¶ REPORT NO. NADC-77202-30 RE-587 LEVEL7 EVALUJATION OF DYNAMICALLY RIVETED JOINTS by 0 ~Basil P. Leftheris H. Eidirloff*J and R.E. Hooson* 19 9 Research DepartmentA E Grumman Aerospace Corporation Bethpage, New York 11T14# July 19T9 4, Final Report E Contract No. N62269-T7-C-04T8 Approved for Public Release; Distribution Unlimited -U', Prepared for CS*~ Naval Air Development Center Warminster, Pennsylvania 189T4 +Principal Investigator .- ItGrumman Engineering Department I 791121021.
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
¶ REPORT NO. NADC-77202-30RE-587
LEVEL7
EVALUJATION OF DYNAMICALLY RIVETED JOINTS
by
0 ~Basil P. LeftherisH. Eidirloff*J
and
R.E. Hooson* 19 9
Research DepartmentAE Grumman Aerospace CorporationBethpage, New York 11T14#
July 19T9
4,Final Report
E Contract No. N62269-T7-C-04T8
Approved for Public Release; Distribution Unlimited-U',
Prepared for
CS*~ Naval Air Development CenterWarminster, Pennsylvania 189T4
+Principal Investigator
.- ItGrumman Engineering Department
I 791121021.
N rTIC ES
REPORT NUMBERING SYSTEM The numberkig of technical project reports issued by the Navae Air DevelopmentCenter is arranged for specific identification purposes. Each number consists of the Center acrnym, the calendar 1yew in which the number was assigned, the sequence number of the ,eport within the specific calendar year, andthe official 2.digit correspondence code of the Command Office or the Functional Directorate responsible for therepol. For example: Report No. NADC-78015.20 indicates the fifteeth Center report for the year 1978, and prepared "by the Systems Directorate. The numerical codes are as followv,:
CODE OFFICE OR DIRECTORATE
00 Commander, Naval Air Development Center01 Technical Director, Naval Air Development Center02 Comptroller10 Directorate Command Projects r20 Systems Directorate J30 Sensors E Avionics Technology Directorate40 Communication & Navigation Technology Directorate50 Software Computer Directorate SaJ60 Aircraft & Crew Systems Technology Directorate70 Manning Assessment Resources -
80 Engineering Support Group
PR'OUCT ENDORSEMENT - The discussion or instructions concerning commercial products herein do not constitute -'an endorsement by the Government nor do they convey or imply the license or right to use such products.
APPROVED BY: • DATE: z"
11
UNCLASSIFIED 3F3.FRSECURITY CLASSIFICATION OF THIS PAGE Mkx0W O Xf Utm4___________________
(I~ REPORT DOCUMENTATION PAGE R INUTRUCTIONS
0 N2.GOVT ACCESSION NO: S RECIPIENT'S CATALOG NUMMER
*j L R/> i0DC72-3 -
4. TITLE (daid SuUNIfe) S. !y!LfAt"RQBL1~r5"_IQPOVgRgD
J;~VALUATION OF DYNAMICALLY NIBEPT, JOINTS,
RCOTAC 0R GRANT NUMB *S)
Bas~ilinf (i ) N62269-77-C-,6478R. E./Hooson .. RABW~ M IMW
9 . PERFORMING ORGANIZATION NAMM AND ADDRESS10PRGA ZMY0
* ~GRUMMAN AEROSPACE CORP. - '"iAREAsKWI1RESEARCH DEPT. / ,7m.V./ /IR WO KUNT 4ABETHPAGE,_N.Y._11714 WORK__UNIT_________
~~ L II.11 CONTROLLING OFFICE NAME AND ADDRESS______
Naval Air Development Center J1?
Warminster, Pennsylvania 18974 -. ~SE 11. MONITORING AGENCY N AMC ADORESSW(I dEIEm hunb C600101W Ofice0h) IIS. SECURITY CLASS. (of this Adbere)
UNCLASSIFIED
16.~~S. DISTIRMSIPOCASATEMEN (of NesAOIP
r6 ITIUINSAEET(5msReaft Approved for Public Release; Distribution Unlimited
17. DIST RISUTION STATEMENT (oMetU. abolrast oiimd hin Week 20, it E4bth"i *001 Repa)
r 18. SUPPLEMENTARY NOTES
C is. KEy WORDs (Contiuiu on toysiso ~id ft it novesa ve kEiwaiif by Nook .. lbor)
Fastenings Fatigue-Improvement Fasteners Stress Wave RiveterVFatigue Interference Fit Fasteners
Fatigue Life JointsFatigue Tests Festened Joints
Aluminum JointsIAUSTR ACT (Cmtifbusan ,.yero aid It nmeoeaep OWd fdmnitt hr Nook 000b..)
j - The performance of stress wave driven rivet installations is evaluatedJ in terms of crack growth arresting capability, using precracked speci-mens subjected to constant amplitude and spectrum fatigue loading.'C Rivets installed in pre-cracked holes using the Grumman Stress WaveRiveter (SWR) are shown to be effective in arresting crack growth.Residual stresses resulting from ri t coldwork are determined and theeffective stress intensity factor, KI is calculated. It is concluded - V&
DO FOR 1473 EDITION DOF INOV6 ii i OBSOLETE • UC SIFE L} 1 /6'JAN 72 /N 0102-014-6601 1 NLSSFE
SECURITY CLASSIFICATION OF THIS PAGE (Men ai 15 Eniered.)
A.~~~~~~~~~I VI- LSIIAINO HSPQ MMD *164*
that stress wave rivet ins~tallations offer significant potential weight
savig fr stuctres esinedto adamtge ole~nt ritria
jD1,,t
NKX-AM1
UNCLASSIFIE11CRIYCLSIICTONO HI Alii 01 XWO4
K
i; I
Grumman Research Department Report RE-587
EVALUATION OF DYNAMICALLY RIVETED JOINTS
by
Basil P. Leftheris+,
H. Eidinoff
and A
R.E. Hooson
Final Report on Contract No. N62269-77-C-O478
July 1979
Approved by:
Rich~rd S heuingoDirector esearch
+
Principal Investigatcr
Grumman Engineering Department
NADC-77202- 30
FoR EWOR D
This research and engineering program has been conducted by the Fluid
Dynamics Directorate of the Grumman Research Department, Grumman Aerospace Corpor- .
ation, Bethpage, New York, under Contract No. N62269-77-C-0478. Dr. Basil P.
Leftheris was Principal Investigator. The contract was administered by the Air
Vehicle Technology Department, Naval Air Development Center, Warminster, Pennsyl- [jvania, with Paul Kozel providing technical liaison. The report summarizes work
performed during the period October 1, 1977 throuh September 30_. _.,9-.
The test portions of the program were performed by Mr. R. Wigger of the
Elements and Materials Test Section under the direction of Mr. D. Layton. LInspection of the specimens was done by the Grtuman Quality Control Department.
Mr. E.R. Ranalli of the Engineering Structural Analysis section made some
valuable contributions to the program.
tiLI
i
u NADC-17202-30
U ~SLMARY
The purpose of this program was to evaluate the advantages of dynamically in-
stalled rivets in aircraft structures. This was accomplished primarily by testinga: large number of three-rivet specimens made of aluminum alloy 2024-T81 material
and riveted with the Grumman Stress Wave Riveter (SWR). To establish the suitability1'', + of riveting aircraft struActures with the SWR in accordance with recent danmge
tolerance criteria, each hole was precracked using constant amplitude loading. The
I i S.R method of riveting expands the rivet radially at high accelerations that
results in plastic deformation of the surrounding material. The process "cold works"
the material and leaves compressive residual stresses u.ound the rivet to prevent
crack growth wider loading.
Precracked specimens riveted with the SWR were fatigue tested under both spec-
trum and constant amplitude loadings. Spectrum tests were conducted using a reduced
F-14 wing fatigue spectrum which included both positive and negative loading
cycles, The greater part of the spectrum tests were conducted with a net stress at
limit load equal to 35 ksi (')41 MPa). Complete arrest of crack growth was evident
* +in all these tests. Such behavior was theoretically predicted by calculating the
stress intensity factor K.. Explicit equations for the residual stresses permitted
integration of the K1 equation. It was found that for the precrack sizes of 0.05 -
0.025 in. (1.27 - i.016 ms), the KI values were near zero, an indication of no
crack growth.
Other tests were carried out at constant amplitude loadings that supplied crack
growth characteristics of the specimens riveted with the SWR. CompArison of the
results was made with open hole and Hi-lok fastened specimens.
A limited evaluation performed to establish design values for static, fatigue,
and damage tolerent (precracked structure) allowablf stress levels indicated asignificant potential weight saving when allowable stresses for precracked holes
were based on SWR-riveted values.
The results obtained from this program indicate that the SWR can arrest cracksS i consistently, while simultaneously installing a relatively inexpensive rivet. It.
is recommended that consideration be given to the use of the SWR process in
meeting damage tolerance requirements for aircraft structures.
LiL iii
-~~~~1 0091110oiN~l"~. ~ --
NADC-77202- 30
TABLE OF CONTENTS
[1 Section Page
1 Initroduction and Background . . . .. ... .. . .. *. .. . 1
2 Specimen Preparation . . . . .. . ... . .. .. .*.. .. .. 1
3 Programi Overview ... 9. .. ... . .. .. . .. . .. . .. 17
14 Fatigue Testing........ . . . . .*.* .... * . . 20
Specimen Testing.......... ....... 20
StressWaveRvetingeti.g. ................. 20
5 Description ot' K1 Calculation Method . ..... 9 214
Introduction ...... . .. .. .. .. .. ...... . . . 214
Stress Intensity Equations . .... .... 25
K Calcu~lations .... *27II6 Fatigue Crack Propagation .. ............ .... ... .. .. ... 31
Constant Amplitdeietsde .est..... .. . ...... 31
Spectrum Tests . . . . . . . . . . . . . . . . . . .* . 36
Stress Intensity of Joints Riveted with the Streis Wave
Effect of Stress Wave Driven Rivets on Fighter Wing Lower
Cover AllowableStresess.e.. . . .. . . . . .. .. .. 48
7 Conclusions and Recommendations . . . . . . . . . . . . . . .. 52
8 References . . . . . . . . . . . . . . . . . . . . . . . . . 514
Appendix
A Equations of ResidualStresess.s.. . .. . .. . .. .. .. 57
LAv
NADC-77202- 30
F }LIST OF ILLUSTRATIONS
Figure J
1 Stress Wave Riveter Being Used in Production of F-I4 TitaniumWing Skins ... . . . . . . . 3
2 Residual Hoop Strain vs Radial Distance, AR = 0.011 Radial
3 Residual Hoop Stress vs Radial Distance, AL ZOZ4-T3. . . . 5
4 Residual Radial Stress vs Radial Distance, AL 2024-T3 . .... 6
5 Residual Hoop Stress vs Radial Distance, AL 2024-T3 . . .... 7
6 Residual Radial Stress vs Radial Distance, AL 2024-T3 .... . . 8
7 Residual Hoop Strain vs Radial Distance, At, 2024-T3 ... ...... 9
"• •• • 8 Residual Principal Strain Distributions, Measured with the ;
Moir' Fringe Technique (Ref. 4), in a Specimen Riveted withthe SWR . . . . . . . . . . . . . . . . . . . . . . . . . .. . 10
... 9 General Open Hole Specimen .. .. .. .. .. .. .. .. ... 12 )
1 0 Two-Piece Specimen with Simple Unloaded Fastener . . .. ... . 13 •
11 Two-Piece Specimen with Tree Unloaded F~steners .. .. .. 4
12 Single Sheet Specimen with Three Fasteners and Washers . . . . . 15
13 Section View of Precracked Specimen [0.040 In. (1.01 mm) Crack]with No Subsequent Growth . . . . . . . . . . . . . . . . . 21
14 Section View of Precracked Specimen [0.070 In. (1.77 mm) Crack]with Subsequent Growth .... . .. .. .... . . 21
15 Test Setup with Parallel Beams to Prevent Bending DuringCompression Load Cycling. . . . . .. .. .. .. .. . . . .. . 22
16 Telescope Observation of Crack Growth During Testing . . . . . . 22
17 Distribution of Residual Hoop Strain of 2024-T81 AluminumSpecimen SWR Riveted with 3/16 In. (4.83 mm) Diameter A-286Rivet with 0.006 Tz.. (0.152 mm) Padial Displacement . . . . . . 28
vii
NADC-77202-30
LIST OF ILLUSTRATIONS (contd)
Figure Page
18 Distribution of Residual Radial Stress of 2024-T81 AluminumSpecimen SWR RivetedI with 3/16 In. (4.83 mm) Diameter A-286Rivet with 0.006 In. (0.152 mm) Radial Displacement ...... 26
19 Distribution of Residual Hoop Stress of 2024-T81 Aluminum ASpecimen SWR Riveted with 3/16 In. (4.83 mm) Diameter A-286Rivet with 0.0%6 In. (0.152 mm) Radial Displacement ..... • 29 2
20 Stress Intensity Factor K vs S oCrack u Leng 29
I R (Hole Radi!s ) . . . . . 2aICrack Lengthb LI
21 Stress Intensity Factor Hole RadiusK* .... .... 30
I~ (Crack Length•22 Stress Intensity Factor Kvs a (CRack L.. .. . 30
23 Constant Amplitude Crack Propagation (Open Hole) . . . . . ... 34
24 Constant Amplitude Crack Propagation (Stress Wave Rivet) . . .. 34
2 5 Constant Amplitude Crack Propagation (Stress Wave Rivet) . . .. 35 Li26 Stress Intensity ....... e................. ... 35
27 Spectrum Crack Propagation (Open Hole) .... . . . . . . 40
28 Spectrum Fatigue Life u Open Hole Specimens . . . . . . . . . . 40
29 Spectrum Crack Propagation (Stress Wave Driven Rivet) . .... 42
30 Spectrum Crack Propagation N Hi-Loks vs SWR 2024-T851Aluminum . .. .. .. .. .. .. .. .. .. .. .. .. ... 44il
31 Spectrtm Crack Propagation ^- Hi-Loks vs SWR 2024-T81Aluminum . . . . .......... . . . . . . . . . . . . . . . 45
32 Spectrum Fatigue Life N Stress Wave Rivet Specimens2024-T85 Aluminum ....................... . 46
33 Effect of Initial Crack Length on Spectrum Fatigue Life2024-T81 Aluminum.......... . . . . . . . .......... . . 46 i
34 F-l4 Lower Cover Section Simulated in 2024-T851 AluminumAlloy . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 !
