I tend to be overly complicated and thus often end up doing things the hard way ...
Tom Kasmer wrote:Hi Ciro, I think you have done a really great job with your writings about the Hydristor. I would be pleased to work with you and explain the magic of the Hydristor. The Hydristor is a machine which converts shaft rotation into variable hydraulic delivery of fluid volume. Think of a hydraulic cylinder pushing a load when hydraulic pressure is applied to the push piston in the cylinder.
Now, imagine a second cylinder plumbed to the first cylinder via a pipe. The second cylinder has a hand lever to push on it's piston. Lets add a sliding tube over the hand lever so that the effective length of the hand lever can be varied, say 10:1 in length. Assume both cylinders have the same diameter of piston.
Case 1: slide the hand lever to max length and push down on the end to cause the second cylinder piston to push in and expel fluid to the second cylinder. The force exerted on the end of the fully extended lever is magnified by 10 times and that 10X force appears at the first cylinder's push output. The hand lever motion is 10X the motion seen at the load.
Case 2: slide the hand lever tube to the 1X minimun extreme. The force applied to the hand lever is directly applied to the piston of cylinder #2 and thence to the output load.
To recap, case 1 sounds like a mechanical lever with a 10:1 advantage and case 2 sounds like directly doing the work. One difference is that I can remotely locate the result of doing the work because I am using hydraulic pipes to transmit the work from point 'A' to point 'B'.
There is another subtle difference here. Using a mechanical lever, I can actually create an infinite ratio. That is done by placing the fulcrum exactly at the point of the load. No matter how much you stroke the lever, no force is required and no load motion ensues.
If you were to move the fulcrum 1/8 inch off the load position, and the lever was 12 inches long, the actual ratio would be 8x12=96:1. A 1 pound force applied to the lever end would create a load force of 96 pounds and a 1 foot stroke (ignore the geometry discussion) would produce a 1/8 inch load motion. If the fulcrum location were 1/16 inch,
the 1 foot and 1 pound of force would equal 192 pounds of load force but the output motion is cut in half to 1/16 of an inch. The point of this discussion is that the hydraulic system described is an imperfect lever and fulcrum as described.
Now, let's consider a mechanical, ratcheting bumper jack and let's add a telescoping handle to this. Each time I stroke the handle to raise the bumper load, the detent clicks in. I can then re-stroke the lever another click to get another bit of bumper height. If I stroke away with a repetitive rhythm, the bumper seems to magically rise above the road. If I were to extend the length of the lever to maximum, the stroking force would be less, but my level of work expended would be the same. It is definitely easier to stroke the longer lever since the force required places a workload on my hand which has it's own force limits. The Hydristor solves this entire dilemma.
In the animation I saw on the website, there is a hydro-mechanical force developed in the direction of the low pressure in both cases. The amount of force is equal to the pressure times the equivalent area of the rotor and vanes viewed along the vertical axis. The original vane pumps were 2 chamber pumps and developed huge side forces on the rotor and vanes and then to the rotor support bearings. The housings had to support these unbalanced forces as well. That design could be easily varied by designs similar in principle to your animation. A 5 Hp variable vane pump in that early design might weigh 300 pounds due to all the metal required to hold it together. Bearing life was short. In 1925, Harry F Vickers invented the dual, pressure balanced' pump which had a modified elliptical chamber (cam ring) to guide the vanes in and out twice per revolution for a total of 4 radial vane motions per revolution. There needs to be a 'kidney port' between the high and low vane motion extremes. The rotational space between any 2 kidney ports must be slightly greater than the rotational space between any two adjacent vanes on the rotor. This prevents oil from bypassing and guarantees that the displacement of oil is directly related to the rotation of the rotor. This design solved the awesome bearing load problem by creating a diametrically opposite and equal force to that of any chamber; hence the name 'dual pressure balanced'. This also created two separate fluidic common but hydraulically separate pumps within the same housing. The rotating vanes have a small but non-trivial mass and the centripetal force of the vanes is speed squared dependent. This is a significant part of the historical inefficiency of the Vickers pump and that number is around 80%. The lifetime according to a Vickers (the corporation) chart is typically 10,000 hours with a 5 micron filter when operating within speed and pressure limits. The pressure is around 2,000-2,500 at speeds up to 5,000 Rpm.
The two thumbnail sketches below your animation show the Hydristor basics of rotor, vanes, and pistons in the two extreemes of piston motion. In the Hydristor, the historical vane tip friction contacting the underside of the belt causes the belt to rotate at near the rotor speed. There is a 'walking mechanism' of contact of each vane acting against the belt where the angularity of force changes 8 times per revolution and each change causes the vane tip friction to 'slip' very slightly. The result is that the belt position relative to the rotor slips slightly behind the rotor speed and the belt life is extended due to the vane friction contact point always moving and no wear spot is generated. The belt also confines the centripetal forces collectively of all vanes. The Hydristor operating speed is very much higher in theory and I imagine designing the flow paths like porting and polishing a racecar intake manifold.
