SECTION REPORT)
Prepared by A. S. Niles
Engineering Division, Air Service
McCook Field, Dayton, Ohio
November 27, 1922 Ralph Brown Oraug!rnr1
LIBRAHY
WASHINGTON
GOVERNMENT PRINTING OFFICE
1923
MAY 1 0 2013
Non•Oepoitory
Auburn University
CERTIFICATE: By direction of the Secretary of War the matter contained herein
is published aB administrative information and is required for the proper transaction
of the public business.
(u)
..\
J
J
J
THE STRENGTH OF WING RIBS.
This report summarizes the results of a st.udy of all the I be salvaged for a separate test of that portion, whi~h will
available American data on the strength of wing ribs in indicate the strength of- the rib in high incidence.
order to develop a standard of c'omparison for new designs. The manner of loa.d distribution has an important effect
Curves that are based on the strength-weight ratio are dee on t.he strength-weight ratio, but not as great as the locaveloped
for this purpose. tion of the center of pressure. If the load is divided be-
Nearly all of the avaifable data has con!:'ist€d of rests tween the upper an(i lower chords, the strength should be
made at the Forest Products Laboratory and at McCook greater than if all is applied to the lower chord. C-0ncenField.
The Forest Products data is more important than tration of the load near the spars cuts down the bending
the McCook Field, owing to the difference of procedure. moments applied and increases the strength-weight ratio.
When a-rib is submitted to the Forest Products Laboratory Six loading curves were used in the tests recorded in
it i3 their cmtom to test not only the rib received bu-t .to Table3 1 and 2 and Figw:es land 2. ' Three of these curves
try alternative de~igns until they believe that the best represent low or medium incidence, The oldest of these
p03sible rib for the given conditions has been developed. is shown in Figure 3-a and is a medium.incidence curve
At McCook Field it is possible only to test the ribs sub- with the center of pressure at about 40 per cent of chord.
mitted, with occasional patching. In short, the Forest It is not a very severe loading, owing to the uniformity of
Product3 Laborat-0ry is able. to try to answer the question load distribution, and gives unduly large strength-weight
of what is the best rib for the given conditions. McCook ratios. The loading shown in Figure 3-b is a low incidence
Field is limited to answering the question of how good the curve with the center of pressure at about 47 per cent of
rib submitted is and in what direction attempts to better the chord, This also gives unduly high strength-weight
it should b~ made. ratios, owing to the negative load on the nose and the
Figures 1 and 2 show the more important test results po3ition of the center of pressure. Figure 3-c is the standwith
the chords plotted as abscissas and str/lngth-weight ard loading curve for medium incidence and is the
ratios in pounds per ounces as ordinates. Figure l shows most severe of the three. The apex of the triangle of load
the result.s of tests under medium and low incidence load- is 25 per cent of the chord to the rear of the leading edge.
ing3,. and Figure 2 under high incidence. Forest Products This loading gives a center ot pressure at 41.6 per cent of
results are indicated by an O and McCook Field by an X. the chord to the rear of the leading edge. In developing
Test3 made at other places are .indkated by an ®· Many the standard curve for comparison of new designs, test.'!
of the result3 are uot plotted, as they ehow poor strength- made with this loading were given the greatest weight, as
weight ratios. it is the one now being used , In Figure 1 the letter a, b,
The .. mo.it important facton which determine the or c near a plotted point indicates the loading from Figun,
strength-weight ratio are condition of loading, · shape of 3 that was used in making the given test.
aerofoil section, location of span, length of chord, average Three loading curves have been ueed in high incidence
load per inch of chord, and type of construction. and are shown in Figure 4. Figure 4--a shows a curve u.~ed
The effect of each of these factor3 will be discussed in on some o.f the early tests. The load is almost uniformly
turn. di~trihuted and gives unduly large strength-weight ratios.
CONDITION OF LOADING. The center of pressure is too farback,and ·the curve give.~
almost a medium incidence loading. The curve shown in
At large angles of attack the center of pre3sure is far Figure 4-b has been very extensively used. The center
forw.trd and the raquired load fact-0r i3 high. As the angle of pre3Sure at one-third the chord to the rear of the leading
of attack decreases, the center of pressure move3 to the edge i3 too far back, however, and the loading is unduly
rear, and the required load factor decrea9e3. The m03t severe for high incidence. The curve in Figure 4-c is the
severe condition of loading occun with th!\ center of pres- present standard, and results obtained by its use were
sure about midway between the wing spaq, and i9 called given the most weight in developing the standard curve
the ·'medium incidence " condition of loading in this of strength-weight ratios in high incidence. The center
report. This loading c:i,use3 a large bending moment on of pressure varies from about 27 t.o 30 per cent of the chord
the portion of the rib between the spars, and limit.s the to the rear of the leading edge. This loading is severe
strength oi nearly the whole rib except the nose. The only on the nose section of the rib, owing to the concentrano3e
and sometime3 the web member3 adjacent to the tion ofload near the front spar, and gives very highstrengthfront.
spar are limited by the high incidence loading with weightnatios. This is proper, since the test is primarily
the center of pre3Sure far forward. With a well-designed to determine the strength of the n\>se section, and the
rib the high incidence test will give higher values of strength of the remainder of the rib is determined by the
etrength-weight ratio than the medium incidence test. medium incidence test; The points in Figure 2 are
This is as it should be, as the required load factor is higher. referenced by letters a, b, and c in the same manuer as
Where only one test of a rib is to be marie, it should be those in Figure 1, except that the letters refer to loodings
that for medium incidence. Often the nose portion can shown in Figure 4 instead of Figure 3.
45593-23 (1)
In determining the load factor that should be carried by
a rib, the following procedure should he used: For hlgh
incidence add 0.5 factor to the requirement for wings in
high incidence. Thus, for Type I airplanes the ribs
should be designed for a load factor in.high incidence of
8.5+0.5=9.0. For medium incidence add 0.5 factor to
the mean of the factors required for the high and low inci- '
dence conditions on the wing. Thus, for Type I airplanes
the ribs should be designed for a load factor in medium
incidence of 0.5 (8.5+5.5)+0.5=7 .5.
SHAPE OF AEROFOIL SECTION.
