P273135

19

AGARD-AG-304

<
Q
ec

<
<

AGARDograph N o . 304

Standard Fatigue Test Specimens

for Fastener Evaluation

ORIS
.RD

DISTRIBUTION A N D AVAILABILITY
O N BACK COVER

si?

708848

AGARD-AG-304

NORTH ATLANTIC TREATY ORGANIZATION

ADVISORY GROUP FOR AEROSPACE RESEARCH AND DEVELOPMENT
(ORGANISATION DU TRAITE DE L'ATLANTTQUE NORD)

88 / C O 4 ^ A ^

As/W
COOAT ;

0L,,n

AGARDograph No. 304

STANDARD FATIGUE TEST SPECIMENS
FOR FASTENER EVALUATION
by
Robin Cook
Materials and Structures Department
Royal Aircraft Establishment
Farnborough, Hampshire GUI4 6TD, UK

This AGARDograph was sponsored by the Structures and Materials Panel of AGARD.

g-g

THE MISSION OF AGARD

The mission of AGARD is to bring together the leading personalities of the NATO nations in the fields of science and
technology relating to aerospace for the following purposes:
Exchanging of scientific and technical information;
Continuously stimulating advances in the aerospace sciences relevant to strengthening the common defence posture;
Improving the co-operation among member nations in aerospace research and development;
Providing scientific and technical advice and assistance to the Military Committee in the field of aerospace research
and development (with particular regard to its military application);
Rendering scientific and technical assistance, as requested, to other NATO bodies and to member nations in
connection with research and development problems in the aerospace field;
Providing assistance to member nations for the purpose of increasing their scientific and technical potential;
Recommending effective ways for the member nations to use their research and development capabilities for the
common benefit of the NATO community.
The highest authority within AGARD is the National Delegates Board consisting of officially appointed senior
representatives from each member nation. The mission of AGARD is carried out through the Panels which are composed of
experts appointed by the National Delegates, the Consultant and Exchange Programme and the Aerospace Applications
Studies Programme. The results of AGARD work are reported to the member nations and the NATO Authorities through
the AGARD series of publications of which this is one.
Participation in AGARD activities is by invitation only and is normally limited to citizens of the NATO nations.

The content of this publication has been reproduced

directly from material supplied by AGARD or the author.

Published November 1987

Copyright AGARD 1987
All Rights Reserved
ISBN 92-835-0429-1

Printed by Specialised Printing Services Limited

40 Chigwell Lane, Loughton, Essex IG10 3TZ

PREFACE
Aircraft fatigue is a very expensive phenomenon in terms of inspection, of maintenance and repair and of decreased
aircraft availability. Hence, design for good fatigue performance remains of great importance. Mechanically fastened joints
quite often turn out to be the fatigue critical elements in aircraft structures. Recently, an extensive co-operative programme
coordinated by the AGARD Structures and Materials Panel was completed in which the fatigue performance of a wide range of
fastener systems were evaluated.
During this programme, which also used a variety of specimen configurations, the desirability of a limited number of
standard specimens to facilitate fastener evaluation in the future, became apparent. Hence, the Structures and Materials Panel
decided to set up a Working Group that would try to define a limited number of recommended specimen configurations, on the
basis of a co-operative test programme.
This programme, which was very ably co-ordinated by Mr. R. Cook of the Royal Aircraft Establishment has now been
successfully completed and the present report contains the results of this collaborative effort.
J.B. de Jonge
Chairman, Working Group on Standard
Fatigue Test Specimens for Fastener Evaluation

SUMMARY
An AGARD coordinated programme which examines the fatigue performance and joint characteristics of a number of
mechanically fastened joints has been completed. This report describes the programme which examines mechanically fastened
joints with 1) no or low secondary bending and 2) with high secondary bending. In part 1, three types of joint are assessed which
exhibit no, low and high amounts of load transfer by the fastener. The no load transfer joint was rejected and the low and high
load transfer joints were considered to be equivalent in rating fastener systems. In part 2, three types of single shear joint are
considered. They are compared on the basis of load transfer and secondary bending characteristics and also on the fatigue
endurance with a range of fastener systems installed. Only one joint, the UK designed Q-joint, adequately fulfilled the
requirements of a standard joint for fastener evaluation purposes.

L'AGARD ("Advisory Group for Aerospace Research and Development" = groupe consultatif pour la recherche et les
realisations aerospatiales) a mene a bien un programme coordonne d'examens des performances en fatigue et des
caracteristiques de solidite de tcnuc d'un certain nombre de liaisons par fixation mecanique.
Le present expose decrit le programme qui traite des jonctions mecanique presentant
(1) une manifestation faible ou nulle de flexion secondaire,
(2) une flexion secondaire importante.
Dans la premiere partie, on examine trois types de liaison qui revelent un niveau de transfert de charge par la fixation
respectivement nui, faible et eleve. On a ecarte la solution d'une liaison sans aucun transfert de charge et estime qu'un transfert
de charge faible ou eleve constituait un bon equivalent pour revaluation des systemes defixation.Dans la deuxieme partie, on
traite trois types de liaison a cisaillement simple. On les compare entre eux en etudiant le transfert de charge et les
caracteristiques en flexion ainsi que la resistance a la fatigue d'une serie de dispositifs de fixation posee. Un seul systeme, le
modele a pression dynamique concu par les Britanniques, a satisfait convenablement aux conditions imposees a un assemblage
c<uiiant destine a 1'evaluation des fixations.

i\

CONTENTS
Page
PREFACE

iii

SUMMARY

iv

1.

INTRODUCTION

2.

PROGRAMME OVERVIEW

2.1 No or low secondary bending programme

2.2 High secondary bending programme

1
1

STANDARD SPECIMEN DESIGNS

3.1 No load transfer specimens
3.2 Low load transfer specimens
3.3 High load transfer specimens
3.4 Specimen designs with high secondary bending
3.5 Specimen requirements

2
2
2
3
3
3

FASTENER SYSTEMS

3.

4.

4.1

111 I O K i n p l a i n h o l e ( 1 A and 2 V)

4.2 HI-LOK in cold-worked hole (IB and 2B)

4.3 HI-TIGUE in plain hole (1C and 2C) and HI-TIGUE in cold-worked hole (2D)
4.4 HUCK-EXL(ID)

4
4
4

5.

TESTING PROGRAMMES
5.1 No or low secondary bending test programme
5.2 High secondary bending test programme

4
4
5

6.

RESULTS AND DISCUSSIONS

6.1 Results and discussion of the no or low secondary bending tests
6.1.1 Low and high load transfer joints
6.1.2 No load transfer joints
6.2 Results and discussion of the high secondary bending tests
6.2.1 Load transfer and secondary bending results

5
5
5
6
6
6

6.2.2 Fatigue test results

7.

CONCLUSIONS AND RECOMMENDATIONS OF PART 1, NO OR LOW SECONDARY BENDING

8.

CONCLUSIONS AND RECOMMENDATIONS OF PART 2, HIGH SECONDARY BENDING

References

Tables

10

Figures

27

Annex 1

Fastener

fits

39

Annex 2

Pin, collar and nut part numbers used in investigation

41

Annex 3

Load transfer and secondary bending measurements

42

Annex 4

Importance of scatter

43

Annex 5

Primary fatigue crack origins

45

Annex 6

Q-joint assessment

57

INTRODUCTION

The most common site of fatigue crack initiation in aircraft structures is from a fastener hole. In consequence there have
emerged numerous fastener systems which claim to improve the fatigue life of the joined components. The design engineer
requires to know how these fastener systems will perform in his particular application such that he may choose a safe but
economic solution. This requirement was recognised by AGARD and two coordinated programmes have been completed
which have addressed this problem. The 'Critically Loaded Hole Technology'( 1) programme was a pilot programme which
established that consistent fatigue data can be generated in complex fatigue tests between a number of participating countries. It
examined principally the effects of hole quality on the fatigue performance of open hole and low load transfer joint specimens.
A comprehensive follow on programme, the Fatigue Rated Fastener Systems (FRFS) programme(2), has now been completed.
This examined the fatigue performance of a range of fastener systems in many different joint configurations in different
materials with a selection of hole preparation techniques and installation parameters. A large amount of valuable design data
was generated on the different specimen types. Comparison of different test data was achieved using core programmes of
specified parameters in which all countries participated.
In order to facilitate such comparisons in future work it is necessary that a number of standard fatigue specimen
geometries are defined and used in conjunction with the standard loading sequences that have been developed in recent years.
Accordingly an AGARD group entitled Standard Mechanical Joint Fatigue Specimens' was established(3) with the task of
defining a number of standard joints for fastener evaluation. The aim was not to design new joints, but to assess in some detail a
number of the joints used in the FRFS programme. This is the final report of the AGARD working group, the participants of
which are given in Table 1.

PROGRAMME OVERVIEW

The most important requirement of any standard specimen is that it should be representative of the structural feature
which it is to simulate. In the case of joints, the main parameters which need to be represented are the amount of load
transferred by and bypassing the fastener, the amount of secondary bending and the way in which these are controlled. The
FRFS programme concluded that the primary parameter is in fact secondary bending. Accordingly the standard mechanical
joint programme is split into two parts. Part 1 considers joints with no or low secondary bending and is described in section 2.1.
Part 2 considers joints with high secondary bending and is described in section 2.2.
2.1 No or low secondary bending programme
In aircraft wing construction there are many areas which exhibit no or low secondary bending. Examples of this are spanwise joints, either skin-to-spar or skin-to-stiffener. The skin-to-stiffener joint will contain very low load transfer and should not
be a fatigue-critical area. The skin-to-spar joint will for the most part be a low load transfer situation but depending on local
design may have a high load transfer near the wing root and become fatigue critical.
Chord-wise joints are on the other hand predominantly high load transfer situations. Where double shear butt joints occur
(i.e. no secondary bending) fatigue resistance is generally good. It is arguable therefore whether this type of joint is fatiguecritical. Chord-wise joints with single shear fasteners generally exhibit significant amounts of secondary bending and are
considered in Part 2 of this exercise (section 2.2).
It is therefore necessary in this part of the programme to consider a number of joints which exhibit various degrees of load
transfer and to pose the question:
Do all of the joints considered, rate fatigue resistant fastener systems in the same way?
There was conflicting evidence from the FRFS programme on this question. If this criterion is satisfied for all of the joints
considered then only one specimen type needs to be defined as a standard. If this criterion is not satisfied, then a minimum
number of joints need to be selected.
Many laboratory joint specimes with no or low secondary bending are intended to be as realistic as possible. Thus many
contain several fastener rows and several fasteners per row. On the other hand there are many simple and much cheaper joints
which only contain one or two fasteners. The main requirement for a standard specimen is that it should produce a rating of
fastener systems. It was considered by the working group that the simple joints were capable of performing this task. The
complex joints which more accurately represent the lateral stress gradients and load transfer distributions between fastener
rows were considered to be unnecessary. The working group also considered that if complex joints were defined as standard
designs, most researchers would not use them because of their cost. This would defeat the main aim of the exercise.
2.2 High secondary bending programme
The structural feature of main interest in this part of the exercise is a single shear chord-wise wing joint. The specimen
design may be relevant to other single shear connections such as lap joints in a pressure cabin, but the essential features of a
chord-wise wing joint must be represented in the specimen design. The main features to be modelled are the amount of load
transferred (LT) by the fastener and the amount of secondary bending (SB) of the joint. There is currently very little data
available on these values in real structure. However LBF reported in 1974(4) that some 60% of aircraft joints studied had an
SB ratio of 0.1 0.4,16% had a ratio from 0.4 0.8 and a further 15% in the range 0.81.4. The range of SB values for wing
skin attachments was 0 0.4. Specimens with SB values of 0.1 or less are considered in Part 1 of this exercise (section 2.1).
Joints with very high SB(> 0.8) are not considered in this programme. In view of the results of the FRFS programme, it is felt that
there is no requirement for a standard fatigue test specimen for fastener evaluation with such a high SB ratio since lifeenhancing fastener systems did not produce significant life improvements. Emphasis in part 2 of this exercise is therefore

