Journal of Advanced Mechanical Design,

Systems, and

Manufacturing

Vol. 1, No. 3, 2007
358
A Study of Self-Loosening of Bolted Joints Due
to Repetition of Small Amount of Slippage at
Bearing Surface
*

Shinji KASEI
**

**Shinshu University,
4-17-1, Wakasato, Nagano, 380-8553 JAPAN
E-mail: skaseid@shinshu-u.ac.jp


Abstract
This paper deals with the mechanism and relating matters on self-loosening rotation
of bolted joints in the cases when very small slippages occur repeatedly at bearing
surfaces under transverse loads. It can be supposed that accumulation of this kind
of loosening finally produces serious states for bolted joints. Based on an
assumption that a restoring force caused by an elastic torsion of a bolt shank which
arises from a relative displacement between bolt and nut threads drives loosening
rotation, a theoretical explanation is presented about how loosening rotation occurs
and grows larger. Experiments in quasi-static states show results which agree with
the theory of this paper. Additionally, consumption of transverse work and
anti-loosening performance being taken up as relating subjects, more information
about self-loosening is shown through examinations on some kinds of test samples.
Key words: Bolted Joint, Transverse Load, Loosening Rotation, Minute Slippage,
Bolt-Shank Torsion, Loosening Mechanism, Consumed Work,
Anti-Loosening Performance
1. Introduction
Self-loosening of screw threads is one of major subjects relating directly to reliability of
bolted joints. Accordingly, many efforts have been made to understand this phenomenon
and prevent its occurrence. It is well known that there exist two kinds of self-loosening, the
one due to rotation of a bolt or a nut and the other not due to it. Their respective features
and mechanisms have been known gradually
(1)~(4)
. Especially, many researchers have
exerted themselves for self-loosening due to rotation since it has a dangerous inclination to
bring greater loss of a fastening force than the case not due to rotation. G.H.Junker
(5)
pointed
out clearly that loosening rotation occurs easily under repetition of transverse loads and thus
we should take notice of this case. A.Yamamoto and S.Kasei
(6),(7)
presented a hypothesis of
the loosening mechanism, in which behavior of an elastic torsion of a bolt shank performs
an important role, and verified it experimentally. As recent achievements, many papers have
been published by R.I.Zadoks & X.Yu
(8)
, P.Wolfsteiner & F.Pfeiffer
(9)
, N.G.Pai &
Hess
(10),(11)
, Y.Jiang, et al.
(12),(13)
, S.Izumi, et al.
(14),(15)
and so on. Each paper has respectively
contributed to explication of the self-loosening problems. It is noticeable that three
dimensional FEM was used as a tool of analyses in some papers
(10)~(12),(14),(15)
. This fact
makes us feel that an era in which problems like self-loosening can be numerically analyzed
has come. Among them, S.Izumi, et al.
(14)
showed the analytical results supporting the
results by the author, et al.
(6),(7)
.
The author, et al.
(16),(17)
pointed out that loosening rotation can be induced even when
*Received 24 Jan., 2007 (No. 07-0054)
[DOI: 10.1299/jamdsm.1.358]


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
359

slippages at bearing surfaces are very small and its situation should be considered as a sign
capable to connect to a serious state. And a theoretical consideration about the loosening
mechanism was developed by basing on the hypothesis
(6),(7)
mentioned above. Subsequently,
several researches intending to this case were performed
(10)~(13),(15)
. In particular, S.Izumi, et
al. classified contacting situations of a mating surface on threads and a bearing surface with
nine kinds of combinations and explained loosening behavior in each combination.
This study was planned to follow up the previous report
(16)
, especially about
experimental considerations which have been remained not enough. So, it is the main
purpose to prove the theory experimentally by using an improved experiment apparatus and
consequently contribute explication of the self-loosening mechanism. In addition, to present
more information about self-loosening of bolted joints, consumptions of transverse work
and anti-loosening performances are examined about some test bolt-nut samples.
2. Theory
Along the hypothesis of the previous report
(16)
, a theoretical approach will be promoted.
Basically, the author considers that loosening rotation is caused by a restoring action of an
elastic torsion of a bolt shank which is due to a relative motion at a mating surface on
threads. Figure 1 shows an explanatory illustration about this relation in the case of
right-hand threads. Speaking relatively, it can be said that whichever side is fixed between a
bolt and a nut, it does not make an essential difference.
2.1 Behavior of elastic torsion of bolt shank
Figure 2 illustrates slide loci
(6)
at the mating surface on threads under the transverse
load. The side of bolt-threads is replaced with a concentrated point and the slide loci which
the point draws on a thread surface of the nut are expressed schematically in processes of
sliding-up and -down (the nut side is fixed). Where, : a half angle of thread, : a lead angle
of a helix, r: a radial position from the bolt axis, u: a circumferential displacement
(corresponding to the torsion), R
r
: a force which moves the bolt-threads up or down along
the thread surface of the nut, F
a
: a force which moves the bolt-threads down along the helix,
and F
t
: a restoring force due to the torsion along the helix. The direction of a resultant force
of R
r
, F
a
and F
t
is considered to coincide with the tangential direction of the loci and to be
opposite to the direction of a frictional force on the thread surface. In the mating surface on
threads, parts of sliding-up and sliding-down coexist and the slide brings the torsion angle