Experience has· shown that better values of strengthweight
ratio can be obtained with thin sections than with
thick ones. This is undoubtedly due to the fact that with
the deep sections the amount of material in the web is disproportionately-
increased. The chord members might be
expected to show: a corresponding decrease, but they do
not do so for two reasons. The unsupported length of a
section of chord member of a thick wing rib is usually
much greater than for a thin one. This induces larger
bending nioments, which neutralize the decrease in direct
stress. The second reason is that even the shallow rib is
seldom limi~d in strength by the chord members, which
are of minimum allowable size. This same size must also
be used for the deep section, so there is no saving. In
plywood web ribs, wh.ere the section is made deep vertical
stiffeners become necessary, though they can be
omitted in the design of shallow ribs.
LOCATION OF SPARS.
The location of the points of support, i. e., the spars, has
a very important effect on the strength-weight ratios of a
rib. If the spars are close together the medium incidence
value should be high, as most of the load comes on a short
span between the spars. On the other hand, the closeness
of the spars is usually attained, in part, by increasing the
length of the forward overhang, and lowers the possible
strength-weight ratios in high incidence. The very high
strength-weight ratios in high incidence obtained with the
present s.tandard loading (fig. 4-c) are largely due to the
large proportion of the load coming <;>n the rib very close to
the spar, on ribs with spars fairly far apart.
CHORD.
As the chord increases, the str~ngth-weight ratio decreases.
This can be seen to be inevitable by the foJlowing
reasoning: Assume two geometrically similar ribs, A
and B, such that the chord of B equals twice the chord of
A. Assume also that the total loads 0n the two ribs are
the same and that the load curves are geometrically similar.
Due to this load corresponding members of A and B
will carry equal axial loads. In tlie case of a tension member,
the sectional areas of the corresponding :in.embers of
the two ribs will be equal, but as the length of the member
in rib B will be twice that of the member in rib A, its
yolume and therefore its weight will be twice as great. In
the case of a compression member the· weight will be more
than doubled, as the length of the member will be doubled,
and the area will also have . to be increased owing to the
greater unsupported column length. Members stressed
in bending will also be more than doubled in weight, as
both the length and the bending moments will be doubled.
From the above discussion it will be seen that the weight
of the rib required to carry a given load will increase
fll8ter than the chord, but probably not as fast as the
square of the chord, ·
In mathematical_ terms W oc c'
Where Wis the weight of rib to carry a given load.
C is the chord.
xis a value between l and 2, depending on
the type of construction.
Similarly the strength-weight ratio L/W oc 1/c'
While this study gives an indication as to the probable
shape of a curve of strength-weight ratio plotted against
chord, it can not be used !!,lone 118 a criterion of rib strength.
The precise value of x is not known and undoubtedly
varies with different types of construction. Then, ,too,
the sizes of rib members are so small that many of them are
determined not by the unit stress but by practical limitations
as to minimum sizes. This is why the curve falls off
comparatively .slowly in the region of short chords, and
approaches the theoretical curve only in the region of very
large ribs. A third mctor is that as the size of the rib
changes, the most economical type of construction changes.
AVERAGE LOAD PER INCH OF CHORD.
In the design of a rib the required average load per inch
of chord is known, arid it is desired to obtain the lightest
design that will carry this load. If the loading is light, it
is probable that a rib will be developed which will be very
light and ~11 carry the load required, but will not have a
very good ~trength-weight ratio. By adding material at
the proper points, the strength-weight ratio will be increll8ed,
but so will the weight. As the original rib carried
the required load, this increase in weight would be undesirable
and useless, as the spars would fail long before the
ribs. The desirable procedure would be to decrease the
size of the original rib so that the required strength could
be obtained with the higher .strength-weight ratio. This
is· not always pOBBible, however, as the members concerned
may already be 118 light as it is considered advisable to
make a rib member, or the strength may decrease so fast
with decrease in weight that any cutting down of the sizes
ofmembers "'ill not be worth while. In such a case the
best rib for the airplane will b'e the one originally developed
in spite of its low strength-weight rati<;>.
TYPE OF CONSTRUCTION.
In the practical design of a rib the most important
factor influencing the strength-weight ratio is the type of
construction used. The other variabl~hord, loading,
aerofoil, and spar location-are usually determined by
other considerations, leaving the type of construction to
he used as the main question for decision at the time the
rib is designe!l. No one ty.pe of construction is best for
all cases, so several types will be discussed and their range
of economical use noted.
PLYWOOD WEB TYPE.
One of the most popular types of rib consists of a plywood
web with spruce cap strips. This type of rib has been
used for so many airplanes that it might almost be called
the standard type. As usually designed, the lightening
h9les a.re cut in the web a.nd vertical stiffeners a.re omitted.
This type of rib, -when properly designed, is unexcelled
for thin aerofoils like the R. A. F . 15, where the chord is
less than a.bout 80 inches. These ribs a.re used on the
DH-4, VE-7, SE-5, MB-3, a.nd numerous other a.irpla.nes.
. The following features should be incorporated in their
design.
The web should be ma.de of three-ply material rather
than of five ply or of single ply with vertical stiffeners.
The grain of the face plies should be perpendicular and
of the core parallel to the chord.
The lightening holes should be small in regions of large
shear, as near the spa.rs. a.nd should increase in size. a.s the
shear decreases. The rib shown in Figure 6 is a.n example
of excellent design in this respect.
The cap strips ma.y be either of the one-piece type
grooved to engage the web, or ma.y be of the split type,
separate strips of spruce being glued a.nd nailed to ea.ch
side of the plywood web.
The web should be ma.de of low density plywood, such
as spruce, mahogany, or poplar. Spanish cedar gives
good test results when new, but is ma.de brittle by the glue,
a.nd should not be used. Usually a. web of l-irn::h plywood
will be found sa.tisfactory for chords up to 66 inches.
When the wing section is deep, as with the U. S. A. 27
aerofoil, the design described above is no longer economical.
It is then a.d visa.hie to a.dd vertical stiffeners, a.s
shown in Figure 7. The web should be of three ply in
this ca.se instead of single ply. It is often possible with
this type of rib to use very thin plywood a.nd omit the
lightening holes.
REINFORCED PLYWOOD WEB TRUSS TYPE.
