ducting optimization – a case study
TRANSCRIPT
![Page 1: Ducting Optimization – A Case Study](https://reader031.vdocuments.site/reader031/viewer/2022012319/589d75d41a28abbd498b94b8/html5/thumbnails/1.jpg)
International Journal of Applied Research In Mechanical Engineering (IJARME), ISSN: 2231 –5950, Volume-1, Issue-2, 2011
31
Ducting Optimization – A Case Study
C. Hemalatha1, T. Radhakrishnan2 & K. Sankaranarayanasamy3 1&2 Bharat Heavy Electricals Limited, Tiruchirappalli, India
3National Institute of Technology, Tiruchirappalli, India E-mail: [email protected], [email protected]
Abstract - Ducting system in a power plant is a conduit connecting one system to another through which either cold air or hot air or flue gas with ash is conveyed for continuous operation of the plant. To perform this function properly, careful design of ducting is required to take care of the stresses and forces exerted by various loadings and thermal expansion of the ducting and equipment it connects under given pressure and temperature. Ducting can be of any configuration viz. square, rectangular, circular, straight or bend, elbow or any other irregular shape. The design of large air and gas ducts is an extremely complex task which should meet both mechanical and structural design criteria. The purpose of ducts is to efficiently convey the air or flue gas from one location to another while maintaining the pressure drop and temperature drop as low as possible. This paper discusses about the optimization in the design engineering of the duct system.
Keywords - CFBC, Primary Air, Secondary Air, Duct, Stiffeners, aspect ratio, optimum, weight, cost
I. INTRODUCTION
Engineering Design optimization both shortens design cycle time and finds new designs that are not only feasible, but also optimal based on the design criteria. Traditional engineering design processes involve strategies such as trial and error, use of previous experience etc., until the requirements are either met or changed to fit the performance. Often, the process is time consuming and does not produce the best design but just a feasible one.
Since the process of design and engineering of ducting is time consuming, the process is automated by developing software. The software generates complete set of manufacturing and erection drawings for Circulating Fluidised Bed Combustion Boiler [1]. For the development of software the duct design is required to be parameterised and optimised.
Engineering Design Optimization can both reduce the cycle time for the design iterations and find the best (optimal) design for the specifications. This process differs from the traditional process in that the iteration loop is computerized. An optimization problem is posed for which the design variable, the design objective and all constraints are specified. The analysis code performs the test phase of the iteration loop. The optimizer may function by perturbing each design variable to determine how each affect the performance and then seek a solution that optimizes the objective. While considering the design engineering of ducting system, the analysis
can be done in various areas like fixing duct configuration or sizing, selection of material thickness, stiffener sizing, selection of expansion joint, maximum size for shipping etc.,
Among these, this paper deals with duct configuration, material thickness and stiffener spacing and sizing. Since, the total weight of the ducting and hence the cost for the customer mainly depends on the above parameters, they are taken up for optimization.
II. DUCT DETAILS
Any duct assembly is a combination of straight duct, transition duct, bend duct, metallic expansion joints or non metallic expansion joints, supports, damper, gates and man hole doors [2]. Straight duct can be of square, rectangle or circular cross section with adequate stiffeners on outside. For air duct and gas duct, carbon steel is used as duct material. Normally, square corner expansion joints are provided in the ducting assembly and while carbon steel is used as expansion joint material for air ducts, corten steel (carbon steel with minimum percentage of copper) is used for gas ducts. Wherever removal of duct is required for maintenance purpose flanges are provided with packing rope and in all other places ducts are welded permanently with each other and the gas tight joint is ensured. Manhole doors are provided in the ducting assembly preferably in a straight duct portion with appropriate access from the nearby floors. Care must be
![Page 2: Ducting Optimization – A Case Study](https://reader031.vdocuments.site/reader031/viewer/2022012319/589d75d41a28abbd498b94b8/html5/thumbnails/2.jpg)
International
taken that thelocation of duc
Ducts are or in combinaare type of boin fixed suppobottom supposupport memband allowed accommodate thermal expanwith the bottosupporting strueither with hconstant load supports to reduct [3].
