www.nature.

com/scientificreports

OPEN Stress analysis and applicability

analysis of the elliptical head
Zhanhui Wang1,2*, Zhifang Zhang1,2, Jinzhong Chen1 & Jinjun Bai1

As the main pressure components of pressure vessels, the mechanical performance of cylinders and
heads affects the normal operation of pressure vessels. At present, no unified theoretical formula
exists for the connection region between an elliptical head and the cylinder. Therefore, the authors
consider the standard elliptical head as the research object. First, the theoretical stress calculation
formula is deduced according to the deformation continuity equation. Second, the stress is
experimentally measured using an internal-pressure thin-walled-vessel stress measurement device,
and the theoretical and experimental stress values in the discontinuous region between the elliptical
head and cylinder are analysed and compared to verify the accuracy and applicability of the theoretical
stress calculation formula. The results show that the theoretical stress calculation formula in the
discontinuous region between the elliptical head and cylinder is valid. By comparing and analysing
the theoretical and experimental stress values, the accuracy and applicability of the theoretical stress
calculation formula in the discontinuous region are verified. The findings can provide guidance for the
stress measurement of internal-pressure vessels.

With the continuous progress of science and technology and the economy, the role of chemical products in
people’s daily life is becoming increasingly ­prominent1. Processes in the chemical industry are generally charac-
terized by high temperatures, pressures, and corrosivity; therefore, pressure vessels are widely u ­ sed2. According
to the magnitude of the pressure, pressure vessels can be divided into internal and external pressure v­ essels3,4. In
the chemical industry, internal-pressure vessels are commonly used. As the main pressure bearing components,
the mechanical performance of cylinders and heads affects the normal use of pressure v­ essels5. In particular, a
pressure vessel bears the combined action of the internal pressure, edge force and edge moment, which causes
the stress to increase in the connection region between the cylinder and head. This phenomenon is known as
the discontinuous effect: the total stress near the connection region between the cylinder and head increases,
rendering it prone to strength f­ ailure6–10. In this context, analysing the stress near the connection between the
cylinder and head is of economic and social s­ ignificance11,12.
Under internal pressure, the cylinder and head individually bear only the membrane s­ tress13,14. Under mem-
brane stress, the corresponding membrane deformation is generated, which is examined in terms of the parallel
circular radial displacement and meridian a­ ngle15. This research method is simple and reliable and pertains to
the non-moment theory. The non-moment theory is applied considering that the mid plane curvature of the
shell has no sudden change, and the boundary is not affected by the transverse shear, bending moment or torque.
However, the curvature radii of the cylinder and head are different, and if the cylinder and head are regarded
as free bodies, the parallel circular radial displacement and meridian angle differ at the connection boundary
under the action of internal pressure. Notably, the cylinder and head are welded together; in other words, at the
connecting boundary, the parallel circular radial displacement and meridian angle must be identical, leading to a
restraint caused by the transverse shear and bending moment at the connection b ­ oundary16–19. This scenario does
not satisfy the application conditions of the non-moment theory, and thus, the moment theory must be adopted.
The elliptical head is composed of a half ellipsoid and straight edge s­ ection20. Scholars have performed con-
siderable research on the internal stress distribution of the elliptical head, specifically, to compare the membrane
stress with finite element analysis or experimental results; however, this comparison is not entirely a­ ccurate21–23.
First, the elliptical head is only studied as a free body without considering the influence of the edge force and edge
moment on the stress distribution. Second, the influence of only the internal pressure is considered, and that of
the bending internal force, such as the transverse shear and bending moment, is ignored. Therefore, the moment
theory should be used to perform the stress analysis because it considers not only internal membrane forces such
as the normal force but also internal bending forces such as the transverse shear and bending moment. The influ-
ence of the internal bending force can be ignored far from the boundary of the elliptical head and cylinder. In the
area near the boundary, the internal bending force must be considered, that is, moment theory must be used. In

1
School of Chemistry and Chemical Engineering, Yulin University, Yulin 719000, Shanxi, China. 2Key Laboratory of
Low Metamorphic Coal Clean Utilization, Yulin 719000, Shanxi, China. *email: [email protected]

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 1

Vol.:(0123456789)
 www.nature.com/scientificreports/

(1 )

(2 )

(a) 2D diagram (b) 3D diagram

Figure 1.  Structural diagram of the elliptical head.

