Geared Crankshafts in IC engines 2

[factsb4pride]

I have been working on a project to develop an IC engine that uses gears to translate reciprocating motion to rotary motion. I have heard many posts here, as well as information posted elsewhere, that says current gear technology is not strong enough to allow gear teeth to survive combustion pulses without damage or undue wear in the long run.

However, another member recently posted info in a thread about Neander motors, which has 2 counter-rotating "crankshafts" that are directly geared to one another. Then power is taken from one crank by a CHAIN drive to the tranny. So combustion pulses are definitely being transferred across the gear tooth faces and through the gear teeth. AND this is a diesel engine, NOT a gas motor, so the forces the gear tooth must handle are very high.

So I guess my question is, do current gear tooth profiles and fabrication materials allow for this now? Are objections to this design of engine based on current realities, or outdated facts and opinions?

Yes, I understand no current engines use this method (except the neander motor), and that it is not a common practice. But my question is: Is it a viable research path to take, and if not, why not. Please be specific.

 

[PJGD]

On the contrary, geared cranks have been very common over the years.

Just a few examples: the Arial Square Four [SI], the Coventry Climax Flat 16 [SI], Junkers Jumo 205 [CI], Napier Deltic [CI], Napier Sabre [SI], Sulzer 12LDA28 [CI], and on and on.

PJGD

 

[factsb4pride]

Thanks for your response.

In my research I came across a paper written back in '88 titled "A Critical Evaluation of the Geared Hypocloid Mechanism for Internal Combustion Engine Application". The focus of this paper was engines using an internal ring gear and planetary gear with a ratio of 2:1 to translate reciprocating motion into rotary motion. One of its conclusions was that the tangential gear tooth loads were within an allowable range for the gear teeth to handle, but the tangential dynamic loads were right on the borderline of acceptance.

I have also seen many objections on this site over the years to engines that have gears receving direct combustion pulses and loads. So I thought I would just ask if there were still valid objections to them based on verifiable evidence, especially since viable engines are using them that are for sale today. And please understand my approach is not manufacturing costs, or any others, but simply whether or not current gear technology can handle that kind of service without undue wear or gear tooth damage.

 

[BrianPetersen]

I'm thinking that as long as there is enough flywheel effect in the crankshafts, and the velocity profiles on each side are not too different from each other, and possibly with some compliance in the way the gears are mounted, it will be OK.

Most transverse motorcycle engines use a direct gear drive from the crankshaft to the clutch basket, and that clutch basket has a set of springs between the actual gear and the clutch basket itself to allow some compliance. Same type of mechanism that car engines with manual clutches use to absorb some of the vibration between the crankshaft/flywheel and the transmission input shaft.

If you are using the gears directly in place of the crank and con-rod in some fashion, so that the piston loads are being transferred directly into the gear without any interposing rather-high-inertia flywheel, it might be a different story, though.

 

[factsb4pride]

Hi Brian, That is exactly the situation I am enquiring about: gas loads from pistons are being directly transferred to gears, although hydrodynamic bearings are in place to maintain proper gear positioning. Only tangential forces are being applied to the gear teeth.

 

[factsb4pride]

I should qualify by that further by stating their is a hydrodynamic "big end" bearing between the gearing and the piston in the design I am looking at. But is that relevant (or not} when it comes to the magnitude of the load pulses being transferred through to the gear teeth?

Another conclusion reached in the paper I mentioned previously was that a certain ratio of bore to stroke could be applied to geared "crankshafts", and they determined it was 3:4. The idea being that if the large internal ring gear was always kept a certain ratio larger than the bore, the subsequent combustion gas loads would never exceed the strength of the gear system, and more specifically, the individual gear teeth. Anyone have any thoughts on that?

 

[factsb4pride]

oops, forgive my use of "their" when it should have been "there".

 

[MikeHalloran]

It is possible to engineer gears to withstand arbitrary levels of shock and vibration for a specified lifetime, within a solution space bounded in part by material limits and physical space constraints associated with the mechanism under development.

The process of actually doing that engineering comprises a lot of calculations conducted with differing values of many of the possible parameters, and may include graphical layouts or their mathematical representations, and perhaps a few physical prototypes or test articles associated with experiments conducted to answer idle speculations or unresolved issues.

The process may include some of the 'bench racing' you are attempting, _after_ at least some of the calculations have been done, but the bench racing does not comprise the whole of the process.