viii
I ~NADC-77202- 30
LIST OF TABLES
I Table Page
1 Test Matrix: Spectrum Fatigue Loading . . . . . . . .. . 18
2 Test Matrix: Conatant Amplitude Fatigue Loading ...... . 193 Summary of Constant Amplitude Fatigue Data 3/16 In. Dia
Fastener Holes in 2024-T81 Sheet. ... . . ......... 32
4 F-14 Wing Spectrum - 15 Layers* ................ 37
5 Slummary of Spectrum Fatigue Data for Dynamically Installed3/16 In. Dia A-286 Interference Fit Fasteners in 2024-T81Aluminum . . . . . . . . . . . . . ........... . . .. . 38
6 Fighter Wing Lower Cover Allowable Stresses . . . . . . . . . 50
ixi
•. i
ix• II i
NADC-77202-30
1 . INTRODUCTION AND BACKGROUND
It has been known for some time that riveted joints of aircraft tructures are-
frequently critical points where structural cracking occurs. Various causes were
identified over the years that attributed early failures to both the itructural de-i[J sign (selection of materials und design allowables) and manufacturing methods. •
Many causes of structural failure are hidden, that is, not identifiable by ectab-S lished inspection procedures. For example, the buckling of the rivet inside the
hole can not be seen from the outside, although it can cause uneven load distribution {•
along the shank of the rivet. Other e~amples of hidden causes arn the ovality of
straight holes, and the lack of precision of the taper holes that could not be con-
trolled adequately through inspection. The viscous sealants used in squeezed riveting
methods also create serious problems when trapped hydrostatically between the rivet
and holes and then dry out, leaving voids that cause fuel leakages, stress corrosionI and galling.
The use of various types of intei.ference fit fastener systems in aircraft struc-
rural. applications has been extensive in recent years, and a variety of these fast-
eners are presently available to the industry. The advantaeges obtained by the use
of such fasteners lies principally !.n the area of fatigue design. Depending on the
type of fastener, the hole is either "propped" or "cold worked" or subjected to a
combination of these operations, and as a result, either or both the effective
alternating qtrd mean stresses in an applied fatigue loading cycle are reduced, with
a subsequent improvement in fatigue life.
In the structural design of aircraft joints, all "interference fasteners" do
not possess a common remedy for improving fatigue life, and considerable differences
exist among fastening systems that: a) volumetrically fill a hole; b) expand the
hole radially, resulting in tension hoop stress and compression radial stress, and
[ • c) expand the hole radially until the material yields plastically, but then surround
the rivet, through elastic recovery, with compressive residual stresses in both the
[ ! hoop and radial directions. The most beneficial effects are those of (c). To dis-
tinguish these effects from others, the words "cold working" the hole were given to
LI any method that left each hole with residual compressive stresses.
One system that requires nonprecision holes, utilizes inexpensive rivets, and
[ -7 cold works the holes, leaving compressive residual stresses behind it, is the Stress
7 " 77-..b i
NADC-77202- 30
Wave Riveter (SWR) (see Fig. 1). The capacitor bank at left supplies the electrical
energy that is converted to a stress wave in the tool. The black control box sitting
on the capacitor bank ensures that each pulse delivered is within specifications
for the particular fastener being upset. The rivets in this case are bucked at the
tail, the countersunk head being supported by a specially designed bucking bar.
The SWR has been used in the Grumman F- 1 4 since 1970, installing straight pins with
interference in the wing box beam, and also repairing rivets in the titanium wing
Sstructures. It was investigated by the USAF in 1972 (Ref. 1) where in fatigue
rating of precracked specimens it was found to possess clear process advantages* IAover other fastening systems.'
3 • The SWR's principle of operation was analyzed both theoretically and experi- L-mentally (Ref 2). Explicit equations to calculate the biaxial (radial and hoop)
residual stresses were deiived (Ref 3) for SWR-installed rivets that can easily pro-
vide the necessary information in riveting applications (see Appendix A). The re-
sults of a typical case are shown in Figs. 2 through 8 for 2024-T3 aluminum. Analyt-
ical results were checked with the Moire' fringe method (see Ref. 4 and Fig. 8).
Furthermore, results of fatigue testing of aluminum specimens with radial cracks,
* riveted with the SWR, showed that cracks of 0.05 in. (1.25 mm) or less, did not grow
during spectrum fatigue. Other authors (Ref. 5) have also shown the advantages of
cold working the hole, especially the improvement of fatigue life where cracks are
present. The SWR method combines: a) the fatigue advantages of cold working
the hole, b) the cost advantages in the use of ordinary rivets instead of the
more expensive precision fasteners, c) the manufacturing advantages of non- *1
precision holes and "forgiveness" in the shop whenever oval or nonaligned
holes pass inspection undetected, and d) the advantage of relaxed inspectionprocedures. , I
*The SWR-driven rivet expands radially with high velocity, cold working the material
around it as it deforms to the shape of the hole. There is no need to cold work thehole and then install a fastener; instead, all the effort is combined in one opera-tion with the SWR.
2
S• •- _ • . -r •--• • • < .. • m • • . : . f • • % --= '• .• -- • • -- . =' , '•
~'A
A,,
@0!
Ik Urn tesW v iee Ie0gUe l rdcin fF1 laimWn kn
""I ,
Li
N ,ýDC- 1720?-30
0. 1 IMATERIAL 2024.T3
RIVET 1 / IN.A-286 L; AR - 0.011 IN.
-- THEORETICAL j+EXPERIMENTAL
0.006 In.
HI I
I iL~
1 2.0 3.0
1136-001W DIMENSIONLESS RADIUS R0 /A
iIJ
_ _nE M
___ .. .. 2Reiua oo tmnvsRdilDitnc.AR001 ada Epnso
Ii NADC-77202-30
...... ..... ...... . .- ..... ... ... .....
-6i
* II
......... . ................
LA.4
~~~~~~~~~~~~~.. . . . .. ........ ........ ......................... .. ..
9 C?
A
* ~ ~ ~ ~ ~ ~ S M . IS d2O .2..................
i I 5
NAXC-77202-30
. ....... .. .............. ....... -- -
°•i \ l"• "............... .... ......... 4! ........... *............ ;l.......... .. .........
S......... ............ ............ ............ .......... ......... ......... ......-.............. 1j..
... .. . ..... ......................Si I I -
I° I
.... .........
............ .............. ........I. ..... .-... ....... ....... .... ...-... . ... ....-
........... i .............
...... j.. ... K K >..... .. I In.............. ........... .... ..... .
........ ..i ..... . . ...
... .......... ...... ... ........... ........
9 9AD
SSo. samia vtov ivnaIS3Wi
NADC-77202-30
I]
....... -... ....- -
.I ............
..... .....-..... .---........ ........... I ..> A-~t----S. .... ................ .................. i..... t -......... ! ........ ... .,........... 4. ....... ....... ........ ,....
> >
S..................-....... . .!.i... .......
........... ... ... ............ ............ t ........... i ............ .............. I...... ....... '1..... ..... . ... ........... ..
- . ................ I"1"...........l .... ....... 1"............ i ............ '. . . .. .. . . . . . .. "
S.. .......... 1.. ............ ............ 4t, ........... ............ i . ........ l..... ...... •... - - ,D A/.,S• I 2 ,/"
. . . .I ..... 4 4
I ......................... ......... .
S .... ' ........... ! . . . . ." . . ............ . . .T . ...... .... ! ....... ....• ........ . .r " • .#i --.. .
1.....r ...... ......... . ...... .. '.
9
-iA0 I V flOOHivnc
NADC-77202-30 K
I ii
r .... . .n
"" •~ .... ... ~.... -.... ... -t\I.................. .... ...... L•<> p.[
S.I .... .. . ......... . .> .. ....... .... .......... ....... . ... ... ..... ... -..... . .... .. . . .. ..........
1 4.I ............ a..n..*l i
t...... ............ ....... .. . H'I-:..~........ ........ L
... . ...... . ................. 1
. .... ......... ..... .. . ......... ............ . H. ............ ...... 5 to
.... ........... ............-....
.1 I............. ............. 1 1 1 4 .
q~~t e ' q'i''
II I00sC.as Iviaivu , aisa ij
[INADC-77202-30[I
I! - . , ~ I -
S................. ............ ....." i........... ............... ............... ......,. ............... ,........... ..............
- !a ,
U* ......... ........ . . *-
......... t. ....... .... .... .... ............... t ..........1> >
.. .. ...... I I ...
................ ............. .... ......... . .... ..... ...S.... i .... .......... .
SI I ,I.
ii i.* I I.•
I ! I 'I................................ . ... .... .....
......... ......... ..................... .......
NIVWIS dOOH ivnaisau
Li
NADC-77202-30
0.10 x
0.06__ _ _ _
RIVET IN
S0,06 (RIVET IN) RIVET OUT U
0.04 L
0.04
6.02
"� P O S IT IO NN
EER
(RIVET IN) E
ER
.0.0 , (
it.1136-007W
Filg. 8 Residual Principal Strain Distributlons, Mesured with the Moljr fFrinp Technique (Ref. 4). In a Specimen Riveted with the SWR '4.
10
L ?NADC-77202-30
2. SPECIMEN PREPARATION1iAll test specimens were prepared from 0.125 in. (3.18 amm) thick sheets of
V • 20241-T81 aluminum alloy, the grain running with the long dimension. The tensile
properties of the sheet material were evaluated and the results summarized below.
I FYty Psi 59,200 (408.2 MPa)
FtupSi 66,900 (461.3 MPa)
%e 8.4 (in 2 in.)
%RA 20.8
E, 106 psi 10.68 (73.6 MPa)
. Rupture Stress, psi 73,300 (505.4 MPa)
Single sheet plain open-hole and 1000 countersunk open-hole dog-bone shaped
fatigue specimens of the general configuration shown in Fig. 9 were prepared. A
number of the plain open-hole specimens were provided with fatigue generated pre-
cracks at the test hole edges.
Filled-hole specimens, consisting of two sheets (i.e., a "head" and "tail"
sheet) joined together at mid-section with a single fastener, or with multiple
fasteners, were prepared. In addition, single sheet filled-hole specimens with
* • back-up washers were also prepared. Specimens of each type were provided with
I ] fatigue generated precracks at the fastener hole edges while others were left un-
cracked. Furthermore, individual two-piece laminated specimens were provided with
precracks in either the "head" or the "tail" sheet, or in both sheets.
The specimens containing a single fastener had a constant radius reduced test
* section as shown in Fig. 10. The multifastened specimens were dog-bone shaped as
shown in Fig. 11 and 12.
Two countersunk fastener types were used, the subject GR501W A-286 steel rivets
and, to provide baseline data, clearance fit GB51OB3-4 steel Hiloks. The A-286
rivets (3/16 in. diameter) were dynamically installed using the Stress Wave Riveter
(SWR). A predetermined interference fit of 0.005-0.006 in. (127-152 imn) was used.Final hole preparation in these specimens was such that a sliding fit existed
between rivet and hole prior to the riveting operation. The Hilok installations
were made in accordance with the Manufacturer's installation procedures.!I
NADC -TT202-30
II It.
2.00 U)g~p1.100 .100
11.1001.100~
3/16" DIA, HOLES(3 PLACES)
.Z00 SEE NOTE 3
NOTES:1. MATL: 2024-TOI2. SURFACE ROUGHNESS RMS 126 ALL OVER ~.
EXCEPT WHERE NOTED.y3. DEaURR EDGES CF HOLES .OIOR NOTCH FREE.4. NO DISCONTINUITY IS PERMITTED AT THE TANGENCY
POINTS OF THE 2 INCH RADII WITH THE REDUCEDSECTION.