T
. By taking a pitch diameter of thread d
p
into account, an equation
p T
d u 2 = is obtained.
If the nut side is fixed throughout the processes, u can reach to a critical value u
cr
(or
Tcr
, a
critical value of
T
) in a critical condition
t a
F F = . However, once the nut slides at the
bearing surface, the torsion is released by a rotation of a unified body composed of the bolt
and the nut
(6),(16)
. Since the bolt-threads descend along the helix of the nut-threads with
Ft
Fa
Rr
A
B

a
Friction force
Screw surface
of nut
Circumfer-
ential dir.
Axial dir.
Radial dir.
u
r
In sliding-up
Slide locus

a
u
r
A'
B'
Slide locus
In sliding-down
Fig.2 Slide loci on mating surface on threads
Fixed plate
Movable plate
Bolt shank
Torsion Restoring
force
Fig.1 Transverse displacement
and torsion of bolt-shank


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
360
increasing of u and thus an axial tension F
f
reduces by a quantity corresponding to a
deformation of tan u , the above rotation becomes a loosening one in the result.
The slide locus is formulated by the following equation
(6),(7)
.



cos cos
cos
d
d
t a
r
F F
R
u
r

= (1)

Denoting a part of F
f
as F
f
', a coefficient of friction as
s
and a part of a spring constant
of the bolt K
u
as K
u
', the following relations are induced for Eq.(1)
(6)
.
( ) ( )
2 ' '
'
,
'
) cos 4 ( , sin
cos tan 1 cos tan
down) (sliding up) (sliding
p u t f a
s
f s
r
s
f s
r
d u K F F F
F
R
F
R





= =
+
= =

Consequently, the critical torsion angle
Tcr
is
known as
tan
2
tan
2
'
'
u
f p
u
f p
Tcr
K
F d
K
F d
= = (2)
where K
u
and K
u
' are quantities with the unit of
[Nm/rad].
2.2 Progress of loosening rotation
Typically, the case where an initial torsion
u
0
due to a bolt clamping is greater than u
cr
is
discussed. Figure 3 conceptually illustrates two
typical relations between the number of
loading cycle n and a behavior of u. The one
denoted as (a) is in a state where the nut does
not slide and conversely the other (b) is in a
state where the slide of the nut exists at the
bearing surface little by little. In the case (a), u
approaches gradually to u
cr
from an upper side
because of a gradual release of u at the mating
surface on threads, but no loosening rotation
occurs. In the case (b), at an initial stage of the
transversely repeated loading u goes down
beneath u
cr
due to a slide at the mating surface on threads and/or the bearing surface of the
nut. Afterward, u increases toward u
cr
when no slippage occurs at the bearing surface and
conversely decreases when the slippage occurs. In Fig.3, a zigzag line expresses a process
of progress of the loosening conceptually and means that the loosening rotation occurs
cyclically in sections from u
un
to u
ln
. It is considered that u
un
and u
ln
depend on the axial
tension, the transverse load, frictional conditions, etc. Moreover, the difference between u
un