III. CFBC DU
Ducts connot necessarilymay be circulthey may be upiece of majoheater or eletemperature cstructures whsuspended by
In a fluidisystem consisfan and Force(SA)) fan thatthe boiler. Thiboiler, the outpre heater wprecipitator oparticles are reflue gas to atm(ID) fans aredownstream ochimney. Theythrough the pthrough the ch
Fig.1: Sch
l Journal of Appl
ese doors arect stiffeners.
supported eithation. Fixed sttom supports.orts, the supprting member
ber is placed oto slide over
the movemensions. The slidm support platuctural membe
hanger rod ashangers. Res
estrict the mov
UCTING
nnect pieces oy considered plar or rectanguunusual in shaor power planctrostatic precchanges, ductshich “float” oa hanger system
ized bed boilerts of a combin
ed Draft (FD) t supply the aiis duct extendstlet flue gas uswhere it is r bag filter wemoved from thmosphere throue located withof the boiler by suck the fluprecipitator an
himney or stack
hematic of Duc
lied Research In
e provided co
her from bottomsupport and sli The basic diffort is directlyand in slidin
over the slide bthe slide bea
ents of the dde bearing platte and then we
er. The top supsembly or wi
straints are provement of the
of major equippart of the equiular in shape ape as they trant equipment cipitator. To s are usually
on slide bearim.
r system, (Fig.nation of Prim(also called S
ir needed for cs up to the air sually passes tcooled; an
where most ohe flue gas; theugh chimney. Ihin the duct setween the pr
ue gas out of tnd out to thek.
ct arrangement
Ducting Optim
n Mechanical En
onsidering the
m or from top iding supports fference is that y welded with ng support the bearing plates aring plate to ducts due to tes are welded elded with the
pport can be of ith variable / ovided in the e support and
pment, but are ipment. Ducts or sometimes
ansition into a like fan, air
accommodate y independent ing plates or
.1) air ducting mary Air (PA) Secondary Air combustion to pre heater. In
through an air electro static f the fly ash e ID fans vent Induced Draft system in the recipitator and the boiler and e atmosphere
of CFBC
mization – A Case S
ngineering (IJAR32
f
In CtemperatCFBC d
Fig
IV. CO
Opteffectivemost oroptimizatoggle componerequiremknowledknowledfeedbackstiffeneranalyseddesign pconsidervarious weight oand manthe calcuoptimumstiffener
V. DU
Whtypes of locationconsiderencounteduct andsituationsquare limitatioupon thethe sizindiscusse
Study
RME), ISSN: 223
CFBC boiler ture and pressu
ducting system
g. 2: Temperat
NCEPT OF O
timization is toe as possible wr use the beation, differendetails, varioents, stiffen
ments are codge and site wodge and acck, the use of r sizing and spd. As per the opparameters likred for any spduct configuraof the duct winufacturing cosulated values,
m value for drs, stiffener size
CT DESIGN
ile doing ducf ducts used dep. Circular ductring performanered during thd difficulty in pns the circuladuct is prefe
on of space rece size of the dung and spacinges about param
31 –5950, Volum
different zonure. The tempeis plotted as sh
ture profile of d
OPTIMIZATI
o make the sywith available est of the synt duct typesous support tning elementompared withorking experiencording to tvarious duct spacing with duperating conditke temperaturepecific systemation, stiffener th stiffeners anst are studied. G
the analysis iduct sheet thie, weight and c
OPTIMIZAT
ct engineeringpending on flowt is considered nce. But becaushe manufacturproviding supp
ar duct is avoerred. In placctangular duct uct and thickneg of the stiffenemeterizing and
me-1, Issue-2, 20
nes are at diferature profile hown in Fig.2.
ducting system
ION
ystem as perfefeatures to ge
ystem.[4]. In s, expansion types and suts and splh practical dnce. Based on the manufactsheet thicknesuct sizing casetion of the plane and pressur
m. Details regaspacing and s
nd duct designGraphically plis made to finickness, numbcost.
