Automatic regulation of clearance in a tilting pad journal bearing/Automatinis tarpelio reguliavimas atraminiu ideklu veleno guolyje.
Marcinkevicius, A.H.
1. Introduction
Hydrodynamic tilting pad journal bearings are widely used in
turbines, compressors, pumps, machine tools and other machines where
rotors are revolving for a long time without frequent start and
stoppage. Their advantage is in smooth and accurate revolving motion,
good vibration damping properties, high stiffness, service durability.
But the main problem of such bearings is the value of clearance between
the journal and pad. First of all, at assembly of such bearings in many
cases there are not adequate means to measure the clearance, the second,
at operation of different bearings, the temperature in different cases
after starting from 18-20[degrees]C at longer work increases (in
researched bearings [1] up to 41-43[degrees]C, at other experiments
[2]--up to 55-65[degrees]C; at approach to failure the temperature may
increase up to 85[degrees]C). Thermal deformation of different elements
of bearings is different, it may hardly influence on clearance value,
and at absence of means for clearance control it is necessary at
assembly to set the maximal clearance for the worse expected work
conditions of the bearing. Increased clearance increases shaft
eccentricity at revolving and decreases bearing stiffness, decreases
dynamic stability of the system. Influence of clearance on properties of
bearings also is analyzed in works [3, 4]. One can come to conclusion
that without knowledge of clearance between the journal and pad, though
accuracy of journal rotation is measured, it is difficulty to solve
about work quality possibilities of bearings [5, 6].
2. System of automatic regulation
For solving the problem of clearance control the patent application
of multi pad self-aligning automatically regulated bearing was proposed
by us [7]. Fig. 1 shows the scheme of automatically controlled three pad
bearing. The self-aligning pads 1 are placed on pins around the
revolving journal 2 in a body 3. Pin 4 is conventional, fixed in body 3
by the screw and counter nut. Pin 5 is assembled on balls 6 with
possibility of axial motion in body 7. For that purpose from lower side
pin 5 supports to dish type spring 8, from the upper side it by rolls 9
supports to wedge 10, from the other side wedge 10 by rolls 11 is rested
on self-aligning support 12. The wedge 10 is driven by servo drive 13
(e.g. by pjezoelectric or other actuators, which are widely used as
driving elements [8, 9]). At forward motion of wedge pin 5 moves down
decreasing the clearance between pads 1 and journal 2, and tightening
the spring 8, at backward motion of wedge spring 8 picks up the pin 5
increasing clearance between pins and journal. The hydrodynamic force
acting in the bearing is measured by pin 14 which by thread is connected
with bushing 15 by thread assembled in body 3. For measuring purpose the
pin 14 is hollow inside, and the measuring stick 16 is assembled in it.
The stick 16 is riveted in the head of pin 14, the head is connected
with the pin body by the neck eccentric to axis of pin by value e (Fig.
1). Because of eccentricity the neck at action of axial force bends, the
bending displacement is enlarged by stick 16 and transmitted to
transducer 17. Transducer 17 is electronically connected with the
clearance control system, which by control the axial position of pin 5
keeps the set value of hydrodynamic force acting between the pads and
journal. Assembly of pin 14 in additional bushing 15 is necessary for
the reason that at assembly of journal 2 in two bearings (one in the
front part of body 3, the other in the rear part of body) it is
necessary to keep together an accurate position of journal axis in the
body 3 and to keep accurate position direction of eccentricity e: it
must be in the imaginable plane of action of resulting hydrodynamic
radial and tangent forces loading the pad, it is perpendicular to the
journal axis. If eccentricity e will be turned in accordance with that
plane, the radial force will be measured with an error. For this case
the pin 14 at its upper side has two cut plane sides 18, direction of
eccentricity e and axis of transducer 17 are parallel to these sides
(Fig. 1).
Meaning of other parts of bearing design (Fig. 1) there are not
explained because it is understandable from the drawing: there are
counter nuts for fixing of set position of pins, connecting elements of
servo drive 13 with the wedge 10, etc. Only one support with the pin 5
is used for regulation because at three pads 1 the self centering of
pads is achieved.
