EXPERIMENTAL ON-STREAM ELIMINATION OF RESONANT

WHIRL IN A JARGE CENTRIFUGAL COWPBESSOR

G. I. Bhat and R. G. Eierman Exxon Chemical Company Baytown, Texas 77520

In October 1982, a severe resonant whirl condi t ion was experienced when a m u l t i - s t a g e centr i f b g a l compressor was f i r s t o p e r a t e d a t h i g h e r t h a n o r i g i n a l l y a n t i c i p a t e d s p e e d s and l o a d s . D i a g n o s i s o f t h i s c o n d i t i o n was made easy by a large-scale computerized Machinery Condition Monitoring System ("MACMOS") . This computerized s y s t e m was- immediateiy a b l e to- v e r i f y t h a t h e predominant subsynchronous whirl frequency locked i n on t h e f irst resonant frequency o f t he compressor r o t o r and d id no t v a r y with compressor speed.

Compressor s t a b i l i t y c a l c u l a t i o n s showed t h e r o t o r s y s t e m had excessive bearing s t i f f n e s s as well as inadequate e f f e c t i v e damping. An optimum bearing design was developed t o minimize the unbalance response and t o maximize t h e s t a b i l i t y threshold.

The above experience is not unusual and p a r a l l e l s t h a t o f many process p l a n t s using l a r g e c e n t r i f u g a l g a s cmpress ion machinery. O f i n t e r e s t , however, is t h e approach taken by t h e p l an t t o f ind a temporary remedy. The e f f e c t i v e compressor bearing loading and e f f e c t i v e clearance c h a r a c t e r i s t i c s were modified with t h e machine continuing its process operat ion a t normal load and speed. This approach involved t h e control led app l i ca t ion o f heat t o t h e compressor support l e g s while c lose ly monitoring machine behavior. The experiment e s t ab l i shed t h e f e a s i b i l i t y o f extend- ing t h e onset o f r o t o r i n s t a b i l i t y i n t h e event t h a t p l an t ope ra t ions would ca l l f o r higher speeds before t h e optimized bearings became avai lable .

DISCUSSION OF PROBLEM U N I T

Description of Unit

The compressor is a two s t a g e e i g h t impeller ho r i zon ta l ly s p l i t machine dr iven by a steam turb ine . O i l seals are used a t both ends and l abyr in th seals are used a t t h e center . The compresser runs a t 175 ps ig i n l e t and 575 ps ig discharge. The r o t o r is supported by f ive shoe t i l t i n g pad load between pad bearings with 0.5 preload. The r o t o r has an unusually l a r g e r a t i o o f bearing span t o s h a f t diameter (99 in/5 in ) . The compressor t r a i n is safeguarded aga ins t high v ib ra t ion with dua l voting l o g i c v ib ra t ion t r i p circuits which shu t t h e compressor down i f v ib ra t ion amplitude exceeds 4.5 m l s a t any time (Figure 1).

A Kingsbury t h r u s t bearing is used t o absorb t h e th rus t . The compressor is driven through a Bendix type diaphragm coupling incorporat ing a torque meter and

81

h o t al ignment tubes. The compressor and the d r i v e r a re supported on a s o l i d concrete foundation.

P r o b l e m

P r i o r t o the f i r s t h a l f o f 1982 t h e compressor was running with s a t i s - f a c t o r y performance a t base capaci ty opera t ing cond i t ions o f 6000 rpn as the maximum speed (F igure 2). The r o t o r was subsequently mod i f ied f o r a d i f f e r e n t mole weight gas with a rev ised maximum speed o f 6500 rpm. The manufacturers' l a t e r a l c r i t i c a l speed and r o t o r s e n s i t i v i t y s tud ies d i d n o t r e v e a l any p o t e n t i a l problems (F igure 3 ) . The r o t o r was h i g h speed balanced a t 6700 rpm with an except ional balance q u a l i t y o f 2 oz-in. r e s i d u a l unbalance (F igure 4).

However, when an attempt was made t o increase the speed above 6000 rpm w i t h an increase i n load, a severe change i n s h a f t v i b r a t i o n was n o t i c e d (ampl i tude increase from l e s s than 1 m i l t o 3.6 mils) . Below 6000 rpm the compressor was running a t l e s s than 1 m i l v i b r a t i o n . Analysis o f v i b r a t i o n spect ra obtained from a computerized Machinery Condi t ion Moni tor ing System ( "MACMOS") revealed t h a t the h i g h v i b r a t i o n ampli tudes- o c c u r r e d a t a predzminant frequency o f 46.5 Hz (2760 cpm), which co inc ides w i t h t h e f i r s t resonant frequency ( c r i t i c a l speed) o f the compressor r o t o r (F igure 5). Varying the r a t e o f b u f f e r gas i n j e c t i o n t o the sea ls produced s i g n i f i c a n t changes i n the v i b r a t i o n ampl i tude suggesting t h a t t h i s so c a l l e d resonant wh i r l c o n d i t i o n was aerodynamical ly induced. Aerodynamic impulses created by gas e x i t i n g i m p e l l e r vanes ( l o a d dependent) p rov ide e x c i t a t i o n fo rce which, f o r systems wi th inadequate damping, can ampl i f y t h e v i b r a t i o n behavior a t t h e f i r s t c r i t i c a l speed o f the r o t o r . I t i s a l s o be l ieved t h a t the low r e s i d u a l unbalance i n th i s machine al lowed a f o r c i n g impulse o ther than unbalance t o pre- dominate and thus e x c i t e the f i r s t resonant mode o f the r o t o r .