TION
, there are vaw area, velocitto be the best se of the difficring of the cirports, at most ooided, whereinces where the
is used. Depeess of the duct ers vary. This optimizing th
011
fferent of the
m
ect or et the
duct joint,
upport litting design
these turing s, the es are nt, the re are arding sizing, n cost otting
nd the ber of
arious ty and while
culties rcular of the n the ere is ending
plate, paper
he use
![Page 3: Ducting Optimization – A Case Study](https://reader031.vdocuments.site/reader031/viewer/2022012319/589d75d41a28abbd498b94b8/html5/thumbnails/3.jpg)
Ducting Optimization – A Case Study
International Journal of Applied Research In Mechanical Engineering (IJARME), ISSN: 2231 –5950, Volume-1, Issue-2, 2011
33
of stiffener spacing and sizing with the use of duct size and configuration so as to reduce the total weight of ducting system.
Objective of function:
Duct size and aspect ratio are required to be analysed to have the optimum duct configuration. The effect of thickness of the duct plate on stiffener sizing and spacing are analysed considering the manufacturing feasibility, because increase in duct weight almost due to duct sheet thickness and stiffeners.
Duct sheet thickness lies within the limited range and number of stiffeners required depends on the thickness of the plate. Based on the maximum stiffener spacing calculated, number of stiffeners is found and the stiffener size suitable to stiffen the duct size is selected. Number of stiffeners decides the number of welding which decides the maximum portion of the duct cost. So it is required to reduce the total cost by optimally selecting the duct plate thickness, stiffener size and spacing.
Case study:
To explain the optimization procedure followed in the present analysis, a portion of the ducting in CFBC Boiler, Induced Draft fan outlet duct is taken as a case study.
Parameters considered for optimization:
• Duct sheet thickness
• Duct configuration - aspect ratio
• Number of Stiffeners
• Stiffener size
• Duct material cost
• Duct fabrication cost
While designing the duct panel the loads considered are,
• Positive and negative pressure load from mechanical design data. (PPL & NPL)
• Dead load according to the system (DL)
• Wind load (WL)
• Ash load(AL)
• Live load (LL)
The limiting conditions considered for the present case are
Design temperature: 200 deg C
Positive Pressure Load (PPL):+20 mbar
Negative Pressure Load(NPL): -10 mbar
Dead load (DL) : 100 kg per m2.
Wind load (WL) : 200 kg per m2
Ash load (AL): 300 kg per m2
Live load (LL): 100 kg per m2
By calculating the occurrence of different possible combinations of the loads the numerical maximum is taken as the design load of the respective panel.
A duct shell is considered to consist of at-least four duct panels (two sides, top and bottom). Each of the four panels may be designed independently to achieve the most economical design for a given application. If different stiffener spacing is considered for all the panels then the minimum of the values is considered as the maximum stiffener spacing. Stiffener sizing will vary depending upon the load acting over the top, side and bottom walls of the shell. Always the bottom wall is highly loaded because of the ash load and water wash load. For economic stiffener size, different sizing and spacing shall be maintained for the different side of the duct shell walls.
The stiffener span is calculated using Stress criterion (Eqn.1), Static Plate deflection (Eqn.2), Dynamic plate deflection (Eqn.3) and Vibration criterion (Eqn.4) methods [5].
