CONSIDERATIONS IN HOT AND COLD
ALIGNMENT AND COUPLINGS
by
Charles Jackson
‘Turbomachinery Consultant
Charles Jackson, PE,
‘Texas City, Texas
Charles Jackson is « Turbomachinery
Consuttané “and Owner” of “Charles
Jackson, PE, founded January 1386,
after serving 38 years tlth Monsanto
Company retiring asa Distinguished Fe.
law tn Corporate Engineering, reporting
10 St. Lous. Heisa Charter Founderand
‘Adelsory Committeeman of the Texat
AGM. Turbomachinery “Symposium,
Charter Director and Lecturer ofthe Vi
ation Institute, and @ Fellow in ASME
He has receoed the ASME Petroleum Dicisions “Oil Drop”
‘Acard, ASME Allan Chapman Award, ond is he frst retp-
fent of ASME’ Frederick .Smarro Awardin Plant Engineering
and Maintenance, He served on the ABI Subcomtce for
Mechanical Equipment for 2 years
He has a BS.ALE, degree fom Texas AGM University and
an A.A. degree in Electronics Technology from the College of
the Mainianel He's « member of Pt Tau Sigma and Taw Beta Pi
He has published over 50 techmiest papers and is the author of
ane book, The Protea Vibration Primer
ABSTRACT
Hot and cold alignment techniques ate addressed, with
zaphical plotting of reverse dial indicator alignment maps. At
alignment program witha two sets of datafor ross checking, and
Suggested proper tooling and procedures are offered. One ease
history on steam turbine to barrel pump is presented where the
turbine shaft vas fated by reverse bending fatigue, The cow
pling portion ofthis study deals with the oupling's importance
In the rotor effects and defects. Certain balance and asveinbly
Solutions and standards are suggested along with the proper fit-
ting of a coupling. A procedure for re-positioning certain cow
plings to match the eoupling’s and rotor’ residual imbalances.
‘Two ense histories ave included,
ALIGNMENT
It seems amazing to me that we can pass laser beam 200 eet,
barring heat wave distortion, and hold one to two mil aecuraey.
Further, we can attach a laser gun to one shaft and passa beam
@ to MB) to the ether shaft, return it via miczor and have an
fagenraey within one mil and angslarity within one minute. Yr,
‘no one seems to know the true heat vse of« particular machine
train. Ifa figure is tossed ont, it seems to be phis or minus 0
mils. Machinery builers sive charts that are not clear-whether
the cold alignment s tobe read in actual offsets or total indicator
readings (HAR), That would only be off by 2:1. The data ofest-
mated thermal movements, farvarded with aproject, seems the
same whether the machinery train #s m @ heated power plant
building, offshore on a open platform, a the Equator, or at
Prudhoe Bay
T would lke to define Coll Alignment asa” process of misalign=
Ing two or more casings ina train, so thatthe shafts connecting,
‘each pai of casings are collinear when a stabilized operating con
dition has been reached.”
1 would like to define Hot alignment as a “process of measur
{ng to oF more casings in a train to accurately determine total
casing moveinents in at least three planes." Normally, measure-
‘ments will continue until measurements stop changing, Le.
‘ess than one il in one hout or until the measurements begin
to vacllate, The eainimuan timeframe for this stable equilibrium
condition of both thermal snd other forces, e.g, hydrauli
could be tw hours (small/medium dnty pomgs wo gears); the
maximum timeframe has been 48 hours
COLD ALIGNMENT DEVICES
‘There are several cold alignment techniquet/devices avail~
able to perform alignment between two casing (excluding inter
ral alignment within casings)
+ Reverse Indicator Brackets: Two ses of ars extending fom
‘one shaft over to the apposite shaft, utilizing dal indicators, to
indicate the total swept readings as one rotation, minimum, is
completed. The bars should be sufficiently rigid and the reach
should be practical, e.g., within three feet. The sag of the bar
‘must be measured and that sag subtracted from the readings
taken, From a realistic stand point, itis better to graph those
readings and to cross check with a programmable calculator
+ Optical alignment using precision ptical equipment in
combination with internal targets, external targets, rotating mnit-
rs, ete
+ Laser equipment with or without reflecting mirrors to trans-
late the data points as ether a fll revolution is made 0 ¥, ¥,
for Me revolutions are made. These lasers and photoelectle re
ceivers can be mounted to the shafts that are being rotated and
requive very small axial spacing on the shafts and ean be used
tap o about 30 femaximum spats
+ Mountings for eradlnga standasd accurately machined ref
erence rod in a four way rectifier type of instrumeist readout
‘Thisallows shafts to be aligned without rotation of the shaftsbut
requiring slightly over one inch, axial length, of clean diameter
shafling
+ Micrometer sewlings taken between shaft ends on various
‘quadrants, The face ahd rim inetbnd could be included with this
type of measurement as they are both considered less accurate
methods,
+ Less accurate methods need not be discussed.
