applied

sciences
Article
Study on Residual Stress of Welded Hoop Structure
Wenbo Ma 1, *, Heng Zhang 1 , Wei Zhu 2 , Fu Xu 1 and Caiqian Yang 1, *
1 College of Civil Engineering and Mechanics, Xiangtan University, Xiangtan 410075, China;
zhangheng940610@163.com (H.Z.); xufu@xtu.edu.cn (F.X.)
2 School of Mathematics and Computational Science, Xiangtan University, Xiangtan 410075, China;
zhuwei@xtu.edu.cn
* Correspondence: mawenbo@xtu.edu.cn (W.M.); ycqjxx@seu.edu.cn (C.Y.);
Tel.: +86-1587-483-1227 (W.M.); +86-1896-361-2628 (C.Y.)




Received: 17 February 2020; Accepted: 10 April 2020; Published: 20 April 2020


Abstract: Residual stress is inevitable during welding, which will greatly affect the reliability of the
structure. The purpose of this paper was to study the residual stress of the hoop structure caused
by the cooling shrinkage of the weld when the outer cylinder was wrapped and welded under the
condition of the existing inner cylinder. In this paper, the “method of killing activating elements” of
ANSYS was used to simulate the three-dimensional finite element of the hoop structure. In the case
of applying interlayer friction, the welding-forming process and welding circumferential residual
stress of the hoop structure were analyzed. The blind hole method was used to test the residual
stress distribution of the hoop structure, and the test results were compared with the finite element
simulation results to verify the reliability of the simulation calculation method and the reliability of
the calculation results. Then, the influence factors of the maximum welding residual stress of the
hoop structure were studied. The results show that the maximum residual stress of the outer plate
surface of the hoop structure decreases with the increase of the welding energy, the thickness of the
laminate, the width of the weld seam, the welding speed, and the radius of the container. Based on the
results of numerical simulation, the ternary first-order equations of the maximum residual stress of
the hoop structure with respect to the welding speed, the thickness of the laminate, and the width of
the weld seam were established. Then, the optimal welding parameters were obtained by optimizing
the equations, which provided an important basis for the safe use and optimal design of the welding
hoop structure.

Keywords: welding; hoop structure; numerical simulation; finite element model; residual stress

1. Introduction
The layered cylindrical vessel has wide engineering applications [1–3]. It can be divided into
several categories depending on the layered vessel type, such as concentric wrapped shell, coil-wound,
shrink fit, and spiral wrapped [4,5]. It is a common problem to crack propagation in the circumferential
weld of a layered cylindrical vessel. It is easy for such cracks to cause structural failure when they
are subjected to local stress caused by an external load, thermal stress, and residual stress [6–9].
Therefore, it is necessary to predict residual welding stress and evaluate it effectively [10].
The welding hoop structure is a kind of constriction structure produced by welding shrinkage in
the layered cylindrical vessel during layered binding welding. This tightening structure will generate
a large circumferential residual tensile stress on the laminate, especially around the weld. This kind
of residual stress greatly increases the risk of cracks in the weld joints and reduces the reliability
of the welded structure. However, for large weldments, it is complicated to carry out an overall
tempering treatment, and a large number of residual stresses are difficult to be eliminated [11,12].

Appl. Sci. 2020, 10, 2838; doi:10.3390/app10082838 www.mdpi.com/journal/applsci

Appl. Sci. 2020, 10, 2838 2 of 16

Therefore, it is very important to study the residual stress and optimize the welding process of the
welding hoop structure.
Combining the experimental measurement with numerical simulation is the most effective solution
to investigate the residual welding stress [13]. In recent years, the finite element method has been
proved to be useful and compelling to evaluate the welding temperature field, residual stress field,
and the effect of post-weld heat treatment [14–21]. There is much software that can perform finite
element simulation of the welding process, such as SYSWELD, Simufact Welding, ANSYS, ABAQUS, etc.
They can all simulate the welding process very well [22–24]. At the same time, software such as
ANSYS can also simulate the multilayer structure very well [25]. In the case of layered pressure vessels,
Xu and Wang et al. improved the welding structure of the layered urea reactor and evaluated the
modified structure by ANSYS software [26]. Then, they used ABAQUS to predict the residual stress in
layer-to-layer joints [27]. Xu and We et al. used ABAQUS to simulate the welding temperature and
predict residual stress in welding and the heat-affected zone between nozzle and nozzle [28]. Xu and
Wang et al. used ABAQUS to study the distribution of residual stress in the weld and heat-affected
zone of the delamination joint with the cladding layer [29]. However, their research ignored the
characteristics of the hoop structure.
According to the actual manufacturing process and structural characteristics of the wrapped
pressure vessel, the inner and outer surfaces of the inner cylinder and the laminate cannot be processed
into absolutely smooth surfaces, but instead surfaces with a certain roughness. Therefore, during the
bandaging process, whether under the action of the bandaging force or when the longitudinal seam of
the laminate cools and contracts, there is interlayer friction in the relative movement of the laminate [30].
In many welding residual stress studies, the simplification of the numerical model generally ignored
the characteristics of the interlayer friction and the clamping structure, which would inevitably lead to
the gap between the simulation results and the actual data. It could not provide the correct theoretical
basis for the analysis and prediction of residual welding stress. Therefore, the analysis of the residual
stress distribution rule of the welding hoop structure and the influence of the welding process on
the residual welding stress is of great significance for the formulation of the correct welding process,
the improvement of the reliability of the welding structure, and the extension of the equipment life.
In this paper, a banded pressure vessel was used as the modeling prototype, and ANSYS was used to
carry out a three-dimensional finite element numerical simulation of the hoop structure. The welding
residual stress distribution trend was studied. The simulation results were compared with the welding
test results under the same process conditions, and the reliability of the simulation calculation method
and the correctness of the calculation results were verified. Then, the influence factors of the model’s
maximum welding residual stress were studied to provide a basis for the safe use and optimal design
of the welding hoop structure.

2. Finite Element Model

2.1. Material Physical Properties

The change of thermophysical properties and mechanical properties of materials with temperature
will significantly affect the accuracy of the welding simulation calculation results. Therefore, a bilinear
follow-up elastoplastic constitutive model is established to simulate the change of elastic modulus
caused by metal exciting heat and cooling during welding. It is assumed in the numerical simulation
that all materials conform to the Mises yield criterion and the bilinear follow-up hardening model [31,32].
The constitutive relationship parameters are shown in Table 1. The thermophysical performance
parameters are shown in Table 2 [33].
Appl. Sci. 2020, 10, x FOR PEER REVIEW 3 of 16
Appl. Sci. 2020, 10, 2838 3 of 16

Table 1. The parameters of the constitutive relation of bilinear dynamic elastic–plastic.

Table 1. The parameters of the constitutive relation of bilinear dynamic elastic–plastic.
Temperature Elastic Modulus before Yield Yield Stress Elastic Modulus after Yielding
T(°C)
Temperature E(GPa)before Yield
Elastic Modulus σ(MPa)
Yield Stress E’(GPa) after Yielding
Elastic Modulus
◦
T20( C) E203
(GPa) σ
356
(MPa) E’ (GPa)
20.3
10020 194
203 321356 19.420.3
200
100 188
194 321321 18.819.4
200
300 188
186 312321 18.618.8
300 186 312 18.6
400 174 283 17.4
400 174 283 17.4
500
500 161
161 249249 16.116.1
600
600 136
136 170170 13.613.6
1000
1000 1010 1111 1 1
1500
1500 1010 55 1 1
2000 1 0.5 0.1
2000 1 0.5 0.1

Table2.
Table Thermal–physicalproperties
2.Thermal–physical propertiesparameters
parametersof
ofQ345R
Q345R steel.
steel.

