See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/328074656

Evaluation and Completion the Design Methods of Pressure Vessels Flange

Joints

Article in Revista de Chimie · August 2018

DOI: 10.37358/RC.18.8.6453

CITATIONS READS

7 647

3 authors, including:

Valeriu Jinescu Georgeta Roman

Polytechnic University of Bucharest Polytechnic University of Bucharest
91 PUBLICATIONS 394 CITATIONS 8 PUBLICATIONS 18 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Georgeta Roman on 07 May 2023.

The user has requested enhancement of the downloaded file.

 Evaluation and Completion the Design Methods
of Pressure Vessels Flange Joints

VALERIU V. JINESCU*, GEORGETA URSE*, ANGELA CHELU

Politehnica University of Bucharest, Industrial Process Equipment Department, 313 Splaiul Independentei, 060042, Bucharest,
Romania

We have provided a comparative analysis of the current international computing standards (European
Standard EN 13445-3; ASME-Code, Section VIII, Division 2; British Standard (PD 5500: 2009)) that take into
consideration only the static loading of flanges and bolts, if the temperatures of the flanges, bolts and sealing
gasket are equal to each other and lower than creep temperature. The paper has put forth relations for
completion the calculation method for flange joints in situations not taken into consideration by standards,
namely: -static loading if flange temperature and bolt temperature are different; -thermal transient loading;
-creep stress; -fatigue loading in the general case of a sequence of blocks of normal stresses. Furthermore,
relationships have been proposed for the calculation of the maximum allowable difference between a
flange and a bolt so as to ensure both the tightness and the mechanical strength of the flange joint.
Keywords: flanges, joints, bolts, thermal transient, fatigue, tightness

