Research on Self-Loosening Mechanism of Bolted Joints ...
Transcript of Research on Self-Loosening Mechanism of Bolted Joints ...
RESEARCH ON SELF-LOOSENING MECHANISM OF BOLTED JOINTS
UNDER TRANSVERSAL VIBRATION
Wei Wang
State Key Laboratory for Manufacturing System
Engineering, Xi'an Jiaotong University, 710049,
School of Air and Missile Defense Air Force
Engineering University, 710051,
Xi'an ,China
Qingli Wang
School of Air and Missile Defense,
Air Force Engineering University, 710051,
Xi'an ,China
Shuai Zheng
State Key Laboratory for
Manufacturing System Engineering,
Xi'an Jiaotong University, 710049,
Xi'an ,China
Xiaowei Liu
School of Air and Missile Defense ,
Air Force Engineering University,
710051,Xi'an ,China
Haiping Liu
School of Air and Missile Defense,
Air Force Engineering University,
710051,Xi'an ,China
ABSTRACT
Bolted joints are widely used in the auto industry, energy
field and Construction, and so on. Due to the wide use of the
bolted joints, the degradation of bolts has significant effect on
the performance of a whole machine. Under transversal
vibration, the self-loosening of bolted joints, which is the
biggest form of failure ranked only second to fatigue failure
[1], will happen, due to the cyclic shear load. This paper is to
study the mechanism of bolted joints’ self-loosening.
Aiming at analyzing the self-loosening mechanism of
bolted joints under vibration, a three dimensional FEA model
of bolted joints , which had taken thread into consideration,
was built with the application of APDL, and the preload was
applied on the bolted joints by dropping temperature, then
FEA transient analysis of the bolted joints under transverse
cyclic excitation was conducted. Effect of transverse cyclic
excitation’s amplitude, initial preload, thread and bearing
friction coefficients, the joints’ surface friction coefficient, the
thread pitch and the hole clearance on self-loosening was
investigated. The results show that the complete thread slip
occurs prior to the complete bearing surface slip under
transverse vibration; the smaller amplitude, the smaller thread
pitch and the smaller hole clearance is, and the greater initial
preload, thread and bearing friction coefficients are, the more
difficult self-loosening is to happen; the joints’ surface friction
coefficient has little relationship with self-loosening, however,
the larger joints’ surface friction coefficient makes the needed
shearing force, which induces the transversal vibration, larger.
These are of great significance for understanding of fasteners’
self-loosening and designing of bolted joints’ anti-loosening.
Keyword: Bolted joints; Vibration; Transient Analysis;
Preload; Friction Coefficient
INTRODUCTION
The self-loosening of bolted joints will cause leakage,
failure due to the uneven stress, the lowering of the complete
machine dynamic performance, and other bad effects, even
lead to accident. Thus it is of significant for studying on
self-loosening.
1 Copyright © 2014 by ASME
Proceedings of the ASME 2014 International Mechanical Engineering Congress and Exposition IMECE2014
November 14-20, 2014, Montreal, Quebec, Canada
IMECE2014-37971
The self-loosening and anti-loosening of bolted joints has
been studied by many scholars. Junker [2]designed the famous
tester, called Junker tester, to study the loosening behavior of
bolted joints and found that the bolted joints under transversal
vibration relaxed more easily, due to the transversal cyclic
shearing load, than under other conditions. Hess and his
co-workers [3] [16] studied the self-loosening of bolted joints
under transversal cyclic shearing load by using experimental
and FEA method, and they put forward that there are four
apparent stages in the self-loosening. Nassar, Housari and
other co-workers [4]-[6] studied the effects of the thread pitch,
the initial preload, the hole clearance, the thread and bearing
friction coefficients on self-loosening through the application
of line linear model. Then, Nassar, Yang and other co-workers
[1], [7]-[11] proposed a more accurate analytical model to
analyze the self-loosening under transversal vibration, and
designed a series of corresponding experiments to validate the
analytical model.The scholars mentioned above got plenty of
achievement on self- loosening study, however, both their
theoretical models and the experimental models were based
on the Junker tester, which did not match the actual for
ignoring the frictional condition of the joints’ surface.