For designs in which the chord is too great or the wing
section too thick for the plywood web type to be economical,
the reinforced plywood web truss type ha.s proven
verysa.tisfactory. This type is.shown in Figures 8, 10, 11,
a.nd · 12. In some early tests of ribs of the plywood web
type, .the lightening holes were ma.de triangular instead of
circular or oblong in order to make a truBB of t.he web
material. This did not prove satisfactory at first, but by
gluing small spruce reinforcing strips to the web, a very
good rib wa.s developed for thick sections and large chords.
Tests have shown that it makes little or no difference in
the strength-weight ratio, whether the rib is ma.de with
cap strips and web reinforcing members all on one side of
the plywood web or divided between the two sides. The
grain of the face plies of the web should be para.Ile! to the
chord.
The best type oftruBB for this construction is the Warren.
with the diagonal members adjacent to the spars normally
in compression. If these members are in tension it is
more·difficult to attach the rib firmly to the spar and the
extra. weight in the joint is likely to be more than that in
the diagonal due to its being a compreBBion rather than a
tension member. Where the depth of the rib is small, the
simpie Warren truBB, a.s shown in Figures 10 and 11, is
satisfactory, but where it is large, the truss must be subdivided
a.s shown in Figures 8 and 12.
In the portion to · the rear of the rear spar the trussing
ma.y be continued, Figure. 12, the plywood web type of
construction with lightening holes, Figure 8, or a combi-
3
nation of the two, Figure 10 , ma.y be used. A solid ply
wood web is sometimes used in the nose section also.
BUILT-UP TRUSS TYPE.
Some designers prefer a built-up type of truBS rib to the
reinforced plywood truBB web type described above .
Examples of such ribs are shown in Figures· 9, 13, ~4,
and 15.
The truBBes shown in Figure 9 are indeterminate. The ·
rib W 51- 2 showed the best strength.of those illustrated in
Figure 9 and would probably have given even better
values if the diagonals ha.d be.en placed to carry compreBSion
under normal loads. The Martin type shown as
W 51-4 and W 51-7 showed up poorly in the ribs shown
which are designed for a short chord and a thick section,
~ut showed up very well for a. 95-inch chord and the
R . A. F. 15 section. It seems to have a very limited field
of economical use.
Severa.I types of statically determinate ' truBBed ribs
have been tested _md proved sa.tisfactory. Most of the
succeBSful designs have ·used Warren trusses, but some
have employed the Pratt type, The chief difficulty to be
solved in the design of a trussed rib is that of connecting
the members. In the reinforced plywood web truBS type
· this is done by the plywood web which acts a.s a. large
gusset plate at every joint. The C0-2 rib, shown in
Figure 14, and the Barling rib used small plywood guBBets
at ea.ch joint. The C0-2 rib did not show up very well in
tests, due to the weakness of the connection to the spar.
The Barling ribs, however, showed very good strength
properties. In the Martin ribs, shown in Figure 15, the
web members are glued directly to the chord members.
Care must be taken in auch designs to see that sufficient
gluing area is provided. In this ca.se the web members
are double, one piece being on each side of the chord
member. In the Wa.qen truBB ribs developed a.t the
Forest Products Laboratory, shown in Figure 13, the joints
are wrapped. While this is fairly easy with a simple
Warren truss, it would be quite difficult if the truss were
subdivided by vertical members.
METAL RIBS.
The metal ribs teste<l. have been of the subdivided
Warren truss type. As rt is easy to make metallic joints to
carry tension, the diagonals a.dja.cent to the spa.rs should
normally carry . that type of streBB. The rib is greatly
weakened if they have to carry compreBBion. The metal
ribs tested have not show.n very good strength-weight
ratios, but it is not known how much this is due to some
inferiority of meta.I to wood for .rib construction, and how
much is due to inefficient design or the fact that .the ribs
tested have had very deep sections. Typical metal ribs
are shown in Figures 14, 16, and 17 ...
DESIGN OF TRUSS RIBS.
TruBB ribs should be designed originally according to
the method described in Information Circular No. 312
and tested. The tests will show what revisions a.re necessa.
ry. The analytical d·esign of a rib should never be considered
as a substitute for a static t03t, but it is a great a.id
in obtaining a well-balanced design .with a minimum ·Of
trials.
4
DEVELOPMENT OF STANDARD CURVES. Columns 9 and 10 give the distance of the spars from the
leading edge in per cent of chord.
In order to determine what strength-weight ratios had
been attained in the past and what should be attained
in the future, the results of the better and more important
tests were plotted in Figures 1 and 2. From these
points curves were drawn to indicate the values that
it is believed should be attained in new designs. These
curves are plotted in Figures 1 and 2 with the points on
which they are based. They are plotted again in
Figure 5 without those points. Owing to the number
of factors which influence the design, the curve
can not be taken alone as the standard of comparison.
For instance, if the aerofoil section used is thick. the
strength-weight ratio indicated by the curve probably
can not be attained, and this factor must be taken into
account. On the other hand, with favorable conditions
and good design, higher values may be attained. As
more data are obtained. it is expected that these curves
will be revised with reference to the new results, and it
may be pOSBible to draw separate curves for thin and
thick sections, or for· metal and wood construction. At
present, however, the data is too me.ager.
TABULATION OF TEST RESULTS.
The principal results of tests on ribs with good strengthweight
ratios are shown in Figures 1 and 2 and Tables l
and 2. The letters near the plotted points in Figures 1
and 2 refer to the loading diagram used in the test, as
explained ·on page 1.
The following data is given in Tables 1 and 2:
Column 1 gives the reference number of the correspond-ing
plotted point on Figure 1 or 2.
Column 2 gives the chord length in inches.
Column 3 gives the ultimate load on the rib, in pounds.
Column 4 gives the average load per inch of chord, and
is obtained by dividing the ultimate load from column 3
by the chord from column 2. The result is in pounds per
inch run.
Column 5 gives the weiglit of the rib in ounces.
Column 6 gives the strength-weight ratio of the rib in
pounds per ounce. It is obtained by dividing column 3
by column 5.
Column 7 gives the initials of the airplane for which the
rib was designed. " Exp." indicates that the rib was
experimental and designed without reference to any particular
airplane.
Column 8 gives the maximum depth of aerofoil section in
terms of per cent of chord.
As much of the data in columns 8, 9, and 10 had to be
scaled from photographs. it is not very precise, but is
sufficiently accurate for the purposes of this report.
