Anda di halaman 1dari 8

European Journal of Mechanics A/Solids 29 (2010) 675–682

Contents lists available at ScienceDirect

European Journal of Mechanics A/Solids


journal homepage: www.elsevier.com/locate/ejmsol

An investigation about interference fit effect on improving fatigue life of a holed


single plate in joints
T.N. Chakherlou a, M. Mirzajanzadeh a, B. Abazadeh a, *, K. Saeedi b
a
Faculty of Mechanical Engineering, University of Tabriz, Iran
b
Mechanical Engineering Department, University of Concordia, Canada

a r t i c l e i n f o a b s t r a c t

Article history: In this article, the effect of interference fit on fatigue life of holed plate of mechanical joints was
Received 18 October 2008 investigated both experimentally and numerically. In the experimental part, fatigue tests were carried
Accepted 31 December 2009 out on the holed specimens of Al 7075-T6 alloy in which oversized steel pins were force fitted to them.
Available online 18 January 2010
These fatigue tests were conducted on open hole specimen and specimens with 1, 1.5, 2 and 4% nominal
interference fit sizes at different cyclic longitudinal loads. From these tests the stress-life (S–N) data for
Keywords:
different interference fit sizes were obtained. The results show that interference fit increases fatigue life
Interference fit
compared to open hole specimens. In the numerical part of the investigation, 3D finite element simu-
Fatigue
Mechanical joint lations have been performed to obtain stress (or strain) histories and distributions around the hole due to
Finite element interference fit and subsequent cyclic longitudinal loading using FEM package. The stress history from
finite element (FE) simulation was used to explain the reason for fatigue life improvement in the
interference fitted specimens.
Ó 2010 Elsevier Masson SAS. All rights reserved.

1. Introduction fastener hole of the components or plates is cold expanded using an


oversized mandrel or pin before assembling. During cold expansion
Fatigue is one of the main agents for failure of mechanical process the oversized mandrel is inserted to the hole from one side
components or plates in aerospace and automotive industries of the plate and removed from the other side. This process creates
when the components are subjected to cyclic loading (Wuanlin and compressive residual tangential stress around the fastener hole
Xiangjiong, 1991). Different components in aerospace structures (Chakherlou et al., 2008). This stress reduces the resultant tensile
are jointed together using rivets or bolts so as to be easily stress around the fastener hole when the components are subjected
dismantled, repaired or replaced by a new part. Furthermore to tensile cyclic loading. However, in the interference fit, the
experiments show that the fatigue resistance of rivet or bolt joints fastener that could be an oversized pin or bolt is inserted to the hole
are better than welded joints (Valtinat et al., 2000). However, in during component assembling (unlike cold expansion technique).
bolt joints the components must be drilled to have holes. These This process leaves mainly tensile tangential residual stress around
holes cause geometrical discontinuities and entail local stress (or the hole (Dowling, 1999). Although interference fit (especially for
strain) concentration during loading. Furthermore drilling may low level) creates mainly tensile tangential stress and so increases
create rough surface or damage the components thus making them mean stress in cyclic tensile loading, but as shown in Fig. 1, it
more prone to fatigue crack initiation and propagation (Chakherlou considerably reduces the stress (or strain) amplitude, which is
et al., 2008; Crew, 1975; Iyer et al., 2001; Rich and Impellizzeri, more influential than mean stress in fatigue crack initiation or crack
1977; Chakherlou and Vogwell, 2003). In order to compensate propagation (Wuanlin and Xiangjiong, 1991).
these drawbacks and increase fatigue resistance in holed compo- There are scatter studies about the effect of interference fit on
nents, different fatigue life improving techniques such as cold fatigue life of bolted or pinned joints in literature (Crew, 1975; Iyer
expansion or interference fit are used. In these techniques either et al., 2001; Rich and Impellizzeri, 1977; Mann and Jost, 1983;
compressive or tensile tangential (circumference) residual stress is Sundarraj et al., 1995; Jost and Carey, 1984; Finney, 1993). However,
introduced around the fastener holes. In cold expansion, the in these studies the direct (or isolated) effect of interference fit was
shadowed by either cold expansion or bolt clamping effect or both
of them. Therefore, it is not clear what part of fatigue life
* Corresponding author. Tel.: þ98 914 4002567. improvement was only due to interference fit. As the isolated effect
E-mail address: abazadeh@tabrizu.ac.ir (B. Abazadeh). of cold expansion and bolt clamping force on fatigue life were