Jiang and his co-workers [12]-[14] designed a new
experimental facility to study on self-loosening. They studied
the self-loosening of bolted joints under transversal cyclic
load by using experimental and FEA method and proposed
that self-loosening should be divided into two stages: the first
loosening due to the short-time and acute plastic ratchet
effect of the material of thread, the second loosening due to
the clamping force loss caused by the nut’s turning back. Jiang
and his co-workers considered the frictional condition of the
joints’ surface, nevertheless, they didn’t consider the spiral
effect of the thread in their finite element simulation,so their
study cannot explain the self-loosening due to nut’s turning
back clearly. Yang Guangxue and her co-workers [15] studied
the effects of the addition bending moment, initial preload and
anti-loosening nut on loosening of bolted joints under
transversal vibration with the application of 3D FEA method
in their research on anti-loosening mechanism of a new type
of nut.
On the foundation of the former sholars, this paper built a
three-dimensional finite element model of bolted joints,
applied preload on the bolted joints model by using cooling
pre-tightening algorithm, then conducted the FEA transient
analysis of the bolted joints under transverse cyclic excitation,
and investigated the effects of the transverse cyclic
excitation’s amplitude, the initial preload, the thread and
bearing friction coefficients, the joints’ surface friction
coefficient, the thread pitch and the hole clearance on
self-loosening.
CORRELATIVE ALGORITHM
Augmented Lagrangian contact algorithm
Contact algorithm is used to dealing with the relationship
of the interaction between the contact body. Effective contact
algorithm should be able to pass the contact force to ensure
that there is no penetration between the two contact surfaces.
General contact algorithms in ANSYS are Penalty Function
algorithm, Lagrange algorithm and Augmented Lagrangian
algorithm. The Augmented Lagrangian algorithm is an
algorithm with easily converging and accurately calculating
based on the former kinds of algorithms.Before calculating
each load step, it will check the penetration between the
contact surface and the target surface and manage the
penetration. when there is a penetration , it will set up a
normal contact force between the contact surface and the
target surface, the size of the force is determinded by the
component proportionately to the contact stiffness and
penetration size, and the Lagrange coefficient, being
equivalent to set a spring which can be adjust its stiffness
between the two surfaces to limit the size of the penetration
and to prevent pathological integral stiffness, which will lead
to the solving divergence, caused by excessive contact
stiffness matrix. The algorithm is described as follows.
The finite element contact equation:
[ ][ ] { } { }K u P R (1)
In Eq.(1), [ K ] is the stiffness matrix, [ u ] is the unit
displacement matrix, {P} is the unit load vector, {R} is the
unit contact force vector.
The portion iR of {R} is determined by Augmented
Lagrangian contact algorithm, as followed:
1
0 , 0, 0i
n i
Rk
(2)
In Eq.(2), nk is normal contact stiffness, i is the
space between the contact surface and target surface, 1i is
Lagrange coefficient. If 0 , which means penetration,
coefficient 1i will modify nk according to the actual
contact stiffness, the concrete can be determined by Lagrange
iterative method:
1
,
,i n
ii
k e
e
(3)
2 Copyright © 2014 by ASME
In Eq.(3),i is the Lagrange coefficient of the previous
step, e is the contact tolerance (the allowed maximum
penetration for convergence).
So we can solve equation (1), and control penetration
and transfer contact force between two contact surfaces.
