You are on page 1of 8

JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN

MECHANICAL ENGINEERING
THERMO STRUCTURAL ANALYSIS AND
OPTIMIZATION OF REACTANT
COLUMN OF PRESSURE VESSEL BY USING FINITE
ELEMENT ANALYSIS
1
Mr. RAHUL KULKARNI,
2
PROF .PRADEEPKUMAR PATIL
1
M.E. (Des!" #$ Me%&'"%'( S)s*e+,
-
Ass*. P.#$ess#., De/'.*+e"* #$ Me%&'"%'( E"!"ee."!,
S0'+ V1'2'"'"3 C#((e!e #$ E"!"ee."!,
V1e2'"'"3 K"#0(e3!e C*), K&'"30' R#'3, I"3#.e, M'3&)' P.'3es&, I"3'
1
rrk.dypsoe@gm!".#om ,
2
prdeepp$!"@s%#e!&dore.#om
A'(TRA)T: This paper describes a project work of thermo structural analysis and optimizing of reactant
column of pressure vessel by using finite element analysis. A reactant column is typically a vertical duct, from
which reactant at high pressure is pushed from top to bottom. At the bottom of the column there are holes
allowing the reactant to flow out, and mix with the other reactant and chemical process is underway. The
advantage of such a design is that the rate of reaction is directly controlled by the pressure in the column, as
this indirectly controls the mass flow rate. o, as more the number of holes more will be efficiency of the
process. The design problem is that, typically the holes are the weak regions structurally and in addition they
also experience high thermal zones, making them more vulnerable.
The paper focuses on:
!. "ptimizing the location of these holes for maximum structural safety,
#. To analyze reactant column and determine their strength,
$. To optimize the design in such a manner that, all failures can be avoided.
Key*ords+ No& ,L!&er, Re#$&$ )o"-m&, T.ermo / ($r-#$-r" FEA, ($$!# ($r-#$-re A&"ys!s.
1 . INTRODUCTION4
A pressure vessel is a closed container designed to
hold gases or liquids at a pressure substantially
different from the ambient pressure. The pressure
differential is dangerous and many fatal accidents
have occurred in the history of pressure vessel
development and operation. Consequently, pressure
vessel design, manufacture, and operation are
regulated by engineering authorities backed by
legislation. In general, Pressure vessels are design
with A!" #oiler and Pressure $essel Code,
ection $III., %ivision & and do not require a detailed
evaluation of all stresses. They are used in a wide
variety of industries 'e.g., petroleum refining,
chemical, power, pulp and paper, food, etc.(
")amples of pressure systems and equipment are
boilers and steam heating systems* pressuri+ed
process plant and piping* compressed air systems
'fi)ed and portable(*pressure cookers, autoclaves and
retorts* heat e)changers and refrigeration plant*
valves, steam traps and filters* pipe work and hoses*
and Pressure gauges and level indicators.
-. MAIN COMPONENTS OF PRESSURE
VESSELS4
-.1 S&e((4
The shell is the primary component that contains the
pressure. Pressure vessel shells are welded together
to form a structure that has a common rotational a)is.
!ost pressure vessel shells are cylindrical, spherical,
or conical in shape.
-.-. S2.* S5//#.*s4
Tall, vertical, cylindrical pressure vessels 'e.g., the
tower and reactor are typically supported by skirts. A
support skirt is a cylindrical shell section that is
welded either to the lower portion of the vessel shell
or to the bottom head 'for cylindrical vessels(. The
skirt is normally long enough to provide enough
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e ??
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
MECHANICAL ENGINEERING
fle)ibility so that radial thermal e)pansion of the
shell does not cause high thermal stresses at its
,unction with the skirt.
-.?. He'34
All pressure vessel shells must be closed at the ends
by heads. -ear heads are typically used flat head.
-.A N#BB(e4
A no++le is a cylindrical component that penetrates
the shell or heads of a pressure vessel.. .o++les are
used for the following applications/
Attach piping for flow into or out of the vessel. Also
provide for direct attachment of other equipment
items, 'e.g., a heat e)changer or mi)er(.
-.9 S5//#.*4
The type of support that is used depends primarily on
the si+e and orientation of the pressure vessel. In all
cases, the pressure vessel support must be adequate
for the applied weight, wind, and earthquake loads.