0997-7538/$ – see front matter Ó 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.euromechsol.2009.12.009
676 T.N. Chakherlou et al. / European Journal of Mechanics A/Solids 29 (2010) 675–682

Fig. 3. s–3 curve of Al 7075-T6.


Fig. 1. Stress variations before and after interference fit in hole edge.

studied in the previous researches by the authors (Chakherlou however, standard tests were performed on tensile specimens to
et al., 2008; Chakherlou and Vogwell, 2003), in the case of knowing obtain s–e curve and Poisson’s ratio. The true stress versus the true
the isolated effect of interference fit, it will be much easier to strain curve obtained from tests is shown in Fig. 3. The elasticity
a mechanical joint designer to decide which combination of them modulus, Poisson’s ratio and hardness were measured to be
will create the best joint from fatigue life point of view. So in this E ¼ 71.5 GPa, y ¼ 0.33 and 188 HV, respectively. Also the yield
paper in order to investigate the isolated effect of interference fit on strength and ultimate strength of the Al alloy were 503 MPa and
fatigue life improvement, a holed plate with an oversized pin force 600 MPa, respectively.
fitted was studied experimentally and numerically. In the experi- In experimental tests, in addition to Al-alloy 7075-T6, steel pins
mental part four batches of Al 7075-T5 alloy specimens with four were also used. These pins, which force fitted to fatigue specimens’
different sizes of interference fit also one batch of open hole were hole to create interference fit, were made from AISI-D2 steel with
fatigue tested to obtain S–N data. In the numerical part, contrary to mechanical properties of E ¼ 207 GPa, y ¼ 0.26, hardness ¼ 62 HRC
2D FEM models investigated in the literature, 3D finite element (746 HV) and yield strength of more than 1000 MPa. The dimen-
method (FEM) models were used to capture stress and strain sions of the pin which caused a 2% interference fit are shown in
variations through the plate thickness around the hole and also to Fig. 4. The big diameters of used pins to create various percentages
explain fatigue crack initiation location for different sizes of the of interference fit are shown in Table 1, the tolerance of the pins
interference fit. Although the effect of interference fit was studied diameter was (0.000 and þ0.003 mm) during manufacturing. The
on only one holed plate in this research, its results can be used in interference fit size was defined as
much complicated practical joints to some extent.
Dd
I ¼  100;
d
2. Experimental test specifications
where D is the big diameter of pin and d is the diameter of fatigue
In experimental part of the research to observe interference fit specimen hole.
effect on fatigue life improvement, fatigue test specimens, their To force fit the pins into the fatigue specimens’ hole, a 60 kN
dimensions are shown in Fig. 2, were prepared. These specimens capacity Roel pull–push test machine was used. To do so, the con-
were made from a 4.5 mm thick Al-alloy 7075-T6 plate. They were tacting surfaces, including the hole inner surface and the pin outer
cut from the plate with planar dimensions of 1.2 m  1.2 m. The surfaces, were smeared with MoS2 lubricant (a special lubricant
longitudinal direction of the specimens was aligned in rolling which is used in metal forming and has a friction coefficient of 0.1).
direction of the plate. As the figure shows the specimens have Then flat end of the pin was located inside a specially pre-designed
a 5 mm diameter hole in the centre, which was drilled and reamed fixture that was fixed to the upper grip of the test machine. After
to achieve the precise size of hole (i.e. 5  0.005 mm). that a washer with 8 mm inner and 12 mm outer diameters was put
Although the mechanical properties of Al-alloy 7075-T6 were around the hole under the specimens (as a support) on the lower
available in many references (Wuanlin and Xiangjiong, 1991; Shah
and Ramachandran, 2002; Mechanical Engineer’s Handbook, 1998)

