Pump cuts friction by eliminating side loads

Figure 1 — Line drawing of a singlepiston Sanderson Rocker-Arm Mechanism illustrates the method of turning rotation into reciprocation. Additional pistons could be positioned in a circle, still using only one u-joint, one nose pin and one crank would be required.

Figure 2 — Hydraulic motor starting torque demonstrated with weight-lifting tests. Linear best fit relationship shows no apparent drag at low pressure.

Figure 3 — Hydraulic motor free running with no apparent frictional load since data curve passes through zero flow at zero stroke.

The application of a recent invention, the Sanderson Rocker-Arm Mechanism (SRAM), reportedly holds promise to improve efficiency of hydraulic pumps and motors, engines, compressors, and other positive displacement pumping devices. The mechanism uses a new principle to efficiently convert reciprocating motion to rotating motion, and vice versa. The low friction exhibited in hydraulic devices holds promise for high overall efficiency, low heat generation, and virtually zero stick-slip operation (stiction).

Description in a nutshell
A simplified drawing of a singlepiston SRAM is shown in Figure 1. A piston is attached to one arm of a 90° rocker arm through a joint that only transmits force parallel to the piston axis to reach the piston. Side load forces on the piston, therefore, are zero by definition. The center point of the rocker arm is attached to a conventional universal joint grounded on one side. This ground supports all output torque — none is transmitted to the pistons.

The other end of the rocker arm, which carries a nose pin, has to follow the reciprocating motion of the piston as projected on the reciprocating plane. However, the universal-joint also allows the nose pin to move at right angles to the reciprocating plane. The nose pin can trace a circle, and, therefore, can be connected to a crank on the output shaft to complete the mechanism.

This description applies to one piston, but as many pistons can be arrayed into a circle as will physically fit. They all share the same universal joint and nose pin, but each has its own 90° rocker arm and piston joint at its own phase angle with respect to the first piston. Plus, each piston can drive two pumping chambers — one at each end — which doubles the work capability without increasing the working diameter of the SRAM assembly.

Like many hydraulic mechanisms, the SRAM can exhibit variable stroke to zero with a single-point mechanical control. The stroke varies about its midpoint in the simplest version, which can drive both single-and double-acting configurations.

The radial position of the nose pin determines the stroke of the SRAM. Placing a cam face radially between the nose pin and the case allows moving the nose pin in and out radially to vary stroke of the piston(s). Because radial position of the nose pin varies piston stroke, displacement can be varied within a single revolution. This arrangement can be configured such that the flow or torque ripple from the piston(s) and the universal joint cancel each other out.

Advantages of the mechanism
The biggest advantage of the SRAM is the near elimination of friction and stiction, according to Albert Sanderson, vice president of R&D, Sanderson Engine Corp., Upton, Mass. The number of journal bearings in the SRAM is five, two in the Cardan-type universal joint, one at the nose pin, and two on the shaft that supports the crank arm. This is true regardless of number of pistons, or whether single-or double-acting cylinders are used.

Sanderson explains, "We don't as yet know the limiting factors, but the mechanism itself seems to be as scalable as the crankshaft, for instance. Applications that recirculate hydraulic fluid are certainly going to generate heat at lower power levels, but based on projections of our 50-hp hydraulic motor, SRAM pumps and motors of 2000 hp or more seem practical." He estimates that the SRAM hydraulic pump would require roughly 30% less input power than to produce the same output as a conventional pump. Doing so would not only reduce initial cost of a prime mover, but energy consumption for the life of the drive.

He adds that the motion of Sanderson pistons is sinusoidal for small angular swings of the rocker arm, which produces low torque ripple in motors and pressure pulsations in pumps. He also says that by eliminating frictional losses of sliding contacts at cylinder walls and bearing pads, the SRAM hydraulic pump heats up only a few degrees under full load.

Low-speed performance
Sanderson says starting torque efficiency of a SRAM will be well over 90%, with no stiction. For example, he explains that the torque required to rotate the shaft of a 50-hp hydraulic motor on a test bench with was less than 3 lb-ft over a full rotation of the shaft. In Figure 2, the plot of torque required to lift given weights from a static start passes effectively through zero when replaced with a best-fit straight line.

Under the no-load conditions illustrated in Figure 3, the motor operated with slight cogging down to 50 rpm. The ability to smoothly run at low speed would extend the speed range of vehicles without having to downshift gears.

"During the tests of our hydraulic motor as a pump," says Sanderson, "volumetric efficiencies to 98% were observed at pressures below 500 psi, and they stayed above 90% up to 2000 psi. Case leakage was low to negligible from 25% to 95% stroke and from 0 to 2000 rpm. These figures closely match the best performance available today mainly because no basic changes in the pistons, cylinders, or valve-plate configurations are necessary."

Sanderson concludes, "We are still in the testing stage with this technology, and our test results need to be extended to measure torque input for pumps and torque output for running motors, both numbers essential for calculating torque efficiency and overall efficiency.

"Many engineers are asking for one-rpm measurements of torque and pressure. We are actively seeking test labs that can make these particular measurements, but not many exist in this country. In spite of that, the results from the measurements we have made so far show the ability of the Sanderson Rocker-Arm Mechanism to overcome some of the limitations of current technology. So we look forward to more progress and to future performance reports within the next six to twelve months."

This information was provided by Albert E. Sanderson and Robert A. Sanderson, vice president of engineering, Sanderson Engine Development Corp. To view or download the complete manuscript from which this article was excerpted, visit www.hydraulicspneumatics.com. For additional information, e-mail sandersonengine@cs.com.