Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

Design of Obround Flange for Pressure Vessel

Application by Analytical Method and FEA to
Comply with ASME code
Y. P. Shah1, M.N. Pradhan2
1
Research Scholar ME student, Department of Mechanical Engineering,
Maharashtra Institute of Technology, Pune, Maharashtra, India
2
Professor, Department of Mechanical Engineering, Maharashtra Institute of Technology,
Pune, Maharashtra, India

ABSTRACT
Flanged joints are essential components in nearly all pressurized systems and piping; however they constitute one of
the most complex parts of design. A wrong design can lead to major leakage, affect the system performance and can
be hazardous for operators. Various factors influence the successful design and operation of a flange joint in service.
The bolted flange joint involves interaction between bolting, flange, and gasket. For various industrial applications
such as vapour absorption chillers, unfired pressure vessels having flange size more than 600 mm; there is a need of
non-circular flanges having reduced heights than traditional flanges. The shape considered for flange (Slip-on type
without hub) is obround shape; whose investigation involves finding an appropriate analytical method to be used for
design of this obround flange and which can also comply with ASME code. This obround flange subjected to internal
pressure is designed using equivalent circular flange method. The finite element analysis (FEA) is used to predict
levels of stress and deformation of a particular flange and stresses are linearized for stress categorization. These FEA
results are compared with ASME allowable limit and are on safe side. The analytical design method is approximate
method and results on positive error side. For selected application; there is 22% reduction in height observed with
the use of obround flange.

Keyword: - Flange, Obround, Non-circular, Pressure vessel, Linearization, FEA

1. INTRODUCTION
Flanged joints with gaskets are very common in pressure vessel and piping systems, and are designed mainly
considering internal pressure. Prevention of fluid leakage is the prime requirement of flanged joints. Many design
variables affect joint performance and it is difficult to predict the behaviour of joints in service. The ASME boiler and
pressure vessel code (BPVC) contains rules for non-circular pressure vessels of unreinforced and reinforced
construction and their end covers. There is no standard procedure available for bolted flange connections; so these
flanges cannot be designed directly with the rules of ASME BPVC Section VIII div.-1 due to complexity of the shape.
Hence the method is formulated for manual design of obround flange which can use ASME BPVC section VIII div.-
1. Guidelines of the ASME BPVC Section VIII div.-2 are to be used with the allowable stress limits of ASME BPVC
section VIII div.-1 and Finite Element Analysis (FEA) is done to meet requirements of ASME BPVC Section VIII
Div-2 as specified in U-2(g).

2. LITERATURE REVIEW
Structural integrity and leakage tightness of bolted flanged connections are one of principal factors to ensure a safe
and extended service life of critical engineering structures such as reactors, steam generators, boilers, heat exchangers,
piping systems, and others that operate under critical process conditions including internal pressure and a variety of
operating temperatures. From structural integrity point of view safe design of the bolted flange connections (BFC) has
been solved and satisfactorily standardized by American Codes which is based on Taylor forge method of flange
design [5]. One of the most common methods used for flange design is found in ASME BPVC section VIII Div.1,

