quiet gear cam drive 4

[ivymike]

Hello

I was wondering if anyone has any knowledge of techniques used to make "nodal" cam gear drive systems quieter?

 

[patprimmer]

Make the cam gears out of nylon 4.6, possibly carbon fibre reinforced

Regards
pat

 

[EnglishMuffin]

I realize that this site isn't intended for general instruction - but just for the benefit of us non-automotive dummies, could you explain what a "nodal cam gear drive" is ?

 

[ivymike]

Well, perhaps I picked a poor way of phrasing it, but what I intended to communicate was that I'm interested in noise reduction for timing drive systems that use gears to drive the cam, and are "nodal" in that they are driven from the "node" end of the crank (flywheel end).

The idea behind the nodal drive is that crankshaft vibrations would have very little magnitude close to the node, resulting in very little vibratory input to the timing drive. Just ahead of the flywheel is where you would find a crank vibratory node, in most cases.

 

[EnglishMuffin]

Got it - thanks.
So I would guess that anything that made a conventional gear drive quieter would be efficacious in this case also, such as using helical gears with optimised addendums to minimise the slide roll ratio. But even though you are at a node, I wonder if some residual torsional oscillation is still a problem ? If that were the case, possibly gear rattle could be a problem. It's possible that this could be prevented by using some type of mesh preloading system, of which there are many. But this costs money of course. It's not generally known, but there have even been people in the machine tool field who have designed high power gearboxes to run with radially preloaded gears - and no special mechanisms or excessive compliance whatsoever. Possibly a flexibly mounted idler gear arranged to apply radial tooth preload might work. Ideally that needs a high pressure angle and very good lubrication. Just some thoughts.

 

[ivymike]

I guess I should have mentioned - the fuel injectors are "pushed" by the cam, so the cam torque reverses several times per revolution. Gear rattle excited by the cam will almost certainly be a problem, and there will certainly be some sort of torsional vibration damper or absorber involved.

I thought of helical gears too, but there is about 95% certainty that I'll be stuck with straight-cut gears.

The radially preloaded concept sounds interesting... I'll have to see whether people look at me like a two-headed or three-headed monkey when I mention it... If it scores only two monkey heads then I'll end up investigating in much more detail.

So far I've got
* pre-loading the gears
* improving the meshing of the gears
* reducing vibratory excitation
on my list. How about designing the gear casing, are there any tried and true methods for making a low noise radiation gear housing?

 

[ivymike]

sorry pat, I forgot to include "making a gear of a less-noisy material" on the list.

 

[PJGD]

If you must have a camshaft (and we are not quite ready for the camless revolution yet), then a nodal drive is the only way to go. The old, and very lame argument against them was that they are harder to work on when maintenance is required. With heavy-duty engines nudging the million mile mark between overhaul, cam drivetrain maintenance is a non-issue. Besides, there should be less stress on the gears of a nodal drive, than in the case of non-nodal, so they should last longer.

If you are driving unit injectors, there will be interesting reversals involved, particularly for widely separated multiple injections. Probably the direction to move in is to incorporate a damper into the cam gear; for instance the Cat 3176 used a pendulum damper there.

If that solution is too easy, you could always contemplate a 3-throw eccentric drive with coupling rods, as used on the 4.5 L Bentley engine!

PJGD

 

[EnglishMuffin]

The damper idea is good - I've applied a viscous torsion damper to a gear train which was experiencing cyclic load reversals, with good results. It was a Houdaille type with silicone oil.

 

[patprimmer]

Isaac
I know nylon sounds a bit light duty, and it might well be inadequate for your application, but type 4.6 is pretty new, and performs quite a lot better than type 6.6. Many people are still unaware just how much it lifted the bar in engine internal applications.

I am a bit out of date now, as my last data is from 1997, but MB, Porsche, Saab and BMW were all looking at it for timeing chain tensioners. It replaced molly and graphite filled nylon 6.6 in a few of these, giveing a minimum of 3 times the service life. The Porsche trials were on an air cooled motor. Even though it gave 9 times the life of molly filled 6.6, it was still not used, and a metal part was specified. I expect the air cooled motor had very high oil temperatures compared to the other engines.