placed on joints with an SB ratio in the range 0.2 to 0.5. The load transferred by the fastener has been shown to be of lesser
importance than the amount of secondary bending (2). However in a chord-wise wing joint the load transferred by a fastener is
likely to be significant. LT values in the range 20% 50% are therefore considered in the exercise.
The LT and SB values discussed in the above paragraph are not absolute values but depend on loading conditions.
Similarly in a laboratory joint these values will depend on the load applied, load sequence and load history already applied. The
LT and SB values however will be predominantly dependent upon specimen geometry and fastener flexibility and fit. The
experimental joints are broadly sub-divided into two groups, those in which LT and SB are significantly altered by the fastener
fit and flexibility (fastener-dominated joints) and those which are not (geometry-dominated joints). In aircraft structure it is not
clear which class of joint predominates, but current opinion is that geometry-dominated joints are more common among those
where fatigue may be critical.
In view of the load transfer and secondary bending considerations detailed above the high secondary bending phase of the
programme is quite complex. There are a number of joints under consideration which must be assessed in a number of ways.
Firstly it must be confirmed by measurement that the LT and SB requirements are fulfilled i.e. the average SB ratio is in the
range 0.2 to 0.5 and the average LT is in the range 20% 50% over a range of applied loads and a range of fastener
installations. These criteria must apply when the joint is in a stabilised' condition i.e. after a period of loading when movement
in the joint has stabilised. From these measurements, with a range of fastener installations, we can identify whether particular
joints are fastener or geometry dominated. Characteristic values of LT and SB can also be assigned to each joint. These
characteristic values must be considered in conjunction with the results of a fatigue testing programme. As discussed in the
previous section a fatigue testing programme is required to establish if all of the joints considered, rate fatigue resistant fastener
systems in the same way.
From the results of the LT and SB measurements and from the fatigue test programme, a number of joints must then be
selected as standard specimens. In order to determine which joint or joints should be defined as standards, a selection
procedure was defined by the working group. This selection procedure starts by considering whether all of the joints yield
similar results in both ranking and fatigue rating, with the use of fatigue resistant fastener systems. It may be that a number, but
not all of the specimens produce similar results. If this is the case then one joint may be selected from this common group.
Further considerations should be made of the remaining joints to assess their importance. It is possible that the fastenerdominated joints will yield different results to the geometry-dominated joints, in which case one joint from each group must be
selected. It may also bc the case that the value of SB may overshadow any other factors in determining the relative life
improvements. In this case it would be preferable to select one joint which could produce different values of SB by simple
geometric changes.

3.

STANDARD SPECIMEN DESIGNS

As discussed in the introduction (section 1), the scope of the programme was to look in more detail at the joints used in the
FRFS programme. A large number of specimens with no or low secondary bending were used which exhibit various amounts of
load transfer. In order to consider which joints should be evaluated in detail for this programme, specimens were sub-divided
into three groups: no load transfer, low load transfer and high load transfer. Specimen geometries considered and those chosen
for evaluation in the three categories, are described in sections 3.1,3.2 and 3.3 respectively.
In contrast however few laboratory joints were tested in the FRFS programme which contained secondary bending in the
range 0.2 to 0.5 and were relatively simple. The joints considered in detail for this programme are described in section 3.4. The
plate materials and hole preparation procedures for both parts of this programme are described in section 3.5.
3.1 No load transfer specimens
Two designs of no load transfer specimen were considered, both of which were tested in the FRFS programme, one design
was from France and the other from Sweden. The French design was chosen for inclusion at the AGARD FRFS meeting in San
Antonio and is shown in Fig 1. The most striking feature of the specimen is the offset fastener hole resulting in different stress
gradients on either side of the hole. This represents the end fastener in a row where the stress gradient is asymmetric. The
overall fastener load transfer is zero, though frictional load transfer may occur through the sideplate. The secondary bending is
considered to be negligible. The specimen consists of a dogbone with a small non-load carrying element attached via the
fastener. The small element and the dogbone undergo the same hole preparations and surface treatments, as described in
section 3.5.
3.2 Low load transfer specimens
Two designs of low load transfer specimens were considered, both of which were tested in the FRFS programme. Both
designs are reverse double dogbone specimens, one previously used by AGARD in the critically loaded hole technology
programme and one developed in the UK. They are similar in concept and geometry, but the UK joint is significantly smaller.
The UK joint (Fig 2) was selected for the following reasons.
1.

Buckling problems associated with the AGARD joint anti buckling guides may be necessary. These are undesirable in
that if the bending is constrained, the LT and SB of the joint will be altered.

2.

Cost.

3.

Data was soon to be available on the UK joint using the same material, fastener systems, surface and hole preparations as
the high load transfer joint.

Measurements of LT and SB on the AGARD joint were made in the FRFS programme(2). It was shown that each of the
two fasteners transfer about 5% of the load. The secondary bending was measured and found to be in the range 0.1 to 0.25.
3.3 High load transfer specimens
Eight different designs of double shear medium or high load transfer joints were tested in the FRFS programme, each joint
containing between two and sixteen fasteners. Complex joints with multiple fastener rows and multiple fasteners in each row
were rejected for the reasons described in section 2.1, namely that the important requirement for a standard joint is its ability to
rate fastener systems and it was considered that a simple two fastener joint was adequate for this purpose. A load transfer of 30
50% in the test section is the main requirement. The selection was thus based on simplicity and cost. Test data on the UK
joint, shown in Fig 3, was already available from the FRFS programme and it was selected for comparison with the no and low
load transfer joints.
3.4 Specimen designs with high secondary bending
Four specimen designs were reviewed, though only three were considered for use as standard joints. The reason for
including test data on the fourth is for comparison, since there is little data available on joints which meet the LT and SB
requirements of this programme. The Swedish X' joint shown in Fig 4 is the joint which was considered unsuitable as a
standard. The joint contains 16 fasteners and has very high lateral stress gradients both of which are undesirable features in a
standard joint. It was used in the AGARD 'Fatigue Rated Fastener System' (FRFS) programme and the results presented here
were obtained as part of that programme. Fig 5 shows the commonly used 11/2 dogbone specimen which was also tested as
part of the FRFS programme by the Netherlands and USA. Fig 6 shows the UK Q-joint which is a modified version of the joint
used in the FRFS programme having 1 /4" diameter fasteners in the controlling section. Fig 7 shows the detailed design of the
Swedish U-joint which was not tested in the FRFS programme. This joint is a derivative of the X-joint used in the FRFS
programme and was specifically designed for this investigation. It is essentially a single column X-joint but with a U-channel
splice plale instead of the flat plate used in the X-joint construction. Two and four column U-joints have been used successfully
in the past, but this is the first assessment of the single column variant.
3.5 Specimen requirements
Joints in Part 1 of the exercise were manufactured from a common batch of 7010 T7651 material. Joints in Part 2 were
manufactured from 7050 T76 material from the same batch as that used in the FRFS programme. The chemical composition
and mechanical properties of both materials are given in Table 2. Holes were produced by the general procedure:-pilot drill,
drill, ream, cold-work, ream, deburr, measure hole diameter, countersink. Variations to this procedure for individual fastener
systems are given in sections 4.1 to 4.4. All specimens were wet assembled using PR-1422 for Part 1 specimens and PR-1431G for Part 2 specimens.
4.

FASTENER SYSTEMS

In order to assess whether the joints described in the last section rate fastener systems in the same way, they must be tested
with a range of fastener systems. Fatigue resistant fastener systems rely on one, or a combination of two or three mechanisms.
These are clamping, interference fit and cold-working. The fastener systems chosen to assess the joints must therefore cover a
range of combinations of these parameters which are typically used in practice. Accordingly four cases were chosen for each
part of the exercise which cover this range. The four cases are described below and are based on the systems used in the FRFS
programme. Systems 1A and 2A arc identical to FRFS-A and systems 1B and 2B are identical to FRFS-B. Systems 1C and 2C
are similar to FRFS-C, which specified an interference fit of 90 10 um.
Part 1 No or low secondary bending
COLD-WORKED

FASTENER

1A

NO

HI-LOK

FIT
Clearance
20 10 (im

IB

YES

HI-LOK

Interference
25 10 urn

IC

NO

HI-TIGUE

Interference
110 10um

ID

YES

HUCK-EXL

Interference
120 lOfim

To check these fits, measurements of hole and fastener diameters were made for each joint and are summarised in Annex
1. The four fastener systems chosen for conditions 1A to ID were HI-LOK, HI-LOK in BOEING CX cold-worked hole, HITIGUE and HUCK EXL respectively.

Part 2

High secondary bending

COLD-WORKED

FASTENER

FIT

2A

NO

HI-LOK

2B

YES

HI-LOK

Clearance
20 10 um
Interference
25 10 mn

2C

NO

HI-TIGUE

2D

YES

HI-TIGUE

Interference
70 10 tun
Interference
70 10 urn

To check thesefits,measurements of hole and fastener diameters were made for each joint and are summarised in Annex
1. The four fastener systems chosen for conditions 2 A to 2D were HI-LOK, HI-LOK in BOEING CX cold-worked hole, HITIGUE and HI-TIGUE in BOEING CX cold-worked hole.
Details of each fastener system and hole preparations are described in sections 4.1 to 4.4. Sketches of the fastener systems
are shown in Fig 8.
4.1 HI-LOK in plain hole (1A and 2A)
The HI-LOK fastener can be installed with light clearance or interference fits. It is available in steel and titanium with a
variety of coatings. HI-LOKS used in this investigation were steel, 6.35mm dia. pins installed with a light clearance fit and
assembled with shear-off type collars, or K-fast nuts. The pin and collar part numbers are given in Annex 2.
4.2 HI-LOK in cold-worked hole (1B and 2B).
The BOEING CX split sleeve process cold-expands the fastener holes prior to assembly. A mandrel is inserted through
the fastener hole and a split sleeve passed over the mandrel, into the fastener hole. The mandrel is then pulled through the
sleeve using a compressed air powered puller. The sleeve is discarded and the hole reamed to size. Specimens are then
deburred and countersinks drilled. The HI-LOK fastener (as described in 4.1) is then installed with a light interference fit and
assembled with a shear-off type collar, or K-fast nut.
The cold-working was carried out using F.T.I, standard tooling to the 8-0-N specification for Part 1 (the no or low
secondary bending joints), and to the 6-3-N specification for Part 2 (the high secondary bending joints).
4.3 HI-TIGUE in plain hole (IC and 2C) and HI-TIGUE in cold-worked hole (2D)
The HI-TIGUE fastener is an interference fit fastener. The pin has conventional parallel sides of larger diameter than the
hole, but has a small lubricated bead at the threaded end which expands the hole as it is assembled, allowing the parallel pin to
be drawn through the hole, resulting in an interference fit. The pin must be drawn through the hole using arivetgun and then the
nut assembled and torque tightened to 10.2 11.3Nm.
The hole diameters required to give the fits described in section 4 are presented in Table 3. For the case of HI-TIGUE in
cold-worked hole (2D), cold expansion was carried out using F.T.I, standard tooling to the 8-0-N speicfication using a common
mandrel supplied by FOKKER (Q-joints and 1 1/2 dogbone only). A final reamer was also supplied by FOKKER to give the
required fits in the Q-joints and 1 1/2 dogbone specimens. The pin and collar part numbers are given in Annex 2.
4.4 HUCKEXL(ID)
This fastener system combines all three fatigue life improvement mechanisms. It is a two part fastener pin, the first part
cold working the hole as it is drawn through, the second part being a parallel sided pin which when installed produces an
interference fit. A collar is placed over the interference fit pin and swaged into locking grooves, whilst the cold-working part is
gripped and pulled until it separates from the installed pin. The complete operation is carried out using a special HUCK pulling
tool.
The part numbers of the pins and collars are given in Annex 2. It should be noted that this fastener type was not available in
the -6 length, consequently for the no load transfer specimen, the -8 length was used in conjunction with a thicker (7mm)
washer.
5

TESTING PROGRAMMES

For both parts of the programme fatigue tests were carried out using the FALSTAFF loading sequence. Five fatigue tests
per condition were performed at each of two stress levels. The details of the testing for Part 1 and Part 2 of the exercise are given
in sections 5.1 and 5.2 respectively.
5.1 No or low secondary bending test programme
The fatigue tests were carried out at two sites; the no load transfer specimens were tested at CEAT, France, and the low
and high load transfer specimens were tested at British Aerospace Woodford, UK. The specimen blanks were all manufactured
by Cleveland Guest, UK and assembled at British Aerospace, Warton. The stress levels chosen were 280MPa and 350MPa on

the net section for the peak FALSTAFF level. The high load transfer specimens were tested at 280MPa and 375MPa, the life at
350MPa being calculated assuming a linear relationship between log stress and log life.
Tests were carried out using servo-hydraulic fatigue machines, these were:UK

Mayes lOOkN capacity.