most of the existing studies, the moment theory has not been used to determine the stress. In addition, no uni-
fied theoretical calculation formula exists for the connection region between the elliptical head and c­ ylinder24–27.
Therefore, in this study, the standard elliptical head is considered the research object. First, the theoretical
stress calculation formula in the discontinuous region between the elliptical head and cylinder is deduced accord-
ing to the deformation continuity e­ quation28. Second, the experimental stress is measured using an internal-
pressure thin-walled vessel stress measurement experimental device, and the theoretical and experimental stress
values in the discontinuous region are compared and analysed to verify the accuracy and applicability of the
theoretical stress calculation formula in the discontinuous region between the elliptical head and ­cylinder19,29.
The findings are expected to have guiding significance for the stress measurement of internal-pressure vessels.

Theoretical stress calculation formula in the discontinuous region

between the elliptical head and cylinder
First, the edge force Q0 and edge moment M0 are calculated according to the deformation continuity e­ quation30,
and subsequently, the internal force and internal moment are determined according to the edge force Q0 and
edge moment M0. Finally, the theoretical stress calculation formula in the discontinuous region between the
elliptical head and cylinder can be obtained by substituting the internal pressure, internal force and internal
moment into the calculation formula of the membrane stress and bending s­ tress31. The specific analysis process
is described in the following text.

Solution of the edge force Q0 and edge moment M0 in the discontinuous region of the ellipti-
cal head. From the perspective of stress and manufacturing, an elliptical head represents the ideal structural
form and is widely used in the petroleum, chemical, medicine, machinery, construction, nuclear power and
other industries. The connection structure between the elliptical head and cylinder is shown in Fig. 1.
The connection boundary deformation diagram is shown in Fig. 2, in which (a) shows the overall deforma-
tion diagram, (b) shows the membrane deformation caused by the internal pressure P, (c) shows the bending
deformation caused by the edge force Q0, and (d) shows the bending deformation caused by the edge moment
M0. According to the deformation continuity condition, after the superposition of the membrane deformation
(b) and bending deformation (c) and (d), the total deformation values (a) of the elliptical head and cylinder at
the connection boundary are equal.
Under the action of the edge force and edge moment, the rigid body motions of the elliptical head and cylinder
are constrained, and they are viewed as rigid bodies. The deformation continuity equation of the elliptical head
and the cylinder can be expressed as

Q Q0
�P1 + �1 0 + �M 0 P
1 = �2 + �2 + �2 ,
M0
P Q 0 M0 P Q 0 M0 (1)
θ1 + θ1 + θ1 = θ2 + θ2 + θ2 .

where subscripts 1 and 2 represent the elliptical head and cylinder, respectively. Superscripts P, Q0 and M0 rep-
p p Q Q
resent the internal pressure, edge force and edge moment, respectively. 1 , θ1 , 1 0, θ1 0, M M0
1 , and θ1 represent
0

the parallel circular radial displacement and meridian angle produced by the elliptical head under the action of
p p Q Q
the internal pressure P, edge force Q0 and edge moment M0. 2 , θ2 , 2 0, θ2 0, M M0
2 , and θ2 denote the parallel
0

circular radial displacement and meridian angle produced by the cylinder under the action of the internal pres-
sure P, edge force Q0 and edge moment M0.
According to the l­ iterature13, the membrane deformation equations of the elliptical head can be expressed as
 p
�1 = Pa2 (2 − a2 /b2 − µ)/(2Et1 ),
p (2)
θ1 = 0.

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 2

Vol:.(1234567890)
www.nature.com/scientificreports/

Figure 2.  Deformation diagram of the connection boundary.

where t1 is the thickness of the elliptical head, a is the length of the semi-major axis, and b is the length of the
semi-minor axis. P is the internal pressure, μ is Poisson’s ratio, and E is the elastic modulus.
The bending deformation equations of the elliptical head can be expressed as
 Q0 � √
 �1 = 2 4� 3(1 − µ2 ) R/t 1 RQ0 /(Et1 ),
 Q0

θ1 = −2� 3(1 − µ2 )RQ0 /(Et12 ),
M0 2 2 (3)
 �M1 0 = 2 3(1 − µ2 )RM 0 /(Et1 ),


3/4 √ 2
θ1 = −4[3(1 − µ )] R/t1 M0 /(Et1 ).
where R is the second radius of curvature at the boundary.
The membrane deformation equations of the cylinder can be expressed as
 P
�2 = PR2 (2 − µ)/(2Et2 ),
(4)
θ2P = 0.
where t2 is the thickness of the cylinder.
The bending deformation equations of the cylinder can be expressed as
4 3(1 − µ2 )√R/t RQ /(Et ),
 Q0 �
 �2 = −2� 2 0 2
 Q0

