Sie sind auf Seite 1von 10

Divisions or Sections, or printed i n its publications.

$1.50 PER
75C TO ASME
COPY
MEMBERS
Discussion is ~r i n t e d only i f the paper is published
i n an ASME journal or Proceedings.
Released for general publication upon presentation
Analytical valuation of
Compressor-Manif old
J. C. WACHEL
F. R. SZENASI
Engineering Dynamics, Inc.
16117 University Oak
San Antonio, TX 78249
This paper discusses the types of vibration and stress problem encountered i n recipro-
cating compressor-manifold piping systems and mathematical analysis methods and digital
computer programs developed i n the research sponsored by the Pipeline and Compressor
Research Council of the Southern Gas Association. These techniques can be used to
calculate natural frequencies, mode shapes, vibrational amplitudes, and dynamic stresses
caused by acoustical and mechanical excitation forces in the system, and they have been
used i n the design of domestic and foreign compressor installations representing well over
10 million installed horsepower. Comparison of these predictions with experimental field
results has shown that accurate estimates of dynamic stress and impending failures can be
obtained.
Contributed by the Petroleum Division of The American Society of Mechanical Engineers for presenta-
tion at the ASME Petroleum Mechanical Engineering Conference, Tulsa, Okla., September 21-25,
1969. Manuscript received at ASME Headquarters June 27, 1969.
Copies will be available until July 1, 1970.
THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS, UNITED ENGINEERING CENTER, 345 EAST 47th STREET, NEW YORK, N.Y. 10017
J. C. WACHEL
Excessi ve vi br a t i ons and s t r e s s e s i n t he
pi pi ng of r e c i pr oc a t i ng compressor u n i t s a r e a
Droblem i n t he n a t u r a l gas and pet r ochemi cal i n-
dus t r y. Compressor c yl i nde r s and t h e i r mani fol d-
i ng b o t t l e s form a complex mechani cal system sub-
j ect ed t o dynamic pul s a t i on f or c e s i n t he b o t t l e s
and c yl i nde r s . Southwest Research I n s t i t u t e has
developed a n a l y t i c a l t echni ques f o r s ol vi ng t hes e
vi br a t i on and s t r e s s probleiiis. Thi s r e s e a r c h was
sponsored by t he Pi pe l i ne and Compressor Research
Council (PCRC) of t he Sout her n Gas As s oci at i on.
The complete compressor-manifold system i s
si mul at ed mat hemat i cal l y by de f i ni ng t he s t i f f h e s s
and mass c ont r i but i on of a l l s t r u c t u r a l members.
Nat ur al f r e que nc i e s and mode s hapes , bot h p a r a l l e l
and per pendi cul ar t o t he engi ne c r a nks ha f t , a r e
c a l c ul a t e d. Vi br a t i ons and r e s u l t a n t s t r e s s e s
i nt r oduced i n t o t he nozzl es can a l s o be det ermi ned.
Recent work on t he f l e x i b i l i t y a t noz z l e - bot t l e
J unct i ons has s i g n i f i c a n t l y improved t he accur acy
of t he a n a l y t i c a l c a l c ul a t i ons . Compressor i n-
s t a l l a t i o n s r e pr e s e nt i ng wel l over 10 mi l l i on
i n s t a l l e d horsepower have been desi gned u t i l i z i n g
t hes e t echni ques . Thi s paper w i l l de s c r i be t he
b a s i s of t he a n a l y s i s and how i t i s used i n i n-
dus t r y t o des i gn compressor syst ems.
COMPRESSOR-MANIFOLD SYSTEM
A t y p i c a l compressor-manifold system shown
Fi g. 1 Typical compressor-manifold system
i n Fi g. 1 i s a gas engi ne which i s dr i vi ng f our
compressor c yl i nde r s i n t he hor i z ont a l pl ane.
The ga s e n t e r s t he r e c i pr oc a t i ng compressor c yl -
i nde r s t hr ough t he s uc t i on mani f ol d b o t t l e ; i t i s
t hen compressed and di schar ged i n t o a di s char ge
mani f ol d b o t t l e . These s uc t i on and di s char ge
mani f ol d b o t t l e s a r e us ua l l y desi gned t o reduce
t he pul s a t i ons i nt r oduced i n t o t he gases by t he
r e c i pr oc a t i ng compression pr oces s . The nozzl es
which a t t a c h t he mani fol d b o t t l e s t o t he com-
pr e s s or c yl i nde r s a r e us ua l l y desi gned t o minimize
t he l oadi ng on t he c yl i nde r s and pr ovi de pr oper
a c ous t i c a l f i l t e r i n g , wi t h c ons i de r a t i ons a l s o
gi ven t o t h e i r f l e x i b i l i t y and pr es s ur e dr op char -
a c t e r i s t i c s .