Column 11 gives the type of construction. The meaning
of the letters used to refer to the various type3 is gi..-en
below
Column 12 gives the number under which the report
containing the data is filed by the technical data section
at McCook Field.
TYPES OF CONSTRUCTION.
The letters in column 11 of Tables 1 and 2 refer· to the
following types of construction. The· figurei, referred to
below illustrate typical ribs of the various types.
A-1. Plywood web type with lightening holes. Figure
6 shows ribs ot this type constructed for the MB-3 by the
Boeing Co. , which gave excellent strength-weight ratios.
A-2. Plywoud web type with lightening holes and vertical
stiffeners. Figures 7 and 8 show ribs of this type used
on various airplanes.
A-3. Plywood web type with vertical stiffeners but no
lightening holes.
A-4. Single-ply web with lightening holes and vertical
stiffeners. Rib W 51-6, shown in Figure 9, is of thi1' type.
Ribs of types A-1 to 4 may have either single piece or
split cap strips.
B-1. Built-up Pratt truss lacking certain diagonal
members. Rib W 51-2, shown in Figure 9 is of this type.
B-2. Martin type incomplete truss. Ribs 51-4 and 51-7
in Figure 9 are of this type.
G-1. Simple Wal'l'.en truss wi~h reinforced plywood web.
This type is illustrated in Figures 10 and 11.
G-2 .. Subdivided Warren truss with reinforced plywood
web as shown in Figures 8 and 12.
C-3. Same as C-2 except that all reinforcing strips are
-on one side of the plywood web.
C-4. Built-up Pratt truss with crossed diagonals and
reinforced plywood web, as in Figure 13 ..
D- 1. Buj\t-up Warren truss. This type is illustrated in
Figure 13.
D-2. Built-up subdivided Warren truss as shown in
Figure 14.
D-3. Built-up Pratt truss with single diagonals as in
Figure 15.
E-1. Built-up subdivided Warren truss constructed of
duralumin. Figures 16 and 17 show ribs of this type.
No division of the built-up types is made_ here on the
basis of method of joining the members.
5
TABLE 1.-Data on tests of ribs in medium and row incidence.
1 2 3 4 5 6 7 8 I 9 I 10 11 12
--------- ------
Point. Chord. Load. Average Weight. Ratio. Airplane. Depth. Spar load. location. Type. Report.
---------·------
I
1 57 650 11. 40 8. 5 76. 5 T . S .. .... . .. 11. 1 6.9 73. 0 C-3 D 52.332/22
2 55 326 5. 94 4.8 67. 7 E xp ....... __ 5. 7 18. 2 64.5 A-1 D 52.337/30
3 55 335 6. 10 5. 1 · 65. 3 Exp .. .... . .. 8. 1 10. 5 69.8 A-1 D 52.332/30
4 54 435 8. 05 6. 8 63.8 T . P.L .... 15.1 12.0 61.0 A-2 D 52.332/53
5 54 390 7. 22 6.3 62.0 T . A.3 . . ... . 11.'l 15. 2 69. 0 B-1 D 52.332/26
6, 60 283 4. 72 4.2 67. 2 S. E.5 .. .. . . 5. 7 12. 2 63.3 A-1 D 52.332/32
I
7 6.3 490 7. 78 8. 2 59.5 P. S. 1.. ··· - 13.0 13. 3 62.2 C-2 D 52.332/55
8 65 485 7. 46 9.5 51. 0 P . W.1. . .. . 11. l 13. 0 63.0 A-3 D 52.332/60
9 68 475 6.99 8. 1 68.5 P . S. 1. .. .. . 8.5 13. 5 62.0 C-2 D 52.332/55
10 69 500 7. 25 7. 5 66.6 P . S.1. . . ... 7..f, 14. 8 65.6 C-1 D 52.332/56
11 72 570 7. 92 9.3 61.3 H . A . . . . . ... 5. 0. 13. 1 58.9 A-4 D 52.3.32/18
I 12 75 460
I
6.34 9. 5 48.3 C. 0. 2 ...... 1 11. 1 11. 0 69.0 D-2 D 52.332/47 13 90 1,074 11. 93 15. 2 70. 7 Douglas.·- . . 11. 1 13. 3 67. 7 D-3 D 52.332/51 i
I
14 94 575 6.12 11. 3 50.8 Martm . . .. _. 6.4 10.6 63.9 B-2 D 52.332/17
15 95 775 8.15 15.0 51. 7 Martin . . __ .. ,5. 7 10.6 63.9 C-1 D 52.332/59
I 16 9fi 1, 237 12. 88 18. 2 68. 0 M. 81.. . . . . .. 7. 7 18. 2 61.0 C-3 D 52.332/21
I
17 96 773 8. 05 12. 8 60.4 M. 81-. ...... 7. 7 18. 2 61.0 C-3 D 52.332/21
18 103 750 7. 28 i6. 5 45. 5 C. 0.3 . . . .. . 11. 5 20.0 60.0 E-1 D 52.332/16
19 108 940 8. 70 15. 5 60.6 T . F . ... ... . 6.6 10.0 70. 0 D-1 D 52.332/43
20 108 589 5.45 13.3 44. 2 T . F ... . .. . . 6.6 10. 0 70.0 C-4 D 52.332/43 I 21 160 1,100 6.88 52. 5 21.0 D.B. l.. . .. . 18. 0 8. 2 61. 7 . E-1 D 52.332if
22 162 1,050 6.49 34.0 30.9 Barling _. _ . . 6. 5 10. 5 60.5 D-2 D 00.12 /1308 I 23 180 1,200 6. 66
I
4~. 5 24. 8 EXP-·- -- ·-·- 6. 4 13. 0 65.0 C-2 D 00.12 M/1425
I 24 180 742 4. 12 35. 7 20.8 Exp .. .. ..... 5. 7 14.7 72. 5 D-1 D 11.1/107
25 180 710 3. 94 38. 8 18. 3 Exp ......... 5. 7 14. 7 72.5 D-3 D 52.332/23
TABLE 2.-Data on tests of ribs in high incidence.
!
1 I 2 3 · 4 5 6 7 8 9 I 10 11 12
P oint. I Chord.
--- --- - -
Load. Average Weight. Ratio. Airplane. Depth. Spar Type. Report.
I load. location.