Cooling pre-tightening algorithm
In FEA, the general pre-tightening algorithm are
Preload Element algorithm, Equivalent Force algorithm and
Cooling algorithm.The pretightening unit can't bear shear
load,so it cannot be used in the transient analysis under
transverse vibration; the loading by Equivalent Force
algorithm is complex, and it cannot simulate the response of
the joints under preloading [17]. To pre-tighten the bolt joints
with shear force, the Cooling algorithm must be used. Cooling
algorithm: setting the bolt material thermal expansion
coefficient, then lowering the temperature of the
pre-tightening bolt part, the bolt will elongate, while the joints
will restrict the deformation of the bolt,which will cause the
inside tension of the bolt ,so the pretightening is simulated.
The algorithm is described as follows.
As shown in figure 1, in the free stage, the temperature
of the bolt part (length l ), which will be pretightened, is
lowered for T . There will be a contraction deformation,
whose size is 0l , due to the heat bilges cold shrink effect. In
the cooling preloaded stage, the existence of joints reduces the
deformation of bolt, there will be a clamping force between
bolt and joints, this force decrease bolt shrinkage deformation
for 1l and cause the joints deformation for 2l .
l
1l
2l0l
Initial stage Cooling pre-tightening Free cooling Fig.1: The deformation corresponding relationship of bolted
joints under cooling pre-tightening
As the figure shows, the axial deformation of bolt joints
has a deformation corresponding relationship:
0 1 2l l l (4)
According to the heat-expansion and cold- contraction,
there is a relationship:
0 0l l T (5)
In Eq.(5), 0 is the linear expansion coefficient of bolt
material.
As the deformation relationship of the bolt and joints
under pre-tightening shows:
1 / bl F K (6)
2 / cl F K (7)
In the above , F is the pre-tightening force on bolt
connection, bK is the stiffness of bolt,
cK is the stiffness of
the joints.
Taking Eq.(4), (5) and Eq.(6) into Eq.(7), and calculating
them , the lowering temperature of bolts when loaded
prestressing force F will be given:
0
1 1( )
b c
FT
K K l (8)
STUDYING ON THE SELF-LOOSENING SIMULATION
Finite element analysis modeling
In order to study the self-loosening mechanism of bolted
joints under transversal vibration based on FEA, the
corresponding finite element model must consider the spiral
effect of the thread, in other words, a three dimensional FEA
model of bolted joints , which takes thread into consideration,
must be built.
In this paper, the ANSYS parametric language is used
to build the FEA model of bolted joints. As Fig.2 shows, the
bolt is made up of two parts: haulm and thread, and is meshed
after using the command VGLUE. The FEA model of the nut
can be obtained in the similar way. The sizes of the model are
follows: the nominal diameter D is 12mm, the bolt head
diameter D1 is 16.6mm, the height of the bolt head KW is
7.5mm, the height of the nut H is 13mm, the thread pitch p is
1.75mm, the length of the bolt L is 64mm, joints’ sizes is
80mm×80mm×20mm(both boards are the same), the hole
clearance ' / 2 is 0.5mm, the profile angle is 60 degree. In
the model, the bolt material is high strength steel, whose
Young’s modulus 1E is 210GPa, Poisson ratio 1 is 0.3,
density 1 is 7.9×103Kg/m3, yield limit is 640Mpa. While the
joints material is Q235 steel, whose Young’s modulus 2E is
210GPa, Poisson ratio 2 is 0.3, density 2 is 7.9×103Kg/m3,
yield limit is 235Mpa.
3 Copyright © 2014 by ASME
Fig.2: Three dimensional FEA model of bolted joints
Constraints applying
This paper consider the slippage between contact
surfaces, so the contact pairs is built by using TARGE170 as
the target unit and the CONTA173 as the contact unit. When
building the contact pairs, the friction coefficients of the
thread engagement surfaces, bearing surfaces and joints
surfaces is defined as 1 、 2 and 3 by defining material
friction factor. Each friction coefficients is obtained from
literature [6] and literature [16], namely the paraffin
lubrication (0.05), the MoS2 grease lubrication (0.1),
mechanical lubricating oil (0.17) and dry friction (0.2).