11I36-OO8W
Fig. 9 Open Hole Specimen
12
I NA.�
KCI:
I I VI: I iiii 2'
1. 1
I-. I
L"1; 1
I A
I-
I'.0 2I:
ii. 'V
I; A
1. "4'
IiL _______
13
NADC-77202- 30
4.7
U31
z~ I
LLI
ZHi
14I
NADC-77202-30
A1*
•[
Ii K
NN i4•15
1!' L) i] • i
1; 3i,.::
I .-:• ,-, ,Il .. .•.,.: •:•,,•. .. •:: ., '. ..... . ..
NADC-77202- 30 -
The spectrum fatigue loading was applied using a servo-hydraulic system con- itrolled by a preprogrammed, electronic, load feedback system utilizing a precali-
brated load cell and a punched paper tape sequencer. A constant force fatigue A
machine equipped with a 5:1 load multiplier and automatic preload control unit was 1used for most of the constant amplitude fatigue work. The servo-hydraulic system
described above was also used. The static loading was applied with a universal
testing machine equipped with an autographic load-strain recorder and a precali-
brated load indicator.
Sheets to be precracked were initially provided with 0.00 in. (2.54 mm)
diameter pilot holes located at the positions of final fastener installation. One
edge of each pilot hole was notched with a sharp edged tool to facilitate crack
initiation. Crack initiation was then accomplished in a direction normal to the
load application using an applied cylic net stress of 20 ksi (138 MPa), R = +0.05. U l
The cracks were then extended to the desired length using a net stress of 15
(103 MPa), R - +0.05. The precracked holes were then opened up to a nominal 0.189
in. (4.8 mm) diameter leaving a through-the-thickness crack measuring from 0.025
to 0.040 in. (635-1016 mm) from the hole edge. Extended crack lengths measuring up
to 0.080 in, (2.03 mm) from the hole edge were also provided, In most cases, J15,000 cycles at 20 ksi (138 MPa) were required to initiate the crEcks at all
three hole locations. An additional 30,000 cycles at 15 ksi (103 MPa) were Jthen required to extend the cracks to the desired lengths. For specimens with
three precracked holes having one crack growing more rapidly than the others,
a clamp was applied to the region of the faster growing crack to retard its
crack propagation rate, and thus ensure that all three precracks had nearly
uniform lengths.3
1.3
16
NAtC-'(7"10"~-30
3, PHO ORAI4 0V HV 1Ih
,,ju PV'Moni'tredtu1g the fat~t g%1 Kdantva~getn oft' ,AI1 rltti rd with thle ZWti wnt' A-let
[J pt~irpoe oft the work pci uwiet luider Milo pi\'gram. We ('.i't' ttot two oorek oft A
teate'. (ai) ""Peotrum fatigile lioading". andi~ b) "oonozt~ant awgpilA t~ue fat Igue loat uig.
Ii ~ ~In the first ser'ies our, aiim waitt to entabliaih it oait~ut-toul~ out' lyirto arrealtitg (,raic'k in spec~imens rbovt.et with the S'WR.Yc, o~vtiot''ona~~pw'oy (if rwtt
Ij we malintained the fol1lowing parnmet~era oonatauti
0 Material .4tIl
Is Speoimea geomet~ry
0 Rivets - 3/116 ill. 4.8 nunI) diavueter made %:f A.286 (Oniutnuau aptio tj lo0'W)
I *~~~ Radial displacement - (umoult. Of cotld wo)rk) 0.006' Au. (, >'n
I* 18pectrtim louding - V-~14 reduoed ' te vC'eationl totr F-4 i* Maixitntua Iu'ecraok oA e rnuge 0. 0t-111 0.040 A,(0.0,0~' 1.01(:' nul)
StIiiia * anql4e Oiu~ -5
* taIt-re at. limit lond in the upootriwun 3": kiti. (.14A.3 MV%~) ao.
Judgigng fromV1 t~ht anaflytical evaiuint.loti of thts rot dald~n ott eaaea and tho ait veiA
*intonsity factoir K1 , (oailulnted froat exidic It eunt~icnti detrived in zloot tol 1)), the
crnoke were not. expected to growt during toating, b wevaii the K1 valme for onoko
o.o0.5 - 0.04 A, (0.6.1 - 1.016 anui) waa tietkir gero. Inidet-d, the tsit. resutmi.
sliowed that, thero was nvocrack growth afftet .14 ,000 hours or' V'-14s ktiilt~ed t'lei ht.
(the F-14 expected lire is 6000 hiours). ttitloiytheref'ore, the resulto
present aasua'auce tha0. undetr tho 0conU t kloam Mata::: tho o'ra:'k WA I A lhvnre t o
.111O A1110ut were te kted 1t. It' keA ( Vt 10 IPit) They kiturv .1voki wt~hout. o'rnk grovwt h
to (1OPOOO F-114 aimulated hours. In Otte of thev tcipeolus, vrnoh pa'op~igt~oaa k did
I develop anid failure occurred ait 61,000 houiro
LIi
~Lki
NADC-77202-30
The complete matrix of specimens tested under the same conditions stated for
spectrum fatigue is given in Table 1 below: Li
TABLE I TEST MATRIX: SPECTRUM FATIGUE LOADING
PECIMN ... SPECIMENTYPE/PATENIR DISCRIPTION QUANTITY
CONTROLS A. OPEN-HOLE 2(SINGLE SHEET) S. COUNTERSINK OPEN-HOLE 6 L
C. PRECRACKED OPEN4HOLE 7
STRESS WAVE A. TWO-PIECE LAMINATED WITHDRIVEN RIVETS PRECRACKED HEAD SHEET ONLY [3(GRSO1W) (1) 0.25-0.040CRACK 1I
42) 0.060 IN. CRACK 6
S. SINGLE PRECRACKED SHEETWITH WASHERS 3
CLEARANCE FIT A. TWO-PIECE LAMINATED WITHG151063-4 PRECRACKED HEAD SHEET ONLYSTEEL HILOKS (1) NORMAL CLAMP-UP 6
(2) NO CLAMP-JP(3) 0,025 - 0.060 IN. CRACK RANGE
8. SINGLE PRECRACKED SHEET WITH
S13.6,-12w WASHERS (NORMAL CLAMP-UP)
Constant amplitude tests were carried out in the second series. Crack propaga- Ltion rates were observed closely for comparison between precracked specimens riveted
with the SWR and precracked specimens riveted with Hi-loks or with open holes. In ]t addition, specimens with a single precracked hole and a single sheet were tested for
comparison with the standard three-hole, two-sheet specimens. The test matrix for Ithese tests is presented in Table 2. -+
Additional specimens were tested with both sheets cracked, or either the rivet
head sheet or the rivet tail sheet cracked, to find any effects in fatigue life at-
tributable to the geometry of the specimens. None was found. Finally, tests
were carried out with a three-hole single-sheet configuration, with a washer attached L Ito each hole and riveted with the SWR to eliminate the carry through load of the
second sheet during loading, while maintaining the local effects of the second
sheet. An engineering analysis of the crack propagation results is given in Section
6. The KI values are estimated from the constant amplitude load fatigue curves
and compared with the KT values derived theoretically in Section 5. Furthermore,
a simulated F-14 lower wing cover is analyzed with damaged tolerance criteria.
The results are tabulated (Table 6) showing the weight penalcies with several
fastener methods, including the SWR method evaluated in this program. In Section
7, the results of the program are summarized and briefly discussed.
18
• I I•
" IT j .
S°-M
V TABLE 2 TEST MATRIX: CONSTANT AMPLITUDE FATIGUE LOADING
SPECIMEN TYPE DESCRIPTION SPECIMEN QUANTITY
CONTROLS A. PRECRACKED OPEN-HOLE 3(SINGLE SHEET)
STRESS WAVE A. TWO-PIECE LAMINATEDDRIVEN RIVETS (SINGLE FASTENER) .(GRS01W) (1) NO PRECRACK IN EITHER SHEET 3
(2) PRECRACKED "HEAD" SHEET ONLY 3
(3) PRECRACKED IN BOTH SHEETS 3 -
B. TWO-PIECE LAMINATED(MULTIPLE-FASTENER)
(1) PRECRACKED "HEAD" SHEET ONLY 342) PRECRACKED "TAIL" SHEET ONLY ,
C. SINGLE PRECRACKED SHEET WITH WASHERS(MULTIPLE FASTENERS) 4
CLEARANCE FIT A. TWO-PIECE LAMINATED WITH PRECRACKEDGB510834 "HEAD" SHEET ONLY (MULTIPLE FASTENERS) 3STEEL HILOKS 0
1136-023W-1
-'1
I
NADC-77202-30
4. FATIGUE TESTING
SPECIMO-N TESTING
Specimens were examined prior to testing to establish hole surface finish and
pre-crack length but standard inspection techniques were generally found to 'e un-
satisfactory due to the small size of the hole.k 4
The initial crack length was therefore measured visually with the specimen
under load, prior to the rivet installation. A sampling of specimens was also
subjected to post-failure metallurgical examination to establish initisl crack
lengths. After testing, a sawcut was made along the crack axis from left to
right, up to the end of the vertical tool marks. The specimen was then pulled toV k -failure in a'tensile test unit, failing at the zone of random fine structure in the
center of the picture. The vertical band at the edge of the hole is the crack placedin the specimen before fatigue testing. An example is provided by Fig. 13 which [
shows a sectional view of a pre-cracked specimen broken statically after spectrum
R fatigue testing for 24000 equivalent flight hours without failure. The initial crack
length was 0.04 in. (1.01 mm) in this case.
Figure 14 shows a pre-cracked specimen that was tested to failure in spectrum
fatigue. This specimen failed at 26000 equivalent, flight hours at a net limit load Listress of 40 ksi, with a pre-crack length of 0.07 in. (1.77 mm). Subsequent crack
growth from the initial pre-crack shown in Fig. 14 is typical of these test speci-
mens.
Figures 15 and 16 show the test setup and visual crack growth monitoring using Ja telescope. Specimens were prevented from buckling by the use of a stabilization
frame. All tests were conducted in a laboratory air environment, J
STRESS WAVE RIVETING
The SWR-driven rivets were of F-14 production quality. Drilling and
riveting was done under accepted Grumman manufacturing procedures.*
SManufacturing Technology MEPS 15000-066
20 5
a. NADC-7720:?-3u
Susqun Growt
11604
Fig 14 Seto iwo-rca dSeien1.7 n 17 m rc]wtSusqetGot
i 21
'Ii
7' INAW W: "
• II
1136-015W-
Fig. 15 Test Setup with Parallel Beams to Prevent Bending DuringCompression Load Cycling
I I~
I136-016W
1606 Fig. 16 Telescope Observation of Crack Growth During Testing
223
I I NADC-77202-30
Rivet installation with the SWR was performed using the latest manufacturing~ I
standards. The holes were drilled with a tolerance of 0.0015 in. (0.0381 mm).•.|
Such tolerances are accepted routinely in aircraft manufacturing at Gruman, since
drilling is accomplished with automated machines. The rivets were first installed
with a snug fit, prior to riveting. Then they were upset using the SWR (SWR setting
of 7.5 kv which produces 0.006 in. (0.152 mm) radial displacement). In practice, =
this procedure prevents gaps under the prefabricated head and provides more uniform
cold working of the hole. Furthermore, a slightly domed rivet-set was used in the
bucking bar side to expand the prefabricated rivet head. This technique is espe-
cially beneficial with GR5OIW rivets which are made with a slightly domed head. A, I flat set was used in the SWR tool side (rivet tail.).
Thd SWR method of riveting is capable of accepting considerable variation in
hole tolerances ("forgiveness") as was discussed earlier. In this program, how-
ever, the purpose was to establish its capabilities under existing manufacturing
procedures.
r 1
H4I I
NW.
L _ 23
rm~
I *NADC-77202-30
IF-5. DESCRIPTION OF K CALCULATION METHOD
I
INTRODUCTION
Whenever a crack appears in a stress field, there is a redistribution of
stress near the crack tip that permits the transmission of the remotely applied
force through the tip region (Ref 6). The remotely applied forces affect only
the intensity of the stresses near the tip, but not their distribution. The stress
intensity factor KI, therefore, represents the amplitude, or coefficient, of the
equation of the stress distribution near the crack tip. The stress intensity fac-
tor can also be seen as a limiting case of the stress concentration factor, Kt, for
I an elliptical hole.
k =The need to know the stress intensity factor (KI) in predicting the fatigue
life of structural components is well known. With the extensive use of residual
I stresses in riveted Joints, new methods are required to calculate the stress in- Ltensity factors which relate the crack length, remote loading, local loading, ana
I structural geometry.
Grandt (Ref 7) and Impellizzeri (Ref 8) used the method of linear superposi-
tion to relate local and remote loadings. Both used a weight function similar to .
the one developed by Bueckner (Ref 9) for evaluation of the integral that gives the
values of KI. In addition, both used a series expansion to relate the edge crack jequation to a single crack emanating radially from a rivet hole. Finally, bothused numerical integration to evaluate the KI integral.
1
In the evaluation of stress intensity factors of holes with cracks, and resid-
ual stresses due to SWR-driven rivets, we used the Bueckner function, the explicit jlequations for residual hoop stress distribution given in Appendix A, and a func-
tions *i suggested by Impellizzeri, which corrects for the hole and edge distance
geometry in each application.
