LAMINATED-RUBBER BEARINGS: FROM HELICOPTERS TO THE DEEP SEA
NEW POTENTIALS FOR UNDERSEA CONTROLS
William L. Hinks
Randolph
Research Co. (RRC)
Presented at UUST11,
(Note: This is the Powerpoint presentation with transcripted text, expanded beyond the original UUST11 paper. Viewing video content requires unblocked access to YouTube)
Hello!
I want to aquaint you with Laminated Rubber Bearings!
They've done well in the helicopter industry. What about the undersea world?
We developed laminated
rubber bearings originally for use on helicopters. But -- what ARE laminated
rubber bearings, and what do helicopters have to do with undersea operations, AUVs, ROVs and other
submersibles?
In brief, we've taken a
device used on helicopters, and developed it into a robust new bearing + seal
technology. We believe it can open new doors for undersea design applications.
I'll get to that in a
moment, but first, let me tell you a little more than you may want to know
about helicopters:
Everyone is familiar
with their rotor blades whirling around at hundreds of revolutions per minute.
This figure shows the central section of a two-bladed helicopter rotor as seen
from above. It rotates counter-clockwise, while the whole thing flies forward
(that is, up on the picture).
Those blades (in
yellow) are heavy, and they create a lot of centrifugal force, CF. Even a small
helicopter blade may develop a centrifugal force up to 10 tons or so, pulling
outward on the central hub.
More than that, each
blade's pitch angle is oscillated about its own longitudinal axis. It pitches
up and down, once per revolution of the rotor. For instance, the blade on the
right is held at a low pitch angle as rotation moves it toward the direction of
flight, while a blade on the left is held at a high pitch angle, to catch more
air on the "backstroke", as it were.
In the figure, the red
shaft at the root of each blade passes through and pulls on a thrust bearing
that connects it to the blue rotor hub. The thrust bearing has to support all
that centrifugal force, while allowing the pitch angle to be cycled during each
revolution.
In that environment,
and without frequent lubrication, ball or roller bearings tend to fail quickly,
due to spalling or "fretting corrosion".
Something new was needed.
At that point, our
innovation was to replace the rolling element bearings on the basis of a new
idea, laminated rubber bearings. These devices support high loads while
permitting angular oscillations for a long time, without lubrication. This is a
technology that’s now common in the helicopter industry, being used on many or
most helicopter designs since our original patent some years ago.
Here's a picture
representing a laminated rubber (or elastomeric) thrust bearing. It's shown
here with a vertical axis. In the helicopter case, that axis would be
horizontal, as we've just seen.
What is it? -- it's simply a stack of alternate rubber and metal washers,
all bonded together. The metal layers or washers are white, and the alternate
rubber layers are grey. The layers are greatly exaggerated in thickness for the
sake of clarity; in some designs, there may actually be 50 or more layers.
In service, this
bearing unit would be seated upon a sturdy base or housing. Tons of force are signified by the black arrow, acting downward upon
the top face - in the horizontal helicopter case, that would be the centrifugal
force from a rotor blade.
Obviously, the
laminated bearing stack must withstand and support this high compressive load.
How can it do that without squeezing the rubber out from between the metal
layers? The answer is that each rubber layer is held back by its bond to the
neighboring metal layers, and is relatively very thin, in some cases thin as a
sheet of paper. The metal layers can be equally thin.
It turns out that these
bearings can support very high loading pressures, such as 10,000 psi or more.
But the actual compressive deflection is very slight, since the rubber layers
can't squeeze out. Loading causes tensile stress in the metal layers, but with
high-tensile metals, that stress is not at all extreme. Safety factors are
quite reasonable despite high compression pressure.
But
what about the need for angular oscillatory motion?
In the illustration, a
vertical line drawn on the edge of the bearing stack in its rest position, becomes a slanted staircase as the whole bearing
is twisted either way. The lines are still vertical on the metal layers, but
slanted one way or the other in the rubber. So there is a distribution of the
angular movement - each thin rubber layer allows a small sideways shift or
shear strain due to its resilience. And the accumulation of
all those parallel motions add up to the total angular movement.
(Obviously, these bearings do NOT allow continuous rotation -- the rubber can
stretch to its limit, and that's it!)
A characteristic of the
rubber is that it provides a wide range of linearity between the shear strain
and the shear stress, over the working range. An opposing torque develops due
to the shear stress, and the individual strains add up to the overall angular
deflection. So, the torque is pretty much proportional to the angle of
deflection. In other words, it's a torsional spring,
and it has little frictional resistance or hysteresis.
Now, finally, here's
where we get down to the undersea action!
The Undersea
Bearing-Seal -- it's the same thing! - only
pressurized by seawater around it, and hydrostatic force on its top instead of
centrifugal force.
It turns out that that
these devices can actually function as hermetic seals! The bearing stack has a
central hole, and obviously, there's a circular body of solid rubber and metal
between the interior hole and the exterior of the bearing. It's
a barrier all around - seawater can't move from the outside to the inside, or
vice-versa.