θ2 = −2� 3(1 − µ2 )RQ0 /(Et22 ),
M0 2 )RM /(Et 2 ),
(5)
 �M2 0 = 2 3(1 −2µ 3/4 √0 2


θ2 = 4[3(1 − µ )] R/t2 M0 /(Et22 ).
When t1 = t2, substituting Eqs. (2), (3), (4) and (5) into Eq. (1) yields the edge force and edge moment as
follows:
 √  2
Q0 = 4 √P Rt 2 ab ,
8 3(1−µ ) (6)
M0 = 0.

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 3

Vol.:(0123456789)
 www.nature.com/scientificreports/

Q0 Q0
M0 (1) p (1) p
M0
M0 M0
Q0 Q0
=
(2) (2)
p p

(a) Total load (b) Internal pressure

Q0 Q0
(1) M0 (1)
M0
M0 M0
Q0 Q0
+ +
(2) (2)

(c) Edge force (d) Edge moment

Figure 3.  Deformation diagram of the connection boundary.

Solution of the theoretical stress in the discontinuous region of the elliptical head. The con-
nection boundary load diagram is shown in Fig. 3, in which (a) is the overall load diagram, (b) is the diagram of
the load caused by the internal pressure P, (c) is the diagram of the load caused by the edge force Q0, and (d) is
the diagram of the load caused by the edge moment M0. The internal pressure P produces the internal membrane
force, which generates the membrane stress, and the edge force Q0 and edge moment M0 produce the internal
bending force, which generates the bending stress. The total stress can be achieved by superposition of the mem-
brane stress and bending stress.
According to the l­ iterature30–32, the total stress in the discontinuous region can be divided into the membrane
stress and bending stress, and the theoretical stress equation of the elliptical head can be expressed as

(Q +M ) (Q +M )
σω = σωP + Nω 0 0 /t ± 6Mω 0 0 /t 2 ,
P (Q0 +M0 ) (Q0 +M0 ) 2 (7)
σθ = σθ + Nθ /t ± 6Mθ /t .

where subscripts θ and ω correspond to the tangential direction of the parallel circle and longitude, respectively.
σω and σθ denote the total longitudinal stress and circumferential stress, respectively. σωP and σθP denote the lon-
Q Q
gitudinal stress and circumferential stress under internal pressure P, respectively. Nω 0, Nθ 0, NωM0, and NθM0 are
the internal forces along the tangential direction of the longitude and parallel circle under the action of the edge
Q Q
force Q0 and edge moment M0; Mω 0, Mθ 0, MωM0, and MθM0 are the internal torques along the tangential direction
of the longitude and parallel circle under the action of the edge force Q0 and edge moment M0.
The internal force and internal torque equations of the elliptical head can be expressed as

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 4

Vol:.(1234567890)
www.nature.com/scientificreports/

Property Value
inner radius R/mm 200
thickness t/mm 4
material 0Cr18Ni9
Poisson’s ratio μ 0.3
elastic modulus E/GPa 206

Table 1.  Material and mechanical properties of the prototype.

Q
 (Q +M )
 Nω
 0 0
= Nω 0 = tan ωe−k1 ω Q0 (cos k1 ω − sin k1 ω)R/R2 ,
 (Q +M ) Q 1 ω cos k ω(a/b)2
√
Nθ 0 0
= Nθ 0 = 0.25PRe−k� R/R2 ,

1


 −k ω
 (Q0 +M 0 ) Q 0 e 1 Q0 R t
 Mω

 = Mω = 4 √ R2 sin k1 ω,

 3(1−µ2 )
Q0
 (Q +M )
0 0 Q0
Mθ = Mθ = µMω , (8)

 � �
4
� √
 k1 = 3 1 − µ2 R1 / R2 t,




R1 = (a4 y 2 + b4 x 2 )1.5 /(a4 b4 ),





R2 = (a4 y 2 + b4 x 2 )0.5 /b2 .