HOT ALIGNMENT DEVICES
‘There are several techniques/devices for measuring the ther
mal movements of machines while these machines reach
‘equilibrium temperatures and forces8 PROCEEDINGS OF THE SEVENTH INTERNATIONAL PUMP USERS SYMPOSIUM
+ Eddy curtent probes mounted on water stands displaying
the shaft postion in three planes or the easing postion in tee
planes. These stands are generally mounted to separate routed
baseplates or to stable sole plates of machinery. Water, taken
froma stable temperature circulating ystems, Le., cooling to-
wers, is used to keep the stands at axed reference position
+ Optical aligament equipment using precision levels and
tvansits, This type of equipment is special and utilizes “split r-
flex bubble” leveling of the transit o level, on each reading,
with the “gun” turning on jeweled bearings. Single or “combina
tions of instruments-levels and “plunging” transits, bench
smatks and mitrors, pls “see through’ targets allow diferent
readings tobe taken at proper time intervals
+ Benchmark gauges, which allow direct measurement in in-
ches and mils to be physically measured between two reference
tooling balls which are mounted in pairs at (1) the bearing hous
ing (or easing location) and, (2) at the machine base, generally
in the mounting platform, mezzanine, or grouted baseplate. An
“inclinometer” plus calibrated distance measuring gauges are
used in combination and can be graphed or programed on aeal-
culatr or computer
+ Mounted laser instruments attached to one machine and re-
ceiving or reflecting laser beams atan adjacent machine.
+ Instrumented couplings abe to record and transmit data as
the articulation of the coupling is forced during. transient
movements.
+ Eddy current probe instrumented reverse bas(cooled
Invar) attached atone beaving housing and indicating to the 2d>
Jacent bearing housing. These bars represent a dynamie version
ofa "reverse dil indicator bracket set”
+ Stiobe ted “eezing”ofverniermessured variations fone
shaf's movements relative to another sensed in the coupling
Jal windows in the coupling guards [Mucra,
ALIGNMENT GRAPHING (OR) CALCULATIONS
‘There are various types of graphs, plotting charts, ratio charts,
ete, in order to present two oF more shafts in some graphical
form of understanding, to allow one to move equipment around
correctly. OF equal importance, in my opinion, is conveying
‘some understanding to the machinists or millwrights.
"The advantage of a linear scaled graph showing the total
machinery scale should be obvious, if not
+ It allows one to make judgments (adjustments) in total
‘machinery move logic to best accommodate large changes, e.8
‘whereas, the motor was to be moved, it seems best to move the
‘pump. Whereas, the turbine and the pump were to be aligned
to the gear (nocmal) its apparent that “dropping” the gear box
staight down by 20 ils accomplishes the tolerances required,
* There is a normal logic to selecting the machine to be
aligned. It is very simple-move the less complicated machine.
Examples are:
“between a motor and pump, move the motor,
+ between a steam turbine and pump, move the pump.
+ between a steam turbine, gear, and pump; nove the tur-
bine and pump,
between a steam turbine and compressor, move the com=
pressor
+ between a gas turbine, gear bot, and pump; move the
gear and pump
+ Often the methodical process of moving down the train via
{wo coupled sh ata time would render the st Boy “elt
+ Regardless which type of hot alignment device isto be used,
‘acold position chart is necessary in the beginning, in the end,
‘and normally throughout the process.
+ Ifone is to use a calculator to make the alignment caloula-
tions, the calculator program does not know any orientation, un-
less the programmer gives tone. The calculators generally im-
ited to two machines, although it does not need to be. Two
‘machines will be shown hereto illustrate the relationships of the
“similar tuiangle relationship ofen used.”
ALIGNMENT FIGURES AND EXPLANATION
Referring to Figure 1, a type! alignment plot ofa pumplor
gnpenon sen Ling lene a He rine For
{lostration,asimpleeod alignment ofthe pumptsrequired
tothe turbine, Le. this is notanormal fel stuatfon that wl
be shown in Figure 4. Generally, the alignment programs wil
fix “orientation” loldng over the stationary machine (rbine)
looking a the moveable machine (pumproompresor). Further,
this means thatthe ‘reverse dial indleatr readings horizon-
talreadingsmoves willbe defined as "Te oii," sice there
Seems tobe litle confusion on top" and “bottom.” In Figure 1,
LS signifies the “Tet eld reading’ taken vith the dal indicator
a the turbine while the each bars attached to the pump
Shaft. LA is the corzexponding “lef” dal indicator reading with
the dil indiator a the pump shalt, with the “reach bar sup-
ported tothe turbine shaft. The Li and L2 values, inthis exam-
Dle would be desired reading aud are"rero"for Li and L2 (reter
{oaligament progaa)
cee Braet
@ 2
=
L bee. seul | _ver. f
t a ‘Saat
] feo.)