1162 www.ijariie.com 211

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

NOMENCLATURE
a: Nominal diameter of bolts (mm) n: Total number of bolts
A: Outside diameter of flange (mm) N: Gasket width (mm)
A: Area of cross-section (mm2) P: Design pressure (MPa)
b: Effective gasket seating width (mm) P: Wetted perimeter (mm)
B: Inside diameter of flange (mm) R: Radial distance from bolt circle to point of intersection of
bo: Basic gasket seating width (mm) hub and back of flange (mm)
C: Bolt circle diameter (mm) Sa: Allowable bolt stress at gasket seating temperature
c: Clearance between OD of shell and ID of flange (mm) (atmospheric temperature) (MPa)
d: Bolt hole diameter (mm) Sb: Allowable bolt stress at operating temperature (MPa)
D: Hydraulic diameter (mm) Sfa: Allowable stress for flange material at gasket seating
D’m: Equivalent mean gasket diameter assumed for design temperature (atmospheric temperature) (MPa)
purpose (mm) Sfb: Allowable stress for flange material at operating
DL: Maximum inside diameter of obround shell (mm) temperature (MPa)
Dm1: maximum mean gasket diameter for non-circular flange (mm) SH: Calculated longitudinal stress in hub (MPa)
Dm2: minimum mean gasket diameter for non-circular flange (mm) SR: Calculated radial stress in flange (MPa)
DS: Minimum inside diameter of obround shell (mm) ST: Calculated tangential stress in flange (MPa)
E: Radial distance from bolt circle to outside diameter of STS: Minimum tensile strength (MPa)
flange (mm) SY: Minimum yield strength (MPa)
G: Diameter at location of gasket load reaction (mm) T: Design temperature (°C)
hD: Radial distance from bolt circle to circle on which H D acts(mm) t: Flange thickness (mm)
HD: Hydrostatic end force on area inside flange (N) tg: Thickness of gasket (mm)
hG: Radial distance from gasket load reaction to bolt circle (mm) tn: Nominal thickness of the shell, pipe or nozzle to which the
HG: Gasket load = Wm1 - H for operating condition (N) flange is attached (mm)
hT: Radial distance from bolt circle to circle on which H T acts(mm) W: Design bolt load for the gasket seating condition (N)
HT: Difference between total hydrostatic end force and the w: Width of the straight portion of obround flange (mm)
hydrostatic end force on area inside flange = H-HD (N) y: Gasket / joint contact surface unit seating stress (MPa)
J: Flange rigidity z: Factor for conversion of obround shape to circular shape
m: Gasket factor

Appendix 2, rules for bolted flange connections with ring type gaskets. Australian Standard AS1210 also follows this
approach. These methods is adapted from of the Taylor-Forge method developed by Waters, Wesstrom, Rossheim
and Williams of the Taylor-Forge company in Chicago in the 1930's and subsequently formed the basis of the ASME
code for flanged joint design[5]. The assumptions made by this method are now generally regarded as simplistic. This
method gave rise to the ‘m’ and ‘y’ gasket factors in ASME section VIII as well as other codes. The calculation is
based on the axial forces balance between the bolt load, the resulting axial force due to the end thrust effect of the
internal pressure and the reaction on the gasket.

Adolf E. Blach [1] in one of his work describes the two design methods for bolted flanged connection of non-circular
cross-section of obround and rectangular type. One method is applicable to unreinforced “almost square" rectangular
shapes, using an "equivalent circular flange", and standard flange design methods. The other is based on a
decomposition of frame and flange bending stresses and may also be used for rib-reinforced pressure vessel flanges.
Calculations, experimental values and finite element results were obtained for flange with ring gaskets (gasket fully
inside the flange bolt line) and full face gaskets. Comparing numerical values with experimental data, he proved that
the method of equivalent circular flanges is suitable for obround and rectangular pressure vessel flanges within certain
limits. The results are on the safe side and become increasingly conservative as the length to width ratio increases.

Muhsen Al-Sannaa and Abdulmalik Alghamdi [3] studied the results obtained using Finite Element Analysis (FEA)
of large diameter welded neck steel flanges under different loading conditions. They give the stress analysis of flanged
joint made up of the flange and the gasket for large diameter steel flanges. They showed that clamping pressure is a
determinate factor for the sealing condition and that clamping pressure needs to be carefully selected to get proper
sealing of the flange-gasket assembly. Increasing the clamping pressure will result in better contact pressure but at the
cost of higher flange stress. Gasket has to be made of soft material with low modulus of elasticity to ensure better
sealing of the assembly. Axial end load may results in gasket leakage if the clamping pressure is not sufficient.

M. Murali Krishna, M.S. Shunmugam and N. Siva Prasad [4] worked on the finite element analysis of bolted flange
joint considering non-linearity of the gasket material under various loading and operating conditions. Gaskets
behaviour is complex due to nonlinear material properties combined with permanent deformation. They found that

1162 www.ijariie.com 212

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

variation of contact stresses due to the rotation of the flange and the material properties of the gasket play important
roles in achieving a leak proof joint. Flange rotation causes variable compression across the gasket from the inner
radius to the outer radius. Due to the variation in compression, the contact stresses also vary along the radius.