24
.m.,.,. .~. & . . .. $t •:_ •_•. - -• ••' ,
[1 NADC-77202-30
STRESS INTENSITY EQUATIONS
'rhe stress intensity factor in a hole riveted with the SWR and loaded at in-finity can be found through the principle of superposition.r $
KI Kresidual stress + Kload
Kioai is given by the equation (2)
4Fl Kload Bowie
0.73 + 0.6762where, $Bowie0.8 ,o,,i Lo0.32,45 + S6
where "c" is the crack length and "R" is the hole radius. The stress intensity fac-
tor (K was derived analyti,'ially by integrating the following equation:
Kresidual stress ý1 4 f o(x)M(x)d~x (3)
!;2 2
where resilual stresses (Ref. 8)
OW 1 +(x) = + ()4)
and
M~x) "o.6147 (c - 2 + 0.2502 (c -X)3/2)
a weight function suggested by Bueckner. (Ref. 9) The geometry of the crack and
the hole are shown below
where R * a general posttion in the
stress field
l3-004W :5
72
-' NADC-T7202-30
Iii'1The various parameters are
RE q( 2)()
2o LIV/ 3T(7)
and
2A) (8)
whereU
F(9) E+ L
F a radial displacement (amount of cold work)
a yield stress of the material
E a Young's modulus
= Poisson's ratio 11f b *edge distance or 10R, whichever is least
(AP) -n•"- Irnt *-**i ) 1 1 + •l ) 'E (1o)
where the subscript 1 refers to the material around the rivet and subscript 2
refers to the rivet material.
I . .6449 (j} + 0.8964 ( 0.7327(+) + 0.3335 (c) 1I
is a correction factor for the effect of finite width.
26
NADC-77202-30
Integration of Eq (3) results in
reidul stress 0i±51 -5 /c- -A 3
++ [.1 34 + + RL6
.. + RLI+(- ~+ (c I )tanh-' . (2
~fII Restrictionsi The maximum crack size that the above equations can be used with is
S '< 1. For > 1, there is an error due to the divergence of a logarithmic series
introduced in the integration.
1 CALCULATIONS
The compressive residual stress distributions for 0.006 in. (0-152 mm) radial
displacement were calculated for 3/iu in. (4.82 mm) holes and 2024-T81 material.
The results are given in Figs. 17, 18 and 19. The dimensionless radius is the ra-/R) radial daisance
tioofhole radius ' ( The significance in the plots is that the bi-
.i, ,. axial compressive field (cold work) extends up to R /R - 2.4 (see Fig. 19). This isdependent on the radial displacement which, in this program, was maintained constant
at 0.006 in. (0.152 mm).*
The KI's were calculated from Eqs. (2) and (12) using a computer. The resultsIIare shown in Figs. 20, 21 and 22. In each figure there are two plot lines: the
upper graph corresponds to the KI without residual stresses, a•d the lower corres- A
ponds to the K with residual stresses. Considerable reduction in the K valueI
of the residual stress case is shown in Figs. 20 and 21.
*In this program the radial displacement vas chosen first before the fatigue
evaluation of the specimens. In design analysis the fatigue requirements will
determine the radial displacement. Further discussion on this approach is given
in Section 6.
27 . 'i
.........................................................1:i11
NADC-77202-30
0450.. ... .......... ......'........ .....
0..20. .. .............. t ........ ....0.1 1c *5G.Opsi 1403 MPQ) L I'
......... 4 . .... .. ......R 0.9 N (2.41 MM)
1136-017W DIMENSIONLESS RADIUS
Fig. 17 Distrubution of Reelduel Hoop Strain of 2024-T81 Aluminum SpecimefiSWR Rivetd with 3/16 In. (4.63 mm) Diamete A-ý6 Rivet with 0.006LIn. (0.152 min) Radial Displacement
..........-..-..-. - 4 ... -..
. ..... .....
4 .. ... t..... t..............t.....
.... . ... ....... ....... .... ..... ......
IA .... .. + ...... ......
.V .4 . 1 ps ( 403M
Sp........... ... R 0,096IN .2,4 1 mmUSOP~ M)
R,DIMENSIONLESS RADIUS-
1136-014W R
HII. 18 Distribution of Residual Radial Stress of 2024- TBI Aluminum SpecimenSWR Riveted with 3/18 In. (4.83 mm) Diameter A-296 Rivet with 0.008In. (0.152 mm) RadUi Displacement
28
CI
~bI) 1..............i*- 0.2 ..~..... .............. *.............
I.. w ....... ...... ........ ......
1. 3
I. (.152~ mm Raia 0.095o IN.(241m
1)38-019WBOWI DIMNSONUTSIAIUN)
Fig. 19 Di....ib.t...... .......u. .op.tr..of20 4.......nu S ecme
150 Riveted wit ........ n. ............3m ) D ...et... ..... Ri... wit.0.00
In.~...... .0...m) ...a..iplmmn
LOAD WIT RSDASTRESS -3,0 S 27Ma E
0L00S IN. (0.152mm) COLD WORKIRIVETED WITH SWR
1 136-020W 0rc 0,2e a p ~Fig. 20 Stre= I ntensity Factor KI vs Cad ent
R Hole kkadius
ii 29Ls:: 1
NADO-7f7202-30 .
300LOAD STRESS -35,000 psi (241 MPs) NET ______
20000 ...............- C..---ý--. . ýWITHO`UTRESIDUAL STRESSE'S.-.(BOWIE SOLUTION)
10000 -- - - .- .- -. ................- ..
500M WITH RESIDUA'L STRESE 0.0 SIN,
(1.52 mm)I COLD WORK R IV ETED I0 ~C [
1136-021W R
Fig. 21 Stress Intensity Factor K - C /Crack Length,
LOAD STR ESS - 45,000 (3 10 MPm) N ET35000
30D00
1500d
0.0 IN.0 3012m CLWR
08 RIEE I IH W
Fi.22 Stress Intensity Factor K, .. rak- agtR kol RadiusI
30
j INADC-77202-30
-~ 6. FATIGUE CRACK PROPAGATION
CONSTANT AMPLITUDE TESTS
A group of twenty-five specimens were subjected to constant amplitude fatigue
; loading, in order to obtain some basic data on the growth of cracks at holes con-j taining rivets driven with the Stress Wave Riveter (SWR). All of these specimens
were made of 2024-T81 aluminum sheet, having a test section 1.50 in. (38.1 mm) wide,-A 1/8 in. (3.18 mm) thick, with central 3/16 in. (4.83 mm) fastener holes. Some speci-
mens had single fastener holes in the test section (see Fig. iO), and others had
three in-line fastener holes in the test section (see Fig. ii). Four of the speci-
mens were tested without initial cracks, with the remainder having through-the-
thickness radial precracks at each fastener hole. The initial crack lengths ranged
from 0.020 in. (0.51 mm) to 0.040 in. (1.02 mm). All of the tests were run at a
maximum gross stress of 30.6 ksi (211 MPa), and a stress ratio (R) of +0.05 in a lab-oratory air environment. The results of all the constant amplitude fatigue tests are
summarized in Table 3.
A control group of three specimens having precracked open holes were tested and
I j found to have an average life of 3130 cycles. Next, fifteen specimens having p.'e-
cracked holes containing SWR-driven rivets were tested, and the fatigue crack propa-
gation life increased to values ranging from 77,000 to 200,000 cycles, a factor of
Li between 25 and 65. The average increase in life for all the precracked SWR speci-mens is a factor of 45.
Most of the test articles containing SWR-driven rivets consisted of two 1,8 in.
(3.18 mm) sheets fastened together. There was no significant difference in life be-
tween specimens having single fasteners in the test section and those having three
in-line fasteners. Specimens were tested in the following three categories: precracks
II in head sheet only; precracks in the tail sheet only; and precracks In both head andtail sheets. No significant difference in life was observed between these configura-
[I tions. The SWR installation method effectively retards crack growth in both the head
and tail sheets. However, when both sheets were precracked, failure always occurred
In the head sheet.
U311
FR
NADC-77202-30
TABLE 3 SUMMARY OF CONSTANT AMPLITUDE FATIGUE DATA
3/16 IN. DIA FASTENER HOLES IN 2024-T111 EE
Pow TestPwtrarlesr Die,
0.030it 1a1l1m. Fia Matto Onole.tes$Poo~nLIwllgO -od - oi - su 8"0__ to_
OP oeP2.17 0.040 35.20 301.64 3M0 Significant groWth 0t all thin hOut.&nor; She" ~~~~0.0305alr tbto oe
0.040 ' 0.
P2.20 0.030 35.0 30.67 3300 Significant growth at all throw holw.- - -. 0.021 Failure at lop hote..
Aveage, 3130
wnAV.2 I T 0 35.00 30.57 2116.000 All crack growth occurred in "head" shoot.
No Pracruacks A-3 00 0.205 I10 0 34.90 30.48 1160111,100 Failed vie initiation at location other thanr at
Single Frastenier IFIg. 101 A.? 0202 0.206 1~ 0 35.0 _3D.57 149,0100 telt hole.
WilA K.1 IX01 0204 0&030 35.001 30.5? 91301110 Failed via Initiation at location other than at2 Shet AK. 0 2 CL2001 0.203 0.040 35,00 30.5 7 1121.000 PreCroackad tIM hole. No significant CrackPs so bad Hood $heet Only I I gowth In haad ee.SW40ol aetesisi IF*g 101 AK-3 0.200 0.203 0.025 311,00 30.57 138,000 All track growth occurred in "proc, acksid"
-I-- 1~ _ - head shem.Average *152,670
ownl A 11-4 020 0 2 004 3490 30.40 103.960 All crack growth occurred In "hee ehow
Shaaao AKpK4 00 0.201) &03 0,015 34.30 30.48 1 U.200 =1. No significant growIth in "toll sheat".Sin4l Fag slnack(ig. t AKIC.S 0201 01204 0.040 38.00 30.57 11114.1110 Failed via Initlation at site other than Firs,
Sino Fra- (F 10)crack. No significant grnowth of either crack.
a"P2623 0.139T 0.202 a.0ow 34.80 30.30 204.300 All crack ii.owth In head ~het at middle2Sat0.200 1A 0.203 0.021 lastener. No siltinlficant growth at other
Plucrackad Head Dowa Only C.1n8 _ 0.2(d1 aW02 fapeterver,.P2.24 0.201 T a.203 0.036 34.20 23.57 121.00 A"l growt h at pond Idduesnar Siead e~a only. A
0.200 M4 0.230 0.040 No eignificent goarowth at other leeteinart.0.201 a am20 ftr'35
P2-27 0.2011 T &M20 0.020 34.011 30.30 1511,41M All crack growIth In head sheet, ar middle
0.201 M4 0.203 0025 .frattmner. No significant growth at other____ 0.2908 0.2030 M021 Average 10.30D Inteners.
MIS1 0.131 020 0.020 34.90 30.48 121,1111 Crack growth occurred both in the "tlal"'and
Shet 0201 *4 0Z20 0.030 "heod" ehaeits at different locations. FaillurePrw eeaad Toll Shaet Only 0.20 a 1120111 0.021 at A~ddle liseneratar. "tall" cet
InUeFrpa F 1 P220 0.200 T' 0.204 0.025 34.00 30.48 IMAM50 Failed via initiation at locatloon other then01911111 0.20 0.040 prectucked hole.
0. Sig 0.205 0.025
P2.29 0.200 T' 0.205 &.026 35.00 30.57 1410.1111 Sgnltrlfieant crack growtlh occurred both In L0.202 M4 0.205 0.030 "tall" andl "heaed' shesaft at differmnt ftee-MISS9 a 0.204 0.025 lteter localont. Failure at middle listenar of
Average 149.000 "et
P2-21 0.199 T' 0.211 0.025 35.40 30.92 69,700 Ail crick grtowIth at top lacntaner. No growthSinop Sheoo with Sack-l.P 0.2W0 M 0.207 &0.00 at other flel -neto
wafercad (NoTel oh )_0200_ __011_ _
Felow era(oTl hed020S 026 006____ ___ ___________
71*nsrn Frtonters Plg. 121 P2.40 0.1191T 0.20? 0.040 36.40 30.92 11111111 All growthawtmiddle frtener. No growtheato.200 MI 0.206 0.035 other fi mener.
9j.0.1398 0.205 0.035
P2-42 0.199 T' 0.206 0.021 35.60 31.00 177.000 All growth at mickdle feeteoner. No growth atOASIS M4 0.207 0.030 other laenetrse.0.196a 0.206 0.030____ _______________
Averaga * 94,070
Steal Hihllo. 101amonce fit) P24 0.040 35.10 30.65 4850 All Cris k growth at bottomn uletener. No328Sat 0.030 fsignlifcant growth at other lestenerePreorcruacd Head Sheeit Only 0.0403 In-ULw eae thnart(Fig. 11l - - - -. 1
P2.9 0.035 30.10 30.66 5700 All crack growth at bottom hrtoaer, No
0.030 significant growth at other fllst1nar.