Thermal Thermal
ThermalExpansion
Expansion Specific
Specific Heat Poisson’s
Poisson’s
TemperatureDensity
Temperature Density EnthalpyEnthalpy Thermal Conductivity
Conductivity Coefficient
Coefficient Capacity
Capacity Ratio
Ratio
T (◦ C)
T(°C) ρ (kg·m
ρ(kg·m −3) −3 )h(J·m ) −3 ) λ(W·m·°C−1◦) −1
h−3(J·m
λ (W·m· C ) α (W·m
α(W·m−2−2·K−1
·K−1)) (J·Kg· ◦ C−1−1) )
CC(J·Kg·°C µµ
00 7800
7800 — — 18.918.9 1.10×10
1.10
-5
× 10−5 580
580 0.35
0.35
400
400 7550
7550 1664 1664 28.35
28.35 1.36×10
1.36 × 10−5
-5 756
756 0.35
0.35
600
600 7550
7550 2779 2779 31.08
31.08 1.39×10
1.39 × 10
-5
−5 840
840 0.35
0.35
800
800 7550
7550 3406 3406 32.76
32.76 1.48×10
1.48 × 10−5
-5 882
882 0.35
0.35
1000
1000 7550
7550 4079 4079 31.531.5 1.34×10
1.34 × 10
-5
−5 756
756 0.35
0.35
1500
1500 7550
7550 4797 4797 28.528.5 1.33×10−5
-5 797
797 0.35
0.35
1.33 × 10

2.2. Model Establishment

2.2. Model Establishment
Using ANSYS software, a banded pressure vessel (Figure 1) was used as a prototype to carry
Using ANSYS software, a banded pressure vessel (Figure 1) was used as a prototype to carry out
out numerical simulation calculations on the welding process of the double-layer hoop structure. The
numerical simulation calculations on the welding process of the double-layer hoop structure. The inner
inner diameter of the structure is R. The welding joint adopts a V-shaped groove, and the angle of
diameter of the structure is R. The welding joint adopts a V-shaped groove, and the angle of the weld
the weld groove is 40°. The thickness of the weld (the thickness of the outer plate) is b, and the weld
groove is 40◦ . The thickness of the weld (the thickness of the outer plate) is b, and the weld width is δ.
width is δ. Based on the actual welding of the layered pressure vessels, displacement constraints were
Based on the actual welding of the layered pressure vessels, displacement constraints were set at the
set at the bottom and left side of the hoop structure model. Set the interlayer friction coefficient to 0.6
bottom and left side of the hoop structure model. Set the interlayer friction coefficient to 0.6 to simulate
to simulate the interlayer friction. The coupling element SOLID5 with temperature and displacement
the interlayer friction. The coupling element SOLID5 with temperature and displacement degrees of
degrees of freedom was selected to perform thermal–structural coupling analysis of the welding
freedom was selected to perform thermal–structural coupling analysis of the welding process by the
process by the direct method. In the element mesh division, considering that there are significant
direct method. In the element mesh division, considering that there are significant temperature and
temperature and stress gradients in the weld area and the near-weld area, the mesh was densified so
stress gradients in the weld area and the near-weld area, the mesh was densified so that the mesh size
that the mesh size of the weld and its surrounding area is not more than 2 mm. The mesh division
of the weld and its surrounding area is not more than 2 mm. The mesh division results are shown in
results are shown in Figure 2. The gray mesh area in the figure is the weld joint.
Figure 2. The gray mesh area in the figure is the weld joint.

Figure 1. Schematic diagram of a bandaged pressure vessel.

Figure 1. Schematic diagram of a bandaged pressure vessel.
Appl. Sci. 2020, 10, 2838 4 of 16
Appl. Sci. 2020, 10, x FOR PEER REVIEW 4 of 16

Figure
Figure2.2.Model
Modelmeshing
meshingdiagram.
diagram.

In
Inthe
thewelding
weldingsimulation,
simulation,the theinner
innerand andouter
outersurfaces
surfacesof ofthe
themodel
modelwerewereconvective
convectiveboundary
boundary
conditions.
conditions.The
Theconvection
convectioncoefficient
coefficient is
is 10 W·m−2 −2·K
·K−1−1
, and
, andthe
theemissivity
emissivityisis0.85.
0.85.The
Theinitial
initialambient
ambient
temperature
temperaturewas ◦
was2020°C.C.InInthe
thesimulation
simulation process, thethe
process, welding
weldingseam
seamis divided intointo
is divided several layers,
several so
layers,
that the the
so that depth
depthof each layer
of each layeris very small,
is very small, and
andthe
theGaussian
Gaussianheatheatsource
sourcecancansimulate
simulatethis
thissituation
situation
well
well[34].
[34].So,
So,ininthe
thethermal
thermalanalysis,
analysis,ititisisassumed
assumedthat thatthetheheat
heatinput
inputof ofthe
thewelding
weldingarcarcmeets
meetsthe
the
Gaussian
Gaussiandistribution.
distribution.TheTheexpression
expressionof ofheat
heatflow
flowdistribution
distributionisis
( − Kr )
2

qq((rr) )== qqm

me
2
e(−Kr ,) , (1)(1)

where qq((r)r )is is

where thethe heat
heat flowflow density
density at the
at the center
center of the
of the heatheat source,
source, qm is
qm is the the maximum
maximum heatdensity
heat flow flow
at the center
density at theof the heat
center of thesource, and K isand
heat source, the heat
K isconcentration factor.
the heat concentration factor.
In actual welding, the base metal will melt, change from solid solid
In actual welding, the base metal will melt, change from to liquid,
to liquid, and absorb
and absorb energy,energy,
while
while
the the molten
molten pool partpoolwill
partsolidify,
will solidify, change
change fromfrom liquid
liquid to solid,
to solid, and and emit
emit heat.Therefore,
heat. Therefore,the
thelatent
latent
heat of phase transition also has an influence on the temperature change,
heat of phase transition also has an influence on the temperature change, which must be considered which must be considered
whencalculating
when calculatingthe thetemperature
temperaturechangechangein inthe
thecrystallization
crystallizationtemperature
temperaturerange.range.Otherwise,
Otherwise,ititwill
will
causeaalarge
cause largedeviation
deviationin inthe
thecalculation
calculationresult.
result.ForForthe
thetreatment
treatmentof oflatent
latentheat
heatofofsolid–liquid
solid–liquidphase
phase
change, there are mainly equivalent specific heat methods: the enthalpy method
change, there are mainly equivalent specific heat methods: the enthalpy method and the temperature and the temperature
risemethod.
rise method.Among Amongthem,them,the theequivalent
equivalentspecific
specificheat
heatmethod
methodisisrelatively
relativelysimple
simpleand
andpractical
practicaland
and
hashigh
has highcalculation
calculationefficiency.
efficiency. This
Thismethod
methodisisused usedin in the
the simulation
simulation of of this
this article.
article. The
Theso-called
so-called
equivalent specific heat method essentially converts the latent heat of crystallization
equivalent specific heat method essentially converts the latent heat of crystallization into specific heat into specific heat
andadds
and addsititto tothe
theactual
actualspecific
specificheat
heatasasthe
thecorrection
correctionvalue
valueof ofthe
thespecific
specificheat
heatof
ofthe
thealloy
alloyininthe
the
crystallization temperature range. The calculation formula
crystallization temperature range. The calculation formula is as follows: is as follows:

∆Q
Δ Q
=
CCee = , (2)(2)
Δ∆T
T
Cp = Ce + C, (3)
C p = Ce + C , (3)
where ∆Q is the amount of heat change during a phase change (J/kg), ∆T is the amount of temperature
changeΔ(Q
where isCthe
◦ C), amount
is the of heat
specific heat change during
when latent a phase
heat is not change
considered(J/kg), ΔT◦ C)),
(J/(kg· is the
Ce amount of
is the specific heat
◦ ◦
change (J/(kg·
temperature C)), and
change Cp Cis the
(°C), equivalent
is the specific
specific heat whenheat (J/(kg·
latent heat is not considered (J/(kg·°C)), Ce
C)).
The APDL language of ANSYS and the “method of killing
is the specific heat change (J/(kg·°C)), and C p is the equivalent specific activating elements” technology
heat (J/(kg·°C)).
was used to simulate the multiple metal filling and heat source movement in the welding process.
The APDL language of ANSYS and the “method of killing activating elements” technology was
Welding process simulation is shown in Figure 3, where the arrow represents the welding sequence.
used to simulate the multiple metal filling and heat source movement in the welding process.
Based on the actual welding process, under the cooperation of the rectangular coordinate system and
Welding process simulation is shown in Figure 3, where the arrow represents the welding sequence.
cylindrical coordinate system, the cyclic dynamic thermal analysis and stress analysis were performed
Based on the actual welding process, under the cooperation of the rectangular coordinate system and
on the welding process of the welded joint structure, and the distribution of residual stress field after
cylindrical coordinate system, the cyclic dynamic thermal analysis and stress analysis were
welding was obtained.
performed on the welding process of the welded joint structure, and the distribution of residual stress
field after welding was obtained.
Appl.
Appl. Sci.
Sci.2020,
2020, 10,
10, x2838
FOR PEER REVIEW 5 5of
of 16
16
Appl. Sci. 2020, 10, x FOR PEER REVIEW 5 of 16