This paper aims to find the appropriate analytical method to be used for the obround flange which can comply with
ASME code. The obround flange is designed using equivalent circular flange method. The finite element analysis
(FEA) is used to predict levels of stress and deflection of a particular flanged joint and stresses are linearized. These
FEA results are compared with ASME allowable limit and are on safe side. The analytical design method is
approximate method which results on positive error side.

3. FLANGE DESIGN
The obround flange designed here is for the particular application in generator shell of vapour absorption chiller unit.
The chiller unit is mainly used in industries and hotels for food storage or air-conditioning purpose. Mostly the fitment
of this unit is in parking area or basement where the height is restricted. So to reduce the machine height; one of the
option is to reduce the height of the generator shell and its flange with the use of non-circular shape.

3.1 Forces Acting on Flange

The forces acting on flange joint subjected to internal pressure with ring type gasket i.e. gasket is wholly within the
circle enclosed by a bolt hole and no point of contact beyond this circle [7],[8] are as shown in figure 1. The various
forces acting on flange can be represented on cross-sectional view in the required directional sense [7], [11] as shown
in figure 2

Fig -1: forces acting in a bolted flange joint assembly Fig -2: Forces represented on flange (Slip-on flange
without hub)
The initial bolt load generated upon tightening is transferred to the gasket via the flanges. This initial seating stress
compresses the gasket and tightens it within itself. The hydrostatic force generated by the system pressure, tends to
‘unload’ and reduce the stress on the gasket. The stress remaining on the gasket is considered to be the ‘operating’ or
‘residual’ stress. It should be seen that on a raised face assembly as shown in figure 1, there will be some deflection
of the flanges themselves (‘flange rotation’). This is a function of the load applied, the flange material and the geometry
of the flanges. Thus, the operational stress towards the outside edge of the gasket tends to be greater than on the inside
edge.

The calculations use four loads bolt loads, gasket load, face pressure load and hydrostatic end force represented in
figure 2 and two conditions seating and operating [11]. Load HD is created by the pressure on the pipe attached to the
flange. During operation, Pressure is applied to the exposed edge of the gasket and gasket tries to expand but is held
in place by the flange faces. The flange faces push back and gives rise to uniformly varying pressure along gasket
width whose average value is represented by load HT. Load HG is the force required to seat the gasket into the flange
gasket face which is based on gasket physical properties.

3.2 Methods for Non-Standard Flange Design

As per the literature survey, the different methods which can be used for non-standard flange design either obround
or rectangular are:

1162 www.ijariie.com 213

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

1) As per Swedish standard for piping

2) Equivalent circular flange method
3) Frame bending method
4) Hydraulic diameter method
These methods either convert the non-circular shape into circular shape and then we can design that circular flange or
considers the change in shape into various formulae in design process for designing flange to be safe.
1) As per Swedish standard for piping [15]:
In the Swedish piping code for flange joint, the design procedure of rectangular, oval and obround flange is
given by converting these shapes into equivalent circular shapes. This method uses the factor k4 multiplied with
the maximum mean gasket diameter Dm1 to get equivalent mean gasket diameter D’m
D′m = k 4 Dm1.......................................................................... (a)

2) Equivalent circular flange method [1], [10], [11]:

The method for design of non-circular pressure vessel flange is developed by Adolf E. Blach [1]. It covers the
design of obround and rectangular shaped flanges which comply with the ASME BPVC codes. It has given two
methods of designing flange as per the shell construction. The first method is for flange mounted on
unreinforced pressure vessels. This method uses the part of the procedure used in the ASME code section VIII
div 1 article UG-34 for the design of non-circular flat covers. In the code, a factor z is defined which relates a
flat cover of obround/rectangular shape to a circular one. The factor z is given by
2.4DS
z = 3.4 − DL
........................................................................ (b)
But factor z should not be larger than 2.5 and length to width ratio should be less than 2 is the requirement.
The square root of this factor is used as a multiplier of the small side of the obround/rectangular flange to
obtain an equivalent circular shape.
B = DS √z...........................................................................(c)
Any obround or rectangular flange which satisfies the above criteria can then simply be designed or analysed
as an equivalent circular flange, and all flange design code rules as per appendix 2 of the ASME BPVC section
VIII division 1 are applicable without modification.