where R1 is the radius of the first curvature, R2 is the radius of the second curvature, x and y are the coordinates
of the measurement point on the elliptical head, a is the long half shaft, and b is the short plate shaft.
Substituting Eq. (8) into Eq. (7) yields the total theoretical stress in the edge region of the elliptical head as
follows:
 √ 2
P tan ωe−k1 ω Rt(cos k1 ω−sin k1 ω)( ab ) R −k1 ω � �2 � R
 σω = PR
 2
2t +
√
4 ± 3PRe√ sin 2k1 ω ab R ,
2
� � 8t 3(1−µ )R2
−k ω
� 4t 3(1−µ )
−k ω
�2 (9)
 σθ = PR R2 PRe 1 cos k1 ω a 2 R 6PRtµe √ 1 sin k1 ω ab 2 RR .
2
� � � �
2t 2 − R1 + R2 ±

4t b 8t 23(1−µ ) 2 2

The theoretical stress equation of the cylinder is expressed as


(Q +M ) (Q +M )
σx = σxP + Nx 0 0 /t ± 6Mx 0 0 /t 2 ,
P (Q0 +M0 ) (Q0 +M0 ) 2 (10)
σθ = σθ + Nθ /t ± 6Mθ /t .

where subscripts θ and x correspond to the tangential direction of the parallel circle and longitude, respectively.
σx and σθ represent the total longitudinal stress and circumferential stress, respectively. σxP and σθP represent the
Q Q
longitudinal stress and circumferential stress under internal pressure P, respectively. Nx 0, Nθ 0, NxM0, and NθM0
denote the internal forces along the tangential direction of the longitude and parallel circle under the action of
Q Q
the edge force Q0 and edge moment M0. Mx 0 , Mθ 0 , MxM0 , and MθM0 denote the internal torque along the tan-
gential direction of the longitude and parallel circle under the action of the edge force Q0 and edge moment M0.
The internal force and internal torque equations of the cylinder can be expressed as
 (Q0 +M0 ) Q
 Nx = Nx 0 = 0,
(Q +M ) Q

N 0 0 = Nθ 0 = −2kRe−kx Q0 cos kx,


 θ(Q0 +M0 )


Mx = −e−kx Q0 sin kx/k, (11)
(Q0 +M0 )


 Mθ = µMx(Q0 +M0 ) ,
� √

 �

k = 4 3 1 − µ2 / Rt.
 �

Substituting Eq. (11) into Eq. (10) yields the total theoretical stress in the edge region of the cylinder.

Stress analysis and verification in the discontinuous region between the elliptical
head and cylinder
To verify the correctness of the theoretical stress in the discontinuous region between the elliptical head and
cylinder, the experimental stress of the elliptical head under different pressures is investigated using an inter-
nal-pressure thin-walled vessel stress measurement experimental device. The accuracy of the theoretical and
experimental stress values in the discontinuous region is verified by comparing the errors of the theoretical and
experimental stress values to provide reference for the stress distribution of the elliptical head.

Measurement of experimental stress in the discontinuous region of the elliptical head. The
material and mechanical properties of the prototype for the experiment are listed in Table 1.
The experimental device is shown in Fig. 4. First, water is added to the pressure pump, and by shaking the
pressure pump longitudinally, the pipeline is pressurized. The pressure is transmitted to the internal pressure

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 5

Vol.:(0123456789)
 www.nature.com/scientificreports/

(a) Physical device

pressure gauge
pressure sensor
strain gauge

pressure pump

internal pressure vessel

(b) Schematic of experimental device

Figure 4.  Internal pressure thin-walled vessel stress measurement experimental device.

Measurement point 1 2 3 4 5 6 7 8 9 10
Distance 0 40 80 120 150 170 190 210 230 250

Table 2.  Distance from measurement point to the vertex of the elliptical head (mm).