Lets go back to the rotating belt. The belt is always immersed in oil. The contact shape of each piston is not a simple curve. The shape is designed to create a 'rotating wedge' of oil continually trapped between the piston curvature and the belt curvature. This rotating wedge is continually squeezed down to a minute thickness forming a self replenishing
hydrodynamic bearing during rotation. The minimizing effect also creates a self replenishing oil seal which prevents oil under pressure from bypassing the rotation mechanics. The hydrodynamic bearing prevents
direct contact of mechanical friction and this is like a connecting rod bearing in an engine. The first Hydristor was tested at Tecumseh test lab in Ann Arbor, Michigan and achieved 95% overall efficiency. That is
higher that an axial piston pump.
Now, let’s consider how the Hydristor varies oil delivery. If each of the 4 pistons is located equidistant from the rotation center AND is slightly in contact with the rotating belt, there is no net output or input of oil from any of the 4 kidney ports during any range of rotating speed. If the trapped wedge of oil between the rotor diameter, the two adjacent vane surfaces, the belt underside and the 2 contacting flat ends of the housing (the 4 'sealing areas' located between the 4 kidney ports at either end) is, say 1 cubic inch for the sake of discussion, and there are 4 such wedges simultaneously moving through the 4 piston sealing areas, then there is no net exchange of oil through any of the 4 kidney ports.
Now, let’s move the 4 piston positions simultaneously. The 12:00 and 6:00 o'clock pistons move out by .001 inch AND the 3:00 and 9:00 O'clock pistons move in equally by .001 inch. Assume the rotor length is 1 inch for the discussion. Also, assume the pistons/belt contact interface diameter is 2 inches. Looking at the net input of the oil wedge rotating under 12:00, the vanes are extended out .001 compared to -.001 at the 3:00 position for a net of .002 inches. This occurs at a radius of gyration of 1 inch and the axial length is also 1 inch. The area patch created in the direction of rotation is .002 times 1 inch axial length or .002 square inch. For one complete revolution, the 'extrusion of this area patch produces a volume of: (2 pi)(r)(.002) or (6.2832)(.002) = 0.0125664 cubic inch for 1 revolution and for one port from 12:00 to 3:00. The exact same volume of oil is simultaneously extruded from port 3 between 6:00 and 9:00.
Think of a window shade 1 inch wide, .002 thick and unrolling a length of 6.2832, and two of them at the same time. Normally, chambers 1 and 3 will be connected to one output and chambers 2 and 4 will be separately connected to the second output simulating a single pump even though the Hydristor is really a double pump. Obviously, doubling the piston motions will double the volume of the extrusions so that is linear. If you had manipulated the pistons oppositely; that is: moved 12 and 6 IN while moving 3 and 9 OUT, all oil flows would have reversed for the same (assumed clockwise) shaft direction and the linear relationship of outputs versus piston position would have held. If a chamber is outputting oil in a certain amount, the rotationally previous chamber has to supply it.
Theoretically, the output of chamber 1 is equal to the input of chamber 2, to the molecule! Same for chambers 3 and 4. So you have two fluidic separate circuits. Let’s connect the chamber 1 output to the input of a fixed displacement hydraulic right side wheel motor whose return is connected to Hydristor chamber 2. Also, connect chamber 3 Hydristor output to the input of an identical fixed left side wheel motor with it's return connected to Hydristor chamber 4. You now have a hydrostatic transaxle perfect for straight line forward/reverse motion. But how do I go around corners??? Simple with the Hydristor! Let’s go back to the equidistant piston case (called neutral). If I move 12 and 6 out while moving 3 and 9 in, I go forward. If I move the pistons opposite, I back up; all hydrostatically! (Let’s go) back to the circle case. Let’ hold 12 and 6 at the neutral position, and then move 3 to the left and simultaneously move 9 also to the left. Wow, the oil in the right circuit causes the right side axle to go forward. However, the left side circuit oil direction is REVERSED!
This causes the left axle to BACK UP! If you connect the right side
fixed motor to an additional front motor in hydraulic series and separately do the same for the left side, you have the basics for a skid steer machine or an off road 4WD that is really 'trick'. This behavior can be seen in the IFPE video of the John Deere tractor video on the Hydristor website. Now, suppose you simultaneously 'left shift' 3 and 9 and simultaneously 'ellipticise' all 4 pistons to go forward. The tendency to back up the left side is subtracted from the forward motion of diametrically moving the 4 control pistons. The tendency of the right side is added to the forward motion. What happens is that the left side turns specifically slower and the right side turns specifically faster depending on the mechanical placement of the 4 control pistons. It is as if you put a large tire on the right and a smaller tire on the left with a solid axle. With the Hydristor dual differential hydrostatic drive, all differential possibilities are possible.
This is the World's first (and only) differential drive and it gives the same drawbar pull going around turns as a solid axle going straight ahead. In regard to 4WD, if the front tire is smaller than the rear tire, you use a commensurately smaller fixed front motor to turn faster so that the tire tread motion is equal front and rear.
That’s good for a start. Let me know what you think and ask any questions. (Best) regards.
Users browsing this forum: CCBot [Bot] and 5 guests