The design pressure of the vessel is not a
consideration in the design of the support since the
support is not pressuri+ed. Temperature may be a
consideration in support design from the standpoint
of material selection and provision for differential
thermal e)pansion.
F!5.e 14 M'" C#+/#"e"*s #$ P.ess5.e Vesse(s
?. BRIEF OVERVIEW OF SOME RESEARCH4
As there are lot of work is done on the pressure
vessels. A brief review of some selected references.
?.1 BY E. 6. BERGMAN, ALHAMBRA, CALIF4
have suggested that, ")ternal loads applied to vertical
pressure vessels produce a)ial loading and bending
moments on the vessel. These result in a)ial tensions
and compressions in the shell, which must be
combined with the effects of the pressure loading to
give the total longitudinal stress acting in the shell.
The design method to be used depends on whether
the longitudinal tress in the shell is tension or
compression, and on whether the vessel is sub,ected
to internal or e)ternal pressure.
?.- PEDRO MARCAL, DENNIS H. MARTENS,
MICHAEL A.P4 This paper represents a practical
review of the use of PC based 0inite "lement
software in the analysis of typical pressure vessel
components. The authors discuss element type
selection criteria and features. ome of the different
element formulations are discussed. !odeling
parameters and convergence procedures are
e)amined. Practical evaluation tolerances are
discussed.
?.? DAVID HECKMAN has suggested that, 0inite
element analysis is an e)tremely powerful tool for
pressure vessel analysis when used correctly. Tested
models were run with errors ranging from seven to
nearly +ero percent error and could be run in a
relatively short time. -owever, even with such results
the operator still is required to be knowledgeable of
not ,ust how to run the finite element analysis, but
also how to read the results. %ata must be verified
with hand calculations to confirm that solutions are
relatively accurate. 1here results are questionable,
such as in the final contact element model, one must
understand ,ust what the finite element model is
modeling and how well this appro)imates the actual
sub,ect.
A. FINITE ELEMENT ANALYSIS4
1hen the 0"! is applied to a specific field of
analysis like stress analysis, thermal analysis, or
vibration Analysis, it is often referred to as finite
element analysis '0"A(. 0"A is the most common
tool for stress and structural analysis. $arious fields
of study are often related. 0or e)ample, distributions
of nonuniform temperatures induce nonobvious
loading conditions on solid structural members. Thus,
it is common to conduct a thermal 0"A to obtain
temperature results that in turn become input data for
a stress 0"A. 0"A can also receive input data from
other tools like motion 'kinetics( analysis systems
and computation fluid dynamic 'C0%( systems.
9. NEED OF ANALYSIS4
In industry the component produced may be of
different si+es, from flat plates of very simple shape
to comple) 2 dimensional solid bodies. %uring
operation they may be sub,ected to various types of
applied loading conditions which include centrifugal
force, pressure and temperature loading and
prescribed boundary conditions. 1ith rising cost of
material over design, the resultant wastage may be
e)tremely costly. 0ailure of component during
service may produce a high service return rate, which
is e)tremely undesirable both from high replacement
cost and damage to the prestige of product. o stress
analysis at the design stage is essential, if service
failures are to be avoided and near optimum designs
are to be achieved for specified.
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e ?A
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
MECHANICAL ENGINEERING
;. SOLUTION PROCEDURE FOR NONLINEAR
ANALYSES4
A.3 employs the 4.ewton5 6aphson4 approach
to solve nonlinear problems. In this approach, the
load is subdividedinto a series of load increments.The
load increments can be applied over several load
steps. 0igure 4.ewton5 6aphson Approach4
illustrates the use of .ewton56aphson equilibrium
iterations in a single %70 nonlinear analysis.
F!5.e A4 ?@D M#3'( #$ Re'%*'"* C#(5+"
T'C(e I4 M#3'( S/e%$%'*#"
F!5.e -4 Ne0*#"@R'/&s#" A//.#'%&
#efore each solution, the .ewton56aphson method
evaluates the out off5 balance load vector, which is
the ifference between the restoring forces 'the loads
corresponding to the element stresses( and the
applied loads. The program then performs a linear
solution, using the out5of balance loads, and checks
for convergence. If convergence criteria are not
satisfied, the out5of5balance load vector is
reevaluated, the stiffness matri) is updated, and a
new solution is obtained. This iterative procedure
continues until the problem converges.
