Casting a windmill gear in grey iron

Discussion in 'Sand Casting' started by Mister ED, Nov 17, 2018.

  1. Mister ED

    Mister ED Silver Banner Member

    As I mentioned in an older thread, I have a couple of 100 year old windmill gears to replace or fix. In the past, I have always brazed up teeth and then recut with a gear cutter on my mill. Brazing that up is a PITA on this coarse of a gear. And since it is a very coarse gear, I don't usually completely fill the space between the teeth. This creates a very interrupted cut when milling (which is rough on my smallish mill).

    So the thought crossed my mind ... since brazing rod is brass or bronze, I wonder if I couldn't pour brass or bronze right around the gear. This would infill all the teeth and give a nice uninterrupted cut when milling.

    Obviously, the gear would need to be:
    • preheated
    • well fluxed
    • quickly put into a mold and poured
    Any thoughts? Obviously with the cold and snow, my mind is already wondering towards next spring, LOL.
  2. PatJ

    PatJ Silver Banner Member

    Why not just pour some new gears or gear blanks in gray iron?

  3. Mister ED

    Mister ED Silver Banner Member

    Pat, that is definitely an option. I am a bit cautious only because to date, I have only had a couple of aluminum pours ... so I am not sure I am ready. I have the furnace for it, and have a new pressurized oil/diesel tank so I can have a more steady fuel flow. Even have a brand A10 new super salamander waiting for iron. Casting the gear in iron was actually part of an earlier thread.

    Like I mentioned, the idea just crossed my mind since brazing is sticking molten brass to iron (in this case) .

    Whichever way I do this, I had my neighbor print a couple gears up and we will probably do the same, up-sizing a bit, to allow for shrink and final machining. We will also have to play around to try and give it some draft.

    I have been watching your furnace thread closely, and trying to learn from your findings. If you need something to practice an iron pour on ... let me know, LOL.
  4. PatJ

    PatJ Silver Banner Member

    I am game, assuming the gear blanks are not 0ver say 6" diameter and can ship in a flat-rate USPS box.
    I do need practice with the iron.
    No guarantees, but I am pretty sure I can do it.
    Not positive of the exact surface finish, but you would generally machine them anyway, and they would have ferrosilicon so should be machinable.
    It is wet season, but I think I could cast them on a good day.
    Send them on.
    Tobho Mott likes this.
  5. Jason

    Jason Silver Banner Member

    Ed, I'd take that offer and RUN WITH IT!
  6. Mister ED

    Mister ED Silver Banner Member

    Oh, Pat and I have been messaging back and forth. ;)

    But, I still have to wonder if a person couldn't overlay brass/bronze onto iron by pouring instead of dabbing it drip by drip via torch.
    Last edited: Nov 20, 2018
  7. HT1

    HT1 Silver

    explain what you are trying to do? if you want to make the Poured metal an actual become a homoginous part with the original piece... that will not work. the part would have to be preheated to the Melting point, or extremely close to it, your Flux would have no place to off gas so your metal would look like swiss cheese, you might be able to make it work under some sort of inert
    you could place dovetails or holes on the Part and get an inter-cast interlock, I have made Aluminum Handles for Knives this way .
  8. Mister ED

    Mister ED Silver Banner Member

    This is what a typical 100 year old iron windmill gear looks like. Generally I will braze up the teeth to the red line (excuse the lack of artistic ability) and then recut the profile on the mill. Pinion gears are one thing, the bull gears are 3-4x bigger and generally have more unique features built into them.

    Now what I was referring to, was molding up a round cylinder, centering the heated & fluxed iron gear in the cylinder and pouring brass or bronze around it (below pic). This would actually make the machining process much easier ... turn down on the lathe then start machining teeth in a nice uniform blank. The unevenness of the braze on a brazed gear makes machining a bit difficult. For brass or bronze, I don't think the base metal should be molten (it isn't in the brazing process). Certainly the temp of the iron would need to be controlled to properly wet the iron. But that is a good point about the off gassing from the flux.

    Just for reference, this is what a new gear would have looked like. My neighbor printed this one in plastic (opposite side of the gear from the first pic).

    Maybe it is not possible to do, and I can understand with most metals it would not be. What got me thinking was just the fact that brazing is layering molten bronze or brass on a heated substrate.