- 1- ---- - - ----------
l 55 363 6. 60 5.1 ~§ Eli:P- · - · - ··-- 8. 1 10. 5 69.8 A-1 D 52.332/30
2 57 800 14. 05 8.2 T . S ... . ... . . 11. 1 6.9 73.0 C-3 D 52.332/22
3 60 482 8. 05 7.9 61 N . 9 . . . . .... . 6.6 10. 4 65.0 A-1 D 52.6 N/l
4 63 830 13.18 8. 1 103 M. B.3 ..... . 5. 7 1~. 9 54. 0 A-1 D 52.332 61
5 66 325 4. 92 5.6 59 D.H. 4 .. . . . 5. 7 10.6 68. 2 A-1 D 52.1/36 DH-4
6 I 72 S76 13. 56 9. 2 106 H . A . .. ..... 5. 0 13.1 58.9 A-4 D 52.332/18
7 90 1,363 15.17 15. 1 90 Douglas._ .. . . 11.1 13.3 67. 7 D-3 D 52.332/51
8 94 672
I
7.15 11. 4 59 Martin .. ... . 6. 4 10.6 63.9 B-2 D 52.332/17
9 ~ 683 7. 11 12. 5 55 M. 81: .... . . 7. 7 18. 2 61.0 C-3 D 52.332/21
10 106 925 8. 73 18.0 51 ¥~i-·.·::::::: 8. 0 15.1 70. 5 C-1 D 00.12 M/1306
11 108 1,612 14. 92 16. 0 101 6.6 10.0 70.0 D-1 D 52.332/43
12 108 1,212 11. 22 15. 7 78 T.F ....... . 6.6 10. 0 70.0 D-1 D 52.332/43
13
I
131 800
I
6. 10 26. 4 30 C. 0.3 . . .. .. 16.3 20.0 60.v E-1 D 52.332/46
14 180 974 5.40
I
36.2 27 EXJ;> ... . ... . . 5. 7 14. 7 72. 5 D-1 D 11.1/107
DISCUSSION OF INDIVIDUAL TESTS. 55-INCH EXPERIMENTAL RIBS.
!<'AVY T. S. RIBS.
These ribs were tested by the Navy in August, l!l21.
The chord is small, 57 inches, but the section is very thick,
so that a Warren truss plywood web with stiffeners was
used. Both web stiffeners and cap strips are on one side
of the plywood, making the section unsymmetrical. The
spars are very far apart. The webs were of birch plywood,
which caused difficulty in obtaining reliable glued
joints, but added greatly in obtaining the high ~trengthweight
ratios obsei;ved. The high strength-weight ratios
obtained in both high and low incidence tests are remarkable
when it is remembered that the ribs are very deep
and the spars far apart. The favorable factors were the
use of birch and the loading curves used. These ribs
show that truss type ribs can be used economically with
thick aerofoils for small chords . . The average load per
inch run on these ribs was very 1igh.
The Forest Products Laboratory teste:;. 12 wing ribs of
55-inch chord in the early part of 1919. Six of the ribs
used the R. A. F. 6 wing section, and six the R. A. F . 15.
Three ribs of each design were tested in high and three
in medium incidence. As the chord is short and the
aerofoil section thin, the plywood web type with lightening
holes was used. Split cap strips were used in both
designs. The strength-weight ratios obtained in medium
incidence are very nearly equal, with a slight advantage
for the R. A. F . 15 ribs, due probably to the smaller distance
between spars. In the high incidence tests the
R. A. F. 6 ribs showed up much better than the R. A. F.
15. The low strength of the latter was due to the excessive
size of. the lightening hole between the leading edge
and the front spar. Failures were about equally divided
between web and cap strips, showing consistent design.
Tests on the R. A. F. 6 ribs are recorded by point 3,
Figure 1, and point 1, Figure 2. Tests on the R. A. F.
15 ribs are recorded by point 2, Figure 1.
T. P. 1 RIBS.
In April, 1922, four-maple plywood ribs for the T. P. 1
were tested at McCook Field under the medium incidence
loading. These ribs are shown in Figure 7. The reinforcing
strips are on both sides of the web and the cap
strips are split. The strength-weight ratio obt~ined with
the short ribs was good and is plotted as point 4, Figure I.
The strength-weight ratios for the long ribs were not very
good and are not plotted in this report. Althrugh the
required strength was attained, with the above ribs, it
was believed that they could be improved and four more
ribs were built and tested, using birch plywood instead of
maple and incorporating minor changes in design. These
new ribs carried heavier loads than the original ones, but
weighed more, and the strength-weight ratios were slightly
smaller .in the case of the 54-inch and slightly greater in
the 78-inch ribs. These tests indicate that for thick sections
the type of construction used is very good for chords
of about 54 inches, but that a trussed type should be used
where the chord is as great as'78 inches.
T. A. 3 RIBS.
6
Three types of wing rib were tested at McCook Fielg in
October, 1921, for use in the T ; A. 3. These ribs were all
tested in medium incidence. The best type was a built-up
Pratt truBB with four bays between the spars, but with no
diagonal members in the two center bays. The result of
the test on this rib is plotted in Figure 1 as point 5. A
photograph of the rib is shown in Figure 9 where it is
called W 51- 2. The other two types, also shown in
Figure 9, are the Martin type of incomplete truss and the
single ply web type with lightening holes and vertical
stiffeners. The two latter types were much poorer than
the first, the last mentioned being parJ~cularly weak.
These tests are a further indication of the possibility of
using trussed type ribs with economy for airplanes with
thick sections and short chords.
S. E. 5 RIBS.
These ribs were developed by the Forest Products
Laboratory in 1918 and 1919 for use in the American
S. E. 5. The original ribs were very weak and flimsy
and a plywood web rib with lightening holes "7'-B developed
in their place. The ribs gave very good strength values
in the medium incidence test. As the chord is only 60
inches and the R. A. F. 15 aerofoil section was used,
these ribs help bear out the theory that the type of construction
used is the most economical for short chords and
shallow sections.
N-9 RIBS .
These ribs developed by the Forest Products Laboratory
in 1919 are of the same type as the S. E . 5 ribs mentioned
above. Five designs were tried and tested in both high
and medium incidence. One design had a single-ply web
with lightening holes and vertical stiffeners. The other
four had plywood webs with lightening holes and split
cap strips. The best results were obtained with the high
incidence loading, due partly at least to the load curves
used.