Flange joints for pressure vessels must withstand
mechanical and thermal stresses, must be rigid enough
lest their deformation should affect tightness and ensure
sealing without loss of fluid throughout their service life.
The present paper deals with the calculation of flange
joints whose sealing gasket is located inside the bolts holes
circle (fig. 1).
Fig. 1 Flange with gasket
inside bolts holes circle: 1,
2 -flanges; 3 – bolt; 4 -
washer; 5 -nut; 6 -sealing
gasket
Currently accepted computing methods for the
calculation of flange joints in pressure vessels [1 - 3] involve
the calculation of strength, which includes the choice of
the sealing gasket; the calculations envisage neither the
required rigidity of the flange joint nor its sealing.
Calculations are made:
- for monotonic/steady state loading under operating
conditions featuring calculation parameters and under
pressure test conditions. In all calculations it is considered
that the flange and bolt temperatures are equal, but below
creep temperature;
- for fatigue loading, considering the linear - elastic
behavior of the material used in the making of flanges and
bolts.
Works [4 - 19] go beyond official standards by taking
into consideration the:
-temperature difference between a flange and a bolt
under operating parameters, as well as in the transient
conditions and the influence of this difference upon
tightness;
-influence of flange rigidity on tightness;
-correlation between flange joint strength, rigidity and
tightness;
-calculation of flange joints under load at temperatures
higher than creep temperature;
-calculation of the components of shock loaded flange
joints;
* email: vvjinescu@yahoo.com, Phone: (+40)722436555; ursegianina@yahoo.com., Phone: (+40)749092895
1954 http://www.revistadechimie.ro
-calculation of flanged joints with ribs.
The works cited as dealing with the calculation of flange
joints strength do not deal with the:
-non-linear behavior of the design materials used in the
make of the joint components, both under static and fatigue
loading;
-influence of damage on the strength of the flange joint
components;
-influence of residual stresses on the strength and rigidity
of flange joint components.
Problems not considered so far in the calculation of
flange joints are dealt with in the papers [20 - 28] and may
be adapted to their calculation.
Some recent research refers to the:
-influence of flange geometry (connection radius, flange
neck length) on flange strength [29];
-influence of the variation of elastic washer deformation
on the clamping force of bolts under creep conditions [30];
-tightness of flange joints under internal pressure and
heat load [31-34], or in transient thermal mode [35].
In addition to these researches, for example, some
works feature experiments with flanges without sealing
gaskets [36].
A literature survey points to the fact that, in addition to
official normative provisions, there are now:
- results that could complement norms and allow for a
closer to reality calculation of flange joints; the effect will
be to reduce flange weight;
- general results adaptable to the calculation of flange
joints (non-linear behavior, damage, residual stress, etc.).
The paper compares the methods of calculating the
flange joints used in official calculation methods and makes
recommendations on complementing them with the
results published in literature.
Comparative analysis of official methods of calculating ring
flange joints under static loading
For low and medium pressure, insufficient flange rigidity
has the following effects:
- insufficient circumferential stiffness may cause
deformation between two bolts with deflection w (fig. 2,
a). In this case, the solution to the problem may come
from thickening the flanges;
REV.CHIM.(Bucharest)♦69♦ No. 8 ♦ 2018
 - insufficient axial or longitudinal rigidity (especially due
to the shell it is assembled onto) determines the rotation
of the flanges with angle 2 . θ (fig. 2, b). In this case,
increasing the rigidity of the flange joint can be done by
using hub flanges (fig. 2, d), and in the case of flat ring
flanges, the shell end is thickened at s2>s1 (fig. 2, c) or
longitudinal ribs, 2, are provided between two successive
bolts (fig. 2, d).
Fig. 2 Deformation of the components of a flange joint subjected to
low and medium pressures due to: a – bending deffection, w, of
each of flanges 3 and 4 between two adjacent bolts, 5 and 6; b -
insufficient stiffness of shells, 1 and 2. Methods for increasing
( )
flange joint rigidity: c- by thickening the end section, 4 s2 > s1 ;
d -the use of ribs, 2, welded to the
flat flange 1 and the shell, 3 [8].
One considers that a flange joint that is very rigid
circumferentially does not deform in the middle between
two clamping bolts. Furthermore, if the flange and the
component it is assembled to feature longitudinal rigidities
above a certain value, then it may be assumed that they do
not rotate. That is how flanged joints are treated in official
standards. That means that they have a relatively high
stiffness and therefore a great weight.
REV.CHIM.(Bucharest)♦69 ♦ No. 8 ♦ 2018 http://www.revistadechimie.ro
Calculations based on official regulations [1 - 3] start
from the assumption that the flanges bend under the
influence of bending moments due to the bolt force. The
calculation refers only to two cases of loading, namely;
- when pre-clamping or initially clamping (on mounting)
with temperature at T=T0 and internal pressure p=p0;
- in operating conditions at T=Tb=Tf and p=pc, (pc -
calculation pressure; Tb - bolt temperature and Tf - flange
temperature, but only under the creep temperature (T<Tc).
Calculation of flange joints at operating mode parameters
Computation with ASME Code [1], European [2] and BS
[3] standards breakdown flanges into integral type (fig. 3,
a,b), loose type flanges (fig. 3, c,d) and loose type lap joint
flanges (fig. 3, e).
The ASME Code [1] breaks down flanges as either
integral (fig. 3, a and b) or loose type flanges (fig. 3, c and
d).
For calculation purposes, one determines the effective
width of gasket b, depending on the reference width bo. For
example, for flat gaskets bo=Bo / 2 where Bo is the gasket
contact width. In the case of a ring seal (where bo=Bo / 8),
Table 1 lists the expressions of bending forces and
bending moments according [2].
Table 2 shows the expressions of the stresses in the
hub, longitudinal (σ H ) , radial (σ r ) and tangential (σθ ) .
Some of the notations in table 1 and table 2 are to be
read as follows [2];
A- is the outside diameter of the flange or, where slotted
holes extend to outside of flange, the diameter to bottom
of slots; B - the inside diameter of flange; B0 -gasket contact
width, limited by gasket width and flange surface
configuration; b -effective gasket or joint seating width; b0
-is the basic gasket or joint seating width; C -bolt circle
diameter; e -minimum flange thickness, measured at the
thinnest section; G - diameter of gasket load reaction; go-
thickness of hub at small end; g1- thickness of hub at back
of flange; ; h -hub length; hD -radial distance
from bolt circle on which HD acts; HD- the hydrostatic end
force applied via sell to flange; hG -radial distance from
gasket load reaction to bolt circle on which HG acts; HG-
the compression load on gasket to ensure tight joint;hT -
Fig. 3. Flange types according to the ASME
code [1]: a -integral flange without a hub;
b -hub integral type flange; c-loose flange
with a hub; d -loose type flange without
a hub; e -lap type joint; loose flange with
a hub
1955
 Table 1
