Sei sulla pagina 1di 11

International Journal of Pressure Vessels and Piping 80 (2003) 831841

www.elsevier.com/locate/ijpvp

Comparative study of the behaviour of conventional gasketed and compact


non-gasketed flanged pipe joints under bolt up and operating conditions
M. Abida,*, D.H. Nashb
a

Faculty of Mechanical Engineering, Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Topi-23460, NWFP, Pakistan
b
Department of Mechanical Engineering, University of Strathclyde, Glasgow, UK
Received 11 November 2002; revised 11 October 2003; accepted 10 November 2003

Abstract
Bolted flanged joints comprise an assembly of a number of important individual components, which are required to perform well together
in service. The ideal requirement for a bolted flange joint is a zero-leak condition. However, whilst recommended design procedures for
bolted flange joints are available in international codes and standards, leakage problems are still faced by industry. These are common in both
normal operating (internal pressure loading) and critical event conditions. The drive is, therefore, to find a flange joint assembly, which
provides zero-leak condition and requires little or no maintenance and handling. Considerable investigation in the area of optimised bolted
joints has been in progress for the past 10 years comparing traditional gasketed joints and compact non-gasketed joints, using both
analytical and experimental approaches. In this present study, two-dimensional non-linear finite element studies have been performed for
both gasketed and non-gasketed bolted flange pipe joints. Based on the stress results for the flange and the bolt and the flange
rotation/displacement, compact non-gasketed flange joints are shown to be a viable and preferable alternative to the conventional gasketed
flange joints. Recommendations are made for a best-fit flange model for static load conditions with zero-leak sealing in a flange joint.
q 2004 Elsevier Ltd. All rights reserved.
Keywords: Gasketed; Non-gasketed; Leakage; Flange joints; Finite element

1. Introduction
All modern international design codes and standards
provide calculation rules for the design of the flanges, which
are principally based on the Taylor Forge method [1,2].
Flange joints designed in accordance with international
codes such as ASME [3], PD [4], and CEN [5] experience
leakage and this problem is continuously faced by industry.
In the present study, detailed analysis of flange joints is
undertaken in order to observe the actual stress behaviour in
flange, bolts and gasket. The reason for such a study is that a
bolted flange joint is a combination of different elements
which are inter-linked with each other to perform as a unit
for zero-leak condition. Mathematically, the zero-leak
condition is taken to be zero flange displacement of the
flange at the inside diameter of the mating flanges in the
flange joint. In conventional flange analysis, no consideration has been given to this important issue, i.e. the flanges,
* Corresponding author. Tel.: 92-938-71858~61; fax: 92-938-71889.
E-mail address: abid@giki.edu.pk (M. Abid).
0308-0161/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpvp.2003.11.013

bolts and gasket are discussed and are treated entirely


separately. Therefore, with the availability of modern
analysis tools, it is possible to adopt a design-by-analysis
approach for the bolted flange joint in order to provide
accurate values for the stresses in key areas as required; for
example, at hub-flange fillets, in the bolts and in the gasket
element itself.
In addition for conventionally designed flanges, where
there is a measure of inconsistency of the dimensions
between flange sizes, it has been difficult to standardise on a
single procedure to overcome the stress and deflection
design versus leakage problem. Therefore, by means of
comparison to the conventional flange design, a new novel
flange design is also studied which is dimensionally very
compact, is light weight and is without a gasket. This is
designated the non-gasketed bolted joint. Stresses in the
flange ring, bolts and gasket, various displacements and bolt
bending results are studied for both the gasketed and nongasketed flange joints.
Previous work by Power [6] used a full three-dimensional (3D) non-linear analysis for comparative study of

832

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

hod
hub outside diameter
gth
gasket thickness
gid
gasket inside diameter
god
gasket outside diameter
gcringid gasket centring ring inside diameter
gcringod gasket centring ring outside diameter
gsrght gasket seal ring height
bcd
bolt circle diameter
fod
flange outside diameter
bd
bolt diameter
bhd
bolt head diameter
bhh
bolt head height
f
element contact stiffness factor