and u
ln
is considered to have a close relation with the magnitude of loosening rotation.
2.3 Relation between transverse load and transverse displacement
Figure 4 shows two basic models expressing the relation between the transverse load W
and the transverse displacement x. The models correspond to cases with some losses of
energy at the mating surface on threads and/or the bearing surface. S
0
is a starting point of
the loading. In the case (a), the slippage at the mating surface on threads only exists (S
4
'S
5
'
Number of cycle n
T
o
r
s
i
o
n

o
f

b
o
l
t
-
s
h
a
n
k

u
loosening
Occurrence of
loosening
Occurrence of
u
ln
u
un
u
cr
(a)
u
0
(b)
Fig.3 Explanatory figure of relation between
torsion and number of loading cycle
W
S3
S5
S0
S1
S2
S4
S5
S6
S7 x
S7
S2
S4
W
x S0
S1
(a) (b)
Fig.4 Two basic models between
transverse load and displacement


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
361
and S
7
'S
2
'), but there is no
slippage at the bearing
surface. The case (b) has, in
addition to the part of the
slippage at the mating surface
on threads (S
4
S
5
and S
7
S
2
),
the part of the slippage at the
mating surface on threads
plus at the bearing surface
(S
2
S
3
and S
5
S
6
). The case (a)
corresponds to the one of
Fig.3(a) and the case (b) to
the one of Fig.3(b).
Incidentally, sections of S
0
S
1
, S
2
'S
4
' and S
5
'S
7
' in the former and sections of S
0
S
1
, S
3
S
4
and
S
6
S
7
in the latter express processes in which the bolt and the nut displace together like a
unified body.
3. Experiment apparatus and experimental
conditions
Fig.5 shows an outline of the experiment
apparatus. Roughly speaking, it consists of a main
body, a driving part and a system for measurement
and control. In the main part, many steel balls being
inserted in-between, a movable plate is clamped to a
fixed plate by a test bolt and a test nut under a
required force. In the driving part, through a
differential screw mechanism a servomotor gives
the movable plate a transverse alternating
displacement. The servomotor has a reduction gear
drive and a rotational encoder, and its rated output
performances are 111W and 30rpm. In addition, the
differential screw mechanism in which a
combination of pitch of 1.5 and 1.25mm is
employed is used to convert a rotational motion to a
translational one and make a small and smooth
quasi-static transverse displacement. A motion
control is performed for the amplitude of the
transverse displacement of the movable plate.
Figure 6(a) shows a detailed drawing of the side
view of the main body. Parts numbered are: the
test bolt, the test nut, a attachment for
measuring the bolt shank's torsion angle or the nut's
rotation angle (additionally, the nut's slippage at the
bearing surface), the movable plate, the fixed plate, a load-cell for measuring
the axial tension of the bolt, a fixture for fixing firmly the test bolt to a inner side of .
A bearing plate with which the test nut keeps contact was heat-treated and its both end
surfaces were ground.
Quantities to be measured are the axial tension F
f
, the transverse displacement of the
movable plate x, the transverse load W, the torsion angle of the test bolt shank
T
, the
rotation angle of the test nut and the slippage at the bearing surface of the test nut s. As
shown in Figure 6(b), fixing the attachment to the test bolt shank,
T
is obtained basing
on subtraction of outputs of two gap sensors A and B, and instead fixing to the test nut,
Pulse Board

Strain Amp
A/DBoard
Strain Gage
Control Unit
Fixed plate
Bolt/nut
Servo motor
Torsion,
Slippage,
Rotation
Differential
screw mech.
Axial force
Ext.
force
Displacement
Movable plate
Electronic
micrometer
Non-contact
sensors
Strain gages
Strain amp.
A/D board
Control unit
Pulse board
Fig.5 Outline of the experiment apparatus
(a) Side view of the main body
(b) Attachment for measurement
of torsion, etc.
Fig.6 Detailed explanation on the
apparatus