A d e t a i l e d a n a l y s i s o f the rotor -bear ing system i n d i c a t e d excessive bear ing s t i f f n e s s as w e l l as inadequate e f f e c t i v e damping. A computerized study revealed t h a t a four pad load between pad bear ing i s an optimum design f o r the machine which would maximize t h e s t a b i l i t y th resho ld speed wh i le y i e l d i n g accepta- b l e unbalance response (Table 1).

However, s i n c e the new f o u r pad bear ings were n o t r e a d i l y ava i lab le , t h e approach taken by t h e p l a n t t o f i n d a temporary remedy with t h e machine cont inu ing i t s process opera t ion a t normal load and speed, i s innovat ive and worth sharing.

EXPERIMENTAL PROCEDURE

The goa l was t o e s t a b l i s h the f e a s i b i l i t y o f inc reas ing the r o t o r s t a b i - l i t y threshold, as p l a n t operat ions would r e q u i r e inc reas ing the r o t o r speed be fore t h e opt imized bear ings became ava i lab le .

It was a n t i c i p a t e d t h a t the goal cou ld be achieved b y modi fy ing e f f e c t i v e compressor bear ing load ing and e f f e c t i v e c learance c h a r a c t e r i s t i c s w i thout s h u t t i n g t h e compressor down.

For the purpose o f t h i s experiment, t h e four compressor support l e g s were provided with i n d u c t i o n heat ing elements which were wrapped with i n s u l a t i n g mater ia l . I n add i t ion , d i a l i n d i c a t o r s were s e t up t o moni tor v e r t i c a l and h o r i - z o n t a l movement o f the compressor casing. Power supply and temperature c o n t r o l

82

r e s p o n s i b i l i t y were assigned t o the truck-mounted laboratory , which was pos i t ioned near t h e compressor p la t form.

Dur ing t h e t e s t r e l e v a n t process parameters were h e l d constant and were cont inuously monitored by MACMOS on a second-by-second bas is and then converted t o s ix-minute averages, The v i b r a t i o n behavior o f t h e machine was, o f course, a l s o logged by an automated computer and a FM tape recorder dur ing t h e e n t i r e tes t .

The a c t u a l t e s t was conducted by step-wise heat ing o f compressor l e g s i n t h e f o l l o w i n g sequence: (1) inboard l e g s only , (2) outboard l e g s only, (3) a l l four legs, and (4) two d iagona l ly opposi te l e g s a t a t ime. Temperatures s t a r t e d a t 150°F and were r a i s e d a t 50°F increments u n t i l f i n a l support l e g temperatures o f approximately 400'F were reached. The v i b r a t i o n ampl i tude behavior was c o n t i n u a l l y monitored on platform-mounted v i b r a t i o n moni tors and readings conf i rmed by simul- taneously observing t h e computer-generated log. The taped v i b r a t i o n s i g n a l s were subsequently analyzed and t h e i r spect ra p l o t t e d as shown i n Figure 6.

The onset o f h i g h v i b r a t i o n was noted on MACMOS p r i n t o u t s as t y p i c a l l y shown i n F igure 7, Whenever a p reset l i m i t i s exceeded f o r a g iven parameter, an alarm occurs and da ta logging i s i n i t i a t e d . Relevant da ta can then be r e t r i e v e d and examined.

F igure 7 shows excessive averaqe v i b r a t i o n o f the compressor outboard bear ing as monitored by an eddy c u r r e n t probe. A la rm- in i t ia ted da ta logg ing was a c t u a l l y t r i g g e r e d be fore 10:27:39 by v i b r a t i o n spikes which must have exceeded the preset l i m i t o f 3.2 m i l s . Note t h a t outboard eddy c u r r e n t probe readings are represented by the numberal 1t8tt whose second-by-second values a r e hand-connected f o r eas ier v iewing i n F igure 8 .