[FIGURE 1 OMITTED]
3. Analysis of bearing properties
At design of the bearing it is desirable to find the more sensitive
one which would give the biggest elastic displacement of stick 16
measured by transducer 17 and together with it would guaranty enough
strength of the pin which could keep the load of hydrodynamic force
produced at revolving of the journal 2. For that reason it is necessary
to know the maximal load which can be created by the hydrodynamic force
and strength of pin neck. The hydrodynamic radial force of oil film
wedge [F.sub.0] is expressed by equation [10, 11]
[F.sub.0] = 5.1 x [10.sup.-11]
[mu]nD[B.sup.2]L[C.sub.L]/[[DELTA].sup.2], N (1)
where [mu] is oil viscosity coefficient in centipoises (cP); n is
revolving frequency of the journal rev/min; D, B, L are diameter of the
journal, width and length of the pad accordingly; [C.sub.L] is
coefficient, [C.sub.L] = 1.25/[1 + [(B/L).sup.2]]; [DELTA] is value of
clearance between the journal and the pad, in mm. Diameter of the pin
head sphere of such a bearing is [d.sub.s] = 24 mm.
Lubrication oil for three pad journal bearings at large revolving
frequency of a journal in common is accepted with very small viscosity,
approximately 4 cP at temperature 55[degrees]C. It coincides with the
lubricant viscosity ISO 6, a little below the Shell Tellus oils Do 10
lubricant. Such is the oil [??]5A produced in Ukraine. Further
calculations we will do on the ground of that oil. Let for calculations
we will accept the bearings of cylindrical grinder which has three pad
bearing with journal diameter D = 70 mm and pad dimensions B = 36, L =
55 mm, number of journal revolutions 3000 rev/min. Fig. 2 shows
dependence of force [F.sub.0] on clearance [DELTA] for such a bearing.
Calculated value of force [F.sub.0] at clearance [DELTA] = 10 [micro]m
is 26722 N.
[FIGURE 2 OMITTED]
Fig. 3 shows the scheme of cross-section of the neck eccentric to
internal hole of the pin (stick 16 is not shown there).
[FIGURE 3 OMITTED]
At eccentric position of an external diameter [d.sub.1] of the neck
according to an internal hole [d.sub.2] the neutral axis x-x of inertia
moment of the neck cross-section will be displaced upward to value
[e.sub.1] from axis of symmetry of external diameter [d.sub.1] and to
value [e.sub.2] from the axis of symmetry of internal diameter
[d.sub.2]. Eccentricity of external diameter according to internal
diameter is equal e. At marking the area moment of inertia of external
diameter [d.sub.1] to upper side from axis x-x by letter [I.sub.1],
moment of inertia of internal diameter [d.sub.2] by [I.sub.2], moments
of inertia of cross-section [d.sub.1] and [d.sub.2] from axis x-x to
axis of symmetry of these cross-sections by [I.sub.3] and [I.sub.4]
accordingly, and area moments of inertia of lower halves of circles
[d.sub.1] and [d.sub.2] by [I.sub.5] and [I.sub.6], one can see that
common area moment of inertia of the first part of cross-section placed
over the axis x-x is equal [I.sub.c1] = [I.sub.1] - [I.sub.2], while of
the cross-section placed under the axis x-x is equal [I.sub.c2] =
[I.sub.3] + [I.sub.5] - [I.sub.4] - [I.sub.6]. Diameters [d.sub.1] and
[d.sub.2] and eccentricity e are known from the pin design. Eccentricity
[e.sub.2] is equal [e.sub.2] = [e.sub.1] + e, so there is left not known
eccentricity [e.sub.1]. Because
[I.sub.c1] = [I.sub.c2] (2)
by insertion of values from [I.sub.1] to [I.sub.6] into Eq. (1) it
is possible to find the value [e.sub.1], necessary for the calculation
of area moment of inertia of the neck cross-section.
It is possible to write that moments of inertia from [I.sub.1] to
[I.sub.6] are equal
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (3)
there [r.sub.1(2)] are radiuses of diameters [d.sub.1] or [d.sub.2]
respectively; [e.sub.1(2)] are their eccentricities according to axis
x-x.
Common area moment of inertia [I.sub.c] of the neck cross-section
according to axis x-x is equal
[I.sub.c] = [[pi]/4]([r.sup.4.sub.1] + [e.sub.1][r.sup.3.sub.1] +
2[e.sup.2.sub.1][r.sup.2.sub.1] - [r.sup.4.sub.2]
[e.sub.2][r.sup.3.sub.2] - 2[e.sup.2.sub.2][r.sup.2.sub.2]) (4)
Because the pin is loaded by the load going through the pin axis
and this axis coincides with the hole axis it is necessary to find not
coincidence of axial load going through [e.sub.2] with the neutral axis
x-x. Inertia moments from [I.sub.1] to [I.sub.4] are expressed by
complicated equations, for that reason value e1 was calculated by the
method of approximation.
