single-enveloping worm gear meshed with a cylindrical worm - lead angles vs center distance 2

[Tmoose]

A gear supplier made us a single enveloping worm gear. 14.5 ° Pressure angle, 3" circular pitch, single RH thread, 36.29" Pitch diameter. The customer reports the new gear contact is awful. There are questions about the installation that will be answered in the next few weeks.

In the mean time the new large worm gear has been re-inspected by the supplier.
It was tested deeply meshed with a new worm ( by another supplier), with center distance about .13" closer than the design Cd of 21.770".
In that condition the perpendicularity of the two gears' axes was off by 0.8 degrees.

The supplier said when the gears' faces are held in contact, the gear axes twist as the center distance is increased. Similar to the attached sketch.

My question is - Does a change of 0.8° sound reasonable for a center distance change of .13" ?

Thanks,

Dan Timberlake

 

[spigor]

As per my calculations, a center distance change from 21.77" to 21.64" means 0.28 degrees angle change.
IMHO the test should be made with design center distance to be meaningful.

 

[Tmoose]

Thanks spigor.

I was afraid of that.
Sounds like the gears will still be skewed .5 degrees when installed, which by my naive calculation, with a 4.75" wide worm gear would burn up 0.041" backlash and still run with a cross contact condition with tiny contact pattern at the very edge of the worm gear face.

I will likely download the demo copy of KissSoft and try it on my own.

The customer has no faith this gear will work for them, and right now I am most reluctant to try to convince them otherwise. This is despite some equally grave concerns about issues with the installation's current condition.

 

[lbmakem]

Are you running the new Gear with the original Worm and did your supplier have this worm when they made the new Gear, if he did then the change in the Centre distance would have no effect on the angle as this is controlled by the Worm PCD which would be 7.25".I agree with spigor that a change in the centre distance of .13 would change the angle by .28 degrees but this would change the Worm PCD to 6.998". If your supplier made a Worm to run at the new centre distance and then you used the original worm on assembly you would get the error your customer is getting.

 

[Tmoose]

Both gears were made to original drawings that by all appearances have sufficed to build MANY stoker drives since the 1920s, with no revisions to PD, center distance, pitch, etc.

The installed center distance should be 21.770". The gear housing is very likely one of the three originals supplied in 1965, two of which do not have this problem, reportedly.
The mention of other center distances only arose because the gear house that made the big gear, and recently checked the almost un-used big gear with a brand new worm did of their measuring with the fears deeply meshed. The reported center distance in that condition was about .13" closer than the design Cd of 21.770", and that is when the gear axes are skewed .8° . This is why it is so important to be confident what will happen to the perpendicularity when the gears are installed on the 21.770" center distance.

thanks all,

Dan Timberlake

 

[Tmoose]

Attached are pictures of the "new" large worm gear. An attempt was made to capture the "coast" face of gear teeth opposing teeth with ugly wear towards one edge on the "drive" side. The load is highly one directional, so it is my belief the "coast" side should never touch the worm in service. The rotation of the worm makes the scuffed areas the "exit" side of the worm gear tooth.

My question is, in the pictures does it appear that the gears have experienced "cross contact", as might be expected from .5 degrees or more shaft angle error?

thanks,

Dan T

 

[spigor]

Thanks for keeping us in the loop. It certainly looks like a cross contact to me, with some tooth edge breakage already visible because of local overloading in point contact.

 

[Tmoose]

Man oh man.

This project continues to bear an inexplicable resemblance to this-

https://upload.wikimedia.org/wikipe...0px-Hannibal_crossing_the_Alps_into_Italy.jpg

Anyhow, as I continue to read about worm (gear) manufacturing and inspection I stumbled on this statement in one of Dudley's books -
Several different kinds of worm-thread shapes are in common use. These are -
• Worm thread produced by straight-sided conical milling or grinding wheel
1- Worm thread straight-sided in the axial section
2- Worm thread straight-sided in the normal section

My new question is, does the attached drawing provide information that makes it clear if the worm shape is type 1, or type 2?

thanks,

Dan T

 

[BrianE22]

I was waiting for someone else to post first so I could learn. I ran across the same problem when I needed to have hobs made to replace an existing worn hob. Your drawings don't make mention of the tooth profile or the manufacturing method (like ours didn't). I sent the hob maker a worm and he was able to measure the tooth form. What he came up with matched how we had them made (straight sided grinding wheel).

If you have one, I'd send out an existing worm for inspection. Lacking that can you find out how it was made (ground, milled, straight sided cutter, lathe???)

 

[spigor]

Tmoose
These three types you have listed are the most common indeed. I list them again to have them numbered:
0- Worm thread produced by straight-sided conical milling or grinding wheel
1- Worm thread straight-sided in the axial section
2- Worm thread straight-sided in the normal section
Looking at the drawing, with tooth height of 2" you probably have type 0. Type 1 is made with turning and usually used on small worms and helix angles up to around 2 degrees. The exact geometry of a type 0 worm depends on tool diameter. Type 0 made with a infinitely large tool is equal to type 2.
I have once calculated and compared the exact geometries of these 3 types for a worm very similar to yours. The differences between geometries were up to approx. 20 micrometers. It means they are unlikely to be the cause of the problems you are reporting with this worm. I do investigate the microgeometry of the worm only when I need to have control over the contact pattern it will produce with its mating worm gear. Here it looks like there is a problem with the macrogeometry, like a 20 degrees cutter being used, or a normal pitch cut instead of the axial pitch being the most common.
Time to get a better gear cutter?

