spur gear box or planetary gear box in light winch

[Rinsuke]

Looks like I found the right place...

We are designing a small winch to be used in forestry operations. I consult with end users and report incorporate their wishes into the new product design guidelines. The winch is meant to be portable , therefore light nevertheless we want a max 2.7 metric ton pull on the wire.
The gear box on the prototype is a 2 stage spur gear design for a 1:114 reduction. Input rpm is 6500 to 9000 with max 5kw power at max rpm. What would be the advantages of going planetary?

I find that weight in this case could be drastically reduced with a planetary design. Could it be a relatively simple one stage system? What kind of weight savings could we expect (percentage wise)? And what kind of cost comparison (general rules) is there between spur and planetary type gearboxes?

The additional compactness of the planetary design, in my opinion, would allow for weight savings elsewhere (gearbox would be shorter , the current spur gear design is double stage in-line and the planetary gearbox would certainly be more compact thereby allowing a shorter tranmission roller chain ...etc... )

I know these are many questions for a first post, but Idon`t have the technical background to go to bat for the end-users against the engineer here and I keep getting a flat answer to my suggestions of a planetary design : "Too expensive!"

TIA

Stephan

 

[plasgears]

At best the planetary would be two-three stages. In a spur gear parallel shaft arrangement with approx 5:1 per stage you have three stages. Proceed with the parallel shaft system. Increase the module every stage, and likewise increase the face width successively every stage. Lilmit pinion tooth count to 10-12. Your stress analysis should aim at near equal tooth stress thru the system. As a refinement, endeavor to extend the addendum in the pinions by a factor of 1.20, and decrease the addendum in the gear by a factor of 0.80. Of course whole depth remains equal for pinion and gear. Also, use a thickness ratio in the pinion of .56, thus favoring the pinion with higher thickness. This will give you low noise and manageable stresses. You will find that the pinion stresses and gear stresses run about equal. Lower stress in the pinion is even better for fatigue strength. It sees more cycles.

 

[Rinsuke]

plasgear, thank you for your reply. I am not an engineer so I had to reread your post a number of times with my reference books to finally get most of it ... I think. So, what you are saying is go with a spur gearbox (parallel shafts) , increase the module for each stage, and keep the input gear to no more than 10-12 teeth (pinion) . The present gearbox has first stage m1.5 x 6 spline then M1.5 x 60, second stage is M2.0 x 12, then M2.0 x 48. the ouput is a 10T sprocket to a 30T sprocket giving us the final reduction of 130:1

I thought a planetary gearbox would be both more compact and lighter but it appears according to your post, that in this case the spur gearbox would do just as well as the planetary and be a lot cheaper.



regards,


Stephan

 

[MattRR]

Two spur gear stages = lowest cost and largest design. Three simple spur gear stages will be lighter and somewhat higher cost. Helical gears have higher capacity and smaller space claim, but then you need bearings that handle end loads.
There are other possibilities that are lighter yet, but would increase cost. For example, a first stage simple spur gear reduction and then a two planet planetary with a 10:1 reduction ratio. The load sharing reduces weight. Input and output are offset though. You could use two stages of spur gearing, and then a planetary with reduction ratio of say 5:1 with more planet gears for more load sharing. Now cost is higher, space claim and weight is less. Using automotive type bearings can result in a low cost planetary. The bearings are then low cost. Use stamped sheet metal carrier and hobbed and shaved planet gears (no grinding) to keep cost low.
Another possibility would be a 'cycloidal' drive by Sumitomo, Fairfield, or other supplier. This has only one internal gear and one external gear, with the external gear mounted on an eccentric. Input and output are concentric. Your reduction ratio may be pushing the limit of a cycloidal drive.
Make sure the minimum number of teeth used for any pinion is at least 15 to stay away from contact near the base circle; and to avoid undercutting, depending on final tooth geometry.

 

[ielivaz]

Hello Rinsuke

I am agear maker, if you wish we could discuss designs.

ielivaz@oxxo.femsa.com.mx

 

[Rinsuke]

MattRR, thank you for the reply. Your suggestions for the design are all very interesting. Right now weight is the focus but no doubt additional advantages would come out of a more refined system, for example shock loading and wear of the gears. Offsetting of input/output is not a problem. Can you be more specific as to the way the spur gear output is linked to the planetary and how the output is handled?
Right now, the output shaft of the gearbox has a 10T sprocket linked by roller chain (Tsubaki #40 chain) to a 30T sprocket which gives an additional 3:1 reduction (although it is fairly heavy and must be further enclosed as per safety rules ).

I was perhaps a bit misleading in the way I first stated the problem since the gearbox itself only need a ratio of roughly 43:1 , the roller chain/ sprocket combination giving us the other part of the ratio needed. The final ratio can be changed as well, in fact the winching speed should probably be slowed down a bit and of course the drum size has an effect as well.