10215 0.021 31.00 310.87 606 All track growth at bottom fletenr. No
lu~e~u*PW 10.025 ___ significant growth at other teletane
32
4 .. 7
NADC-77202-30
A group of three specimens, consisting of single sheets with precracked holes
d"I LI containing SWR-driven rivets backed up with washers, were tested. This group had 1an average crack growth life of 94,070 cycles, about 60% of the average life of the
I i double sheet specimens. The reason for this is not clear. It was initially con-
sidered that in the double sheet configuration, load transfer from the cracked to
unrracked sheets could possibly extend its life. However, when both head and tail
sheets were precracked, the resulting life was nearly identical to the case where
only one of the two sheets was precracked. The double sheet specimens, having only
j L one of the sheets precracked, did exhibit longer critical crack lengths than the
single sheet specimens, indicating that some unloading of the cracked sheet at the
large crack lengths (greater than 0.20 in. or 5.1 mm) occurred. However, the crack
propagation life at large cracks lengths is very short, and therefore, insignificant.
SComparison of the uncracked fatigue life (12,000 cycles, Ref. 3, Fig. 20)
for the open hole configuration with the uncracked fatigue life for specimens
with SWR fasteners, which averages 223,500 cycles (see Table 3), inditcates a
factor of 19 increase in life for the SWR.
Crack propagation analysis was performed for the open hole specimens , and is pre-sented in Fig. 23 along with test data for three specimens. The agreement between
analysis and test here is excellent Next, crack length versus cycles test data for
single and double sheet specimens with SWR-driven rivets is presented ii Figs. 211 and
25, respectively. By comparing the test data with the open hole crack propagation anal-F ' ysis, it is seen that the crack propagation life is increased on the average by fac-
tors of 30 times for single sheet specimens, to 50 times for double she-t specimei,s
by the installation of rivets with the SWR. In additien, test data for specimens
with clearance fit Hi-Loks are also shown in Fig. 25, and it Is observed that they
increase the specimen life over the opern hole configuration by a factor of about two. A
A crack propagation analysis of a cracked hole having an SWR-driven rivet was
attempted, utilizing the stress intensity solution developed in Section 5. However,
the solution (see Fig. 21) indicates that under an applied gross maximum stress of :130.6 ksi (35.0 ksi, net), the stress intensity is reduced to a level below the threshold 14-
stress intensity (w3.0 ksi 1 .). Hence, no crack growth should occur. Since craok
'I'
33
'~ _7
NADC-7' 202- 30
0.5
ANALYSIS
0.40 3/16 IN. OIA. HOLE D
'Co CC 0.030 IN.
SMAX -30.57 kol (GROSS) 0a 0.30 R "+ 0,05 13
•Z:1.50 IN. 4& : OIE"
3 0
00.20
00 04
0 P2-I?A P2.26 SPECIMEN NOS
. 010 3 P2-22
0 1.0 2.0 3.o 4.0
1136'024W CYCLES- 1000
Fig 23 Constant Amplitude Crack Propagaion (Open Hole)
$MAX 30.57 ksI (GROSS)
0.50 P~ +0.05
0.40 W L 3/16 IN. DIA. HOLE WITH SWR0 0
Cm.aS 0Q 0
G 0.30 f 0
0.20 SINGLE SHEET 0SPECIMEN 0
0 0 10 A 02!(OPEN HOLE 3P
0,10' CO 0.030 IN.) 0 0 A P2-21 43.
1AAA# 0 P2WA cA LY035 0N.I0 P24 C•, 0o 0.030ON, i
20! 40 90 tO 4 T o1 136-095~W CYCLES-, 1000
Fig. 24 Constant Amplitude Crack Propagation (Stress Wave Rivet)
34 I-
NADC- 772302-30
0.60SPECIMEN Co
SYM NO. FASTENER INCHESIi__ SMAK -30,57ksi (GROSS)4 AKK-8 SNR 0,015 0 R -'4.005
0.50 " AK.3 SWR 0.0250 0 AKK.4 SWR 0.040
P2-1 8 HI.LOK 0.030 A
*P2-9 HI-LOK 0,030 0z 040 P2.6 HI-LOK 0.040 0 A
> A
00.30 • 3/16 IN. DIA. HOLE WITH SWR 0 A
Lu 0.j
0 Az 0 x00.,0 0 a !
So DOUBLE SHEET 0 .
0.10 SPECIMEN
0 20 40 60 80 100 120 140 160 1801136-037W CYCLES - 1000 18
Fig, 25 Constant Amplitude Crock Propagation (Stress Wave Rivet)
30
25O
~20-
H/O IEDA OL/;PEICE) BACKFIGUREDS• FROM CONSTANTI b AMPLITUDE DATA -.
T WITH SWRS
-SR- 30057 ksl
0 0,06 0.103116 IN, DIA HOLE0.S• WITH SWR WITH SWRi / (PREDICTED)
110 THRESHOLD STRESS l
-- INTENSITY RANGE .0INS~LR - +0,05)'
U50 0.10 0.15 0.20 0.25 0.30
1136-038W CRACK LENGTH (C) IN.
Fig. 26 Stres Intensity
1...- - 35
NADC-77202-30
growth did occur, it was decided to backfigure the stress intensity from the fatigue L Icrack propagation rates, ia the same manner as Anderson and James (Ref. ii) and Grant
and Hinnericks (Ref. 5). The result of this analysis is presented in Fig. 26, where
it is noted that the general trends exhibited by the theoretical analysis are found,
but that the slope of the test-data-derived stress intensity as a function of
crack length is different. The methods yield K's that are significantly different
across the range of crack lengths considered with the backfigured results being
the more credible. The theoretical analysis attempted here will require further
development as discussed later in this report.
The installation of rivets with the SWR produces a residual stress field that Uretards the crack propagation rate at large distances from the hole. For example,
the retarded growth extends out to a distance greater than three radii, as can be
seen by examining Fig. 26. Previous studies with coldworked holes (Refs. 5 and 10)
indicate that the retarding effect on crack propagation disappears at a crack length
to radius ratio (C/R) of about two. The importance of this effect is that the SWR
should be able to retard the crack propagation rates with longer initial crack Lilengths, than can be done by other fastener hole treatments.
SPECTRUM TESTS
A series of tests were performed under fatigue spectrum loading. The test ar-
ticles are identical to the previously described constant amplitude specimens, ex- V1cept that they are 2.0 in. (50.8 mm) wide instead of 1.5 in. (38.1 mm). The applied
fatigue spectrum is a block fighter wing spectrum obtained from Ref. 12, where each
block represents 100 flight hours with loads arranged in a low-high-low sequence.The spectrum is presented in Table 4. Fifty-three specimens were tested, and a sum-
mary of these test results is presented in Table 5.
A crack propagation analysis was made for a cracked open hole under fatigue
spectrum loading at a gross limit stress of 31.68 ksi (218.3 MPa), which in this case
is equivalent to a net limit stress of 35.0 ksi (241 MPa). The analysis is presented
along with test data in Fig. 27, where good correlation is shown. For an initial@Lcrack length.(Co) of 0.035 in. (0.89 mm), the analysis is conservative by about 15%.
The Grumman crack closure model, as described in Ref. 13, was used to make the crack
propagation prediction shown in Fig. 27, and all subsequent predictions. Analytical
predictions and test data for cracked open hole specimens over a range of applied
limit stresses are presented in Fig. 28, where it is noted that the analysis is ac-
curate for test data shown. Also shown are test data and predictions for open hole 5A36
I--,
K NADC-77202-30
K
TABLE 4 F-14 WING SPECTRUM - 15 LAYERS*,I
Level+* Layer -€-L~imt ý' Limit 100 flightes
1 1 o.26 0.12 500.2 2 0 .3 6 0.12 50.
43 3 o.14 -0.18 1.4 4 o.41 -0.10 14.
o5 0.41 -0.03 35.6_6 0..005 •OO.7 7 o.47 0.12 200.8 8 0.53 0.12 150.90.60 0.12 70.o1 10 0.68 0.12 25.
12 12 0.84 0.12 10. l-13 13 0.92 0.12 4. .114 14 0.98 0.12 1. '15 3.5 1.o4 0.12 1. , ;
I 16 14 0.98 0.12 1.17 13 0.92 0.12 3.18 12 0.84 0.12 10.19 11 0.77 0.12 25.S10 0.68 012 90.
* -9 0.60 0.12 70.8 0.53 0.12 150.
23 7 o.47 0.12 200." -24 6 o.41 0.05 100.
25 5 0.41 -0.03 35.K26 4 o.41. -0.10 13.
27 3 o.41 -0.18 2.28 2 0.36 0,12 50629 1 0.26 0.12 500.
P, ~ (c: pondi~ng to LIKvIT Stress/Load) = 1.00
._. Designation used in fatigue analysis computer programs
Ref - ADC -76383 -30, (Ref. 12).Note: 100 f!" '-4 hour block or 100 flights arranged iii
lcw-hi, ow order.Thiz -, .ctrum is reduced from 212 layers bý equivalent
! 1exceedances.
L,
37
NADC-7202-3TABLE S SUMMARY OF SPECTRUM FATIGUE DATA FOR DYNAMIC N1.01 INSTALLED 3/10 IN. DIA A-296* INTERFERENCE FIT FASTIENERS IN 2024-TBI ALUMINUM
TV"e soemn . Pe" trst Wootil Nt40 111ulaewd Cormmentsfills PFollow Ole. Creek LImit Pfi161st W4e L
Hoead 8Sht/Tel 1t Lenifth stress To fralurC fi.) Kai
9Owen-Hole 014.1 411 5,800 Failure at mniddle hole. Signi.
0 0.109 in.01in 2.0in. 1ircant growth at top hole. No
WideWkl hole. i9Three Hoe N240.6 11,400 Failure at top hole. No Signit.-
In-LW* - cant growth at other holes.
aCountairounk 0H144 45 6,000 Failure at top hole. Significant[ joe4sole growth at all three hales.
* 01,N in.0 in 2.0in. CO11.10 48 5,500 failure at middle hale. No111111e $@nowi signiflicant growth at other L
a Thre Holes COH-1 40 9,800 Failure at top hole. No signifi-tut.Lino cant growth at other holes.
0014.3 40 10.100 Feilure at middle hole. No [significant growth at Otherholes.
0014.2 32 31,6.0 Failure lot top hole, No growthat other holes.
CONS 32 32,600 Failure at bottorn Weo. No__________________growth at other hotes.
as Prec akad 014K-2 0.028 40 4,400 Failure at bottom hole. NoOpen-Hlole 005significant growth at top hole.
Wideampl hol,. ailue attopAolea 0.I0 in. D In 2.0 In. 014K.3 0.035 40 4.400 Significant growth at all three
Wn.ide 111111110111 0.040 holes. Failure at top hole.
Ol1K4-Hle 0.03 30 14,700 Significant growth at all thraeI-ieNK10.030 holes. Failure at top hole.
0.020
P2.37 0.035 3.0 14,700 Significant growith at alltheM0.025the holes,. Failure at tphl.L90.020 ~de oe
P2-31 T 0.035 35.0 7 200 Significant cak growth at oiM0.040 althreeahole&. Failure at
em 0.028 mottom hole.
P2.45 T 0.025 35.0 7,100 Crack growth at all threeM 0.030 holet. Failure at bottom*: 1.045 hole.
0 street Wave W11) 0.020/0.030/0.040 45 60.00 No FailuraAivst43101") PI1.111) 45 60,000 No Failurej
P1.2(111 0.020/0.030/0.040 44.3 67,100 Failure at Middle fastenera Two Shoet Laminate, I'll.24 0.025/0.030/0.035 36.1 34.600 No Failura
2 In. Wide I Except P1.11 0.020/0.030/0.025 35.8 24.000 No FailureWheretNoted P1.18 0.000/0.025/0,025 35.8 24,000 No Failure .
Pi.21 35.6 24,000 No FailuraeIa he nadds PA35.7 33.200 No FAilure .
In-Line frosteonen P1.16 387 28,400 No FailureP148 35A 24.000 No Failure
a Procrackad "Head" PI.17 35.4 24.000 No FailureSheet Only P1 .20 35.A 24,000 No Failure
P1 . 35.3 24,000 No FailurePI.12 38.3 24,000 No FailureP1.13 35.3 24,000 No FailureP1.14 35.3 24.000 No FailureP1419 35.3 24,000 No FailureP10,14 0.025/0.025/0.015 35.1 24,000 No Failura
Note. (1) Spacirran width reduced from 2.0 in. to 1.80 into accomnmodate load control equipment.
1130-0OsW1Ž
383
TABLE 5 SUMMARY OF SPECTRUM FATIGUE DATA FOR DYNAMICALLY INSTALLED 3/16 IN. DIA A-2SINTERFERENCE FIT FASTENERS IN 2024-T81 ALUMINUM (CONT)
Type Illa"tttn Spa, Poet Teat Initial Nbt Uimtala4ed commentsutNo. Feltu lotow Clook Lim"t Flo i
HOWd shit"T hit Legteh 11iota To Folihil
0 SWH P2 1.1 02011. 02" 0 01% A611 20.04,10 Tlotisiit hi4W it,40202 0.20b 0060111ei "ot
a Two sheet LailttIillt 0.202 0,20? 000 sttO Ihf 411""A4 at Whutlli
to tonali NOa Clock ill -talk"
atThtoleUnloathod In- P2.14 0202 0.206 0.0h) 3b2 22.600 kliotllitl tit "ho4ill show% at ailtLin Fawws0.21 0w 00"thlois.'tsoliot% I eiiui, at
01202 0.206 0060I t1010111a I~gtskl41 Nit 0110b.
