Spiral bevel gears n gearing

Home Model Engine Machinist Forum

Help Support Home Model Engine Machinist Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
You are correct in that most metals shrink when they cool.
There is always that oddball metal that expands when cooled, such as that used for casting type, using in printing presses, which expands (I found that out the hard way).

Shrinkage varies between metals, but is somewhat close between aluminum 356 and gray iron.
I have been using a multiplier of about 1.015 for gray iron, which is mostly what I use these days.

Pattern making is a bit of an art in that you are adding material to surfaces that have to be machined, to build them up slightly, adding draft angle so the pattern can be removed from the sand mold (some patterns do not have draft angle), and shrinkage.

Shrinkage is not necessarily uniform in all directions, but I have found that it is uniform enough for small engine parts that genally the overall shrinkage can be considered to be the same in all directions. For a long narrow part ,this may not be the case.

Minor variations in shrinkage don't affect most cast parts, but for gears, close consideration would have to be given to some things.

For my Frisco Standard, the helical gears will both be cast, and they are the same diameter gears, so the shrinkage will be the same for both gears.
There will not be any draft angle or machining allowance on the Frisco Gears; they will be cast as if they were made like the investment process, but using resin-bound sand.

I will start with the 1.015 multiplier, and see if that produces a usable set of gears, with "usable" being a set of gear castings that have centers that are very close to the "as-machined" size.

I am not aware of any adjustment allowed in the position of the vertical Frisco Standard side shaft either at the head or where the gears are near the crankshaft.
Probably what I will do is install one gear on the cranshaft, mesh the other gear next to it on a flat surface, and measure the crankshaft center to gear center distance.
Then I would drill the hole in the base for the veritical side shaft.

The gears would be gated from the bottom in the mold.
The sprue is vertical passage from the top of the sand mold down to the horizontal runner(s).

Here is a windmill gear that I attempted and failed to cast.
The pattern was not mine, and so I could not heat the pattern to remove it from the bound sand, and did not think to do so either.
I came close to pulling the pattern from the bound sand with no draft, and with a little heat (which would ruin the 3D printed pattern), I think I would have succeeded in making an gear that was accurate enough to work well.

Edit:
Shown is my preliminary sprue/runner/gating diagram.
I no longer use a basin at the bottom of the sprue, I use smooth transistions in the runners while maintaining their constant section, and I use spin traps at the ends of runners. I use straight gates now, not tapered ones.


Capture1.JPG



GEAR-FLASK-LAYOUT-01.jpg
rImg_2494.jpg



rImg_2495.jpg



rrImg_3933.jpg



r20190420_224740.jpg
 
I think this mold ended up being a three-part (3 levels too it, each adhered together with foundry cement).
I came very close to casting an accurate gear, but the person wanted two gears, and so I would have needed two sacrificial patterns.
This was a good learning exercise, and from this exercise, I think I learned eough to cast some helical gears for the Frisco Standard marine engine.

The molds for this gear were a bit rough, but I think I can improve greatly on this process in the future.
The intent is to make gears that will function as-cast, with just some run-in with lapping compound, and perhaps light surface cleanup/buffing.

People will insist that you MUST have draft angle on a pattern, but with bound sand and the right release agent, this is not necessarily true.
Most people use green sand, which is mostly sand and clay, and for those molds you do need draft angle.

rImg_3973.jpg


rImg_3974.jpg


rImg_3976.jpg


rImg_3977.jpg
 
I wonder if this method could be adapted - using a lathe as a shaper. In contrast to the video below, the blank would have to be mounted at an angle (with an offset tailstock, or perhaps by mounting a tailstock centre in a boring head fitted to the tailstock, to enable it to be set across to the required taper - an idea I read recently in an old book which I rather like). But maybe the tapers required are rather extreme, and this would not work.


Beautiful work, well done.
 
It is indeed beautiful. In case of any doubt, I feel it important to direct your 'well done' at Mr Hansen, not me!
 

Latest posts

Back
Top