3) Frame bending method [1]:

The other method given by Adolf Blach is frame bending approach for non-circular pressure vessel flanges.
This method is applicable to the flanges with obround or rectangular shapes mounted on reinforced vessels.
The method uses a combination of frame analysis for the ability of the flange to retain its shape, and bending
of an infinitely long flanged section in a perpendicular plane with respect to the frame.

In this case, the flange must also act as a stiffener for the vessel side plates, in addition to providing a tight seal
between components. Thus, such flanges have to resist frame bending stresses, stresses which occur when a
frame is subjected to internal pressure. These stresses cause deflections in a plane perpendicular to the vessel
axis. In addition, flanges also have to resist flange bending stresses in planes parallel to the axis of the vessel,
stresses which occur when a flange is bolted-up about the gasket, or when internal pressure effects tend to open
up the bolted connection.

4) Hydraulic diameter method:

This method uses the conversion of non-standard shape into circular shape using the logic of hydraulic diameter.
Hydraulic diameter is the commonly used term when handling flow in non-circular tubes and channels. Using
this term one can calculate many things in the same way as for a round tube. This is given by
4𝐴
𝐷 = B = 𝑃 ........................................................................................... (d)
This gives the equivalent inside diameter of flange which can be used in design calculations as per ASME
BPVC section VIII appendix 2.

3.3 Input Parameters for Design

The generator on which this obround flange is going to be mounted contains hot water on tube side and weak solution
of Li-Br on shell side. The working parameters needed for design of flange are as given in table I.

1162 www.ijariie.com 214

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

Table -1: Input Parameters for Flange Design

Parameter Label Value Unit
Design Pressure P 1.054 MPa
Design temperature T 200 °C
Corrosion allowance 1.5 mm
Vessel or nozzle wall thickness tn 16 mm
Minimum inside diameter of shell DS 500 mm
Maximum inside diameter of shell DL 856 mm
Clearance between OD of shell and ID of flange c 1 mm
Flange thickness (assumed) t 70 mm
The existing circular flange outer diameter is 873.76mm which is needed to be replaced for height reduction. The
bolting is selected as in [7] and its dimensions are taken from [17] which is given in table below. Similarly for the
given operating parameters, gasket is selected and which has following parameters required in design as specified in
table III.
Table -2: Bolting Specifications [17] Table -3: Gasket Specifications [11]
Selected Bolt Parameters Selected Gasket Parameters
Parameter Label Value Unit Selected Gasket Spitman AF 154
Selected Bolting M30 Parameter Label Value Unit
Bolt diameter a 30 mm Gasket factor m 1
Bolt hole diameter d 33 mm Gasket contact surface
y 1.4 MPa
Radial distance from bolt circle to point of unit seating stress
R 33.34 mm
intersection of hub and back of flange Gasket width N 20 mm
Radial distance from bolt circle to outside Gasket thickness tg 5 mm
E 33.34 mm
of flange
Root area of selected bolt 502.96 mm2
Total bolts being used n 28 Qty

3.4 Material of construction

The material required for standard component is selected as per the operating temperature to which it is subjected and
the form of the material. The working temperature range for generator is from 150C to 200C i.e. 302F to 392F.

1) Flange and bolt material:

The material for flange is selected as per guidelines given in ref. [6]. The material used for the flange is SA-
516 Gr70 which is in the plate form and for bolts; it is SA-193-B7. Material properties are as shown in table 4.
The maximum allowable stress value is the maximum unit stress permitted in a given material used in a vessel
constructed. The criteria for maximum allowable tensile stress values permitted for different materials are given
in mandatory appendix 1 & 2 of ASME BPVC Section II, Part D. As per it, the allowable stress value for
selected flange material is 138 MPa and for selected bolting material is 172 MPa.

