Question and Answer;
Is it patented?
An international patent application has been submitted and is pending. A positive search report has been issued by the Eurpoean Patent Office.
Have you built a prototype?
No. Our main goal at this stage is to license or sell the rights to the invention in order to facilitate its development.
Is there a theory behind non-circular gears?
Non-circular gears are designed under the same principles as modern circular gears. They have been known for almost the same time, but their use has been very limited, most probably due to their difficult manufacturing before CAM systems were invented. The most used have been elliptic gears. You can check for yourself the site of this Chinese company which specializes in manufacturing non-circular gears to see some amazing setups in motion.
With so many gears meshing with one another, there will be big friction losses.
There are many gears, but many of them are redundant. They’re built into the system to balance the forces. The gear oscillator should not have more friction losses than a typical epicyclic gear train. In an epicyclic gear train, the power transmission is effected in two steps: from the ring gear to the planet gears and from the planet gears to the sun gear. In the gear oscillator, there are also two steps: from one ring gear to the planet gears and from the planet gears to the other ring gear. Furthermore, the number of planet gears should have no negative effect on the overall friction losses.
In a sense, this is similar to an electrical circuit where some resistances are placed in parallel. The equivalent resistance is smaller, not bigger.
In the gear oscillator, or in an epicyclic gear train for that matter, the planet gears also work in parallel, meaning that the power is transmitted along multiple parallel paths, and although this is not to imply that the friction losses will be much smaller, they certainly should not be bigger.
A reasonable estimate of efficiency in conventional meshes is 1-2% loss per mesh. We don’t expect the non-circular gears to behave very differently.
The cylinders on the main page design will be diffficult to build to precision and difficult to seal.
As regards the sealing problem, it can be basically solved using cylinder sleeves. As regards the precision problem, it can be greatly reduced if the position of the sleeves relative to the block or of the pistons relative to the rotor can be adjusted. These are details that must be worked out in the engineering process.
In these types of gear sets, there’s a set of teeth that always takes the load from the combustion pulses in engines. This teeth will be prone to failure and will suffer degradation problems.
This statement must be qualified. Suppose that a gear oscillator machine is working at a steady speed and load. Due to the flywheel effect, the load is spread throughout the whole cycle and thus throughout many different teeth. This is the main reason why this effect is important. Still, it’s true the load is uneven and that some teeth will sustain a higher load than others on a permanent basis. This is an inherent trade-off of this mechanism. Of course, wear can be minimized with a fine lubrication as in any other gear set. It’s worth noting that this is the only real source of wear in the whole mechanism, so the lubrication system can be focused on it.
During a planet gear revolution, the loads on its teeth change from one side of the teeth to the other. This will cause dynamic impacts on the teeth and noise because of the backlash/tolerance that the mesh must always have.
The gears must be built with minimum backlash to begin with. This can be said of any two engaging parts so it’s hardly an added requirement. The oil or grease film that will form on the flanks of the teeth will partially absorb the shock. If it’s not enough, we can resort to spring-loading the planet gear axles to push each planet gear pair against the ring gears. There are no forces between the gears when the loads reverse, and the centrifugal forces due to the rotating planet carrier will help push the planet gears radially outwards, so the stiffness of the spring can be reduced.
In the flywheel effect page, the parameters of the rotor are assumed to be such that they maximize the effect in a particular mode of operation. The operation is not so optimal in the other modes.
This is true to a large extent, although we must bear in mind that the moment of inertia can be changed dynamically, in the same way that a skater changes dynamically her moment of inertia when she pulls in her arms and legs to spin faster. Notwithstanding this subtelties, many applications require that the compressor or engine work at a given rate and can be optimized at that rate. This is the case of some hybrid cars, for instance.
If inertial forces aren’t so bad, why do engineers try to minimize them in conventional engines?
