Damaged Gear

aeichman · Dec 28, 2017

Hey guys,

I'm working on an old aircraft instrument that uses a series of PCB Gears as encoder rings. The problem is that a PCB board as a gear isn't really a great idea. The material isn't strong enough to hold up to regular usage over 30+ years.

Right now I'm working on reverse engineering the gear, which has gone on pretty smoothly. Using the mating gear, I was able to determine all the appropriate measurements for the gear and confirm it (roughly) with a pair of calipers.

The problem I'm having is with the center hole diameter. Since the material isn't that strong, it's been worn down in three general three areas (where the asymmetrical nylon pulleys hold onto the gear) and is no longer perfectly round. This inner diameter is unfortunately critical for the part. There isn't any documentation for the gear and the manufacturer has been out of business for the better part of 20 years. If you guys have any suggestions on how to measure the original diameter (if it's possible), it would be greatly appreciated.

scojen · Dec 29, 2017

The center hole diameter what does it fit on that's where I would start I am guessing this is an encoder like device possibly for a navigation aide. Center to center distance for the gear mesh is as important as the mesh itself but I would not rule out a 3D printed part.

DDoug · Dec 29, 2017

aeichman said:
I'm working on an old aircraft instrument that uses a series of PCB Gears as encoder rings. The problem is that a PCB board as a gear isn't really a great idea. The material isn't strong enough to hold up to regular usage over 30+ years.

Says who ?

Micarta gears are standard in allot of applications.

I would send to a gear expert (like a couple of members on this board) and pay
them to get the measuring right.

Scottl · Dec 29, 2017

If it's worn in three areas does that mean there are three areas that still have the original dimensions? I was thinking in that case a plug gage could determine the original size. If you don't have anything that size perhaps turn an aluminum plug and keep test fitting between very light cuts?

steve45 · Dec 29, 2017

I'm a little confused. You're saying that the gear doesn't actually fit on a shaft, but rather there are 3 rollers that hold it in position?

What exactly is this used on?

Zahnrad Kopf · Dec 29, 2017

Maybe I missed it, but exactly why can't the mating part be measured to determine what size your bore needs to be? It would seem the simplest avenue of approach and least susceptible to introducing speculation and uncertainty.

gbent · Dec 29, 2017

The simplest way to measure a tri-lobed hole is a 3 point hole mic. A low contact pressure method that should have greater accuracy would be a coordinate measuring machine(CMM). A CMM can measure an arc segment and determine the radius and centerpoint location. The smaller the arc segment, the greater the error.

aeichman · Jan 2, 2018

The PCB board gear on this has never held up. We used to just replace them, but the manufacturer and anybody who knows anything about these instruments are long gone. The reason we're replacing them is that the teeth keep breaking.

The mesh with the mating gear is too tight, so we're going to reduce the pitch circle, but that inner diameter is crucial for this part. We have the center to center distance, but that doesn't help with the fit on this.

It's off a fuel boom position for a refueling tanker. The gear used to fit perfectly, but after years of use the weak board was worn down. We don't know where the board was worn down, so we're just spitballing by this point and trying all our options. I pitched the idea of sending it out, but I work with a bunch of old, stubborn engineers. I don't think that's happening anytime soon.

I think our best (and cheapest) bet would be to turn down an aluminum plug and continue light passes until we have something that fits.

I'm brand new to the world of gears, so any other thoughts or wisdom is greatly appreciated. Thanks everyone!

SIP6A · Jan 2, 2018

You say that it rides on three rollers. Could you measure the center distance of the shafts where the three rollers ride and then do the math to get the od of the gear?

michiganbuck · Jan 2, 2018

likely the ID needs to be Concintric to the OD..so just fitting the bore would not make that.. Perhaps true the Id to the OD then make the plug..With finding a way to attach the plug to the gear .

Making a whole new gear should not be a huge problem.

