Worm Gear Gearbox

[Jizvonius]

I will vouch for the workability of the worm gear gearbox in a drivetrain. In 02 my high school team tested it. Just make sure that you pay attention to the forces that your gearbox will be subject to, and lube like a madman. The forces don't work out the same as with spur gears as there is more axial force on the worm gear and worm, so putting a large amount of force through a worm gear gearbox can cause an improperly considered structure to flex and bind even if you were precise in your machining. (it can still work just design the box structure and shafts to take the forces)

Also make sure that you check the torque ratings on your gears. The fact that they don't backdrive also means that much of the shock of hits goes soley to the teeth of the worm gear.

We also got our gears from boston gear.

[Peter Matteson]

Last year we used worm gears to power our shoulder mechanism for the first part of the UTC regional. We removed the arm primarily because it unbalanced the robot.

We adapted the FP motors to go into the Bosch transmissions. On the output of the transmission we had a six point socket attached. The socket interfaced with a bolt that the worm was welded to. The worm assembly was setup with 4 bearings, 2 flange type radial and 2 thrust. The worm wheels were soid mounted to our shoulder shaft and the motors and transmissions were mounted to the arm. It worked well for moving the arm and locking it in place with no back drive, but alas few ever saw it work properly for many reasons.

Pete

[Jim Schaddelee]

Team 107 also used a worm drive transmission a few year ago,it work ok but there are things to be aware of. The main issue is your reduction should not exceed 50 to 1 on the worm ,for the reason that it will not back drive or coast. You can compensate for this by adding a ramp down in your code, but in humble opionion worm tranmisson is better suited for arms or lifting or any application where there are not sudden loads ,like stopping a moving robot.
I may still have one of the transmission we used if you would like me to sent it to you to look at I would be happy to help. I will try to post some pictures of a new worm tranny I designed over this summer If anyone would like to see it.

thank jim schaddelee

team 107 Holland Christian / Metal Flow

[dddriveman]

Originally Posted by Jim Schaddelee

Team 107 also used a worm drive transmission a few year ago,it work ok but there are things to be aware of. The main issue is your reduction should not exceed 50 to 1 on the worm ,for the reason that it will not back drive or coast. You can compensate for this by adding a ramp down in your code, but in humble opionion worm tranmisson is better suited for arms or lifting or any application where there are not sudden loads ,like stopping a moving robot.
I may still have one of the transmission we used if you would like me to sent it to you to look at I would be happy to help. I will try to post some pictures of a new worm tranny I designed over this summer If anyone would like to see it.

thank jim schaddelee

team 107 Holland Christian / Metal Flow

I would love to take a look at that worm tranny. If you do not mind sending it to me. You can find my email in my profile.

[Matt Adams]

Originally Posted by dddriveman

I would love to take a look at that worm tranny. If you do not mind sending it to me. You can find my email in my profile.

I don't want to jump in and be a nay-sayer... but I think one of the things you need to really weigh in before considering a design using a worm gear is the immense efficiency loss you're going to have simple due to the nature of the gear set.

We're talking in the range of efficiencies as low at 30% !

The advantage you mentioned, about not being able to back-drive the gear set, is not always true. There is some critical pressure angle and coefficient of friction that leads to the point were you actually can back-drive this gear set. Jim said that it's above a reduction of 50-1. I can't verify this either way.
[edit: But Paul definitely can. Rock on!]

However, simply because of the efficiency issue, I would steer clear of them in a drive-train gearbox unless there's a really good engineering justification. For a large reduction in a small amount of space, an off-the-shelf planetary gear set would probably hold better efficiencies.

That's my two cents. Good luck

Matt

[Paul Copioli]

Back Drive-ability is a function of surface finish, lubricity, and worm lead angle. Using a bronze worm gear and a hardened steel worm (the small screw looking one), the max lead angle is 5 degrees to insure self locking. This is not a guarantee! What is a guarantee is that above 5 degrees, the worm gear set will not lock. The lead angle is the reason the window motor locks and the van door motor does not lock. It really has nothing to do with the gear ratio. A gear ratio of 50:1 with a lead angle of 15 degrees will not lock.

However, the more thread starts on the worm, the higher the lead angle and the lower the gear ratio is. Remember, gear ratio for a worm gear is # of teeth on the gear divided by number of thread starts on the worm.