    I may try it next spring, just for the heck of it. I need to get my but in gear, finish the wife's kitchen, so I can get a bit more experience and progress to casting iron.
  9. HT1

    HT1 Silver

    Um if you have a lathe and A mill, why dont you just procure a piece of stock and machine a new gear??? all I can think of is you are trying to maintain the "original" but it seems like you are trying a difficult route.. if the gear has some special features, then cast your own gear blank and machine it .. I've been involved in gear repairs in the past ... Braze repair, full nickle weld ups on cast Fe, casting Blanks, and turning Blanks...
    the thought patterns on whether to repair or replace normally involve: unusual material, if the gear is some obscure material... repair. Size , small gears replace large repair... Complexity if the gear is going to require some serious machining just to get the blank to shape... repair .. .

    But your concept of "cast brazing" could only work with a permanent mold probably open faced, the blank would have to come out of an oven at the appropriate base metal heat for Brazing (around 1300F or cherry red), go into the Mold and then be poured quickly, since you cannot work the heat or look at individual work areas, there is no chance of you seeing any unbonded areas till you hit the Mill.

    BTW the process you are thinking about is similar to the process used to reline babbit bearings... go look at that . its hateful, dangerous and you are talking about over doubling the heat involved, I'll pass

    if you are doing many of these gears I would recommend you improve your brazing process, by putting the gear on a rotary wheel, with fixed preheating torches. you bring the tooth in question under the preheat torches, when it comes to temp rotate, while you are brazing that tooth, the next is preheating , you keep rotating and brazing, you need a quick kill valve for the preheat, and adjustment valve readily at hand for finishing up the gear, you might have to arrange a cooling method for finished teeth ...lp air say 20PSI ... I would even fixture the braze torch so it has a restricted range of motion and is mostly self supporting , so you wont get tired while working ...I have alot of experience in this sort of thing semi-production tungsten carbide brazing. if the work is repetitive jigs and fixtures are your best friend ever!

    V/r HT1

    [QUOTE="Mister ED, post: 10621, member: 65"
    Now what I was referring to, was molding up a round cylinder, centering the heated & fluxed iron gear in the cylinder and pouring brass or bronze around it (below pic). This would actually make the machining process much easier ... turn down on the lathe then start machining teeth in a nice uniform blank. The unevenness of the braze on a brazed gear makes machining a bit difficult. For brass or bronze, I don't think the base metal should be molten (it isn't in the brazing process). Certainly the temp of the iron would need to be controlled to properly wet the iron. But that is a good point about the off gassing from the flux.


    Maybe it is not possible to do, and I can understand with most metals it would not be. What got me thinking was just the fact that brazing is layering molten bronze or brass on a heated substrate.

  10. PatJ

    PatJ Silver Banner Member

    We are going to try and cast it with the teeth on it, and no draft.
    I think I can do it if we can get the shrinkage right.
  11. Mister ED

    Mister ED Silver Banner Member

    Thanks HT1. Yes, in any of these restorations we do try and keep as much original, as possible. For the gear in the pic above, procuring some Dura-bar (cast iron) and machining was one of my options. For this one, Pat J is going to try his hand at casting it.

    Thanks for putting my mind at ease that the "cast brazing" could potentially be possible (which is what I was struggling with), but probably not feasible (at least in the home shop).

    As for pouring babbitt, I actually do that a couple times a year. These old windmills typically have babbitt bearings, although once in a while you'll find one with oil impregnated wood bearings. Sometimes the bearings are poured directly into the iron casting, sometimes they are poured into a mold and machined (basically a replaceable insert), and sometimes the babbitt is in a replaceable iron shell (that is the situation for my current mill, that the above gear came from).

    Your setup for brazing is a great idea and would definitely make things easier. Something to think about if I start doing more than a couple of these a year (maybe after I retire).
  12. HT1

    HT1 Silver

    I do Piece work at a machine shop that specializes in replacing tungsten, some of the Jigs and fixtures the Owner has produced are absolutely ingenious, , from several grinders, that while hand controlled, have limits to control angles, and everything is supported and designed to reduce vibration... to a welder with Pneumatic controls to rotate the work with your foot while you weld with a fully supported Mig gun, that you give just the slightest bit of direction to using motorcycle handlebars... all while comfortably sitting arms on rests, with an air conditioner blowing cool air under your leathers and up to your face... I'm an absolute believer in making jobs as easy as possible on the Labor... since that is normally me

    V/r HT1
  13. Jason

    Jason Silver Banner Member

    Sneak us a few photos.. I could use some automation ideas. I've seen the spinning foot powered thingy for welding round crap. txsviking on YT has them on his bench.
  14. PatJ

    PatJ Silver Banner Member

    Here is the pattern I received from Ed.
    Apparently it was printed on a Prusa MK3 3D printer, and it is a very nice print.

