Gears for a 150 year old mill

[shapeaholic]

A very good friend has a contract to do some repair/restoration on the "historic" Watsons Mill in Manotic Ontario ( near Ottawa)
One of the many issues that need to be addressed are the worn/broken/loose/mis-shapened teeth on the drive wheel from the turbine to the shaft of the mill grinding wheel.
The setup is; 51"dia - 70 tooth wheel on the turbine shaft driving a 22"dia-30 tooth wheel on the grinding wheel shaft.
From what I can figure, the gears are approximately 1.4dp, with the teeth on the drive wheel being wood inserted into an iron hub, and the driven wheel being entirely iron.
One of the features that can be seen is that the teeth on the iron wheel have a reduced thickness (75%) and the wood teeth are increased in thickness (125%) undoubtedly to provide extra strength.

The wood teeth are "supposed to be" tight, and held in by a dovetailed arrangement, with some sort of tapered dovetailed wedges.
As can be seen this was poorly done during a previous restoration, and now needs some attention.

Can anyone help with some questions we have?
1) what wood should the new teeth be made of? I suspect the current are maple.
2) is the grain orientation important? what should it be ( dumb question but I gotta ask)
3) How tight should the teeth be in the iron wheel ( I suspect a slight drive fit, then wedged)

Sorry the pictures are not better, the lighting is a challenge.

Peter

 

[Joe Michaels]

I'd suggest you get in touch with SPOOM (Society for the Preservation of Old Mills). Their membership deals with this sort of wooden-tooth mill gearing on a regular basis. I have some familiarity with this type of mill gearing from working at Hanford Mills Museum in East Meredith, NY.

The wooden teeth are often made by first making up a set of moulder knives. The teeth are then made using a moulder (or woodworking shaper), in long lengths of wood.

For choice of wood, hard maple is a good choice. If a tooth were positioned at 12:00 on the gear rim, the grain of the wood should run vertically. This puts the shearing force from the mating gear teeth across the long grain. If the teeth were made with the long grain running parallel to the rim of the gear, the wooden teeth would break off as soon as they came under load. I have heard of teeth being made of white oak, which is a longer-grained wood than maple, but not so hard. The belief is the longer grain gives better resistance to the bending load from the mating gear teeth. The wood selected should be a clear piece of hardwood, free from knots or other defects. Old mills are often damp places. Wood to be used for the gear teeth should be air-dried and let stand in the mill for as long as possible before making into the teeth.

Teeth should be a driving fit in the dovetails. The mill gearing I have seen has solid wood teeth, no wedging. If the arrangement of the teeth in your mill gearing uses wedging, the final result must have the teeth solidly anchored in the rim of the gear. The base of each tooth must seat solidly on the rim of the gear.

Your mill gearing uses a different arrangement than some that I have been around. What I see from the pictures is that the teeth and the wedging did not fit tightly, and what look like cardboard shims are seen. This points to some attempt made to tighten the teeth in the gear rim. My other observation is that the shape of the wooden teeth is more of a square 'cog' type of tooth while the iron gear teeth are involute profile. This may have been part of the problem in running that gearing. Iron teeth with an involute profile running against square wooden cogged teeth may have resulted in the loosening of the wooden teeth. From your post, it seems the wooden teeth were made extra heavy vs what the iron teeth are. I'd suggest checking the gear data on the iron gear teeth for chordal thickness at the pitch point vs DP, depth of tooth, and the basic gear parameters. It may be that the wooden teeth, being made 'squarish' are running too tight on the iron teeth. As I wrote, above, I've seen a number of mill gears with inserted wood teeth. Never seen one with square cogged teeth, and the wooden teeth had a more-or-less involute profile (or something that looked like it). If the 'squarish' cogged wood teeth are running on involute iron teeth and setup too tight in terms of gear lash & engagement depth, broken wood teeth and loosened wood teeth (hence the shimming on the wedging) is the likely outcome. The wedging should have held those teeth in solidly.
Wood also swells in a moist environment, so the wooden teeth, if initially set tight in the iron gear rim, should get tighter due to the swelling.

