2 Stage Reduction Vs Planetary Gear System

A cycloid reducer might work for this. The output is in line with and the same rotation as the input, and they are fairly compact and shock resistant. One brand I know of (no commercial interest) is Sumitomo "SM-Cyclo."

Thanks. That will be considered.

Say if I leaned towards the first option and went with a double strand roller chain, I was wondering what kind of sprocket selections I would be looking at for the 3:1 stage alone. Where can I find an RPM-horsepower rating chart for double strand chains? Would you know? I have found plenty for single strand.

Sorry, let me clarify. I was asking what would be an equivalent pair of sprocket sizes needed if I went with a double strand chain, for the same kind of 3:1 reduction for the first stage. Input RPM is like 2500. If I can reduce my sprocket sizes that way, from what would otherwise be needed with a single strand chain, this option would be most attractive. However, I'm not sure how to go about selecting a double strand chain neither have I seen any literature with HP charts for them.

I don't have a chain book at hand. Depending on how many side/middle plates there are, double strand chain would handle about (but maybe not quite) twice the HP of single strand. Chain will need to be lubed, which may create some issues. You might also take a look at "silent chain," which will minimize noise. Belt drive could also be a candidate for this moderate HP. With 3600 rpm, check that chain/belt linear speed is within limits.

Thanks, the multiple strands factor would be 1.7. Chain needed would be an ANSI No. 35-2. But minimum sprocket size is ...hmm, puzzling me.

Several factors can influence minimum sprocket size. The sides of the chain must clear the sprocket hub, which in turn must have sufficient thickness over the shaft diameter. Tooth wear is greater on smaller sprockets. For sprockets with fewer than about 12-13 teeth, chordal action can cause chain whip.

We're thinking of a 12T for smallest sprocket mounted on a spline shaft. But design literature says we need a minimum of 17 tooth, but I understand this guideline is stated giving consideration to the fact that you're running a single thread roller chain. How about double strand? Would we fare well with, say a 15T but on a double strand chain?

Ya got me there; I don't know if the minimum sprocket differs between single and multi-thread chain. Off hand, I don't see why it would.

Once again, I don't have a book here, but if the pitch of #35 chain is 7/16", 0.4375 ÷ sin 12° ≈ 2.104" pitch diameter. [Not sure about the pitch--just estimating.] Whether this restricts the hub diameter too much I don't know....

The silent chain mentioned earlier is the kind that timing chain is, so it can handle the engine rpm. I think the smaller sizes of roller chain can, too, but I would want to check. Most of the chain drives I have worked with are downstream of a reduction gear, hence low rpm, and I don't recall the limits. A Dodge, Browning, Martin, or Tsubaki book would likely have this info.

Have you in your experience found cycloid gear reduction application in automobile transmissions? Would it add a lot of weight?

They are fairly compact, but I don't have a spec sheet handy to make a selection and get dimensions and weights.

I don't know if you have room to maneuver as for using in-line versus offset shafts. Cycloids and planetaries are typically in-line; belts, chains, and other gearing are typically offset.

We'll have to offset it with the orientation of the engine's PTO shaft and the axle. Will be tuning the CVT soon. Its a good design exercise.

With regard to hub, I'm thinking of implementing a 12T gear with a polygon spline. It would give me more space between the shaft and teeth (stronger sprocket) and would allow me to carry greater loads on the shaft itself. How much higher I don't know. But I believe I can get away with it.

The efficiency of transmission of planetary chain is one of the highest and for your case it may be >95%.

The advantage are the compactness and the rigidity.

The direction of the output and the input will be same if you go for fixed ring model ie the ring will be stationary on the body the carrier will be rotating (and will be the output shaft attached to it), input from the sun gear. The ratio of 6:1 may be available off the shelf.

Only there can not be any offset at the center line of input and output. You can check the globalspec for the planetary gear boxes manufacturer, and there are plenty.

Check a reputed one, since the quality of the gears matter in these. Load taking capacity will never be a problem, since the loads are distributed over multiple planets in the low speed range and that creates the compactness. (We make in that speed range or almost ie 7:1 around, only the capacity is too high- ours are of multiple 100s kf KW range and the efficiency is > 96% on our machines)

advantages- High efficiency, Rigid, Reliable (if from good source), compact.

Disadvantage: No center line offset, unless designed for it (it is possible- we have an HMC where one of the attachments have a centerline offset from the drive spindle and is planetary reduced with rotating ring gear ) , Cost (comparable).