ShootOnly 112-43 0A90 0.203 0070 311.0 M11AX) 1 alliv at thttiloo takihiiifl
0Initial Clock Length 0.190 0.203 o00e aath.' loaaolnnar it Noiw(aitend to 10.060 in. in "ld thoill
P2 44 T go10 0.2011 0.06 40.1 22,300 I oilul **I ilokialol iati'iii'oN 0.199 0.204 0.010) Noa sitipidivoitt gitowth at1It 0A9111 0.206 0.0115 ltholl tilt,10wi .
P2-47 T 0.106 0204 0 06?. 40AI 26,100 141M 1Iiiu kW1to Laaittain at.ic
N 0190 0.202 0010 Noi tiorlobis-.lkvAt 4 awtti0.01111 0.203 0.0611 twhat 14419wnti
swIi.2.0 in.,d PK 124111 T 0.126 0.20 0.020 3b.6 24,000. No iitaleut No as.'lyoM:N 0.190 0,2 0.010
*Thtw Unlooted In- 6: 0 10 0.204 0.02?.Line FaeteNteu
*Single Precteckeld P2-49 T 0.106 0.20 0030 3b.1 24.00 Nolitaito Novilait~li-vtol-Sheet with Sock-up N. 0.10 0.206 0 02b ~.f~~i
P2-50 T, 3.51 24,000 Nodualao Nainitllt~I vl-k
* Cleatrance Fit NL4 Il) 0.040 44.3 8,100 80iglilll-it kitol~iatt M1 allht"G8510634 Steel (1102-31 0 04b. tostoilalvi. Failure a 41t:4Hiloks 0.040 lostollf. - .1
* 7 Shoa Laminateas HI- (11 0.0"0 44.2 1111i0o d0 atw. 411 mialola tai*t.'wi
2.Oin. wide leacept 11124) .5N 141t4ilwon Viwt14 h a thvii
Inlhosnoed 0,02ttotelitsla"oUlotl I0,030 19 latv Sinticam ioth *1hli
mtco osd L2002?.0312bx Sigijmtiant wiaawh 414oil thitooSheet only P2.81 0020 1a.iiia 1114 iaiii
0.02b. 'Aei
141.1 0.030 3b,$ 214.20 s'ill-iiitgiwilt 'j 41* all 01.90(P2.7l 003?. leteiiaI ilii' t likl
0.03b141-6 .35.2 t40 itilklia at imtkiall tostoniia Noi
(P2-12) 0.0.1t0Ioiial tlii ttto
* Clearance Fit P2-lb 7 0026 .160 81.4m IL 1#11#1101iidl laa~ie
G8 5610834 Steel N 0.026 No aiqiisivwam iaot ia~~ttiHiloks. No Clemp-Up I.000ohl ottit
* Two Sheet Latrninstot. P2-35 T 0.113?. .1.0 I.")0 sigllitwlict il vikil atll 12 in. wade11 N 0.03 ttiviu latottaivi. r#liii. 41
* Three Unloaded In aI 0.040 %kll. boattaili C1111#001Line Fasteners T 0,02B W52 81,400 kltick Viowath at 411 Ihi"
0Precracked "Heed" P02-36 N 0020 tostoil.'i in ii 5heedehet otilySheet Only a 0.02?..5nh lat teaiai i floo
Fl * Cloearnce F it P2-30 T 93.0.10 3b 6 14,300 ali 1ti ltw"NtaSle1 1 1113k W 0,020 skomfaico ltot tohi
lb Three Unlo01tiltl In P2-39 1 10.)(M .16 J110 (Cq*A- li'wthkvla ." m oaii l M11Line. Fostonsit N 0.04b v11fl~.iaaIilli~tiv4IIaI 0 (1.110 .11ta000 lalaiis.
IIeSminle Pioctwaked P2.46 T DAM2 35.11 WAX).1 1.'4WkIted wowth xvuliottat ll441L~aShoot With Back-Up M. 0.020 thtfaivv lotsi~ Foillmo'at
weashew. 2 In. 0; 0,026 aaiddll tos114114
Not. (l) Specimena width r01laac.d Itoit 2.0 in. to 1.60 .1 t to~ilit aa24elo colati ol ealutitia.ntin[I1136-03oW 2.2 1
39
NADC-77202-30
OPEN HOLE iCo 0.035 IN.o.50 SLIM 3 1.70ki LIM
(GROSS) / O3/16 IN. DIA, HOLE..• • i
t 0,40 ANALYSI
ah0.20 0
. A P2-37
0.10 0 P2:41 SPECIMEN NOS
,, A . si1 . . L .LI
0 2 4 6 8 10 12 14 16-13642awFLIGHT HOURS -1000 jS.; ~11L36-024W L
Fig. 27 Spectrum Crack Propagation (Open Hole) 2024-T81 Aluminumr
45
"INITIAL40 CRACI-
LENGTH -C (in.) TEST DATA
0.006 ONO PRECRACK
35 0.020? ,PRECRACK0,025IN.<C <0.0451IN.
22
FATIGUE 0.190 IN. DIA.ANALYSIS 0* HOLE
25 IOPRECRACK T% 2.90
PRECRACKED \ NL
10 104 105 106
EQUIVALENT SERVICE LIFE HOURS1 136-03oW .. .
Fig. 28 Spect m Fatigue Life - Open Hole Specimen. 2024.T81 Aluminum I
40!
.F•7"" • -"
ii NADC-77202-30
LIspecimens that are not precracked. These predictions are made with cumulative dam-
age calculations bascd on strain-cycling. Again, good correlation is exhibited.
Fifteen double sheet test articles (Fig. ii) were prepared with stress wave
: 1 1- rivets placed in each hole, and with each hole precracked prior to the installation
of the rivets. The precracks were located in the head sheets, and were through-the-
SVthickness radial cracks having initial lengths ranging from 0.020 to 0.045 in. All
fifteen specimens were tested to 24,000 equivalent flight hours at 31.70 ksi (218.3
MPa) gross limit stress, and no crack growth was detected. Since the fasteners were
LK countersunk, there could have been some crack growth under the rivet head, but the
crack length could not have exceeded 0.060 in., since no cracks were observed on the
specimen surface. However, four of the specimens were failed statically after com-
= ipletion of the fatigue spectrum cycling. Examination of these specimens subsequent
i to failure revealed no measureable crack growth during the 24,000 equivalent flight
hours of cycling.
An additional three precracked specimens containing rivets driven with the SWR
were tested at a gross limit stress of 39.30 ksi (270.8 MPa). Two of the specimens
achieved 60,000 hours of fatigue spectrum loads with no cracks detected on the sur-
'face, and the tests were terminated. A third specimen failed at 67,100 equivaleint
failed at about 3,000 hours (see Fig. 28), making the increase in life due to the SWR Iinstallation a factor of at least 20,
J--• Three test articles were prepared with initial crack sizes of 0.070 in. (1.78 _
mm) to 0.080 in. (2.03 mm), and were tested under fatigue spectrum loading at 30.57r 8ksi (210.6 MPa) gross limit stress. These specimens were run to failure. The crack
length versus time test data are presented in Fig. 29. Here the life from a 0.080
in. (2.03 mm) initial crack to failure is between 24,000 and 28,000 hours. This is 4
somewhat reasonable, as the total stress intensity at a 0.080 in. (2.03 mm) crack
length is about twice what it was at 0.035 in (0.89 mm), and therefore considerably
more damage will be done by the spectrum loads. These results indicate that an ini- jtal crack of 0.080 in. (2.03 zmm) or less could survive for 24,000 flight hours
at a gross limit stress or 31.70 ksi (218.2 vPa), and that a 0.050 in. (1.27 mm)
initial crack could survive even longer with SWRs installed.
41
NADC-77202-30 LJ
ifLi
0.00J
IL
0,0- SLIM " 3170 kal (GROSS[
S0,40- ANALYSIS 0 I[
(OPEN HOLE) 0 ,IN OIA.0 IHOLE WITH
o.30o °
s 'WR l-
0.20- 0 0 91}
A [0
Fig. 29 Spectrum Crack Propopltion (Strom Wave Driven Rivet) 2024.1T8I
-II
ii
0.A"
0 4 6 2026 4_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _3t
FLIHT OUS -02
I NADC-77202-30
II
Specimens were tested with clearance fit Hi-Lok fasteners, in order to show
i f a comparison of the SWR fasteners with a conventional fastener. Test data at
two limit stress levels, 31.68 ksi (218.3 MPa) and 39.30 ksi (270.8 MPa), are
U presented in Figs. 30 and 31, respectively. Three Hi-Lok specimens were
tested without fastener clamp-up. The detrimental effect of lost clamp-up
is illustrated in Fig. 30. Clamp-up is not generally considered to be areliable life enhancement technique for threaded fasteners of this size and type
so the unclamped values would likely be used in design. However, against
1 either the clamped or unclamped data the SWRs have an apparent advantage for
fatigue or damage tolorant design.
J All of the precracked specimen test data are shown in Fig. 32 in order to com-
pare them with each other and with open hole fatigue life predictions. The no
clamp-up Hi-Lok data fall between the crack propagation life predictions for
initial crack lengths (C ) of 0.05 in. (1.27 mm) and 0.005 in. (0.127 mm). The0average initi.l crack length for this group of specimens is 0.038 in. •0.965 mm).
Test data for Hi-Lok specimens with proper clamp-up tend to lie along the analy-tical curve for specimens with no precracks. The SWR specimens with initial
crack lengths less than 0.045 in. (1.143 zmm) are shown in the figure,
[ but are again inconclusive, as they are not tested to failure. One
exception is the specimen plotted at a gross limit stress of 38.80 ksi(267.3 MPa) and a life of 67,100 hours. This data point looks considerably superior
to both the Hi-Lok and open hole data. Finally, the specimens with SWR-installedL rivets with large initial crack lengths (CO > 0.065 in. or CO > 1.651 mm) are shown,
*1 "- and it is observed that they are equal to or better than the Hi-Lok specimens withinitial cracks less than 0.045 in, (1.143 mm).
I Since the test data presented in Fig 32 does not accurately present the differ-ences between specimens having large and small initial crack lengths, the data was
- replotted using the product of the gross limit stress (SLIM) and the squvre root of
the initial crack length times w, or, SLIMVW7. The result is shown in Fig. 33,
where it is now observed that the SWR data and the Hi-Lok data are separated. Av-erage lines are drawn through these groups of data, and it is noted that for a con-
stant initial crack length and limit stress, the SWR data show about seven times the
life of the Hi-Lok data. On this basis, the SWR-installed fasteners are clearly
0 ~431
NADC-77202-30
111
S%- ~oao•N. QLIM a 31.70 kal U
2 SHEET (GROSS)0 .50 SPECIMEN 0M
0.40 ANAL 1818 0EI- WEN HOLE)
A3 CLEARANCE FIT~0.30 - 0 HI-LOK SPECIMENS
G P2-79* P2-1 5 (NO CLAMP-UP)
P2-35 (NO CLAMP-UP) S0,20 /0 0 P2-6 (NO CAMP-UP)
0 15 SWR SPECIMENS-- ~0CYCLEO FOR 24,000 HOURS
13 WITH NO MEASUREABLE0.10 90CRACK GROWTH
(0.025 IN. < C0 <0.040 IN,)
00
0 4 a 12 16 20 24 28 32ElUJIVALENT SERVICE LIFE - 1000 HOURS
Fig. 30 Spectrum Crack Propagation iH-LOKS vs SWR, 2024-TB1 Aluminum I
I4
I-
NADC-'(202-30
Co000IN.I 5~LIM-39.30 kol (GROSS)
10.50 2 $MEETSPECIMEN
IANALYSIS CLEARANCE FIT040 (OPEN HOLE1) HI.LOK SPECIMENS W0
0.4 A P24ow* P2-3PCIE
P1-2
A 0
04 I]~ ~~~ ~ A4 6 2 2 0 4 6 6
E CUIV LEN D SERVC LVE000( HOURS
011!
45 NOTE: ALL SPECIMENS ARE PRECRACKED
40 ONFo ~ ~SPECIMENSj
0A
, DNF....sw15SWR SPECIMENS
CTEST DATA'1, a o OPEN HOLE 0.035 IN. < Co < 0.045 IN.
46 SWR - 0.025 IN, < Co < 0.045 IN.
30 -\ SWR - 0.065 IN. <,Co < 0.080 IN.
0 CLEARANCE FIT HI-LOK~0 0.026 IN, < Co < 0.045 IN.
SFATIGUE ANALYSIS C CLEARANCE FIT HI.LOKFI - ANO CLAMP UP) L-N ~O PRECRACK(NCLMUP
K - 2.0 N 0.025 IN. < CO < 0.045 IN.i -- ,,,PRECRACKED
20 I
o4 o• 108 w
EQUIVALENT SERVICE LIFE ~ HOURS1138.038w
Filg. 32 Spectmm Fatigue Life Stiress Wave Rivet Specimens, 2024-TOI Aluminum
24
TEST DATA
20 AEGAVERAGE:SWR O OPEN HOLELAVERAGE INSTALLED A STRESS WAVE RIVETC E RIVETS CLEARANCE FIT HI-LOKS' 10 FIT HI-LOKS CtCLEARIANCE FIT HI-LOKSNoCAPU)1
010 10
EQUIVALENT SERVICE LIFE-"HOURS
Fig. 33 Effect of Inittal Crack Lenglth on S8ecrum Fatigue LIfe 2024-T81 Aluminum
06
ANALSIS
....... ----- N~Tf~t .