Bolted joints pre-tightening
In the static solver, the normal temperature is setted by
using command TREF, the temperature load according to
Eq.(8) is loaded on the bolt by using command BFV to
pre-tighten the bolt. Fig.3 shows the stress nephogram of nut
thread after preloaded.As shows in the figure, the first circle
thread stress is the largest, the followed decreases one by one,
which fits to the actual stress distribution of bolt joints.
The
first
circle
Fig.3: The stress nephogram of nut thread after preloaded
Transversal vibration transient analyzing
In order to conduct the FEA of bolted joints self-loosening,
the transversal excitation x should be loaded and the transient
analysis should be conducted. The x is determined by the
formula:
0 sin( )x t (9)
In Eq.(9), the excitation amplitude is 0 ('
0 / 2 ),
is angular frequency.
Concretely, the joint surface friction coefficients are
setted as 1 2 3 =0.1, the FEA model of bolted joints,
showed in figure 1, is pre-tightened by the load of
F(F=10730N) to conduct static analysis. After the solution, the
displacement load 0.2sin(3600 )x t is loaded, and the
transient analysis is conducted for time step t=0.0125s and
end time=0.325s.
ANALYSIS OF SELF-LOOSENING MECHANISM
Analysis of self-loosening process
The curve in the Figure 4 shows the shearing load
changing over transverse displacement, called shearing load
hysteretic curve. As the transversal stiffness of bolted joints
can be explained by 0/transverse shearingK F ( shearingF means
the biggest shearing load, which equals to the half the vertical
span of the curve), thus the transversal stiffness of bolted
joints can be measured by the vertical span of the curve: the
bigger the vertical span is, the bigger the transversal stiffness
is.
O
A
B
C
D
-0.2 -0.15 -0.1 -0.05 0 0.05 0.1 0.15 0.2-2
-1.5
-1
-0.5
0
0.5
1
2
1.5
Th
e sh
eari
ng
lo
ad/K
N
Transverse displacement/mm
Fig.4: The relationship between shearing load and transverse
displacement
What shows in Figure 5 are the contact status of thread
engagement surface and bearing surface at point A, B, C and
D in Figure 4. In the figure, we can know that the thread
engagement surface slips completely, while the bearing
surface slips incompletely at point A and C, and the slippage
parts of bearing surface at point A and C are opposite,
suggesting that the bearing surface can slip in a whole circle.
However, both thread engagement surface and bearing surface
slip incompletely at point B and D. It suggests that the
slippage at the amplitude of transverse displacement is the
acutest, and the complete slippage of thread engagement
surface happens before that of bearing surface.
Near contact slip stick
(a)Point A (b)Point B
4 Copyright © 2014 by ASME
(c)Point C (d)Point D
Fig.5: The contact status of thread and nut bearing surface
The Curve in the Figure 6 shows the residual preload
change over time. In the figure, we can know that the preload
loses a little in a circle (for decreasing only 0.7% per circle).