You insult our education and experience with the phrase "outdated facts and opinions". I will refrain from a complementary slur.

It has been pointed out in a separate discussion that Watt's planetary linear to rotary converter was invented in order to get around a patent that covered the slider-crank. Both are now in the public domain, and have been analyzed to death, and can be used to make durable engines. Slider cranks dominate right now because they are much less expensive to manufacture. That is also a valid engineering constraint on a problem's solution space.








Mike Halloran
Pembroke Pines, FL, USA

 

[patprimmer]

The Mazda Rotary (Wankle) engine also uses a geared crank I think because it needs to because of the motion of the piston relative to the crank.

Why add extra complication if it's not needed.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[izzmus]

The gear is there strictly to keep the rotor in the correct position relative to the peripheral housing, although its job does mean that some amount of torque is resisted by it. One gear is stationary and the other gear is mounted on the rotor.

The rotor gear is a light press fit and is located rotationally with spring pins (split roll pins). Interestingly, Mazda reduced gear loads by switching from nine pins to twelve.

 

[MikeHalloran]

The very first Wankels did not have anything resembling a crank. They had a triangular rotor rigidly connected to the output shaft, that rotated 2 revolutions for every 3 revolutions of the entire remainder of the engine, like a LeRhone rotary, including the shell, sparkplugs, carburetors, whatever. Or maybe it was 3 for 2.

The orbiting rotor connected to an output shaft by a 3:2 planetary gearset was the result of a kinematic transformation, which allowed the engine shell to remain stationary, greatly simplifying the fuel system, etc.

The output shaft geared to the rotor represented complexity added because it _was_ necessary.



Mike Halloran
Pembroke Pines, FL, USA

 

[factsb4pride]

@ Mike Halloran: I am not sure why you perceived some kind of insult in a question I was asking:

"Are objections to this design of engine based on current realities, or outdated facts and opinions?"

While I appreciate your input, no insult to anyone was written or intended on my part.

To others:

I have also analysed the use of gears in the mazda rotary engine, as well as others, and thus my question to the forum: Has gear technology reached the level where gears can directly absorb the kind of punishment that IC engines dish out, both gas and diesel? If the consensus is no on this forum, that is information worth knowing considering my current project at work.

Based on my research and calculations so far, the answer depends on how many gears are used to direct combustion pulses from a single cylinder to the rest of the drivetrain, what the ratio of bore to stroke is, crank throw length, and other factors as well, such as tooth profile design and gear module used. If an internal ring gear and pinion are used, and rotating forces are balanced out properly, the only loads the gears see are tangential in nature, and comprise the gas loads from the piston.

Please keep your input coming, appreciate it all!

 

[BrianPetersen]

You CAN engineer the gears to withstand the loads. This does not imply that the outcome will be practical or cost effective compared to a conventional crank and con-rod mechanism ... but you CAN do it.

If the gear mechanism is big enough, and the cylinder is small enough, and the revs are low enough, at some point the stresses will be OK.

You will have to analyse the forces involved in your proposed cylinder size, taking into account the peak cylinder pressure and the various inertial loads, find out how much torque is being applied to the gear (and along with that, how much force is being applied to every other part of the mechanism), and then size the gears accordingly and design the rest of the mechanism accordingly.

There is no substitute for crunching the numbers. No amount of us telling you whether it is okay or not will take the place of that!

 

[factsb4pride]

@ Brian: I HAVE crunched the numbers, and have come to my own conclusions based on those calculations. No substitutions were implied. I am simply here seeking out alternative viewpoints, and hopefully supporting references to those viewpoints.

I already stated my research has revealed a bore-to-stroke ratio of 3:4 has been proposed as a guideline to designing IC gear driven systems that employ internal ring gears. This relates directly to your comment that such systems can be designed effectively.

I am hopeful that some of you readers have real world experience you are willing to share on this topic. Hopefully with experimental data or supporting references...

 

[BrianPetersen]

A fixed bore/stroke ratio and not dependent on any other factors is not a plausible outcome. What about the peak cylinder pressure? What's the compression ratio? Is the engine going to have forced induction? What's the effect of combustion on the peak cylinder pressure? At what point in the crank rotation is the peak cylinder pressure going to occur? At what point in the crank rotation is the maximum torque going to be imposed on the gear mechanism?