Nomenclature
A
B
C
g0
h
t
jh
pid
pod
wth
sh
sod
hubht

outside diameter of the flange


inside diameter of the flange
bolt circle diameter
wall thickness of basic shell/pipe
taper-hub length
non-gasketed flange thickness
joint height
pipe inside diameter
pipe outside diameter
pipe thickness
shoulder height
shoulder outside diameter
hub height

three different types of joints for the internal pressure


loading case only. Using 3D modelling the variation of
contact pressure and stress behaviour in the flange and the
bolts was studied. The contact, or interface pressure was
used as the main quantitative measure of the sealing ability
of the three joint styles. More recently, Cao [7] also
performed 3D FEA considering gasket non-linearity under
both bolt up and operating conditions and studied the joint
strength and sealing capability. Similarly, Fukuoka [8] has
studied the bolt stress scatter, effect of tightening sequence,
bearing surfaces and number of tightening passes. However,
flanged pipe joints have already been modelled using 2D
finite element analysis by Hyde et al. [9], Shoji [10] and by
one of the present authors [11].
In the present study, in order to compare the two flange
styles and their relative performance, a flange size of 4 in.,
900# class was selected. 2D axi-symmetric non-linear finite
element analysis for both the joints was carried out for
internal pressure loading only. During the analysis, the
intention was not to study the variation in contact pressure
and contact area but rather to study the stress behaviour in
the flange and the bolts, especially the effect of bolt bending,
bolt pre-load and flange displacement/rotation. It is noted
that variation in contact pressure and contact area is
theoretically of importance but any deformation on the
flange surface, specifically for the metal-to-metal contact
non-gasketed flange joint, cannot stop leakage even with a
very high contact pressure. This has been noted during
published experimental tests [11]. However, it is recognised

that the contact pressure study is of course important from


the design point of view, in order to take account of bolt
behaviour and the contact force of the flanges, but will not
be reported herein.

2. Allowable stresses and flange joint configuration


2.1. Allowable stresses
The yield stress of the flange and pipe material selected
was 372 N/mm2, giving a nominal design stress of
248 N/mm2. High strength bolts of 30 and 10 mm were
used in the gasketed and non-gasketed joints. Material
properties for flange, pipe, bolts and gasket selected are
given in Table 1.
2.2. Flange geometry
The configurations under examination and parameters
used for modelling are shown in Fig. 1a and b for the
gasketed and non-gasketed joints.
Gasketed flange. For the flange and the gasket geometry,
the dimensions were obtained from BS1560 [12]. A washer
is not included in the present work, since for gasketed joints;
it is not a mandatory requirement of the PD or ASME code.
Gasket was modelled as a solid ring for the gasketed joint. In
actual practice during the bolt up condition, it is necessary to
compress the raised sealing part down to equal to

Table 1
Material properties of gasketed and non-gasketed flange joints
Part

Flange and pipe


Bolt
Gasket

Gasketed flange joint

Non-gasketed flange joint

E
(N/mm2)

Design stress
(N/mm2)

As per code

E
(N/mm2)

Design Stress
(N/mm2)

As per code

173,058
168,922
164,095

0.3
0.3
0.3

248.2 2=3sy
723.9 sy
206.8 2=3sy

ASTM A350 LF2 or A105


ASTM SA193 B7
ASTM A182 F316

203,395
204,000

0.3
0.3

248.2 2=3sy
640.0 sy

ASTM A350 LF2 or A105


ISO 898, Grade 8.8

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

833

Fig. 1. Parameters used in the finite element model of flange joints (a) gasketed, (b) non-gasketed.

the centring ring thickness in order to energise the proper


sealing capability of the gasket as recommended by the
gasket suppliers. During the previous experimental studies
[11], an inconsistency in the gasket outside diameter
dimension were noted and the outside diameter of the
gasket was subsequently machined to allow fit up into the
flange joint. Flange parameters used are given in Table 2.
Non-gasketed flange. A washer with the bolt is
recommended for the non-gasketed joint but it was observed
in preliminary studies that it does not have a significant
effect on stress and deformation results for the axisymmetric model. It was, therefore, decided not to include
the washer in the analytical model. For the non-gasketed
joint, to achieve better joint strength and sealing ability, a
higher initial pre-load and a positive taper angle of 0.038

on the flange surface is applied. Flange parameters used are


given in Table 3.