 





 E
 D




¶h

Figure 5.  Distribution of measurement points.

vessel to produce stress and strain. The stress is measured using a pressure gauge and transmitted to a computer
through a pressure sensor. The strain is measured using a strain gauge. Ten strain gauge measurement points are
set. The distance from each measurement point to the vertex of the elliptical head is shown in Table 2. Measure-
ment points 1 to 9 are located on the elliptical head, measurement point 10 is located on the cylinder, and the
distribution of the measurement points is shown in Fig. 5.
Considering constant geometric dimensions of the elliptical head, the internal pressure P value is set as
0.1 MPa, 0.2 MPa, 0.3 MPa, 0.4 MPa and 0.5 MPa. By changing the internal pressure P, the influence of the inter-
nal pressure on the experimental values of the longitudinal stress and circumferential stress at each measurement
point is investigated, as shown in Figs. 6 and 7. When the internal pressure P is unchanged, the longitudinal stress
value is always greater than zero, corresponding to tensile stress. Measurement point 1 is located at the vertex, and
measurement point 10 is located at the equator. The longitudinal stress gradually decreases from measurement
points 1 to 10. The longitudinal stress is the maximum at measurement point 1, and the longitudinal stress at
measurement point 10 is half the stress value at measurement point 1. The reduction range from measurement

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 6

Vol:.(1234567890)
www.nature.com/scientificreports/

25 0.1 MPa
0.2 MPa

the longitudinal stress(MPa)

0.3 MPa
20
0.4 MPa
0.5 MPa
15

10

0
0 2 4 6 8 10
measuring point

Figure 6.  Influence of the internal pressure on the experimental values of the longitudinal stress at each
measurement point.

30

25 0.1 MPa
the circumferential stress(MPa)

0.2 MPa
20 0.3 MPa
0.4 MPa
15
0.5 MPa

10

-5
0 2 4 6 8 10
measuring point

Figure 7.  Influence of the internal pressure on the experimental values of the circumferential stress at each
measurement point.

points 8 to 10 is less because these measurement points are located in the discontinuous region, which bears
not only the membrane stress but also the discontinuous stress. In terms of the circumferential stress, the vari-
ation trend is more complex and can be divided into three s­ tages33. In the first stage, from measurement points
1 to 6, the circumferential stress is greater than zero, corresponding to tensile stress; in the second stage, from
measurement points 7 to 9, the circumferential stress is less than zero, corresponding to compressive stress; in
the third stage, the circumferential stress is greater than zero at measurement point 10, corresponding to tensile
stress. From measurement points 7 to 10, the stress first decreases and later increases, similar to the longitudinal
stress, affected by the discontinuous ­stress34. At measurement point 1, the longitudinal stress and circumferential
stress are approximately equal. At measurement point 6, the circumferential stress changes from tensile stress
to compressive stress. This phenomenon occurs because the stress distribution of the elliptical head is complex.
The value of the longitudinal stress is always greater than zero, corresponding to tensile stress, and it gradually
decreases from the vertex to the equator. The circumferential stress is related to a/b. When a/b < ­20.5, the stress
value is always greater than zero, corresponding to tensile stress, and this value gradually decreases from the
vertex to the equator. When a/b = ­20.5, the value is zero at the equator, and when a/b > ­20.5, compressive stress
occurs. In this study, the a/b value of the experimental device is 2, and thus, the longitudinal stress is always
tensile stress, and the circumferential stress is compressive s­ tress35–38.

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 7

Vol.:(0123456789)
 www.nature.com/scientificreports/

Measurement point 1 2 3 4 5 6 7 8 9
R1 400 382 260 253 185 138 96 61 53
R2 400 394 347 343 309 281 248 214 204

Table 3.  Distance between measurement point and edge of the elliptical head (mm).

Measurement point 1 2 3 4 5 6 7 8 9
K 5142 4875 3117 3015 2086 1491 968 577 491

Table 4.  Stress enhancement coefficient for each measurement point.

25 0.1 MPa
0.2 MPa
the longitudinal stress(MPa)

0.3 MPa
20 0.4 MPa
0.5 MPa
15

10

0
0 2 4 6 8 10
measuring point

Figure 8.  Influence of the internal pressure on the theoretical values of the longitudinal stress at each
measurement point.

Calculation of the theoretical stress in the discontinuous region of the elliptical head. Accord-
ing to the arc length formula, the meridian angles (radian system) of measurement points 1 to 9 starting from
the boundary are 1.57, 1.43, 0.62, 0.60, 0.43, 0.33, 0.23, 0.11 and 0.06, respectively. Measurement point 10 is
distributed on the cylinder. According to the conversion, the distance in the meridian direction from the bound-
ary of measurement point 10 is 10 mm. The distances between the measurement points and edge of the elliptical
head are listed in Table 3.
The stress enhancement coefficient of each measurement point, calculated using the abovementioned formula,
is presented in Table 4:
By substituting these data into the total stress calculation formula of the elliptical head and cylinder edge
region, the theoretical longitudinal stress and circumferential stress at measurement points 1 to 10 can be
obtained, as shown in Figs. 8 and 9. The findings are consistent with the variation law of the experimental stress:
For the same internal pressure P, the theoretical longitudinal stress is always greater than zero, corresponding to
tensile stress, and its value gradually decreases from measurement points 1 (vertex) to 10 (equator). The theoreti-
cal circumferential stress is compressive ­stress39. At the vertex, the theoretical longitudinal stress is approximately
equal to the theoretical circumferential s­ tress40.