The maximal strain [sigma] received in the neck of a pin and
angular deflection of the pin head are evaluated accordingly by the
equations
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (5)
there [l.sub.k] is the neck length.
Displacement [delta] at the point measured by transducer 17 will be
proportional to length ratio of measuring stick [l.sub.s] to neck length
[l.sub.k], it is [l.sub.s]/[l.sub.k]. Because [theta] is proportional to
[l.sub.k] and deflection measured by the transducer depends on [l.sub.k]
one can go to conclusion that length of the neck [l.sub.k] does not
influent on deflection measured by the transducer, but length [l.sub.s].
Table shows dependence between neck diameters [d.sub.1], [d.sub.2],
[e.sub.1], strain [sigma] at eccentricity e = 2 mm and load of the pin
with force [F.sub.0] = 26722 N, cross-section area A in [mm.sup.2], and
neck stiffness c, N/[micro]m; displacement [delta] is for load 1 N.
As it is seen, such diameters of the neck assure necessary strength
of the neck. Apart of that, modern sensors can measure displacements in
nanometer limits. In such a case the measurement method enables to
measure load on a pin 14 (Fig. 1) in limits of 1-10 N. At increase of
diameter [d.sub.1] measurement sensitivity decreases, for that reason
for proposed dimensions of the bearing Fig. 1 the neck diameters and
eccentricity e should be got of that kind.
Force [F.sub.0] shows the hydrodynamic force which acts on the pins
at journal rotation. If the journal is at center of the bearing, it is
from all pad directions is acted by the same force and is in equilibrium
state, the load carrying force ([F.sub.3]) comprises only at dislocation
of the journal from its equilibrium (centered) position to eccentricity
[e.sub.j]. This force for a three-pad journal bearing is expressed by
the equation
[F.sub.3] = [F.sub.0](1/[(1 - 0.5[chi]).sup.2] - 1/[(1 +
[chi]).sup.2]) (6)
or stiffness [c.sub.j] of lubricant hydrodynamic wedge film is
[c.sub.j] = [F.sub.3]/[e.sub.j], N/[micro]m. There [chi] =
2[e.sub.j]/[DELTA], where [DELTA] at this case is in [micro]m.
Fig. 4 shows the dependence of force [F.sub.3] on value [DELTA] at
the eccentricity [e.sub.j] = 1 [micro]m. It is seen that at smaller
values of [DELTA] carrying force of the bearing smartly increases. The
stiffness cc of the contact between the spherical pin head and the pad
sphere can be expressed by the equation
[c.sub.c] = 9.81[d.sup.2.sub.s]/16[k.sub.s] (7)
where [d.sub.s] is diameter of the sphere; [k.sub.s] is a
coefficient, [k.sub.s] = 0.5 [mm.sup.2][micro]m/N. Assuming [d.sub.s] =
24 mm, we obtain [c.sub.c] = 706 N/[micro]m.
[FIGURE 4 OMITTED]
The common stiffness [c.sub.com] of the pin head contact and the
head neck can be found from the equation 1/[c.sub.com] = [1/[c.sub.c]] +
[1/[c.sub.p]] and for the neck with [d.sub.1] = 16, [d.sub.2] = 11
because of its high stiffness the common stiffness leaves practically
the same, in the limits [c.sub.com] = 705 N/[micro]m. At not
automatically controlled bearing the third pin would also be of the same
stiffness. At load with the force [F.sub.0] = 26722 N elastic deflection
in mechanical contacts of the bearing would be 37.9 [micro]m. If to
assemble the bearing with such interference (that at revolving it could
work with the clearance 0.01 mm), after stoppage of rotation the pins
will remain prestressed with interference of 27.9 [micro]m, or the
interference force will be [F.sub.in] = 19670 N, the journal could not
start revolving for such interference. For that reason the bearing could
not work with the clearance of 0.01 mm at rotation.
Fig. 5 shows the minimal value of set clearance [DELTA] with which
the bearing at revolving of the journal could work keeping initial
interference force after stoppage (the same--at starting) equal zero
([F.sub.in] = 0). This force is defined by the equation
[F.sub.in] = [F.sub.0] - [c.sub.com][DELTA] = 0 (8)
[FIGURE 5 OMITTED]
The minimal clearance which is possible to achieve without initial
interference at not revolving journal is [DELTA] = 0.0156 mm. One must
keep in mind that this clearance must be kept at cold bearing. At its
heat, e.g. if temperature between the bearing body and the journal would
change to 1 [degrees]C, at the radius of external diameter of the body
equal to 90 mm, the clearance would change to 1.1 [micro]m. After
stoppage of heated system a new start of it would be with an
interference force. It influences hardly on the bearing wear.