Table -4: Mechanical Properties of Flange and Bolt Material [13] Table -5: Properties of Spitman AF 154 Gasket
Parameter Value Unit Parameter Value Unit
Selected Material SA-516 Gr70 SA-193-B7 Max operating pressure 150 Bar
Carbon Steel Max. Short term service
Nominal Composition Carbon steel - 450 C
1Cr-1/2Mo temperature
Product Form Plate Bolting - Max. Continuous service C
250
Size / Thickness - ≤64 mm temperature
Max. Operation temp.
UNS No. K02700 G41400 - 290 C
Tensile Strength 485 860 MPa for steam
Yield Strength 260 725 MPa Density 1800 Kg/m3
Max. Temperature Tensile strength 15 MPa
538 538 C
limit Compressibility 7-11%
Density 7750 7750 kg/m3 Recovery 50%
Modulus of Elasticity 185 185 GPa Water absorption <5%

1162 www.ijariie.com 215

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

2) Gasket Material:
Material used for gasket compressed non-asbestos fiber reinforced type (CNAF). We have considered Spitman
AF 154 (Champion seals Pvt. Ltd.) as it has superior performance compressed jointing sheet incorporating a
blend of special heat resisting aramid fibres with a high quality nitrile elastomer binder. Properties of gasket
are as given in table 5.

3.5 Selection of Appropriate Design Method

The selection of the appropriate method to design the obround flange will affect the flange thickness considerably.
One of the key requirements of the work is that the design should comply with the ASME code.

We are having 4 methods for the approximate design calculation of obround flange and each method gives different
stress values. This graph is plotted with 65 mm flange thickness, M30 (28) bolting and 20 mm gasket width. We plot
the stresses in gasket seating and operating condition of flange as shown in chart 1.

It is clear from the chart that stresses generating in

equivalent circular flange method are maximum of all.
So if we design the flange as per equivalent circular
flange method, we will design it for maximum stress
condition and it will be safe for other methods. The
other advantage of using this method is; this method
uses the ASME code for flange design with slight
modification so it will comply with ASME codes as
well. So the most appropriate method for designing of
obround flange is equivalent circular flange method.

Chart -1: Trend of the stress vs. various methods

3.6 Design by Formula (Analytical Method)
The manual design of the obround flange is done with the equivalent circular flange method as its give the safest
design. The manual design of slip-on type flange considers only tangential stress ST in flange which is the combination
of membrane + bending stresses. The hub stress SH is considered zero as slip-on flange has nearly hub-less design and
radial stress SR is consider to be entirely carried away by shell so they are ignored.
With the use of the given basic dimensions we found the factor z to be 1.9567 which is below the limit of 2.5. Then
with the help of equation (b), the inner diameter of equivalent circular shape of flange is calculated as 751.181mm. It
is just the virtual dimension of circular flange required for finding the approximate stresses in the actual obround
flange. This dimension along with the required specifications of bolting, gasket and their materials is used to find the
stress occurring in flange with the help of procedure given in appendix 2 of ASME BPVC section VIII division 1. The
flange thickness appropriate for the given loading condition and dimension is found to be 70mm. The max stress
generating in this flange is 127.93 MPa in operating condition and 118.8 MPa in gasket seating condition. The height
of the designed flange is 680.68mm. The flange dimensions also satisfies the rigidity criteria with J = 0.987 which is
below allowable limit of 1. Cover flange is also design based on the UG-34 given in [11] and the minimum thickness
required for flange is 45mm.

4. FINITE ELEMENT ANALYSIS

The modelling of generator flange for chiller is carried out in SolidWork, Release 2015 (Student Edition) and analysis
on its Cosmos solver. Flanges, nozzle opening on shells are important for illustration, process and inspection. They
will not only weaken the strength of shell but also generates boundary stress on the joint of vessel flange, leading to
severe stress concentration. So the joint is most vulnerable part to failure. It’s of great importance to study the influence
of various parameters on stress distribution of the Flange. Due to different loadings applied to flange, a local stress
state of the flange connection characterized by high stress concentrations occurs in the intersection region. The
significant stress concentration almost always occurs in the vicinity of the flange-to-shell junction due to the inherent
structural discontinuity that is formed by the intersection.