Comparison of two analysis methods for the elliptical head. To compare the accuracy of the two
analysis methods of the elliptical head, the internal pressure load of 0.1 MPa is considered as an example, and
the comparison of the theoretical stress and experimental stress is presented in Figs. 10 and 11. The variation law
of the theoretical stress of the elliptical head is consistent with that of the experimental stress, which verifies the
correctness and reliability of the theoretical method. For the longitudinal stress, a large error is noted between
the theoretical and experimental values at measurement point 1, which can be attributed to the inaccurate loca-
tion of the strain gauge or sensitivity of the temperature and test instrument.

Finite element analysis of the elliptical head. Considering the elliptical head as the research object,
an FEM analysis is conducted using ANSYS software under an internal pressure load of 0.1 The model is shown
in Fig. 12a, and the distributions of the longitudinal stress and circumferential stress are shown in Fig. 12b,c,

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 8

Vol:.(1234567890)
www.nature.com/scientificreports/

25

the circumferential stress(MPa)

0.1 MPa
20
0.2 MPa
0.3 MPa
15 0.4 MPa
0.5 MPa
10

-5
0 2 4 6 8 10
measuring point

Figure 9.  Influence of the internal pressure on the theoretical values of the circumferential stress at each
measurement point.

6 20
the longitudinal stress(MPa)

the theoretical stress

error percentage(%)
5 10
the experimental stress

4 0

3 -10

2 -20
0 2 4 6 8 10
measuring point

Figure 10.  Longitudinal stress obtained using the two analysis methods.

80
the theoretical stress
6 the experimental stress
the circumferential stress(MPa)

60
error percentage(%)

4
40

2
20

0 0

-20
0 2 4 6 8 10
measuring point

Figure 11.  Circumferential stress obtained using the two analysis methods.

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 9

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 12.  Finite element analysis of the elliptical head.

20
6

10
the longitudinal stress(M Pa)

the FEM stress

error percentage(% )
5
the experimental stress 0

-10
4

-20

3
-30

-40
2

-50
0 2 4 6 8 10
measuring point

Figure 13.  FEM-obtained longitudinal stress of the elliptical head.

respectively. The longitudinal stress is always greater than zero, corresponding to tensile stress. The circumferen-
tial stress decreases gradually from the apex to the equator, and compressive stress occurs.
To compare the accuracy of the FEM-obtained stress and experimental stress of the elliptical head, the
internal pressure load of 0.1 MPa is considered as an example, and the comparison of the FEM-obtained stress
and experimental stress is shown in Figs. 13 and 14. The FEM-obtained longitudinal stress first decreases, sub-
sequently increases, and later gradually decreases from measurement points 1 to 10. This phenomenon is highly
inconsistent with the variation trend of the experimental longitudinal stress, and the relative error percentage is
up to 45%. The FEM circumferential stress first decreases and later increases from measurement points 1 to 10,
and compressive stress occurs, consistent with the change trend of the experimental longitudinal stress. However,
the relative error percentage is extremely large, up to even 600%. The comparison results show that the theoretical
formula is superior to the FEM for solving this problem.

Conclusion

1. The theoretical stress calculation formula for the discontinuous region between the elliptical head and cyl-
inder is obtained considering the deformation continuity equation, edge force, edge moment, internal force
and internal moment.
2. When the internal pressure load is constant, the theoretical meridian stress is always greater than zero, cor-
responding to tensile stress, and this value decreases gradually from the vertex to the equator. The variation
trend for the theoretical circumferential is relatively complex and can be divided into three stages; moreover,
this stress is compressive. At the vertex, the longitudinal stress and circumferential stress are approximately
equal. The change from measurement points 8 to 10 is influenced by the discontinuous stress, and the vari-
ation trend is abrupt.
3. By comparing and analysing the theoretical stress and experimental stress in the discontinuous region
between the elliptical head and cylinder, the accuracy and applicability of the theoretical stress calculation
formula in the discontinuous region are verified.