Integrating with the Figure 5, we can know that the
incompletely slippage can also cause relaxation
0 0.05 0.1 0.15 0.2 0.25 0.3 0.3510
10.1
10.2
10.3
10.4
10.5
10.6
10.7
10.8
Res
idue
pre
load
/KN
Time/s
Fig.6: Residue preload VS time
Effect of excitation amplitude on self-loosening
In order to investigate into the effect of amplitude of
transversal vibration on self-loosening, referring to the
simulation of section 3.4, we only modify the parameter 0
to 0.02, 0.1, 0.2, 0.3 and 0.4. Then the comparison simulation
of bolted joints with different excitation amplitude under
transverse vibration is conducted. The curves in the Figure 7
show residual preload change for different excitation
amplitudes. In the figure, we can know that the curves of 0.02
and 0.1 are nearly parallel with the coordinate axis, which
indicates that there is little preload loss when the excitation
amplitude is 0.02 and 0.1. However, the preload loss is very
visible when the excitation amplitude is 0.2, 0.3 and 0.4. It
suggests that the self-loosening of bolted joints can happen
only when the excitation amplitude is big enough. When
comparing the preload loss of 0.2, 0.3 and 0.4, we can know
that the greater the amplitude is, the more likely to happen the
self-loosening is. The Figure 8 shows shearing load hysteretic
curves under different amplitudes. In the figure, we can know
that the greater the amplitude is, the bigger the shearing load
needed is.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.359.4
9.6
9.8
10
10.2
10.4
10.6
10.8
Pre
load
/KN
Time/s
0.020.10.20.30.4
Fig.7: Residue preload VS time under different amplitudes
-0.4 -0.3 -0.2 -0.1 0 0.2 0.3 0.40.1-3
-2
-1
0
1
2
3
The
sher
ing l
oad
/KN
Transvers displacement/mm
0.020.10.20.30.4
Fig.8: Shearing load hysteretic curves under different
amplitudes
Effect of initial preload on self-loosening
In order to investigate into the effect of initial preload on
self-loosening, referring to the simulation of section 3.4, we
only modify the parameter F to 3, 10.73, 21.32 and
29.93.Then the comparison simulation of bolted joints with
different initial preloads under transverse vibration is
conducted. The curves in the Figure 9 show the losing preload
percentage change for different initial preloads. In the figure,
we can know that the greater initial preload is, the smaller
percentage of preload loss is, the less likely to happen the
self-loosening is. The Figure 10 shows shearing load
hysteretic curves under different initial preloads. In the figure,
we can know that the greater initial preload is, the bigger
vertical span is. It suggests that the greater initial preload is,
the greater transversal stiffness of bolted joints is.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
s-1
0
1
2
3
4
5
67
Per
centa
ge
of
losi
ng p
relo
ad/%
Time/s
3KN10.73KN21.32KN29.93KN
Fig.9: Curves of losing preload percentage under different
initial preloads
5 Copyright © 2014 by ASME
-0.2 -0.15 -0.1 -0.05 0 0.1 0.15 0.20.05-5
-4
-3
-2
-1
0
1
2
3
45
The
sher
ing l
oad
/KN
Transvers displacement/mm
3KN10.73KN21.32KN29.93KN
Fig.10: Shearing load hysteretic curves under different initial
preloads
Effect of coefficients on self-loosening
(1) Effect of coefficient1 on self-loosening
In order to investigate into the effect of thread
engagement surface friction coefficient (1 ) on
self-loosening, referring to the simulation of section 3.4, we
only modify the parameter 1 to 0.05, 0.1, 0.17 and
0.2.Then the comparison simulation of bolted joints with
different thread engagement surface friction coefficient under
transverse vibration is conducted. The curves in the Figure 11
show residual preload change. In the figure, we can know
there are little preload loss when the coefficient is 0.17 and
0.2. However, the preload loss is very visible when the
coefficient is 0.05 and 0.1. Because the bigger coefficient is,
the bigger frictional force is, and the less likely to happen the
self-loosening is. The Figure 12 shows shearing load
hysteretic curves under different thread engagement surface
friction coefficients. In the figure, we can know that the
shearing load hysteretic curves are nearly the same. It suggests
that the thread engagement surface friction coefficient has
small effect on the transversal stiffness of bolted joints.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35