France CE AT 1 OOkN capacity.

Testing on the UK machines was carried out at a mean cyclic frequency of 11 Hz which gives a frequency of 1.8Hz for the
maximum load excursion. Testing on the French machines was carried out at a mean cyclic frequency of 12Hz, giving a
frequency of 2Hz for the maximum load excursion.
5.2 High secondary bending test programme
As described earlier, the information required from the test programme is twofold; the load transfer and secondary
bending characteristics of the joints and the fatigue endurances. A number of specimens were strain gauged and measurements
of LT and SB made using the procedures described in Annex 3.
The fatigue test stress levels varied from joint to joint, depending on the amount of secondary bending. The proposed two
stress levels are defined as those levels which produced fatigue endurances of 5000 and 15000 FALSTAFF flights with the
datum (clearance fit HI-LOK) fastener installed. These stress levels were not universally used in the programme. The net
section and gross section stress levels for the peak applied FALSTAFF load used by each participant are given below:-

Net section

Gross sectin

U K - Q joint

280MPa and 350 MPa

210MPa and 263MPa

N L / I - 1 1/2 dogbone

268MPA and 335MPa

200MPa and 250 PMa

S - X joints

200MPa and 267MPa

150MPa and 200 MPa

S - U joint

276MPa and 345MPa

200MPa and 250 MPa

RESULTS AND DISCUSSIONS

The results and discussions of the two parts of the programme are presented in this section. The no or low secondary
bending part of the exercise is discussed in section 6.1 and the high secondary bending part in section 6.2.
6.1 Results and discussions of the no or low secondary bending tests
The fatigue test results from the no load, low load and high load transfer joints are given in Tables 4, 5 and 6 respectively.
The relative life improvements over the datum system (HI-LOK fastener in a clearance fit hole) are given in Table 7. The life
improvement factors are based on the log mean endurances for each test condition. In view of a number of problems involved
with the testing of the no load transfer joint, these results are discussed separately. Section 6.1.1 discusses the results of the low
and high load transfer joints, and section 6.1.2 the no load transfer joint.
6.1.1 Low and high load transfer joints
As discussed earlier in section 2.1, the joints have to be assessed in terms of life improvement factors. As can be seen from
Table 7, all of the fastener systems are rated in a similar way by the two joints. The life improvement factors are consistently
higher at the higher stress level, markedly so for the HUCK and HI-TIGUE fasteners. This is perhaps to be expected with
interference fit fasteners, the magnitude of any beneficial compressive residual stress being controlled by the peak of the
applied loading. This is not expected however for the case of cold-working, where more benefit is expected at lower stresses.
The ranking of the fastener systems is consistent for the two joints considered and is summarised below:Life improvement ratios, based on log mean lives.
1

3 4

5 6

7 8

10

r '
i

12

.1

'

HI-LOK
HI-TIGUE
I
I

11

FASTENER
SYSTEM
LLT
FAST
HI
HLT
J
i

n
HI-LOK AND C. W.
I

I
I
I

HUCK EXL C. W.
1

The absolute fatigue lives of each joint are also very similar under the same test conditions. Log mean lives of each
specimen type are within 35% of each other under identical test conditions. It was noted however subsequent to testing that the
low load transfer joints had been assembled and tested with titanium fasteners instead of steel. The test data are plotted in Figs 9
and 10 for the low and high stress levels respectively. In view of the high scatter in the data, comparing the life improvement
ratios on a log mean basis is not considered sufficient, as it may give a false impression of expected life improvements. The effect
of scatter on life improvement ratios is therefore discussed in Annex 4. The effect of fastener type on fatigue crack origins must
also be considered and detailed failure sites are given in Annex 5. General observations are also made on the effect of fastener
type in the same annex.

6.1.2 No load transfer joints

The no load transfer joints showed scratches and surface dents received in transit from the UK to the testing laboratories
in France. These damaged areas proved to be the initiation sites in over one third of the specimens. Of the remaining specimens,
one half of the failures initiated in the test section and one half initiated in the dogbone radius. The fatigue test results are given
in Table 4, from which a number of observations can be made. Failures from the dogbone radius are independent of the
fastener system installed and hence the fatigue lives are also independent. Failures at the high stress level initiating from surface
scratches produce fatigue lives lower than those failing from the dogbone radius. This is not the case at the lower stress level. In
view of the fact that one third of the specimens failed from no apparent defect in the dogbone radius it must be concluded that
this specimen is not suitable for fastener evaluation. The stress concentrating effect of the hole is readily overcome by fatigue
resistant fastener systems. Fatigue failures therefore occur at an alternative site of stress concentration. A general conclusion
can therefore be drawn, that a stress concentration greater than that of an open hole is required in a standard joint for fastener
evaluation. The greater concentration can be simply achieved by using a joint in which some load is transferred by the fastener
system. Other designs of no load transfer joint have been successfully used for fastener evaluation(2) but have not been tested
with such extremes of interference fit and cold-working.
6.2 Results and discussion of the high secondary bending tests
A change in design of the UK Q-joint has meant that the test results have been obtained on two different designs of joint.
An assessment of the importance of this change is made in Annex 6. The results of the LT and SB measurements on each joint
are discussed in section 6.2.1. The fatigue test results are discussed in section 6.2.2.
6.2.1 Load transfer and secondary bending results
The results of the load transfer and secondary bending measurements on each of the joints are presented and assessed in
this section. The results of the O-joint, 1 1 / 2 dogbone, U-joint and X-joint are presented in subsections a, b, c and d
respectively. Both the LT and SB measurements vary to some degree with applied load. The values of SB at the peak applied
load is important in determining if residual stresses are formed (or existing residual stress fields are modified) and if so their
resulting magnitude around the fastener holes. This has a significant effect upon the damage done by the ensuing load cycles.
Most fatigue damage however is done by the relatively lower load cycles, typically the maximum damage occurring at a stress
range of about 1 /3 of the level 32 peak stress. The SB values at the damaging 1 /3rd peak stress level are also calculated as a % of
the SB values at the peak load, i.e. if the SB value for a joint is 0.5 and the 1 /3rd ratio is 80% then the peak SB is 0.5 and the SB at
l/3rd FALSTAFF peak load is 0.4.
a)

O-joint measurements
Measurements have been made on both variants of the O-joint (3/16" and 1/4" fasteners in the controlling section) with
HTLOK fasteners installed in plain and cold worked holes. The results are presented in Tables 8 and 9 for the 1/4" and
3/16" fasteners respectively. Measurements of LT and SB with HI-TIGUE fasteners installed in plain and cold worked
holes are presented in Table 10. A comparison of the results with different diameters of HI-LOK fasteners is made in
Annex 6.
For each set of test data it is apparent that cold working does not significantly affect the SB ratio but does affect the LT.
The effect of fastener fit however is quite marked. Comparing Tables 9 and 10, it can be seen that high interference fit
fasteners (2C and 2D) produce lower LT and SB values in the test section than the light clearance/light interference fit
fasteners (2 A and 2B). This variation in values however is quite small compared with other fastener-dominated joints (e.g.
1 1/2 dogbone) The O-joint is therefore classed as a fastener-dominated joint, but with a low fastener dominance. A
summary of these measurements is presented in Fig 11.
The effect of applied load level on the LT and SB values is similar with any of the fastener systems installed, the 1 /3rd load
ratio being between 63 73% for the four fastener systems.

b)

11/2 dogbone measurements

Measurements have been made on the 11/2 dogbone specimen with HI-LOK and HI-TIGUE fasteners installed in plain
and cold worked holes. The results are presented in Tables 11 and 12 respectively. Values of load transfer vary little with
either applied load or fastener fit. Load transfer values at peak applied load vary only from 24% to 31% for the four
fastener systems, LT increasing with fastener interference. The secondary bending ratio however varies both with applied
load and fastener fit. The variation of secondary bending with applied load shows a reversal of the bending direction with
both the HI-LOK fastener installations (2A and 2B). In the HI-LOK in a plain hole case, the rate of change of SB ratio with
applied load is quite extreme. With HI-TIGUE fasteners installed, however, very little variation of secondary bending
with applied load is found, the l/3rd load ratio being about 80% for both plain and cold worked holes. Comparing the SB
ratios for the four fastener installations shows a large dependence on the fastener system. The 11/2 dogbone specimen is
therefore classified as a fastener dominated joint. A summary of the measurements at peak FALSTAFF load is presented
in Fig 12.

c)

U-joint measurements
The results of secondary bending measurements on the U-joint with HI-LOK and HTTIGUE fasteners in plain holes are
given in Table 13. Load transfer measurements were made on this specimen but evaluation of the results WEIS not realistic
with so few strain gauges; the LT was assumed to be 50 55%. The SB measurements were made with and without midside supports. A considerable difference in SB values was observed when comparing the results with and without support.
A peak value of 0.3 without support and 0.4 with support was measured with HI-LOK fasteners installed and 0.26 and
0.34 respectively with HI-TIGUE fasteners installed. SB values vary little with applied load, the l/3rd load ratio varying

between 82% and 100% for the four cases. The variation of SB values with fastener fit is similar to that found in the Qjoint. The reduction in SB values in the U-joint with increasing fit is 13% to 16% compared with a reduction of 20% to
27% in the O-joint.
d)

X-joint measurements
Measurements of SB have been made on this joint, with the datum fastener system installed. The results of the SB
measurements give values between 0.4 and 0.8 (depending on location) at the peak applied load. The average value across
the test section is 0.61. Load transfer measurements were not evaluated as too few strain gauges were used to give realistic
results.