Now, by making use of
this sealing property, we can provide a safe way for a shaft to penetrate the
hull of a submersible craft. With that, we could
control the pitch angle of an external hydrofoil from an internal actuator, for
instance.
"PENETRATE THE
With those words, some
of you may turn me off right now! You may consider such an idea as an impending
CATASTROPHE - ASKING for TROUBLE. And experiences of the past may well bear out
your concerns. I hope some facts about this new technology may lead you to
reconsider.
Here's what I'm talking
about: a bearing-seal as it might be used in an AUV. This cross-section represents the top of its
circular pressure hull. The hull is light gray in color, and the down-curving
edges of the hull are part of the complete circle. There's a gray inward bulge
of the hull at the center, and it encloses a cylindrical cavity or receptacle.
A light yellow flanged shaft passes through a bearing-seal within the
receptacle. The bearing-seal has white metal and black rubber layers and blue
top and bottom end parts. The end parts are sealed by O-rings - black spots - relative
to the shaft flange and the bottom of the receptacle.
The large arrows at the
top represent intense hydrostatic pressure, pushing down on the shaft and its
flange, and therefore upon the bearing-seal. Seawater also presses into the gap
around the sides of the flange and surrounds the bearing-seal. But seawater
can't get inside the bearing-seal nor thence into the
vessel. And as I said before, the shaft can still be oscillated.
The shaft is maintained
on center by two radial bearings. One (or maybe both) is a bushing or needle
bearing in the air environment below the bearing-seal. The other bushing may be
above the bearing-seal in seawater as shown: a low-friction Teflon bushing
that's red in color, supporting the shaft or its flange.
So what we have is a
bearing AND a hermetic seal, that permits limited
angular movement of a shaft that penetrates the hull of a subsea vessel. It's
under high force due to intense hydrostatic pressure.
What it has in common
with the helicopter application, then, is high force capability with limited
angular oscillatory movement, as well as the capacity for long life.
Let me give
you some examples of tests we've run on small bearing-seals. At the lower right
center of the picture is a hydraulic press, and the
shiny cylindrical pressure chamber that is centered in it. The hydraulic press
forces a piston into the cylinder of the pressure chamber, and thereby
pressurizes the seawater and the bearing-seal configuration inside. A shaft
passes through the internal bearing-seal and extends out into the air. A
vertical lever arm is attached to the end of the shaft, so it can be oscillated
back and forth.
The internal
configuration is just as shown in the previous AUV figure, only inside out, if
you will. ABOVE the bearing-seal, in the seawater, would be the INSIDE of the
pressure chamber, and BELOW it, the open AIR. In both this figure and the
pressure chamber, the outside of the bearing-seal sees pressurized seawater,
while the inside sees air. In both cases, the hydrostatic pressure pushes down
on the shaft flange, and the shaft extends axially through the bearing-seal and
into the air environment.
We ran such a pressure
test on a very small bearing-seal, only 17 mm OD, about the size of a dime,
with a 6 mm hole. It was oscillated +/- 15 degrees at 16,000 psi for many hours.
In the video below, the pressure gage is almost maxed-out, indicating that the internal seawater
pressure is about 16,000 psi. Start the video to tweak the lever arm, and you can see the resilient action of the small bearing-seal, even under all that pressure. It's free and springy! Pressure does NOT affect it.
Here's another thrust
bearing-seal that we tested, It has an outside
diameter of 2.06 inches, an ID of 1 inch, and about 1/2 inch height. It has a
steel center plate and end plates. On each side of its center plate, it has 40
laminations of high-tensile cartridge brass, with 39 intervening rubber layers,
to make a total of 80 brass layers. Both the rubber and the brass layers are
only .002 of an inch thick.
This bearing-seal was
placed into the tester under 10,000 psi pressure as will be seen on the pressure gage, and was linked up to a
crankshaft. Starting the video, it runs at an accelerated speed of about 600 cycles per minute.
While under that 10,000
psi pressure, it was oscillated at +/-15 degrees for over a million cycles
without failure. The lever arm could also be easily moved by hand through the
same angle. Some tests on that unit later went up to the crushing pressure of
17,000 psi -maximum sea depth, under continuous oscillation.
So a shaft penetrating
a hull using this new technology is NOT comparable with a delicate
Teflon/rubber seal, rubbing on a polished shaft. There is no rubbing or sliding
action anywhere or any corrodable surfaces. And
tearing or damage from incursions of sand or whatnot can't happen. And unlike
the friction torque on an ordinary sliding seal, the bearing-seal does NOT
become stiffer with added depth; torque vs. angle remains essentially
unchanged, as you saw in the previous video. And depending upon the design,
more than a million cycles are possible.
Now that you have an
idea of what these things can do, let me summarize so far, and discuss some of
the design factors that we can deal with.
The various factors,
including the dimensions, hydrostatic pressure, torsional
stiffness, shear strain, fatigue life, and angular range can all be balanced to
achieve various end results. There are numerous relationships between these
factors. For instance, torsional stiffness is
inversely proportional to the total height of the laminate stack, and fatigue life
is dependent upon shear strain, pressure and other factors.