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
362
and s are obtained basing on subtraction
and sum of outputs from A and B,
respectively.
Table 1 shows the summary of
conditions for the experiments. Test thread
sizes are M10 and M101.25, and a grip
length is 28mm. An initial axial tension of
the bolt is always set to be 10kN and
machine oil added MoS
2
powder is applied
to the thread surfaces and the bearing
surfaces as a lubricant.
4. Experimental results and
considerations
4.1 State of progress of loosening
Figure 7 shows changes of the
rotation angle of the nut under
transversely repeated loads.
Amplitudes of the transverse
displacement x are 15m to 40m.
The negative side of means the
loosening direction. Changes of in
these cases are small and especially
in ~ the changes are smaller
and can be recognized to stop on the
way. In the cases of ~, it is
supposed that the changes are not due
to loosening rotation, but due to
releases of initial torsions of the bolt
shanks and initial deflections of the
bolt shanks which are forced to
generate in clamping operations.
These kinds of rotations only connect
to initial tiny losses of F
f
. On the
other hand, in and , real
loosening rotations probably add to
some extent and so their values
steadily grow greater little by little.
According to Fig.7, it can be
supposed that loosening rotation will
go forward when the amplitude of x
is greater than around 30m in the
present experiments.
4.2 Torsion of bolt shank and
loosening rotation
Figure 8 shows experimental
curves of the torsion angle
T
and
corresponding curves of F
f
when the
amplitudes of x are 15 and 60m.
About
T
, the direction indicated with
-30
-25
-20
-15
-10
-5
0
0 100 200 300 400 500 600 700 800
Number of cycle n
R
o
t
a
t
i
o
n

o
f

n
u
t

B:15m A: 15m
A:30m B: 30m
A: 35m A: 40m
Fig.7 Progresses of the nut rotation (A: M10 hexagon nut,
B:M101.25 hexagon nut with flange)
0
2
4
6
8
T
o
r
s
i
o
n

T



(

)
15m
60m
Tcr
2
4
6
8
10
0 50 100 150 200 250 300
Number of cycle n
A
x
i
a
l

t
e
n
s
i
o
n


F
f




(
k
N
)
15m
60m
Fig.8 Changes of torsion angle and axial tension (M10
hexagon nut)


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
363
the arrow mark in Fig.1 is employed
as the positive side. It is known that
these curves of
T
show a similar
tendency to Fig.3. In the case of
15m,
T
approaches gradually to a
value regarded as
Tcr
from an initial
value forced by the clamping
operation and only the tiny loss of F
f

can be recognized in the early stage
as mentioned in 4.1. On the other
hand, in the case of 60m,
T
drops
down rapidly just after a start of the
transverse loading and afterward
repeats up and down in the area between zero and the value regarded as
Tcr
. Moreover, it
can be seen that progress of loosening corresponds to the behavior of
T
. By substituting
F
f
=10kN, =3.03, d
p
=9.03mm and K
u
=1.5910
3
Nm/rad to Eq.(2),
Tcr
=1.5210

3
rad=5.22is found. Here, the value of K
u
is from a calculation about the test bolt. This
value agrees fairly well with the experimental value indicated in Fig.8.
To examine in detail, a behavior of
T
was detected in a process which consists of a
stage of a manual clamping by a spanner, five cycles of loading (the amplitude of x is
100m) and a manual loosening by the spanner. Figure 9 shows its result. In the figure, A
1

is equivalent to a point at which the fastening torque reaches a maximum, A
2
a removal
point of the fastening torque, A
3
a point of a maximum loosening torque and A
4
an ending
point of the loosening. According to Fig.9, in the process of transverse loading except an
initial part,
T
behaves up and down similarly and periodically every half cycle. Such
feature is because the test bolt-nut is almost under the same kinetic conditions every half
cycle in whichever direction the transverse load acts. However, it is supposed that if there
are some reasons like geometrical deviations (e.g. perpendicularity between a bolt's axis
-10
-5
0
5
10
15
20
25
0 100 200 300 400
T
o
r
s
i
o
n
a
l

a
n
g
l
e

T





(

)
A
1
A
2
A
3
A
4
100m5cycles
Time (s)
Fig.9 Behavior of torsion angle of bolt-shank in
manual-fastening plus transverse-loading plus
manual-loosening (M10 hexagon nut)
-0.8
-0.4
0
0.4
0.8
T
r
a
n
s
v
e
r
s
e

l
o
a
d


W

k
N

0
2
4
6
T
o
r
s
i
o
n

T

(

)
-35
-30
-25
-20
-15
-100 -50 0 50 100
N
u
t

r
o
t
a
t
i
o
n

Transverse displacement x (m)