Figure 3.
Figure Schematic diagram
3. Schematic diagram of
of the
the welding
welding process.
process.
Figure 3. Schematic diagram of the welding process.
3. Results
3. Results
3. Results
3.1. Analysis of Temperature State in the Welding Process of Hoop Structure
3.1. Analysis of Temperature State in the Welding Process of Hoop Structure
3.1. Analysis
Based on of the
Temperature State in of
welding process theaWelding
pressureProcess
vesselofinHoop Structure
practical engineering as a prototype, a finite
Based on the welding process of a pressure vessel in practical engineering as a prototype, a finite
element
Basedmodel
on theof welding
the welded hoopofstructure
process a pressure was established.
vessel in practical In the simulation,
engineering as athe inner anda outer
prototype, finite
element model of the welded hoop structure was established. In the simulation, the inner and outer
cylinders were set to Q345R steel. The inner diameter of the laminate
element model of the welded hoop structure was established. In the simulation, the inner and outerwas 1800 mm, the thickness
cylinders were set to Q345R steel. The inner diameter of the laminate was 1800 mm, the thickness of
of the laminate
cylinders were set was to12 mm, steel.
Q345R the weldThewidth δ was 14 mm,
inner diameter of theand the length
laminate of themm,
was 1800 weldment was 1 m.
the thickness of
the laminate was 12 mm, the weld width δ was 14 mm, and the length of the weldment◦ was 1 m. The
The welding head
the laminate was 12 adopted
mm, the a V-shaped
weld width groove,
δ wasand14 mm,the angle of the
and the weldofgroove
length was 40 .was
the weldment The1welding
m. The
welding head adopted a V-shaped groove, and the angle of the weld groove was 40°. The welding
current was
welding head460 A, theaarc
adopted voltagegroove,
V-shaped was 30 and V, andthe the welding
angle of the speed was 10was
weld groove mm/s.40°. The
The welding
welding
current was 460 A, the arc voltage was 30 V, and the welding speed was 10 mm/s. The welding seam
seam was
current wasdivided
460 A, into six voltage
the arc lanes. Figure
was 304 V,is the
andtemperature
the welding program
speed was of10
themm/s.
second pass
The of welding.
welding seam
was divided into six lanes. Figure 4 is the temperature program of the second pass of welding. Figure
Figure 5 is the temperature cycle curve 1 mm away from the fusion line
was divided into six lanes. Figure 4 is the temperature program of the second pass of welding. Figure and 1 mm away from the
5 is the temperature cycle curve 1 mm away from the fusion line and 1 mm away from the lower
lower surface of the upper plate. It can be known from Figure 5 that
5 is the temperature cycle curve 1 mm away from the fusion line and 1 mm away from the lower when the first layer of the weld is
surface of the upper plate. It can be known from Figure 5 that when the first layer of the weld is
welded,
surface of the temperature
the upper plate. around
It canthebe weld
known reaches a maximum,
from Figure 5 that and when then
thethe
firsttemperature
layer of theisweldcooledis
welded, the temperature around the weld reaches a maximum, and then the temperature is cooled
to room temperature. As the distance from the heat source increases,
welded, the temperature around the weld reaches a maximum, and then the temperature is cooled the temperature of the node
to room temperature. As the distance from the heat source increases, the temperature of the node
gradually
to decreases. As the distance from the heat source increases, the temperature of the node
room temperature.
gradually decreases.
It candecreases.
gradually be seen from Figure 4 that with the activation of the unit, the heat source moves in a
It can be seen from Figure 4 that with the activation of the unit, the heat source moves in a
predetermined
It can be seen direction, and the
from Figure temperature
4 that with the of the heat of
activation source and the
the unit, its surroundings
heat source moves decreases
in a
predetermined direction, and the temperature of the heat source and its surroundings decreases with
with increasingdirection,
predetermined distance from
and thethetemperature
center of the ofheat source.
the heat source It can
andbe itsknown from Figure
surroundings decreases4b,cwith
that
increasing distance from the center of the heat source. It can be known from Figure 4b,c that with the
with the movement of the welding heat source, the temperature distribution
increasing distance from the center of the heat source. It can be known from Figure 4b,c that with the gradually reaches the
movement of the welding heat source, the temperature distribution gradually reaches the quasi-
quasi-steady-state stage, and the temperature field moves toward the
movement of the welding heat source, the temperature distribution gradually reaches the quasi- welding line regularly until the
steady-state stage, and the temperature field moves toward the welding line regularly until the end.
end. It can bestage,
steady-state known andfrom
theFigure 4d that field
temperature as themoves
heat source
toward moves to the end
the welding of regularly
line the weld, theuntilwelding
the end.is
It can be known from Figure 4d that as the heat source moves to the end of the weld, the welding is
about to end. The temperature here changes rapidly, which will directly affect
It can be known from Figure 4d that as the heat source moves to the end of the weld, the welding is the size and distribution
about to end. The temperature here changes rapidly, which will directly affect the size and
of the residual
about to end. welding stress at thehere
The temperature end changes
of the weld. rapidly, which will directly affect the size and
distribution of the residual welding stress at the end of the weld.
distribution of the residual welding stress at the end of the weld.

(a) (b)
(a) (b)
Figure 4. Cont.
 Appl. Sci. 2020, 10, x FOR PEER REVIEW 6 of 16

Appl. Sci. 2020, 10, 2838 6 of 16

Appl. Sci. 2020, 10, x FOR PEER REVIEW 6 of 16

(c) (d)
Figure 4. Temperature cloud diagram of the second layer weld welding. (a) t = 205 s; (b) t = 252 s; (c)
t = 300 s; (d) t = 365 s. (c) (d)
Figure 4. Temperature
Temperature cloud
clouddiagram
diagramof
ofthe
thesecond
secondlayer
layerweld
weldwelding.
welding.(a)(a) = 205
t =t 205 s; (b) = 252
t = t252
s; (b) s; (c)
s;
t(c) t = 300
= 300 s; (d) t = 365
t = 365
s; (d) s. s.

Figure 5. Temperature
Figure 5. Temperaturecycle curve
cycle at aatpoint
curve 1 mm
a point from
1 mm thethe
from fusion lineline
fusion andand
1 mm below
1 mm thethe
below lower
lower
surface of the laminate.
surface of the laminate.

3.2. Analysis
Figure 5.ofTemperature
the Distributioncycleofcurve
Circumferential
at a point 1 Residual
mm fromStress in theline
the fusion HoopandStructure
1 mm below the lower
3.2. Analysis of the Distribution of Circumferential Residual Stress in the Hoop Structure
surface of the laminate.
After welding and cooling, the distribution of the overall circumferential residual stress on the
After welding and cooling, the distribution of the overall circumferential residual stress on the
external surface of the structure is shown in Figure 6, where Φ represents the angle from the point
3.2.external
Analysissurface
of the Distribution
of the structure of Circumferential
is shown in Figure Residual 6, Stress
wherein Φthe Hoop Structure
represents the angle from the point of
of the laminate surface to the center of the weld, as shown in Figure 1. The top right corner of the
theAfter
laminate
welding surface to the
and cooling, center of the weld, as shown in Figure 1. The top right corner of theonpicture
picture is the stress cloud diagramthe of distribution
the weld andof thethe overall circumferential
surrounding area. The locationresidual stress
of nodes used the
to
is the
external stress
surface cloud
of thediagram
structure of the
is weld
shown and
in the
Figure surrounding
6, where Φ area. The
represents location
the angleof nodes
from used
the to study
point of
study residual stress is shown in the black line in the figure. It can be seen from the figure that the
the residual
laminate stress
surfaceis shown
to the in the black
center of the line
weld,in the figure.in
as shown It can be seen from
top the figure that the maximum
maximum value of residual welding stress occurs on both Figuresides of1.the
The right
interface corner
between of
thethe picture
weld and
is value
the of
stress residual
cloud welding
diagram of stress
the occurs
weld and ontheboth sides
surrounding of the interface
area. The between
location of the
nodesweldused andto the
studyweld
the weld plate. The stress value exceeds the yield strength of the base metal. With the increase of the
plate.
residual The stress value exceeds the yield strength of the base metal. With the increase of the distance
distance stress
from is theshown
weld,inthe theresidual
black line in the
stress onfigure.
the outer It can be seen
surface offrom the figure
the hoop that the
structure maximum
tends to zero
from
value ofthe weld,welding
residual the residualstressstress
occurson on theboth
outer surface
sides of interface
of during
the the hoop between
structurethe tends
weld to and
zerothegradually.
weld
gradually. Analysis of the stress distribution shows that welding, the thermal expansion strain
Analysis
plate. The of the
stress stress
value distribution
exceeds the shows
yield that of
strength during
the welding,
base metal. the
Withthermal
the expansion
increase of thestrain of the
distance
of the metal in the weld and near the seam area may be constrained by the surrounding colder metal,
frommetal
the in the the
weld, weld and near
residual stress theonseam
the area surface
may beofconstrained by the surrounding zero colder metal,
resulting in compressive plastic strain. In outer
the process the hoop
of welding structure
cooling, tends toshrinkage
the volume gradually.
of the
resulting
Analysis in compressive plastic strain. In the process of welding cooling, the volume shrinkage of the
weld andof the stress distribution
heat-affected zone produces shows that stress,
tensile duringsowelding,
the tensilethestress
thermal expansion
in the weld area strain
showsof thean
weld
metal and
in trend,heat-affected
the weld and zone produces tensile stress, so the tensile stress in the weld area shows an
upward while thenear the stress
tensile seam can areaonlymayoffset
be constrained by the surrounding
part of the compressive stress, socolder
there metal,
is still
upward
resulting intrend, while the
compressive tensile stress
plastic canprocess
only offset part of cooling,
the compressive stress, so there is
thestill
residual compressive stress afterstrain.
welding. In the
The structural of welding
characteristics ofthe
thevolume shrinkage
hoop structure leadofto a
weldresidual
and compressive
heat-affected stress
zone after
produces welding.
tensile The structural
stress, so the characteristics
tensile stress of
in the
the hoop
weld structure
area shows lead
an to
certain residual tensile stress in the cold metal region of the entire container during cooling shrinkage,
a
upward certain residual tensile stress in the cold metal region of the entire container during cooling
and withtrend, while the
the increase tensile
of the stress
distance can the
from only offset
weld, thepart of thetensile
residual compressive stress, so there is still
stress decreases.
shrinkage,
residual and withstress
compressive the increase of the distance
after welding. from the
The structural weld, the residual
characteristics of the hooptensile stress decreases.
structure lead to
a certain residual tensile stress in the cold metal region of the entire container during cooling
shrinkage, and with the increase of the distance from the weld, the residual tensile stress decreases.
Appl. Sci. 2020, 10, 2838 7 of 16