FORCES AND BENDING MOMENTS (fig. 3), [2]
radial distance from bolt circle to circle on which HT acts;
HT -the hydrostatic end force due to pressure on flange
face; W - design bolt load for assembly condition; CF -bolt
pitch correction factor; M - the moment exerted on the
flange per unit of length; WA -the minimum required bolt
load for assembly condition; y = q -minimum gasket or
joint seating pressure;βF, βV, ϕ- factors for integral type
flange calculation derived from European Standard EN
13445-3 [2] diagrams; K - the ratio of the flange diameters;
βFL, βVL -βFL, βVL - factors for the calculation of the
loose hubbed flange derived from European Standard EN
13445-3 [2] diagrams;λ - factor to be calculated according
to EN 13445-3 [2]; formula; β T, β U, β Y-βT, βU, βY-
calculation factors according to EN 13445-3 [2] relations.
Calculation of flange joint under pressure test
Vessels undergo pressure tests at p p > pc , where pc is
the the design pressure. The flange joint must be checked
under pressure test conditions when the gasket has to be
tightened with gasket pressure p e=m . p p. Since the
pressure test is generally carried out at T=To, the total force
in the bolts is calculated with the relation:
(19)
where,
and
The whole bending moment în this case is,
(20)
where FD,p and FT,p are calculated with relations,
1956 http://www.revistadechimie.ro
Table 2
FLANGE STRESSES [2]
(21)
whereas aD ≡ hD, aT ≡ hT and aG ≡ hG (fig. 3).
Conclusions regarding the calculation of flange joints based
on current standards
By summarizes the calculation formulas for flange joints
according to the regulations [1 - 3] it has been found that:
- there are no differences between calculation relation-
ships for:- bending forces and bending moments; - effective
stresses; -allowable stresses;
- there are however the following differences: standards
make use of different types of symbols for forces, bending
moments, stresses, lengths, correction coefficients etc.
Proposals for calculating flange joints in situations not
provided by official regulations
The case of stationary conditions when the flange
temperature and bolt temperature are different
Sometimes bolted flange joints -such as in steam pipes
[8]- are subjected to high temperatures. Experimentally, it
has been shown that under such circumstances, pipe walls,
flanges and bolts have different temperatures. Since there
is a layer of air between the flange and the bolts (obstructing
metallic continuity), the temperature in the bolt, Tb, is
different from the flange temperature, Tf. Sometimes the
thermal expansion coefficients for flange, αf, and bolts,
αb, are different. Consequently, there is a tendency for
different flanges and bolts to expand, resulting in additional
stresses in bolts and flanges. For example, in pressure
vessels operating at T>T0, (T0 -environmental temperature),
the flange temperature is higher than the bolt temperature.
REV.CHIM.(Bucharest)♦69♦ No. 8 ♦ 2018
If αf = αb, then the flanges tend to further expand thermally;
as their expansion is being prevented by bolts, the latter
will undergo further elongation stress, while the flanges
will undergo further compression and bending stresses.
The proposed calculation method is further given below
Fig. 4. Forces loading a
flange joint with
sealing gasket.
for the cases when the temperature difference between
flange, Tf, and bolt, Tb, is different from zero [8].
Due to the heating from T0 to Tf, the flanges tend to
expand with (fig. 5)
and the bolts tend to dilate with,
Under these conditions, the length of the bolt in service
at environment temperature , is,
(22)
where h is the thickness of the flange plate, hg,e -gasket
thickness and ∆ f - flange deformation.
If the temperatures of the flange and bolts are different
from , instead of relation (22), we may write,
(23)
where the elastic elongations in the presence of thermal
deformations (marked as *) are separated from the
thermal extensions. The thermal expansion of the gasket
has been neglected since it has a small thickness
compared to ls ( hg , e << lb ).
The variation in the tendency of bolt expansion, between
the running state at T ≠T0 and the assembly state at T0, is
obtained by subtracting, each term at a time, from
relationship of bolt length at environment temperature,
lb,0 = 2h + hg ,0 − 2∆ f ,0 (bolt length in initial clamping
state at T=To) from relation (23). There was obtained,
(24)
The variation in elastic deformation of the flange-shell
system is negligible compared to its variation caused
by flange rotation, so that
(25)
where ∆f - ∆f,o - is the variation between the operation state
∆f , and the initial clamping ∆f,o , of the elastic deformation
of the flange - shell system and ∆f - ∆f,o - is the variation
between the operation state ∆’f , and the initial clamping
∆’f,o, determined by the flange plate rotation.
Under operating conditions at pressure p ≠ p 0 (p 0-
environmental pressure) and flange and bolt temperatures
, the total bolt force is,
(26)
where and .
REV.CHIM.(Bucharest)♦69 ♦ No. 8 ♦ 2018 http://www.revistadechimie.ro
The bending moment applied to full type flanges under
operating conditions with ∆T f , b = T f − Tb = 0 , is:
. (27)
If b
T ≠ T f an additional force emerges, namely FG (∆T ) ,
the additional force on the gasket due to the difference
between bolt and flange expansion [8] is,
(28)
where:
yb - bolt elasticity; yg- sealing gasket elasticity;
yf - flange elasticity;
ϕ - flange rotation caused by a bending moment equal
to unit;
lb - bolt length;
αb - coefficient of liniar thermal expansion of bolt
material;
αf - coefficient of liniar thermal expansion of flange
material;
Tb - bolt temperature; Tf - flange termperature.
After several transformations [8], one obtained the
expression for the clamping force of bolt in service ,
when ∆T f , b = T f − Tb ≠ 0 and p ≠ p 0 , in the form,
(29)
where the sealing force required by the gasket when
[8],
(30)
It means that the effect of the differing tendencies in
thermal expansion has a bearing upon the sealing force.
This must be checked in accordance with the operating
conditions (T≠To and p≠po ). In order to reduce the value of
FG(∆T)in a given case, it is possible to enhance the bolt
elasticity yb , by increasing length lb with the aid of some
distance sleeves, or / and by reducing the diameter of the
un-threaded rod.
The temperature difference between the flange and the
bolt in service, in stationary mode, depends on the
environmental working temperature and the type of flange
joint. The thermal insulation of the flanges decreases the
value of the temperature difference ∆T f,b =T f, -T b
∆T f , b = T f − Tb and also the additional stress F (∆T).
G
Force Pb,e determines the amount of flange loading under
the meridian bending moment
(31)
or
The bending moment supplementing M s,e
corresponding to ∆T f , b = 0 , is
(32)
Calculation of flange joints in a thermally transient mode
In case T ≠ T0 , the temperature difference between the
flange and the bolts in a transient mode ∆Ttr is higher than
in the stationary mode, ∆Ttr ≥ (∆Tf,b)st.
The flange and bolt temperatures evolve differently
during the transient (heating or cooling) mode.
Consequently, the temperature difference ∆Tf,b increases,
reaches a maximum level (∆Tf,b)max, then decreases until
it reaches the minimum value (∆Tf,b)st corresponding to the
stationary mode (fig. 5). The length of the steady state for
a bolt, ts, is generally, higher than for the flange. The
1957
 On this basis one obtained the relation for the calculation
of the bolt stress at t = 0, depending on the stress required
at a given moment t [9],