Fig. 4. Dimensions of the pin for 2% interference fit size (mm) and force fitting it into
Fig. 2. Fatigue test specimen dimensions (mm) without inserted pin. a specimen hole.
T.N. Chakherlou et al. / European Journal of Mechanics A/Solids 29 (2010) 675–682 677

Table 1 and loading (during pin force fitting) are symmetric respect to X–Z
Big diameter of the pins for various sizes of interference fit. and Y–Z planes, only a quarter of the plate and the pin were
Pin diameter (mm) 5.050 5.075 5.100 5.200 included in the FEM simulation.
Interference fit size %, I 1 1.5 2 4 To mesh the plate and the pin, linear isoparametric cubic
Solid45 elements were used. This 2  2  2 integration point
element has 8 nodes (one node at each vertex) and each node has 3
Table 2 DOF (degree of freedom) displacement in X, Y and Z directions. This
Maximum applied tensile loads to specimens and their equivalent remote sress element was preferred to 20 node quadratic isoparametric cubic
amplitudes.
Solid95 as it has the same accuracy in plasticity but well-conju-
Pmax (kN) 29 31 33 35 37 39 41 gated with contact elements (Swanson Analysis Systems Inc,
Corresponding remote stress 128.85 137.75 146.65 155.55 164.4 173.3 182.2 2007b).
amplitude (MPa)
To include contact between the pin and the hole surfaces,
CONTA 173 and TARGE 170 were used to mesh the pin outer surface
grip of the machine. In the last stage the lower grip was moved up and the hole inner surface. These contact elements allow pressure
to insert the pin into the hole (see Fig. 4). to be transferred between the contacting surfaces but without
After interference fitting the specimens, fatigue tests were penetrating each other (Swanson Analysis Systems Inc, 2007a). In
performed for four different interference fitted and open hole the model the friction between contacting surfaces was included
specimens. Fatigue tests were carried out at seven different load using the elastic Coulomb friction model with friction coefficient of
magnitudes (as shown in Table 2) to obtain S–N data. In conducting m ¼ 0.1, this value was obtained from experimental test (Chakherlou
fatigue tests, sinusoidal cyclic loads were applied using a Zwick and Vogwell, 2003).
Roel Amsler HA250 fatigue machine (see Fig. 5) with zero load ratio In the FEM model, an elastic–plastic kinematic hardening
(R ¼ Pmin/Pmax ¼ 0 as shown in Fig. 6) and frequency of 15 Hz. material behavior was considered for the Al-alloy plate in which
Experimental fatigue test results are shown in Fig. 7 as semi-log stress–strain curve (see Fig. 3), elastic modulus and Poisson’s ratio
(S–N) chart where Sa ¼ Pa/A (Sa is the remote stress amplitude, Pa is were provided by experimental test results. For the steel pin elastic
the applied load amplitude and A ¼ 25  4.5 mm2 is the remote material behavior was used as the pin deforms elastically during
cross-sectional area). It is worth mentioning that a number of interference fit and subsequent longitudinal loading.
fatigue tests were carried out at load levels lower than 29 kN for Regarding to the model constraints, symmetric (displacement)
interference fitted specimens, but in all of them the fatigue crack boundary condition was applied to the plate and the pin in the X–Z
initiated and propagated at region far from the hole at fillet area and Y–Z planes due to the model symmetry.
(their S–N data were not shown in Fig. 7). As explained in the previous section a steel support washer was
placed the test specimen underneath around the hole during the
3. Finite element models experimental pin force fitting to restrain it in Z direction. In the FEM
simulation to account for this support washer, the nodes on the
3.1. Modeling and solution process plate bottom face, which located on the area between the radii
8 mm and 12 mm from the hole centre (showing the washer inner
There are a number of researches about interference fit (Crew, and outer diameters), were constrained in the Z direction.
1975; Shah and Ramachandran, 2002) but these studies were based The FEM solution process was conducted in three load steps
on 2D models in which the effect of stress variations have not been (stages). In the first load step, the pin upper surface nodes were
considered through the plate thickness. However, in this paper a 3D displaced 8 mm in -Z direction through the hole to simulate the
model is used to include stress (or strain) variations in the plate experimental pin force fitting. In the model as shown in Fig. 8, the
thickness. To do numerical simulation for the interference fit, only conical part of the pin was modeled through the hole in a near
the middle part of the 4.5 mm thick fatigue test specimen (see contact condition so as to avoid extra and unnecessary displacing
Fig. 2) was considered to have an optimum model size (i.e. the plate the pin during solution process. After that, in the second load step
model size is 25 mm  40 mm). As the geometry of the specimen a tensile stress (remote stress) was applied to the right end of plate