Stress Criterion:
S=2tP
Ys2
+2” (1)
Static Plate deflection:
S = 44304.29
PEt
+2” (2)
Dynamic plate deflection criterion:
S = 4'2
3
)1()12(384
PEt
γ−Δ
+ 2’’ (3)
Plate vibration criterion:
S = f
Et893 + 2” (4)
where
S - stiffener spacing
![Page 4: Ducting Optimization – A Case Study](https://reader031.vdocuments.site/reader031/viewer/2022012319/589d75d41a28abbd498b94b8/html5/thumbnails/4.jpg)
mspwcS
S
S
S
International
t - plate thi
Ys- yield s
P - design p
E - Young
γ = Poisson
∆= plate de
p´=expecte
f = natural
Among tminimum valuepacing. Then f
width) to be scriterion (EqnStiffener vibrati
Stiffener stress
L ≤8PSMB
Stiffener deflec
L ≤ 3465
384x
Stiffener vibrati
L ≤ 183
Where
M= combin
B = allowa
L = Maxim
I = combin
K = lengthk=2 for mu
A = combi
A table hastiffener for thI sections) for
l Journal of Appl
ickness
strength
pressure
’s modulus
n’s ratio
eflection
ed pressure pul
frequency
he values froe is identifiedfor each stiffestiffened can
n.5), Stiffenerion methods (E
criterion:
3
2SS
MB+
ction criterion:
644
xPSEIK
ion criterion:
fA
EI
4
ned section mo
able bending st
mum allowable
ned moment of
h factor; k=1 fultiple lengths
ined cross secti
as been preparhe give span of
various duct p
lied Research In
lsation
om these fourd as the maximner size, the sbe calculated
r deflection Eqn.7).
odulus of (plate
tress
stiffener lengt
f inertia of (plat
for uniform sti
ional area (plat
red which givef the panel (botplate thickness.
Ducting Optim
n Mechanical En
r criteria, the mum stiffener span (depth or d using stress
(Eqn.6) and
(5)
(6)
(7)
e + stiffener)
th
te +stiffener)
iffener length,
te +stiffener)
es size of the th channel and
mization – A Case S
ngineering (IJAR34
Table 1:
Thickness
in mm
3
4
5
6
8
For differentselected
TakstiffenermaximuApplyinPlate defcriterionobtainedmm. Thstiffenerconsider
a) If un
Stif
Stif
Nu
b ) If n
Max
Num
Fig.3:
Study
RME), ISSN: 223
Selection of c
ISMC75
ISMC100
ISM12
2289 2657 29
2265 2635 29
2220 2600 29
2170 2560 29
2061 2446 28
each individut channel / bfrom the stiffe
king a duct mr spacing and sum transport
g Eqn.1,2,3 aflection, Dynam
n for the sd are 1962mmhe minimum var spacing whred for analysis
niform stiffene
ffeners = lengt
ffener spacing (
mber of stiffen
on-uniform sti
x. Stiffener spa
mber of stiffen
Duct panel w
31 –5950, Volum
channel size fo
Channel siz
MC25
ISMC150
ISMC200
990 3281 3766
975 3275 3770
950 3255 3760
910 3220 3740
812 3134 3670
ual panel of beam size, thener table.
made of 4 mmstiffener sizes length of du
and 4 for Stremic plate deflespecified con
m, 1093.6 mm,alue is found thich is 698 s.
er spacing is co
th/Max_sp=>ro
(S1) = 2500/4
ners (n) = 3
ffener spacing
acing (N1) = 6
N2 = 6
ers (n) = 3
with uniform sti
me-1, Issue-2, 20
or span of duct
ze
C ISMC250
ISMC300
6 4142 4561
0 4155 4575
0 4160 4580
0 4160 4575
0 4129 4542
the duct shelhe stiffener si
m plate for anaare calculated
uct as 2500 ess criterion, ection and Vibrditions the v,1195 mm andto be the maxmm for the
onsidered,
ound(3.5)=4
= 625 mm
is considered,
630 mm
620 mm
iffener spacing
011
size
ISMC400
5217
5235
5250
5255
5243
ll, for ize is
alysis, d for a
mm. Static ration values d 698 imum
case
![Page 5: Ducting Optimization – A Case Study](https://reader031.vdocuments.site/reader031/viewer/2022012319/589d75d41a28abbd498b94b8/html5/thumbnails/5.jpg)
International
Fig.4: Duct pa
Sl. N
o
Thic
knes
s m
m
1 3
2 3.5 3 4 4 5
5 6
6 8
From Taband 6 mm, edifferent, whstiffener spacimm plate thicksheet weight iof 3 mm plaincreased andMoreover weldifficult. For aconditions, bysuitable stiffenstiffener spacitabulated.