1162 www.ijariie.com 216

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

4.1 Design by Analysis (DBA) Approach [12]

Traditionally, to determine the acceptability of the design, formulas and charts based on analytical solutions and
empirical data have been used. This is today called as Design by Formula (DBF) which is described by ASME VIII
Division 1. In the 1960’s an alternative to DBF was established known as Design by Analysis (DBA). It served as a
complement for the design cases which were not covered by DBF, and was based on a method where stresses are
classified into different categories. Formulas are not always applicable and therefore design by analysis serves as a
complement within both the ASME code as allowed by U-2(g).
ASME BPVC section VIII division 2 Part 5 [12] states the design by analysis requirements and includes several
methods for evaluating the design against four different failure modes. Methods are described for evaluating against
plastic collapse, local failure, buckling, and cyclical load. The “Stress classification” method, which this work is
limited to, is included in the analysis for plastic collapse and thus, this is the only failure mode that will be considered.
For the elastic stress analysis method a linear elastic model is used. This means that the stress in the material is assumed
to be linearly proportional to the strain. The “Stress categorization” or “elastic stress analysis-method” is used since
it is a relatively straightforward way to obtain a result.

4.2 Analysis
1) Basic Assumptions:
In order to simplify the analysis of the flanged joint, a number of assumptions were made. These basic
assumptions are:
 Gasket material was assumed to have linear properties with the non-linear behaviour of the gasket section
ignored.
 All materials for the model of slip-on flange, shell, bolts, gasket and blind flange, are assumed isotropic.
 Analysis will be linear static analysis.
 Temperature effects will not be considered (already considered while selecting material).
 Bolt loads will be averaged over the area where the bolt head are located in the circular ring.

2) Boundary conditions and Loads:

The fix support is applied to the tubesheet which is going to be attached to evaporator shell and back side of
the generator shell to restrain it in all direction as shown in figure 3. The pressure of 1.054 MPa is applied at
the inside of the obround shell. Then the gasket load reaction (HG) of 1449.81 N & difference between total
hydrostatic end force and hydrostatic end force acting inside the shell (HT) of 1091.67 N is applied at distance
of 26.4 mm and 32.45 mm respectively from bolt circle position in inward direction on flange face. The load
imparted on the flanged joint by the twenty eight bolts is calculated using the bolt load calculation during flange
design. That bolt load calculated is applied on the effective bolt contact area near the every bolt holes. The total
value of bolt force 538272.530N is divided by number of bolts and then applied. The model is fine meshed with
the solid elements. Mesh used is the standard mesh using parabolic tetrahedral solid elements defined by four
corner nodes, six mid-side nodes, and six edges. For the optimum selection of the mesh size and plotting of the
results, the stress convergence is established and the percentage variation must be kept under 5%.

Fig -3: Boundary conditions and loading Fig -4: von Mises stress plot of obround flange (a)

1162 www.ijariie.com 217

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

Fig -5: von Mises stress plot of obround flange (b) Fig -6: Deformation plot of obround flange

3) Results:
The von Mises stress plots of flange analysis are given in figure 4 and figure 5. The maximum stress generated
is 181.615 MPa. This stress value is higher than the allowable stress but below the yield strength of material.
The stresses are localized at the shell and tubesheet junction and also near the bolt holes at straight to circular
transition of flange; so these need to be linearized to compare with allowable limit. The stress in the other
regions is within allowable limit.
The deformation plot of obround flange subjected to given loading conditions is as shown in figure 6. The
maximum deformation obtained is 0.901 mm which is in the straight portion of shell. Maximum deformation
occurring in flange is within the range of 0.600 mm to 0.715 mm. The deformation allowed is 5 mm; so the
design is safe for deformation & also deformation is negligible as compared to overall size of the flange. It is
occurring because of the non-symmetric shape of the shell. The shell straight portion tries to adopt more stable
shape at these locations subjected to internal pressure and hence the deformation is occurring maximum there.

5. LINEARIZATION
In the finite element method, when continuum elements are used in an analysis, the total stress distribution is obtained
i.e. von Mises stress. The ASME code does not use principal stress or von Mises stress as comparable. These values
cannot be directly compared with the analytical values of stresses as calculations give the membrane and bending
stress. Therefore, to find membrane and bending stresses, the total stress distribution shall be linearized on a stress
component basis and used to calculate the equivalent stresses.

Linearization is a decomposition of the stress distribution we see in FEA of pressure vessels. It decomposes a basically
parabolic distribution into a uniform value (membrane stress), a linearly changing value (bending stress), and possibly
an extra component (peak stress) [14].By doing this, we can use finite element distribution and pick one or more stress
classification lines to decompose the stresses such that we can apply the code. A Stress Classification Line or SCL is
a straight line running through thickness of a vessel. It is perpendicular to both the inside and outside surfaces. The
guidelines of linearization of stress results as per annexure 5-A given in [12] are used for stress classification.