(39)

The relationship shows how high the stress should be at

the initial moment σ(0) >σ(t), so that at the end of the
operating period, t, there still remains stress σ(τ) in the
projected element.
The time interval t∆σ, until stress relaxation with ∆σ(=
σ(0) - σ(t) is obtained from relation (39) in the form [9],

(40)

Fig. 5. Variation over time during the transient heating of the

vessel: a - temperatures of flange Tf, and bolt Tb; where
b -temperature difference ∆Tf,b=Tf-Tb [8]
For the stabilized creep,
minimum temperature difference (∆Tf,b)st is reached when
t ≥ tb. (41)
The total force in the transient mode bolts is calculated where a and k are constants of the bolt material; in general,
with a relationship similar to relationship (29), constant a is negligible for very high time, t, intervals.
(33) From relations (40) and (41) for the stabilized creep
where the additional sealing force corresponding to the zone, one obtains duration t(o)∆T, in the
transient mode is ideal case, until stresses get relaxed by ∆σ=σ(0)-σ(t),

(34) (42)

where Tf and Tb are taken from diagrams of the kind shown Expression does not take into account the
deformation of the tightened parts, i.e. bolts and flanges.
in figure 5. One usually calculates Pst,r for . From Because the elements that must maintain total
relations (26), (33) and (34) there results: deformation constant are also deformed, one writes [9],
(43)
(35)
where [44],
where,