10.3
10.4
10.5
10.6
10.7
Res
idue
pre
load
/KN
s
Time/s
0.050.10.170.2
Fig.11: Residue preload VS time under different 1
-0.2 -0.15 -0.1 -0.05 0 0.1 0.15 0.20.05-2.5
-1.5
-1
-0.5
0
0.5
1
1.5
22.5
-2
0.050.10.170.2
Th
e sh
erin
g l
oad
/KN
Transvers displacement/mm
Fig.12: Shearing load hysteretic curves under different 1
(2) Effect of coefficient 2 on self-loosening
In order to investigate into the effect of bearing surface
friction coefficient ( 2 ) on self-loosening, referring to the
simulation of section 3.4, we only modify the parameter 2
to 0.05, 0.1, 0.17 and 0.2.Then the comparison simulation of
bolted joints with different bearing surface friction coefficient
under transverse vibration is conducted. The curves in the
Figure 13 show residual preload change. In the figure, we can
know that the bigger coefficient is, the greater preload loss is,
when the bearing surface friction coefficient is 0.1, 0.17 and
0.2. However, there is little preload loss when the bearing
surface friction coefficient is such small as 0.05, which is
contrary to other scholars’ results, but is also correct. Because
the bolt head’s and nut’s DOFs are not constrained in this
paper, but constrained in other scholars’ analysis to fit with
Junker test. When the bearing surface friction coefficient is
very small, the slippage, caused by transversal movement of
the board, will be very slight, thus the self-loosening speed is
very slow. The Figure 14 shows shearing load hysteretic
curves under different bearing surface friction coefficients. In
the figure, we can know that the bearing surface friction
coefficient has as small effect as thread engagement surface
friction coefficient on the transversal stiffness of bolted joints.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.3510.3
10.4
10.5
10.6
10.7
10.8
s
Res
idu
e p
relo
ad/K
N
Time/s
0.050.10.170.2
Fig.13: Residue preload VS time under different 2
-0.2 -0.15 -0.1 -0.05 0 0.1 0.15 0.20.05-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
0.050.10.170.2
Th
e sh
erin
g l
oad
/KN
Transvers displacement/mm
Fig.14: Shearing load hysteretic curves under different 2
(3) Effect of coefficient 3 on self-loosening
In order to investigate into the effect of joints’ surface
friction coefficient ( 3 ) on self-loosening, referring to the
simulation of section 3.4, we only modify the parameter 3
to 0.05, 0.1, 0.17 and 0.2.Then the comparison simulation of
bolted joints with different joints’ surface friction coefficient
under transverse vibration is conducted. The curves in the
Figure 15 show residual preload change. In the figure, we can
know these curves are nearly the same. It suggests that the
joints’ surface friction coefficient nearly has no effect on
self-loosening under the same transversal vibration amplitude.
The Figure 16 shows shearing load hysteretic curves under
different joints’ surface friction coefficients. In the figure, we
6 Copyright © 2014 by ASME
can know that the bigger the joints’ surface friction coefficient
is, the bigger vertical span is. It suggests that the bigger the
joints’ surface friction coefficient is, the bigger transversal
stiffness of bolted joints is, the less likely to happen
self-loosening is.
0 0.05 0.1 0.15 0.2 0.25 0.310.3
10.4
10.5
10.6
10.7
s0.35
0.05
0.2
0.10.17
Res
idu
e p
relo
ad/K
N
Time/s
Fig.15: Residue preload VS time under different3
-0.2 -0.15 -0.1 -0.05 0 0.1 0.15 0.20.05-3
-2
-1
0
1
2
30.05
0.2
0.10.17
The
sher
ing l
oad
/KN
Transvers displacement/mm
Fig.16: Shearing load hysteretic curves under different 3
Effect of thread pitch on self-loosening
In order to investigate into the effect of thread pitch on
self-loosening, referring to the simulation of section 3.4, we
only modify the parameter p to 1.5(Fine thread) and
1.75(Coarse thread).Then the comparison simulation of bolted
joints with different pitches under transverse vibration is
conducted. The curves in the Figure 17 show residual preload
change for different thread pitches. In the figure, we can know
that the preload of fine thread loses slower than the coarse
thread’s. It suggests that the anti-loosening performance of
fine thread is better than coarse thread. The Figure 18 shows
shearing load hysteretic curves under different thread pitches.