LI NADC-77202-30
superiorto t he Hi-Loks. Additionally, the SWR data demonstrates about 17 times the
life of the open hole and "no clamp-up" Hi-Lok data. In order to improve the basis
for the SWR curve in Fig. 33, some of the SWR specimens tested to only 24,000 hours
should be tested to failure. These data will provide points at an SLIM I of-- about 10.0 KSi Vin. Extrapolation of the SWR curve indicates that the life for
these specimens should be about 160,000 hours.
L An analytical curve based on an open hole crack propagation analysis is also
* shown in Fig. 33. The analysis assumes that the initial crack length is 0.038 in.
I (0.965 mm), an average value for the open hole specimens, and varies the limit stress.
The curve shows good correlation with the open hole and "no clamp-up" Hi-Lok test
data.
STRESS INTENSITY OF JOINTS RIVETED WITH THE STRESS WAVE RIVETER
The stress intensity calculated for the problem posed here, and previously de-
scribed in Section 5, does not show agreement with the stress intensity derived from
test data using the Anderson and James procedure (Ref 10). The differences could be
attributed to the following:
(1) Theoretical stress intensity solutions employing linear superposition,
such ae performed here and elsewhere (Refs. 7, 14), elastically combine theI-i L]residual stress field due to plastic straining (cold working) and the re-motely applied stress field. These solutions usually predict complete crack
LI arrestment for small crack length. However, experiments here and else-where (Refs. 10, 14) show crack growth for small crack lengths, indicating
that apparently a relaxation of the residual compressive stress near thehole circumference takes place, permitting crack growth to occur. In the
LI case of non-precracked test articles, crack initiation and crack propagation
I at small crack lengths takes place similarly.
(2) Residual stresses are calculated for the uncracked configuration, when for
Smost of the specimens, the residual stresses due to the installation of the
stress wave driven rivet are introduced into the cracked configuration.
LI The present effort to predict stress intensity by SWR-produced residual stresses
and remotely applied loads is considered an initial attempt and not suitable
for design use in its present form. Further effort along these lines will be
expanded in the future.
147
T , - I Er . . ' . .. . . . -
NADC-77202-30
EFFECT OF STRESS WAVE DRIVEN RIVETS ON FIGHTER WING LOWER COVER ALLOWABLE STRESSES
An investigation into the effect of the SWR fasteners on the allowable stresses
on a fighter wing lower cover was made, ubing the fatigue spectrum used in the
present SWR study. The structure is assumed to be the riveted sheet and stringer
structure shown in Fig. 34, and is made of 2024-T851 aluminum. Design life is
assumed to be 12,000 hours, and a scatter factor of two is required. The allowable iistresses are presented in Table 6 for a variety of conditions. First, a static al-
lowable limit stress of 37.0 ksi (255 MPa) (limit) is established. Next, a fatigue Ir1
allowable of 28 ksi (220 MPa) is determined for an effective stress concentration
.factor of 3.0, i.e., for an uncracked structure. This would be representative of
either open holes or a conventional fastener system, and results in a loss of about
14 percent in allowable stress and 11 percent increase in structural weight local-
ly. Here it is assumed that the structural weight changes by 80 percent of the a)-
lowable stress change. Use of SWR driven fasteners can increase the fatigue life at'
the fastener holes sufficiently to raise the allowable stress above the 37.0 ksi
LIMI• ~static value (see Fig. 32 and consider SWR specimens with SLI = 39.30 ksi).
Considering a damage tolerant design such as is currently required by the Air
Force, and following the slow-crack-growth structure tenets of MIL-A-83444 (Ref. 15),
the designer must assume that the structure is initially cracked, and that critical *1! , holes have an initial crack length of 0.050 in. (1.27 mm). For a structure thicker
than 0.050 in. (1.27 mm), a radial circular corner crack at the hole circumference
must be assumed. For open holes or holes with conventional fasteners, and plates Ithicker than 0.17 in. (4.32 mm), the allowable stress is 23.0 ksi (158 M]Pa) based
on crack propagation life from the initial crack size of 0.050 in. (1.27 mm) to the jcritical crack size where failure occurs. If SWR fasteners are introduced, the
allowable stress will be increased to above 40.7 ksi (280 MPa), and the static a]-lowable stress will govern (see Table 6).
According to specification MIL-A-83444, the use of fatigue enhancing fastener
systems such as cold worked holes, interference fits, stress wave driven rivets, etc.,
requires that an element test program be performed to demonstrate the increased
structural life. In addition, the allowable stresses to be used for design must be
based on a crack propagation analysis for a crack from the holes in question, as- usuming a conventional fastener system is installed. However, an initial crack size
of 0.005 in. (0.127 mm) is permitted for this analysis. If the analytically deter-
mined allowable limit stress is smaller than the test data indicates, the analytical
48
~ ....-- 'a~v~ i2 .T .• • -•" .-- • ...
jaii
NADC-77202-30
TABLE 6 FIGHTER WING LOWER COVER ALLOWABLE STRESSES l2024-T151 ALUMINUM
DESIGN LIFE " 12,000 FLIGHT HOURS
1 2 34 5 6INITIAL WEIGHT iDESIGN CRACK ALLOWABLE PENALTY
CRITERION FASTENER LENGTH STRESSCONFIGURATION SYSTEM (IN.) (Kil) (PERCENT)
Static 37.0 BASELINEStrength
0,34IN. ~-+95L
2CLEARANCE _ _. _ _9.
Fatigue WINTERFERENCE >37,0 0-F >
SWR >>37.0 0
S~LESS THAN 0.10 IN.3/16 IN. DIA.
Co CONVENTIONAL 0.06 20.0 +37,0
SWR (. 25.0 +26.00,05 >>37.0 0
3/16 IN. DACoCONVENTIONAL 0.06 23,0 +30.0
00.24 IN. V) 28.0 +19.5
T 0.06 >>37.0 0
0.23 IN. o0.2 20.0 +37o0
- 0.10 35.0 +4.3
(1) MIL-A-83444 REQUIRES AN INITIAL CRACK OF 0.006 IN. BE USED WITH CONVENTIONAL CRACK PROPAGATION ANALYSISFOR CALCULATING DESIGN ALLOWABLE STRESSES.
1136-036w
5I0[
K NADC-77202-30
Inumber must nevertheless be used for design purposes. Therefore, these numbers are
also presented in Table 6, and it is noted that their allowable stresses are con-
siderably lower than those demonstrated by test data for the SWR system. For ex-U ample, consider the case in Table 6 with the corner crack at a hole with SWR-driven
rivets. Here, the MIL-A-83444 required analysis provides 28.0 ksi (192 MPa)
allowable limit stress for a plate thickness of 0.23 in., whereas the test data
LI generated indicate an allowable stress greater than 40 ksi (.275.6 MPa).
Specification MIL-A-830)4 also requires that cracks in the structure away from
rholes must also be considt-d. ,i'r standard non-destructive inspection (NDI) pro-
cedures, a semi-circular inisiil surface crack having a surface length of 0.25 in.
(6.35 mm) must br assumred to be present, and the limit allowable stress for this
crack will be 20.0 ksi (137 MPa), far less than the static allowable stress. This
stress can be increased by reducing the initial crack length to 0.10 in. (2.54 mm)
and introducing an NDI demonstration program to establish that the NDI procedures canIi • find the smaller cracks with sufficient reliability. This approach has been imple--. mented in the F-16 program. Using a 0.10 inch initial surface crack length (2C ) in-
0creases the allowable stress to 35.0 ksi (241 MPa), and this value will now govern
the design. The resulting weight increase is less than 5% over a static design.
The table shows a maximum weight penalty of 37% for a structure with conventional
clearance-fit fasteners designed to damage tolerant criteria. Using a stress
wave driven rivet fastener system and following the procedures of specification
MIL-A-83444 at holes and improved NDI away from holes, the weight penalty is
reduced to 19.5%. Furthermore, by using a statistical sample of test data
i for the SWR fastener treated holes having 0.05 in. (1.27 mm) initial cracks
rather than the analytical procedure required by MIL-A-83444, the weightpenalty can be reduced to 4.3%. It should also be noted that straight and
tapered shank interference fit fastener systems in precracked structures havenot been included in the present program.
51
NADC-77202-30 i
7. CONCLUSIONS AND RECOMMNDATIONS
The major conclusion of this study is that the installation of stress wave
driven rivets is extremely effective in retarding the propagation of cracks from "
holes. For the case considered, 2024-T81 alumiinum specimens subjected to typical
f' ighter wing fatigue spectra and a typical gross limit stress (31.70 ksi or 218.3
"MPa) , precracked holes were shown experimentally to experience no detectable crack
growth during 24,000 equivalent flight hours of spectrum loading. The specimens
tested had precracks ranging from 0.020 to 0.045 inches in length, and had A-2186
steel stress wave driven rivets installed with 0.006 inch (0.15 mm) radial displace-
ment. There were enough specimens tested under these conditions to provide a valid
statisticRl sample, indicating that this result could be applied to a similar design H .-
situation provided that there are no damaging environmental factors present. The L -5-_
crack propagation rates were reduced to the extent that the cracks could be considered
[ to be arrested. .. ;
A few specimens identical with the previous group, were tested at a higher gross
limit stress (39.30 ksi or 270.8 MPa), and here crack growth retardation from the
stress wave driven rivets was sufficient to provide lives in excess of 60,000 equiva-
lent flight hours. This was not a statistical sample but indicates a good potential Lfor application at this stress level. In addition, a few specimens were tested at a
gross limit stress of 31.70 ksi (218.3 MPa) but with initial cracks about 0.080 inches
(1.78 mm) in length, and here significant crack growth retardation was also exhibited.
Lives for this group of specimens averaged about 26,000 equivalent flight hours. It jshould be noted that for all these cases, the stress wave induced radial displacementwas limited to 0.06 in (0.15 nun). A higher radial displucement (say O.01." "'1. Or
0.30 mm) would have provided higher residual stresses and Increased crack growth
retardation. The increase in radial displacement will "cold work" the plate material
to greater radial distances from the rivet, providing crack growth retardation for
longer crack lengths. This would be particularly effective for the case of long
initial crack lengths.
Results obtained during the present program indicate that significant weight
in the manufacture of damage tolerant airframe structures. At the present time
the Stress Wave Riveter is not a fully automated production unit, but the syatem 5 *
521
to NADC-T1202-3CIicould easily be incorporated in a Drivematic type riveting machine and used at
normal production rates.
Other important recommendations are:
Li *Partial load transfer specimens riveted with the SWR, should be testedin addition to the zeo load transfer specimens evaluated to date.
LI • Specimens riveted with the SWR should also be tested in controlled
corrosive environments to establish whether there is any degradation
of the residual stresses.
* Specimens of various thickness and materials should be tested to
establish SWR settings before automated applications are made.
• Additional spectrum testing should be performed using load
spectra with higher levels of compression loading.
o Higher levels of radial displacement (cold work) and wider
L variations in hole tolerance should be investigated.*] The repeatability and consistency of tile SWR process in a pr'o-
duction environment should be examined.
i IV
L53
il I]
~ -r 'L
NADC-77202-30
8 REFERENCES
I. Tiffany, C. F., Stewart, R. P., and Moore, T. K., "Fatigue and Stress - Corro-
sion Test of Selected Fasteners/Hole Processes," ASD-TR-72-111, Jan. 1973.
2. Leftheris, B. P., "Stress Wave Riveter System Analysis," Grumman Research fl i
Department Report RE-503, July 1975.
3. Leftheris, B. P., "Advantages of Residual Stresses in Dynamically Riveted uJoints," Grumman Research Department Report RE-552, Feb. 1978.
4. Horsch, F. J. and Schwarz, R. C., "Moire Fringe Data Handling System for Ap-
plication in a Industrial Laboratory," Grumman Engineering Test Operations,
presented at the Sixth International Conference for Experimental Stress Analy- L <•
sis, Sept. 1978.
5. Grandt, A. F. Jr., and Hinnerichs, T. D., "Stress Intensity Factor Measure-
ments for Flamed Fastener Holes," U.S. Air Force Materials Laboratory Wright-
Patterson Air Force Base, Ohio; presented at the Army Symposium of Solid LiMechanics, 10-12 Sept. 1974.
6. Paris, P. C. and Sih, G. C., "Stress Analysis of Cracks," ASTM Special Tech-
nical Publication 381, 1970.
7. Grandt, A. F. Jr., "Stress Intensity Factors for Some Thru-Cracked Fastener
Holes," Air Force Materials Laboratory, Wright-Patterson Air Force Base, Ohio;
presented at the Seventh U. S. National Congress of Applied Mechanics, June
19714.8. Impellizzeri, L. F. and Rich, D. L., "Spectrum Fatigue Crack Growth in Lugs,"- °
Symposium in Fatigue Crack Growth Under Spectrum Loads, hSTM STP 595, American
Society for Testing and Materials, p.p. 320-326, 1976. U9. Bueckner, H. F., "Weight Functions for the Notched Bar," ZAMM 51, 97-109,
1971. S:U!!
10. Chandawanich, N. and Sharpe, W.N., Jr., "An Experimental Study of Fatigue
Crack Initiation and Growth from Coldworked Holes," Engrg. Fract. Mech., iI f
vol. 11, pp. 609-620, 1979.