6.2.2 Fatigue test results

The fatigue test results are presented in Tables 14 to 17 for the O-joint, 1 1 /2 dogbone, X-joint and U-joint respectively
and discussed in sections a) to d) respectively.
a)

O-joint endurances
Fatigue testing of the Q-joint, as discussed earlier, has been performed with two variants, one with 3/16" diameter and
one with 1/4" diameter fasteners in the controlling section. The HI-LOK fasteners in plain and cold worked holes (2A and
2B) were tested with the smaller fasteners in the controlling section. The HI-TIGUE fasteners in plain and cold worked
holes (2C and 2D) were tested with the larger fasteners in the controlling section.
Comparing first the results for HI-LOK fasteners in plain and cold-worked holes (see Table 14), the expected benefit
produced by cold-working alone is offset by the relatively high secondary bending in the joint. The resultant stresses at the
interface of the joint were some 40% higher than the axial stresses applied. The high stress levels in the FALSTAFF
sequence thus quickly wipe out the beneficial compressive residual stresses induced by the cold working process.
In contrast, however, the interference fit HI-TIGUE fasteners provide considerable improvement in fatigue performance.
The reason for this benefit is threefold. Firstly the fretting damage is considerably reduced due to the lower relative slip
caused by the interferencefit.Secondly a compressive beneficial residual stress is formed by the application of the highest
load in the FALSTAFF sequence. The secondary bending causes the surface stress at the joint interface to be 40% higher
than the axially applied stress. Thus a large compressive residual stress is formed at the interface on unloading. Thirdly the
stress concentration factor is considerably reduced by the high interference, thus reducing the damaging effect of
subsequent alternating loading . The combination of these three effects causes a considerable increase in the fatigue
endurance. The combination of an interference fit fastener and a cold worked hole appears to give an even greater
enhancement in fatigue endurance. This enhancement however does change the failure mode of the joint. Three of the
specimens tested with HI-TIGUE fasteners in cold worked holes, failed away from the test section, one at the edge of the
gripped area, one at the controlling section, and one at the end of the lap plate. All of the remaining specimens with cold
work and interference fit had at least one crack initiating due to fretting in the test section, but not from the fastener holes.
It must be concluded that the limit of life enhancement of this joint is being approached as failure is beginning to occur
remote from the fasteners. However the joint is considered to be capable of rating most fastener systems under conditions
of high secondary bending.

b)

1 1/2 dogbone endurances

Testing of the 1 1 /2 dogbone specimen has been undertaken at two sites. Testing with HI-LOK fasteners installed in plain
and cold worked holes (2 A and 2B) was undertaken at NLR as part of the FRFS programme. Testing with HI-TIGUE
fasteners installed in plain and cold worked holes (2C and 2D) was undertaken at the University of Pisa. In view of the
possibility of high secondary bending stresses occuring in this specimen, it was required that the testing was performed
using fatigue test machines of a similar stiffness at both sites. To ensure that a change in test machine did not affect fatigue
lives, a number of check tests were undertaken at the University of Pisa. These results are presented in Table 15, along
with the other fatigue test results of the 1 1 /2 dogbone specimen.
The compatibility tests show that with clearance fit HI-LOK fasteners installed and tested at the higher stress level,
consistent data was produced at both sites. The 1 111 dogbone specimen is a fastener-dominated joint, so that the amount
of secondary bending is dependent on the fastener fit. The secondary bending in the cold-worked light interference fit
joint (2B) is some four times greater than that in the clearance fit joint (2A). The beneficial compressive residual stress
field formed by the cold working process is offset by the detrimental increase in local surface stress. As a consequence
there is no increase in fatigue endurance over the datum system using cold-working and light interferencefit.The data on
joints with HI-TIGUE fasteners in plain holes (2C) shows that no improvement in fatigue endurance is achieved with the
use of interference fit fasteners. The results of the fatigue tests with HI-TIGUE fasteners installed in cold-worked holes
(2D) also show no life improvement over the datum HI-LOK system. Both of these observations are attributed to the
increase in secondary bending of the joint with increased fastener fit. The SB ratio with high interference fit fasteners has
increased to a peak value of 0.45 compared with 0.22 for light interference cold-worked holes and less than 0.1 for
clearance fit holes. In fact, the increased SB ratio for the HI-TIGUE fastener in a plain hole (2C), causes a reduction in
fatigue life compared with the datum HI-LOK system. Cold working of the hole before installation of a HI-TIGUE
fastener (2D), however, causes an increase in fatigue life over the HI-TIGUE in a plain hole (2C). This increase only
brings the fatigue life of the HI-TIGUE fastener in a cold worked hole back to that of the datum system. It is considered
that from these results, the 1 1/2 dogbone specimen is not suitable for the evaluation of fatigue resistant fastener systems.

c)

X-joint endurances
Fatigue testing of the X-joint was undertaken as part of the FRFS programme. The results obtained are presented in Table
16. The beneficial effect of cold working is clearly demonstrated in this joint, with life improvement factors of about two
being obtained at both stress levels. It was noted however that cold worked specimens tested at the low stress level all
failed in the splice plate; life improvement factors are therefore not relevant. Measurements of load transfer and
secondary bending have not yet been made with both fastener systems installed. It is not possible therefore to assess the
importance of these results. From the measurements on the datum system however it is surprising to find that cold
working doubles the fatigue endurance despite the high value of secondary bending (0.6 with the datum system).

d)

U-joint endurances
The fatigue test results of the U-joint specimen are presented in Table 17. Results were obtained at two stress levels with
both the HI-LOK and HI-TIGUE fasteners installed in plain holes. Three failures with HI-LOK fasteners installed,
occurred in the splice plate. All of the failures with HI-TIGUE fasteners installed, occurred in the splice plate. One non
splice plate failure occurred with HI-TIGUE fasteners installed, but this was when the splice plate was of a different
material (7010-T73651). All other splice plates were made from 7075-T6 material. It was decided to abort further testing
of the U-joint specimen, on the assumption that most failures would occur in the splice plate. It was therefore concluded
that the U-joint specimen was not a suitable design for fastener evaluation. It should be noted however that specimens with
two or four U-channels fastened to the base plate have been used successfully in the past. A future development of the
single column U-joint will probably involve thickening the U-channel web from 5 to 6mm.

CONCLUSIONS AND RECOMMENDATIONS OF PART 1, NO OR LOW SECONDARY BENDING

1)

The low and high load transfer joints tested in this programme under FALSTAFF loading both produced similar fatigue
lives under the same test conditions.

2)

The ranking of all four fastener systems was the same in the low and high load transfer joints.

3)

The low and high load transfer joints are considered equivalent in rating fatigue resistant fastener systems.

4)

The no load transfer joint is not recommended as a standard joint for fastener evaluation.

CONCLUSIONS AND RECOMMENDATIONS OF PART 2, HIGH SECONDARY BENDING

This investigation has shown the difficulty in designing a joint for fastener evaluation which exhibits typical amounts of
load transfer and secondary bending found in practice. All of the joints tested produced a number of failures originating away
from the test section under certain conditions. Failure sites are detailed in Annex 5 and are summarised below.
Failure away from test section using
fastener system
2A

2B

2C

Q joint

1 1/2 dogbone

Xjoint
Ujoint

2D

Clearly the U-joint would be rejected as not being able to rate the simpler fastener systems in the designated test section
where the load transfer and secondary bending are being controlled. This was also the case for the X-joint. The 1 1/2 dogbone
specimens exhibit a percentage of failures away from the test section with most fastener systems. The O-joint behaves similarly
when high interference fit fasteners are installed. It is considered that these joints are therefore approaching their limits for
assessing fatigue rated fastener system.
The 11/2 dogbone specimen has been shown to be a fastener-dominated joint in which the secondary bending in the test
section is highly influenced by the fastenerfit.The expected increase in fatigue life with an increasing degree of interferencefitis
offset by an increase in secondary bending ratio. In fact a slight reduction in fatigue life was obtained with high interference fit
fasteners when compared to clearance fit fasteners. In a real structure it is anticipated that the secondary bending ratio at any
location will to some degree be dependent on the fastener fit, but will be predominantly dependent on the local geometry. The 1
1/2 dogbone specimen is rejected as not being able to rate fastener systems in a useful way.
The Q-joint is the only specimen which consistently produces failures in the test section. The values of load transfer and
secondary bending are predominantly governed by the geometry of the joint and to a lesser degree by the fastener fit. Of the
joints considered in this exercise, the Q-joint is recommended as the most suitable for fastener evaluation purposes.

REFERENCES
T.COOMBE and R.B.URZI. Critically Loaded Hole Technology Pilot Collaborative Test Programme. Final Technical
Report, AGARD Report No. 678, (November 1980)
H.H. van der LINDEN. Fatigue Rated Fastener Systems An AGARD Coordinated Testing Programme. AGARD
Report 721 (November 1985)
3.

R.COOK and H.H. van der LINDEN. Standard Fatigue Test Specimens for Fastener Evaluation. Paper presented at
AGARD meeting, SAN ANTONIO, (April 1985)

4.

D.SCHUETZ and H.LOWAK. The effect of secondary bending on the fatigue strength of joints. RAE Library
Translation No 1858,(1974)

10

TABLE 1
PARTICIPANTS IN THE STANDARD SPECIMENS PROGRAMME
Country

Participants

France

Centre D'essais Aeronautique

de Toulouse - CEAT

I P A Liberge

Italy

University of Pisa

G Cavallini

The Netherlands

National Aerospace
Laboratory -NLR

H H van der Linden

Sweden

Saab-Scania

L Jarfall

Sweden

Flygtekniska Forsoksanstalten
- FFA

B Palmberg

United Kingdom

Royal Aircraft Establishment

R Cook

11
TABLE 2
A.

TYPICAL CHEMICAL COMPOSITION AND MECHANICAL PROPERTIES

OF 7010-T7651

CHEMICAL COMPOSITION - UNCLAD

Cu

Mg

% Min

1.5

2.2

% Max

2.0

2.7

Si

Fe

Nl

Mn

Zn

Pb

Sn

Ti

Zr

5.7
0.1

0.15

0.3

0.11

6.7

0.05

Cr

0,05

0.05

0.17

0.05

0.05

Remainder Al
MECHANICAL PROPERTIES - MINIMUM REQUIREMENTS - L DIRECTION
Tensile strength

530 MPa

0.2% Proof Stress

450 MPa

Elongation %
gauge length 50 mm

B.

TYPICAL CHEMICAL COMPOSITION AND MECHANICAL PROPERTIES OF 7050-T76

CHEMICAL COMPOSITION - UNCLAD

Cu

Mg

Si

Fe

Mn

Zn

Ti

Zr

Cr

% Min

2.00

1.9

5.70

0.08

X Max

2.60

2.6

0.12

0.15

0.10

6.70

0.06

0.15

0.04

Remainder Al

MECHANICAL PROPERTIES - L DIRECTION

Min

Max

Tensile strength
(MPa)

573

592

0.2% Proof stress

(MPa)

521

552

Elongation %
Gauge length 50 mm

12

12.5

12

TABLE 3
Hole Dia
(Remed)

CSK Dia

Hole Dia
(After CW)

Remarks

1A Hi-Lok in
Plain Hole

6.35
6.37

9.96
10.06

Torque tighten
to 6.8 - 9.1 Nm

2A Hi-Lok in
Plain Hole

6.35
6.37

9.96
10.06

Record Value at
which collar
shears off

IB Hi-Lok in
CW Hole

5.97
6.04

9.96
10.06

6.30
6.32

Torque tighten
to 6.8 - 9.1 Nm

2B Hi-Lok in
CW Hole

5.71
5.79

9.96
10.06

6.30
6.32

Record value at
which collar shears
off

IC & 2C Hi-Tigue
in Plain Hole

6.21
6.25

9.83
9.93

ID Huck EXL

6.045
6.17

2D Hi-Tigue in
CW Hole

5.97
6.04

Produced by Huck
Installation Tools
9.83
9.93

Using Reamer
supplied

Torque tighten to
10.2 - 11.3 Nm
Automatic clamping
by Swaging Collar
Torque tighten to
10.2 - 11.3 Nra

13

TABLE 4
FATIGUE LIVES OF NO LOAD TRANSFER SPECIMENS

INITIATION SITES AND FALSTAFF' FLIGHTS

1
.0 FAILURE AND
| LOG MEAN VALUES AT PEAK APPLIED NET .ECTION STRESS (MPa)
FASTENER SYSTEM
TEST SECTION
280

1A HI-LOK IN
PLAIN HOLE

IB HI-LOK IN
COLD WORKED
HOLE

SURFACE DEFECT

350

280

350

280

350

22232
26173

89525
112973
115773

23140

63240

28081
23032

24122 |

105400

23140

63240 |

25431 |

77632
129281

37973

69232
94325
68831

16481
14126
17344

1100182 1

37973 I

76602 1

15924 |

32173

19432

96360

18930

32173

119432 '

ID HUCK EXL IN
COLD WORKED HOLE

| 63538

20240
22032
21730
21319 |

96360
28373
61373

I 41729 1

38730
22360

I 29428 !
24222
29530
31200

18930 '

87130
74232
65032
57530
| 70137 |

19190

65646

26631

19190

65646 |

IC HI-TIGUE IN
PLAIN HOLE

62773
46330
88200

DOGBONE RADIUS

28154 1

| 26631

14

TABLE 5
FATIGUE LIVES OF LOW LOAD TRANSFER SPECIMEN

FASTENER SYSTEM

1A HI-LOK IN
PLAIN HOLE

IB HI-LOK IN
COLD WORKED
HOLE

FALSTAFF FLIGHTS TO FAILURE AND

LOG MEAN VALUES AT
PEAK APPLIED NET SECTION STRESS
280 MPa
350 MPa
15372
6031
7772
14329
5424
20172
22680
9239
16572
8572
| 17560

| 7257

49231
118194
203993
115231
99031

18759
61325
46586
46796
83591

I 106257

IC HI-TIGUE IN
PLAIN HOLE

46163 i

55351
54772
53172
42172
74772

17080
17729
22959
21572
23372

55109 !