The pri mary s t r e s s problem i n compressor-
mani f ol d i n s t a l l a t i o n s i s t he nozzl e s i nc e i t
Fi g. 2 Comparison of measured vi br a t i on and
pul s a t i ons
PULSATI ON
e UNmALANCcD 4 DYNAMIC rAl LURC
"!.RATION
MCCYANICAL r o ~ c r s 7 S T RC~ S PROBA~ILITY
CNCROY
I I I k I 1
I I I I
I I f I ! I
NATURAL
&rLkINGs CRLOUCNCbLS
Fi g. 3 Flow char c f o r a na l ys i s of system r e l i a -
b i l i t y
connect s t he maj or masses of t he system i . e . , t he
s uc t i on b o t t l e , di s char ge b o t t l e , and compressor
c yl i nde r s . When a vi br a t i on resonance of one of
t he s e masses oc c ur s , r e l a t i v e d e f l e c t i o n s be-
tween t he masses cause dynamic s t r e s s e s i n t he
nozzl es . The r e s u l t a n t s t r e s s i n t he noz z l e s i s
a combi nat i on of t he f l e x u r a l bendi ng and t or -
s i ona l s t r e s s e s .
Experi ence based upon appr oxi mat el y 400
f i e l d e va l ua t i on s t udi e s a t r e c i pr oc a t i ng com-
pr e s s or syst ems has shown t h a t , whenever f a i l u r e s
occur , t he cause can us ua l l y be t r a c e d t o t he
e x c i t a t i o n of a mechani cal r esonance. Reci pro-
c a t i ng p i s t o n mot i on ge ne r a t e s pul s a t i on ener gy
a t ever y engi ne harmonic; however, a c ous t i c a l
r esonances due t o t he combi nat i on of mani fol d
volumes, noz z i e s , chokes, i n t e r n a i passage geome- '
t r y , and t he s t a t i o n pi pi ng can cause c e r t a i n f r e -
quency components t o be a mpl i f i e d. Qui t e of t e n
t he a c ous t i c a l resonance w i l l coi nci de wi t h a
mechani cal n a t u r a l frequency of t he system. Thi s
resonance c ondi t i on us ua l l y caus es exces s i ve vi -
br a t i ons . However, hi gh ampl i t ude pul s a t i ons
away from a mechani cal n a t u r a l frequency may not
cause exces s i ve vi br a t i ons . These f a c t s a r e borne
out by t he da t a gi ven i n Fi g. 2.
When a mechani cal n a t u r a l frequency i s ex-
c i t e d by a c ous t i c a l pul s a t i ons r e s u l t i n g i n ex-
NOMENCLATURE
A~~~
= c r os s - s e c t i ona l a r e a of cr osshead gui de
ADP = c r os s - s e c t i ona l a r e a of di s t a nc e pi ece
Ai = c ons t a nt s
BJ = bol t e d j oi nt i n e r t i a of cr osshead gui de
1
BJ2 = bol t e d j oi nt i n e r t i a of di s t a nc e pi ece
D = di amet er of b o t t l e
d = di amet er of nozzl e
E = modulus of e l a s t i c i t y
FCHG = s hear form f a c t o r of cr osshead gui de
FDP = s he a r form f a c t o r of di s t a nc e pi ece
G = modulus of r i g i d i t y
I CH~ = moment of i n e r t i a of cr osshead gui de
.IDP = moment of i n e r t i a of di s t a nc e pi ece
K. = ge ne r a l i z e d s t i f f n e s s e s
1
Ki j = s t i f f n e s s ma t r i x
K - ~ = f l e x i b i l i t y ma t r i x
i j
L = l e ngt h of nozzl e
- l e ngt h of cr osshead gui de
L~~~ -
LDP = l e ngt h of di s t a nc e pi e c e
M = ge ne r a l i z e d moments
M. = ge ne r a l i z e d masses
P , F = ge ne r a l i z e d f or c e s
j
x = de f l e c t i on p a r a l l e l t o x a x i s
xi = ge ne r a l i z e d di spl acement s
y = de f l e c t i on p a r a l l e l t o y a x i s
z = de f l e c t i on p a r a l l e l t o z a x i s
,B = j oi nt f l e x i b i l i t y f unc t i on
= Kronecker d e l t a
[ = damping r a t i o
h = ei genval ues of st i f f ' ness- mass mat r i x
$ = f unc t i on
SUBSCRIPTS
C = c yl i nde r
CHG = cr osshead gui de
DB = di s char ge b o t t l e
DN = di s char ge nozzl e
DP = di s t a nc e pi ece
SB = s uc t i on b o t t l e
SN = s uc t i on nozzl e
7[ ( = dummy i ndi c e s
Fi g. 4 SGA compressor i n s t a l l a t i o n des i gn f a c i l i t y
c e s s i ve vi br a t i ons and s t r e s s e s , a l t e r a t i o n of t he
des i gn i s neces s ar y t o e l i mi na t e t he problem. Thi s
s ol ut i on can be accompl i shed i n e i t h e r of two ways.