When fibre reinforced, nylon 4.6 has usefull properties up to about 290 deg C, vs Nylon 6.6 at about 200 deg C

Regards
pat

 

[Warpspeed]

Would it be possible to use two sets of gears side by side sufficiently preloaded in torsion, so that tooth contact was maintained?

I like the idea of the Bentley eccentric drive, but if I remember correctly this also had a 2:1 gear reduction associated with it.

 

[EnglishMuffin]

If it's a four stroke with a conventional camshaft, what's wrong with a 2:1 gear reduction ? Your split gear certainly works - it's done in a lot of zero backlash gear drives (it was one of the "many" that I mentioned in an earlier post). It is probably quieter and more problem free than radial preload, which I suspect would have a whine associated with it, plus high wear. But you do have to make two gears instead of one.
Regarding plastic gears, I wonder if anyone has ever experimented with using a Wildhaber-Novikov tooth profile?
This tooth profile, which is almost unknown in this country, has been used for some high power (steel)
applications in Russia, and is nothing like an involute at all. The teeth have conjugate circular arc profiles which conform very closely with one another, making for very strong teeth with very low contact stresses. But the gears have to be helical to get the correct constant velocity meshing action. You can generate Novikov teeth by hobbing just like involute teeth, and in the case of plastic gears, once you've made the mold, who cares whether it's an oddball ?

 

[EnglishMuffin]

One further note about WN gears - here is a guy on a web site I just found that says WN gears are not all they are cracked up to be - but says at the end that you can get 40 -60% increase in load capacity, which is not to be sneezed at.

http://www.zakgear.com/WN.html

 

[shabba]

Has anyone ever heard of a torque balancing cam? That could remove a great deal of your cyclic loading, what stage are you at with your engine design?

 

[GregLocock]

Isaac, check out Honda's scissor gears for a camshaft. These use a torsionally preloaded second gear set to reduce the effective clearance to zero. A cunning engineer might include some damping in the compliant feature.





Cheers

Greg Locock

 

[ivymike]

what stage are you at with your engine design?
Don't want to be too specific, but early enough to do almost anything, late enough that there is pressure not to do it unless it will definitely work.

check out Honda's scissor gears
Thanks for the tip, I'll do that.

 

[shabba]

Hmm, scissor gears are a good idea, but not quite sure how the honda setup works (second gear set?) Usually to set up a scissor gear train, all the gears (except the driving gear) are split type with only one being spring loaded. This gear will preload the whole train eliminating backlash.

One thing to watch (as greg pointed out) is vibration. You may need to carfully examine the potential harmonics and perhaps introduce some kind of damping.

Another (similar) method I have come across used a separate thin disc, preloaded axially against the side of a gear. This disc was cut with one extra tooth compared to the driven gear and therefore rotated at a different speed, using the friction between it and the gear as a damper. Don't do it.

 

[PJGD]

Some further comments on this topic:-
You recall that in a separate thread, I recommended "Valve Mechanisms for High-Speed Engines", Philip H Smith, 1967, GT Foulis, (Not to be sold in the USA). I would strongly recommend it again, for ideas that may spark solutions.

For instance, it shows the pre-war MG engine with a vertical bevel driven OHC drive with the generator integral to the shaft, ie, the rotating armature acts as a high inertia damper in the drive. Staying with your gear drive, you will I am sure have looked for a half-engine-speed accessory with a high inertia or other opposite torque canceling signature, such as a hydraulic pump that you could drive from the cam gear. Even if there were such an accessory, I doubt that your engine would require it anyway!

Another possibility is to arrange your pumping cams to have relatively long top dwell, and then have the return flank coincide with the pumping flank of a different cylinder. That will help to cancel out some of the instantaneous drive torque spike.

Then again, it may be instructive to contemplate the reduction gear drive used on the Napier Sabre aero engine. The H24 cylinder engine (two flat 12's, one above the other) joined the two crankshafts to the propeller output gear through helical compound gears with a balancing feature which cancelled out the torsionals from the two cranks. I assume that in this case, the propeller and it's output gear had high inertia and thus was a nodal point in the middle. If you will be using the flywheel/clutch mass as your nodal inertia on the crankshaft, then an arrangement as wonderfully elegant as the Napier solution, will be difficult to arrange.

PJGD