Fig. 5. Interference fitted specimen before and after fracture in fatigue test machine.
678 T.N. Chakherlou et al. / European Journal of Mechanics A/Solids 29 (2010) 675–682

Fig. 8. Meshed model using FEM.

Fig. 6. Sinusoidal applied cyclic loads.

(see Fig. 8) to simulate the longitudinal loading during experi-


mental fatigue tests. The remote stress magnitude (Smax ¼ 2  Sa)
was chosen so as to create the load level that was applied to the
specimens during fatigue testing. In the last load step, this remote
stress was removed from the model to simulate unloading.
Therefore, with these three load steps the experimental process of
pin force fitting and one cycle of cyclic loading during experimental
fatigue tests were implemented in the FEM simulations. However,
to investigate that if the number of cyclic loading affects the stress
(or strain) distributions in the plate further cyclic loading was
Fig. 9. Longitudinal stress, sx (MPa) distributions around the interference fitted hole
continued in the other load steps. The results showed that the for I ¼ 1% (pin was not shown).
stress (or strain) distributions did not alter by the second or third
cyclic loading (see the following sub-sections).
initiates and propagates, three paths are chosen and shown in
3.2. Stress distribution due to interference fit Fig. 11.
The results of longitudinal stress distributions after first load
The distributions of longitudinal stress which are left from 1% step (interference fitting) at the smallest cross-sectional area for
and 4% interference fit sizes after first load step are shown in Figs. 9 the defined paths and for different interference fit sizes are shown
and 10 as contours. As the figures show interference fit (depends on in Figs. 12–15.
its size) can leave tensile or compressive longitudinal stress or both As the graphs show the created tangential stress (or longitudinal
of them around interference fitted hole. stress at the smallest cross-sectional area) due to pin force fitting is
In order to indicate stress distribution as graphs in the following not uniform through the plate thickness and its value depends on
discussions at the smallest cross-sectional area, where fatigue crack the interference fit size. This stress is tensile at the top, middle and

Fig. 10. Longitudinal stress, sx (MPa) distributions around the interference fitted hole
Fig. 7. Fatigue lives obtained from experimental tests. for I ¼ 4% (pin was not shown).
T.N. Chakherlou et al. / European Journal of Mechanics A/Solids 29 (2010) 675–682 679

Fig. 11. Paths chosen for stress distribution as graphs in the following discussions.

Fig. 14. Longitudinal stress, sx distributions in the defined paths for the interference fit
size of I ¼ 2%.

plane stress distribution for 4% interference fit size shows


compressive stress a little distance away from the hole. This could
be due to the plate material bulge out that compresses the elements
at that region.