One i s t o reduce t he magnitude of t he pul s a t i ons
and unbal anced f or c e s s u f f i c i e n t l y t o lower t he
vi br a t i on ampl i t udes t o a s a f e l e v e l . The second
method i s t o change t he mechani cal n a t u r a l f r e -
quency so t h a t coi nci dence wi t h t he maj or a c ous t i c
f r equenci es does not occur . Thi s w i l l r e s u l t i n
a r e duc t i on of t he v i b r a t i o n ampl i t udes propor-
t i o n a l t o t he mechani cal a mpl i f i c a t i on f a c t o r ( Q ) .
Changing t he nozzl e s i z e , l e n g t h , wal l t hi c kne s s ,
o r t he masses of t he mani f ol d b o t t l e s t o change
t he mechani cal n a t u r a l f r equency of t he system
may a l s o a f f e c t t he a c ous t i c r es pons e. Whenever
changes a r e made i n t he syst em, t he e f f e c t s of t he
a c ous t i c r esponse must be r e- eval uat ed t o be s ur e
t h a t no new coi nci dence of t he a c ous t i c and me-
chani cal f r equenci es occur . As di s cus s ed i n a
1
pr evi ous paper (L),
i t i s ne c e s s a r y t o make a
l unde r l i ne d numbers i n par ent hes es de s i gna t e
References a t t he end of t he paper .
complete system a na l ys i s i n or de r t o i ns ur e t he
adequacy of t he system r e l i a b i l i t y . Fi g. 3 gi ve s
t he flow c ha r t f o r a n a l y s i s of t he system r e l i -
a b i l i t y . The ul t i ma t e concern of t he engi neer
i s whether o r n o t t he system w i l l f a i l . In or der
t o det er mi ne t h i s f a i l u r e pr oba bi l i t y o r t he
s a f e t y f a c t o r of t he i n s t a l l a t i o n , i t i s neces s ar y
t o de f i ne t he pul s a t i on and mechani cal energy i n
t he syst em, t he n a t u r a l f r e que nc i e s , t he unbal -
anced f or c e s , t he v i b r a t i o n ampl i t udes , and t he
r e s u l t a n t dynamic s t r e s s e s . These a r e a s of con-
c e r n have been emphasized by t he Sout hern Gas
As s oci at i on' s PCRC r e s e a r c h over t he pa s t 15 ye a r s .
The Sout her n Gas As s oci at i on Compressor I n s t a l -
l a t i o n Design F a c i l i t y has become t he st andar d
des i gn t echni que f o r det er mi ni ng t he a c ous t i c a l
ener gy i n pi pi ng syst ems. Thi s f a c i l i t y , shown
i n Fi g. 4, has been des cr i bed i n s e ve r a l paper s
(2-2).
Thi s system us e s an e l e c t r i c a l anal og
model of t he a c ous t i c pul s a t i ons i n t he pi pi ng
syst em, t her eby enabl i ng one t o a c c ur a t e l y meas-
ur e pul s a t i ons and unbal anced f or c e s i n t he l ab-
or a t or y. Thi s i nf or mat i on can t hen be used wi t h
K A - AXI AL STIFFNESS
KF - FLEXURAL STIFFNESS
KT - TORSIONAL STIFFNESS
E N G I ~ ~
Fi g. 5 Spr i ng and mass nodel of compressor-mani-
f ol d system
+
MOTION
I
rnii +P- Fl - Fz - 0
mi.- Fs+Fq+FeO
&B+M- PO+T, +T~- S, ~, - F ~ ~ - O
Fi g. 6 Free-body diagram of c yl i nde r
t he mechani cal n a t u r a l f r equenci es and mode shape
da t a of t he compressor-mani fol d system t o pr e- COMPRESSOR CYLINDERS
d i e t vi br a t i on and s t r e s s ampl i t udes .