    The idea is to cast this gear in iron, and come up with a casting that includes the teeth, and one that is accurate enough to use with some light machining or general cleanup.

    To make this part, I will be using bound sand (probably resin-bound sand) since it works very well with iron.

    I bought some mold wash, and will be using that to hopefully help get a clean surface finish.

    There is no draft angle on this pattern, but in the past I have been able to pull patterns from resin bound sand sometimes even if they have no draft angle.
    I have been told that pulling patterns from bound sand without draft angle will damage the mold, but it can sometimes be done without damage, and I have verified that with actual moldmaking.

    The pattern does have a shrinkage allowance, and so lets hope that the correct factor is used (I will dig up the exact shrinkage factor that was used and post that).

    So now I need to get my molding shop set up again, and get the heater and powered respirator working in there again, so I can get some molds made.

    The part will be cast in gray iron, with ferrosilicon added for machinability.

    This is a neat 3D pattern, and it would be great to be able to reproduce this part in a usable form with the teeth cast in place.
    Lets hope we can pull this off with the new iron furnace.
    Keep your fingers crossed.

    I have not set up the flask yet, but I have some ideas.
    I will probably make three molds, and cement them together in a radial fashion, so I can cast three gears at once.
    As long as I am casting, I may as well cast three, and perhaps one or more of them will turn out.

    I have heard arguments about the shrinkage of iron, with ironsides saying that some parts don't show any shrinkage, and some do.
    I will err on the side of caution and probably allow for a healthy riser on the top of the part, and even if it is not needed, it will be an insurance policy for shrinkage issues.

    It would be totally cool to be able to get a good enough finish to cast nearly usable gears with teeth.
    With bound sand, the part will come out with pretty much the exact dimensions as the pattern (minus shrinkage).
    Molding with bound sand does not use the typical green sand pattern rapping technique, which creates a distorted cast part.
    The part will be removed from the bound sand with an automotive slide hammer and a lot of wax, and the mold will basically be an exact mirror of the 3D pattern.

    This is the 3D screen capture of the gear.
    I told Ed to omit the counterbore, and machine that into the part later, assuming we have a successful pour.



    And here is the pattern that was 3D printed from the model.
    The tolerances on the 3D printed part seem very good.
    I will probably sand it slightly more with 320 grit sandpaper, and then wax it generously.
    Wax is a must with bound sand, else you will end up with a pattern that is permanently glued into the sand.



    I will put a core in it, which will be slightly smaller than the finished shaft size.
    I am thinking perhaps a bottom feed, with a large riser on the top.
    I may have to turn a piece of tubing/pipe to get an exact core diameter.
    Last edited: Dec 5, 2018
  15. Melterskelter

    Melterskelter Silver Banner Member

    I think it is extremely likely that this part will need to be fed with a riser. That is based on the fact that it is quite chunky---very thick wall and nearly a solid cylinder. My limited experience with iron is that thin-walled parts fill and then freeze and, I reason but do not actually know, because they are more or less flat planes the to opposed surfaces can just move together to accommodate shrinkage due to cooling and solidification. At the other extreme, by the same unverified reasoning, a sphere fills and then freezes on the surface defining its hot volume as a fluid filled but hard-walled sphere. It gradually cools but can not just collapse in on itself so it can either draw a vacuum or just pull in the weakest section of it surface to allow collapse of that area and to "feed" the sphere. Or, alternatively, if it has a liquid weak spot, a riser, it can and will feed from the riser.

    As a rule for grey cast iron "liquid cast iron has a density of 7.65 g/cm3. Immediately after solidification, the density of the solid cast iron is found to be 7.71 g/cm3." It contracts due to solidification more than just due to normal thermal contraction observed in its always-solid state. The quote is from:

    But the more I read about this, the more complex it gets. In some cases inoculants (FeSi) can significantly increase shrinkage especially if used in excess and in certain cases if the carbon/inoculant ratio is out of whack it can actually cause expansion of the iron. I have not yet observed expansion. My castings seem to follow the 1/8th inch per foot of shrinkage rule to the letter. I do think I have seen excess shrinkage due to excess FeSi and have thus been gradually throttling back on it and also have been very sure to add it only a minute or two before the pour due to loss of effectiveness after 5 to ten minutes from addition.