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 10

Vol:.(1234567890)
www.nature.com/scientificreports/

the FEM stress

6 200
the experimental stress

the circumferential stress(MPa)

4

error percentage(%)
0

2
-200

0
-400

-2
-600

-4
0 2 4 6 8 10
measuring point

Figure 14.  FEM-obtained circumferential stress of the elliptical head.

Received: 3 August 2021; Accepted: 15 November 2021

References
1. Zheng, J. Y., Dong, Q. W. & Sang, Z. F. Process Equipment Design (Chemical Industry Press, 2010).
2. Wang, Z. W. & Cai, R. L. Chemical Vessel Design (Chemical Industry Press, 2005).
3. Heng, J. Y., Wu, L. L. & Shi, J. F. Extreme pressure equipment. Chin. J. Mech. Eng. 24, 202–206 (2011).
4. Li, Y., Zhang, X. F. & Cai, Y. B. The influence of transition zone between cylinder and head of high pressure vessel on stress con-
centration and its optimization design. Chem. Mach. 43, 320–323 (2016).
5. Han, C., Kim, M. & Yoon, E. S. A hierarchical decision procedure for productivity innovation in large-scale petrochemical pro-
cesses. Comput. Chem. Eng. 32, 71–75 (2008).
6. Parnas, L. & Katirci, N. Design of fiber-reinforced composite pressure vessels under various loading conditions. Compos. Struct.
58, 83–95 (2002).
7. Zu, L., Koussios, S. & Beukers, A. Shape optimization of filament wound articulated pressure vessels based on non-geodesic
trajectories. Compos. Struct. 92, 339–346 (2010).
8. Skopinskii, V. N. et al. Plastic limit loads for cylindrical shell intersections under combined loading. Int. J. Press. Vessels Pip. 126,
8–16 (2015).
9. Leua, S. Y. et al. Plastic limit pressure of spherical vessels with combined hardening involving large deformation. Int. J. Press. Vessels
Pip. 114, 16–22 (2014).
10. Neilson, R. et al. A comparison of plastic collapse and limit loads for single mitred pipe bends under in-plane bending. Int. J. Press.
Vessels Pip. 87, 550–558 (2010).
11. Leu, S. Y. Limit analysis of strain-hardening viscoplastic cylinders under internal pressure by using the velocity control: Analytical
and numerical investigation. Int. J. Mech. Sci. 50, 1578–1585 (2008).
12. Lei, Y. A review of limit load solutions for cylinders with axial cracks and development of new solutions. Int. J. Press. Vessels Pip.
85, 825–850 (2008).
13. Fu, L. et al. Stress distribution of elliptical head using three stress analysis methods. Modern Manuf. Eng. 15, 71–75 (2015).
14. Zheng, J. et al. Effect of shape imperfection on the buckling of large-scale thin-walled ellipsoidal head in steel nuclear containment.
Thin-Walled Struct. 124, 514–522 (2018).
15. Li, K. et al. Buckling behavior of large-scale thin-walled ellipsoidal head under internal pressure. Thin-Walled Struct. 141, 260–274
(2019).
16. Li, K. et al. Experimental investigation on buckling of ellipsoidal head of steel nuclear containment. Int. J. Press. Vessel Technol.
139, 061201–061209 (2017).
17. Roche, R. L. & Autrusson, B. Experimental tests on buckling of torispherical heads and methods of plastic bifurcation analysis. J.
Press. Vessel Technol. 108, 138–145 (1986).
18. Wang, H., Ding, S., Li, L., Sang, Z. & Krakauer, B. W. Buckling of thin-walled torispherical heads in water heater tanks. Thin-Walled
Struct. 85, 156–164 (2014).
19. Magnucki, K., Jasion, P. & Rodak, M. Strength and buckling of an untypical dished head of a cylindrical pressure vessel. Int. J.
Press. Vessels Pip. 161, 17–21 (2018).
20. Muscat, M. & Camilleri, D. Comparison between different design approaches to prevent buckling of torispherical heads under
internal pressure. Int. J. Press. Vessels Pip. 4, 61–66 (2013).
21. Xi, W. & Kui, Z. Theoretical solution and finite element solution for an orthotropic thick cylindrical shell under impact load. J.
Sound Vib. 236, 129–140 (2000).
22. Haojiang, D., Huiming, W. & Weiqiu, C. A theoretical solution of cylindrical shells for axisymmetric plain strain elastodynamic
problems. Appl. Math. Mech. 23, 138–145 (2002).
23. Kabir, M. Z. Finite element analysis of composite pressure vessels with a load sharing metallic liner. Compos. Struct. 49, 247–255
(2000).
24. Duffey, T. A. & Romero, C. Strain growth in spherical explosive chambers subjected to internal blast loading. Int. J. Impact Eng.
28, 967–983 (2003).
25. Skopinsky, V. N. Strain stresses in ellipsoidal pressure vessel heads with noncentral nozzle. Nucl. Eng. Des. 198, 317–323 (2000).
26. Tangbanjongkij, C., Chucheepsakul, S. & Jiammeepreecha, W. Large displacement analysis of ellipsoidal pressure vessel heads
using the fundamental of differential geometry. Int. J. Press. Vessels Pip. 172, 337–347 (2019).
27. Jiammeepreecha, W. & Chucheepsakul, S. Nonlinear static analysis of an underwater elastic semi-toroidal shell. Thin-Walled Struct.
116, 12–18 (2017).