For t he c yl i nde r equat i ons of mot i on, s t i f f -
CALCULATIONS OF NATURAL FREQmNCIES
A mat hemat i cal model of t he e l a s t i c and mass
pr ope r i t e s of a compressor-mani fol d system has
been devel oped which s i mul at es t he response of t he
phys i cal system. The system i s concept ual l y r e-
duced t o masses and s pr i ngs a s shown i n Fi g. ?.
The noz z l e s and cr osshead gui de a r e r epr es ent ed
a s f l e x u r a l , a x i a l , and t or s i ona l s pr i ngs whi l e
t he c yl i nde r s and b o t t l e s a r e masses. The de-
gr e e s of freedom of t he b o t t l e s and c yl i nde r s a r e
i l l u s t r a t e d i n Fi gs . 6, 7 , and 8 . The n a t u r a l f r e -
quenci es and mode shapes of t he system can be ob-
t a i ne d by s ol vi ng the d i f f e r e n t i a l equat i ons of
motion f o r t he masses of t he syst em.
The equat i ons of motion which were wr i t t e n
f o r t he system a r e of t he form:
M. z. + C i i i t Ki x i = 0
I 1 (1)
where :
ne s s of t he cr osshead gui de and di s t a nc e pi ece
must be de f i ne d. The s t i f f ' ne s s e s of t he cr os s -
head gui de , di s t a nc e pi e c e , and bol t e d j oi nt s a r e
c a l c ul a t e d a s s pr i ngs i n s e r i e s t o obt a i n t he
e f f e c t i v e s pr i ng c ons t a nt s f o r t he f l e x u r a l , ro-
t a t i o n a l and e xt e ns i ona l s pr i ngs ( Fi g . 5 ) . The
expr es s i on f o r t he combined s t i f f h e s s of t hes e
beams was det ermi ned by cons i der i ng t he r e s u l t i n g
de f l e c t i on caused by a u n i t f or c e a c t i ng a t t he
c e nt e r of gr a vi t y of t he c yl i nde r ( Fi g. 6) . The
c a l c ul a t i on of de f l e c t i on c ons i de r s t he s hear
f or c e s and t he s l ope s a t each of t he j unct i ons
between t he d i f f e r e n t components. The r e s ul t i ng
expr es s i ons f o r de f l e c t i on and sl ope can be com-
bi ned i n t o t he f ol l owi ng equat i ons :
Mi = ith v i b r a t i n g mass
where
K- l K - l K- l
Ci = damping on i
th
mass
w 15' 61. ~2
For s i mpl i c i t y i n s ol vi ng f o r t he n a t u r a l
The f l e x i b i l i t y ma t r i x, ~51, which r e l a t e s
1
f r equenci es t he damping t erms a r e al l owed t o van-
t he de f l e c t i ons x and 6 t o t he f or c e s P and M i s
i s h . The p e ~, c e n t c r i t i c a l damping i n compressor- symmetric. In mat r i x form t he equat i on becomes:
mani f ol d systems has been det ermi ned by exper i -
ment al f i e l d t e s t s t o be appr oxi mat el y 0. 05
A, , = K - I F
r l 5 5 ( 4 ) 1
( Q - 1 0 ) . The di f f e r e nc e between t he damped and
undamped n a t u r a l f r e que nc i e s i s t r i v i a l .
where A,, r e pr e s e nt s t he di spl acement ve c t or ; ~ii,
* M,e Y s s
U, - Xsm- (Xc+b,8+ FM)
'T+= - J ; ,
J+-M,-M2-F,R+F2R+Kr+- 0
DISCHARGE BOTTLE EOUATIONS
'Fx,, - MD,~,,
'FY~, - Msa YDB
X F ~ - O
rn2
XM,,- 0
r n2, - Fq=O
Ji -TI -Feat- 0
~ig.8 Free-body diagram of bottles
I XDB
Fig.7 Free-body diagram of system determined as a function of the stiffnesses. The
resulting equations for zSB and asB can be solved
as in equations (2) and ( 3 ) . For these vectors
the flexibility matrix; and F;. , the force and
moment vector.