is the additional sealing force required under the transitional

mode. In general,
For flange joints
The bending moment during the transitional regime is,
(36)
The extra bending moment caused by the transient
mode is obtained from relations (27) and (36):
(37) D2 - circumference diameter of bolt setting area.
From relations (39), (40) and (41) one obtains the stress
Calculation of flange joints in creep conditions
at a given moment, under creep conditions,
The long-term loading of bolts or bolts in flange joints at
temperatures exceeding creep temperature (T>Tfl) raises
the following problems:
-the determination of the initial stress value so that at
the end of the operating period, t, a certain stress value is
provided, corresponding to the tightening force necessary
to ensure tightness; For long durations t∆σ, a may be neglected. In order to
- the duration of the stress drop from σ(0) to σ(t) should maintain, after a determined time, a relatively large value
be longer than the interval between two wedged bolt for σ(t) and as close as σ(0) , it is necessary that Φ should
clampings. be high, that is to say kfl should be high while kelshould be
One analyzed the behaviour of the bolt material under low, respectively. This is achieved by using long bolts,
creep conditions given by Norton s law, possibly with spacer rings between the flange and the nut.
(38)
Calculation of flange joints under fatigue load
At present, the calculation of flange joints under variable
where is the deformation velocity; B(t) and n stresses with several successive blocks of stress (fig. 6) is
done by using the Palmgren - Miner relationship that has a
constants of the material under creep conditions; σ - bolt number of drawbacks [21]. This relationship has been
stress.
1958 http://www.revistadechimie.ro REV.CHIM.(Bucharest)♦69♦ No. 8 ♦ 2018
corrected by replacing the right-hand member with variable
. One obtained,
(45)
where ni is the actual number of loading cycles whose
stress amplitude is σa,i, and Ni is the number of stress
cycles whose stress amplitude is σa,i until failure. Ni is
obtained for the given value of σa,i on the fatigue diagram
(Wohler), while
Fig. 6 Fatigue load with several successive blocks of stress
Constant C takes different values in official regulations
[21; 37]:
C=1 in the ASME Code, since Ni is the number of
allowable stress cycles in the case of stress amplitude
σa,i ;
C=0.8;0.4 or 0.3 in EN 13445 and PD 5500, depending
on the equivalent number of load cycles calculated
according to these standards
Relation (45) does not take into account the influence
of the medium stress σm, of the behavior - sometimes non-
linear- of the material, of the damage and residual stresses.
A relationship recently proposed by Jinescu V.V. [22; 37;
38] considers the nonlinear behavior of the material under
static loading (exponent α=1/k) and fatigue loading
(exponent m in Basquin law), mean stress (σm), residual
stress (σres) and total damage at time t, (∆T(t)). The relation
for the life calculation under loading with successive blocks
of cyclical stresses has the following expression [21; 37],
(46)
where, for steels
Fig. 7. Curve of fatigue strength depening on loading cycles (Wohler)
with k=0.16...1.0 and m = 3...5 - for different steels; α=0
-for shock loading and α=1 for linear-elastic behaviour
under monotone loading.
The right hand side member, CD, depends on medim
stress, σm, deterioration DT(t) and participation of residual
stress, σres [21;38],
(47)
REV.CHIM.(Bucharest)♦69 ♦ No. 8 ♦ 2018 http://www.revistadechimie.ro
where is the medium stress; -
tensile strength; σ res - residual stress; - the value
of this ratio in the last (final) block of cyclic stresses.
Fatigue strength depends on the number of loading duty
cycles. Three successive cyclical loads are considered,
corresponding to the three domains of the Woler curve
(fig. 7), namely:
-a block of stresses with a small number of cycles (nL)
and maximum stress σmax,L, between yield point σy, and
tensile strength σu;
-a block of stresses with a finite number of cycles (nF)
and maximum stress, σmax, f between yield point σy, and
fatigue strength σR, where ;
-a block of stresses with a number of cycles (nH) whose
maximum stress is lower than the fatigue limit, σR.
In this case, the total participation of the specific energies
corresponding to the three blocks has the expression [39],
(48)
where u is the Coffin- Manson law exponent, corresponding
[ )
to the deformation in the plastic state for
σ max, L ∈ σ u ; σ y ,
(49)