3.1. James, L.A. and Anderson, W.E., "A Simple Experimental Procedure for Stress 3Intensity Factor Calibration," J. of Eng. Fracture Mechanics, vol. 1,
April 1969, pp. 565-568.
54
NADC-77202-3.
12. Broek,, D. and Smith 6% H1., "Spectrum Loading Fatigue Crack Growth Predic-
tions and Safety Factor AngayS13," Report No. NADC-76383-30, Battelle
Columbus Laboratoriest 214 September 1976.
13. Bell, P'.D. and Creager, M.,, "Crack Grovth Analysis for Arbitrary Spectrum
Loading," vol. 1., AFF1)L-TR-T14-129, October 1974.
114. Haul T.M., McGee, W.M., and Aberson, JA., "Extended Study of Flaw Growth
at Fastener Holes," AFFDL..TR-TT-83, April 1978.
15 MIL-A-834144 (USAF), "Airplane Damage Tolerance Requirements," July 1974.
55.
NADC-77202-30
APPENDIX A
I EQUATIONS OF RESIDUAL STRESSES
The residual radial and hoop stresses and strains around an SWR-riveted fas-
tener are given by the followii!R equations (Ref 3).
aI + 2K 1n . - 1-
I- _. • =+ U-
+t2KinK + 1t 1 (A-2)
L i _+ _R)r 2 F(t) - RKln + (A-3)
F 2) ( ln + u F(t)- (-
where :
IiI"S.... K = uniaxial yield strength
= ... •iFt radial displacement (extent of cold work) Ii,•R 1 E l + F(t) E'+1A
H = radius of hole before riveting '•
R E R +E
]. C=~1 + •2.~.--iwhere subscript 1 refers to the material surrounding the ':
S..... ' rivet and subscript 2 refers to the material of the rivet ';
[ 'I ~B =reference radius such that B >>R, or riveting of joirts equal to the edge ••
Ldistance0H = radius (variable) such that R < (0 < B
0 0
r (A-4)
where
HADC-77202-30U
The above equations do not include secondary yielding. Reference 3, how-
ever, gives the complete derivation. f
LI
58
NADC-77202-30
D:STRIBUTION LIST
ri Government ActivitiesNo. ofCopies
7 NAVAIRSYSCOM, AIR-50174(2 for retention, 2 for AIR-530, 1 for AIR-530215, 1 for AIR-530221C,2 for AIR-320B) . . . . . . 8
NAVAIRTESTOEN, Patuxent River, Maryland . . . . . . .. . 1NAVAVNSAFECEN, NAB, Norfolk, Virginia . . . . . . . . . iCRAVANTRA, NAS Corpus Christi, Texas 1CNABATRA, HAS, Pensacola, Florida . . . . . . . . . 1
I• CNARESTRA, NAS, Glenview, Illinois ICNATRA, NAS, Pensacola, Florida . . . . . . . . . . 1NAVAIRSYSCOMREPLANT . . . . . . . . . . . . 1NAVAIRSYSCOMREPCENT. . . . . . . . . . . . . .NAVAIRSYSCOMREPAC . . . .. 1NAVAIREWORKFAC, NAS, Alameda, California. . 1NAVAIREWORFKAC, NAS, Jacksonville, Florida . . . . . . . . 1NAVAIREWORKFAC, NAS, Norfolk, Virginia . . . . . . . . 1"NAVAIREWORKFAC, NAB. Pensacola, Florida . . . . . . .. 1NAVAIREWORKrAC, NAS, Quonset Point, Rhode Island . . . .. 1NAVAIREWORKFAC, NAS, San Diego, California . . . . . . . .1NAVAIREWORKFAC, NAS, Cherry Point, North Carolina . . . . . .1COMNAVAIRLANT . . . .. ..... . . . 1
L • COMNAVAIRPAC. . . . . . . . . . . . . . .NWL, Dahlgren, Virginia (Attention Mr. Morton) ISUSAF AF•DL, Wright PatLterson Air Force Base, OH 4533 (Attention
Dr. R. Steward Code ) . . . . . . .USAF Systems Command, WPAFB, Ohio 45433
Attention FBR . .1Attention . . . . . . . . . . . . . 1Attention LLD . 1Attention SEFS . . . . . . . . . . . . 1Attention FYA . . .. . . . . . . . . 1Attention LAM . . . . . . . . . . . I 1Attention FBA . . . . . . . . . . .. 1.Attention LPH . . . . . . . . . . . . 1
USA AMMRC, Watertown, Massachusetts . . . . . . . . 1 .USA APG, Aberdeen, Maryland. .. . . . . 1USA AMRDb, Fort Eustis, VA 23604 (mi. Berrisford) . . . . . . iUSA AVSCOM, St. Louis, Missouri (Attention AMSAV-GR)Defense Research and Development Staff, British Embassy
Washington, D.C., via NAVAIR (ATR-5302) . . . . . .1Canadian Joint Staff, NaNT Member
Washington, D.C., via NAVAIR (AIR-5302) . . . . .Technical Advisory, AFlAS-B, Directorate of Aerospace Safety,
Norton AFB, California . . . . . . . . . . . . 1DDC . ........ 12NAVSLASYSCOM, Washington, D.C. 20362' (Mr. C. Pohler, Code 035) 14
NAVSHIPRADCEN, Bethesda, MD 20034(Attention Mr. A. B. Stavovy 730, . . . . . .
59
p .. ..- - • '. . - - - ;
NAD-77202-30 1
DISTRIBUTION LIST (cont)No. ofCopies
Aircraft Industry
Bell Aerosystems Co., Buffalo, Now York 14205 . . . .. . . . 1Bell Helicopter Co., Fort Worth, Texas 76101.. . . ..- 1Boeing Co., Airplane Div., Seattle, Washington 98124 (Mr. T. Porter and
Mr. Joseph Phillip .). .. .. . . . ..- 1 aBoeing Co., Airplane D1,r., Wichita, Kansas 67210 1Boeing Co., Vertol Div., Thila., Pa. 19142 . ... . . . .. . 1McDonnell Douglas Aircraft Corp., Aircraft Div.,
Long Beach, California 90801 . . .1.. . . . 1General Dynamics/Convair, San Diego, California 92112 . .1. . . 1General Dynamics Corp., Fort Worth, Texas 762301 . . . . . .1Goodyear Aerospace Corp., Akron, Ohio 44305 . .. . • • • 1 IGrumman Aerospace Corp., Bethpage, Long Island, New York 117.14
(Mr. B. Beal and Dr. B. Leftheris) . . . . . .. . .Aircraft-Missiles r v., Fairchild-Hiller Corp.,
Hagerstown, Maryland 21740 . . . . • .* • . . .Kaman Aircraft Corp., Bloomfield, Connecticut 06002 . 1Lockheed Aircraft Corp., Lockheed-California Co.,
Burbank, California 91503 . . . . . . . . .0 1 ULockheed Aircraft Corp., Lockheed-Georgia Co., Marietta, Georgia 30061 . 1LTV Aerospace Corp., Dallas, Texas 75222 . . . . ., . . . 1Martin Co., Baltimore Md. 21203 . . . . . . . . . . 1McDonnell Douglas Aircraft Corp., St. Louis, Missouri 63166 . • . 1Rockwell International, Columbus Aircraft Div, Columbus' OH 43216! ~ ~(Mr. 0. Acker) ....- 1Rockwell International, Los Angeles, CA. 90053 (Mr. 0. Fitch) . . . 1Northrop Corp., Aircraft Div., Hawthorne, California 90250Republic Aviation Div., Fairchild-Hiller Corp.,
Farmingdale, Long Island, New York 11735 . . . . . . 1Sikorsky Aircraft Co., Stratford, Connecticut 06497 . " . . " 1Standard Pressed Steel Corp., Jenkintown, PA 19046 (Mr. Garreth)
Research Organizations
Drexel University, Philadel-hia, PA 19104 (Attention Dr. H. Harris)Hdqtrs., R&T Div., AFSC, Bo~iing AFB, District of Columbia 20332. . 1 IOffice of Aerospace Research, Arlingtoi., Virginia 22209 . . . . 1FAA (FS-120), Washington, D.C. 20553 . . . . . . . . . 1Scientific and Technical Information Facility, U
College Park, Maryland 20740 (NASA Rep.) 2Administrator, NASA, Washington, D.C. 20546 . . . ..- 1NASA-Langley Research Center, Materials Division Hampton, VA 23365 3
(Mr. H. F. Hardrath) . . . . . . . . . . . 1National BuStds, Washington, D.C. 20234 . . . IOffice of Naval Research, Washington, D.C. 20362 Dr. N. Perrone) . I 1Director, Naval Research Lab, Washington, D.C. 20390 . . . . . 1Midwest Research Institute, Kansas City, M ,souri 64110 . . . . . IUniversity of Illinois, Urbana, Illinois 61803 (Prof. T. J. Dolan
ar' Prof. J. Morrow) . . . . . . . . . . . 1 each
603
IINAD-TT202-30
DISTRIBUTION LIST (cont)..
No. ofCopies 77
D University of Kansas, Lawrence, Kansas 660o44. . . . . . .1University of Michigan, Ann-Arbor, Michigan 48105 . . . . . . 1University of Minesota, Mineapolis, Minnesota 55455Alcoa, Alcoa Labs, Alcoa Center PA 15069 (Mr. J. G. Kaufman). 1.'LeHigh University, Bethlehem, PA 18015 (Prof. G. C. Sih) . . .. . 1University of Dayton Research Institute 300 College Park
Dayton, OH 46469 (Dr. J. Gallagher). . . . . . . 1Belfour Stulen, Inc., Traverse City, Michigan 49684 . . . . . . ICornell Aero. Lab, Buffalo, New York 14221 . . . . . . . . 1Metals & Ceramics Information Center, Battelle, Columbus Laboratories,
Columbus, Ohio 43201 1NASA, Levis Research Center, Cleveland, OH 44153 (Tech Library) . . . 1Naval Postgraduate School, Monterey, California 9394011 (Prof. Lindsey) . . . . . . . . . . . .. . 1
Government Agencies
U ASD, WPAFB, OH 45433 (Attn: Dr. J. Goodman). . . . . . . .AFFDL, WPAFB, OH 45433 (Attn: M4r. R.M. Bader) . . . .. 1NASA, Langley Research Center, Hwnpton, VA 23665 (Attn: Mr. H. Hardrath). 1USAAVRDC, Applied Tech. Lab, Fort Eustis, VA 23604 (Attn: Mr. H.
Reddich, Jr.) . . . . . . . . . . . .1British Defence Staff, Defence Equipment Staff, British Embassy,
3100 Massachusetts Aye., N.W., Wash. D.C. 20008(Attn: Mr. F.S. Wood) . . . . . . . . . .
Distribution Outside United States
National Aero. Establishment, National Research Council, Montreal Rot d,*' Ottawa KIA OR6, Ontario, CanadaLAttn: Mr. John Dunsby. . . . . . . . . . 1Royal Aircraft Establishment, Farnsborough, Hants GU146TD, England
SItructures Department (Attn: Messrs: P. Guyett, R. Maxwell,E W. Kirkby). ...... . . . .. 3
Aeronautical. Research LabsStructures DivisionBox 4331 PO, Melbourne, Victoria 3001 Austrailia(Attn: Dr. Alfred Payne)
Defense Scientific Establishment11MNZ Dockyard, Aukland 9, New Zealand (Attn: Mr. H. Levinsohn) . 1
6
I r
Related Documents
![Index [link.springer.com]978-0-85729-507-1/1.pdf · High school personality questionnaire (HSPQ), 130 Hoffa’s fat pad extraskeletal osteosarcoma, 148 localized pigmented villonodular](https://static.cupdf.com/doc/110x72/5be095d809d3f2de4d8c5b79/index-link-978-0-85729-507-11pdf-high-school-personality-questionnaire.jpg)
![bioszfera.combioszfera.com/downloads/Tamasztoszovetek-fakt.pdf · 2019. 9. 27. · pp pdjdv khqjhu pho\qhn n|]hspq d fhqwuiolv fvdwruqiedq +dyhuv ipoh fvdwruqd ypupu ixw $] rv]whrqedq](https://static.cupdf.com/doc/110x72/60ac6a000830c13e037ed9e4/2019-9-27-pp-pdjdv-khqjhu-phoqhn-nhspq-d-fhqwuiolv-fvdwruqiedq-dyhuv-ipoh.jpg)
![Journal für Reproduktionsmedizin und Endokrinologie · HSPq, PolyAr [21] Na/K α1r, Zyklin Ar [22] FSHRr, Cx43r, COX2r, EP2r, EP3r [23, 24] FAS q, FASLGq, STAT3q [25] Qualität der](https://static.cupdf.com/doc/110x72/5eaba0d3fd2c3f41d33c3c62/journal-fr-reproduktionsmedizin-und-endokrinologie-hspq-polyar-21-nak-1r.jpg)