3. Finite element modelling


Previous work by Spence et al. [13] and Nash et al. [14],
showed the viability of a 2D axisymmetric model, for what
is essentially a 3D component. In neglecting the holes in the
flange and the presence of individual bolts round the flange,
the system can be considered as a continuous bolt ring
located at the bolt centre and running round the circumference of the bolt circle. Schematic quarter models of both
flange joints are shown in Fig. 2a and b. The ANSYS finite
element code v5.7 was employed throughout this work.

834

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

Table 2
Gasketed flange joints geometric parameters
Flange parameters
Flange hub length, Flange joint
Bolt circle dia., Flange outside
hubht (mm)
height, jh (mm) bcd (mm)
dia., fod (mm)
69.9

71.67

Gasket, pipe and


bolt parameters
Gasket inside
dia., gid (mm)

Gasket outside
dia., god (mm)

120.65

149.35

235

292

Gasket thick.,
gth (mm)

Gasket seal
ring inside dia.,
gcringid (mm)
106.43

4.5

Flange inside
dia., pid (mm)
87.3

Flange thick.,
fh (mm)

Flange hub
dia., hod (mm)

44.4

159

Shoulder height, Shoulder outside


sh (mm)
dia., sod (mm)
6.4

Gasket seal
Pipe inside
Pipe outside dia., Pipe thick.,
ring outside dia., dia., pid (mm) pod (mm)
wth (mm)
gcringod (mm
206.50
87.3
114.3
13.5

157.2

Bolt dia.,
bd (mm)
30

Table 3
Non-gasketed flange joints geometric parameters
Flange hub
length, h (mm)

Pipe thickness,
g0 (mm)

Joint height,
jh (mm)

Bolt circle
dia., C (mm)

Flange outside
dia., A (mm)

Flange inside
dia., B (mm)

Flange thick.,
t (mm)

Flange surface
angle, taper (deg)

Bolt dia.,
bd (mm)

34

13.5

102.6

146

171

87.3

30

0.03

10

Fully parametric finite element models were used throughout so that the time involved in building the scaled geometry
models of other different sizes is minimised.
3.1. Element selection
Since the flange, bolt and gasket stresses were the
required outputs from this study, it was necessary to use two
classes of element: solid element PLANE82 to model the
solid entities and contact element CONTACT48 to measure
contact pressure or stress variation between the mating
surfaces. Contact stiffness (KN) for CONTACT48 element
quantifies the level of penetration that the target surface
allows. Over penetration results in inaccuracy if it is too
small a value and non-convergence of solution results if it is
too large a value. From the earlier study by Spence et al.
[14], it is noted that only a lower value of f ; the element
contact stiffness factor, affects the maximum axial stress at
the intersection of the taper hub and the vessel. When
varying the value of f between 0.1 and 100, at the value of
f $ 1 stress was almost levelled out, and therefore, a value
was selected of f 10 for further analysis. The effect of the
variation of contact stiffness on stress is shown in Fig. 3.

3.3. Boundary conditions


The boundary conditions applied for both the joints are
shown in Fig. 4. Internal pressure is applied at the inside
diameter, and the loading due to the head is directly applied
as nodal forces across the wall of the pipe which are
obtained using the internal axial force Ppri2 divided by 2prt
(i.e. Pri2 =2rt). The flanges are free to move in either the
axial or radial direction and rotate, providing the exact
behaviour of the stresses in the flange and the bolt.
Symmetry planes are constrained in both the axial and
radial directions. For the gasketed joint, due to the gap
between the flanges, internal pressure is also applied on
the gasket inside diameter as well as on the flange shoulder
in the gap. The gasket is constrained in the radial direction
to prevent it moving radially since, in practice, it is located
interior to the bolts. The bolts for both flanges are
constrained at the mid diameter in the radial direction as

3.2. Mesh refinement


In order to ensure accurate results, a reasonably refined
mesh was applied in the regions of interest such as at the
hub-pipe and hub-flange intersections. The mesh for the
gasketed flange was also refined at the flange shoulder to
measure the contact stresses at the gasket. Twenty-one and
seven elements across the pipe and hub thickness, 10 and 13
elements across the flange thickness and 10 and 12 elements
across the bolt width were defined for the gasketed and nongasketed flange joints. Illustrations of the meshed models
are shown in Fig. 4a and b.

Fig. 2. Schematic quarter models for (a) gasketed, (b) non-gasketed flange
joints, showing bolts as continuous bolt ring (as used in axi-symmetric
representation).