S3
S6
S4
S7
S7
S4
S3
S6
S2
S2
S2
S5
S5
S3
S6 S7
S5
S4
Fig.10 Mutually related diagrams of
transverse load, torsion angle of
bolt-shank and rotation of nut
(M10 hexagon nut)
0
2
4
6
-120 -60 0 60 120
x (m)
0
2
4
6
-120 -60 0 60 120
x (m)

Fig.11 Additive diagrams of torsion angle of

bolt-shank (M10 hexagon nut)
Table 2 Slip widths and rotation angles of nut in
a half cycle (n=5, M10,hexagon nut)


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
364
and a bearing surface), relating conditions will change and consequently
T
will show
some different behavior. And, during transverse loading
T
exhibits a bias with a positive
value. This is considered to meet the hypothesis that an elastic torsion of a bolt shank plays
a role to drive loosening rotation.
Figure 10 shows diagrams of W,
T
and relating to x by arranging these in parallel.
These diagrams were obtained from a simultaneous detection in the same experiment (the
amplitude of x is 100m). Since the case is supposed to correspond to the state shown by
Fig.4(b), the symbols used there are attached at the corresponding points supposed. During
sections of S
3
S
4
and S
6
S
7
in which the bolt and the nut are regarded to be unified,
T

changes just only a little, but increases during sections of S
4
S
5
and S
7
S
2
in which the mating
threads are regarded to be forced to slip. And it is known that the nut rotates in the manner
synchronizing with sections of S
2
S
3
and S
5
S
6
which are regarded as sections involving the
slippage at the bearing surface. Figure 11 shows the diagrams of
T
versus x when the
amplitudes of x are 60 and 80m. Similarity is clearly recognized between the diagrams of
Figs.10 and 11. It is considered that such behaviors and relationships show a validity of the
theory explained in 2. It is supposed that geometrical and tribological conditions at the
mating threads and the bearing surface are influence factors which bring fluctuations to
these relations.
In the experiments, progresses of loosening rotation are clearly recognized when the
amplitude of x is greater than 60m. Table 2 shows experimental values of the width of the
slippage s at the bearing surface and the increment of during a half cycle at the fifth cycle
(n=5). We can confirm that s is very small on the whole, loosening rotation occurs even in
the case where s is less than 10m and its value grows larger with increase of the magnitude
of slippage.
M10, hex. nut M101.25, hex. nut M10, hex. nut with locknut
Fig.12 Changes of axial tension and consumed work per a cycle (amplitude of x is 80m)
n = 2
n
=100
n
=300
- 0. 8
- 0. 4
0
0. 4
0. 8
- 100 - 50 0 50 10 0
x (m)
W



(
k
N
50 100 -100
0.8
0.4
-0.4
-0.8
W



(
k
N
)
0
-0.8
-0.4
0
0.4
0.8
-100 -50 0 50 100
x (m)
W
(
k
N
)
n = 2
100
n = 2
100
300
x
(m)
-1
-0.6
-0.2
0.2
0.6
1
-100 -50 0 50 100
x (m)
W
(
k
N
)
n = 2
100
300
M10, hex. nut M101.25, hex. nut
M10, hex. nut with locknut
Fig.13 Diagrams between transverse load and transverse displacement corresponding to the cases of Fig.12