20368 J

115021
97031
79972
ID HUCK EXL IN
COLD WORKED HOLE
96281 |

66231
33431
47943
69611
52132

15

TABLE 6
FATIGUE LIVES' OF HIGH LOAD TRANSFER SPECIMEN

FASTENER SYSTEM

1A HI-LOK IN
PLAIN HOLE

FALS'FAFF FLIGHTS TO FAILURE AND

|LOG MEAN VALUES AT
PEAK APPLIED NET SECTION STRESS
280 MPa

375 MPa

37898
23172
30839
3492_14821

7572
4031
3559
3431
2959

1 26875

IB HI-LOK IN
COLD WORKED
HOLE

51172
76759
224420
166196
93021
L06386 |

IC HI-TIGUE IN
PLAIN HOLE

66624
32796
52172
123227
41024
56510

ID HUCK EXL IN
COLD WORKED HOLE

1 4060 |

350 MPa

6346 |

31759
34525
27772
27359
| 30212 !

40680 I

34224
27749
27031
12972
20924
23368 !

211711
155031
155172
96996
146572

87511
58880
55172
59929
52972

148580 |

61814

28786

76057 |

Estimated figure assuming that a log-log S-N curve is linear

in this region.

16

TABLE* 7
FATIGUE LIFE IMPROVEMENT FACTORS

FASTENER
SYSTEM

LLT

NLT
280 MPa

350 MPa

280 MEa
H7560I

1A HI-LOK IN
PLAIN HOLE

IB HI-LOK IN
CW HOLE

HLT
350 MPa
72571
1

280 MPa
(268751
1

350 MPa
6346)
1

(106257

46163]

1106386

406801

6.05

6.36

3.96

6.41

IC HI-TIGUE
IN PLAIN
HOLE

! 55109

ID HUCK EXL
IN COLD
WORKED HOLE

962811

3.14

5.48

J20368]

56510J

287861

2.81

2.10

4.54

52137]

148580

[76057]

7.18

5.53

11.99

17

TABLE 8
LT AND SB MEASUREMENTS - Q- JOINT U " DIA) HI-LOK
SPECIMEN TYPE: O-JOINT WITH %" DIA HI-LOK FASTENERS
System 2A Plain hole - clearance fit
System 2B Cold worked hole - light interference fit
MAX LOAD (kN):

67

MIN LOAD (kN): -20.5

2B

2A
% of the max load in
FALSTAFF
LT %

0
16.7
33.3
50
66.7
83.3
100
83.3
66.7
50
33.3
16.7
0
Minimum load
0

0
22.3
28.3
32.3
35.2
36.2
37.6
38.9
40.0
40.5
39.6
39.3
0
21.9
0

SB ratio

LT %

SB ratio

0
39.4
44.7
46.7
48.6
49.7
50.3
51.2
51.7
52.1
52.6
54.3
0
44.4
0

0
.131
.321
.457
.495
.498
.494
.473
.450
.417
.341
.218
0
.236
0

18

TABLE 9
LT AND SB MEASUREMENTS - Q-JOINT (3/16" DIA) HI-LOK
SPECIMEN TYPE:

O-JOINT WITH 3/16" DIA HI-LOK FASTENERS

System 2A Plain hole - clearance fit

System 2B Cold worked hole - light interference fit
MAX LOAD (kN): 67
MIN LOAD (kN): -20.5

2A
% of the max load in
FALSTAFF

0
16.7
33.3
50
66.7
83,3
100
83.3
66.7
50
33.3
16.7
0
Minimum load
0

2B

LT %

SB ratio

LT %

SB ratio

0
30.4
36.1
41.3
44.3
46.4
49.6
51.3
52.5
54.5
56.8
65.3
0
48.3
0

.192
.341
.350
.376
.405
.443
.406
.348
.318
.309
.336
0
.418
0

0
28.0
37,5
42.3
44.9
47,3
48.7
49.8
50.3
51.3
53.3
59.3
0
52.0
0

0
.264
.297
.353
.393
.422
.441
.402
.352
.302
.270
.419
0
.290
0

19

TABLE 10

LT AND SB MEASUREMENTS - Q-J0INT HI-TIGUE

SPECIMEN TYPE:
System 2C
System 2D

O-JOINT WITH Jf" DIA HI-TIGUE FASTENERS

Plain hole - high interference fit

Cold worked hole - high interference fit
MAX LOAD (kN):

67

MIN LOAD (kN): -20.5

2C

2D

% of the max load in

FALSTAFF

0
16.7
33.3
50
66.7
83.3
100
83.3
66.7
50
33.3
16.7
0
Minimum load
0

LT%

SB ratio

0
21.96
25.95
27.68
34.95
38.48
40.67
40.54
38.12
34.60
28.63
24.20
0
18.78
0

0
0.120
0.195
0.248
0.284
0.308
0.324
0.313
0.289
0.263
0.236
0.171
0
0.152
0

LT%
0
14,11
17.15
21.93
26.40
29.66
32.71
32.09
29.89
27.23
22.46
18.32
0
17.82
0

SB ratio
0
0.170
0.212
0.262
0.314
0.344
0.355
0.349
0.333
0.300
0.235
0.162
0
0.151
0

20

TABLE 11
LT AND SB MEASUREMENTS 1% DOGBONE, HI-LOK
SPECIMEN TYPE:
System 2A
System 2B

1% DOGBONE WITH

YA" DIA HI-LOK FASTENERS

Plain hole - clearance fit

Cold worked hole - light interference fit
MAX LOAD (kN):

60.5

MIN LOAD (kN): -16.2

2A

% of the max load in

FALSTAFF

0
16.7
33.3
50
66.7
83.3
100
83.3
66.7
50
33.3
16.7
0
Minimum load
0

2B

LT %

SB ratio

LT %

SB ratio

0
25.9
24.7
23.8
24.6
25.1
25.7
25.5
24.3
23.1
22.4
20.9
0
12.7
0

0
0.041
0.018
0.027
-0.005
-0.055
-0.095
-0.058
-0.028
0.001
0.026
0.026
0
0.054
0

0
22.0
22.1
22.1
22.2
23.0
23.8
23.3
22.9
23.1
23.6
24.9
0
19.5
0

0
-0.086
0.031
0.142
0.187
0.210
0.223
0.211
0.180
0.140
0.060
-0.023
0
-0.055
0

21

TABLE 12

1% DOGBONE, HI-TIG*UE

LT AND SB MEASUREMENTS -

SPECIMEN TYPE: 1V-, DOGBONE WITH JJ" DIA HI-TIGUE FASTENERS

System 2C
System 2D

Plain hole - high interference fit

Cold worked hole - high interference fit
MAX LOAD (kN): 60.5
MIN LOAD (kN): -16.2

2D

2C
% of the max load in
FALSTAFF

LT%

SB _atio

LT%

SB ratio

0
16.7
33.3
50
66.7
83.3
100
83.3
66.7
50
33,3
16.7
0
Minimum load
0

0
28.8
28.3
27,8
27.7
27.7
27.8
27.5
27.7
27.8
27.3
26.3
0
24.0
0

0
0.265
0.424
0.504
0.502
0.477
0.455
0.412
0.387
0.352
0.266
0.060
0
0.111
0

0
30.5
31.0
31.8
31.5
31.4
31.4
31.1
30.9
30.8
30.4
29.2
0
27.2
0

0
0.425
0.455
0.483
0.517
0.506
0.481
0.455
0.429
0.394
0.310
0.136
0
0.311
0

22

TABLE 13
SB MEASUREMENTS U-JOINT, HI-LOK AND HI-TIGUE

SPECIMEN TYPE: U-JOINT WITH \ INCH DIA FASTENERS

SYSTEM 2A HI-LOK IN PLAIN HOLE
SYSTEM 2C HI-TIGUE IN PLAIN HOLE
(a) NO SIDE SUPPORT AND (b) MID SIDE SUPPORT
1

System 2A

% of the max load

in FALSTAFF

Type (a)
SB ratio

System 2c

Type (b)
SB ratio

Type (a)
SB ratio

Type (b)
SB ratio

16.7

0.360

0.485

0.260

0.305

33.3

0.305

0.455

0.250

0.300

50

0.305

0.445

0.270

0.320

66.7

0.295

0.420

0.270

0.325

83.3

0.290

0.400

0.255

0.305

100

0.300

0.395

0.260

0.335

83.3

0.295

0.380

0.250

0.310

66.7

0.255

0.355

0.235

0.300

50

0.250

0.335

0.265

0.285

33.3

0.245

0.310

0.260

0,275

16.7

0.335

0.375

0.340

0.325

Minimum load

0.140

0.135

1
After 10,000 cycles

23

TABLE 14
FATIGUE LIVES OF Q-JOINTS

FALSTAFF Flights to Failure and

] Log Mean Value s at Peak
Applied Net Sect.ion Stress
Fastener System

|
280 MPa

2A HI-LOK IN
PLAIN HOLE

2B HI-LOK IN
COLD WORKED HOLE

12128
14431
12160
13831
14031
|l3280|

9631
12424
12329
16224
17631
|13337

350 MPa

3925
2929
3444
4336
|3639

3801
3172
3624
5323
|3905 |

2C HI-TIGUE IN
PLAIN HOLE

18530
78032
25225
30860
|32572f

21730
15825
13573
16385
11173*
8232
14448)

2D HI-TIGUE IN
COLD WORKED
HOLE

25573
77330
45632*
91225
111393*
| 56504|

38981
32695
16840*
11159
7997
(18364}

* FAILURE AWAY FROM TEST SECTION RESULT NOT INCLUDED IN LOG MEAN VALUE

24

TABLE 15
FATIGUE LIVES OF V/-. DOGBONE SPECIMENS

FALSTAFF Flightsi to Failure and

Log Mean Values at Peak
Applied Net Sect:ion Stress
FASTENER SYSTEM
268 MPa

2A HI-LOK IN
PLAIN HOLE

2B HI-LOK IN
COLD WORKED
HOLE

2C HI-TIGUE IN
PLAIN HOLE

2D HI-TIGUE IN
COLD WORKED
HOLE
2D HI-LOK IN
PLAIN HOLE
COMPATIBILITY
TESTS AT PISA

335 MPa

18411
60372*
56972
63831
|44893|

9559
15419
23373
22231
]16635|

29572
40431*
58231
35759*
39722J

13524*
14231
17962*
19172
16045]

30764
36572
|33542|

9983
16734*
14062*
11780
|12899|

37146
42640*
38265*
(392801

19007*
17308*
14946
14564*
jl6358|
11165
13826
19955
(145501

FAILURES INITIATING AWAY FROM TEST SECTION

25

TABLE 16
FATIGUE LIVES OF X-JOINTS

FALSTAFF Flights to Failure and

I Log Mean | Values at Peak
Applied Net Section Stress
FASTENER SYSTEM
200 MPa

2A HI-LOK IN
PLAIN HOLE

2B HI-LOK IN
COLD WORKED
HOLE

SPLICE PLATE FAILURE

267 MPa

13860*
16972
13180
13772
|14375(

5329
5590
6280
5425
|5646|

42772*
30224*
35631*
27630*
(33588

10929
11972
11372
6172
9789J

26

TABLE 17
FATIGUE LIVES OF U-JOINTS

FALSTAFF Flights to Failure and

I Log Mean Values at Peak
Applied Net Secti on Stress
FASTENER SYSTEM

2A HI-LOK IN
PLAIN HOLE

276 MPa

345 MPa

7831
19311*
10631*
23631
|l396lf

5558*
5729
5431
5571|

24372*
17834*
16031*
19572*
|19217|

15424**
13520*
13969*
|14282

2B HI-LOK IN
COLD WORKED
HOLE

2C HI-TIGUE IN
PLAIN HOLE

2D HI-TIGUE
IN COLD WORKED
HOLE

* SPLICE PLATE FAILURE (7075-T6)

SPLICE PLATE MADE FROM 7010 - T73651(base plate failure)

27

60

o
CM

CN
CM

O
CV.