K& 1 ~ ; 5 , K5& Kt ; = + (LsN. ISN. ASN, DSN. PSB. E, G) (7)
The calculation of the flexibility matrix
inverse which is the stiffhess matrix K allows
'l [
All of the equations of motion are now in
the solution of the force vector. the form:
The forces and moments resulting from de-
flections of the crosshead guide and distance
piece spring are now determined for use in the
equation of notion.
The differential equations of motion were
converted into matrix form for simplicity of
solution. The general form of the equation is
MANIFOLD BOTTLES
For the manifold bottles the major springs
Are the suction and discharge nozzles. The noz-
zles have forces and moments which cause torsional
and bending deflections (Fig.7) . An expression
for the deflections xsB and +C can be written in
terms of the forces and moments imposed upon the
nozzle from the cylinder similar to equations (2)
and (3). For these displacement vectors
K&
KC:, K t 5 = +( Ls N. ISN, FSN, DSN, PSN, E, G)
(6)
The equations for the discharge bottle can be
written similarly.
The other degrees of freedom of the bottles
are illustrated in Pig.8. The motions are a
combination of linear and angular movements from
the equilibrium position. The calculation of the
forces and torques on the bottles require the
determination of the torsional and flexural spring
stiffnesses. The forces acting on the bottles as
a result of the linear and angular motion can be
where M is the diagonalized mass matrix and K is
the stiffhess matrix. The form of these equations
lends itself quite readily to the eigenvalue-
eigenvector solution method.
2
The relationship between x and y( i i = - w x)
is used to simplify the equation of motion:
where the w 2 represents the diagonalized eigen-
value matrix which is called lambda ( A ) .
The diagonalized mass matrix inverse is ob-
tained by forming a matrix whose diagonal is the
individual inverse of the individual masses.
Multiplying both sides of the matrix equation by
M inverse (iT1) gives
CYL l CYL 2
Fi g. 9 Exc i t a t i on f or c e s i n compressor-manifold Fi g. 10 Comparison of c a l c ul a t e d and measured
system s uc t i on b o t t l e vi br a t i ons
Rearrangi ng gi ve s :
Thi s form of a ma t r i x equat i on i s t he f ami l -
i a r ei genval ue e qua t i on. The val ues of A f o r
which t he equat i on i s s ol ubl e a r e known a s t he
c h a r a c t e r i s t i c va l ue s o r ei genval ues of t he ma t r i x.
The problem of f i ndi ng t he ve c t or s which s a t i s f y
t he equat i on i s t he r e f or e c a l l e d t he ei genval ue
problem f o r t he gi ven ma t r i x. Cor r espondi ngl y,
t he ve c t or s ol ut i ons a r e t he ei genvect or s of t he
ma t r i x ( T' K) , which i s r e f e r r e d t o a s t he s t i f f -
ness-mass ma t r i x.
Phys i c a l l y, t he ei genvect or s r e pr e s e nt t he
mode shapes of t he p a r t i c u l a r vi br a t i on c or r e s -
ponding t o t he ei genval ue which r e pr e s e nt s t he vi -
b r a t i o n a l n a t u r a l f r equency. The equat i on ob-
Table 1 Cal cul at ed ver s us measured n a t u r a l f r e -
quenci e s
Mode f cal c. - tmeas.
Unit A r i g ~ d body (x) 25 2 4
suct i on bottle ( c ) 42 3 8
cyl i nder (x) 58 58
Unit B r ot ar y (x) 6 7 73
cyl i nder (x) 39 39
suct i on bot t l e (2) 28 2 9
di schar ge bottle (2) 19 2 0
Unit C r i g ~ d body (x) 18 15
r ot ar y (x) 76 72
di schar ee bottle (2) 33 3 1
t a i ne d when t he det er mi nant of t he c o e f f i c i e n t
mat r i x vani s hes i s known a s t he c h a r a c t e r i s t i c
equat i on of t he ma t r i x and t he val ues of A f or
which t he equat i on i s s a t i s f i e d a r e t he de s i r e d
ei genval ues . I n ge ne r a l , t he c h a r a c t e r i s t i c equa-
t i on w i l l have "n" r o o t s wi t h "nu ei genvect or s .
The s ol ut i on f o r t he ei genval ues of a system
wi t h "nu degr ees of freedom would be an 'nth'' or der
equat i on whose s ol ut i on would gi ve "nu r oot s of
t he c h a r a c t e r i s t i c equat i on. The "n" r oot s would .
t hen r e pr e s e nt t he vi br a t i ona l f r equenci es squared.