 

[Jim Christie]

While their gearing may be different you might consider consulting with Upper Canada Village and how they maintain their gearing if you haven't already done so .
I posted some pictures of some of the gearing there in this old thread.

Fw: Upper Canada Village and The Time Travelers

Upper Canada Village and The Time Travelers I recently visited Upper Canada Village an 1860 s period pioneer village that was assembled mainly from older buildings that would have been flooded for the construction of the St Lawrence Seaway that has been operating for 51 years . You can read...

www.practicalmachinist.com

My pictures are being held for ransom there now so I have added a couple here and can add some more later if needed.

A few years ago the Algrove publishing arm of Lee Valley Tools republished an edition of this old book

The Young Mill-wright and Miller's Guide: Illustrated by Twenty-eight ... : Oliver Evans, Thomas P. Jones, Cadwallader Evans, Thomas Ellicott : Free Download, Borrow, and Streaming : Internet Archive

I have a printed copy and it mentions how to treat the wood for the gear teeth before they are used
I didn't check the current Lee Valley site to see if the reprint was still in stock there.

algrove publishing the young millwright - Google Search

This video mentions using maple for gear teeth although they are different from the ones you have shown.

PCN Millwright Gristmill Plimoth Plantation : PCN : Free Download, Borrow, and Streaming : Internet Archive

PCN PACTV Community News...The Plimoth Gristmill’s mill recently had a tune up from highly respected millwright Jim Kricker. Mr. Kricker was working on...

archive.org

I'm not sure if Arva Flour Mills are still running their older equipment

Arva Flour Mills - New home of Red River Cereal

Arva Flour Mills - the finest artisan flour from Canada’s oldest continually operating flour mill. The new home of Red River Cereal!

www.arvaflourmill.com

There was an older thread about it here

Flat belt flour mill in modern standards safety shocker

Check this out: 'My heart just sank' Ontario flour mill may shut down over safety concerns - Home | As It Happens | CBC Radio A 193 year old flour mill with 1903 equipment runs into safety code issues. Folksy video to be found here: The Mill Of Arva - YouTube I think the interesting thing here...

www.practicalmachinist.com

Jim

 

[TaperPin]

Looking at the broken teeth, it looks like they are made from white oak based on the coarseness of the grain and radial rays in the grain. White oak is strong and holds up in high moisture areas, whereas maple is not known for keeping its dimensions with humidity changes.

The teeth shouldn’t look square - it puts all the pressure out on the outermost points and doesn’t do as good a job distributing load between multiple teeth further increasing the risk of tooth failure.

My career for 25 years has been primarily a woodworker - I don’t see any decent furniture or architectural woodworker having a problem duplicating those teeth and the tapered locks on not much more than a table saw and planer. It would be tempting to have planer or shaper knives cut for the sides of the teeth, but some teeth should be made and tested for proper contact first. Once the tooth shape of the test pieces is roughed out on the table saw, it doesn’t take that much longer to just make all the teeth. The original teeth were probably just cut on a saw with multiple passes since knives that can only be used for this one project are expensive.

The last person to attempt the teeth clearly knew little about gear teeth and was not detail oriented enough to get the tapers correct - that looks like something a high school kid would make. The angles and tapers are not that complicated, and they should fit with no gaps.

 

[TaperPin]

There are a few other things that came to mind. Wood teeth are not all that durable and would be more prone to cracking if the runout of the two gears isn’t decent - same for bearing slop. Hopefully bearings are tightened up and the wooden teeth should be adjusted to make up for out of roundness of the base iron wheel - it’s not hard to spin the thing and get measurements for each tooth - that should easily get the tooth edges within 1/16” if the bearings aren’t shot.

Notice where the old teeth broke - out on the ends where they have no strength - any woodworker would recognize that these ends aren’t for generating power as much as they are shoulders to hold the tooth against the iron - each tooth should be sanded so it has only light contact out on the ends and it will last much longer. Also, as the teeth wear the wood teeth should be adjusted periodically to limit the pressure applied to the ends.

In case whoever works on this doesn’t know, kiln dried wood lacks much of the internal resin that makes it water resistant. Dry wood is important, but if it can be air dried rather than kiln dried that’s best.