Appl. Sci. 2020, 10, x FOR PEER REVIEW 7 of 16

Figure
Figure 6. The
6. The distribution
distribution curve
curve of circumferential
of circumferential residual
residual stress
stress inhoop
in the the hoop structure.
structure.

Figure
Figure 7 is residual
7 is the the residualstressstress distribution
distribution diagramdiagramof theof outer
the outer surface
surface of clamp
of the the clamp structure
structure at at
the value
the value of Φof(Figure
Φ (Figure 1) between
1) between 0 and 20◦20°.
0 and . In In Figure
Figure 7, A,
7, A, B, B,C,C,D,D, andand EE respectivelycorrespond
respectively correspond to
thewelds
to the weldsand andthethesurrounding
surroundingnodes nodesA,A, B, B,C,C,D,D,and andE. E.
It canIt can be seen
be seen from fromthe the
figurefigure
thatthat
the the
maximum
maximum residual
residual tensile
tensile stressstress appears
appears at the atedge
the edge of weld
of the the weld (point(point B), and
B), and the maximum
the maximum residual
residual
compressive
compressive stressstress
appears appears
at the at theofedge
edge of the welded
the welded plate
plate (point D).(point
Welding D).isWelding
a fast andisnon-uniform
a fast and non-
uniform
thermal cycling thermal
process.cycling
When the process.
welding When the welding
heat source contactsheat source
the weld area,contacts the weldaround
the temperature area, the
temperature
the area around
rapidly rises to thethemelting
area rapidly
point of rises
thetolaminate
the melting pointcausing
material, of the laminate
the soldermaterial,
to fuse withcausing
the the
weldsolder to fuse
plate [35]. Whenwiththe the weld plateis[35].
temperature closeWhen
to the the temperature
melting point of the is close to the
material, themelting point of the
elastic modulus
material,
of the the elastic
steel decreases modulus
sharply. of the
Hence, thesteel
meltingdecreases
zone and sharply.
its edge Hence,
regions theare
melting zone and
more plastic than itsinedge
regions
other regions. areAtmore plastic
this time, than in the
although other regions.
melting zone Atisthis time, although
thermally expanded, theplastic
melting zone is thermally
deformation will
expanded,
occur, which will plastic
not deformation will occur, which
cause high compressive stress.will not cause
However, inhigh compressive
the near-weld stress.
area, the However,
material in
can the
stillnear-weld
maintain the area,basic
the material
properties. can During
still maintain
thermal theexpansion,
basic properties. During thermal
it is restricted and squeezedexpansion,
by it
the is restricted and
surrounding squeezed by thearea,
low-temperature surrounding
which causes low-temperature
local compression area, which causes
plastic local compression
deformation and
plastic deformation
compressive stress in the and compressive
near-seam area,stress
as shown in theinnear-seam
the CD section.area, asDuring
shownthe in the CD section.
cooling processDuring
of
the cooling the
the weldment, process of the weldment,
temperature around thethe weldtemperature
decreasesaroundslowly, the andweld decreases
the weld slowly, and the
and heat-affected
zoneweld and heat-affected
contract, which causeszone contract,
stretching which
to the causesarea,
far seam stretching
as shown to the far seam
in section DEarea,
of theas figure.
shown in
The section DE of the figure.
plastic deformation The plastic
of the metal in the CD deformation
segment due of tothe metal
local in the CD
compression segment
cannot due to local
be contracted
compression
freely, cannot be
but it is stretched in contracted freely, but zone.
the low-temperature it is stretched
Hence, in thethe low-temperature
increasing zone. Hence, the
trend of compressive
increasing
stress becomestrend slower. of At
compressive
the edge ofstress becomes
the weld, it is slower.
in contact Atwiththe edge of the weld,
the welding plate,it with
is in fast
contact
heatwith
the welding
dissipation, plate, with drops,
fast temperature fast heatanddissipation,
rapid shrinkage fast of
temperature
the solder volume, drops, and rapidinshrinkage
resulting high tensile of the
solder
stress volume,
between resulting
the units, in highintensile
as shown section stress
BC. between
The heat the units, as at
dissipation shown in section
the center of theBC. Theisheat
weld
dissipation
slow, at the center
so the temperature of the
drops weldresulting
slowly, is slow, in so the
theslow
temperature drops slowly,
volume shrinkage of theresulting
solder. in Duethetoslow
volume shrinkage
the symmetry of the solder.
of the structure, Due tostress
the tensile the symmetry
is small, as of shown
the structure,
in the ABthe section
tensile stress
of theisfigure.
small, as
shownthe
Therefore, in the AB section
residual stressof the figure. is
distribution Therefore,
shown inthe residual
Figure 7. stress distribution is shown in Figure 7.
Appl. Sci. 2020, 10, 2838 8 of 16
Appl. Sci. 2020, 10, x FOR PEER REVIEW 8 of 16

Appl. Sci. 2020, 10, x FOR PEER REVIEW 8 of 16

7. Distribution
Figure
Figure of circumferential
7. Distribution residual
of circumferential stress
residual around
stress the the
around weld of the
weld hoop
of the structure.
hoop structure.
Figure 7. Distribution of circumferential residual stress around the weld of the hoop structure.

Similar
Similar to tothetheoverall
overall circumferential
circumferential residual stress distribution
distributionofof the outer
outer layer
layerpipe,
pipe,
Similar to the overall circumferentialresidualresidual stress
stress distribution of
thethe
outer layer pipe, the the
thecircumferential
circumferential residual
residual stress stress ofof the
the inner
inner layer
layer pipe
pipe hashas a alarge
large variation
variation near
nearthetheweld,
weld,so the
so the
circumferential residual stress of the inner layer pipe has a large variation near the weld, so the
surrounding
surrounding of the
surrounding weld
of the
of theis taken
weld weld as
is isthe
taken main
takenasastheresearch
the main object.
main researchFigure 8 is a diagram
object.Figure
research object. Figure8 is of the
8 is circumferential
a diagram
a diagram of the
of the
residual stress
circumferential of the
circumferential inner
residual
residualpipe
stresswith
of of
stress the
the Φinner
inner
the value
pipe between
pipe with
with the0◦ΦΦand
value
value10◦between
.between
At the0°top0°
and right
and of the
10°.10°.
At picture
At top
the the is
top right
right
theofnephogram
of the of circumferential
picture is the nephogram residual
of stress around
circumferential the
residual inner
stress
the picture is the nephogram of circumferential residual stress around the inner tube weld. The tube
aroundweld.
the The
inner black
tube line
weld. in the
The
figure is the
black black
linelocation
in in of
linethe thethe
figure selected
figure is the
is the point,
location
location and the
ofof the
the arrow is the
selected
selected direction
point,
point, and
andtheof
thethe selected
arrow
arrow is the point.
direction
is the As of
shown
direction the
of the
in the selected
figure,point.
selected point.
due toAs As
theshown shown
influence in the
of the
in the figure,
welding
figure, due to
due process,the influence
the maximum
to the influence of the welding process,
of thecircumferential
welding process, the maximum
residual tensile
the maximum
stress circumferential
of the inner tube residual
is located tensile
at the stress
center ofofthe
theinner
weld,tube is located
which exceeds at the center limit
of the weld, which
circumferential residual tensile stress of the inner tube is located atthe
theyield
center ofofthe theweld,
material.
which
exceeds the yield limit of the material. As the distance from the center of the weld increases, the value
As exceeds
the distance fromlimit
the yield the center
of the of the weld
material. Asincreases,
the distance thefrom
valuethe of center
the circumferential residual the
of the weld increases, stress
value
of the circumferential residual stress rapidly attenuates; then, it gradually attenuates and tends to be
rapidly
of theattenuates; then, itresidual
circumferential gradually attenuates
stress rapidly and tends tothen,
attenuates; be stable.
it gradually attenuates and tends to be
stable.
stable.