M. B . a RIBS,
Tests were made in October, 1921, of six ribs constructed
by the Boeing Co. for the M. B. 3. · These ribs were of the
plywood w'lb type with lightening holes and grooved cap
strips. They are shown in Figure 6. The strengthweight
ratios obtained ran as high as 109. The value
plotted as point 4, Figure 2, is the average for the five
ribs tested . ·
D. H . 4 RIBS .
These ribs are the result of a large seri~ of tests by the
Forest Products Laboratory. The rib finally decided on
is of the plywood web type with lightening holes and split
cap strips. It is very .similar to the 55-inch experimental,
S. E. 5, N. 9, and M. B. 3 ribs previously described. The
t ests recorded do not show very good results, however,
when compared to some of the other designs. Nevertheless
they are much better than the earlier ribs and mark
a stage in the d evelopment of rib design.
P. S. I RIBS.
Six ribs for the P . S. 1 were tested at McCook Field in
Jun e, 1922. Owing to the wings tapering in both plan
and elevation, two ribs were submitted for each of three
chord lengths, 63, 67.75, and 69.25 inches. ' ,These ribs
are shown in Figure 8. The 63-inch and 68-inch ribs,
which are of the reinforced plywood web truss type, gave
satisfactory strength values and very good strength-weight
ratios. They are represented by points 7 and 9 of Figure I.
One cause for their falling below the standard curve is
the depth of the section. The 69-inch rib was ot' the
plywood web type with circular lightening holes and vertical
stiffeners. The strength of the rib was very poor
and it had to be redesigned. In the new ribs the same
type of design was used as for the 63 and 68 inch ribs of
the original tes t, except that a simple Warren truss was
used instead of a subdivided one. These ribs are shown
in Figure 11. As a matter of fac t the truss ribs shown in
Figure 8 were constructed originally as simple Warren
trusses and the vertical members were added during the
tests. The test results on the revised 69-inch ribs are
plotted as point 10, Figure 1. These tests indicate that
the reinfor~ed plywood web trul!B type of rib is economical
for short 'chord lengths with both thick and medium
aerofoils. The diagonal members should make an angle
of about 45° with the chord. If a thick aerofoil is used ,
a subdivided Warren tl'UBB should be employed. If the
aerofoil is thin , however, the vertical members may be
omitted. All of the P. S. 1 ribs were tested in medium
incidence.
P . W. I RIBS
Nine ribs of five different designs were tested at McCook
Field in January, 1921, for use in the thick wings for the
P. W. 1. The five types were Martin incomplete truss.
reinforc.ed plywood truss web, and plywood web with
vertical stiffeners and round, oblong, and no lightening
holes. The plywood web designs had divided cap strips.
The best results were obtained with the rib having no
lightening holes. The results for this rib are plotted as
point 8, Figure 1. The strength-weight ratio of this rib
was not particularly good, but it is so much petter ihan
the other types tested that it is recorded in this report.
H. A. SEAPLANE RIBS.
These ribs were tested by the Forest Products Laboratory
in 1921. These ribs were very shallow and had singleply
mahogany webs with lightening holes and vertical
birch stiffeners. The results of most tests indicate that
for ribs of this size a plywood web without stiffeners is
superior to the construction used. Nevertheless excellent
strength-weight ratios were obtained, particularly under
the high incidence loading. Diagonal stiffeners on both
sides of the spars undoubtedly aided greatly in obtaining
the high strength values recorded. The cap strips of these
ribs were grooved.
C. 0. 2 RIBS.
Two wooden and two metal ribs, shown in Figure 14,
designed for the C. 0. 2 were tested at McCook Field in
January. 1921. The two types carried approximately the.
same total loads, but those made of wood were the lighter,
and therefore had better strength-weight ratios. Although
the strength-weight ratio of the wood ribs was not pa1·ticularly
high as originally designed, it is believed that it was
increased by improving the spar connections ,on the ribs
actually used in the construction of the airplane. It is
not likely that the metal ribs could be improved to the
slme extent. These tests indicate that for ribs carrying
light loads wood is a better ·material than meta1. All the
ribs were tested in medium incidence, the results of the
tests on the wood ribs being plotted as point 12, Figure 1.
DAVIS-DOUGLAS RIBS.
Thirteen Davis-Douglas wing ribs were tested by the
Forest Products Laboratory in January, 1922. Six of them
were manufactured by the Davis-Douglas Go. and the
others by the laboratory. The ribs were all of the same
design, a built-up Pratt truss, similar to the Martin ribs
shown in Figure l>. Split cap strips were used and the
diagonal members had small spacer bl0<~ks at their centers.
Excellent strength-weight ratios were obtained in both
high and low incidence tests. The type of rib used is very
economical in weight. but is harder to build than the
plywood web type.
MARTIN RIBS.
A large number of ribs have been tested for the Army
and Nav:r Martin bombers. The early Army bombers used
the R. A. F . 15 aerofoil. The best ribs for this aerofoil
were of the reinforced plywood web truss type tested at
McCook Field. PQint 15, Figure 1, shows the results on
· these ribs. In the later airplanes a thicker aerofoil was
employed, necessitating ~he design of new ribs. The Army
is now using the rib shown in Figure 15, although the
strength-weight ratio is a little lower. This was to be expected,
however, with the deeper section. In June, 1921,
the Forest Products Laboratory published the results of
tests on two types of ribs for the Navy Martins. One of
them was on ribs of the Martin type of incomplete truss
similar to the ribs W-51-2 and W-51-7 in Figure 9. The
other report was on built-up Pratt truss ribs similar to the
Army ribs shown in Figure 15. The weights of the two
types were about the same. In high incidence th.e Martin
type showed about 50 per cent more strength than the
Pratt truss type. In low incidence .the best Martin type
rib was but little better than the best Pratt truss type rib.
606860-31--2
7
In the average results, however, they were not quite ~o
good. .The range of results in either type, however. was
much greater than the difference between types. The
indications of the Forest Products tests are that the Martin
type is the stronger, but not so reliable. The resulte
of the best tests on the Martin type rib are plotted ae point.
14, Figure 1, and point 8, Figure 2.
LOENING M 81 RIBS .