3.3. Longitudinal stress history from FEM simulation

Stress distribution due to different interference fit sizes was


shown for three paths at the smallest cross-sectional area in the
previous section. However, in order to have an overview of the
effect of interference fit on the stress history at some critical loca-
Fig. 12. Longitudinal stress, sx distributions in the defined paths for the interference fit
size of I ¼ 1%. tions (nodes in the FEM model around hole), longitudinal cyclic
loading was continued on the FEM model (this could be a simula-
tion of cyclic loading in fatigue tests). The positions of nodes to
bottom planes (for the smallest cross-sectional area) for the smaller which the stress history is shown are illustrated in Fig. 16.
interference fit size. However, with increasing the interference fit The longitudinal stress history for three nodes at the hole edge
size the longitudinal stress becomes compressive at the middle on the top, middle and bottom planes is shown in Fig. 17 for 1.5%
plane around the hole (this is similar to tangential stress created interference fit size with maximum 29 kN longitudinal (zero to
around a hole in an infinite holed plate with elastic–plastic material tension) cyclic loading. As the figure shows the stress amplitude is
behavior, that is subjected to a big internal pressure at the hole almost the same for the three nodes, but the mean stress at node 1
surface (Kim and Lee, 2006; Hsu and Forman, 1975)). At the bigger is the biggest. For this model the stress history is also shown in
interference fit sizes, the longitudinal stress still remains tensile for Fig. 18 for four adjacent nodes at the top plane at the smallest cross-
the top and bottom planes at the hole edge. Scrutinizing the bottom sectional area. Node 1 again has bigger stress amplitude and mean

Fig. 13. Longitudinal stress, sx distributions in the defined paths for the interference fit Fig. 15. Longitudinal stress, sx distributions in the defined paths for the interference fit
size of I ¼ 1.5%. size of I ¼ 4%.
680 T.N. Chakherlou et al. / European Journal of Mechanics A/Solids 29 (2010) 675–682

Fig. 16. Node locations at the smallest cross-sectional area in the FEM simulation.

Fig. 19. Longitudinal stress, sx history of node No.1 for different interference fit sizes
and open hole models for maximum 29 kN of cyclic loading.

Fig. 20. Fatigue crack growth in interference fitted specimens for 29 kN and more
cyclic loading.

Fig. 17. Longitudinal stress, sx (MPa) history at nodes No. 1, 100 and 200 for model
in Section 3.1) is shown in Fig. 19 for maximum 29 kN cyclic
with 1.5% interference fit size which subjected to a maximum longitudinal load of
29 kN cyclic loading. loading. As the figure shows interference fit reduces stress ampli-
tude and increases mean stress compared to open hole model.
stress compared to the other three nodes as can be seen from the However, with increasing interference fit size the rate of stress
figure. amplitude reduction decreases so that the stress amplitude
The longitudinal stress, sx history at node 1 for different reduction for 2% and 4% interference fit sizes is almost the same, but
percentages of interference fit and open hole model (not explained increasing interference fit sizes entails mean stress increase.

4. Fracture section survey

In order to investigate the effect of interference fit on fatigue


crack initiation and propagation region in test specimens, fracture
sections were shown to a number of specimens in this section. Figs.
20 and 21 show the locations of fatigue crack growth for different
range of load from top view. As it is shown in Fig. 21 for cyclic
loading less than 29 kN fatigue failure occurs somewhere far from
hole edge regardless of interference fit size whereas in the speci-
mens subjected to 29 kN cyclic loading and more, fatigue failure
occurs from hole edge (see Fig. 20).

Fig. 18. Longitudinal stress, sx history at nodes No. 1, 2, 3 and 4 for model with 1.5%
interference fit size which subjected to a maximum longitudinal load of 29 kN cyclic Fig. 21. Fatigue crack growth in interference fitted specimens for cyclic loading less
loading. than 29 kN.
T.N. Chakherlou et al. / European Journal of Mechanics A/Solids 29 (2010) 675–682 681

Fig. 22. Crack initiation, growth and fracture locations of an open hole specimen for
35 kN cyclic loading. Fig. 24. Crack initiation, growth and fracture locations of 2% interference fit size
specimen for 35 kN cyclic loading.