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 11

Vol.:(0123456789)
 www.nature.com/scientificreports/

28. Zingoni, A. Stress analysis of a storage vessel in the form of a complete triaxial ellipsoid, hydrostatic effects, pressure vessel heads
due to thrust loads. Int. J. Press. Vessels Pip. 62, 269–279 (1995).
29. Magnucki, K. & Lewiński, J. Fully stressed head of a pressure vessel. Thin-Walled Struct. 38, 167–178 (2000).
30. Chen, X. Fundamentals of Process Equipment Mechanics (Chemical Industry Press, 2006).
31. Ding, B. M. & Huang, Z. L. Chemical Containers (Chemical Industry Press, 2003).
32. Jin, L. et al. Axisymmetric model of the sealing cylinder in service: Analytical solutions. J. Mech. 37, 404–414 (2021).
33. Magnucki, K., Szyc, W. & Lewiński, J. Minimization of stress concentration factor in cylindrical pressure vessels with ellipsoidal
heads. Int. J. Press. Vessels Pip. 79, 841–846 (2002).
34. Jasion, P. & Magnucki, K. Elastic buckling of clothoidal-spherical shells under external pressure—Theoretical study. Thin-Walled
Struct. 86, 18–23 (2015).
35. Zheng, J.-Y. Design philosophy of flat ribbon wound layered pressure vessel. Int. J. Press. Vessels Pip. 75, 57–61 (1998).
36. Senalp, A. Z. Investigation of the effects of perturbation forces to buckling in internally pressurized torispherical pressure vessel
heads. Adv. Eng. Softw. 45, 232–238 (2018).
37. Miller, C. D. Buckling criteria for torispherical heads under internal pressure. J. Press. Vessel Technol. 123, 318–323 (2001).
38. Mackenzie, D., Boyle, J. T., Boyle, J. T. & Hamilton, R. The elastic compensation method for limit and shakedown analysis: A review.
J. Strain Anal. 35, 171–187 (2000).
39. Athiannan, K. & Palaninathan, R. Buckling of cylindrical shells under transverse shear. Thin-Walled Struct. 42, 1307–1328 (2004).
40. Li, H. & Mackenzie, D. A characterising gross plastic deformation in design by analysis. Int. J. Press. Vessels Pip. 82, 777–786 (2005).

Acknowledgements
This work was supported by the National Natural Science Foundation of China (21763030), Shanxi Provincial
Innovation Capacity Support Plan (2020TD-032), Shanxi Provincial Department of Education Key Scientific
Research Plan (Key Laboratory) (19js069), Yulin Science and Technology Plan in 2019 (2019-83-7), and Shanxi
Science and Technology Department Project (2021JM-522).

Author contributions
Z.W., Z.Z., J.C. wrote the main manuscript text and J.B. prepared Figs. 1, 2, 3, 4, 5, 6, 7, 8 and 9. All authors
reviewed the manuscript.

Competing interests
The authors declare no competing interests.

Additional information
Correspondence and requests for materials should be addressed to Z.W.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

© The Author(s) 2021

Scientific Reports | (2021) 11:22900 | https://doi.org/10.1038/s41598-021-02397-7 12

Vol:.(1234567890)