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

835

Fig. 3. Effect of contact stiffness on maximum axial stress (N/mm2).

they cannot move in this direction within the bolt hole. An


axial displacement is applied on the bottom nodes of the bolt
shank which was progressively optimised by sequential
iteration, to obtain the required pre-stress.
3.4. Initial bolt up and operating (internal pressure) loads
A nominal pre-load of about 35% (245 N/mm2) of
the yield stress of the bolt (723 N/mm2) was chosen for
the gasketed joint as per the achieved maximum strain
in the bolt at the applied torque of 505 N m during the
experimental study [11]. The ASME code does not
specify a magnitude of pre-load for the bolts, rather only
a minimum seating stress that relates to the gasket style
and composition. In addition, in general practice, the
pre-load for the gasketed joint is based partially on the
practical knowledge that most fitters of flanged joints
manually tighten the bolts as hard as possible. For the
non-gasketed joint a pre-load approximately 80% of the
yield stress of the bolt [11] was applied for better joint
strength due to the fact that the higher the pre-load the
better the joint is and is easily achieved for metal-tometal contact flanges due to no gasket stress limitation.
As the yield strength of the bolt material is 640 N/mm2,
so the applied bolt pre-stress is 512 N/mm2. This stress
is based on the bolt shank area of 78.54 mm2, whereas
the bolt strength is taken as per thread root area
(58 mm2), resulting in pre-stress to be applied was
378 N/mm2.
After the bolt up, an internal pressure of 15.3 N/mm2
was applied as per the pressure rating recommended for
the flange size of 4 in., class 900# as defined by the
codes. This applied pressure is the maximum permissible
working pressure, whereas, in order to compare the stress
pattern or structural integrity of both the flange joints,
the pressure required was raised to the proof test pressure
of 23 N/mm2 and thereafter to a maximum value of
40 N/mm2 (i.e. 2.6 times the normal operating pressure
of 15.3 N/mm2).

4. Results and discussion


4.1. Comparison of calculated and FEA results
To verify the models, the main pipe wall axial and hoop
stresses were compared with the calculated stress results and
were found in good agreement.

Fig. 4. Element plots for flange joints (a) gasketed, (b) non-gasketed with
enlarged hub portion.

836

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

Fig. 5. Non-gasketed flange joint, (a) stress intensity plot for full joint at internal pressure of 15.3 N/mm2, (b) stress intensity plot for flange only at internal
pressure of 15.3 N/mm2, (c) maximum principal stress plot at internal pressure of 40 N/mm2.

4.2. Comparison of results for both the joints


4.2.1. Non-gasketed flange joint
With reference to Fig. 5a c and Table 4, the following
observations may be made. At the maximum allowable
working pressure of 15.3 N/mm2, the maximum stress
intensity in the flange is less than the allowable stress of
the flange material. The stress is higher at the hub-flange
fillet, than at the inside diameter at the hub. A high stress
is also observed at the area under the bolt head, due to the
high-applied pre-load in the bolt. The maximum principal
stress is less than the allowable of the flange material even
at a pressure of 2.6 times the maximum permissible
working pressure. The contact stress variation between the
two flanges is very small from the inside diameter (a
value of 19 N/mm2) to the outside diameter (a value of
35 N/mm2) with the applied internal pressure. The rotation
from inside to outside diameter of the non-gasketed flange
is almost negligible at the maximum permissible, the
proof test and the maximum applied pressures, and is
recorded in Table 4.

4.2.2. Gasketed flange joint


With reference to Figs. 6a,b and 7a,b and Table 4, the
following observations were made. For the gasketed flange
joint, even during pre-loading, stress intensity and principal
stress is significantly more than the allowable of the flange
material at the hub-flange fillet. The application of internal
pressure added to this effect resulting in increase in stress.
This indicates yielding at hub-flange fillet during both bolt
up (pre-loading) and operating (pressure) loading (Fig. 6a
and b). At other locations, i.e. hub-pipe and hub-centre, the
resulting stress was less than the allowable.
It is also important to note that the local yielding is more
than 20% of the hub-flange fillet depth, Fig. 7a. The
maximum stress at the inside diameter of the hub is also
higher than the allowable of the flange material. In addition,
a high stress is present at the edge of the shoulder, which
acts as the pivot point due to the flange rotation. The higher
stresses in this region are possibly due to the geometric
stress concentration and idealised finite element modelling
(Fig. 7a). From Fig. 7b and c stress intensity plots, overall
stress variation in the flange and bolt is obvious. The rotation