aspect ratio a
1.15 2.441
l Journal of Appl
anel with non-u
Max
. Stif
fene
r Spa
cing
(c
alcu
late
d)
mm
553
592.8 630 698
760
870
ble2, it is obseeven the maxhile considerining, number okness, no of stis more compaate thickness,
d this increaselding in 3mmall the above cy changing thners are selecting , total duc
b stiffenea
2.807 ISMC1
lied Research In
uniform stiffen
Uniform stiffenerspacing
spac
ing
(S1)
,
mm
num
ber o
f tif
f(
)
500 4
500 4 625 3 625 3
625 3
833 2
erved that for ximum stiffeneng uniform/ f stiffeners aretiffeners is redared with other
number of es number of
m plate is founcases, under abe aspect ratioed and keepinct weight is c
er_ Stiffener_b
00 ISMC125
Ducting Optim
n Mechanical En
ner spacing
r
stiff
ener
s (n)
spac
ing
mm
N1 553
553 630 698
760
870
ducts of 4, 5 er spacing is
non-uniform e same. For 8 duced but duct rs. In the case stiffeners has welding also.
nd to be very bove specified o of the duct, ng the uniform calculated and
Wt. Panel _a
191.61
mization – A Case S
ngineering (IJAR35
Simsizes wtradition
Tabrequiredas 2500 uniform Table 2:duct plat
Non
Sample1
spac
ing
m
m
N2 420 4
420 4620 3552 3
490 3
380 3
Confor aspefrom taincludescost is cwelding total wemm for s
Table3-1.15
Wt. Panel_b
Tpan
220.349 82
Study
RME), ISSN: 223
milar way the rwere computednally used for f
ble 2 is prepd to stiffen the
mm. The datastiffener spaci
Stiffener requte thickness
Uniform stiffener sp
num
ber o
f st
iffen
ers
spac
ing
m
m
n N1550
550 630 698
760
870
nsidering a samect ratio of 1.1able1. Total ds duct panel wcalculated bascost. Table 3
eight and cost specified aspec
Table showing
Total nel_wt
no ostiffen
s
23.936 3
31 –5950, Volum
requirement ofd for differenfabrication of d
pared for numduct span keepa are found foing.
uirement and sp
pacing
Sample2
mm
spac
ing
m
m
N2 425
425 315 203
110
542 & 273
mple case of 415, the stiffeneduct weight is
weight and stiffed on materialists the samplfor the duct p
ct ratio.
g weight and co
of ner Total
_stiff_wt
360.64
me-1, Issue-2, 20
f stiffeners for nt cases of pducts.
mber of stiffping the duct l
or uniform and
pacing for vari
num
ber o
f st
iffen
ers
n 4 2
4 24 24 2
4 2
3 4 2
4 mm thickneser size is idens calculated wfeners weight. al cost and stifle calculated dplate thickness
ost for aspect r
Total weight tot
1185 16
011
other plates
feners length d non-
ious
Leng
th
m .5
.5
.5
.5
.5
.5
ss and ntified which Total
ffener data of s of 4
ratio
tal cost
65123
![Page 6: Ducting Optimization – A Case Study](https://reader031.vdocuments.site/reader031/viewer/2022012319/589d75d41a28abbd498b94b8/html5/thumbnails/6.jpg)
International
Similarly various duct between 3 mmminimum platmm, for furthVarious comband graphs are
VI. RESULT
Fig.5 relaconfiguration tons, for variosimilar for theaspect ratio ofmm thickness(aspect ratio comparing 3.53.5 mm plate mm, because comparing theaspect ratios, tduct is 1, thedemands diffegraph that 4 mthe duct for all
Fig.