 

[Tmoose]

The folks that made the big worm have made tooling and yesterday tested the contact patterns of their worm gear when meshed with a new worm made by the same folks that have made worms for us for several years. My vague understanding is for the pattern testing the new worm has replaced the hob in the machine used to make the large worm gear.
FWIW they were provided with a brand new worm specifically for testing the worm gear when they made it 2 years ago.

This is a report from late yesterday. - "Once we set the CD (center distance) the mesh pattern has moved to the top of the tooth flank and is no longer in the center as it was where we were at full depth."

 

[Tmoose]

Attached is an image including actual pictures of the recent contact pattern test results and scuffing wear on the gear after test running by the customer.

These are the Gear supplier's manufacturing engineer's most recent comments -
"This was not what I was expecting ........... on this gear set!!!"

"I am going to get the cutter on order and that will give us several days to come up with suggestions on how to repair/recut this (large worm)gear."

======.

From our 2014 PO to the large worm gear supplier.
" XXXXXXX ( the company I work for ) to supply the (large gear ) casting and the mating part ( the worm ) to test fit. "
"Mating part ( worm) to be returned with the finished machine part ( the large worm gear."

Thanks for all your help.

Dan T

 

[Tmoose]

The supplier annealed the gear, "adjusted the angle" and recut the worm gear. The attached image shows the new patterns with the worm and gear at the correct center distance. Recall that My vague understanding was for the pattern testing the new worm replaced the hob in the machine used to machine the large worm gear.

The supplier emphasized that when testing the pattern this way the gears do not drive each other. They are each rotated (by the hobbing machine tool?).

The pattern now looks reasonably well centered, unlike the images in my July 12 post.
But the broad contact is out near the gear tips on one side of the gear teeth, and kind of narrow and hard in a narrow band near the root of the opposite side of the teeth.

Sometimes I can imagine that, with forced synchronized rotation for each gear, the worm //might// be shifted so it is sort of hovering shifted axially "backwards" in a zone somewhere between actually driving in either direction, with a tendency to "touch" with the patterns provided, and we should accept the gears.
Then I think that if the worm is floating in such a "zone" it has no business touching the gear AT ALL, and the "contact patterns" are indicating that now the pressure angle or something is messed up.

All comments appreciated.

thanks

Dan T

 

[spigor]

Something is not right, especially on the side with narrow contact pattern. They probably shifted the worm meshed with the gear until they touched and got pretty close to the true contact pattern. Did they measure the backlash and if so- is it ok? You say that the gear has been annealed- is it a steel on steel application? Is the worm harder than the gear? Is the side with the narrow contact the coast side? I have seen a gear drive with a narrow contact, as soon as we gave a little more power the oil started to burn and first signs of seizure were present. If they did not get the angle right the first time, is there a chance that the other dozens of critical parameters sitting in the cutting process and the tool are kept right?

 

[Tmoose]

Hi spigor,

The gear material combo is kind of bizarre.
The Worm was and is unhardened steel.
In the previous millennium, when the gear drive was recognized as the premium product of its kind, and a big seller, the gear started as "cast iron".
After a few decades, and the demand had shifted to another type of drive, curiously ( to me) the material was revised to flame hardened mid grade ductile iron.
The reasons for the gear material revisions are now lost to history, but I believe the fairly costly changes were likely made in response to a wear problem, despite the resulting highly non-traditional material combination.


I have asked that the supplier evaluate the backlash, suggesting feeler gages on both sides of the gear teeth.
It is a one direction drive with a real heavy, lazy load with lots of friction (traveling grate) so I envision no chance of load reversal in service. Thus I don't see even a bunch of extra backlash by itself being a problem when running.

One new, additional problem I am concerned with is the gear is designed completely symmetrical to be flipped after a bunch of years to provide 2X the life.
On the newly re-machined gear I think I see flats or facets on the gear flanks with nearly line contact near the root of the teeth.
I am imagining they tweaked the hob angle or ?? to broaden and center the contact pattern and then only cut one side of the teeth to 100% clean up. The re-angled hob just carved off material on the "high spot" on opposing tooth down as a coincidence. Maybe that means the backlash will seem low if tested now by rotation, until the opposite flank is cleaned 100% and the facet is removed.

I better send an e-mail emphasizing both sides of each tooth must be machined to clean 95% because the drive is essentially bi-directional ( due to flipping for extra wear life ). I am hoping but not really expecting that that will fix the root contact.

All comments appreciated .

thanks

Dan T

 

[spigor]

Increasing the backlash beyond the designed value by simply cutting deeper usually ruins the contact pattern.
Any steps and edges on the flank surface will be detrimental, reducing the contact area, causing localized power flow, disturbing the lubrication.
It doesn't look like a good done job.