Table 4
Non-gasketed flange joint, maximum principal stress and flange rotation results
Internal Pressure (N/mm2)

15.3
23.0
40.0

Maximum principal stress (N/mm2)

Flange rotation (degree)

Gasketed flange joint (mm)

Non-gasketed flange joint (mm)

Gasketed flange joint (mm)

Non-gasketed flange joint (mm)

309.59
316.62
321.66

94.64
100.31
137.95

0.1589
0.3085
0.3131

0.0300
0.0307
0.0424

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

837

Fig. 6. Gasketed flange joint, (a) principal stress plot after pre-loading or bolt up, (b) stress intensity (exaggerated) plot for flange at a pressure of 15.3 N/mm2.

of the gasketed flange joint is proportional to the applied


internal pressure. For the gasketed flange joint, flange
rotation
means the
flange portion between
the shoulder outside diameter and the flange outside
diameter. Shoulder rotation means the rotation of the
shoulder portion only.

4.3. Bolt stress behaviour of joints


4.3.1. Non-gasketed flange joint
With reference to Fig. 8a,b and Table 5, the following
may be concluded. There is high stress at the top of the bolt
shaft close to the bolt head and the shank corner. However,

Fig. 7. Gasketed flange joint, (a) stress intensity plot showing the location and the depth at which the yielding starts at a pressure of 15.3 N/mm2, (b) stress
intensity plot for full joint model, (c) stress intensity plot for flange only at internal pressure of 15.3 N/mm2.

838

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

Fig. 8. Non-gasketed flange joint bolt only, (a) stress intensity plot for the full joint due to the operating load, (b) stress variation in the bolt due to the operating
load (zoomed bolt portion only).

no yielding resulted and the maximum stress in the bolt is


less than the allowable stress of the bolt material. From
the stress variation in the bolt, the bolt bending is very
small, due to the very small flange rotation with applied
internal pressure.
4.3.2. Gasketed flange joint
With reference to Fig. 9a,b and Table 5, similar to the nongasketed joint bolt, a high stress occurred at the corner of the
bolt head and the bolt shank with no yielding. However, bolt
bending is obvious for the gasketed joint due to a difference
of stress of about 200 N/mm2 at the inside and outside nodes.
This difference is a maximum of 40 N/mm2 for the nongasketed joint, which is very small compared to the values
obtained for the gasketed joint. This stress variation in the

bolt shows the flange rotation during pre-loading (bolt up)


and operating (pressure) loads.
4.4. Contact stress behaviour for gasketed and non-gasketed
joints
Due to the flange rotation, the contact pressure on the
gasket varies from the gasket inside to the outside diameter.
The gasket was modelled as a solid ring. The deformed
shape of the gasket is shown in Fig. 10 and the gasket
contact stress variation is shown in Fig. 11. From the high
stress at the seal ring outside diameter, the flange rotation is
again obvious and shows that during the bolt-up, the seal
ring can be damaged for any applied higher pre-load. As
discussed earlier, it is important for the gasketed joint to
achieve the proper seating stress recommended by the gasket

Table 5
Bolt stress variation for non-gasketed and gasketed flange joints with the variation of pressure
Internal pressure
(N/mm2)

Stress in bolts
(N/mm2)
Non-gasketed flange joint

15.3
23.0
40.0

Gasketed flange joint

Inside diameter
(mm)

Mid diameter
(mm)

Outside diameter
(mm)

Inside diameter
(mm)

Mid diameter
(mm)

Outside diameter
(mm)

392.20
391.88
402.26

374.13
373.10
376.38

356.06
354.30
350.46

343.65
346.86
347.50

242.87
243.17
243.82

141.86
141.86
139.93

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

839

Fig. 9. Gasketed flange joint bolt only, (a) stress intensity plot for the full joint due to the operating load, (b) stress variation in the bolt due to the operating load
(zoomed bolt portion only).

supplier for the no leak condition. It is obvious from the


gasket stress variation that a very low stress value is
achieved on the contact area between the flange and the
gasket. In addition, in axisymmetric modelling, the gasket is
modelled as per the ideal required condition.
In Fig. 12, the stress variation at the contact surface with
the symmetry plane for the non-gasketed joint is shown
which increased gradually from the inside to the outside
diameter during the application of internal pressure.