In the figure, we can know that the shearing load hysteretic
curve of fine thread has the same vertical span with the coarse
thread. It suggests that the thread pitch has small effect on the
transversal stiffness of bolted joints.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.3510.3
10.4
10.5
10.6
s
10.7
Res
idu
e p
relo
ad/K
N
Time/s
fine
coarse
Fig.17: Residue preload VS time under different thread pitches
-0.2 -0.15 -0.1 -0.05 0 0.1 0.15 0.20.05-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
Th
e sh
erin
g l
oad
/KN
Transvers displacement/mm
fine
coarse
Fig.18: Shearing load hysteretic curves under different thread
pitches
Effect of hole clearance on self-loosening
In order to investigate into the effect of hole clearance on
self-loosening, referring to the simulation of section 3.4, we
only modify the parameter ' / 2 to 0.5, 0.75 and 1.25. Then
the comparison simulation of bolted joints with different
clearance under transverse vibration is conducted. The curves
in the Figure 19 show residual preload change for different
hole clearance. In the figure, we can know that the bigger the
hole clearance is, the faster the preload loses, which proves
that the better the assemblage is, the better anti-loosening
performance is. The Figure 20 shows shearing load hysteretic
curves under different hole clearance. In the figure, we can
know that the shearing load hysteretic curve of bigger hole
clearance has the smaller vertical span than the smaller ones’.
It suggests that the bigger hole clearance has smaller
transversal stiffness of bolted joints. At the same time, the
difference between three shearing load hysteretic curves is not
unconspicuous. It suggests that the hole clearance has small
effect on the transversal stiffness of bolted joints.
0 0.05 0.1 0.15 0.2 0.25 0.3 0.3510
10.1
10.2
10.3
10.4
10.5
10.6
10.8
10.7
Res
idue
pre
load
/KN
Time/s
0.751.25
0.5
Fig.19: Residue preload VS time under different hole clearance
-0.2 -0.15 -0.1 -0.05 0 0.1 0.15 0.20.05-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
Th
e sh
erin
g l
oad
/KN
Transvers displacement/mm
0.751.25
0.5
Fig.20: Shearing load hysteretic curves under different hole
clearance
CONCLUSIONS
The self-loosening mechanism of bolted joints under
7 Copyright © 2014 by ASME
transversal vibration is studied based on FEA. And the
transverse vibration amplitude, the initial preload, thread and
bearing friction coefficients, the joints’ surface friction
coefficient, the thread pitch and the hole clearance on
self-loosening is investigated. The results show that:
(1)Both the slippage of thread engagement surface and
the slippage of bearing surface can lead to the preload loss of
bolted joints under transversal vibration. The complete
slippage of thread engagement surface is priority to bearing
surface.
(2)The greater the vibration amplitude is, the more likely
to happen the self-loosening is.
(3)The greater initial preload is, the smaller percentage of
preload loss is, the less likely to happen the self-loosening is.
(4)The greater thread and bearing friction coefficients is,
the less likely to happen the self-loosening is. The joints’
surface friction coefficient has small influence on self-
relaxation under the same transversal vibration. But it has a
big impact on the transversal stiffness of bolted joints.
(5)The self-loosening of bolt with coarse thread is more
likely to happen than that with fine thread.
(6)The paper proves that the better the assemblage, the
better locking performance of fine thread.
In conclusion, when designing bolted joints’
anti-loosening, we should put how to change the thread
engagement surface friction first. We should also consider
bearing surface friction. We should give the biggest preload
within the strength and increase joints’ surface friction without
leakage.
All in all, this paper is of great significance for
understanding of fasteners’ self-loosening and designing of
bolted joints’ anti-loosening.
ACKNOWLEDGMENT
This work is supported by Postdoctoral Science
Foundation of China (No.2014M552430) and a grant from the
National High Technology Research and Development
Program of China (863 Program) (No.2013AA040604).