This part a throw away part to protect the mating part IMHO..what is the mating part made of?

asymmetrical nylon pulleys ..how many pulleys?

aeichman · Jan 2, 2018

The mating part is steel. We have every dimension figured out, but the design of this instrument is complete garbage. Everything is fixed and off center in some way. It makes no sense whatsoever and has resulted in more problems than anything else.

aeichman · Jan 2, 2018

The OD isn't the problem, it's the ID we need of that center bore.

michiganbuck · Jan 2, 2018

I'm thinking if the Id runs perhaps .020 out or round to the Od it will likely hang-up or wear to only one side...
Checking that the ID is to center may be good..or making the new ID true to the OD..

With the Id wore out it may need to be checked that it is still center or made to be that.

tdmidget · Jan 2, 2018

If there are 3 places on the ID that are not worn then a CMM should be able to give you the number. It takes 3 points to define a circle but if you have more then even better.

HuFlungDung · Jan 2, 2018

It's great to be able to ask a question on a difficult topic.

It's even greater when the question is clear and unambiguous to someone who has no clue what you're referring to. A picture of the 'hub' or whatever this thing rides on would be nice to see.

Milland · Jan 2, 2018

Are you describing non-cylindrical pulleys, or three pulleys of different size used as the constraints? If the "asymmetrical nylon pulleys" are not concentric units why is this the case?

Rather than use FR4 for the new board, perhaps a more robust plastic could be used, such as PEEK. If there's certification issues that prevent this, then a regular replacement program should be introduced to prevent degraded performance risking a flight.

aeichman · Jan 2, 2018

My apologies on not being clear with the pulleys. There are three pulleys on the instrument. They are cylindrical, but the position of them on the instrument is not symmetrical in anyway on the instrument or for the gears. I'll see if I can get an image to clarify things a bit more. We were thinking of S1000-2 or an aluminum core with copper coating. The ID roughly measures 1.383" and varies by .03"

HuFlungDung · Jan 2, 2018

Ok, so these 3 pulleys are spinning on some sort of axles. You will need to measure the center distances of all three pulleys with respect to each other. The easy solution is then to fire up your CAD and plot those points and draw an arc on those three points, deduce the diameter of the circle those axles are on, then add the offset for the pulley radii to the radii of the circle the pins are located on. Or you can trig it out the hard way.

Milland · Jan 2, 2018

aeichman said:
My apologies on not being clear with the pulleys. There are three pulleys on the instrument. They are cylindrical, but the position of them on the instrument is not symmetrical in anyway on the instrument or for the gears. I'll see if I can get an image to clarify things a bit more. We were thinking of S1000-2 or an aluminum core with copper coating. The ID roughly measures 1.383" and varies by .03"

More pictures would be very helpful. I'd not use any measurements of the current board as canon, just for gross reference. The measurements you get off the mating part are what count, as well as the distance to the pinion gear the teeth mate with. If anything, this number is most critical to prevent binding or improper readings.

On the board material, if the Al/Cu version could work that might be the way to go. PEEK would have the advantage of being lighter and without risk of shorting to the circuit traces.

michiganbuck · Jan 2, 2018

QT:[have any suggestions on how to measure the original diameter (if it's possible), it would be greatly appreciated.]
Perhaps measure all around and find the thickest area from OD to ID and use that for reference to imagine the part with Id and OD running true to each other and at that diameter decide if that would work.

Seeing now that it is a very small part think I would just make a new one perhaps out of Nylon...Good to make a proper print of the part for reference..

from here it looks like a few teeth missing at 11:00 so trying to save that part not so good IMHO.

If the ride-on pulleys are fixed then centers to centers and add the size of pulleys should get a very close number.
that circle to the mating gear should also be not that difficult..

Measure and draw the assembly is often the best to figure things out..Pulley centers, pulley size = that circle, the gear shown as measured at least wore places, the distance to driving gear center, driving gear diameter and root diameter, the amount of gear lash desired..

Damaged Gear

Plastic

Stainless

Diamond

Diamond

Stainless

Diamond

Diamond

Plastic

Titanium

Diamond

Plastic

Plastic

Diamond

Diamond

Diamond

Diamond

Plastic

Diamond

Diamond

Diamond

Similar threads