5. Conclusions
In view of the stress distributions resulting in the flange
and bolts together with the observed flange rotation, bolt
bending, bolt fatigue and bolt relaxation it is concluded that

a dynamic behaviour in the gasketed flange joint is present


during both the bolt-up and operating conditions. This is
primarily due to the inclusion of a flexible gasket element
between the flanges leading to a continually varying
situation where, when bolt or pressure load changes, the
flange faces move and rotation results in response. This
provides permanent damage in the flange and repeated use
of this joint cannot ensure its performance for zero-leak,
even though the joint may be well assembled. Secondly, due
to limitations of the choice of gasket seating stress to avoid
crushing, a higher initial stress cannot be applied in the
bolts, and therefore, the available strength of the high yield
bolts cannot be utilised.
In comparison, stresses observed in the non-gasketed
flanged joint are within the specified allowable levels even
up to an internal pressure of 2.5 times design pressure.

Fig. 10. Gasketed flange joint, deformed gasket plot.

840

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

Fig. 11. Gasketed flange joint, gasket stress variation plot along the gasket, after the operating load.

Fig. 12. Non-gasketed flange joint, stress variation at the symmetry plane (contact surface).

M. Abid, D.H. Nash / International Journal of Pressure Vessels and Piping 80 (2003) 831841

A non-gasketed metal-to-metal contact joint without a


separating gasket is concluded as having static mode of
load due to no significant movement of the flange faces with
a change in bolt or pressure load. Due to this no bolt bending
and so no bolt fatigue results, therefore, it is considered a
safe joint. Finally due to the compact dimensions, it is
comparatively 10 times lighter than the conventional
gasketed flange joint, resulting in easy handling and joint
assembly.
Hence, as a result of the detailed finite element study
presented herein, compact non-gasketed flange joints may
readily be considered as a viable alternative to the
conventional gasketed flange joints presently in use.

References
[1] Waters EO, Taylor JH. The strength of pipe flanges. Trans Mech
Engng 1927;49:53142.
[2] Waters EO, Wesstrom DB, Rossheim DB, Williams FSG. Formulas
for stresses in bolted flanged connections. Trans ASME 1937;59:
1617.
[3] American Society of Mechanical Engineers. ASME boiler and
pressure vessel code. Section VIII, Division 1. App. Y, 2001 Edition.
[4] PD5500:2000. Unfired fusion welded pressure vessels. British
Standards Institution, London.

841

[5] EN 13445. European unfired pressure vessels, CEN; 2002.


[6] Power DJ. A study of conventional and unconventional flanged pipe
joint styles using non linear finite element analysis techniques. MPhil
Thesis; 1997.
[7] Cao D, Xu H. 3-D finite element analysis of bolted flange joint
considering gasket non-linearity. Pressure Vessel Piping Conf Boston
1999;382:121 6.
[8] Fukuoka T, Takaki T. Evaluations of bolt-up sequence of pipe flange
using three dimensional finite element analysis. Pressure Vessel
Piping Conf Boston 1999;382:2135.
[9] Hyde TH, Lewis LV, Fessler H. Bolting and loss of contact between
cylindrical flat-flanged joints without gaskets. J Strain Anal 1988;23:
18.
[10] Shoji Y, Nagata S. A new analysis method for flange-gasket systems.
ASME Pressure Vessel Piping Conf Boston 1999;382:113 20.
[11] Abid M. Experimental and analytical studies of conventional
(gasketed) and unconventional (non gasketed) flanged pipe
joints (with special emphasis on the engineering of joint strength
and sealing). PhD Thesis. University of Strathclyde, Glasgow;
2000.
[12] British Standard BS1560: Section 3.1. Circular flanges for pipes,
valves and fittings. BSI London; 1989.
[13] Spence J, Macfarlane DM, Tooth AS. Metal-to-metal full-face
taper-hub flanges: finite element model evaluation and preliminary
plastic analysis results. Proc Inst Mech Engrs 1998;212(Part E):
5769.
[14] Nash DH, Spence J, Tooth AS, Abid M, Power DJ. A parametric study
of metal-to-metal full face taper-hub flanges. Int J Pressure Vessel
Piping 2000;77:791 7.

Potrebbero piacerti anche