Figure 8.Figure 8. Distribution

Distribution of circumferential
of circumferential residual
residual stress
stress in the
in the weld
weld aroundthe
around theinner
inner tube
tube weld.
weld.

4. Experimental Verification
4. Experimental Verification
Figure 8. Distribution of circumferential residual stress in the weld around the inner tube weld.
In order In
to order
verifytothe
verify the correctness
correctness of thestructure
of the hoop hoop structure simulation,
simulation, the welding
the welding test under
test under the
the same
conditionssame conditions was carried out, and the residual stress of the specimen was measured
was carried out, and the residual stress of the specimen was measured by the blind hole by the blind
4. Experimental
hole method.Verification
In the test, the inner and outer cylinder was Q345R steel. The inner diameter of the
method. In the test, the inner and outer cylinder was Q345R steel. The inner diameter of the plywood
In order to verify the correctness of the hoop structure simulation, the welding test under the
same conditions was carried out, and the residual stress of the specimen was measured by the blind
hole method. In the test, the inner and outer cylinder was Q345R steel. The inner diameter of the
Appl. Sci. 2020, 10, x FOR PEER REVIEW 9 of 16
Appl. Sci. 2020, 10, 2838 9 of 16

plywood was 1800 mm, the thickness of the plywood was 12 mm, the width of the weld was 14 mm,
and the
was 1800length
mm, theof the pipe is 1ofm.
thickness theAplywood
V groovewas shall12bemm,
adoptedthe widthfor theofwelding
the weldjoint,
was and
14 mm, the angle of
and the
the weld
length groove
of the pipeshall be A
is 1 m. 40°. The welding
V groove current
shall be adopted was for460
theA,welding
the arc joint,
voltageandwas
the30 V, the
angle welding
of the weld
speed was 10 mm/s,
◦ and the diameter of the welding wire was 4 mm. During
groove shall be 40 . The welding current was 460 A, the arc voltage was 30 V, the welding speed was the welding process,
electrode
10 arc welding
mm/s, and the diameteris used as welding
of the the base,wire
and wassubmerged
4 mm. During arc welding is usedprocess,
the welding to fill and cover the
electrode arc
surface. is used as the base, and submerged arc welding is used to fill and cover the surface.
welding
Due totothethe
Due largelarge experimental
experimental materials,
materials, non-destructive
non-destructive residual
residual stress stressis inconvenient.
detection detection is
inconvenient.
Therefore, the Therefore,
YC-III typethe YC-III
static type static program-controlled
program-controlled resistance strainresistance
gauge and strain gauge
ZDL-II typeand ZDL-
residual
II type residual stress drilling device were used to test the residual stress of
stress drilling device were used to test the residual stress of the laminate by the blind hole method the laminate by the blind
hole method (Figure 9). After the laminate was wrapped and welded, in
(Figure 9). After the laminate was wrapped and welded, in the stable area of the weld, strain gauges the stable area of the weld,
strainrespectively
were gauges werepastedrespectively
in the pasted
Φ values in the
of 2Φ◦ ,values
5◦ , 10◦of 2°,◦ ,5°,
, 15 45◦10°,
, and15°,
9045°, and 90° directions.
◦ directions. Then, weThen, used
we used the drilling device to drill the blind hole with the hole diameter of
the drilling device to drill the blind hole with the hole diameter of 1.5 mm and the depth of 2.0 mm, 1.5 mm and the depth of
2.0 mm,
and and we measured
we measured the corresponding
the corresponding strain
strain value. Thevalue.
residual Thestress
residual
valuestress
was value was calculated
calculated according
according
to to theofprinciple
the principle of stress
stress release. Therelease. The calculation
calculation formula was: formula was:

E(ε1( +1 ε2 )2 ) ± EE ( ε1 − ε 2 )22+  2ε 2 − ( ε1 + ε 2 )  2 , 2
E ε +ε
σ =
q
(4)
σ1,2 = 4 B (ε1 − ε2 ) + [2ε2 − (ε1 + ε2 )] ,
1,2 ± (4)
4A 4A 4B
1 2ε −（ε + ε ）
θ =1 tg−1-1 2ε22 − (ε11 + ε3 3 ). (5)
θ = 2tg ε -ε . (5)
2 ε33 − 1ε1
Subsequently, the strain release systems A and B
B were
were modified
modified with
with the
the shape-strain
shape-strain specific
parameter, and
energy parameter, and the
the correction
correctionformula
formulawas
was[36]:
[36]:
( −0.341163 S (10S−(8 )10≤−8
2.00417
−0.341163 ) ≤ 2.00417,
A( με / MPa
A(µε/MPa) = ) =  −8 , (6)
(6)
−0.005223 ×
−0.005223 × SS −−0.330356
0.330356 SS((10 10 )) >>2.00417
−8 2.00417

− 0.763261
−0.763261 SS(10 −8
(10−8) )≤ ≤2.00417
(
2.00417
B ( με / MPa
B(µε/MPa) =  ) = .. (7)
(7)
× S − 0.719503 S 10−8
−8
>
−0.0219526 × S − 0.719503 S (10 > 2.00417
−0.0219526 ( ) 2.00417
The stresses at the measuring points in the x and y directions were:
The stresses at the measuring points in the x and y directions were:
σσ1 ++ σ
σ22 ± σσ11 −−σσ22 cos 2θ.
σx,y
σ x,= (8)
y = 22 ± 22 cos 2θ .
1
(8)

The
The comparison
comparison between
between the the test
test results
results andand the
the finite
finite element
element calculation
calculation results
results are
are shown
shown in in
Table 3 and Figure 10. It is worth noting that because the experimental specimen
Table 3 and Figure 10. It is worth noting that because the experimental specimen is large, there would is large, there would
be
be aa certain
certaindeviation
deviationininthe residual
the residualstress measured
stress measuredat the at corresponding pointpoint
the corresponding according to the to
according angle.
the
Therefore,
angle. Therefore, the residual stress distribution trend was used as the verification standard here.be
the residual stress distribution trend was used as the verification standard here. It can It
seen
can be from the
seen figure
from thethat the that
figure distribution trend of trend
the distribution the residual
of the stress on stress
residual the outer surface
on the outerofsurface
the hoop of
structure
the hoop is basicallyisconsistent
structure basically with the numerical
consistent with thesimulation
numerical results, which
simulation can show
results, which thecan
reliability
show the of
the finite element
reliability calculation.
of the finite element calculation.

(a) (b)

Figure 9. Cont.
Appl. Sci. 2020, 10, 2838 10 of 16
Appl. Sci. 2020, 10, x FOR PEER REVIEW 10 of 16

Appl. Sci. 2020, 10, x FOR PEER REVIEW 10 of 16

(c) (d)
(c) (d)
Figure
Figure9.9.Experimental
Experimentalspecimen
specimenand andexperiment
experimentapparatus.
apparatus.(a)(a)Experimental
Experimentalspecimen;
specimen;(b) (b)Drilling
Drilling
Figure 9. Experimental specimen and experiment apparatus. (a) Experimental specimen; (b) Drilling
equipment
equipmentand YC-III
and static
YC-III stressstress
static measuring instrument;
measuring (c) TJ-120-1.5
instrument; strain gagestrain
(c) TJ-120-1.5 rosette;gage
(d) Drilling
rosette;
equipment and YC-III static stress measuring instrument; (c) TJ-120-1.5 strain gage rosette; (d) Drilling
operation.
(d) Drilling operation.
operation.

Table 3. Comparison between the results of the test and the results of the finite element calculation.
Table 3. Comparison between between
Table 3. Comparison the results of the of
the results test
theand the results
test and of of
the results thethefinite
finite element calculation.
element calculation.