The cost plotted for dexplicitly showfor all the aspe
Fi
Fig.7 showduct with samfor various du
l Journal of Appl
the total weighplate thickn
m and 3.5 mm pte thickness spher analysis 3binations of dae plotted.
TS AND DISC
ates the aspect with correspon
ous plate thickne plate thicknef 1.15 results ies. For 3.5 an
1) is found5 and 4 mm pla
is more whenof increased n
e weight of ththough the opte space availaerent aspect ramm duct gives l aspect ratios.
5 : Duct aspect
of the duct fordifferent aspecws that 4 mm ect ratios of the
ig.6 : Duct thic
ws the relationme flow area buuct thickness. F
lied Research In
ht and cost are ness. Since thplate are very pecified in lit.5 mm plate iata from table
USSION
ratio (depth/wnding weight nesses. The cusses of 4,5,6 an least weight
nd 4 mm platesd to be optiates, weight ofn comparing wnumber of stifhe other platestimum aspect rability at diffeatios. It is fouthe least over
t ratio with we
r various duct ct ratio.(Fig.6)
thickness plate duct.
ckness with co
nship between tut with variousFrom the grap
Ducting Optim
n Mechanical En
calculated for he difference minimum and
terature is 3.5 is considered.
e are analysed
width) of duct of the duct in urve pattern is and 8 mm. An
for 5,6 and 8 s, square duct imum. While f duct made of with that of 4 ffeners. When s with various ratio for 4 mm erent location und from the rall weight for
ight
thickness are ). This graph te is optimum
ost
the cost of the s aspect ratios ph, it is found
mization – A Case S
ngineering (IJAR36
f
that 3.5manufacand stifffor all thoptimum
VII. CO
It optimizafan outle• Wh
decrfounthicsolu
• Thoanalstiffcon
• Thethicthe any
REFER [1]. He
SanEn
[2]. PatAS
[3]. RoGaApCoAm(A
[4]. KeDueng
[5]. RaFoPuEn
Study
RME), ISSN: 223
5 mm plate cturing cost wfeners weldinghe aspect ratio
m compared wi
Fig.7: Duc
ONCLUSION
is evident ation approachet duct system ile it appears threase the numnd from the kness and uni
ution for the givough various stlysis, it is prudfener spacingsidering the ov
e analysis shokness of 4 mmcost of ductinmanufacturing
RENCES emalatha nkaranarayana
ngineering. trick J.Brook
SHRAE Journaonald L.Schneias Ducts for Ppplications, Aommittee of merican SocSCE),1995
erry Penningtouctwork Desigineering, Oct
aymond.J.Roarkrmulas for st
ublishers, 2002ngineers, (ASC
31 –5950, Volum
is the worswhich includes g cost of 4 andos, the 4 mm pith other plate t
ct aspect ratio w
N
from the mh adopted on a of a CFBC boihat increase in
mber and size oanalysis pres
form stiffener ven design contiffener spacingdent to conclu
g is the moverall weight oows that the m, under designg which is theg industry.
C, Sunasamy K, Au
ks, Duct desal, Jan 2010, Vider, The StrucPower Stations
Air and Gas the energy
ciety of
on, Worley Paign and En2008, Vol.12 (k, Warren.Ctress and stra2, American E), 1995.
me-1, Issue-2, 20
st consideringboth material
d 5 mm platesplate is found thickness.
with cost
mechanical dInduced Draft
iler that, n duct thicknessof stiffeners, sented that 4spacing is the
ndition. g are in use froude that the unost preferred f the ducting.ducting with
gn condition ree prime objecti
ndararajan utomation of
sign fundameVol.152. ctural Design os Industrial B
Structural DDivision of Civil Engin
arsons, Econongineering, P(10)
C.Young, Roain, Current
Society of
011
g the l cost s. But to be
design ft (ID)
s may , it is
4 mm e best
om the niform
one
plate educes ive of
S, Duct
entals,
of Air Boilers Design
The neers,
omical Power
oark’s Law Civil