8. DETAILS OF REACTANT COLUMN
ASSEMBLY4
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e ?9
C#+/#"e"*
N'+e
D'+e*e.
" ++
T&%2"ess
" ++
He!&*
" ++
kirt upport 8&9: ;9 &9::
#ase Plate 8&9: &9: &9:
hell 8&9: ;9 99::
Inner
6eactant
<=:: 9: 99::
Cylindrical
-ead
8&9: &9: &9:
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
MECHANICAL ENGINEERING

F!5.e ?4 Re'%*'"* C#(5+"
<. MODELLING AND ANALYSIS4
25% model of reactant column assembly is created in
A.3 1orkbench and CATIA, which is required
for the purpose of further analysis. A.3
1orkbench provides a highly integrated engineering
simulation platform, supports multi physics
engineering solutions and provides bi5directional
parametric associability with most available CA%
systems.
<.1 Me%&'"%'( Se(e%*#" '"3 P.#/e.*es4 -igh
quality carbon steel is used for boiler and pressure
vessel fabrication which is ideally suited to the high
standards set by the oil, gas and petrochemical
industry. 1e select A9&8 >rade ;: material applied
to the body contains/
T'C(e II4 Me%&'"%'( P.#/e.*) #$ M'*e.'( SA91;
G.'3e 86
3ield trength <?: !pa
@ltimate strength ?8: !pa
Thermal Conductivity 8:.9 1Am
:
B
PoissonCs 6atio :.2
<.- C&e+%'( C#+/#s*#"4
T'C(e III4 C&e+%'( C#+/#s*#" #$ M'*e.'(
SA91; G.'3e 86
Carbon 'C( :.2:D
!anganese'!n( :.=:5&.2D
Phosphorous'p( :.:29D
ulphar's( :.:29D
ilicon'si( :.&25:.?9D
<.? Mes&"!4
The 6eactant column assembly model is meshed with
<: node he)ahedron 7EI%&=8 element. It is a
higher order 25% <:5node solid element that e)hibits
quadratic displacement behavior. The element is
defined by <: nodes having three degrees of freedom
per node/ translations in the nodal ), y, and +
directions.

F!5.e 94 F"e HeD'&e3.#" Mes&e3 M#3e( #$
Re'%*'"* C#(5+" Asse+C()
<.A De$""! *&e C#"*'%*s4
-ighly nonlinear problems Contacts are necessary.
o it requires proper attention to define the correct
contact pairs and target and contact surfaces. 1hen
two surfaces are in contact one is target surface and
another is contact. Contact surface is one from which
forces get transfer to target surface. Target surface is
that to which forces get transfer. 6eactant column
assembly consists of different components like top
plate, inner reactant column, shell, bottom plate, skirt
support. These components are connected to each
other have number of contacts pairs. All contacts
defined are of bonded type.
F!5.e ;4 B#"3e3 C#"*'%* Ce*0ee" I""e. Re'%*'"*
C#(5+" '"3 B'se P('*e
<.9 B#5"3'.) C#"3*#"s4
#oundary Conditions are as follows/
&. #ottom there is skirt support which is considered
as 0i)ed.
<. 1eight of cobalt meshing &?:: kg applies on the
base plate. This is equally divided into ? parts that is
29: kg on equally each divided faces.
2. Internal pressure :.< !PA apply on top F bottom
plate also on shell faces.
?. There are pressure differences :.:G !PA during
flow of pressure raised fluid that is apply on internal
faces on inner reactant column.
9. tandard earth gravity G.=& mAs
<
apply at centre of
pressure vessel.
8. Process Temperature 9::k.
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e ?;
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
MECHANICAL ENGINEERING
F!5.e 84 Re'%*'"* C#(5+" 0*& B#5"3'.)
C#"3*#"
7. RESULT AND DISCUSSION4
After giving the solution, the problem is being solved
in analysis software A.3 by giving the number of
sub steps for nonlinearity. The results of stress and
deformation are form in graphs for all boundary
conditions. The stress and the deformation color
counters plots for considering above all boundary
conditions by changing mesh si+e and sub steps.