Experiment
Experiment
Experiment Results σσEE(MPa)
ResultsResults(MPa)
σE (MPa) Numerical
Numerical
Numerical Simulation Resultsσ
SimulationResults
Simulation Results σσNNN (MPa)
(MPa)
(MPa)
206.14 206.14 269.26
269.26
206.14 269.26
58.22 58.22 138.31
138.31
58.22 138.31
−14.94 −14.94 −27.49
−27.49
−14.94
59.26 59.26
−27.49
24.08
24.08
59.26
15.81 15.81 24.08
7.64
7.64
15.81
−5 −5 7.64
5.23
5.23
−5 5.23

Figure 10. Comparison between the results of the test and the results of the finite element calculation.
Figure 10. Comparison between the results of the test and the results of the finite element calculation.
5. Factors Affecting the Maximum Residual Stress of the Hoop Structure
5. Factors Affecting the Maximum Residual Stress of the Hoop Structure
The distribution of residual welding stress depends on several main factors, such as the
The distribution of residual welding stress depends on several main factors, such as the structural
structural dimensions, material properties, restraint conditions, heat input, number of weld passes,
dimensions,
Figure 10.material
and weldingproperties,
Comparison between
sequence. restraint
the factors
These conditions,
results of the
will heat
testthe
affect and input,
the
maximum number
results of weld
of theresidual
welding finite passes,
element
stress theand welding
oncalculation.
laminate
sequence. These
surfacefactors willstudy,
[16]. In this affectthe
thefollowing
maximum welding
important residual
factors stress on
were selected the laminate surface [16].
for research.
5.InFactors Affecting
this study, the Maximum
the following importantResidual
factorsStress of the Hoop
were selected Structure
for research.
The distribution of residual welding stress depends on several main factors, such as the
structural dimensions, material properties, restraint conditions, heat input, number of weld passes,
and welding sequence. These factors will affect the maximum welding residual stress on the laminate
surface [16]. In this study, the following important factors were selected for research.
Appl. Sci. 2020, 10, 2838 11 of 16

Appl. Sci. 2020, 10, x FOR PEER REVIEW 11 of 16

5.1. Effect of the Radius on the Maximum Residual Stress of the Hoop Structure
5.1. Effect
In order of the Radius
to study on the Maximum
the influence of radiusResidual
on theStress of the Hoop
maximum Structurestress of the welded hoop
residual
structure, theInfinite
orderelement
to studysimulation
the influenceofofa radius
hoop structure with a residual
on the maximum differentstress
radius was
of the carried
welded out.
hoop
The simulation results
structure, are shown
the finite elementin Figure 11.
simulation of As the radius
a hoop structureincreases, the maximum
with a different radius wasresidual
carriedstress
out.
The simulation
of the hoop results are shown
structure decreases. in Figure
The reason 11. As
may be the radius
that as theincreases,
radiustheof maximum residual
the container stress
increases,
of theof
the curvature hoop
the structure decreases.
central heating The
area of reason may be that
the weldment as the radius
decreases. of the container
The welding laminate increases,
tends tothe
be
curvature of the central heating area of the weldment decreases. The welding laminate
flat, which reduces the heat dissipation area at the weld joint and reduces the residual stress caused by tends to be
flat, which reduces the heat dissipation area at the weld joint and reduces the residual stress caused
the uneven contraction of the hoop structure during the cooling shrinkage of the solder.
by the uneven contraction of the hoop structure during the cooling shrinkage of the solder.

Figure 11.Figure 11. The distribution

The distribution curve curve of maximum
of maximum welding
welding residual
residual stress
stress ofofthe
thehoop
hoopstructure
structure with
with
different radiuses.
different radiuses.

5.2. Effect5.2.
of Welding Power and
Effect of Welding Welding
Power SpeedSpeed
and Welding on the
onMaximum Residual
the Maximum Stress
Residual Stressofofthe
theHoop
Hoop Structure
Structure

The totalThe total

heat inputheator
input or internal
internal heat source
heat source in welding
in arc arc welding is the
is the productofofarc
product arcpower
power and
andprocess
process
efficiency. In this simulation method, the correct heat input selection and modeling are
efficiency. In this simulation method, the correct heat input selection and modeling are essential to the essential to
the outcome. It is essential to understand the concept of heat input and heat source efficiency, which
outcome. It is essential to understand the concept of heat input and heat source efficiency, which varies
varies between 65% and 85% for gas metal arc welding [19]. Since energy cannot be fully applied to
between 65% and 85% for gas metal arc welding [19]. Since energy cannot be fully applied to the
the heating weldment, the power that is effective for heating welding is [37]:
heating weldment, the power that is effective for heating welding is [37]:
Q = η Q0 = ηUI , (9)

where η is the thermal welding efficiency, Q = ηQ Q00is=theηUI,

arc power (W), U is the arc voltage (V), and I is (9)
the welding current (A).
where η is theAccording
thermal welding
to Formulaefficiency, Q0 is theofarc
(9), the influence power
current and(W), U ison
voltage thewelding
arc voltage (V), and Itoisthe
is proportional the
welding current (A).
welding power. At the same time, at the same welding output power, the transient temperature field
According to Formula
distribution caused by (9), thesource
heat influence of current
movement andaffect
will also voltage on weldingofisresidual
the distribution proportional
weldingto
the welding power.
stress. Attothe
In order same
study thetime, at the
influence same welding
of welding output
power and power,
welding speedthe
ontransient
the maximumtemperature
welding
residual stress
field distribution causedof by
theheat
clamping
source structure,
movementthe welding
will alsoprocess of the
affect the hoop structure
distribution under different
of residual welding
welding powers and welding speeds was simulated. The influence
stress. In order to study the influence of welding power and welding speed on the maximum curve of welding power and
welding
speed on the maximum welding residual stress is shown in Figure 12.
residual stress of the clamping structure, the welding process of the hoop structure under different
welding powers and welding speeds was simulated. The influence curve of welding power and speed
on the maximum welding residual stress is shown in Figure 12.
Appl. Sci. 2020, 10, 2838 12 of 16

Appl. Sci. 2020, 10, x FOR PEER REVIEW 12 of 16

Figure Figure
12. The12. The distribution curve of maximum welding residual stress of the hoop structure under
distribution curve of maximum welding residual stress of the hoop structure under
different welding powers and welding speeds.
different welding powers and welding speeds.
It can be seen from the figure that as the welding energy increases, the maximum welding
It can be seen from the figure that as the welding energy increases, the maximum welding residual
residual stress of the hoop structure decreases accordingly. It should be that with the rise of welding
stress of the hoop structure decreases accordingly. It should be that with the rise of welding energy,
energy, the temperature of the solder and the melting zone increases, and the elastic modulus of the
the temperature of the solder
welding material decreasesand the melting
sharply. Comparedzonewith increases, and the elastic
the surrounding modulus
cold metal, it can of
bethe welding
regarded
material as a plastic material, so its ability to resist the extrusion in the surrounding low-temperature azone
decreases sharply. Compared with the surrounding cold metal, it can be regarded as plastic
material,wassorelatively
its ability to resist
reduced, the extrusion
resulting in the
in reduced surrounding
compressive stress.low-temperature
In addition, with the zone was relatively
increase of the
reduced, temperature
resulting inofreduced
the weld area in thestress.
compressive melting In state, the surrounding
addition, heating
with the increase area
of the became more
temperature of the
substantial, and the overall heat dissipation time becomes longer, so the
weld area in the melting state, the surrounding heating area became more substantial, and the overall stress produced in the
cooling process
heat dissipation tended to longer,
time becomes decrease.soTherefore,
the stressthe increase in
produced of the
welding energy
cooling can effectively
process tended toreduce
decrease.
the generation of residual stress on the surface of the laminate. As the welding speed increases, the
Therefore, the increase of welding energy can effectively reduce the generation of residual stress on
maximum welding residual stress on the surface of the laminate decreases accordingly. This may be
the surface of the laminate. As the welding speed increases, the maximum welding residual stress
because of the addition of welding speed, which makes the temperature around the weld rise more
on the surface of the laminate decreases accordingly. This may be because of the addition of welding
evenly and can effectively change the uneven expansion caused by the irregular temperature change.
speed, However,
which makes the engineering,
in actual temperature thearound
increasethe weld risespeed
in soldering morewillevenly
causeand can effectively
insufficient change
solder filling,
the uneven expansion
so when consideringcaused by the irregular
the soldering temperature
rate, sufficient change.
solder filling shouldHowever,
be satisfied.in actual engineering,
Therefore, under
the increase in soldering
the condition that speed will can
the solder causebeinsufficient
fully filled, solder filling,
the proper so when
increase of considering
the welding the soldering
speed can
effectively reduce the generation of residual welding stress of the hoop structure
rate, sufficient solder filling should be satisfied. Therefore, under the condition that the solder can after the welding is be
completed.
fully filled, the proper increase of the welding speed can effectively reduce the generation of residual
welding stress of the hoop structure after the welding is completed.
5.3. Effect of Weld Width and Thickness on the Maximum Residual Stress of the Hoop Structure
5.3. Effect ofInWeld Width
order andthe
to study Thickness onofthe
influence Maximum
weld Residual
width and Stress
laminate of the Hoop
thickness on theStructure
maximum residual
stress of the hoop structure, the finite element simulation of the welded clamp structure with different
In order to study the influence of weld width and laminate thickness on the maximum residual
weld widths and the thickness of the hoop structure was performed. The maximum welding residual
stress of the hoop structure, the finite element simulation of the welded clamp structure with different
stress curve is shown in Figure 13. It can be seen from the figure that the maximum residual stress
weld widths andwith
decreases the the
thickness
increaseof
ofthe
weldhoop structure
width was performed.
and laminate thickness. TheThe maximum
reason should bewelding
that withresidual
the
stress curve is shown in Figure 13. It can be seen from the figure that the maximum residual
increase of the weld width, the solder area will increase, resulting in a slower heat dissipation speed stress
decreases with
in the weldthe increase
area, easing of
theweld width
residual and
stress laminate
caused by thethickness. The reason
sharp contraction should
of solder duringbethe
that with
heat
dissipation
the increase of theprocess, and at the
weld width, thesame time,
solder thewill
area increase of the resulting
increase, stress areain
at athe weld will
slower heatreduce the
dissipation
speed inextreme value
the weld of easing
area, residualthe
stress. Whenstress
residual the width
caused ofby
thethe
weld is constant,
sharp contractionthe of
amount
solderofduring
solder the
increases with the increase of the thickness of the laminate, which slows down the
heat dissipation process, and at the same time, the increase of the stress area at the weld will reduce heat dissipation
the extreme value of residual stress. When the width of the weld is constant, the amount of solder
increases with the increase of the thickness of the laminate, which slows down the heat dissipation
Appl. Sci. 2020, 10, 2838 13 of 16
Appl. Sci. 2020, 10, x FOR PEER REVIEW 13 of 16