^
&

Fig 1 France no load transfer joint

2S

h1

+
+

*1

Dowel
holes

Fig 2 Low load transfer joint

co

29

20

in
co

in
csi

to

in
CM

IT)

in

Dowel
holes

CO
CD

o
CO

Fig 3 High load transfer joint

mm

80

&

Ad

120

96 - 6 6

F2d

4- 4- 4 4- I
+ h 4- 44-

ti
4- 4-

Ad
Ad

4
Ad

Ad

1 +

4- 4-

72

eJtC

r=A0

41

2d

Fastener

362

1
2dT| [ . A d ^ A d * ^ A d |

, ^

i
[jd

Fig A Swedish 'X' joint

180

Section A-A
(countersinks in
base plale)

300

j_r18
in

in

185
Fig 5

I I dogbone specimen

32

80

-L_u_.

"='Ve

o
CO

L_._....7
/

25

i \

.C'sks
/ t M l
*/
section
eeM

u
c>1

CM

-j-

j\

-.-

C'sks
controlling
section

"

1
i

cn

J'
.

4Q

Fig 6

M'

UK *Q' joint

3...

25

3 jc

1A8

.Umim

/6.3
-n-

12.5

38

12.5

60

12.5
27

25

27

25

12.5

Fig 7 Swedish 'U' joint

L_f

Ul

Hi-Lok fastener with flush head

Hi-Tigue fastener with flush head

Fig 8 Fastener systems

GP L

Lockbolt fastener with flush head

NLT

LLT

XXX

QD

X X

HLT

+ ++

o <>
| o

NLT
__
cn

LLT

HLT

xic

+1 +

2 - two
-

results

log mean endurance

o o

o NLT

c/o

I LLT
Z HLT

|o<

+ + I

8
NLT
|
x

I 000

LLT

o
X

jc

4?

HLT
J

10'

103
FALSTAFF flights
Fig 9 Fatigue endurance at 280MPa net section stress

10

ocf oo

NLT ___:
o

_i

J_

LLT -

X X

HLT
(375)

NLT L.-

| x x x

4+ |f

CD

2 - t w o results

j r LLT

XK

1 - l o g mean endurance

~ HLT (375)

+1
o oo

- g l LLT I

jxxx

NLT _

oo

(|

2
+

CD

ef

+ 2 (f +

* HLT 5 (375)
o
o

ooop

NLT --

3 LLT

X |

XX

10'

+ +-H-)

HLT (375)
I

10
FALSTAFF flights
Fig 10 Fatigue endurances at 350MPa net section stress

1 ,1

10'

37

- 20

-16
o
19
XI

a.
m-*

-a
n
o

-12

-_
(TJ

<u
a.

*in

f-8

m
o

ro

cr
c

<u
c
-J

5
c
_a

I/.
(0 _

80- 0

50

c
o
u

60

in

_c

40 - . 5
A0
30 - .4

4.
O
C

ra

20

w
_3

TJ
C
4*
If

20 - .3

_r
10 - .2
20
0 L .1
2A
Hi-Lok in
plain hole

2B
Hi-Lok in
cold worked
hole

2C
Hi-Tigue in
plain hole

2D
Hi-Tigue in
cold worked
hole

Fastener system

Fig 11 Summary of Qjoint data at 350MPa peak stress

38

0 L -.2
2A
Hi-Lok in
plain hole

2B
Hi-Lok in
cold worked
hole

2C
Hi-Tigue in
plain hole

2D
Hi-Tigue In
cold worked
hole

Fastener system

Fig 12 Summary of ITdogbone data at 335MPa peak stress

39
ANNEX 1
FASTENER FITS
A

Fastener fits of no or low secondary bending joints

In order to cover a range of fastener fits commonly used in practice, four ranges were chosen, two in combination
with cold working. The four ranges chosen were:1A clearance
IB interference
IC interference
1D interference

20
25
110
120

lOum
10 (im (with cold work)
10(im
10 (im (with cold work)

In order to check these fits, both hole and fastener diameters were measured. The fasteners used were all from a common
batch and the diameter variation was very low. The specimens were produced by the same manufacturer and the
tolerances on hole diameter were found to be good, with the exception of system 1D where the hole was produced by a
special tool. The specified and measured fits are given below for the four fastener systems.

1A

Range specified
Hole diameter range
Fastener diameter range
Measured fit range

+ 1 0 + 30um
6 . 3 6 3 - 6 . 3 6 8 mm
6.325 6.337 mm
+ 2 6 + 43 urn

1B

Range specified
Hole diameter range
Fastener diameter range
Measured fit range

15 35 um
6 . 3 1 2 - 6 . 3 1 7 mm
6 . 3 2 5 - 6 . 3 3 7 mm
- 8 - 25 (im

IC

Range specified
Hole diameter range
Fastener diameter range
Measured fit range

- 1 0 0 - 1 2 0 urn
6 . 2 1 8 - 6 . 2 2 3 mm
6 . 3 2 5 - 6 . 3 3 7 mm
- 1 0 2 - 1 1 9 urn

ID

Range specified
Hole diameter range
Fastener diameter range
Measured fit range

- 1 1 0 - 1 3 0 jim
6 . 1 9 8 - 6 . 2 4 8 mm
6 . 3 3 8 - 6 . 3 4 8 mm
- 9 0 - 1 5 0 jim

Fastener fits of high secondary bending joints

In order to cover a range of fastener fits commonly used in practice, four ranges were chosen, two in combination with
cold working. The four ranges chosen were:
2A
2B
2C
2D

clearance
interference
interference
interference

20
25
70
70

10 ism
10 jim (with cold work)
lOum
10 jim (with cold work)

In order to check these fits, both the fastener diameter and hole diameters were measured. The HI-LOK fasteners used by
each participant in systems 2A and 2B were not from a common batch, hence the measured fastener diameters are not
necessarily consistent. The HI-TIGUE fasteners however were from a common batch and the measured diameters are
therefore the same. Variation in fastener diameter for a given batch was found to be very low, usually less than about 5 um. The
hole diameters produced and measured by individual participants also show good repeatability, generally within 15 um. All
hole diameters were measured, and a summary of measurements and fits is given below.
Fastener system 2A (range specified + 1 0 + 30 <im)
Joint
Q

U
1 1/2
X

Hole dia. range

Fastener dia.

(mm)

(mm)
6.330
6.324
6.310
6.330

6.363-6.368
6.315-6.320
6.331-6.341
6.345-6.360

Range of fits

(Urn)
+33+38
-4 -9
+21+31
+15+27

Fastener system 2B (range specified 15 35 um)

Joint

Hole dia. range

Fastener dia.

(mm)
6.312-6.317
6.320-6.327

(mm)
6.330
6.324

(Urn)
-16-21
+3 - 4

6.325-6.339

6.330

+9

1 1/2
X

Range of fits

-5

40

Fastener system 2C (range specified 60 80 um)

Q
1 1/2
U

6.246-6.258

6.325

-67 -79

6.225-6.250

6.325

-75-100

Fastener system 2D (range specified 60 80 (im)

Q
6.246-6.257
1 1/2
6.220-6.235
U
*Holes produced by the same reamer

6.325

-68-79

6.325

-90-105

41
ANNEX 2
PIN, COLLAR AND NUT PART NUMBERS USED IN INVESTIGATION
This annex gives the pin and collar/nut part numbers used throughout the investigation.
Fastener system l A a n d IB
NLT joint
LLT joint
HLT joint

HL19PB-8-6
HL11VAP-8-8
HL19PB-8-11

wirh
with
with

H604-4 nut
H604-4 nut
H604-4 nut

HLT319-8-6
HLT411AP-8-8
HLT319-8-11

with
with
with

H541L-4Fnut
H541L-4Fnut
H541L-4Fnut

GPL3SC-DT08-08AC
GPL3SC-V08-08AC
GPL3SC-DT08-11AC

with
with
with

2SCC-3C08 collar
2SC-3C08 collar
2SCC-3C08 collar

with
with
with
with
with

HL70-8 collar
HL70-6 collar
HL70-8 collar
HL70-8-collar
HL70-8 collar

with

H541L-4Fnut

with
with

H541L-4Fnut
H541L-4Fnut

Fastener system IC
NLT joint
LLT joint
HLT joint
Fastener system ID
M

NLT joint
LLT joint
HLT joint

Fastener system 2A and 2B

O joint
1 1/2 dogbone
U joint
X joint

HL19PB-8
&HL19PB-6
HL19PB-8
HL19PB-8
HL19PB-8

Fastener system 2C and 2D

0 joint

HLT319-8 7
&HLT319-8-9
1 1/2 dogbone
HLT319-8-7
U joint
HLT319-8-7
1
*2

These fasteners were not available in the -6 grip length. The -8 grip length was used with a 7 mm thick washer.
Joints are assembled with a 1 mm thick washer.

42
ANNEX 3
LOAD TRANSFER AND SECONDARY BENDING MEASUREMENTS
The method adopted for measuring load transfer and secondary bending in joints was that used in the FRFS programme.
Strain gauges were attached to the specimens at various locations as described in Annex 4 of reference 2. The SB gauges were
attached on either side of the failing element near to the fasteners from which failure occurs. The SB ratio was simply calculated
as the ratio of the bending strain to the axial strain. The LT gauges were attached to either side of the specimen at two distinct
locations. One location was remote from the test section and measured the total load applied. The other location was on the
failing element beyond the row of fasteners from which the failure occured. This row of gauges measured the load bypassing the
fasteners. The method of averaging these values was not defined. The method used on the Q-joint consisted of averaging
consecutive values across the section. The method used on the 1 1/2 dogbone consisted of integrating afifthorder polynomial
fit obtained from the surface strain measurements. Each method was used consistently and comparisons should only be made
of LT values on each specimen with different fastener installations. Comparisons between LT values for different joints should
only be used as a guide.
Values of LT and SB are dependent on the applied load. Accordingly measurements from the strain gauges were taken
over the range of loads which were applied during the fatigue tests. The range was split into discrete levels and measurements
were taken at each level given below.
% of maximum load
in FALSTAFF
0
16.7
33.3
50
66.7
83.3
100
83.3
66.7
50
33.3
16.7
0
minimum load
0
The LT and SB values were calculated at each discrete level for each strain gauge pair. The average value was calculated
across the section and is presented in tabular form. The data also shows how the LT and SB values vary throughout a loading
cycle. A measurement cycle was made at the start of the test. In order to ensure that measurements were also available when the
joint was stabilised, a bcdding-in procedure was used. This simply involved cycling the joint from zero load to 50% of the
maximum FALSTAFF load for a number of cycles. The measurement cycle was then repeated. This process was repeated until
the measurements of LT and SB had stabilised. The loading sequence used in this programme is given below:*

0 - 1 0 0 % FALSTAFF - MIN FALLSTAFF - 0

0 - 50% FALSTAFF - 0 (5000 CYCLES)
*
0 - 1 0 0 % FALSTAFF - MIN FALLSTAFF - 0
0 - 50% FALSTAFF-0(5000 CYCLES)
*
0 - 1 0 0 % FALSTAFF - MIN FALSTAFF - 0
The * represents the measurement cycle described earlier.
The complete FALSTAFF load range was not used for the bedding-in process. This is because experience showed a high
percentage of strain gauge failures using the complete sequence. Since some of the gauges were adhered to the faying surfaces,
replacement was difficult and time consuming. The compromise bedding-in process however resulted in some strain gauge
hysteresis, during the measurement cycle. The measurements of LT and SB presented in this report are those obtained after the
bedding-in procedure.