    Summary of this windy post (perhaps I should have shrunk it ;-) ): Use a riser. Add FeSi just before pouring. don't use too much FeSi.

  16. PatJ

    PatJ Silver Banner Member

    Here is what I was thinking as far as layout.
    The flask would make 1/2 of the mold, and two molds would be made, and then cemented together.
    The sprue and associated basin would each be in 1/2 of the mold.

    After sketching this, it would made sense to move the sprue and basin over to be inline with the main runner.

    The risers would be perhaps 2" diameter.

    I may use a round flask with a flat on one side to save on sand and binder.
    I have also considered a triple-arrangement, but I really don't need three gears, and actually I don't need two, but best to have a spare in case one does not turn out.

    It is interesting to note that the old flywheels that were large were molded flat on the floor of the foundry, and a core box was make for one spoke and its associated hub and rim section only.
    The complete mold was made by using the pie-shaped core box to ram the required number of spoke/hub/rim sections around in a circle on the floor.
    If you look at the old cast flywheels, you can see faint lines where the sections were adjoined.

    The mold will be as thin as possible, and hopefully will be 1" thick or less in most places.
    Resin bound molds can be as thin as 1/2", or perhaps less in places, since it is very rigid.

    Last edited: Dec 6, 2018
  17. PatJ

    PatJ Silver Banner Member

    Since I am using bound sand, and it tends to be very clean with no lose sand particles, then I could probably omit the dead ends at the ends of the runners, and use a radial runner system off either side of the central basin.
    Typically a radial arrangement uses a ceramic filter at the base of the sprue.
    I have some ceramic filters, but am hesitant to use one since I have had flow issues with them before.

    I would use wedge-shaped gates, tapered in two directions, to try and skim off any slag that may remain on top the molten iron flow.

    A bit of a funnel shape at the top of the sprue seems to help with spillage while pouring, so I may carve out a bit of a basin at the top of the sprue.

    The top of the sprue and the top of the risers would be at the same level.
    The bottom of the molds would be perhaps a 1" flat slab of bound sand, cemented onto the top part of the mold.
    I guess it would make sense to put the runners in this bottom slab.

    I would put a core in each gear.

    Keep in mind that I would only need 1/2 of the runner, and one gate, since 1/2 mold will be made at one time.
    I would have to make the sprue as a two-piece design, and use half of it with 1/2 the basin.

    Last edited: Dec 6, 2018
  18. PatJ

    PatJ Silver Banner Member

    Here is an example of twinning two molds together to cast two parts at once.
    I suppose the two molds could have been cast separately, but in this case I did use a ceramic filter, and so I wanted one filter to apply to two parts.
    And I only had one pattern, so I had to make one mold at a time.

    The sprue is a glued on structure made from the same resin-bound sand.
    And the cores were also made from resin-bound sand.

    The alignment of the twin mold halves was not critical, as you can see from the photos, but it still worked fine.
    The mold parts are cemented together using standard furnace cement applied to the outside seam.
    The only purpose of the cement is to prevent runout.
    The weights function to hold the mold halves together, as the cement will not do that well enough by itself.

    I did not use dead-ends in the runners, since I was using a filter.







  19. PatJ

    PatJ Silver Banner Member

    I bought a 240 volt 4KW heater for my molding shed.
    My daughter "borrowed" my nice 240 volt molding shop heater for her back porch.
    What do you think the chances of me getting that back are? (Ans: zero).

    So with a heater in the molding shed, perhaps I can get some molding done for the windmill gear and the bearing cap.

    We have been getting rain and drizzle almost every day.
    I am considering more and more moving the furnace under a canopy, and putting a heat-deflector shield about 6 feet above the furnace lid, at a 45 degree angle, so I can cast in any weather.
    I am not sure if I would overheat a heat shield or not.
    Perhaps if the heat shield had some strategically locate hard fire bricks on it.

    The canopy is painted steel sheet metal, and I don't want to damage the paint.

    This is where a side exhaust would come in handy, but I do use the lid opening to charge the crucible and remove slag, so I could not do that with a side discharge.

    Last edited: Dec 10, 2018
  20. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Have you ever tried to measure the temp 6ft directly above your vent during a melt? Seems like everything is a different ball game with iron and oil burners but almost every person I have spoken to that runs an LP furnace reports the temperature on a ceiling directly above their furnace vent is a total non-issue. -You may not even need a diffuser.


Share This Page