- 2s
ens
Figure 1, Typica? Turbine Pump Alignment Graphical Plot.
In looking at Figure 1 and the "sag corrected” alignment data
recorded, this condition ean be conected “vertically” by adding
44.30 ml shim at the pump’s near foot and a 10 mil shim under
the pup s far foot. No horizontal misalignment is indicated as
the “Tet” readings equal th “right” readings.
A large reverse dil indicator bar set is shown in Figure 5
beingused ona turbine drive with seven inch demeter coupling
Ihubs, a 16 in diameter coupling and an 28 in spacer. The "Sa
‘Check for Umi ag at 22 in span i shown in Figure 6
However, to understand the technique and the relationship
used, please refer to Figure 2. Now the vertical lines 8 and Id
indicate the reverse dial indicator spans, and the 18 postion be-CONSIDERATIONS IN HOT AND COLD ALIGNMENT AND COUPLINGS 9
‘comes the “common reference plane.” The distance D1 is the
indicator span(I3 to Id). D2 the distance, 13 to the NF (near
foot). D3is the distance 18 to FF (a fot). By enclosing the po-
sition ofthis shat in the “triangles” shown, one can decide how
{an alignment program can be configured, ie., DU is propor
tional to D22..Day3..DUya.. Now, € can be seen that the
‘ear foot correction is (B9-B1)-2), and the fr foot correction
is (BS-BL)3). Now, (B3-B) is the algebra difference in the
“vertical” dal indicator readings divided by 2. And, y2 = (D2/
Dilylsand y8= (D8/DD yl
DESIRED (CESEN)RODNGS
*
vOn aon {a |
ne
& or u On
MI=(BS-£1) (0-82), 0,
Dini ~02/r2 ~03/13
Di/rimp2/v2;01/¥1=03/"3 Ley
ye=(o2/o1)v1
y3=(03/01)"1
‘Y(NLFtcorr)=(83—B1)=¥2
ME Ftcor)=(63-B1)-3
wane 11a
Figure 2. The Logic of Similar Triangles in Reverse Alignment
Calculations
The
&:
+ Allows onetoenterthe dial indicator readings dested, then
enter the dal indieatrs readings from the fe, then be fn four
keys forthe correction necessary at the to fet.
+ Te asks forthe distances DI, D2, and D3
+ Teasks for the dal indestor veadings i sucession TI, Ll,
BL RL
+ Tt assumes that the dial indicator readings inthe feld have
‘been corrected for sag by subtracting out the minus readings
based on a sag chedk of tne bass be used. Example, the THR
sag ofa bar were ~2 mis), then the horizontal readings would
aodified by adding +1 mil = ~ (—1 mil. The bottom vertical
reading would be modified by adding +2 mils = ~ (~2:nib)
Usekson, 1985)
+ Tedoes not requie a printer, but one ean be use. Ie does
require an HP-A calculator
* By entering [NEQ] ALPHA “I” ALPHA, «second set of
readings cam be enered fra “trim,” without entering the di
tances or desived readings
igniment program included in this paper (APPENDIX
In Figure3, one en se what would happen in graphical plot
ting, ithe same vertical reverse indicator readings” were used
asin Figuce I, but the turbine is tobe aligned tothe pump. A
Slight shift in the horizontal values onthe turbine readings i
nade to show the resulting swing ofthe “outboard” end ofthe
toring, hextzontals tothe pump. This two mil difleenee on
centerline forthe one indiator, pls ont as #32 mil (near foot)
‘nd 8.3 mil x foot) conection (right side move).
Figure 3. Alignment Plot Using igure 1 Data but Moving the
Turbine.