-Paul

[greencactus3]

a little off, but for lego, wormgear 24:1 reductions work very well, for lego robots i mean. not too slow, yet (well since lego isnt efficient whatsoever,) efficient (make sure you wupport both axles from both directions, and since lego axles are plastic, support them as close to the gears as possible.
well, yea. worm gear gear boxes (wormgearboxes?) as an idea are very applicable ithink. except as with designing any kind of gearbox, will have difficult parts but if it works well, great!

...and yea, im understanding why my english teachers rarely give me any good grades on my essays.

[Adam Y.]

Originally Posted by Matt Adams

I don't want to jump in and be a nay-sayer... but I think one of the things you need to really weigh in before considering a design using a worm gear is the immense efficiency loss you're going to have simple due to the nature of the gear set.

We're talking in the range of efficiencies as low at 30% !

Actually efficiency might not be an issue if you properly design the gearbox. A little bit of engineering and you should be able to get the efficiency fairly high. Of course a little research shows that the efficiency can reach 60%.
Article
Ps. Please someone build a drivetrain using these gears. These are the funniest sounding words ever. Hypoid and globoid.

[Cory]

Originally Posted by Adam Y.

Actually efficiency might not be an issue if you properly design the gearbox. A little bit of engineering and you should be able to get the efficiency fairly high. Of course a little research shows that the efficiency can reach 60%.
Article
Ps. Please someone build a drivetrain using these gears. These are the funniest sounding words ever. Hypoid and globoid.

60% efficiency is still downright terrible for a drivetrain when you consider spur gears are 98% efficient and chain is 95% efficient

[Adam Y.]

60% efficiency is still downright terrible for a drivetrain when you consider spur gears are 98% efficient and chain is 95% efficient

Read the article. It's actually possible to reach that degree of efficiency using regular worm gears.

[Paul Copioli]

Here is a picture of a hypoid gear. It is used on an automotive painting robot that my company manufactures.

[Paul Copioli]

Here are some facts about worm gear efficiency:


Efficiency = Tan(Lead)*(1-f*Tan(Lead))/(f + Tan(Lead))

Lead = Worm Lead Angle
f = Coefficient of Friction (Mu)
For a bronze gear and a steel worm f = 0.05
Easy to purchase worm gears have lead angles ranging from 4.5 degrees to 22 degrees. Here is the bounding range on efficiencies

Lead angle of 4.5 degrees has an efficiency of about 61%

Lead angle of 22 degrees has an efficiency of about 87%

This assumes everything is lined up!! Tolerances are very critical for a worm gear set-up.

The lead angle of 22 degrees corresponds to worms with 4 thread starts and the lead angle of 4.5 corresponds to worms with 1 thread start. The worm with one thread start has 4 times the gear ratio for the same package size than the worm with 4 starts. A worm gear box has several advantages for FIRST robots. The main two are package size and weight. We have used worm gears on our robots for three years, but only once on our drive transmission (see white papers under CCT).

[Matt Adams]

Originally Posted by Adam Y.

Actually efficiency might not be an issue if you properly design the gearbox. A little bit of engineering and you should be able to get the efficiency fairly high. Of course a little research shows that the efficiency can reach 60%.

I'm going to have to strongly disagree about efficiency not being an issue.

Even with your research, which I won't refute, 40% losses in a single stage of a gear set is still quite large compared to the 10%-15% losses over an entire drive train with spur gears and chain.

It should be noted that Adam's article did cite a 97% efficiency for a multiple stage gear train using worm gears, which I found to be quite impressive. (Costs? Yikes! )

Worm gears have their place, but I would strongly discourage their use on a drive train if you're just looking to reduce your gear ratio. I'd definitely like to see justification addresing the losses, which we can all admit are several times larger than spur gears.

(See Paul's link for an appropriate application.)

Matt

[Peter Matteson]

The worms we used last year were 35:1 on the back of another transmission. I don't recall the ratio for the Bosch tranny we used however. We wanted an extremely high torque setup for our application though.

Pete

[Jim Schaddelee]

Here is a design for positioning worm transmission from team 107 , the nice thing about our design is that there is 50 to 1 reduction on the worm which should make it not back drive. yet you can quickly change speeds and motor combination. It also can have either a pot or encoder mounted simply.



Thanks Jim schaddelee team 107