The maximum stress concentration zones can be identified Table -6: Allowable Stress Values for Localized Stress
from the stress plot of the obround flange as given in
Quantity (MPa) Allowable stress Value
figures 4 & 5. The stress concentration is at transition of
Membrane Stress Sa 138
flange from straight portion to curve portion, flange Membrane Stress + Bending 1.5*Sa 207
thickness near bolt hole and also at the intersection of shell
and tubesheet. Allowable stress values as per category
[14] are as given below in table VIII.

5.1 Stress Linearization / Stress Classification

1) At flange thickness: Firstly we perform the linearization at the flange thickness near the bolt hole in flange
Stress concentration line is marked on the section of stress concentrated zone with following points :
Start point Point 1(218, 290.59, -57.764)mm
End point Point 2(218, 290.59, 11.915)mm

1162 www.ijariie.com 218

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

Fig -7: Stress concentration zone in flange Fig -8: Stress classification line through flange
The maximum values of membrane and membrane + bending stresses as per von Mises is as shown in table 7.
Table -7: Equivalent Membrane and Bending Stresses at Flange
Quantity (MPa) Max. Prin. Mid. Prin. Min. Prin. von Mises
Membrane Stress 19.233 2.524 -4.988 21.474
Membrane Stress + Bending (Point 1) 98.948 4.992 -2.391 97.854
Membrane Stress + Bending (Point 2) 3.128 -4.778 -66.892 69.488

2) At flange to shell transition: The stress concentration is also seen near the shell and flange intersection and
equivalent stress in that location is more than allowable stress value as shown in figure 9. So it needs to be
linearized so that it can be compared to allowable stresses as per code.
Stress concentration line is marked on the section of stress concentrated zone with following points :
Start point Point 1(-255, -238.04, -70.366)mm
End point Point 2(255, -254.47, -70.362)mm

Fig -9: Stress concentration zone near flange to shell Fig -10: Stress classification line near flange to shell
transition transition
The maximum values of membrane and membrane + bending stresses as per von Mises is as shown in table 8.

Table -8: Equivalent Membrane and Bending Stresses near flange to shell transition
Quantity (MPa) Max. Prin. Mid. Prin. Min. Prin. von Mises
Membrane Stress 50.724 17.254 1.232 43.740
Membrane Stress + Bending (Point 1) 29.840 -2.773 -71.966 90.043
Membrane Stress + Bending (Point 2) 114.841 56.618 11.860 89.438

1162 www.ijariie.com 219

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

4) Near shell to tubesheet joint: The stress concentration is maximum at the shell and tubesheet intersection and
equivalent stress in that location is 31% more than allowable stress value as shown in figure 11. So it needs to
be linearized so that it can be compared to allowable stresses as per code.

Fig -11: Stress classification line on shell

Stress concentration line is marked on the section of stress concentrated zone with following points :
Start point Point 1(0, 265.76, -387.48)mm
End point Point 2(0, 250.39, -387.05)mm
The maximum values of membrane and membrane + bending stresses as per von Mises is as shown in table 9.
Table -9: Equivalent Membrane and Bending Stresses at Shell
Quantity (MPa) Max. Prin. Mid. Prin. Min. Prin. von Mises
Membrane Stress 14.568 3.459 -14.708 25.598
Membrane Stress + Bending (Point 1) 0.169 -38.144 -146.042 131.315
Membrane Stress + Bending (Point 2) 123.333 45.052 22.272 91.816

5.2 Results
The stresses at the most stress concentrated zones are compared with the allowable limits as per ASME given in table
10.
Table -10: Results of the Stress Linearization
At flange At flange to shell
Allowable Near shell to tubesheet joint
Quantity (MPa) thickness transition
stress
von Mises stress von Mises stress von Mises stress
Membrane Stress 138 21.474 43.740 25.598
Membrane Stress + Bending (Point 1) 207 97.854 90.043 131.315
Membrane Stress + Bending (Point 2) 207 69.488 89.438 91.816
As the linearized stress values are far below the allowable stress values, the design is safe. The analytical value of
stress is 127.930 MPa and the maximum value found after linearization in flange is 97.854 MPa; this gives the clear
idea that the analytical method formulated gives the safer design and its errors on the safe side.