In Figure 4, the typical “required” (based on hot equilibrium
conditions) plot ean be seen. It simply says that both the turbine
and the pump are going to heat up in service and move to some
equilibrium position. That equilibrium condition should leave
the turbine and pump as near collinear as can be determined
through heat rise measurements, The upper set of numbers
then become the "reverse indicator cold reading” to affect a suc-
cessful hot running train. These readings become the desired
cold. postion “veadings.” For turbomachinery tains (speeds
‘greater than 3600 rpin) the offsets 3 each indicator location
‘should be 14 mil (0,0005 in) per inch of axial span between the:
‘wo indicators, For pumps 3600 rpm and less a 1 mil per inch
tolerance can be accepted. The current actual cold position
(ACP) would require the near foot of the pump tobe raised 20-21
inils and the Far Foot to be raised 40-42 mils. No horizontal
‘moves are required asthe turbine is currently three mile to the
‘ight of the pump, a condition that could have been desired
based on a movement of the turbine “left” three mils due to
Be
Fr neni
wong | BS] wong berg
+
nee,
anne ies
Figure 4. Typical Alignment Plot Based on Thermal Data-
Turbine and Pump,30 PROCEEDINGS OF THE SEVENTH INTERNATIONAL PUMP USERS SYMPOSIUN,
fnleVevhaust Hines entering the turbine from the righthand sid,
for example
“The aumnbers in (parentheses) were outputted from the pro
gram shown in the APPENDIX and are generally more acc
rite, depending on the plotting accuracy. The field readings
taken were already corrected for one mil bar sag (bvo mils TLR
sag). The horizontal readings are corrected by subtracting a(—1
‘mil sag or —(— 1)= +1. The bottom readings are ontected by
subtracting a (—2 ails) sag or ~(~2)=+2. The pre-sug cor
rected readings for L3, B3, R3, LA, Ba, Ré would have been
“+10, +26, +16, 25, ~56, —31 respectively (Figure 6)
Figure 5, Photo of Reverse Indicator Bor Set for One Mil Sag in
22 Inch Span.
Figure 6. Photo of Sag Check for the Brackets in Figure Susing.
Sis Inch Pipe
CASE HISTORY
The case history involves one often boiler feed pumps whick
have steam turbine drivesin the 15,000 np range with the ariel
‘pumps delivering 4,000 psig water. The coupling is a flevible
‘gear type at 18 in spacer length. The steam turbine ean use ei
ther HP or LP steam. The exhaust bearing is overhung. The tur:
bine has turning gear. The turbine shaft "broke though’ near
the turing gear. The overall cold misalignment was 87 mils
total. The offet plotted continued over 150 mus in the same,
i.e., “worst” direction. There war resistance to correct the situs
tion since the inlet and oxtlet dies to the pump are all welded,
‘Continued operation ofthe stubbed turbine shaft would have re:
sulted in another fishure. The vibration level was totally unae
ceptable, Le., a certaiz speeds, over eight mils peak-to-peak.
‘The shalt orbit plotted by two orthogonal. displacement
probes represent the locus ofthe high spotof the shaft, Net only
did this turbine drive exlubit strong “jump phenomenon dat
ing the polar plots, but the circular orbits fist showed the “de-
fated beach ball” pattern followed by the “figure eight” pattern,
‘worst coneition, This can be seen in Figures 7 (deflated beach
ball orbs). The vibration spectrum showed the 2x at 200 per
centafthe 1 value
4C BFPT Orbit
4300 rpm
Figure 7. Orbit of Boiler Feeducater Pump Turbine under Mis-
alignment Preload.
The orbit shown in Figure 8 of the (eight-bovdie pattern)
the preload ineveased with speed and load. The heavy shat pre~
Toad is in the direction 90 degrees to the long part of the figure
eight, One way of looking at the vibration peters shown in the
orbits to consider the shaft under preload, It cannot move in &
circular direetion, but must move wheee the high preload does
rot exist, ie, in the direction ofthe ong motion (ox) the igure
eight.
‘One must always consider thet some uid geroscopfes come
into play, e.2-, push down. .go right. push up.-ga lef. Further
depending on the direction af ad/movement relative tothe lo
Figure 8. Orbit of Figure 7, but with Secere Misalignment 100:
200 mils (Fight Inches).(CONSIDERATIONS IN HOT AND COLD ALIGNMENT AND COUPLINGS 31
cation ofthe vibration sensors, ther one can probably deduce
that both sensors ean show 1’ and 2 motion (in some case),
avd one sensor can see song 1x and the other sensor see
strong] and even sttonger3 x , Wone can cary that argumer
‘one step further, to where the flexible gear teeth ofa coupling
become “locked then another strong component eomes in fiom
De onbalance and 1% and 2x axial contributions
‘The motion of the turbine casing and bearing pedestal was
measured with wodal instrumentation at several points. This,
notion responding from the “preload” ean be see's i Bi
where one would be looking atthe turbine vertical moti
{ng back from the coupling position. The motion ofthe turbine
‘ase and separate overhung bearing pedestal
14s viewed fom alongside he turinecoupling ooking pra
lel to the earth,
Can
the coupling be compressed in place to drop ot the spacer vith
but special tcl Hydrautie dilaton reqites one or two pomp
Stho fo providing that? How much s the draw (advance) What
Is the taper onthe sha? How is that taper being maintained?