43

ANNEX 4
IMPORTANCE OF SCATTER
The fatigue endurances for the no or low secondary bending programme are presented infigures9 and 10 for the case of
low and high applied stresses respectively. The scatter is generally lower at the higher stress level but not significantly. There
appears to be no relationship between fastener system and the amount of scatter. The NLT transfer specimen endurances are
similar with all of the fastener systems and at both stress levels and are therefore not further considered in this Annex.
The fatigue life scatter bands of the three enhanced fastener systems (1B, 1C and 1D) overlap for both joints at both stress
levels. It should be borne in mind that the HLT joint endurances were obtained at 375MPa and the LLT joint at 350MPa.
Scatter bands of the datum fastener system only overlap those of the other three systems in one case. This is the case of the HLT
joint at the lower stress level with HI-TIGUE fasteners. This could potentially result in a lower than unity life improvement with
HI-TIGUE fasteners over the datum system.
To quantify the effects of scatter on fatigue life improvement factors (LIF), the results are analysed in two ways. Firstly the
extremes of LIF are calculated from the data. Secondly each data set is considered to be distributed in some regular way. The
minimum lives of each data set are used to calculate LIF (min) and the maximum lives of each data set to calculate LIF (max).
Assuming the distribution of each data set to be log normal, the log mean values are also used to calculate a factor LIF (Av).
TABLE A1 Absolute Life Improvement Factors
,

- .

HLT JOINT

I FASTENER
SYSTEM

280MPa

LLT JOINT
350MPa

280MPa

I HI-TIGUE

! 1 .86 - 5-22

1.85 - 4-31

0.86 - 8.31

I HI-LOK IN
I C.W. HOLE

| 2.17 - 14-24

2.03 - 15.41

1.35 - 15-14 | 3-61 - 11.67 !

3-62 - 12.83

2.56 - 14.28 | 7.00 - 29-57 I

I HUCK EXL IN | 3-53 - 8.03

I C.W. HOLE

375MPa
1.71 - 11-57 i

TABLE A2 Distributed Life Improvement Factors

1

FASTENER
SYSTEM

LIF |

LLT JOINT
280
350

HLT JOINT
280
375

I HI-TIGUE

MIN |
AV
MAX

2-94
3-14
3-30 I

3.15
2.81
2.53 I

2.21
2.10 |
3-25

4-38
4-54
4.52

1 HI-LOK
IN C.W.
! HOLE

MIN
AV
MAX

3-44
6.05
8.99

3-46 1 3-45
6.36
3-96
9.05
5.92 I

9-25
6.41
4-56

| HUCK EXL IN I
| C.W. HOLE

MIN j
AV
MAX

5-58 |
5-48 |
5-07

6.16
7-18 |
7-53 !

17-90 |
11-99
11-56

6.54
5-53
5-59 |
I

From these two tables it can be seen that there are no systematic differences in the way in which the two types of joint rate
the fastener systems. It does however indicate that the life improvement factor based on the log mean values (Table A2) of the
endurance data provide a good guide to the improvements found in practice. If worst case data is required for minimum life
improvement estimations, then the absolute approach (Table Al) must be used.

45
ANNEX 5
PRIMARY FATIGUE CRACK ORIGINS
Primary fatigue crack origins have been noted by each participant on each fatigue test specimen and are presented in this
annex.
a)

HLT Specimen

6~-_T

s~y

N, d
>

^ rW
12

<i

}"(

Fastener
1

12

Fastener
2

1
1

Maximum net
St r e s s (MPa)

I FASTENER SYSTEM
i

SPEC NO

Flights
to Failure

ORIGINS |

i
i

i
i

i
i

1A.
i

HI-LOK
IN PLAIN HOLE

280
280
280
280
280

HI/1
H1/6
H1/5
HI/10
H1/7

37898
23172
30839
34929
14821

4
10
a
c
10

375
375
375
375
375

H1/2
HI/9
H1/8
H1/3
H1/4

7572
4031
3559
3431
2959

a
a
a
a
5,6,7,111

i
i

46

Fastener system

1B.

1C.

1D.
i
i
i
i
i

HI-TIGUE
IN PLAIN HOLE

HI-LOK IN
COLD-WORKED
HOLE

HUCK-EXL IN
COLD-WORKED
HOLE

Maximum net
Stress (MPa)

SPEC NO

Flights
to Failure

ORIGINS

280
280
280
280
280

H4/2
H4/5
H4/7
H4/9
H4/10

66624
32796
52172
123227
41024

10
6,7
b

375
375
375
375
375

H4/1
H4/3
H4/4
H4/6
H4/8

34224
27749
27031
12972
20924

10
10
10
6,7
6,7

280
280
280
280
280

H6/2
H6/8
H6/4
H6/1
H6/9

51172
76759
> 224420
166196
93021

11
11
-

375
375
375
375

H6/7
H6/3
H6/6
H6/10

31759
34525
27772
27359

11
11
5
6,7

280
280
280
280
280

H3/9
H3/1
H3/3
H3/4
H3/5

211711
155031
155172
96996
146572

d,12

375
375
375
375
375

H3/6
H3/7
H3/8
H3/10
H3/2

87511
58880
55172
59929
52972

b
10
c
b
10

11
11

b
b
d
d

47

Summary of HLT-joints

|Fastener I
System 1

TOP PLATE (CSK)

280
375

(1) 37898 ! (4) 2959

i
37,898

2,958

(9) 1232271 (6) 12972

(8) 20924

IB

EOT PLATE
280
375

CENTRE PLATE
280
375

(6) 23172 I
(7) 14821

(5) 30839
(2) 7572j
(10) 34929 | (9) 4031 I
(8) 3559)
(3) 3431 j

18,532

32,820

4,394

(7) 52172 I (1) 34224| (5) 32796

(3) 27749! (10) 41024 !
(4) 27031!
i
i

123,227

16,475

52,172

28,396
i

27,565

88,276

36,680
i

'

33,113

(9) 2117111 (7) 588801 (9)

(2) 52972! (1)
(3)
(4)
(5)
211,711

(6) 27772 (2) 51172 i (7) 31759!

(10) 27359 (8) 76759 I (3) 34525!
(1) 166196!
(9) 93021

IC

1D

55,848 1

211711
(6)
155031
(8)
155172 I (10)
96996
146572

148,579

66,1411

( ) Specimen numbers
Observations
(1) Failures in the top and bottom plates originate at fastener 1.
(2) Failures in the centre plate originate at fastener 2.
(3) The majority of failures using system 1D are in the centre plate, the longest lives are achieved with this system.
(4) All of the failures using system 1C are in the side plates, the majority in the bottom plate.
(5) The majority of failures using system 1B are in the side plates.

4S

b)

LLT Specimen

cr o
CSK

Hole 1

Fastener System

11 A.

|1B.

HI-LOK
IN PLAIN HOLE!

HI-TIGUE
IN PLAIN HOLE I

I
i

Hole 2

Maximum net
Stress (MPa)

SPEC NO

280
280
280
280
280

166
167
168
169
170

15372
14329
20172
22680
16572

350
350
350
350
350

161
162
163
164
165

6031
7772
5424
9239
8572

280
280
280
280
280

L4/1
L4/9
L4/4
L4/7
L4/10

55351
54772
53172
42172
74772

350
350
350
350
350

L4/6
L4/3
14/8
L4/5
L4/2

17080
17729
22959
21572
23372

Flights to
Failure

ORIGINS

(D b
(2) b
(2) b
(1) c,b,1

(D c,b,1
(2) b

(D b,1
(D c (2) ll
(2) b,1

(D 1 (2) bj
(2) a

(D a
(D 2 (2) a!
(D a 1
(2) a (1) a ,

(D a
(2) a

(D a
K D 2 (1) aj
(1) a
i

49

Fastener System
i

Maximum net

SPEC NO

Stress (MPa)

!1C-

I ID-

HI-LOK IN
COLD-WORKED
HOLE

HUCK-EXL IN
COLD-WORKED
HOLE

Flights to
Failure

ORIGINS

280
280
280
280

186
187
189
190

49231
118194
115231
99031

(D 2,1

350
350
350
350

181
182
183
185

18759
61325
46586
83591

(D
(D
(D
(D

280
280
280

196
198
200

79972
115021
97031

(2) 1
(2) c,b
(2) a

350
350
350
350

191
193
194
195

69611
66231
33431
47943

(D 2

(2) b,2
(2) b,2

(D 3.1

1,2
b,1
4.1
1,3

(D a
(D 2
(2) a,b

Observations of LLT joints

(1) Failure from hole (1) was more common than from hole (2) (60% at (1) 40% at (2)) but in no systematic way.
(2) Failure from the countersinks was observed only with high interference fit fasteners.
(3) All of the failures with system IB originated at the countersinks, only 50% with system ID originated at the
countersink.
(4) All of the failures with fastener systems 1A and 1C originated from multiple origins at the bore of the hole.

so

c)

Q-joint

Position of
f a i l u r e section

AA

Failure
section
i

Fastener system:

2A.

HI-LOK IN
PLAIN HOLE!
I
i
i
i
i
i

Maximum net
(Stress MPa)

SPEC NO

FLIGHTS TO
FAILURE

ORIGINS

280
280
280
280
280

X3
X4
X7
X9
X10

12128
14431
12160
13831
14031

(-1)
(-1)
(-1)
(-1)
(-1)

350
350
350
350

XI
X2
X5
X6

3925
2929
3444
4336

(-1) m,t
(0) K.L.R.S
(0) K,L,R,S
(0) K.L.R.S

j,m,q,t
h,m,q,t !
m.q(0)S
j,m,q,t i
j,q,t

51

Fastener system

Maximum net
Stress (MPa)

2B.

HI-LOK IN
COLD WORKED
HOLE

ii

i
|
i

i
i
i

! 2C.

HI-TIGUE IN
PLAIN HOLE

SPEC NO

FLIGHTS TO
FAILURE

ORIGIN

280
280
280
280
280

CW2
CW4
CW6
CW9
CW10

9631
14424
12329
16224
17631

(-2) g.(-D
(-2) o,p
(-1.5) 0(0)K
(-1 .5) h,q,u
(-1 ) j,m,p,t,u

350
350
350
350

CW3
CW5
CW7
CWS

3801
3172
3624
5323

(0)
(0)
(0)
(0)

280
280
280
280
280

12
13
14
17

78
25
30
18

(+1) B,(0),L
(+1) E,(0),R
+

350
350
350
350
350
350

15
16
19
20
21
22

15825
21730
11 173
16 385
8 232
13 573

032
225
860
530

K.L.q.S
K,L,S
K,L,R
K.L,_.,(-. )o,ul

( DB

(+1) D,E

(+1
(-D
(-3)
(-0

)E
L,q
u
q,t

(
-D n
(+1 ) E
i

280
280
280
280
280
! 2D.