6. CONCLUSIONS
The design of obround flange has be done analytically with the equivalent circular flange method and its finite element
analysis has been conducted which shows the equivalent stresses generating in flange. These stresses are linearized to
obtain the membrane and bending stresses and are compared to allowable limit. The key conclusions are as listed
below-

1162 www.ijariie.com 220

Vol-1 Issue-2 2015 IJARIIE-ISSN(O)-2395-4396

 The equivalent circular flange method is the most appropriate method for obround flange design as it gives the
maximum stress for which we have to design the flange and which also complies the ASME code.
 Equivalent circular flange method is approximate analytical method for design of obround flange with some
limitations that factor z should not be larger than 2.5 and length to width ratio should be less than 2.
 The stress values calculated using with this method falls on the high side as required for a safe design.
 The stress concentration in flange is maximum near bolt hole at the transition area of flange from straight portion
to curve portion and in shell it is maximum at the shell to tubesheet attachment.
 The results from stress linearization of stress concentration zones show that the stresses occurring in the flange
and shell are within allowable limits as specified by ASME.
 Though it is an approximate method, it can be used successfully to reduce the height of the pressure vessel
equipment considerably. The height reduction achieved in our case is 22%.

ACKNOWLEDGEMENT
We thank to Mr. Rahul R. Joshi (Assistant Manager), C&H-Cooling division of Thermax Ltd. who provided insight
and expertise that greatly assisted the research. We would also like to show our gratitude to the Mr. M. Nataraj (head
of cooling engineering) for sharing their pearls of wisdom with us during the course of this research.

REFERENCES

[1] Adolf E. Blach; “Non circular pressure vessel flanges: New design methods”, Fluid sealing, Springer-science &
business Media; pp.247-265.
[2] Hildegard Zerres, Yann Guerout, “Present calculation methods dedicated to bolted flanged connections”,
International Journal of Pressure Vessels and Piping 81 (2004), pp. 211–216.
[3] M. Murali Krishna, M.S. Shunmugam, N. Siva Prasad, “A study on the sealing performance of bolted flange
joints with gaskets using finite element analysis”, International Journal of Pressure Vessels and Piping 84 (2007),
pp. 349–357.
[4] Muhsen Al-Sannaa and Abdulmalik Alghamdi, “Two dimensional finite element analysis for large diameter steel
flanges”, 6th Saudi Engineering conference (2002), Vol.5, pp. 397-408.
[5] Waters EO, Rossheim DB, Wesstrom DB, Williams FSG. Development of General Formulas for Bolted Flanges.
Chicago, IL: Taylor-Forge and Pipe Works; 1937.
[6] Dennis R. Moss, “Pressure Vessel Design Manual”, Fourth Edition, 2013, Gulf Professional Publishing.
[7] Kanti K Mahajan, “Design of Process Equipment”, Second Edition, Pressure Vessel Handbook Publishing INC.
[8] Lloyd E. Brownell and Edwin H. Young, “Process Equipment Design”, Wiley-Interscience.
[9] John H. Bickford, “Gaskets and Gasketed Joints”, CRC press.
[10] B. S. Nau, “Fluid Sealing”, Springer-Science &Business Media.
[11] ASME BPVC section VIII Division-1, “Rules for Construction of Pressure Vessels”, 2013, ASME International.
[12] ASME BPVC section VIII Division-2, “Alternative Rules for Construction of Pressure Vessels”, 2013, ASME
International.
[13] ASME BPVC section II Part D, “ASME Boiler and Pressure Vessel Code – Materials (Metric)”, ASME
International.
[14] ASME PTB-3, “ASME BPVC section VIII- Division 2 Example Problem Manual”, 2013, ASME International.
[15] Swedish piping code for flange joint
[16] British standards PD5500:2012 enquiry case 5500/133; “Flat unstayed ends of non-circular shape and associated
flanges”.
[17] Standards of the tubular exchanger manufacturers association, Eighth edition.

1162 www.ijariie.com 221