Potentially daft question on worms and wormgears

[THT]

This is connected to at best an educational project, not a commercially viable one. This is completely outside my field. Any answers, explanations, and (online) references would be greatly appreciated.

The mechanism has a gearwheel A that needs to drive gearwheel D (and E). A's axis is horizontally 90 degrees from D's axis and they are separated by some space. Can horizontal wormgear A drive a vertical worm B, that in turn drives a vertical worm C, that drives a horizontal wormgear D, that in turn drives a horizontal wormgear E?

Searching, I failed to find illustrative examples of such successions. Individual contact interfaces yes, but combinations is where I failed. I found explanations why a spur wormgear doesn't drive a worm, but I'm unsure of helical wormgears, and I'm still wrapping my head around hyphoid.

The (tentative) reason for two worms is restrictive space considerations. D and E drive parallel-axis cylindrical objects; there is some friction and some other stresses, but torque need not be transmitted further, if (and when) that is a consideration.

Thank you for your patience... I hope this turns out useful for someone beside me. I'm sure this forum has a lot of readers not directly looking for their work stuff. While there's no escape my question is a pretty daft one.

 

[desertfox]

hi

Try posting a sketch it speaks a 1000 words.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[drawoh]

THT,

There is no problem building a multi-stage worm gear. The thing will work, possibly at very low efficiency. You can back drive a worm gear, unless you have a low lead angle in the teeth. I am not aware of anybody doing this systematically.

An alternate 90[°][ ]drive mechanism is crossed helical gears.

Sketch?

--
JHG

 

[tbuelna]

As drawoh noted, it is not easy to drive a wormgear using a worm. The worm/gear mesh would need to have an extreme lead angle. And even then the mesh would have poor efficiency. Once the mechanical efficiency drops below about 30%, the gear drive will become self-locking.

A combination of spur and spiral bevel gears would be better if you can make the geometry work.

 

[THT]

Thank you for the helpful replies, and sorry for the delay. And I profoundly apologize for the dim, crayon class image below. Problems with software updates and scanners. I hope it conveys the idea. There's a driving gear, there's the diagonally shaded area of other inevitable spatial obstruction, then there's the two cylinders that need to move (but no torque forwarded further, while the cylinders have some friction and other stresses to endure). "A" is the worm+worm plan, "B" has a bevel gearwheel. The former would be easier looking at space constraints alone. I understood from the replies that the latter might be the one that actually works efficiently.

Some third solution, or "B" would work fine?

 

[drawoh]

THT,

A worm driving a worm may work, but it won't work well. I am not aware of anybody doing this.

If your input and your first output are bevel gears, you will be fine. Your second output will need an additional drive, two spur gears, or some sort of belt drive. The idler shown in your second view can be made to work, but it will be complicated, and I don't see what it will accomplish.

--
JHG

 

[THT]

> If your input and your first output are bevel gears, you will be fine

Can you explain this to me again? Not trying to be extra obtuse but there's something here for me

 

[gruntguru]

The gear fitted to each output cylinder will need to be a spur gear. A spur gear cannot be driven by a bevel gear (the idler in your drawing).

Suggest you have a bevel gear on the input, driving a bevel gear on cylinder 1, then a spur gear on each cylinder. So cylinder 1 has two gears - a bevel input and a spur output. No idler.

je suis charlie

 

[mechedoc]

I realize it has been a while, but I just came across this today.

Another option would be to skip the intermediate gears and make the input a face gear:

http://swissgears.com/en/index.cfm/treeID/411

Then you would have the input connected directly to the first output in mesh with the second output.