These ribs were of the reinforced plywood web t-ruse
type, and are similar to those in Figure 8. except that. a.ll
reinforcing strips were on one side of the plywood. The
ribs were tested by the Forest Products Laboratory in
September, 1921, in both high and low incidence . . Excellent
results were obtained in low incidence, as shown by
points 16 and 17, Figure 1. The high incidence results._
shown by point 9, Figure 2, are only fair. The ribs would
be greatly improved. if the r..ose section was made stronger.
As these ribs are of the same chord length as t.he Martin
ribs, it is interesting to compare the two types. In hig)1
incidence the Loening ribs were inferior to the Mart.in .
due to the ·weak cqnstruction of the nose section and its
greater length. In low incidence the Loening ribs were
superior to the Martin ribs, as the closeness of the spars
more than compensated for the increased depth. One of
the Loening ribs tested in low incidence had tension diagonals
l1-djacent to the spars instead of compression diagonals.
It failed at about half the load of the other type, illustrating
the advisability of having the diagonals next t.o the spa.rs
in compression in wood-trussed ribs.
C. 0. 3 RIBS.
One 103-inch and one 131-inch meta.I rib for the C. 0. 3
were tested in high incidence and two 103-inch ribs in
medium incidence at McCook Field in 1921. These ribs
were of the Warren ·truss type, and were constructed of
duralumin cha.nnels. as shown in Figure 17. None of the
strength-weight ratios wer,e particularly good. but some
very interesting results wiire obtained. In the 103-inch
rib tested in high incidence the strength was almost
doubled, with no change in weight, by changing from a
subdivided Warre~ truss with the diagonals nearest the
spars in compression to a similar truss r .-ith the diagonals
i,n. tension. In the medium incidence tests the strength
was greatly increased by changing from a simple Warren
truss to a subdivided truss. The test of the subdivided
Warren truss rib in medium incidence is recorded by
point 18, Figure 1. This rib ~hows a strength almost as
good as could be expected fron1 wood. The 131-inch rib
tested in high incidence shows up poorly in comparison
with the standard curve as sllown by point 13, Figure 2.
106-INCH EXPERIMENT.-\L RIBS .
Seven reinforced plywood web truss ribs were t-ested at
McCook Field in June, 1920. These ribs were constructed
to test the value of this type of construction, which had
not yet been thoroughly tried out. All of the ribs were
of the same design, shown in Figure 10, except that the
material and direction of the grain of the plywood was
varied. The ribs were- tested in high incidence, the best
result being plotted as point 10, Figure 2. The strengthweight
ratio obtained was not particularly high, owing to
the loading curve used and the lack of verticals subdividing
the panels of the trllBB. The wing section .used is
slightly too deep for a simple Warren truBB. The tests
indicated that birch plywood webs made a stronger and
stiffer rib than spruce.
T. F . RIBS.
Seventy-five ribs were tested by the Forest Products
Laboratory in developing a 108-inch rib ,or the T. F.
fl)ing boat. The report on these ribs was published in
February, 1921. Three different types, involving in all
15 different designs, were built. and tested . The three
general types were veneer truss, plywood truBB, and Warren
truBB. The veneer trUBB rib was of the Pratt truBB type
with crossed diagonals. The chord members and vertical
posts were made of spruce and the diagonals of birch
veneer. The type of construction was not satisfactory,
being difficult to build, heavy, and weak. The plywood
truBB type is shown in Figure 13. It differs from the type
developed a.t McCook Field and shown in Figures 8, 10,
11, and 12 in the following details: The truss is :rpore nearly
of the Pratt than of the Warren type. The cap strip is
grooved to receive. the web instead of being of the split
type. The best test on one of these ribs is recorded by
point 20, Figure 1.
The best design tested was of the simple Warren truss
type and is shown in Figure 13. This is a built-up rib
with taped joints and showed the very good strength
values recorded by points 19 in Figure 1, and 11 and 12
in Figure 2. This type was recommended for use by the
Forest Products Laboratory. It is believed, however, that
the McCook Field type of reinforced plywood· web truss
rib would give results approximating, if not exceeding,
those obtained with these built-up ribs. In medium incidence
the T. F. ribs were tested under the loading shown
in Figure 3-a, which is not as severe as loading 3- c, which
was used on the McCook Field ribs.
D. B. l RIBS.
Eight subdivided Warren truss ribs of duralumin for the
8
D. B. 1 were tested at McCook Field in July and Septem her,
1922. All members of these ribs, which are shown in
Figure 16, were channels with the legs bent over to make
a square tube with an open seam. The ribs were 102,
103, and 160 inches in length. The best strength--values
were obtained from the 160-inch ribs, and are shown by
point 21, Figure 1. After the first six ribs-had been tested,
two new ribs were constructed of 160-inch length. Several
improvements were made in these ribs, and the total
strength was increased, but the strength-weight ratio was
not changed. The comparatively poor strength of these
ribs was due, to the use of metal and the very deep wing
section.
BARLING RIBS .
These ribs were of the built-up subdivided Warren truss
type. The joints were formed with plywood gusset plates
and wood screws and glue. Except for the wood screws,
and the UB'e of a T section for the chord members, the ribs
are very much the same as those shown in Figure 14.
The strength-weight ratio was high,. as shown by point 22,
Figure 1, but the loading was that shown in Figure 3a,
which is not very severe.
180-lNCH EXPERIMENTAL RIBS.
Experimental ribs of 180-inch length have been built
and tested by both McCook Field and the Forest Products
Laboratory. Only six ribs were tested at McCook Field,
four using the U. S. A. 27 aerofoil and two the U.S. A. 5.
The U. S. A. 5 ribs showed the best strength-weight ratios,
due to the shallow section, as shown by point 23, Figure 1.
Of the U. S. A. 27.section ribs, those in which the greatest
number of members was in tension showed poorer results
by a small margin. The poorest of the U.S. A. 27 section
ribs, however, gave as high a strength-weight ratio as the
best of the Forest Products ribs, which were built to the
shallow R. A. F . 15 section. The McCook Field tests were
made in October, 1920.