Originally the wood chosen for the teeth would be from old slow growing trees with a lot of rings - good luck finding that now, but you definitely want to look through a bunch of wood to hand pick the best grain. The wedges can be from anything, but a woodworker would choose a quartersawn white oak since it’s easy to work with - and these don’t have to be old growth like the teeth.

If good quality white oak can’t be found, there are a number of very durable hardwoods, but the prices will be stiff - Purple Heart, Brazilian cherry, Ipe.

 

[dundeeshopnut]

Lots of ironwood [also locally known as hardhack] here in eastern Canada. Dense, tough and easy to find without knots. Rumor has it that is used to be used to line the bottom of conveyors as chains and paddles could run on it for years.

 

[MrStretch]

I remember reading somewhere that the pinions in American wooden work clocks were made out of mountain laurel because it's similar to boxwood. Don't know if one can get large enough pieces to make what you need.

 

[newtoll]

Hello everyone, I am the gentleman that has been given the work for the restoration of this mill and this beautiful example of 1860s technology . I have been working on the museum for perhaps 4 years and we have rebuilt some major sections already.

The mill was in operation until the early 50s when it would not sell. the machinery was stripped, and then some of the floors boards too. It was purchased by the Rideau valley conservation authority, and brought to its current condition, Machinery was gathered in the early 70s and with collected machinery and gear works, which are generally quite Plantiful. As these things go, I suspect money was a major concern, you can tell as the works move on up from the subbasement /basement to the first floor,second floor, third floor, an attic . the quality of work diminishes the further from the turbines . Considering what sort of information was available back then on the 70s it’s not a bad attempt at trying to establish an 1860s style Grismill ,although this example is quite modern for it’s 1960s timeframe, for instance, the turbines are actually cast in the city of Ottawa and original equipment. Six turbines in total this one being original equipment 60 hp . Still with the original single log Lignumvitae wood bearings . further down the Rideau canal , which is a UNESCO site is the oldest foundry in Canada in Merrickville and I am suspect as to whether they were the actual foundry, Because the mill sits along the Rideau canal which was established by the British Admiralty these sorts of commodities were available to the interior of Canada,

I’ve actually read the book that Lee Valley republished on mill construction and there are some striking similarities . The original owner, Mr. Dickinson, being from New York State, would have access to the books information. Mr. Dickinson, being of British descent did architecturally style the building, much more after the English style, including turned Doric wooden columns of fir logs on the main floor, similar to the historic dockyards in Plymouth England of 1805 . the iron gear works, which were readily available from Canadian manufacturers in the day are most likely from the Cambridge Waterloo region of Ontario, once known as Gault Ont and generally the epicenter of that type of work

I have toured the example at upper Canada village, and it is merely a small arrangement falling off a little pond, not anywhere near the capacity or technology of this Mill .

Maple most certainly would’ve been used, and in my opinion is the appropriate material. Mostly because of its dense, tight grain northern growth and it’s available . Still good quality material can be had today. My opinion of grain direction is the material should be flat sawn tight grain old growth . If the grain is selected this way, moisture expansion and contraction will occur in only the height, and therefore the mashing of the teeth will always remain the same. Quartersawn Greene would be a mistake . Grain direction obviously needs to be in line with the socket. This would negate milling your teeth in long boards and slicing them up. Profiling will be done after the tooth is established and tested on my Robinson spindle molder . All grind the knives and keep them in stock for future work. I have everything I need, including setting stands for the spindle molder to establish the proper knife . Top and bottom shoulders on the iron gear, provide shoulders for the wedges to drive tight currently and that is the only shoulders engaged, and consequently where the teeth are snapping off. The flower grinder routinely has to tighten them and has been grinding on this equipment for 15 years.

What appears to be there right now is perhaps what I would call someone dropping off a sample to an Amature or local cabinet maker in the 70s who rudimentary perceived the teeth as something simple . it looks as if he used a quarter round bit to provide the shape. Obviously not taking into account how wide the tooth should be or its geometry.