where εa,pl is the strain amplitude in the plastic state; NL -
number of load cycles down to failure and C a constant. In
general u = 0.5
In the law of material behavior under static loading with
stress σ,
(50)
ε is strain, Mσ and k are the material constants, and
α =1 k .
Exponents m2 and m3 derive from Basquin’s law [45],
, (51)
where σ a is stress amplitude; A - material constant.
In range II, of finite number of loading cycles, m=m2
while N=NF. In range III, with a very high number of cycles
(straightline CD1) m=m3 and N-NH.
The loading is not harmful if,
(52)
where,
(53)
1959
where is strain, actually its critical value in range
I; is the mean stress, i.e. its critical value in the II
(F) or III (H) range; D(a;c) is the damage caused by a
crack featuring depth a and length 2c.
and
δσm, H are equal to 1.0, if the corresponding average
stresses are positive (elongation) and equal to -1.0, if the
medium stresses are compression stresses.
If,
(54)
the loading is dangerous.
In fracture mechanics, one currently uses the concept
of stress intensity factor Ki, where i=I; II or III for the three
modes of failure.
In the case of simultaneous loads corresponding to the
three modes of failure, based on the results provided in
[40; 41] and the Law of the Equivalence of Processes and
Phenomena [42; 43], one obtained the expression of the
equivalent stress intensity factor,
(55)
where is the fracture toughness, KII,cr and KIII,cr are
the critical values of KII and KIII, respectively, while
α1 = 1 k1 , where k1 is the exponent in the law of
behaviour under shear stress featuring τ ,
(56)
where γ is the shear strain, while Mt and k1 are the material
constants.
The strength condition in this case is
(57)
where , where cK>1 is the safety coefficent in
relation to the stress intensity factor.
On the other hand, by expanding the results obtained in
work [27], one may write that the damage caused by the
crack with depth a and length 2c is,
(58)
Because fracture mode I corresponds to the loading
within the limits of linear-elastic behavior, K=k1=1, while
α=α1=1, the previous relationship becomes,
(59)
With relationships (48) and (52) - (54) and (58), one
can verify the strength of the flange joints of a cyclically
stressed pressure vessel undergoing different stress groups
in the Wöhler diagram zones I, II and III.
The conditions for maintaining the strength and tightness
of a flange joint
For the good operation of flange joints, it is necessary to
ensure their strength and tightness. Further down we are
presenting the strength and tightness condition,
respectively, of a flange joint, in the general case
- Strength condition. In any of the stress conditions
(except for the pressure test), the maximum bolt force
must be less than the maximum allolwable bolt force, Ps,M.
If Ps peaks in the transient mode, then:
1960 http://www.revistadechimie.ro
(60)
Based on this relationship, one obtained the condition
to be met during the transitory regime lest the flange joint
should lose its tightness [8]:
(61)
where one wrote
(62)
the maximum allowable temperature difference resulting
from the strength condition
-Tightness condition. This results from the fact that in
any situation the pressure on gasket, q, is greater than the
crushing pressure and less than the pressure that is
destroying it, namely:
For kg, one should set experimental values depending
on the gasket type and material. Generally kg,=1.5...4.
From the above-mentioned conditions, the variation in
the gasket thickness can be determined on the basis of q,
between its actual value and its minimum value which
still ensures tightness, qo or qo (T).
The allowable temperature difference in the tightness
condition resulted from stating that the maximum bolt
elongation ∆ls,max should be less than the allowable variation
in the gasket thickness,
(63)
One concluded that the maximum allowable
temperature difference between the flange and the bolt
liable to maintain joint tightness, has the expression [8],
(64)
The tightness condition is expressed by the relationship,
(65)
One should consider as maximum temperature
difference between the flange and the bolt the lowest of
all values calculated with relations (62) and (64). From a
practical point of view, one should set the condition that
(66)
By doing so, both the strength condition and the tightness
condition will be met. Each temperature interval should
comply with the condition that heating / cooling be
conducted in such a way that the relationship (66) is
observed.
Conclusions
From the analysis of the current method of calculating