process
process of of the welding
the welding and reduces
and reduces the residual
the residual tensile
tensile stress
stress causedby
caused bythe
thesharp
sharp contraction
contraction ofofthe
the
weld during
weld during cooling. cooling.

Figure 13. The distribution curve of maximum welding residual stress of the hoop structure under
Figure 13. The distribution curve of maximum welding residual stress of the hoop structure under the
the different thickness values of the laminate.
different thickness values of the laminate.
6. Optimal
6. Optimal DesignDesign of Process
of Process Parameters
Parameters
According to the distribution of the maximum welding residual stress of the hoop structure
According to the distribution of the maximum welding residual stress of the hoop structure under
under different welding parameters, the curve fitting of the maximum welding residual stress under
differentdifferent
welding welding
parameters, the curve fitting of the maximum welding residual stress under different
parameters is carried out, in order to obtain the best welding parameters to
weldingminimize
parameters is carried out, in order
the residual welding stress to obtain
within thethe best welding
allowable parameters
parameters of the to minimize
actual workingthe
residual conditions.
welding stress within the allowable parameters of the actual working conditions.
In this study,
In thisthe maximum
study, residualresidual
the maximum stress ofstress
the welded
of the hoopwelded structure was taken
hoop structure wasastaken
the objective
as the
objective function. The welding speed, the thickness of the laminate, and the
function. The welding speed, the thickness of the laminate, and the width of the weld were represented width of the weld were
represented by the design variables x , x , and x respectively. The reglm function
by the design variables x1 , x2 , and x3 respectively. The reglm function of multiple linear regression
1 2 3 of multiple linear
analysis regression analysis regression
and generalized and generalized regression
analysis were analysis
called in were called into
MATLAB MATLAB
performtoternary
performone-time
ternary
one-time regression fitting on the parameters of Table 4. Table 5 shows the primary regression
regression fitting on the parameters of Table 4. Table 5 shows the primary regression parameters.
parameters. The fitting correlation coefficient is 0.9574, which indicates that the method is effective.
The fitting correlation coefficient is 0.9574, which indicates that the method is effective.
The corresponding p values
Table 4.of the constant
Maximum residual stressxvalue
terms 1 , x2 , under
and xdifferent
3 are allparameters.
less than 0.05, indicating that
these terms in the regression equation are significant. The ternary linear regression equation for the
Data Number No Welding Speed v(m) Shelf Thickness δ(m) Weld Width b(m) Maximum Residual Stress σ(MPa)
maximum residual 1 stress σ is: 0.01 0.008 0.008 260
2 0.01 0.008 0.01 247
3 0.01
F = 525.5378 0.008 − 14305x 0.012
− 13099.1667x − 4121.1111x . 237 (10)
1 2 3
4 0.01 0.01 0.008 218
5 0.01 0.01 0.01 208
According 6to Formula (7),0.01 the residual welding
0.01 stress of 0.012
the hoop structure decreases
200 with the
7
increase of the thickness of the 0.01 0.012 of the weld seam,
laminate, the width 0.008 and the welding 189speed when the
8 0.01 0.012 0.01 182
actual working 9condition allows. 0.01 According to China’s current
0.012 0.012pressure vessel welding
175 standards,
the width of the10butt welded joint0.012 (weld seam) should
0.008 be determined
0.008 according to the223groove, and the
11 0.012 0.008 0.01
form of the groove is determined by equipment and technology. According to the212requirements of
12 0.012 0.008 0.012 203
HG3129-1998, when the welding temperature is 1500 ◦ C, the radius of the container is 1.8 m, the solder
13 0.012 0.01 0.008 176
14 the welding0.012
can be fully filled, speed is 0.018 m/s,0.01
the width of the 0.01weld is 0.014 m, and
168the thickness is
15 0.012 0.01 0.012 161
0.014 m, the optimum solution exists. At this time, the maximum residual stress σmin = 31.89 MPa.
Appl. Sci. 2020, 10, 2838 14 of 16

Table 4. Maximum residual stress value under different parameters.

Welding Speed v Shelf Thickness δ Weld Width b Maximum Residual

Data Number No
(m) (m) (m) Stress σ (MPa)
1 0.01 0.008 0.008 260
2 0.01 0.008 0.01 247
3 0.01 0.008 0.012 237
4 0.01 0.01 0.008 218
5 0.01 0.01 0.01 208
6 0.01 0.01 0.012 200
7 0.01 0.012 0.008 189
8 0.01 0.012 0.01 182
9 0.01 0.012 0.012 175
10 0.012 0.008 0.008 223
11 0.012 0.008 0.01 212
12 0.012 0.008 0.012 203
13 0.012 0.01 0.008 176
14 0.012 0.01 0.01 168
15 0.012 0.01 0.012 161
16 0.012 0.012 0.008 163
17 0.012 0.012 0.01 156
18 0.012 0.012 0.012 150
19 0.014 0.008 0.008 196
20 0.014 0.008 0.01 186
21 0.014 0.008 0.012 178
22 0.014 0.01 0.008 165
23 0.014 0.01 0.01 157
24 0.014 0.01 0.012 150
25 0.014 0.012 0.008 144
26 0.014 0.012 0.01 137
27 0.014 0.012 0.012 132

Table 5. Parameter estimation of ternary regression equation.

Variables Constants x1 x2 x3
estimated value 525.5378 −13099.1667 −14305 −4121.1111
p-value 0.0000 0.0000 0.0000 0.0001

7. Conclusions
This study investigated the residual stress of the hoop structure caused by the cooling shrinkage
of the weld when the outer cylinder was wrapped and welded under the condition of the existing inner
cylinder. In this study, the method of combining finite element simulation and experiment was used to
analyze the welding forming process and welding circumferential residual stress of the hoop structure
under the consideration of the friction between layers. Further study will research the influencing
factors of the maximum welding residual stress of the hoop structure. The main research results in this
study are concluded below:

(1) In the welding direction, the residual stress at the starting and ending arc changes significantly,
and the residual stress at the center tends to be stable. In the circumferential direction of the pipe,
the residual stress around the weld zone of the outer pipe changes significantly. The residual
stress of the welded laminates has the highest value on the two sides, where the weld and the
welded plate intersect. The maximum residual compressive stress appears at the edge of the
welding plate. The maximum circumferential residual tensile stress of the inner tube is located at
the center of the weld. With the increase of the distance from the center of the weld, the value of
the external residual stress rapidly decays.
(2) The blind hole method was used to test the residual stress of the clamped structure to verify
the correctness of the model. The finite element simulation analysis was performed on the
Appl. Sci. 2020, 10, 2838 15 of 16

multilayered pressure vessel with different welding parameters. It can be seen that the maximum
residual stress on the surface of the welding laminate shows a decreasing trend with the increase
of welding energy, laminate thickness, weld width, welding speed, and vessel radius. Based on
the overall multilayer clamped high-pressure vessel standard HG3129-1998 as a parameter,
a three-dimensional first-order equation of the maximum residual stress on the welding speed,
the thickness of the laminate, and the width of the weld seam is established according to numerical
simulation results and optimized The optimal welding parameters under the actual working
conditions were obtained, which provided a reference basis for the safe use and optimal design of
the welding hoop structure.