Jack English machines

 

[Jim Christie]

By what I can see on your YouTube site that came up in a search I think Watson's Mill is in good hands

https://www.youtube.com/@jackEnglishmachines/videos

Jim

 

[rustyironism]

It's good to see such interest in the mills.
Next weekend , June 10 and 11, is the Mill Days tour in southeast Pennsylvania.
My friend Jay has done incredible things with his New Park Wind Engines and is part of the tour.
I will unfortunately miss it this year, but if you are in the area, it is a great way to spend a day.

http://www.newparkwindengine.com/wp-content/uploads/mill-days-2017.pdf

Mike

 

[newtoll]

By what I can see on your YouTube site that came up in a search I think Watson's Mill is in good hands

https://www.youtube.com/@jackEnglishmachines/videos

Jim

Most of the videos of the recent work is on my IG feed . Monitor” Jack English machines “ yes, do you like videos there’s lots there . The museum staff appreciates the documentation on my feed, and it keeps them out of my hair.

 

[stephen thomas]

Do you know how old the gears are?
How accurately are the mortises spaced - as cast? or machined?
That is, were they original to the mill?
Or later replacements for all wooden gears?

Are you taking the whole gear to your shop?
By the time iron or steel framed mortise gears were being made, my impression is that many of them were machined much like cast iron gears, in the factory. Carey catalogs from the 20's allude to that.

3 old gear catalogs here that include mortise gears specify maple.
Bond (1925) states: "Our mortise gears are filled with cogs made of specially selected Michigan hard rock maple, water cured and kiln dried."
OTOH, the typical installation would have been a fairly dry factory.

I can vaguely recall catalog cuts showing replacement teeth, but i don't have any paper here on it.

It's easy to imagine having fun setting that up as a (metal) shaper job with an index, & live tooling on the toolslide swinging (wood) shaper cutters.

Or tooling up a versamil, and running it in place on the gear at the flour mill with a simple index.



I used to get slate from a quarry in PA that had a lot of wooden mortise geared equipment, and they were still maintaining it in the late 80's/early 90's at least. You could see very large diameter gears with sections that had new looking teeth in places.

smt

 

[newtoll]

Hi Stephen. The sub basement gear works are original to the building 1860. And from what I can tell machined to the 6” turbine shaft. They run remarkably true for their size, Sockets remain sand casted surfaces, considering the era fairly well done.. most of the smaller gears and right angle gears the teeth are simply castings that employ, even and odd tooth arrangements to self machine in place. The systems run slow in today’s context and it is amazing how horribly something can mash and and still perform rather adequately.

Not necessary to remove the gear and would be a logistical nightmare, requiring major rigging. And there is also the possibility of damaging equipment. Shafting of this particular arrangement is directly off the turbines, lignum vitae thrust bearing bear their weight . One of the turbines bearings was replaced by a water turbine company . One of the smaller turbines add in my opinion were not appropriately selected for the job. Takes a special type of craftsman to have both the ability to work on machinery and wood machining. A Combination that is not common. The vertical position of the shaft is held with wooden bearing blocks that utilize packers made of asbestos. Incidentally, these are in bad condition and Jimmy rig . Tolerances well outside the specifications of the original equipment.

The biggest issues in the lack of a lubrication regime. And recipes of improper, lubrication perhaps, passed down from agricultural remedies of the home farm .

It’s interesting to know that the Timber frame structure within the stone building is circular, saw, and beautiful quality pine. Timber work that holds machinery grear works is of a separate quality. The structure for the machinery is removable and repairable by way of wedges and sockets to the building timber . Theses surfaces are hand planed and painted a Crimson color probably of a milk paint and absolute highly precise double mortise and tenon work in massive 24 x 24“ material of the highest quality pine . Unfortunately, some of the repairs done with chain saw in the past scar this exquisite work.

Jack English machines

 

[newtoll]

Recent work at the mill

 

[newtoll]

Jack English machines

 

[Asquith]

I’ve seen numerous examples of wood/iron gearing in windmills and watermills, and I occasionally run an 1861 steam pumping engine with applewood cogs in an iron flywheel. I hope they remain intact, as re-cogging a wheel with wooden teeth is an immensely skilled job. I have no relevant experience to offer, but I do have some good books.