Fracture sections at the smallest cross-sectional area to show


the fatigue crack initiation locations are displayed in Figs. 22–25 for
open hole and interference fitted fatigue specimens. As the figures
show for open hole fatigue crack initiates around the mid-plane but
for interference fitted specimens it initiates from the top plane.

5. Discussion

The obtained S–N data from experimental tests for open hole
and interference fitted specimens shown in Fig. 7 indicate the effect
of interference fit on increasing the fatigue life compared to open
hole specimen which confirms the results presented in literature.
Also, as the figure shows increasing the interference fit size from 1%
to 1.5% and 2% entails more fatigue life improvement, however,
further fatigue life improvement is not appreciable when interfer-
ence fit size increases to I ¼ 4% and even in a number of specimens
the fatigue life decreases a little compared to the specimens with
I ¼ 2%. Fig. 25. Crack initiation, growth and fracture locations of 4% interference fit size
Also the fatigue life improvement in the interference fitted specimen for 35 kN cyclic loading.
specimens compared to open hole ones is explained using FEM
simulation. As Fig. 19 shows, interference fit generally decreases
cyclic stress amplitude (the most effective parameter in fatigue life) I ¼ 2%, but the mean stress is much bigger in comparison. This could
and increases a little mean stress at the hole edge at the smallest be the reason why no more fatigue life improvement was appeared
cross-sectional area compared to open hole. With increasing the in the experimental fatigue tests for interference fit size of I ¼ 4%.
interference fit size from I ¼ 1% to I ¼ 2% this phenomenon is much Investigating the fracture section at the smallest cross-sectional
appreciable (see Fig.19) thus causing more fatigue life improvement. area of the specimens reveals that the fatigue crack initiates and
However, in bigger interference fit size (i.e. I ¼ 4%) the cyclic stress propagates around the middle plane for the open hole (see Fig. 22)
amplitude reduction is almost the same as interference fit size of but from the top plane for the interference fitted specimens (see Figs.
23–25) at hole edge. This also could be explained from FEM simu-
lation by prying Figs. 17 and 18. As the first figure shows the stress
amplitudes (in cyclic longitudinal loading) for three nodes in the top,
middle and bottom planes at the hole edge (for interference fitted
models) are almost the same, however, the mean stress in node one
on the top plane is bigger in comparison with node 100 and 200.
Likewise such a phenomenon can be observed for node 1 in Fig. 18
when its amplitude and mean stress are compared with those of
nodes 2, 3, 4 at the top plane. So the location of node 1 is the critical
point (or region, loosely saying) to early fatigue crack initiation.

6. Conclusion

Experimental test results show that the interference fitted


specimens have improved fatigue life compared to open hole
specimens.
Fig. 23. Crack initiation, growth and fracture locations of 1.5% interference fit size Experimental results show that with increasing interference fit
specimen for 35 kN cyclic loading. size from 1% to 1.5% and 2% fatigue life increases in the interference
682 T.N. Chakherlou et al. / European Journal of Mechanics A/Solids 29 (2010) 675–682