Anyway, this is of course if you have the time to comment.

Sincerely

Stephan

 

[MattRR]

Stephan,
If you are looking at 43:1 then the cycloidal drive may be ideal, Sumitomo, Fairfield, or Ikona could help you. For a planetary system, if you can live with an offset, then a single spur or helical stage, followed by a planetary/star epicyclic stage would work. To minimize weight, a study would need to be made on the optimum reduction ratio split between the first stage simple gearing and the second stage epicyclic gearing. A reasonable guess would be about 7:1 in the first stage and about 6:1 in the epicyclic.
To answer your question, the output from the gear of the first stage usually drives the sun gear of the epicyclic with a quill shaft coupling, although other couplings may be used. The output of the epicycic can be either the ring gear (star system, rotating ring gear, fixed carrier) or the carrier (plantetary, fixed ring gear, rotating carrier). Usually the sun gear is allowed to "float" radially to improve load sharing among the planet gears. The planetary is more efficient than the star reduction, and also provides somewhat more reduction ratio in the same size package. The advantage of the star system is that it is usually easier to lube the bearings, and there is no centrifugal load on the planet bearings resulting from a rotating carrier. With ring gear output, the ring gear is mounted on bearings, and a fixed spline is machined on the shaft to allow for attaching the sprocket. Your sprocket would have a mating spline to attach to the shaft. For the rotating carrier output design, the carrier is mounted on bearings, with a spline for attaching the sprocket. The sprocket would be straddle mounted between bearings, and the end bearings would be mounted in a cover that is separate from the main housing.

Regards,
Matt

 

[Rinsuke]

Matt
your ideas were discussed at length two days ago and we have two possible design in mind at this point with a three stage spur gear design slightly ahead at this point in the process (cost but also engine attachment to the gear reducer housing, we need a certain size housing to mount the engine and input drive...) Will come back to you when we have a clearer layout in mind.

thanks again

Stephan

 

[Spurs]

Stephan

Are you worried about efficiency in the gear box? If it is not a big issue, why not do the entire 43:1 in 1 stage and use a simple worm drive?

If you use high helix worms, (ie approx. 20 degree helix) you will see efficiencies of about 60%.

 

[Rinsuke]

to all who contributed ideas and advice for this project. Unfortunately the design ( and resulting prototype ) has still not found a perfect and definite solution. Mostly because of lack of funds I suppose... The prototype and concept though has gotten a very good reception from the potential end users and we are still at work on other part of the system ( shaft sizing and materail) drive chain from the engine ( I believe this one is settled, see we have a max rpm of 10,000 ...) and other peripheral hardware to be packaged together with the winch ( if it goes commercial). The transmission remains the biggest challenge . If money was not a hurdle, I have found the Shimpo speed reducer and cyclo drive very well designed, but cost and weight remain question marks.

I certainly do not want to settle for any less that can be achieved and we are asking a few people here whether it would be possible to design a custom cyclo reducer ... I may have some questions in the future again , for now though, I wanted to thank Matt, Spurs, plasgear and others who responded with advice.

Step1

 

[Rinsuke]

Spurs, the worm gear solution was examined but friction is considered quite high and self locking is not an advantage in our case so it was dropped. It is not easy dealing with the input/output shaft angle either, so planetary is still favored, cyloid hoped for.

regards

step1

 

[Spurs]

Rinsuke

If you can allow the system to be backdriven, then you should be able to use a high helix worm with potentially an efficiency of 75% all in one stage.

If you cannot allow backdrive, you could use the same worm but add a clutch mechanism to prevent backdrive.

Thats not that bad an efficiecny for 1 stage, especially when you consider that the drive efficiency of spur gears is approximately 95% per stage.

Therefore over 2 stages, the efficiency would be .95*.95 = 90.25%. It gets a lot worse as you add more.
85.7% over 3 stages and 81.4% over 4, 77.3% over 5 stages.

 

[MattRR]

Rinsuke,

I am aware of someone who is developing a new type of "bearingless planetary" for a similar load, speed, and reduction ratio application. Target cost is $400 per unit, building 100 units at a time. Unit cost could be reduced if larger production is planned. Weight is under 5 lbs, and max od of the drive is 5 inches. Length is about 4 inches. If you are interested, let me know, as some of the details are proprietary. The unit has greater than 95% efficiency overall.

Matt

 

[Rinsuke]

MattRR

Yes I would be interested to know more about this gearbox. Sounds like it is a close fit.

Step1

 

[MattRR]

Rinsuke,

For more info, contact George Gianopulos & Associates, Dayton, Ohio. Phone 937-274-2996.

Matt