Even though mortised cast iron wheels may be fine examples of foundry art, it would be far too much to expect that the mortices are uniform, or that the periphery runs dead true, and consequently an immense amount of highly skilled chisel work was needed to get the cogs to be a tight fit, and then to give them the correct geometry and regular pitch. In the 1970s the volunteers at Claverton Pumping Station were faced with the problem of replacing the teeth of their 16 ft diameter waterwheel-driven pitwheel. They lacked relevant experience and money, but had time, enthusiasm, and a variety of skills. The way they fitted the 408 teeth, with a maximum pitch error of 0.005”, resulting in gearing which ran with ‘uncanny silence’, is described in six pages of a booklet, ‘Claverton Pumping Station’ published in 2003 by the Kennet and Avon Canal Trust. Jack – send me a PM if you want to know more.

Fitting the teeth in their wheel required traditional methods, but profiling was done using a 3-axis router. This sounds more sophisticated than it was: they couldn’t afford to buy equipment, so they made use of an old milling machine-type compound slide combined with a borrowed Myford vertical slide, and a home/apprentice-made router head and cutters.

They used oak blanks which had been seasoned for 2 years, but I found a source which recommended the use of blanks which had been felled over nine years!

 

[stephen thomas]

Jack, that is a very cool project, and nice that the iron frame of the mortise gears is orginal to the 1860's!

I was picturing a mill like the mill here in Robert Treman park, which dates to 1839.
All the primary gearing is wooden framed, with wooden mortise teeth wedged in.
It is missing a lot of teeth though the frame seems complete and solid. It will probably never run again.
Sort of sad, because all the machinery above the turbine is mostly intact and operable or at least restorable with "maintenance" that would have been appropriate to the time.

Enfield Falls Mill and Miller's House - Wikipedia

en.wikipedia.org

Regretfully, there is a lot about the park online but very few photos of the mill machinery.

, but profiling was done using a 3-axis router.

This was my suggestion in post #12 to use a Versamil.
Mine is 3 axis, 16" feed table, & came with one big former gear cutter (modified for something else, apparently). Actually 5 axis consdering to rotational positioning axes. Some have an additonal axis besides those listed (quill) though that would mostly be in the way for gear teeth. They are made for similar tasks on metal machinery. Working on wooden teeth would be a doddle after rigging a suitable way to fix and index it separate from the wheel so the wheel comes out round with accurately spaced teeth. The rpms are low for wood working, (IIRC, a high speed head was available)but that might not be a big problem with a large, sharp, wood cutting shop made shaper cutter.

Home of the Portable Machine Tool

VersaMil multiple-purpose portable machining units. Master Milling machines.

www.versamil.com

The owner/manufacturer, Brian Healy is an original PM'r and sometime OWWM'r.
(Like me, he seldom posts on OWWM due to conflicts with owner.)
I also have not seen much from him recently, hope all is well.

Absent that, Versamil's competitor before modern times was Master. They made a lot of building block, stackable slides. Fairly rugged kit.

I imagine there will be loads of hand work fitting the teeth solidly and accurately. Then a bit more after the teeth are shaped fairly closely even if by machine

smt

 

[TaperPin]

The project is in good hands.

Looking at the wear pattern on the iron teeth (rust vs no rust), it appears the most wear has been to the middle, as one would expect if the original wooden teeth were relieved on the ends for light contact, but the last replacement wood teeth are straight so they are now being driven primarily by the outer portions that are the weakest. Who knew wooden gears would be so interesting! Lol

Right now someone is wondering why not 3d scan both wheels and cnc the replacement teeth. No. Let us old guys have this moment.

 

[John Garner]

My understanding is that Brian bought the Versamil product line from Master when they decided to discontinue it.

 

[shapeaholic]

Thanks all for the discussion! ( please continue )
Jack and I will learn lots.

For those who don't know, Jack is a very talented builder, restorer, maintainer of old wood structures.
AND he loves old machines :-))

Peter