‘Are plug and ring gauges being supplied? Are they made to a
Standard? Does one use «gnge to hp a fll taper? (NO) Are
inpping tools made separately and compared to a standard?
GQESH IT lap a sha taper and a hub bore, do I change the r=
quired spacer length? (YES!) How do I maintain the proper
Spacer length without ming the equipment (Special shim?)
BoThave them? (Probably net) tt have to move the vine
tment analy lf be bole bound? (Maybe )IF ose te pump
or motor sly, do need to recheck alignment? (ESD
+ Most motor drives require some form of “limited end float”
uring startup, How is that provided? Is tthe correct amount?
How will it be maintained?
+ Does the coupling require lubricant? Grease or oil? What
type? How much? Haw i provided? Packed or injected? How
much oil must be injected?
APPENDIX 2
‘An APPENDIX has been added to suggest a method of using
hydraulic dilation iting of couplings osha A specie interfer-
ence [s being used as an example (APPENDIX 2, Figure A). An
interference chart is shown plas a formula for determining the
fit given a set of conditions (APPENDIX 2, Figure B)
® , e a
Seas
@)
ce)
a
APPENDIX 2—Figure A. (1) Remove “O” Rings, Engage Hub
to Shaft, Gage and Set Stop Bing for Advance; (2) Reinstall "O”
Rings and Back-Up Rings, Re-engage Hub, Bleed and Connect
Pumps; (3) Keep Pusher at 800-1000 PSI, Dilate With Bore Pres-
‘sure, Hold @ Stop R: (4) Hold Pusher Pressure, Remove Dila-
tion Pressure, e.g. 40,000 psi, Remove Pusher Pressure, Install
Retention Nut (if any), Recheck/Remove Stop R.38 PROCEEDINGS OF THE SEVENTH INTERNATIONAL PUMP USERS SYMPOSIUM.
caine
paw Oia
eoiras
Sree ess Lin
[Se cr cual sere
ToL -ADWANGE GAUGECFOR 1/2"/FT,
{TAR“oo NOT LAP STH COUPLING RUB)
(8 btad ARTI. OF TE.cUSTOR)
EES DILATON TooisCOESS/A.C->
Por £ Et 0o?-pe
2D, 0
heink Preseure @ Wub-to-Shatt, ze
Unterferenceub-tochore, sachs
nodolus tye
UGoeetde Bia, oF Cooling
(Pal Abe-Seety)
Dye Bis of Coupling Bore, anchor
APPENDIX 2 Figure B. Interference Chart
BIBLIOGRAPHY
cal
at, M., Presentation to Houston ASME Chapter, unpub-
ed (1988),
rosjack, M. J-, Drawing for API 671 Task
J. Bssinger, Chairman
Guy, KR, Eshleman, R. L., and Jackson, C., “Ausibiary Tar
bine Subsynehronous Vibrations,” (Misalignment Shaft Fail
uve) Vibration lastitite (1988),
Jackson, C., "Cold and Hot Alignment Techniques of Turboma-
‘chinery.” Proceedings of the Second Turbomachinery Sym-
>posium, Gas Turbine Laboratory, Departayent of Mechanical
Engineering, Tesas A&M University, College Station, Texas,
asa)
Jackson, C. “Repostoning A Couplings Residual Unbalance
“To Correct A Rotors Residal Unbalance-CPLGBAL, Vibra
tion Institute, Clear Lake, Texas, and Annual Rotordynamies
Balancing Course (1984),
ce Committee,
Jackson, C., "Reverse Indicator Cald Alignment With Graphical
Plotting,” Proceedings of the Second International Bump
Symposium, Turbomachinery Laboratories, Department of
Mechanical Engineering, Tesas A&M Universite, Callege
Station, Texas (1985)
Jackson, C., “Alignment and Baling Notes,” Vibration Insti
ie MVAI Course, Annnal (1956).
Jackson, C., “Flexible Couplings,” Amnnal Mechanical En-
gineering Seminar, “Flexible Couplings.” Minden, Nevada
(1387-9),
Murray, M., Paper submitted for presentation, 19th Turboma-
chinery Semposium (September 1990
Weaver, FL, A.I.Ch.E. 70th National Meeting, “Rotor De:
‘sign and Vibration Response” (197),
Winkler, AR, "High Speed Rotating Machinery Unbalance,
Coupling or Rotor,” Vibration Tnstitute, Nassau Bas, Tesas
(983)