 

[tbuelna]

factsb4pride,

A gear drive connecting two counter rotating crankshafts can be successfully designed. But as BrianPetersen points out, there are lots of issues to consider.

The gears must be designed to handle the peak instantaneous loads throughout an engine cycle, produced by all the cylinders. The gear loads must also take into account any dynamic loads due to crankshaft torsionals. Some teeth on the crank gears will have a shorter fatigue life than the rest, since they are the ones in mesh during periods of peak loads. And if the gears are fastened to the crankshaft, the gear face contacts should be de-rated for any mesh misalignments caused by crankshaft bending.

Finally, if you put any idlers in the gear train, be sure to use a lower tooth bending stress allowable for your idlers, since they will experience full reverse bending loads. Also take into account the fact that engine oil is not a particularly good gear lubricant.

Good luck.
Terry

 

[factsb4pride]

Thanks to all for responding!

@ Brian P: The 3:4 ratio I mentioned takes into account the factors you mention, including peak gas forces, and compression ratios. The idea being that the bigger the ring gear, with the number of gear teeth remaining constant, the more force the gear teeth can handle. Given those constraints, larger ring gears = larger geer teeth. If the ring gear and pinion always stay a certain ratio above cylinder bore diameter, the researchers (that came up with the ratio) claim it will always be able to handle the subsequent dynamic gear tooth loads. I encourage you to read the paper I posted earlier, as it goes into great detail about this topic. I also have another paper where the authors claim the pinion gear face width should be no less than one-third of the pinion gear diameter. So more than one research group has come up with guidelines for constructing this kind of mechanism.

@ tbuelna: I am very familiar with the concepts and points you posted, as I have been researching this topic for quite a while. Setting that aside, my goal here was to find people who could provide any real world experience, or experimental data, that I might have missed in my own research. I am well versed in the loads that gears must endure to achieve the stated goal in this application, I am simply trying to find gear technology that can accomodate those loads, or find people who can offer relevant input.

One thing I need to clarify: size does matter. Of course, just about any engineer worth his salt could design a hugely oversized gear set that would survive direct combustion forces. That is fairly obvious. I am talking about a practical application, not merely a thought exercise, so if I gave the wrong impression earlier on, I apologize.

The engine I am designing is intended for home electricity generation, and use as a hybrid generator. It will run at its most efficient rpm most of the time. Efficiency is more important a consideration than a broad torque band, or high end power output or high rpm. Some additional cost for gear sets (over and above what slider cranks would have called for) is allowable, but within reason. Just as the gear set sizes have some variability, but within reason.

As another poster pointed out, a lot of research has gone into this kind of concept over the last 40 years. The problem I see is that all people do is theorize about the topic, with few if any actual machines built, run, and/or tested tested. Ishida was one of those. I was hoping to come across the few exceptions in this forum, but that may not be a valid expectation.

 

[factsb4pride]

I have been considering Brian P's comment about direct power gear viability:

"if the velocity profiles on each side are not too different from each other".

In the neander motor I mentioned earlier, the crank gears rotate together in the same direction at the same speed and loads. One is not stationary, while the other is in motion, they are both in motion equally at the same time. That would seem to shoot down my train of thought that the neander motor proves it can be done.

However, as near as I can tell the neander motor passes power to the transmission off a single crankshaft cog and through a chain. So the total combustion peak loads of the whole engine would be passed through that one chain cog. If that is possible, it seems it would be possible for a solid gear to absorb those same forces, especially on a single, per-cylinder basis.

 

[PJGD]

It sounds as if you plan to run the engine at its efficiency sweet-spot speed and then gear a fixed speed generator to that.

A few thoughts:
1) If your engine is 4-stroke, you might be able to tune the engine to be efficient at twice the generator speed, and then drive the generator off the same gear as the camshaft. See SAE paper 700205 on the Teledyne Continental "Tiara" engine to see an example of this.

2) A good strategy might be to use the artifice of a "hunting" tooth in your gear mesh to distribute the wear from the pulse loading and any mesh errors across all the teeth.

3) I think that the rule of thumb is that you can expect to loose about 0.5% of transmitted power per gear mesh.

PJGD

 

[BrianPetersen]

In the application in question, what's the proposed advantage of this gear-type mechanism over the tried-and-true normal crankshaft and con-rod?