The c h a r a c t e r i s t i c equat i on of t he example would
be a s f ol l ows:
where Ai, i = 0 , n r e pr e s e nt s f unc t i ons of mass
and s t i f f n e s s .
When t h i s c h a r a c t e r i s t i c equat i on i s sol ved
f o r t he r oot s o r e i ge nva l ue s , t hen t he ei genvect or s
can be obt ai ned. The e i ge nve c t or s a r e obt ai ned
us i ng equat i on ( 1 2 ) . Using t he s t i f f nes s - mas s
ma t r i x ( M - ~ K ) and mul t i pl yi ng by an ei genvect or
X and f or c i ng t h i s t o equal an ei genval ue t i mes
t he unlmown ei genvect or w i l l r e s u l t i n a s e t of
equat i ons . Thi s s e t of equat i ons i s s ol ubl e f o r
a uni que ei genvect or d i r e c t i o n ; however. t he mag-
ni t ude remai ns undet ermi ned. These ei genvect or s
r e pr e s e nt t he mode shape of vi br a t i on cor r es -
ponding t o t he ei genval ue o r n a t u r a l frequency of
t h a t vi br a t i on.
MODES OF VIBRATION
A r es onant mode shape nor mal l y occur s f o r
each degree of freedom. The l owest vi br a t i on
mode i n a t ypi c a l system i s ge ne r a l l y t he z r e -
CYLlNDER
. MEASURED
CALCULATED
Fi g. 11 Comparison of c a l c ul a t e d and measured Fig. 12 Comparison of c a l c ul a t e d and measured
c yl i nde r vi br a t i ons di s char ge b o t t l e vi br a t i ons
sonance of t he mani f ol d b o t t l e s . I f t he b o t t l e
i s n o t compl et el y symmetric wi t h r e s pe c t t o t he
c yl i nde r and nozzl e l o c a t i o n s , a combined a-- z
resonance of t he b o t t l e w i l l occur .
I n t he x d i r e c t i o n which i s p a r a l l e l t o t he
b o t t l e a xe s , s e ve r a l r es onant modes occur which
a r e of i n t e r e s t i n p r a c t i c a l problems. A low
f r equency r i g i d body mot i on of t he mani f ol d b o t t l e s
and a l l c yl i nde r s i s us ua l l y t he predominant mode
shape. Another x r es onant mode i s t he r o t a r y mode
i n which t he two b o t t l e s move i n oppos i t e x d i -
r e c t i ons and t he c yl i nde r s remain a t r e s t . In
a ddi t i on a c yl i nde r resonance mode occur s i n
which t he mani f ol d b o t t l e s remain a t r e s t whi l e
t he c yl i nde r s a r e a t r esonance.
Ot her r esonances which occur a t hi gher f r e
quenci es a r e t he c yl i nde r z r es onant mode which
occur s c ol i ne a r wi t h t hc c yl i nde r l ongi t udi na l
a x i s . The c yl i nde r 4 resonance i s a mode shape
i n which t he c yl i nde r r o t a t e s about i t s l ongi -
t udi na l a x i s .
CALCULATION OF VIBRkiTIClN AND STRESS
A computer program which us e s t he ei gen-
ve c t or method f o r s ol vi ng t he mechani cal n a t u r a l
f r equenci es and mode shapes of t he compressor-
mani f ol d system enabl es one t o c a l c u l a t e t he
f or ced vi br a t i on response of t he system us i ng
va r i ous f or c i ng f unc t i ons a t d i f f e r e n t mass l o-
c a t i ons . A Four i e r expansi on of any complex
f or c i ng f unc t i on, i nc l udi ng t he phas i ng, can be
a ppl i e d a t each mass l oc a t i on. Complex waves of
t ypi c a l unbal anced f or c e s i n t he s uc t i on and d i s -
charge b o t t l e s and c yl i nde r s a r e shown i n Fi g. 9.
s t a l l a t i o n des i gn f a c i l i t y .