REFERENCE
[1] X.Yang, S.A.Nassar, Z,Wu . Criterion for Preventing
Self-Loosening of Preloaded Cap Screws Under Transverse
Cyclic Excitation[J]. Journal of Vibration and Acoustics,
ASME, 2011,133(8):041013(1-11).
[2] G.H.Junker.New Criteria for Self-Loosening of Fasteners
UnderVibration[J].SAE Trans., 1969,78:314–335.
[3] N.G.Pai, D.P.Hess . Three-dimensional Finite Element
Analysis of Threaded Fastener Loosening due to Dynamic
Shear Load[J] . Engineering Failure Analysis,
2002,9:383-402.
[4] S.A.Nassar, B.A.Housari.Effect of Thread Pitch and Initial
Tension on the Self-loosening of Threaded
Fasteners[J].Journal of Pressure Vessel Technology, ASME,
2006,128(11):590-598.
[5] S.A.Nassar, B.A.Housari . Study of the Effect of Hole
Clearance and Thread Fit on the Self-Loosening of Threaded
Fasteners[J] . Journal of Mechanical Design, ASME,
2007,129(6):586-594.
[6] B.A.Housari, S.A.Nassar . Effect of Thread and Bearing
Friction Coefficients on the Vibration-Induced Loosening of
Threaded Fasteners[J].Journal of Vibration and Acoustics,
ASME, 2007,129(8):484-494.
[7] S.A.Nassar, X.Yang.A Mathematical Model for Vibration
Induced Loosening of Preloaded Threaded
Fasteners[J].Journal of Vibration and Acoustics, ASME,
2009, 131(4): 021009(1-13).
[8] X.Yang, S.A.Nassar . Vibration-Induced Loosening
Performance of Preloaded Threaded Fasteners
[J].Proceedings of PVP, ASME, 2010, PVP2010-25811.
[9] X.Yang, S.A.Nassar.Effect of Thread Profile Angle and
Geometry Clearance on the Loosening Performance of a
Preloaded Bolt-nut System under Harmonic Transverse
Excitation[J] . Proceedings of PVP, ASME, 2011,
PVP2011-57690.
[10] X.Yang, S.A.Nassar . Analytical and Experimental
Investigation of Self-Loosening of Preloaded Cap Screw
Fasteners[J].Journal of Vibration and Acoustics, ASME,
2011,133(6):031007(1-8).
[11] A.M.Zaki, S.A.Nassar, X.Yang.Effect of Conical Angle and
Thread Pitch on the Self-Loosening Performance of Preloaded
Countersunk-Head Bolts[J] . Journal of Pressure Vessel
Technology, 2012,134(4):021210(1-8).
[12] Y.Jiang,M.Zhang,C.H.Lee . A Study of Early Stage
Self-Loosening of Bolted Joints[J].Journal of Mechanical
Design, ASME, 2003,125(9):518-526.
[13] Y.Jiang,M.Zhang,T.W.Park,et al.An Experimental Study of
Self-Loosening of Bolted Joints[J].Journal of Mechanical
Design,ASME,2004,126(9):925-931.
[14] Y.Jiang,M.Zhang,C.H.Lee . Finite Element Modeling of
Self-Loosening of Bolted Joints[J].Journal of Mechanical
Design,ASME,2007,129(2):218-226.
[15] Yang Guangxue, Xie Jilong, Xie Yunye . Study on
Mechanism of Anti-loosening of A New Type of Nut Based on
FEM[J] .Engineering Mechanics, 2010, 27(12): 224-228,
243.( in Chinese)
[16] N.G.Pai, D.P.Hess.Experimental Study of Loosening of
Threaded Fasteners due to Dynamic Shear Loads[J].Journal
of Sound and Vibration, 2002,253(3):585-602.
[17] Jin Jing, Wu Xingyue, Zheng Jianhua.Bolt joint prestress
8 Copyright © 2014 by ASME