The Forest Produ~ts Laboratory tests were made in 1918
and 1919. Six types, embracing 11 designs of trussed nbs,
were tested. The earlier tests in 1918 showed that the
plywood web type with lightening holes, which shows up
so well with short chords, was not economical for ribs of
this size. In those tests a Pratt truss type similar in construction
to that shown in Figure 15 gave very good strengthweight
ratios. The best of these tests is recorded by point
25, Figure 1. In the 1919 tests several types of trussing
were tried out, but the best results were obtained with a
truBB of the.same type as was found best .for the 108-inch
T . F. ribs. The average results of the tests on this type of
rib are shown by point 24, Figure 1, and point 14, Figure 2.
I ' I
rt
I
APPENDIX.
DISCUSSION OF RESULTS OF RECENT TESTS
ON RIBS CONSTRUCTED BY THE GLENN L.
MARTIN CO.
Since the ,body of this report was written, data has bel)n
obtained on tests made by the Glenn L. Martin Co. on .
wing ribs designed by them for use in ipree designs under
construction for the Navy. These result.I! are summarized
and discuBBed in this appendix.
MARTIN 13S-INCR RIBS.
Five ribs of 138-inch chord were tested. Three of these
were tested in low incidence with the loading shown in
Figure 3b, and two in high incidenc~ with tp.e loading
shown in Figure 4c. The ·ribs were .of the built-up truBS
type and employed the U. S. A. 27 wiIJ.g section. The
truBB used was of the type shown iri. Figure 18. The web
members were cqnstructed of two strips of sprqce separated
by a balsa filler of varyizig thickneBI! and formed approximately'
a strut of uniform strength. The strength-weight
ratios obtained are very good, especially if account is taken
of the depth of the aerofoil section. The best test results
are plotted as point 26, Figure 1, and point 15, Figure 2.
The ribs weighed 27 .5 ounces each, and gave. the following
strength-weight ratios. The ribs tested in low incidence
gave ratios of 36.8, 35, and 33.9 pounds per ounce, and
those tested .in high incidence 41.8 and 39.1. The low
incidence values are all above the standard curve, but the
loading was leBB severe than that for which the curve was
plotted and the spars fairly close togetlier. On the other
hand, the section was very deep. The high incidence
values fall below the curve_, althougn they were tested
with the loading for which the curve was drawn. This
was due, partly at least, to the deep section and the fact
that the front spar was quite far back of the leading edge.
On the whole, the tests indicate that the design was first
class.
MARTIN 78-INCR RIBS.
Six ribs of 78-inch chord were tested, three in low
incidence and three in high incidence. The aerofoil was
the U. S . A. 27 and the type of construction the same as
for the 138-iach ribs. The best test results are plotted as
point 27, Figure 1, and point 16, Figure 2. The strength.
weight ratios obtained were 75.5, 74.1, and 64.5 in low
incidence and 117. 92.7, and90.8inhighincidence. The
rib,s weighed 8.5 oun.ces apiece. The low incidence values
are all above the standard curve. One high incidence
value and the average of the three are above the curve .
These ribs show up, on the whole, even better than tbA
138-inch ribs.
MARTIN 36-INCR RIBI'.
Six ribs of 36-inch chord were tested, three in low
incidence and · three in -high incidence. These ribs · were
stamped out of duralumin sheet. The outer edges of the
rib and the sides of the circular lightening holes were
flanged over to give strength laten\lly. Theaerofoil sec·
· tion was the U. S. A. 27. The strength-weight ratios
obtained were rem:trkably lugh-132, 126, and 121.in low
incidence and 163, 139, and 128 in high incidence. Non"
of these points are plotted in Figures 1 and 2; but it can
easily be seen that they would come far above the standard
curves. In considering these tests it should be remembered
that the ribs weighed 4 .. 38 ounces, which is quite AA
much as a satisfacj;ory wood rib would weigh, and much of
the strength of the rib is useleBB. It might be possible to
obtain a satisfactory· rib using the same design but n
lighter gauge of metal;- ·but thatfa not probable, owing t.O'
the likeliltood of crinkling. Full advantage of the strength
of these ribs could be taken only-by the use of a strong wing
covering;, such as corrugated duralumin,. '\Wilch wou!d
allow the rib spacing to be jncreased. In conclusion it
m~y be said that these ribs are the most efficient ones
referred to in this report, and the Martin Co. deserve great
credit'foi: their development·. At present, however, much
of thei~ strength is wasted, and research should be continued
to see if this type of design can not be developed
to give a rib of light weight and more near\y the desired
strength, even at the.expense of a small l_oss in strengthweight
ratio.
(9)
10
P~o&a T11-:,laJ
0~ ... at' ,,,,,_,.,da £0-il,n.,.
e8 aHiO!C_ 11o1_t l"i_e'J _
.5rREN6'TH W,E/6H7" R'1T/O.S
"4eDltAHAN4 LPl'II' /HC/~,vce
Ga il:J 44 !10 ,JoP /.,. 40
Curv' ;,. 1.-ofc
FIG. J.
"'
..
,.
11
Proc.e Ts3n,d
0 .,..,,.c,f P',o,,wd.!l /.a borot'tn7n
e~r,~kl ·
0 Oil't1r IAan ~ -
Sr.eE-t«V?t _w'.e,t1t1r.Btn1?~
H,ttN /Nc10M&#:
.. (
· ,
... l - P,fr,J{in dn;M.
FIG. 2.
12
F-/ C1. 4c
F!Gs. 3a, 3b, and 3c: F ms. 4a, 4b, and 4c .
. .
-
t
•!
(
~ -
1
14 ~,l?NCIQet? CV&YoE..5
OA
,J.[:eE-/V/3TH-!'IJ;:,fi:ttz:.£:.nno
... Or ,ee,.:,
"' ~ .,,, "" - , ' - riu.5
F IG. 5.
13
14
FIG. 8.
FIG. 9.
15
FIG. JO.
FIG. 11
FIG. 12.
16
,?c/NrO.eCcO PL YvVOOp ,eu,5,;1 RIB
P.e,qrr r.eu::.:5 rypt:- /fl/TH cm,;,,:ic0 0/6(:;0l"((lL,:,
8UIL T UP TYPe- W?J.eee-lY T£U.5.S Rlf5
FIG, 13.
F IG. 14
17 ·
Fie.... 15
.s«(ion AA
18
FIG. 17
@ . CZ '
~ru~_/ ~7'
.3,iz.,. We;,,, gf' We.& M ...n zkr
.5¢cl<on B-B
FIG, 18.
0