The harmonics of t he complex f or c i ng func-
t i ons s i mul a t i ng t he magnitude and d i r e c t i o n of
t he a c t u a l f or c e s encount ered i n t he system a r e
a ppl i e d t o t he mat hemat i cal model. The dynamic
di spl acement s r e s u l t i n g from t he a ppl i e d harmonic
f or c e s of a l l t he masses can be c a l c ul a t e d. Vi-
br a t i ons a t va r i ous harmonic f r equenci es can be
recombined a s a complex vi br a t i on waveform. S t r e s s
These f or c i ng f unc t i ons can be obt ai ned from ex- Fi g. 13 Experi ment al f a c i l i t y f o r br anch connect i on
per i ment al f i e l d da t a o r t he SGA compressor i n-
r e s e a r c h
val ues based upon t he r e l a t i v e de f l e c t i ons be-
tween connect i ng members a r e c a l c ul a t e d f o r each
harmonic f r equency. Fl exur al bendi ng and t or s i ona l
s t r e s s e s i n t he noz z l e s c a l c ul a t e d f o r each t ype
of motion a r e t hen combined i n t o t he r e s u l t a n t
maximum s he a r i ng and pr i nc i pa l s t r e s s e s by t he
combined s t r e s s equat i ons .
CORRELATION WITH FIELD STUDIES
Mechanical n a t u r a l f r equenci es and mode
shapes measured dur i ng f i e l d e va l ua t i on s t u d i e s
of t he compressor-manifold system have c or r e l a t e d
wel l wi t h c a l c ul a t e d r e s u l t s . A comparison of
measured and c a l c ul a t e d n a t u r a l f r equenci es of
t he most commonly occur r i ng modes i s gi ven i n
Table 1.
I n a d d i t i o n , vi br a t i on d e f l e c t i o n s and dy-
namic s t r e s s e s a r e r out i ne l y measured on f i e l d
e va l ua t i on s t u d i e s and a r e compared wi t h t he cor -
respondi ng v i b r a t i o n s and s t r e s s e s c a l c ul a t e d by
t he pr e vi ous l y des cr i bed a n a l y t i c a l t echni ques .
A comparison of t he c a l c ul a t e d and measured va l ue s
of vi br a t i ons f o r a t y p i c a l f i e l d e va l ua t i on
st udy a r e pr es ent ed i n Fi gs . 10, 1 1 , and 12. While
c or r e l a t i on of i ndi vi dua l da t a poi nt s v a r i e s , t he
ove r a l l t r e nds compare f avor abl y.
When f i e l d da t a on t he v i b r a t i o n a l ampl i -
t udes of t he compressor-manifold system a r e a va i l -
a b l e , t h i s a n a l y t i c a l t echni que can be used t o
pr e di c t t he s t r e s s e s occur r i ng i n t he syst em. The
measured v i b r a t i o n a l ampl i t ude a t a poi nt can be
used t o det er mi ne t he v i b r a t i o n a l ampl i t udes and
s t r e s s e s of t he system a t a p a r t i c u l a r f r equer cy.
FLEXIBILITI I N NOZZm-BOTTLE JUNCTIONS
To a c c ur a t e l y c a j ' ~ u l ? 'e n a t u r a l f r equenci es
and s t r e s s e s i t i s neces s ar y t o cons i der t he
noz z l e - bot t l e j oi nt f l e x i b i l i t y f a c t o r . The s h o r t ,
s t i f f noz z l e s t y p i c a l l y used i n compressor-man-
i f o l d syst ems c,ause t he b o t t l e wal l t o d e f l e c t
and t he j oi nt t o r o t a t e . Anal ysi s of t he s e de-
f l e c t i o n s i s compl i cat ed by t he us e of r e i nf or c e d
br anch connect i ons such a s s a ddl e s , pads , sweepo-
l e t s , we l dol e t s , drawn o u t l e t s , enci r cl ement
s a ddl e s , and t e e s . Experi ment al da t a f o r each
r e i nf or c e d br anch connect i on a r e neces s ar y t o
det ermi ne t he j oi nt f l e x i b i l i t y f unc t i on, s i nc e
t he complex geometry i s not r e a d i l y s ol vabl e by
simple s h e l l t heor y. Data on s e ve r a l t ypes of
br anch connect i on de s i gns has been obt ai ned by
ot he r s ( 6 ) .
The a ut hor s have conducted s t a t i c and dy-
namic t e s t s on s e ve r a l noz z l e - bot t l e r e i nf or c e d
br anch connect i ons i nc l udi ng s a ddl e s , sweepol et s,
f l e x i b i l i t i e s and s t r e s s i n t e n s i f l c a t l o n f a c t o r s
f o r bendi ng, s he a r , and t or s i ona l l oads . I n
a ddi t i on t he e f f e c t of t he va r i ous r ei nf or cement s
upon t he mechani cal n a t u r a l f r equenci es of t he
system were obt ai ned. The exper i ment al f a c i l i t y
used t o make t he s e measurements i s shown i n Fi g. 13.