HI-TIGUE IN]
COLD WORKED
HOLE

4
5
6
7
10

(+2) a
1
(+2)f(0)R
2

25 573
45 632
77 330
111 393
91 225
I

350
350
350
350
350

2
3
8
9
11

38 981
32 695
16 840
11 159
7997

*1 Failure from defect near end plate

*2 Failure from fasteners in controlling section
Q-joint Summary
Fastener system/stress level Failure site
2 A/low On -1 line, just away from edge of hole
2A/high On 0 line, at edge of hole.
2B/low On -1.5 line, well away from edge of hole
2B/high On 0 line, at edge of hole.
2C/low On +1 line, at edge of countersink
2C/high On+1 line, at edge of countersink or
On -1 line, near edge of hole.
2D/low Various edge of specimen
2D/high On -1.5 line, well away from edge of hole.

(-Dv(O)B
(-5) P

(-D h
(-3) u
(-1) n
(-2) u

52

d)

1 1/2Dogbone

4
^S
_ ^

OPQRZX

zzzr

T5 T U V W

^Si

4
55

53

Fastener system

1
I 2A.
i

HI-LOK IN
PLAIN HOLE

1
i
i
i
i

Maximum net
Stress (MPa)

SPEC NO

FLIGHTS TO
FAILURE

ORIGIN

268
268
268
268

18411
60372
56972
63831

S
R
S
E

335
335
335
335

9559
15419
23373
22231

F
S
F
E

268
268
268
268

29572
40431
58231
35759

335
335
335
335

13524
14231
17962
19172

R *V

S
R-*W

S
R-V

2B. HI-LOK IN
COLD WORKED
HOLE

R
P+T

i
i

268

Cl

268

C5

30764
36572

G
H

335
335
335
335

C4
C2
C3
C8

9983
16734
14067
11780

G
U
U
G

268
268
268

D1
D2
D6

37146
42640
28265

A
Y
Q

COLD WORKED
HOLE

335
335
335
335

D7
D5
D8
D3

19007
17308
14946
14564

P
P
B
T

| 2A. HI-LOK IN
PLAIN HOLE
(COMPATABILITY)

335
335
335

A3
A1
A2

11165
13826
19955

F
E

2C. HI-TIGUE IN
i

PLAIN HOLE

2D. HI-TIGUE IN

G
i

54
1 1 /2 Dogbone Summary
Fastener system/stress level Failure site
2 A/low From edge of hole at interface or just forward of hole.
2A/high From edge of hole at interface or just forward of hole.
2B/low Just forward of hole or at edge of side plate.
2B/high At edge of side plate or from edge of hole at interface.
2C/low Just away from edge of hole at interface
2C/high Just away from edge of hole at interface or forward of fastener holes.
2D/low Various away from fastener hole on top surface.
2D/high Just forward of fastener hole on top surface.

55
e)

X-joint

3
c
a
1

5
a

b
2

h
8

I Fastener system

Maximura net

Stress (MPa)

i
i
i

i
i
i

SPEC NO

FLIGHTS TO
FAILURE

200
200
200
200

41 .6
41.7
41 .8
41.2

13860
16972
13180
13772

267
267
267
267

41 .1
41-3
41.4
41-5

5425
5329
5590
6280

200
200
200
200

42.4
42.5
42.6
42.8

42772
30224
35631
27630

267
267
267
267

42.1
42.2
42.3
42.7

10929
11972
11372
6172

ORIGIN

1-6
a,b,d-h
a-e,g

a-f

2A. HI-LOK IN
PLAIN HOLE

| 2B. HI-LOK IN
i
COLD WORKED
i
i
HOLE
i
i
i

i
i

a-f,h

a-h
a-f
a,c-h

1-6,h

1-8
1-8
1,3-6,8

a-h
a-h
a,c-h
1-8

i
i

X-joint Summary
Fastener system/stress level Failure site
2A/low from fastener holes in base plate
2A/high from fastener holes in base plate
2B/low from fastener holes in splice plate
2B/high from fastener holes in base plate

56

f)

U-joint

b .
c
1
2

b
c

1
2

" >

-\

Fastener system

Maximum net

SPEC NO

Stress (MPa)

! 2A.

j 2C.

HI-LOK IN
PLAIN HOLE

HI-TIGUE IN
PLAIN HOLE

FLIGHTS TO
FAILURE

ORIGIN

276
276
276
276

S1-4
S1-5
S1-6
SI-7

7831
19311
10631
23631

b,c
2
1.2
b,c

345
345
345

S1-1
S1-2
S1-3

5558
5729
5431

1.2
a,c
b

276
276
276
276

S3-4
S3-5
S3-6
S3-7

24372
17834
16031
19572

1,2
1.2
1,2
1,2

345
345
345

S3-1
S3-2
S3-3

15424
13520
13969

a
2
1,2

57
ANNEX 6
Q-JOINT ASSESSMENT
The Q-joint has been developed during the past three years. It is anticipated that development will continue to enable the
joint to produce different amounts of LT and SB by simple geometric changes. These changes involve fastener diameter,
thickness of controlling element and spacing of fastener rows. The standard Q-joint design is given in Fig. 3 with 1/4" diameter
fasteners installed. Testing of the Q-joint with HI-LOK (2A) and HI-LOK in a cold worked hole (2B), was performed with
3/16" diameter fasteners installed in the controlling section, and 1/4" diameter fasteners in the test section. In order to assess
whether using the standard joint would produce similar fatigue lives to those obtained with the smaller fastener in the
controlling section, measurements of LT and SB were made on standard joints. The results of these measurements are given in
table 4, and are compared with those obtained using the smaller 3/16" diameter fastener, presented in table 5. Many strain
gauge failures occurred during the testing of the 1/4" standard joint, particularly with the plain hole specimen. Comparisons
can however be made of the cold worked specimens. The load transfer values are very similar but the secondary bending is
somewhat higher in the standard joint over most of the load range. It is expected therefore that the fatigue lives of standard Qjoints with 1/4" HI-LOK fasteners in plain and cold worked holes will be slightly shorter than those with 3/16" diameter
fasteners in the controlling section. A small number of tests have been performed with the standard joint, and the above appears
to be true at the lower stress level. At the higher stress level however the fatigue endurances with the standard joint are slightly
greater than those with the 3/16" fasteners. Since the secondary bending is so. high, it is expected that there would be little
difference between the fatigue lives of plain and cold worked hole specimens. This also appears to be true from the small
number of additional tests.

REPORT DOCUMENTATION PAGE

1. Recipient's Reference

2. Originator's Reference 3. Further Reference

AGARD-AG-304
5. Originator

6. Title

ISBN 92-835-0429-1

4. Security Classification
of Document

UNCLASSIFIED

Advisory Group for Aerospace Research and Development

North Atlantic Treaty Organization
7 rue Ancelle, 92200 Neuilly sur Seine, France
STANDARD FATIGUE TEST SPECIMENS FOR
FASTENER EVALUATION

7. Presented at

8. Author(s)/Editor(s)

9. Date
November 1987

by Robin Cook
10. Author's/Editor's Address

11. Pages

See flyleaf
12. Distribution Statement

64

This document is distributed in accordance with AGARD

policies and regulations, which are outlined on the
Outside Back Covers of all AGARD publications.

13. Keywords/Descriptors

Fastener
Joints
Bend properties

Fatigue properties
Loads (forces)

14. Abstract

An AGARD coordinated programme which examines the fatigue performance and joint
characteristics of a number of mechanically fastened joints has been completed. This report
describes the programme which examines mechanically fastened joints with (1) no or low
secondary bending and (2) with high secondary bending. In part 1, three types of joint are
assessed which exhibit no, low and high amounts of load transfer by the fastener. The no load
transfer joint was rejected and low and high load transfer joint were considered to be equivalent
in rating fastener systems. In part 2, three types of single shear joint are considered. They are
compared on the basis of load transfer and secondary bending characteristics and also on the
fatigue endurance with a range of fastener systems installed. Only one joint, the UK designed
O-joint, adequately fulfilled the requirements of a standard joint for fastener evaluation
purposes.
This AGARDograph was sponsored by the Structures and Materials Panel of AGARD.

Ms.(&S3m
NATO

t $

OTAN

DISTRIBUTION OF UNCLASSIFIED

7 rue Ancelle 92200 NEUILLY-SUR-SEINE

AGARD PUBLICATIONS

FRANCE
Telephone (1)47.38,57.00 Telex 610 176

AGARD does NOT hold stocks of AGARD publications at the above address for general distribution. Initial distribution of AGARD
publications is made to AGARD Member Nations through the following National Distribution Centres.Further copies are sometimes
available from these Centres, but if not may be purchased in Microfiche or Photocopy form from the Purchase Agencies listed below.
BELGIUM
Coordonnateur AGARD - VSL
Etat- Major de la Force Aerienne
Quartier Reine Elisabeth
Rue d'Evere. 1140 Bruxelles

NATIONAL DISTRIBUTION CENTRES

ITALY
Aeronautica Militarc
Ufficio del Delegate Nazionale all"AGARD
3 Piazzale Adenauer
00144 Roma/EUR

CANADA
Defence Scientific Information Services
Dept of National Defence
Ottawa, Ontario K1A 0K2
DENMARK
Danish Defence Research Board
Ved Idraetsparken 4

LUXEMBOURG
See Belgium
NETHERLANDS
Netherlands Delegation to AGARD
National Aerospace Laboratory, NLR
P.O.Box 126
2600 AC Delft
NORWAY

2100 Copenhagen 0

Norwegian Defence Research Establishment

Attn: Biblioteket
P.O. Box 25
N-2007 Kjeller

FRANCE
O.N.E.R.A. (Direction)
29 Avenue de la Division Leclerc
92320 Chatillon

PORTUGAL
Portuguese National Coordinator to AGARD
Gabinete de Estudos e Programas
CLAFA
Base de Alfragide
Alfragide

GERMANY
Fachinformationszentrum Energie,
Physik, Mathematik GmbH
Kernforschungszentrum
D-7514 Eggenstein-Leopoldshafen

2700 Amadora

GREECE
Hellenic Air Force General Staff
Research and Development Directorate
Holargos, Athens
ICELAND

Director of Aviation
c/o Flugrad
Reyjavik

TURKEY
Mill! Savunma Bakanligi
ARGE Daire Baskanligi
Ankara
UNITED KINGDOM
Defence Research Information Centre
Kcntigern House
65 Brown Street
Glasgow G2 SEX

UNITED STATES
National Aeronautics and Space Administration (NASA)
Langley Research Center
M/S 180
Hampton, Virginia 23665
THE UNITED STATES NATIONAL DISTRIBUTION CENTRE (NASA) DOES NOT HOLD
STOCKS OF AGARD PUBLICATIONS, AND APPLICATIONS FOR COPIES SHOULD BE MADE
DIRECT TO THE NATIONAL TECHNICAL INFORMATION SERVICE (NTIS) AT THE ADDRESS BELOW.

r
c

PURCHASE AGENCIES
National Technical
Information Service (NTIS)
5285 Port Royal Road
Springfield
Virginia 22161, USA

ESA/Information Retrieval Service

European Space Agency
10, rue Mario Nikis
75015 Paris, France

The British Library

Document Supply Division
Boston Spa, Wetherby
West Yorkshire LS23 7BO
England

Requests for microfiche or photocopies of AGARD documents should include the AGARD serial number, title, author or editor, and
publication date. Requests to NTIS should include the NASA accession report number. Full bibliographical references and abstracts of
AGARD publications are given in the following journals;
Scientific and Technical Aerospace Reports (STAR)
published by NASA Scientific and Technical
Information Branch
NASA Headquarters (NIT-40)
Washington D.C. 20546, USA

Government Reports Announcements (GRA)

published by the National Technical
Information Services, Springfield
Virginia 22161, USA

Printed by Specialised Priming Services Limited

40 ( higwell Lane, Loughton, Essex IGI03TZ
ISBN 92-835-0429-1