Author Contributions: Conceptualization, W.M. and H.Z.; methodology, W.M.; software, Z.H.; validation, W.M.,
H.Z., and W.Z.; investigation, W.Z.; resources, F.X.; data curation, H.Z.; writing—original draft preparation, H.Z.;
writing—review and editing, W.M.; visualization, W.Z.; supervision, W.Z.; project administration, C.Y.; funding
acquisition, C.Y. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Natural Science Foundation of China (2019JJ50625), the Key
Research and Development Plan of Hunan Province (2017WK2032), the Special project of innovative province
construction in Xiangtan City (CG-TPZJ20191028), the Innovative Venture Technology Investment Project of
Hunan Province (2018GK5028), the Key R&D Program of Hunan Province(2018WK2111), the S&T Innovation
Platform and Talent Plan of Hunan Province (2017XK2048), and the Major Program of Science and Technology of
the Hunan Province (2017SK1010).
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Zhu, X.; Wu, J.; Chen, G. Development on integrated multilayer clamping high pressure vessel and its
application. Petro-Chem. Equip. 1999, 4, 44–47.
2. Liu, D.; Song, P.; Hu, M. The prospect of application of integrated multilayer-wrapped high pressure vessels
for hydrogenation reactor. Machinery 2010, 37, 1–3.
3. Song, P.Y.; Gao, J.; Ma, F.B. Prospects Analysis of Integrated-Wrapped Multi-Layer Vessels Used as High
Pressure Gaseous Hydrogen Storage Vessels. Adv. Mater. Res. 2012, 588–589, 1755–1759. [CrossRef]
4. Hashemi, J.; Rasty, J.; Li, S. Integral hydro-bulge forming of single and multi-layered spherical pressure
vessels. J. Press. Vessel Technol. 1993, 115, 249–255. [CrossRef]
5. Zheng, J. Design philosophy of flat ribbon wound layered pressure vessel. Int. J. Press. Vessel. Pip. 1998, 75,
57–61. [CrossRef]
6. Jojima, T. Urea reactor failure. Ammon. Plant Saf. Relat. Facil. 1979, 21, 111–119.
7. Wang, W.; Li, A.; Zhu, Y. The explosion reason analysis of urea reactor of Pingyin. Eng. Fail. Anal. 2009, 16,
972–986. [CrossRef]
8. Patterson, L.B.; Clinton, I.A. Ammonia separator accident. Ammon. Plant Saf. Relat. Facil. 1979, 21, 95–99.
9. Deliang, R.; Yuncheng, W.; Kai, C. Cracking analysis of multilayer warping cylinder for the region with weld
defects. Chin. J. Mech. Eng. 2000, 36, 109–112.
10. Ghosh, S.; Rana, V.P.S.; Kain, V. Role of residual stresses induced by industrial fabrication on stress corrosion
cracking susceptibility of austenitic stainless steel. Mater. Des. 2011, 32, 3823–3831. [CrossRef]
11. Sharifi, H.; Raisi, S.; Tayebi, M. The effect of stress relieving treatment on mechanical properties and
microstructure of different welding areas of A517 steel. Mater. Res. Express 2017, 4, 126508. [CrossRef]
12. Withers, J.P. Residual stress and its role in failure. Rep. Prog. Phys. 2007, 70, 2211–2264. [CrossRef]
13. Tan, L.; Zhang, J.; Zhuang, D. Influences of lumped passes on welding residual stress of a thick-walled
nuclear rotor steel pipe by multipass narrow gap welding. Nucl. Eng. Des. 2014, 273, 47–57. [CrossRef]
14. Attarha, M.J.; Sattari-Far, I. Study on welding temperature distribution in thin welded plates through
experimental measurements and finite element simulation. J. Mater. Process. Technol. 2011, 211, 688–694.
[CrossRef]
15. Mirzaee-Sisan, A.; Wu, G. Residual stress in pipeline girth welds-A review of recent data and modelling.
Int. J. Press. Vessel. Pip. 2019, 169, 142–152. [CrossRef]
Appl. Sci. 2020, 10, 2838 16 of 16

16. Deng, D.; Murakawa, H. Numerical simulation of temperature field and residual stress in multi-pass welds
in stainless steel pipe and comparison with experimental measurements. Comput. Mater. Sci. 2006, 37,
269–277. [CrossRef]
17. Deng, D.; Kiyoshima, S. FEM prediction of welding residual stresses in a SUS304 girth-welded pipe with
emphasis on stress distribution near weld start/end location. Comput. Mater. Sci. 2010, 50, 612–621. [CrossRef]
18. Luo, Y.; Jiang, W.; Wan, Y. Effect of helix angle on residual stress in the spiral welded oil pipelines:
Experimental and finite element modeling. Int. J. Press. Vessel. Pip. 2018, 168, 233–245. [CrossRef]
19. Manurung, Y.H.P.; Lidam, R.N.; Rahim, M.R. Welding distortion analysis of multipass joint combination
with different sequences using 3D FEM and experiment. Int. J. Press. Vessel. Pip. 2013, 111, 89–98. [CrossRef]
20. Liu, C.; Luo, Y.; Yang, M. Three-dimensional finite element simulation of welding residual stress in RPV with
two J-groove welds. Weld. World 2017, 61, 151–160. [CrossRef]
21. Xie, X.; Jiang, W.; Luo, Y. A model to predict the relaxation of weld residual stress by cyclic load:
Experimental and finite element modeling. Int. J. Fatigue 2017, 95, 293–301. [CrossRef]
22. Kik, T.; Jaromír, M.; Iva, N. Application of Numerical Simulations on 10GN2MFA Steel Multilayer Welding.
In Dynamical Systems Theory and Applications; Springer: Berlin/Heidelberg, Germany, 2017; pp. 193–204.
23. Sajek, A. Application of FEM simulation method in area of the dynamics of cooling AHSS steel with a
complex hybrid welding process. Weld. World 2019, 63, 1065–1073. [CrossRef]
24. Qin, Q.; Zhang, D.; Zang, Y. A simulation study on the multi-pass rolling bond of 316L/Q345R stainless clad
plate. Adv. Mech. Eng. 2015, 7, 7. [CrossRef]
25. You, L.; Yan, K.; Liu, N. Assessing artificial neural network performance for predicting interlayer conditions
and layer modulus of multi-layered flexible pavement. Front. Struct. Civ. Eng. 2020. [CrossRef]
26. Xu, S.; Wang, W.; Li, M. A modified weld structure of layered urea reactor based on stress analysis and leak
detection. J. Press. Vessel Technol. 2012, 134, 031203. [CrossRef]
27. Xu, S.; Wang, W.; Chang, Y. Using FEM to predict residual stresses in girth welding joint of layered cylindrical
vessels. Int. J. Press. Vessel. Pip. 2014, 119, 1–7. [CrossRef]
28. Xu, S.; Wei, R.; Wang, W. Residual stresses in the welding joint of the nozzle-to-head area of a layered
high-pressure hydrogen storage tank. Int. J. Hydrog. Energy 2014, 39, 11061–11070. [CrossRef]
29. Xu, S.; Wang, C.; Zhao, Y. Residual Stress in the Welding Joint of Layered Cylindrical Vessels Including the
Weld Clad Effect. J. Press. Vessel Technol. 2015, 137, 041405. [CrossRef]
30. Ma, W.B. Residual Stress Research of the Laminates of Integral Multilayer-Wrapped Pressure Vessel.
Ph.D. Thesis, Xiangtan University, Xiangtan, China, 2010.
31. Yuan, G.L.; Shua, Q.J.; Huang, Z.H. An experimental investigation of properties of Q345 steel pipe at elevated
temperatures. J. Constr. Steel Res. 2016, 118, 41–48. [CrossRef]
32. Jin, W.; Jia, L.I.; Shaoxiong, L. Numerical simulation analysis of welding residual stresses in spherical tank.
Trans. China Weld. Inst. 2016, 37, 89–94.
33. Chi, L.X.; Ma, Y.L. Test verifies and simulation on welding temperature field of steel plate for pressure vessel.
Trans. China Weld. Inst. 2012, 5, 62–65.
34. Teixeira, P.R.D.F.; de Araújo, D.B.; da Cunha, L.A.B. Study of the Gaussian distribution heat source model
applied to numerical thermal simulations of TIG welding processes. Ciênc. Eng. 2014, 23, 115–122.
35. Wang, P.; Xie, P.; Zhao, H.Y. Fundamental research of welded plastic strain evolution process II:
Evolution process of welded plast. Trans. China Weld. Inst. 2014, 35, 72–74.
36. Ma, W.B.; Chen, S.G.; Liu, H.Q. FEM simulation of calibration on strain release coefficients in blind hole
method. Trans. China Weld. Inst. 2011, 32, 97–100.
37. Chiumenti, M.; Cervera, M.; Salmi, A. Finite element modeling of multi-pass welding and shaped metal
deposition processes. Comput. Methods Appl. Mech. Eng. 2010, 199, 2343–2359. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).