The use of automatically controlled bearing enables to set
necessary clearance between the pads and the journal at its start and at
work and in such a way to achieve necessary working conditions of the
bearing. Research of the bearing with controlled work conditions would
give better illustration of its work accuracy dependencies (revolving
accuracy, damping and vibrations properties, etc.) than at
"blind" research, when the hydrodynamic force conditions in
the bearing are not known. Apart of that, it is possible to show, that
control of the clearance in a bearing at start, at idle run of the
journal and at work enables to control power losses and heat of the
bearing.
3. Conclusion
Analysis of the possibilities of a new design of hydrodynamic
tilting pad journal bearing with self-aligning pads shows that by
control of the clearance between the journal and pads it is possible to
keep in front defined conditions of the bearing operation, increase its
stiffness and at its research to receive the picture, better showing
dependencies of functioning quality of the bearing on its real state.
Received April 15, 2011
Accepted April 05, 2012
References
[1.] Marcinkevicius, A.H. 1995. Stabilization of Grinding Processes
and Machine Tool Work, Vilnius, Technika. 312 p. (in Lithuanian).
[2.] Hargreaves, D.J.; Fillon, M. 2007. Analysis of a tilting pad
journal bearing to avoid pad fluttering, Tribology International 4 (40):
607-612. http://dx.doi.org/10.1016/j.triboint.2005.11.019.
[3.] Reddy, D.S.K.; Swarnamani, S.; Prabhu, B.S. 2000.
Thermoelastohydrodynamic analysis of tilting pad journal bearing--theory
and experiments, Tribology Transactions 1(43): 82-90.
[4.] Hori, Y. 2006. Hydrodynamic Lubrication, Tokio, Springer
Verlag, 360 p.
[5.] Vekteris, V.; Cereska, A.: Jurevicius, M. 2010. Diagnostic
research of adaptive hydrodynamic segmental bearings with elastic ring
in a rotor system, Mechanika 2(82): 58-62.
[6.] Vekteris, V.; Moksin, V. 2010. The modular principle of design
of adaptive film lubrication bearings, Mechanika 4(84): 53-57.
[7.] Marcinkevicius, A.H. 2010. Multi-pad self aligning
automatically regulated bearing, Patent Application No 2010107, Vilnius,
6 p. (in Lithuanian).
[8.] Minchenia, V.T.; Stepanenko, D.A.; Zimmerman, C.K.;
Bansevicius, R.; Bubulis, A.; Kundu, P.; Gishkeliuk, I.A. 2010.
Simulation of vibrational piezoelectric actuators of the micro-robot
using the finite element method, JVE Journal of Vibroengineering 12(3):
314-319.
[9.] Patasiene, L.; Ragulskis, K.; Federavicius, A. 2010.
Development and experimental analysis of layered actuators, JVE Journal
of Vibroengineering 12(3): 333-339.
[10.] Marcinkevicius, A.H. 2010. Hydrodynamic tilting pad journal
bearing with automatic control, Warshaw, The Archive of Mechanical
Engineering 4(57): 343-354.
[11.] Reshetov, D.N. 1972. Parts and Mechanisms of Metal Cutting
Machine Tools, Vol. 2, Moscow, Maschi
A.H. Marcinkevicius
Vilnius Gediminas Technical University, Sauletekio al. 11, 10223
Vilnius, Lithuania, E-mail: andma@vgtu.lt
http://dx.doi.org/ 10.5755/j01.mech.18.2.1567
Table
Dependence of [e.sub.1], strain [sigma], and measurement displacement
on neck diameters
[d.sub.1], mm [d.sub.2], mm [e.sub.1], mm [sigma], MPa
16 11 3.19 261
15 10 2.84 283
14 9 2.48 310
[d.sub.1], mm [delta], [micro]m A, [mm.sup.2] c, N/[micro]m
16 4.8 x [10.sup.-4] 106 2.25 x [10.sup.7]
15 5.4 x [10.sup.-4] 98 2.43 x [10.sup.7]
14 6.1 x [10.sup.-4] 90 2.65 x [10.sup.7]