flange joints, common to official standards [1 - 3], it results
that they only refer to the static loading of flanges and
bolts, in the case when the temperatures of the flanges,
bolts and the sealing gasket are equal to one another and
lower than creep temperature.
Starting from the reported results in the literature, it is
proposed to complete the existing calculation method by
adding the effects caused by: -the temperature difference
between flanges and bolts in operation and in the thermal
REV.CHIM.(Bucharest)♦69♦ No. 8 ♦ 2018
transient mode; - static stress under creep conditions; -
fatigue loading, taking into account the nonlinear behavior
of flange and bolt materials, and accumulation or
superposition of loading with a reduced number of load
cycles, under a finite number of cycles and/or a very high
number of load cycles.
Based on the analysis of the influence of temperature
difference between flanges and bolts, a relationship has
been proposed that limits this temperature difference in
order to prevent losing tightness and/or exceeding the
mechanical strength of the flange joint.
References
1.*** ASME Boiler and Pressure Vessel Code. Section VIII, Division 2,
Rules for construction of pressure vessels, 2010.
2.*** European Standard, Unfired pressure vessels - Part 3: Design -
EN 13445-3, 2009.
3.*** British Standard PD-5500, Specification for fusion welded
pressure vessels, 2009.
4.JINESCU, V.V., TEODORESCU, N. Construcþia si calculul imbinarilor
cu flanse (III), Rev. Chim. (Bucharest), 33, no. 7, 1982, p 671 - 676.
5.JINESCU, V.V., TEODORESCU, N., GARDUS, V.., Constructia si calculul
îmbinãrilor cu flanse (IV), Rev. Chim. (Buchrest), 38, no. 8, 1987, p
727 -730.
6.JINESCU, V.V., TEODORESCU, N., GARDUS, V., Constructia si calculul
îmbinarilor cu flanse (V), Rev. Chim. (Bucharest), 39, no. 1, 1988, p
75 - 78.
7.JINESCU, V.V., TEODORESCU, N., GARDUS, V.., Constructia si calculul
imbinarilor cu flanse (VI), Rev. Chim. (Bucharest), 39, no. 2, 1988,
p 185-188.
8.JINESCU, V.V., Utilaj tehnologic pentru industrii de proces, Vol. 3,
Ed. Tehnica, Bucuresti, 1988.
9.JINESCU, V.V., Calculul si constructia utilajului chimic, petrochimic
si de rafinarii, Vol. 1, Ed. Tehnica, Bucuresti, 1983.
10.TACA, C., PAUNESCU M., Influenta deformatiei circumferentiare a
inelului flansei plate asupra etanseitatii, Buletinul IPB, Seria Mecanica,
Tom L, 1988, p. 91-96.
11.PAUNESCU M., TACA C.D., STEFANESCU M., Flanged joints rigidity
of the pressure vessels, Romanian Academy, Branch office of Iasi,
Modeling and optimization in the machines building field (MOCM-6),
2000, p.45-51.
12.PAUNESCU M., MANOLACHE Z., TACA C.D., Studiul experimental
al asamblãrilor cu flanse plate cu garnituri din marsit, Buletin
I.P.B.,Seria Constructii de Masini, tom XLIX, 1987, p. 83- 88.
13.IATAN R., RENERT M., Calculul si constructia flanselor cu nervuri,
Buletinul Institutului Politehnic Gheorghe Gheorghiu-Dej, Bucuresti,
tomul XL, Seria mecanica, 1978, nr. 2, p. 51-60.
14.IATAN, R., RENERT,M., BOTEA,, N., Metode de calcul pentru
flansele cu nervuri tesite, Rev. Chim.(Bucharest), 29, nr.7, 1978, p.
678-682.
15.PÃUNESCU M. Rigiditatea asamblarilor cu flanse ale recipientelor
sub presiune, Constructia de masini, 53, nr. 4, 2001, p. 9-13.
16.JINESCU V.V., TEODORESCU N., Metoda pentru calculul asamblarilor
cu flanse, în Constructia de Masini, 52, nr. 12, 2000, p. 1- 8.
17.JINESCU V.V., TEODORESCU N., Analiza rezistentei, rigiditatii si
etanseitatii asamblarilor cu flanse inelare, în vol. Industria de Petrol
ºi Gaze. Prezent ºi perspective, I.P.G., Ploiesti, 14-15 mai 1992, p. 161-
166.
18.RENERT M., IATAN R., Cercetari pentru stabilirea metodicii de
calcul a construcþiilor de flanse cu nervuri, Institutul de Cercetari si
Proiectari de Utilaj Chimic si de Rafinarii, 2, 1972.
19.JINESCU V.V., MANOLACHE Z., IATAN R.,PAUNESCU M.,TACA
D.,TEODORESCU N., STEFANECU M., JUGANARU M., SIMA T., DESANU
M., Analiza rigiditatii, etanseitatii si rezistentei asamblarilor cu flanºe
inelare, în vederea optimizarii lor constructive, IUC -Grivita Rosie,
21-4-6/faza I, II, III, 1984 - 1986.
20.JINESCU V.V., Cumulation of Effects in calculation the Deterioration
of Fatigue loaded Structures, International Journal Damage Mechanics,
21, July, 2012, p. 671-695.
21.JINESCU, V.V., Critical Energy Approach for the Fatigue life
Calculation under Blocks with different normal Stress Amplitudes,
Int. J. Mechanical Sci., 67, febr. 2013, p. 78-88.
22.JINESCU, V.V., IORDACHESCU, V.I., TEODORESCU, N., Relation for