Equat i ons which can be used t o c a l c ul a t e
t he j oi nt f l e x i b i l i t y f unc t i ons have been obt ai ned
from a n a l y s i s of t he measured da t a .
CONCLUSIONS
Thi s paper ha s des cr i bed methods which a r e
p a r t of a cont i nui ng e f f o r t by t he Sout hern Gas
As s oci at i on ana Southwest Research I n s t i t u t e t o
pr e di c t t he response of compressor-manifold sys-
tems. Some of t he maj or concl us i ons which can be
made a r e :
1 Comparison between c a l c ul a t e d and rneas-
ur ed mechani cal n a t u r a l f r equenci es has shown
agreement wi t hi n accur acy r equi r ement s needed f o r
des i gn of r e c i pr oc a t i ng compressor i n s t a l l a t i o n s .
2 v i b r a t i o n s of t he compressor-manifold
system can be c a l c ul a t e d by us i ng t he a c ous t i c a l
f or c e s i n t he mat hemat i cal model. Measured and
pr e di c t e d vi br a t i on ampl i t udes compare f avor abl y.
3 The j oi nt r o t a t i o n a t t he noz z l e - bot t l e
connect i on g r e a t l y a f f e c t s t he c a l c ul a t i on of
n a t u r a l f r e que nc i e s , mode s hapes , vi br a t i on, and
s t r e s s e s . Research conducted dur i ng t he pa s t
year has r e s ul t e d i n improved equat i ons f o r pr e-
d i c t i o n of t h i s f l e x i b i l i t y , and methods f o r t he
s e l e c t i on of t he optimum f i t t i n g f o r p a r t i c u l a r
l oadi ng c ondi t i ons have been evol ved.
4 The a n a l y t i c a l t echni ques des cr i bed can
be used i n t he des i gn s t a ge i n conj unct i on wi t h
t he Sout hern Gas As s oci at i on compressor i n s t a l -
l a t i o n des i gn f a c i l i t y t o s t udy t he combined
e f f e c t s of a c ous t i c a l and mechani cal changes.
The dynamic s t r e s s e s and f a t i gue l i f e of t he com-
pr e s s or nozzl es can be pr e di c t e d.
5 In e x i s t i n g i n s t a l l a t i o n s i t i s pos s i bl e
t o e s t i ma t e t he s t r e s s e s i n t he compressor nozzl es
caused by vi br a t i ons measured on t he a c t ua l i n-
s t a l l a t i o n by f or c i ng t he mat hemat i cal system t o
have t he same vi br a t i on ampl i t udes . By us i ng
t h i s t echni que a n i n c i p i e n t f a i l u r e can be cor -
r e c t e d be f or e it occur s .
1 Wachel, J . C. , "Consi der at i on of Mechan-
i c a l System Dynamics i n Pl a nt Desi gn, " ASME Paper
No. 6 7 - ~ ~ ~ - 5 .
pads, and wel dol et s . These t e s t s det ermi ned t he 2 Ni mi t z, Wal t er , and Damewood. Gl enn, "Com-
pr e s s or I n s t a l l a t i o n Design Ut i l i z i n g an El e c t r o-
Acous t i cal System Analog, " ASME Paper No. 61-WA-
290.
3 Henderson, E.N. , "Gas Pul s a t i ons The
Problem, Sout her n Gas As s oc i a t i onl s Approach,
Re s ul t s , " O i l & Gas J our na l , Vol . 56, No. 1 9 ,
1958.
4 Hughes, J.V., and Shar p, J . M. , "The
Design and Oper at i on of a n ~ l e c t r i c a l Analog of
a Reci pr ocat i ng Conpressor I n s t a l l a t i o n , " ASME
Paper No. 56-A-200, 1956.
5 Damewood, Gl enn, "SGA and Southwest Re-
s ear ch Team Up t o Take t he Pul se of Compressor
pi pi ng, " Pi pe l i ne I ndus t r y, J ul y, 1956, p p 30-33.
6 Rodabaugh, E.C., e t a l . , "Eval uat i on of
Experi ment al and The or e c t i c a l Data on Radi al
Nozzles i n Pr es s ur e Ve s s e l s , " Sect i on V I , USAEC,
TID-24342.

Das könnte Ihnen auch gefallen