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
365
4.3 Consumed work
Since occurrence of loosening rotation must be closely related with the slippages at the
mating surface on threads and the bearing surface, a work (energy) consumed by them is
considered to involve information about loosening. For anti-loosening performances, G. H.
Junker, et al.
(19),(5)
employed a quantity relating to the external work of vibration as an
evaluation factor. Here, the work consumed per one loading cycle designated as L will be
examined. L is considered to equal an area surrounded by diagrams of W versus x like in
Fig.4. So, measurements of L are tried every cycle through a numerical treatment with
regard to W and x.
Figure 12 shows experimental examples about F
f
and L versus n. The nut samples are as
follows, (a) a M10 hexagon nut, (b) a M101.25 hexagon nut and (c) a M10 hexagon nut
with a locknut (so-called a "double-nut"). Figure 13 shows diagrams of W versus x
corresponding to the cases of Fig.12. According to these figures, the following can be
stated. In the case of (a), both F
f
and L fall most rapidly. It seems to be caused by a reason
that the slippages are easy to occur at the mating surface on threads and the bearing surface.
In the case of (b), both of F
f
and L shows a tendency of slow falling. It is supposed that the
slippage at the mating surface on threads is hard to occur to some extent. Lastly, in the case
of (c), F
f
shows no falling tendency, but L shows a quite different feature. Namely, although
the value of L is initially in a higher level and keep the level for a while, afterward it falls
down steadily. It can be pointed out that the slippage is forced to occur only at the bearing
surface, protecting strongly the slippage at the mating surface on threads by a complete
locking by means of a locknut. In the above example, it is known that the slippage at the
bearing surface is getting harder to occur at n=300 than at n=2 and 100. Perhaps, such a
situation comes from tribological changing of the bearing surface.
4.4 Tests for anti-loosening
performance
Anti-loosening perform-
ances of several kinds of nut
samples are compared one
another by evaluating a loss
rate of the axial tension .
Table 3 presents M10 or
M10 1.25 test samples in
detail and the marks assigned
for them. In the case of a M10
hexagon nut with a locknut (a
"double-nut"), a locking
operation is completely done
to assure locking effect.
Figure 14 shows the
values of which are
obtained from the axial
tension losses between n=0 to
300 under five different
amplitudes of x up to 100m.
Each point doted expresses
the average of three test
results. Speaking from the
results shown in Table 2, the
slippage at the bearing
0
20
40
60
80
100
Amp. of displacement (m)
L
o
s
s

r
a
t
e

o
f

a
x
i
a
l

t
e
n
s
i
o
n

(
%
)
15 30 60 80 100
Fig.14 Results of anti-loosening tests
With washer
Coarse (1.5)
Fi ne (1.25)
Hex. Nut
Thread
pitch (mm)
Without washer
Hex. nut
Hex. nut
wi th fl ange
Hex. nut
with locknut
(double-
nut)
"Washer" means a special plain washer which was heat-treated and
ground


Table 3 Samples for anti-loosening tests


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
366
surface is estimated to be very small.
According to Fig.14, the "hexagon nut", which is regarded as the basic sample, has the
lowest performance. Other samples, although degrees of the effect are not alike, exhibit
some respective effects. However, an excellent performance is only recognized in the case
of the "hexagon nut with a locknut ("double-nut")". Other cases seem to have insufficient or
limited effects. Progress of loosening rotation is affected by the conditions of the bearing
surface and the mating surface on threads. The performance of the "hexagon nut with a
locknut ("double-nut")" comes from a hard obstruction against the slippage at the mating
surface on threads. Basically, situations in which slippages may occur at both parts of the
mating surface on threads and the bearing surface should not be permitted for preventing
loosening rotation. Fig.14 suggests such a point well.
5. Conclusions
(1) With regard to the cases under repetition of small slippages at the bearing surface, a
theoretical explanation about the mechanism for occurrence and progress of loosening
rotation was presented. It is based on the hypothesis that the basic driving force of
loosening rotation is the restoring force due to the elastic torsion of the bolt shank.
(2) Experiments under repetition of the transverse displacement were performed about M10
or M101.25 samples in quasi-static states. It can be said that experimental results show
well agreements with the theory of this paper and thus the basic mechanism of loosening
rotation can be explained well by this paper.
(3) The consumed work in the mating surface on threads and the bearing surface was taken
up as a factor suggesting the feature of loosening. With some examples detected in the
experiments, relations between its value and tendency and loosening behaviors were
explained.
(4) Anti-loosening performances about some kinds of test samples were inspected under
repetition of small slippages at the bearing surface. Among the present samples, only the
case employing the "hexagon nut with a locknut", which was subjected to a strong
locking operation, showed an excellent performance. It is emphasized that when a high
anti-loosening performance is required a firm obstruction of the slippages at the mating
surface on threads and/or the bearing surface must be ensured.
Acknowledgment
The author expresses sincere thanks to Prof. Emeritus, Akira Yamamoto, Tokyo
Institute of Technology for his guidance and suggestions over long years. The author is
grateful to Mr. Satoshi Yoshida, Mr. Hisanori Ishibashi and Mr. Takato Komura for their
co-operations and contributions to this study as graduate students, and also grateful to other
students who took part in this study. Additionally, the author wishes to thank the Ministry of
Education, Culture, Sports, Science and Technology, and Japan Society for the Promotion of
Science for their financial support (the Grant-in Aids for scientific research) to some parts
of this study.

