Gearbox Reverse Drive Question 1

[kander]

Lets take a gearbox of ration 4:1. 4 in = 1 out.

But say we were to drive the gearbox in reverse, ie 1 in = 4 out. As a rule of thumb, taking an average friction ratio for gearboxes, what is the maximum gearbox ratio before the gearbox will bind up under the friction of the gears?

 

[Metalguy]

For starters, it depends on the type of gears and their arrangement.

 

[kander]

Metalguy. Ball park figure. Lets say single stage planetery. Real rough idea.

 

[Barry1961]

A single stage planetary will not go much over 10:1. I have back driven 100:1 2 stage planetary reducers.

Barry1961

 

[sreid]

Regarding the original question, is the gear ratio limited in any way by teeth engagement friction? My visualization is that if the driven gear becomes infinately large it becomes a "rack" gear. And one can certainly "back drive" a pinion with a rack.

 

[MikeHalloran]

There is no friction involved in tooth engagement of proper gears; the tooth faces roll on each other.

One can indeed back drive a pinion with a rack. However, as the pinion becomes smaller, the force required to generate a given torque on the pinion becomes larger, and eventually reaches a level sufficient to fracture a pinion tooth.

For a single gear pair, there is a ratio beyond which the teeth are not strong enough to transmit the loads on the small gear.

For planetary gearsets, the problem becomes more difficult because the gears must be precisely matched to make them share the load.



Mike Halloran
NOT speaking for
DeAngelo Marine Exhaust Inc.
Ft. Lauderdale, FL, USA

 

[sreid]

In my post above I meant to say "driving gear", not "driven gear."

For years, I too thought that an involute gear mesh had pure rolling friction until I happened to buy Earl Buckingham's book, "Analytical Mechanics of Gears." To quote, "Thus when two involutes are acting against each other, a combined rolling and sliding action takes place between them because of the varying lenghts of equal angular increments on the profiles." He then goes on to analyse the sliding part in detail. It is small (usually less than 1%) but not insignificant and is the primary efficiency loss mechanism in gears.

In fact, the sliding is smaller by half between a rack and a pinion than it is between two "pinions" of equal size.

 

[diamondjim]

There are two types of sliding, one
in the arc of approach and the other
in the arc of recess. The first
does cause friction. Luckily it is
often a small part of the mesh.

 

[kander]

Barry1961. You say that the upper limit to back-drive a planetary gearbox is about 10:1, however you then say that you have back-driven 100:1 planetary, 2-stage gearboxes. This doesn't make sense. Can you please explain.

I now have more information. The gearboxes are 255:1 planetary 3-stage. Guys, putting individual gear teeth meshing friction, arc of approach, etc aside for a moment, is it possible to back drive a 255:1, 3-stage planetary gearbox?

 

[diamondjim]

If you have enough torque, I cannot
see why it would not be possible.

 

[Barry1961]

Sorry for not being clear. When I mentioned the single stage 10:1 I was refering to your reply to Metalguy:

<"Metalguy. Ball park figure. Lets say single stage planetery. Real rough idea.">

I don't think you can get much over a 10:1 reduction with a single stage planetary gear set. A 10:1 is very easy to back drive.

The largest ratio planetary gearbox I have back driven was a two stage 100:1.

On a side note, if you are driving the gearbox with a DC or servo motor you will probably have to disable the dynamic braking to back drive it. Most DC and servo drives default to dynamic braking mode when power is off.


Barry1961

 

[kander]

Barry1961. A two stage 100:1 planetary gearbox. My gearbox is a three stage 255:1. Therefore, if we assume all the gear stages have equal ratio, would each gear ratio be around 6.34:1? If so, it would be possible?

 

[BillPSU]

In the mobile equipment industry, a 24:1 worm wheel gear box is commonly backdriven in the rotation system. Pressure relief's are set and allow the boom to swing back if excess side load occurs.

 

[Barry1961]

If I were going to guess at the ratio of each of the three stages in your gearbox to get a total of 255:1 I would assume they used a combination of their standard ratios they have listed in their single stage units. Often the ratios listed are not exact.

For example it is common to get a two stage 20:1 ratio with a 4:1 and a 5:1, not two 4.47:1.

I don't see why having three sets of 6.34:1 would not be possible, just not likely.

You might be able to get the mfg to back drive one on the bench and measure the torque for you. It would probably be close with any three stage box. You could also measure the torque to back drive your motor then multiply it by the ratio. With such a high reduction I would try to measure the static/break away torque also.

Barry1961

 

[kander]

Thanks Barry1961. The individual gear ratios are going to be fairly close, yes? i.e. 7.5:1, 6.8:1 & 5:1?

Or could you expect something like i.e. 10:1, 12.75 & 2:1 for planetary gearboxes?

 

[Barry1961]

Just from talking with various gearbox manufactures over the years I learned there are some ratios that are intrinsically weak. I am not sure if the weak ratios are the same with all brands but if I remember correctly the most common ratios to avoid were 3:1 and 10:1, low and high end, with the low end being the weaker of the two. They were not too bad when used alone but often the manufacturers would avoid using them together. If you look in the catalog you may be able to spot the gaps in ratios of two stage boxes and deduce the weak ratios.

If all they were worried about was torque rating I would guess they would want to stay in the middle, 5:1, as much as possible. But of course noise is a big issue these days so they probably size the first stage to a quiet ratio even if they have to sacrifice a little torque and efficiency to make it quiet. I don’t know what the quietest ratio would be or if it would be the same across brands. My gut tells me the higher the better for noise but it is probably not that simple.

Some of your less precision industrial boxes such as Rex PlanetGear may not care about noise since they are not designed to operate much over 1750 rpm and are not used in quiet environments like hospitals. I would assume for this grade of box they would size the ratios based on torque and cost only. They may be willing to sacrifice a small amount of torque to use a gear set that they need more volume in to keep cost down or a gear set that uses less material or cheaper bearings or easier to manufacture.

There are good engineers who spend their entire career figuring out the best design for a planetary gearbox and still learn new stuff everyday.

Barry1961