Bilgram bevel gear generator

[Dan Rowe]

My Shay locomotive research lead to researching early bevel gear production methods. Shays were produced from 1880 till 1945 so just how did they make the bevel gears was a leading question in my mind.

My experience in geared locomotive patents led me to try the same procedure for bevel gear patents. Several of the US bevel gear patents are very interesting and could be used in a home shop to produce accurate historical gears or ones that no one could tell the difference.

The early bevel gear machines were template planer machines. One of the earliest one of these I found was a machine designed by George Corliss to make bevel gears for his steam engines. US Patent 6,161
http://www.google.com/patents/about?id=ps1EAAAAEBAJ&dq=6161
These machines were slow and cumbersome.

The fully automatic version of a bevel gear template planning machine was invented by William Gleason. This machine increased the production speed of bevel gears considerably US Pat. 175,895.
http://www.google.com/patents/about?id=XapTAAAAEBAJ&dq=175859

The next major improvement was a radical departure form the existing machines. The simplest form of Hugo Bilgram's gear generating machine is really only an attachment to a shaper. US Pat. 294,844.
http://www.google.com/patents/about?id=n6diAAAAEBAJ&dq=294844
Hugo Bilgram was given the John Scott Award and the Elliott Cresson Medal for the invention of this machine in 1887.

A bit of study of the patent drawings shows that it would be possible to make a bevel gear attachment for a shaper in the home shop. I have generated a spur gear using the method in my shop. I was not all that happy with the results but the disappointing result was a learning curve. I will post some photos of that attempt and I might even give it another go to fix where I went wrong.

An explanation of the spur gear generaton process with a shaper can be found here:
http://myweb.tiscali.co.uk/remark/pages/workshop/projects/geochuck1.html

The bevel gear generation method is very similar. In the spur gear method a drum with the diameter of the pitch circle is attached to the same spindle as the gear blank. A wire cable is wrapped around the pitch circle drum and both ends are attached to something solid. As the shaper table is moved the gear blank rotates the same distance along the pitch circle and the one tooth rack mounted in the shaper clapper generates a proper involute curve.

The bevel gear arrangement is very similar only instead of a cylinder for the pitch circle and a linear table a cone is needed that rotates on a rotary table. The one tooth rack in now a one tooth crown gear that generates the bevel gear.

Hugo Bilgram also invented a very useful valve gear diagram, which I find much simpler to use and understand than a Zeuner diagram. Bilgram's valve gear diagram is in several slide valve gear books and I finally tracked down a copy of "Slide Valve Gears" by Hugo Bilgram ME 1878. This is available on the web but I like real books to study from.
http://www.archive.org/details/slidevalvegears00bilggoog

I have a few engravings of Bilgram machines and I will post the images soon but the first Bilgram patent is the key to understanding the machine.

I uploaded an article about the Bilgram bevel gear machine that is from the American Machinist May 9 1885.

Dan

 

[Dan Rowe]

Here is a simple drawing of the main part of a William Gleason bevel gear planer. I think that it would be a simple machine to construct. The cutter could be replaced with a flexible shaft machine so a rotery cutter could be used. Hand power would move the cutter and hold it to the template.

A cnc mill could be used to make the template. The machine could be used to make wood patterns then dress the final casting.

Dan

 

[Dan Rowe]

I know you guys want photos so this is all I could find of my first shaper gear generation try.

I used a lathe headstock and three jaw chuck to hold a shaft that had the pitch circle drum and the gear blank.

The end view shows my index system that is another 40 tooth gear and a simple index pin.

This is the set up on my South Bend Shaper with the first pass made. I was hopefull that the gear would look like the one on edge.

Well that did not happen. ???

Instead I made this

Well that was a big disappointment. I cut the whole gear with generation only. As I did that for all 40 gaps I must have experimented with different cutter depths from partial depth to full cut. I knew something was going wrong after the first gap but I decided to complete the exercise and then figure out what went wrong.

The cuts are fairly uniform unfortunately they have little to do with gear teeth. I think that if I had cut all the teeth to depth with out any generation motion first it would have worked. The cable was not nearly as rigid as I thought. The cutter force of the in cut pass was way greater than the cutter force on the out pass. This is what I believe caused the odd shape teeth.

Dan

 

[Dave Sohlstrom]

Howdy Dan long time no talk to

I have been playing with a attachment for my shaper to cut spur gears using the same method that you are using. Plan on using music wire just as tight as it will stand. PD templet will be on the ram side of the table as you have done but will put indexing at the other end of the table with the gear blank or blanks will be between the two. Templet and indexing will be linked.

From what I can derive from the patent drawings they are setting the cutter to the full depth of the gear and only running around once.

Looks like that first gear of yours would make a great turbine wheel

Dave

 

[Dan Rowe]

Hi Dave,
Yes a full depth cut is what I see in the drawings and that is what I believe the gent I linked in UK is also doing. I did not have the nerve to do a full depth cut the first time I rolled the blank past a moving shaper ram. If you check the link of the gent in the UK you will notice that he has rigged a very sturdy support for the wire on the pitch drum. I attached one end of the wire to the wall which had 2" x 6" studs with a steel bar spanning several studs. The other end was secured to a 1" pipe that was fitted from the floor to the overhead. The pipe end is the end I suspect flexed on the in-cut. It is proof to me that a very rigid system is needed to make this work with a single generating pass.

Dan

 

[Dan Rowe]

The Bilgram method is not restricted to a shaper only. My next test will be on a 4 axis Taig mill set up.

The sides of a 20 degree rack are 20 degrees from the vertical. This means we can use a 20 degree per side tapered end mill in a vertical milling machine. I plan to try full cnc but there is no reason that the wire method could not be used with a manual mill. The wire would go a along the long table axis and the short axis serves as the shaper stroke.

My test in the shaper used a 29 degree acme thread tool I made on a ship for valve stem use. That way I did not have to grind a special tool. I am very sure that this is not the first time that has been done to generate a 14.5 degree pressure angle gear on a shaper.

Dan

 

[dazz]

Hi
I think the problem with cable is that it can flatten and stretch. I suggest using a ribbon of steel (not braid) or many strands of solid piano wire running on a flat pulley and each individually tensioned.

 

[TSutrina]

My suggestion is to use steel banding for packaging or used for cut cores for electromagnetic components, over a cable or music wire. Stiffness is the tension and the cross section. The bend radius is determined by the thickness. Steel banding is made to bend around a curve and return to the original straight shape

 

[Wizard69]

For those of you interested Gleason Works still operates in Rochester NY!

I know that doesn’t help much as nobody will be buying a new gear generator from them. At least not for the home shop.