Calculation of Critical Stresses in Pressure Equipment with Cracks,
Rev. Chim. (Bucharest), 64, no. 8, 2013, p. 858-863.
REV.CHIM.(Bucharest)♦69 ♦ No. 8 ♦ 2018 http://www.revistadechimie.ro
View publication stats
23.JINESCU V.V., IORDACHESCU V.I., Calculation of deterioration due
to cracks in tubular specimens, U.P.B. Sci.Bull., Series D, vol. 76,
2014, p. 149 -160.
24.JINESCU V.V., NICOLOF V.I., JINESCU C., CHELU A., Superposition
of Effects in Calculating the Deterioration of Tubular Structures and in
Non-newtonian Fluid Flow, Rev. Chim. (Bucharest), 66, no. 5, 2015,
p. 698 – 702.
25.JINESCU V.V., CHELU A., TEODORESCU N., NICOLOF V.I., Strength
of Tubular Samples and Tubular Cracked Junction Under Combined
Loads, Rev. Chim. (Bucharest), 66, no. 11, 2015, p. 1832 – 1836.
26.JINESCU V.V., NICOLOF V.I., CHELU A., MANEA S.E., Critical stresses,
critical group of stresses and strength of tubular structures without
and with cracks, U.P.B. Sci.Bull., Series D, vol. 77, 2015 p. 165-176.
27.JINESCU V.V., NICOLOF V.I., Strength criteria at the tip of the crack,
Engineering Fracture Mechanics, 142, 2015, p. 108-115.
28.JINESCU V.V., NICOLOF V.I., JINESCU G., MANEA S.E., Cap. 1: The
Principle of Critical Energy is a Transdisciplinary Principle with
Interdisciplinary Applications în vol. Proceedings of the International
Conference on Interdisciplinary Studies (editor Valentina Mihaela
Pomazan), INTECH, ICIS 2016, p. 1-23.
29.*** FESSLERAND H., PERRY D.A., Effect of changes in the taper-
hub shape on the strength of high-pressure flanges, Proc Instn Mech
Engrs Vol 212 Part E,1998, pg. 151 - 162.
30.BOUZID A.H., NECHACHE A., The modeling of bolted flange joints
used with disc springs and tube spacers to reduce relaxation,
International Journal of Pressure Vessel and Piping, Vol. 87, 2010, pg.
730 - 736.
31.BOUZID A.H., NECHACHE A, BROWN, W., The effect of steady
state thermal loading on the deflections of a flanged joint with a
cover plate, ASME Pressure Vessel and Piping Conference, Vancouver,
British Columbia, Canada, August 5- 9, Conference paper PVP2002-
1093, PVP-Vol.433,2002, p.153-162.
32.BROWN, W, DERENNE M., BOUZID, A-H., Determination of the
steady state operating temperature of pressure vessel flange
components: Part-1 - Analytic Methods, ASME Pressure Vessel and
Piping Conference, Seattle, Washington, July 23-27, PVP-Vol.405, 2000,
95-104
33.KUMANO, H., SAWA T., HIROSE, H., Mechanical behavior or bolted
joint under steady heat conduction, ASME J. Pressure Vessel
Technology, 115(2), 1994, p.142-146.
34.SAWA ,T. TAKAGI, Y., TATSUOKA, T., Thermal stress analysis and
evaluation of sealing performance in pipe flange connections with
spiral wound gaskets under elevated temperature and internal
pressure, ASME Pressure Vessel and Piping Conference, Conference
paper PVP2005-71425, 2005, p.105-113.
35.BROWN, W., DERENNE, M., BOUZID, A-H., Determination of Gasket
stress levels during thermal transient, ASME Pressure Vessel and
Piping Conference, Vancouver, British Columbia, Canada, August 5-9,
Conference paper PVP2002 - 1078, PVP-Vol.433, 2002, p. 21-28.
36.*** Nashand D.H., Abid M., Surface sensitivity study of non-gasketed
flange joint, Proc. Instn. Mech. Engrs. Vol. 218 Part E: J. Process
Mechanical Engineering, 2004, p. 205 - 212;
37.JINESCU V.V., Fatigue life prediction for simultaneous cyclic loading
with blocks of normal stresses and shear stresses, Jour. of Engineering
Sciences and Innovation, 1, nr. 1, 2016, pp. 1-16.
38.JINESCU V.V., Applications in Mechanical Engineering of the
Principle of Critical Energy, Lambert Academic Publishing,
Saarbrucken, 2015.
39.JINESCU V.V., NICOLOF V.I., TEODORESCU N., Effects superposition
under imposed deterioration and simultaneous fatigue in different
regimes, Int. J. Damage Mech, 26, Issue 5, July 2017, p. 633-650.
40.JINESCU V.V., Principiul energiei critice în domeniul mecanicii
ruperii materialelor, I, Rev. Chim. (Bucharest), 41, no. 3, 1990, p. 209-
2l6.
41.JINESCU V.V., The Principle of Critical Energy in the Field of Materials
Fracture Mechanics, International Journal of Pressure Vessels and
Piping, 53, nr. 1, 1992, p. 39-45.
42.JINESCU V.V., Legea echivalenþei proceselor si fenomenelor si
aplicatiile ei, Rev. Chim. (Bucharest), 52, no. 12, 2001, p. 768-773.
43.JINESCU V.V., Energia, Energonica si Termodinamica, Editura AGIR,
Bucuresti, 2016.
44.KARCIANOV L.M., Teoria polzucesti, Fizmatghiz, Moskva, 1960.
45.BASQUIN, O.H., The exponential law of endurance tests, Proc. Am.
Soc. Test Mater, 10, 1910, p.625-630.
Manuscript received: 24.05.2018
1961