Let’s take an upright straight cylinder. The inertial force they try to minimize is the vertical one. There’s also the horizontal one (i.e. the horizontal component of the inertial force) that acts on the cylinder walls and adds or subtracts to the side-thrust due to the expanding/compressing gas. This horizontal force creates the counter-torque that always accompanies the output torque and that must be sustained by the chassis. You just can’t eliminate this counter-torque in IC engines, so it makes little sense to balance this horizontal inertial force. That being said, in a 4-stoke, every piston moves “idly” during the exhaust and intake strokes, but the inertial forces of a piston and conrod still produce stresses on its bearings and on the crankshaft that wouldn’t exist if the piston was massless. So it still makes sense to reduce the piston-conrod mass. All these drawbacks aren’t present in the gear oscillator: there’s just an inertial torque that always adds/subtracts to the output torque in a quite convenient way. There are no “idling” strokes because the pistons are double acting and anyway, they are solidly fixed to the rotor.
Can helical gears substitute for spur gears?
Not only they can, but they would be the preferred choice. They are not drawn in the accompanying animations for the sake of clarity and because they are more difficult to model than the spur gear type.
You say that more pistons can be added, but the more pistons that are used, the shorter the stroke, and this will limit the efficiency in an engine. Isn’t it so?
No it isn’t. You can add as many pistons as you want and still keep a square cylinder (stroke equal to bore) and the same volume per cylinder. You just get a bigger “doughnut”. What would be reasonable indeed is a gear train configuration with more pairs of planet gears to transmit the increased power.
At one point you say that to increase power density the engine must operate at a higher speed, and then at another point you say that to increase efficiency it must operate at a lower speed. Isn’t there a contridiction?
Yes and no. To get better power density, one must increase the rpms but keep the mean piston speed constant. This should not affect efficiency in a significant way. To increase efficiency, we think that an engine/compressor must work at a low piston speed and this certainly would decrease the output, so it’s basically a trade-off.
This design doesn’t have flywheels. Aren’t they needed?
The planet carriers can be designed to double as flywheels, but the need for flywheels is reduced due to the flywheel effect of the rotor.
The unions between the planet carriers and the planet gears’ axles seem too weak.
Again, the design shown is for illustration purposes. The planet carriers are drawn intentionally small to best allow seeing the operation of the mechanism. In a real world device, they can extend even beyond the planet gears bounds.
How do you plan to set up the valve train in an engine?
The well-trodden path to go would be the same as radial engines. If small cylinders are used, two valves per cylinder shall work fine, so the total number of valves may not increase compared to a conventional engine. However, we feel that there’s also the chance that some sort of sliding valves might work fine. If so, that would simplify greatly their operation.
Efficiency is the ultimate goal in a world of increasingly costly energy supplies, so it’s crucial to measure the gear oscillator system against a conventional one. Of course we can expect to increase efficiency due to reduced internal friction and reduced vibrations, but the question here is: can the gear oscillator increase efficiency in a more meaningful way? To answer this, let’s consider its two main applications separately.
Reciprocating piston compressors are almost the only ones used in high-pressure processes and in special applications like gastight diaphragm compressors. Therefore, it’s important to devise ways to improve their efficiency without fundamentally changing the piston-cylinder embodiment. This is where the gear oscillator comes into play.
From thermodynamics, we know that the most efficient way to compress a certain amount of gas in a cylinder is to do it in an isotherm and reversible way.
In an isothermal compression the gas is kept at a constant temperature. It’s clear that a slow piston speed facilitates the transfer of heat from the gas to the sorroundings and reduces the amount of energy needed to compress the gas. Certainly, some compressors mix the gas with oil or other lubricants to absorb this heat, but this lubricant has to be removed after the process adding costly filters and reducing again overall efficiency. Another approach, for which the gear oscillator is well suited, is to break the process into several stages, each one with a low pressure ratio (discharge pressure / inlet pressure) an intercooling, thus approximating an isothermal process. This also increases volumetric efficiency. The following graph shows the savings potential of isothermal compression in a reversible process when compared to an adiabatic one (worst case).