!a)imum von5mises stress and deformation color
counters plotted as shown in below/
F!5.e <4 V#"@Mses S*.ess P(#* #" S2.* S5//#.*
!a)imum "quivalent tress H <.;9??e= Pa
!a)imum %eformation H &.?=;;e9Pa
The ma)imum stress value shown by the red color
plot is acting at very small portion of skirt support. In
actual practice this type of stresses are due to the
stress concentration and can be neglected. And they
are less than allowable stress of that material.
T#*'( N5+Ce. #$ H#(es #" I""e. Re'%*'"* C#(5+"
'.e E 8-
F!5.e 74 V#"@Mses S*.ess P(#* #" I""e. Re'%*'"*
C#(5+" &'1"! Re%*'"!5('. P'**e."
1e are concentrated on inner reactant column of
pressure vessel. 0rom this analysis we can say that
comparatively very low stress introduce on inner
reactant column. !a)imum vonmises stress is found
on top portion inner reactant column as shown as red
color contour plot above figure G. This is also less
than that of the yield point of that material.
T'C(e IV4 A"'()ss Res5(* #$ Re%*'"!5('. P'**e."
F!5.e 164 T#*'( De$#.+'*#" P(#* #" I""e.
Re'%*'"* C#(5+" &'1"! Re%*'"!5('. P'**e."
16. NEW TRIANGULAR HOLE PATTERNS4
As we know rate of reaction is directly controlled by
the pressure in the column, as this indirectly control
the mass flow rate. o we choose different holes
pattern such a way that increasing the mass flow rate
and also efficiency of process.
F!5.e 114 T.'"!5('. P'**e." H#(es
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e ?8
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
MECHANICAL ENGINEERING
.ow we select the triangular pattern having the
ma)imum number of holes keeping the all
dimensions are same as in previous case.
T#*'( N5+Ce. #$ H#(es #" I""e. Re'%*'"* C#(5+"
'.e E 19;
F!5.e 1-4 Re'%*'"* C#(5+" 0*& T.'"!5('.
H#(es P'**e.".
This new triangular pattern is analy+ed for the same
operating conditions as in previous case and found
out the results.
F!5.e 1?4 Re'%*'"* C#(5+" 0*& T.'"!5('. H#(e
P'**e."s 0*& B#5"3'.) C#"3*#"s
F!5.e 1A4 V#"@Mses S*.ess P(#* #" I""e.
Re'%*'"* C#(5+" 0*& Ne0 T.'"!5('. H#(e
P'**e."s
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e ?<
A"'(
)
N#.
Mes&
SBe
N#. #$
N#3es
N#. #$
E(e+e"*s
S*.ess "
MP'
T#*'(
De$.
"
++
& =9 2=G;88 &:<2?G 9G.&?8 &;.<;
< G: 29&G?2 G9G:& 9=.:99 &;.<G
2 G9 2&2?G9 =<;:9 9=.9=9 &;.2&
?
&:: <=8<9? ;9:G& 98.<<2 &;.<2
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
MECHANICAL ENGINEERING
F!5.e 194 T#*'( De$#.+'*#" P(#* #" I""e.
Re'%*'"* C#(5+" 0*& Ne0 T.'"!5('. H#(e
P'**e."s
T'C(e V4 A"'()ss Res5(* #$ T.'"!5('. P'**e."
A"'(
)
N#.
Mes&
SBe
N#. #$
N#3es
N#. #$
E(e+e"*s
S*.ess
"
MP'
T#*'(
De$. " ++
&
&<:I?
:
??<&=
G
&&922< &88.= &;.288
<
&2:I?
:
?<?&G
2
&&?28:
&82.;
;
&;.2;;
2
&?:I?
:
?:2::
;
&:=&92
&88.9
9
&;.?
? &9:
2G&&8
2
&:9<:2
&88.;
<
&;.?&<
9 &=:
29?<9
=
G&9:2
&88.8
;
&;.2=
11. COMPARISON OF RECTANGULAR AND
NEW TRANGULAR HOLE PATTERN
MODEL4
Comparison of both the models from the above stress
plots is decrypted in the table. 0rom the above
analysis some points are highlighted as below/
Increase in number of holes increased the flow rate
capacity of the process, as in the new triangular
pattern model compare to basic model.