The experiment demonstrated t h a t low-v ib ra t ion opera t ion a t t r a i n speeds o f approximately 6,360 RPM was f e a s i b l e with compressor outboard support l e g s heated t o approximately 375°F. A t these condi t ions, t h e compressor outboard end had grown 32 m i l s i n the v e r t i c a l , and 4 m i l s i n the h o r i z o n t a l d i r e c t i o n . The subsynchronous (2,760 cpm) v i b r a t i o n component was smal ler than the once-per- r e v o l u t i o n component a t t h i s speed. This i s g r a p h i c a l l y i l l u s t r a t e d i n Figure 9, which shows compressor outboard spect ra obtained under s i m i l a r loads and speeds, wi th d i s s i m i l a r support l e g temperatures. A t ambient temperatures, t h e subsynchro- nous frequency r e g i s t e r s an uncomfortable 2.8 m i l s on the v e r t i c a l eddy c u r r e n t probe. When the compressor outboard l e g s were heated t o approximately 375'F, t h e once-per-revolut ion and subsynchronous v i b r a t i o n components dropped below 0.35 m i l s .

CONCLUSION

Severe a e r o d y n a m i c a l l y induced subsynchronous v i b r a t i o n prob lems developed when the normal opera t ing speed o f a l a r g e c e n t r i f u g a l compressor was increased. An experiment was c a r r i e d o u t t o extend t h e onset o f r o t o r i n s t a b i l i t y t o h igher speeds on- l ine, wi thout changing t h e bas ic r o t o r bear ing system charac- t e r i s t i c s . The t e s t r e s u l t s i n d i c a t e d t h a t the e f f e c t i v e compressor bear ing support c h a r a c t e r i s t i c s c o u l d be mod i f ied and t h e s t a b i l i t y th resho ld be increased t o an acceptable l e v e l .

This experiment has demonstrated t h a t with proper inst rumentat ion and monitor ing, i t i s f e a s i b l e t o extend t h e s t a b i l i t y th resho ld speed o f c e n t r i f u g a l compressor, w i thout r e q u i r i n g a shutdown o f the equipment.

a3

ACKNOWLEDGEMENT

5 SLBP L/D = .3/PLF = .5

5 Si8P L/D = .43/PLF = .O

5 SLBP (55 percent o f f s e t )

3 SLOP

5 SLBP

5 SLOP

4 SLBP

7 SLOP TRESS-32

7 SLOP TRESS-68

The authors acknowledge M r . H . P. Bloch, M r . H. G. E l l i o t t , Mr . D, G. Stroud, and Mr . R. H . Schmaus f o r t h e i r support dur ing t h i s t e s t and subsequent data reduction.

11.4

8.10

8.97

11.8

10.96

15.5

6.11

10.96

17.24

REFERENCES

"Four Pad T i l t i n g Pad Bearing Design and Appl icat ion for Mult is tage Ax ia l Compressors", J. C. Nicholas and R. G . Kirk, ASME Paper, 81-LUB-12, 1981.

"Select ion and Design o f T i l t i n g Pad and Fixed Lobe Journal Bearings f o r Optimum Turborotor Dynamics", J. C. Nicholas and R. G. K i rk , Proceedings o f the Eighth Turbomachinery Symposium, Texas A&M Univers i ty , ed, P. E. Jenkins, November, 1979.

"The In f luence o f T i l t i n g Pad Bearing Character is t ics on the S t a b i l i t y o f High Speed Rotor Bearing Systems:, J. C. Nicholas, E. J. Gunter, and L. E. Barnett, Topics i n F l u i d F i lm Bearing and Rotor Bearing System Desiqn and Optimization, an ASME spec ia l publ icat ion, 1978, pp. 55-78.

"Optimum Bearing and Support Damping for Unbalance Response and S t a b i l i t y of Rotat ing Machinery", L. E. Bar re t t , E . J. Gunter, P. E. A l l a i re , ASME Paper NO. 77-GT-27.

TABLE I. - SUMMARY OF POTENTIAL ALTERNATE BEARING PERFORMANCE

-.c c< tl) Bearing Case

1 0) COPT

0.24

0.45

0.40

0.69

0.70

0.63

1.12

0.86

0.59

AMPLZ.

-93.6)

22.0

53.7

15.1

25.7

24.6

8.7

12.6

31.5

-0.034

0.143

0.058

0.208

0.122

0.128

0.362

0.251

0.100

2,451

2,399

2,431

2,403

2,415

2,439

2,421

2,415

2,446

Ref: (1) "Optimun Bearing and Support Damping for Unbalance Response and S t a b i l i t y o f Rotat ing Machinery", L. E. Earret t , E. J. Gunter, P. E. A l l a i re , ASME Paper No. 77-GT-27.

" S t a b i l i t y and Damped C r i t i c a l Speeds o f a F l e x i b l e Rotor i n F lu id-F i lm Bearings",

(2) J. W. Lund, ASME Paper No. 73-DET-103.

