I've heard the Hatch dyno story. (I've heard additional details, too, that kinda question the validity of their assumptions as to cause.)
As to the RWS drives: there are numerous copies that have run for years on 3-rotor 20Bs, one of which was P-ported (likely north of 300 HP) and flying on a Lancair ES. That drive went through some incredible abuse during the install's 'teething' stage, and the only real damage was to the aluminum adapter plate that carries the 'damper' rubbers, and, IIRC, the screws that secured the stationary part of the planetary set to the housing.
I do get it that all engines make pulses. But is it just 'marketing' that a V-8's pulses are easier to tame than a 4 cyl engine? Isn't it easier to tame even if rpm is adjusted to the same number of pulses per minute? Does the peak-to-average torque ratio matter at all?
You're certainly much more knowledgeable than most of us here (certainly than me) on this subject. So I'm hoping you can help me understand why the rotary's 2-per-rev is worse than a flat 4's 2-per-rev.
Thanks for helping me understand what's happening.
Charlie
I knew Ev Hatch. The "whys" of the issues were just out of his comprehension for a while. He may have mis-interpreted the sources of his problems, but in the end he cam up with two very successful solutions. He understood it all well enough to both scheme out a pendulum damper that worked and a stiff system that worked.
Pendulum dampers are one of those things that has to be pretty much right on the money for tuned order and be able to store enough energy so that it can work to settle down the engine after transients. Miss on order by 4% or short on energy storage, and your data looks the same as if it were not there. Want to make it suck up fiing pulses really well? Gotta hit it within about 1-2%. And a stiff system? All the vibrating torque of the output goes through the gears and is reacted in the bearings. Shortfalls here bite too. He got solutions.
It IS just marketing to tell people "And this engine never goes negative torque". They ALL have torsional vibe at firing order.
Is it easier to tame a V-8 over a I-4? Define easier:
To an engineer practiced in the art, they require the same amount of work. You must either be given transmitted vibe targets or come up with them yourself, come up with estimates of the rpm range, estimate the spring rate to isolate to required level of transmitted vibration, estimate the spring travel to absorb the firing accel and re-emit it during the compression decel, find the right package of how many and how big springs with all the rest of the hardware, then package it all in the space you have. If you come up short of space or high on cost, you have to go back to the bosses and get them to free some space or say we can compromise on the transmitted vibe somewhere;
To a person guessing at what they need to make it all work, one may succeed right away and the other may fail, requiring many cycles of changing design features before a solution is achieved.
A three rotor has 50% higher pulse train frequency at any given rpm, and less ripple. Is that easier? I can calculate the changes required of the springs and shafts and gear systems. The damper has two big design criteria. First is high enough torque capacity and second is low enough spring rate. Shafts and gear also must have high enough torque capacity. Assuming a successful 2 rotor, the three rotor has 1.5 times as much mean torque but a little less ripple, and the three rotor makes that ripple 1.5 times faster. So roughly 1.5 times as much torque and only allow 1.22 times the two rotor's spring rate. That is a substantial increase in total stored energy for the soft element - go short and the soft element will be overloaded/overheated/bottomed or otherwise have either short damper life or hammer the gears to death. Then the gearset has to have plenty of capacity to go 1.5 times the torque of the two rotor. If the existing system had that much headroom, terrific. If it did not, they might have to find some room and build stouter parts that store more energy.
Let's go through an example:
I have four of these rubber donuts in my two rotor (or I-4) and I know it is OK at 150 ft-lb but not at 180 ft-lb. My new three-rotor (or V-6) is 225 ft-lb, so I need six of the same donuts to handle the torque without turning them to powder. If this is tried and the spring rate of the donuts was just low enough for the two rotor, the new spring rate is 1.5 times as high and it can only afford to be 1.22 times as high - it will now not isolate well, and is likely to also overheat and turn them to dust, if the gears stay together long enough and the prop does not fly apart due to all the hammering. All three are possible and in a race to see which one dies first. So you need 150% more total capacity with a spring at less than 122% more spring rate, which means more energy storage in the springs. I hope you have space for bigger donuts? Same with coil springs or giubos...
Now maybe your only problem area is at the low rpm end of things. Could you solve it by speccing the idle speed up 25%? Well, let's see what the tails are: If you have the same prop, you will have 56% more idle thrust that makes taxi and approach more challenging, but that bigger engine is likely to need 50% more prop blade area, so that will actually mean 134% more idle thrust - Oh, nobody's gonna buy that... But that larger prop will have more inertia, so that will help a few percent on low rpm isolation. Well it is something,but not much. Then there is the fact that once you get to firing speed (while cranking), the engine does have to accelerate itself and that prop through a now higher resonant rpm before it can get to idle. I got news for you, if resonance rpm is higher, it takes longer to go through the resonance, giving the system time to be captured at resonance rpm where its vibration is growing without an upper bound and can break or fatigue things before it finally goes on up to normal idle speed. Yeah, you really do have to design the damper for your inertia at both ends and to give resonance neatly between firing rpm (during start) and nominal idle speed, so it can accelerate through it quickly.
And we have not even gotten into failure mode management yet. Any cylinder or rotor or maybe even bank of cylinders can stop firing, giving you additional large torsional inputs, but the system has to be secure against tearing itself apart for at least some time while you find an airport and put down, right?
Is one easier than another? Only if you are very lucky. To be assured of having a salable product at the end, each product has to be engineered. If one were willing to guess and play and set a cancel date and be willing to just cut losses, yeah, it might be easy. Most businesses do not like those kinds of gambles, but some might.
Billski