T'C(e VI4 C#+/'.s#" #$ H#(e P'**e."
Percentage increase in flow rate capacity is*
1-. CONCLUSION4
0rom the above problem calculation and solution it is
observed that, 0inite element analysis is an e)tremely
powerful tool for analysis of reactant column
assembly. Tested models are run and reduce the
effort in a relatively short time. -owever, the finite
element model produces results which are
consistently points towards a useful design changes.
The analysis of the proposed reactant column
assembly brought into light a number of inadequacies
in design.
The first thing that was observed that, the new holes
pattern is able to reduce the stresses in the assembly,
as the stresses became well distributed over the inner
reactant column assembly due to more symmetrical
distribution of reactant column assembly.
A new hole pattern was selected, which, allowed =?
more number of holes which lead to increase flow
rate capacity 92.=?D .Also, use of triangular pattern
is more suitable for mass flow rate.
!a)imum stress occurs on skirt support and
minimum on the inner reactant column. All the
component of reactant column sustains the moderate
stresses which are lower than allowable stresses of
the material and the deformation is within the
permissible limit.
1?. FUTURE SCOPE4
Currently 6ectangle Pattern and Triangular Patten is
optimi+ed and results are taking with the help of
analysis software.
The conclusion of study is that, Triangular Pattern
with more number of holes are given good results
compare to rectangular pattern and from that, it is
observed that , more the number of holes more will
be efficiency of the process.
The future scope of the study is that,
&. 7ptimi+e the reactant column for different
pattern of hole and for different material
properties.
<. 0or thickness parameter, optimi+e all $essel
Components.
2. 7ptimi+e the pressure vessel for lowest
weight and ma)imum efficiency.
?. 7ptimi+e the different component of
pressure vessel by considering
environmental conditions such as humidity,
corrosion, wind, acid rainfall etc.
REFERENCES4
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e ?7
S..
N#.
P'**e."s
Res5(*s #$ Re'%*'"* C#(5+"
N5+Ce.
#$ H#(es
M'D. V#"
Mses S*.ess
" MP'
M'D.
De$. "
++
&.
6ectangular
-oles Pattern
;< 9G.&?8 &;.<;
<.
Triangular
-ole Patterns
&98 &88.= &;.288
JOURNAL OF INFORMATION, KNOWLEDGE AND RESEARCH IN
MECHANICAL ENGINEERING
J&K The %esign 7f $ertical Pressure $essels
ub,ected To Applied 0orces #y ". :. #ergman,&
Alhambra, Calif
J<K 7n @sing 0inite "lement Analysis 0or Pressure
$essel %esign !ichael A.Porter ,%ennis -.
!artens ,Pedro !arcal
J2K 0inite "lement Analysis 7f Pressure $essels
%avid -eckman, @niversity7f California, %avis
J?K Predicting The Eife 7f kirt upports In
Pressuri+ed 6eactors @nder Cyclic Conditions #y
!arcos ugaya Petrobras .A., Cenpes
J9K L0inite "lement !odeling 0or tress AnalysisM N
by 6obert %. Cook@niversity of 1isconsin N
!adison
J8K L0undamentals of 0inite "lement AnalysisM N by
%avid $. -utton.
J;K LIntroduction to the 0inite "lement !ethodM N by
"vgeny #arkanov, Institute of !aterials and
tructures,6iga Technical @niversity.
J=K L0inite "lement AnalysisM 5 by %avid 6oylance,
%epartment of !aterials cience and "ngineering,
!assachusetts Institute of Technology Cambridge.
JGK LAnalysis of a !ultilevel Iterative !ethod for
.onlinear finite "lement "quationsM 6andolph ".
#ank and %onald O. 6ose, .aval 6esearch.
J&:K L0ailure of !aterials in !echanical %esignM N
by O. A. Collins, 7hio tate @niversity.
J&&K L0inite "lement Analysis Theory and PracticeM N
by !. O. 0agan, @niversity of -ull.
ISSN 6789 : ;;<=> NOV 1- TO OCT 1? > VOLUME : 6-, ISSUE @ 6- P'!e A6