84

I *

Figure 1. - Compressor X-section.

85

FREQUENCY (EMNTS44IN x lm0)

(a) Cascade plot- Figure 2. - Base operating conditions.

86

CROBE 01 ID: CO?lPREfSOR<OTD VERT) ORIENTATION- 43.

CROBE 02 ID: CO?iPRESSOR(OTB MORT) ORIENTATION= 133.

UNFILTERED

nnx nnr- 0.49 nxLs PK-PK

mnx nnp- 0.62 nILs PK-PK

UT 1 ; : : : : i : : ' i . * ............................................................. .......... .......................

AHP SCALE- 0.10 ?IILS/DIV AHP SCALE= 9.19 ?lILS/DIV TIHE SCALE- 10.99 HSEC/DIV

ROTATION: CU RPfl<STRRT>= 4863 RPM<END)= 4857

PROBE 01 ID: COMPRESbOR(0TB VERT) ORIEMTATION= 4SO i x VECTOR- 0.87 nxLs PK-PK e-1160

?ROBE 02 ID: COf4PRESSOR(OtB MORZ) ORIENTATION= 1330

1 X FILTERED i x VECTOR= 0.16 nxLs PL-PK e-204

K

.... ....

... .... ....

.... ....

.... ... ....

... ....

.... M? SCALE- 0.03 IILS/DIV

ROTATION: CU

.......................+................................... ....................... . ...........-........... ............ ...........-.......... ....... .....-...........(.....*......

....... f....". .............................

........... ........... A ........... .. .....................

.............. . .*......I wv ...+ ....................... ivi .......... ....................... '+ .......................,......... L ........... 2 ........................ ..............

nnp SCALE- 0.03 nxLs/Dxv TIHE SCALE= 10.09 nSEC/DIV R?l(START>= 4069 RPll<END>- 4872

(b ) O r b i t s .

Figure 2. - Concluded.

87

Figure 3. - Critical speed map.

88

Figure 4 . - High speed balance p lo t .

89

f"

(a) Cascade plot.

Figure 5. - Uprated operating conditions.

90

?ROBE 01 ID: COHPREtSOR(0Tl) VERT) ORIENl i l l IOM* 4s.

PROBE 02 ID: COHPRESCOR(0Tl) MORT) ORltNTIlTIOM* 1350

UWFILTERED

m x nnp= 6.60 n i L s PK-PK

mnx nip- 6 . ~ 0 niLs PK-PK

K

IIHP SCRLE= 6.18 HILS/DIV

ROTCITION: CY

IIHP SCRLEI 0.10 HILS/DIV T IHE SCRLE- S.60 HSEC/DIV RPn(STIIRl>* 6 0 2 4 RPtl<END)- 6025

?ROBE 01 ID: COlPRESSOR(0TB VERT) ORIENTRTION* 450 ix VECTOR= e.07 n u s PK-PK e-1480

?ROBE 02 ID: COHPRESSOR<OTB MORT) ORIENTRTION* 1350

1X FILTERED ix VECTOR. e.21 ~ I L S PK-PK 0-740

K

m P 8CIILEm 0.bS RXLWDIV ( I lP SCIILE. 6.W I I I L S / D I V T l l E SCRLE- S.6B HEEWDIV

aotniiom: cu R?l<STCIRT>* Cb24 R?)I(END>= 6026

(b) Orbits . Figure 5 . - Concluded.

91

a. sx 1x ax

FREQLENCY (EKNlS/t i lN x 1000)

Figure 6. - Uprated operating conditions with heated legs: Cascade plot.

92

:?i 2% i i

I

ILa 2s

I I

i i

i ; i'

H I

n t

t CI - a;; 92 n o

1 . 1 ' I '

,

a,

&I W

a, *d U

m N

m N

0 rl

0 U

a3 d

m N

0 rl

U

E

E

..

..

..

..

5 * 0 d

m N

0 r(

0 U

a3 0

m hl

0 d

G

..

..

..

..

.d

M 0 d

a U a U

' 0 . a , m U r l a .d

E U .d hl F: 4 '0

- -

k; a P )

= a , z z ' g

22

.d f - u a a , &

ME- B4 .d

93

i i l

! I I

n

I ;

2

E (d k w

.d U

L n N

co N

0 r(

0 U

m m I- N

0 I4

k 0 cw

.*

..

..

..

4 U c) aJ k .d T3

I x c -d

c 0 *d u (d k .d 3 M c .d k (d aJ

-3 k (d 0 P U 9 0 I

co a, k 3 M

'A h

n

n

94

.. . . - . .. .

v)

M e, d

U f.4 0 a a 3 m a

f.4 0 0

3 0

f:

5 .rl 3

4 m 0 .rl U C e, a .rl

f.4 m e, C u (d

95