But all the discussion
so far represents only the simplest type of bearing-seal. They all had flat
metal and rubber layers. Instead of being planar, the layers may instead have
conical or hemispherical shapes, or may be curved to form a cylinder or other
shape. Such shaped bearing-seals can then be capable of supporting radial, or
combined radial and thrust loads, while permitting angular movement about other
axes.
For instance, in this
configuration, conical bearing-seals are employed, shaped like lamp shades. But
there are two of them, with the central flange of the shaft sandwiched between
their blue conforming end members.
The top bearing-seal
mirrors the features of the bottom unit. It's loaded by the hydrostatic
environment like a piston, pushing down into the cylindrical housing. There are
O-rings to seal the bearing-seal ends between the housing cylinder and the
shaft. A Bellville spring on top contains the assembly when it's not
pressurized.
The previous figure
used a single bearing-seal with the high pressure on its periphery. But in this
case, the top bearing-seal will instead experience the high pressure in its
aperture. Adding the top bearing-seal makes it the primary seal, with a backup
by the bottom bearing-seal. The top seal prevents fluid pressure from acting
directly upon the bottom unit.
But if the top were to
fail for any reason, the bottom bearing-seal would become the primary seal. It
would function in the same manner and with the same loads as the single
bearing-seal seen before. If even more backup was desired, an O-ring could be
used on the bottom of the shaft as shown.
Generally speaking,
potential applications can be found where it's advantageous to do control and
actuation in an air environment. It becomes possible to use non-esoteric, low
cost, off-shelf electric actuators, without draggy
external bulk nor concern about seawater corrosion.
Now, let's talk about
the fishtail application. While a clear need for the fishtail mode of
propulsion and control may or may not have been demonstrated, there has been
the promise of superior maneuverability, efficiency and stealth of underwater
propulsion that mimics the natural motions of fish.
A number of
investigators and institutions have pursued theoretical studies of fish
propulsion, and they have revealed some useful dynamic models. Also, working
prototypes of undersea vehicles have been developed based upon this
understanding. Some of these institutions are MIT, Olin, and Boston Engineering
Corporation, and participation and funding has been provided by several US Navy
agencies. (I must apologize for my omission of the important work earlier presented in this
symposium.)
One way of providing
this kind of locomotion is shown in this conceptual AUV. (Unlike the earlier
papers, there were no hydrodynamic or controls studies made to support this or
the next configuration.)
A stub shaft
penetrating vertically near the aft end of the hull is supported and sealed by
a bearing-seal configuration as we've seen before, either the single or double
type. An external lever arm is attached at right angles to that vertical stub
shaft. The lever arm extends aft beyond the pressure vessel, to end in a compliant
fin.
The tail fin is rapidly
swiveled back and forth in a horizontal arc for propulsion. The pivoting shaft
might be driven continuously by an internal actuator such as a motor-driven
crankshaft and connecting rod. Or it might be used with deliberate motions, as
a fish does, for directional control or acceleration.
Also, I've indicated
lateral hydrofoils on each side, acting as pectoral fins for vertical control.
They would be internally actuated in pitch, through laterally-oriented
bearing-seals.
Here's another possible
implementation of tailfin propulsion. It would use spherical bearing-seals.
These specialized bearing-seals support and seal a yellow shaft that extends
longitudinally from the aft body of the AUV. The extended and moveable shaft is
tipped by black tailfins for propulsion, and is surrounded at the junction with
the aft body by an orange compliant rubber fairing for smooth hydrodynamic
flow.
The details of that aft
bearing-seal configuration are shown in this cross-section. It has two opposed
spherical bearing-seals. They surround, support and seal a ball-like
enlargement that's located mid-length on the shaft.
You can see that the
shaft can move angularly, not only twisting about its longitudinal axis, but
can also teeter up and down within the plane of the picture, as well as back
and forth out of the picture plane. So it has three degrees of angular freedom.
Depending upon the
internal actuating mechanisms, any or all of the three oscillatory degrees of
freedom could be employed. Among other possible configurations, a compliant
cruciform tailfin as shown could permit vertical as well as horizontal flapping
motions for propulsion. It could also be used for pitch and direction control
of the craft. Maybe even oscillatory twisting motions of the shaft, or is that
getting a little too far-fetched?
The side pectoral fins
in this case might also be implemented with spherical bearing-seals. All 3
degrees of freedom could be used. With clever internal actuation, that would
make it possible to provide flapping or rowing motions of the pectoral fins,
including control of their pitch angle -- everything a real fish could do with
them?
Well, these examples
wrap up my case. Our underwater application of laminated rubber bearings is
still undergoing development, particularly the conical and spherical
configurations. Although these applications
have yet to be proven, they are meant to suggest some of the opportunities that
can be opened up by this new approach to design. The technology provides
simplicity, low cost, and reliability for vehicles that probe the depths of the
sea.
Thank you -- and I hope
I've changed some minds!