fitted specimens, however, further interference fit size increase (i.e. Finney, J.M., 1993. Cold Expansion and Interference for Extending the Fatigue Life of
Multi-Layer Metal Joints. Defence Science and Technology Organisation, Aero-
I ¼ 4%) showed no more fatigue life improvement.
nautical Research Laboratories, Australia. Research Report, 17.
As the FEM simulations show longitudinal stress, at the smallest Hsu, Y.C., Forman, R.G., 1975. Elastic-plastic analysis of an infinite sheet having
cross-sectional area, varies through the plate thickness in the a circular hole under pressure. Trans. ASME, J. Appl. Mech. 42, 347–352.
interference fitted specimens. This stress is generally tensile Iyer, K., Rubin, C.A., Hahn, G.T., 2001. Influence of interference and clamping
on fretting fatigue in single rivet-row lap joints. J. Tribol-T ASME 123,
through the plate thickness at the small interference fit size (i.e. 686–698.
I ¼ 1%) but it is compressive around mid-plane for bigger interfer- Jost, G.S., Carey, R.P., 1984. Strains in an Elastic Plate Containing an Interference Fit
ence fit sizes (i.e. I ¼ 1.5%, 2% and 4%). Bolt near a Free Edge. Defence Science and Technology Organisation, Aero-
nautical Research Laboratories, Australia. Structures Report, 400.
The FEM simulations indicate that interference fit (depend on its Kim, S.S., Lee, D.G., 2006. Design of the hybrid composite journal bearing assembled
size) reduces cyclic stress amplitude and increases mean stress. by interference fit. Compos. Struct. 75, 222–230.
Fractured sections at the smallest cross-sectional area show that Mann, J.Y., Jost, G.S., 1983. Stress fields associated with interference fitted and cold-
expanded holes. Met. Forum 6, 43–53.
fatigue crack initiates and propagates mostly around the mid-plane Mechanical Engineer’s Handbook, second ed., 1998 John Wiley & Son Inc.
in the open hole specimens and around the top plane at the hole Rich, D.L., Impellizzeri, L.F., 1977. Fatigue analysis of cold-worked and interference
edge in the interference fitted specimens. This location (node 1, see fit fastener holes. Cyclic stress–strain and plastic deformation aspects of fatigue
crack growth. ASTM STP 637, 153–175.
Fig. 16) was also predicted by longitudinal stress history from FEM Shah, S., Ramachandran, S.T., 2002. Test and Analysis Based Stress Intensity Factor
simulation. for Interference-Fit (Taper-Lok) Fasteners. Structural Integrity Department,
Lockheed Martin Aeronautics, Georgia, USA.
Sundarraj, N., Dattaguru, B., Ramamurthy, T.S., 1995. Analysis of a double shear lap
References joint with interference pin fit. Comput. Struct. 55, 357–363.
Swanson Analysis Systems Inc, 2007. ANSYS, Release 11. Contact Technology Guide.
Chakherlou, T.N., Oskouei, R.H., Vogwell, J., 2008. Experimental and numerical Swanson Analysis Systems Inc (Chapter 3). Surface-to-Surface Contact.
investigation of the effect of clamping force on the fatigue behaviour of bolted Swanson Analysis Systems Inc, 2007b. ANSYS, Release 11. Structural Nonlinearities,
plates. Eng. Fail Anal. 15, pp. 563–574. User’s Guide, for Revision 5, 2 (Chapters 13 and 14).
Chakherlou, T.N., Vogwell, J., 2003. The effect of cold expansion on improving the Valtinat, G., Hadrych, I., Huhn, H., 2000. Strengthening of riveted and bolted steel
fatigue life of fastener holes. Eng. Fail Anal. 10, 13–24. constructions under fatigue loading preloaded fasteners-experimental and
Crews John H. Jr., 1975. Analytical and Experimental Investigation of Fatigue in theoretical investigations. In: Connections in Steel Structures IV. AISC and ECCS,
a Sheet Specimen with an Interference-fit Bolt. NASA Langley Research Center, Roanoke/USA, pp. 464–473.
Hampton, VA 23665. Wuanlin, Guo, Xiangjiong, Fu, 1991. An analysis of the crack propagation life of the
Dowling, N., 1999. Mechanical Behavior of Materials, second ed. Prentice-Hall interference-fitted fastener. Acta Aeronautica et Astronautica Sinica 12, 106G
International, NY, USA, ISBN 013905720X, pp. 220–280. (NASA, Langley Research Center).

Anda mungkin juga menyukai