References
(1)Yamamoto,A., Principles and Design of Threaded Joints, (1995), Chap. 4, Yokendo. (in
Japanese)
(2)The Japan Research Institute for Screw Threads and Fasteners (JFRI) ed., Guide to Fastening
Technology of Bolted Joints, (2004), Chap. 6, JFRI. (in Japanese)
(3)Yoshimoto,I. ed., Points of Design of Threaded Joints (revised ver.), (2002), Chap. 6, Japanese
Standards Association. (in Japanese)
(4)Wiegand,H. et al., Schraubenverbindungen, (1988), Kapitel 9, Springer.


Journal of Advanced Mechanical Design,
Systems, and
Manufacturing
Vol. 1, No. 3, 2007
367
(5)Junker,G.H., New Criteria for Self-Loosening of Fasteners under Vibration, SAE Transactions,
No.690055, pp.314-335.
(6)Yamamoto,A. and Kasei,S., Investigations on the Self-loosening of Threaded Fasteners under
Transverse Vibration A Solution for Self-loosening Mechanism, Bulletin of the Japan
Society of Precision Engineering, Vol.18, No.3, (1984), pp.261-266. (originally in Japanese)
(7)Yamamoto,A. et al., Investigations on the Self-loosening of Threaded Fasteners under
Transverse Vibration Theorization of Locking Performance Curve, Journal of the Japan
Society of Precision Engineering, Vol.43, No.9, (1977), pp.1069-1074. (in Japanese)
(8)Zadoks,R.I. and Yu,X., An Investigation of the Self-loosening Behavior of Bolts under
Transverse Vibration, Journal of Sound and Vibration, Vol.208, No.2, (1997), pp.189-209.
(9)Wolfsteiner,P. and Pfeiffer,F., Selbstttiges Losdrehen von Schraubenverbindungen unter
Dynamischen Belastungen, VDI Berichte, Nr.1426, (1998), pp.191-204.
(10)Pai,N.G. and Hess,D.P., Experimental Study of Loosening of Threaded Fasteners due to
Dynamic Loads, Journal of Sound and Vibration, Vol.253, No.3, (2002), pp.585-602.
(11)Pai,N.G. and Hess,D.P., Three-dimensional Finite Element Analysis of Threaded Fastener
Loosening due to Dynamic Shear Load, Engineering Failure Analysis, No.9, (2002),
pp.383-402.
(12)Jiang,Y. et al., A Study of Early Stage Self-Loosening of Bolted Joints, Transactions of the
ASME, Journal of Mechanical Design, Vol.125, (2003), pp.518-526.
(13)Jiang,Y. et al., An Experimental Study of Self-Loosening of Bolted Joints, Transactions of
the ASME, Journal of Mechanical Design, Vol.126, (2004), pp.925-931.
(14)Izumi,S. et al., Three-dimensional Finite Analysis of Tightening and Loosening Mechanism
of Threaded Fastener, Engineering Failure Analysis, No.12, (2005), pp.604-615.
(15)Izumi,S. et al., Small Loosening of Bolt-nut Tightening System due to Micro
Bearing-surface Slip: Finite Element Method Study, Transactions of the Japan Society of
Mechanical Engineers, Series A, Vol.72, No.717, (2006), pp.780-786. (in Japanese)
(16)Kasei,S. et al., On Self-loosening of Threaded Joints in the Case of Absence of Macroscopic
Bearing-surface Sliding Loosening Mechanism under Transversely Repeated Force,
Bulletin of the Japan Society of Precision Engineering,Vol.23, No.1, (1989), pp.31-36.
(originally in Japanese)
(17)Kasei,S. and Matsuoka,H., Considerations of Thread-loosening by Transverse Impact,
Analysis of Bolted Joints (The 1998 ASME/JSME PVP Conference), (1998), pp.117-123.
(18)Kasei,S. and Sun,P., Fundamental Considerations on Thread-loosening due to Transverse
Impact, Transactions of the Japan Society of Mechanical Engineers, Series C, Vol.60, No.570,
(1994), pp.625-631. (in Japanese)
(19)Junker,G. and Strelow,D., Untersuchungen ber die Mechanik des selbstttigen Lsens und
die zweckmige Sicherung von Schraubenverbindungen (), Draht-welt, Vol.52, Nr.5,
(1969), pp.317-335.