Thank you. We decided to go with planetary gear of 4:1 reduction after another round of research!

Its going to have an input shaft and an output shaft, both hollow, of 21mm outer diameter each. But because the output shaft is seeing "4 times the input torque" from the engine after the CVT in low ratio, we're wondering if 21 mm would be enough to support the torsional shear stress on the output side. After running the numbers on the 21 mm output shaft based solely on the idea that is experiencing only torsion, the shaft would see a torque of 200 ft-lbs, and if we use that in the torsion equation, our torsional shear stress comes out to 136 Mpa.

The ASME says that the maximum permissible shear stress in a transmission shafts with allowance for keyways (which we are going to have) cannot exceed 42 Mpa. So looks like we're definitely off limits there. Should I go with a bigger shaft size to bring the shear stress down or do you think 21mm shaft (O.D) would do fine? The I.D of the shaft is 12.5mm.

Thank you for your help.

Input shaft of the planetary is 12.5 or the output shaft is hollow? Φ21/Φ12.5?

I am not immediately in a position to do the calculations, but based on your initial data (10HP = 746W) and RPM 3600/4=900= around 94 Rad/sec , the torque transmitted will be quite low.

What you have to do is do the stress analysis including keyway (I am continuing here instead of in the other post ) and check the factor of safety. Why your torque data is not matching with mine?

Usually we prefer atleast 8 for fatigues, and our output shafts (these are too large just to give an idea these are 450HP @12" shaft at around 50 RPM so very high torque, with keyway and very heavy lowcycle fatigue) are of AISI 4340 and is one of the better ones (forging, Hardened and tempered) for these applications.

If I downscale it to your (10HP) the FOS seem to be OK, that is at first glance without running the detailed analysis software.

In case you have a problem with keyway stresses, think about hydraulic fitted hubs, as we do for quite a few ones.

Got a good link here, a bit old like me goes back to - well a bit later than my old classroom days still..

May be good enough for hand calculations, though didn't cross check with ANSYS, still it may help.

Thank you for the link. The application of the shaft is in an offroad vehicle which will see a lot of speed fluctuations because of speed bumps, obstacles, hills etc. I have only done a pure torsional analysis as simple as it gets. Perhaps I should do a "fluctuating torsional analysis" for my hollow shaft? What do you think?

Carry out a simple stress analysis (if you can get hold of ANSYS even better) and then put a factor of safety of at least 8 in case of high fatigue (possible) though in case of chain drives, the fatigue may be a bit less due to its elasticity, but you are saddled with a high bending moment at the sprocket joint.

You have to do the complete stress analysis and with a 8 - it is supposed to take care of total fatigue- if i remember that is the value that corresponds to infinite life principle in fatigued shafts.The maximum stress is likely to be on the keyway corners - concentrate on that for the factor of safety calculation.

Also think over keyless joints (taper hydraulic or shrink fit may serve the purpose) In fact as per the literature the hydraulic pressed hubs have more transmission capacity (though may not look like) than keyed hubs.

But is your torque calculation OK? It looks to be a bit high for that RPM

On a dia 21/12.5 hollow shaft how are you going to manage a keyway? the standard keyeay may not suit there, the depth will be too high. Why hollow shaft? lubrication?

Best way to find may be to check the planetary output shaft diameter that should give you sufficient data.

I just feel that the diameter should have been nearer to 100mm (4" than 1")

The sprocket is going to be splined connected. The only keyway joint is at the secondary clutch. The 21/12.5 combo on the input shaft was tested last year with the secondary clutch and it worked great, so we don't find a reason to change that.

Keyless joints are going be a little impractical right now.

4" dia is going to be huge given our design constraints (bearings, attachment points etc).

Sorry can not have a look at the ACAD (my server firewall blocks the personal storage sites with some big bold messages IWSS Security blah blah ).

However as I said is your torque calculation OK? As I said

We have say one gear box of 500KW @40 RPM and the hub dia is 300mm (keyless), and that to of SGI. Loading is most likely worse than your case- cyclic -fatigue, bending etc all there.

Even if you are at that RPM - your size - normal downscaled will be √(500/7.46) = approx 8

And that makes dia 300/8= 37.5mm provided you are at this RPM- and your data shows much higher RPM and that makes the 21mm shaft OK.

Then again as you have already tried with this size shaft at this RPM (?) it should work fine.

Anyway I am looking at bot your posts (so need not repeat the data , I will try to maintain this however since this is a longer link going on)