Casting a windmill gear in grey iron

Andy, I think the stripper plate is the key. In green sand you are able to rap in all directions including vertically as opposed to what pat had to deal with for the resin bonded mold. Like you said it is quick to ram another mold and a deep draw is a test of the sands strength. One suggestion for deep spur gears was to make the pattern hollow with end caps so the molder could reach down and rap to full depth rather than using a drive pin and creating "a bevel gear" My petro bond straight from the bag is almost like play dough and will capture detail that my normal heap won't
As an aside looking back at Pattern making 1905 and some of the other Lindsay books your method of splitting the pattern is mentioned, but discouraged. That was probably to reduce down on pattern complexity and also the chance for mis-alignment in a production setting.

 

Thanks, ESC. I was wishing I could get inside to rap the gear today but I had made a plug and couldn't get it out to rap.



Gear has been sanded and painted.

In the flask.



Great expectations...



Almost there



I could open face cast that and Ed could just clean it up.

Doesn't that hub look nice...

 

Looks pretty good.
.

 

Could you do some rapping in resin sand before it sets and then wait for it to setup before pulling?

 

My experience with resin sand is that if you rap/move/disturb it prior to set time, then you will fracture the sand, and you will most likely have multiple fracture/failure points.

.

 

So immediately after it's placed you can't rap it? Is it real sticky? I've never been around it.

 

I'm learning about gears now. Made the drag half with draft and started printing the cope pattern.



I learned today that a nice sharp file cuts PLA better than sandpaper.



The stripping plate was 102% of pattern size and wouldn't clear after I'd painted and sanded the pattern. I printed a 106% plate.

Painted and sanded with a core pattern.



Still pulled the sand out. The perimeter of each tooth is much longer than the arc across the tooth tips and it wants to fracture instead of pull.



Swinging an iron bar inside the center hole does vibrate it enough to do some good, and with the sand lightly rammed I got a fair pull.



But the tips of the mold (roots of the teeth) failed slightly.

Plan now is to sand the teeth a bit more and wax them. Is auto wax good or bad for helping release green sand?
Them gently ram each tooth in, vibrate, and use the stripping plate. I think I'll have to get cheeky to bottom pour this.

Results are guaranteed, but success is not. Here's hoping for good results.

 

After resin sand is mixed, it is pretty loose and has a wetted feel to it.
It is bad about sticking to the pattern unless you have a lot of wax or other release agent, and some draft angle.
I have pulled patterns that were perhaps 1/2" tall with no draft, but that is as tall as I have successfully pulled with no draft.

Once you dump the sand in the mold and ram it, you cannot touch it until it sets, and then there is a strip time after which you can pull the pattern.
Once the sand sets, it is hard, but the mold needs to remain flat because the sand is not at full strength, and it will warp if you pull it out of the snap flask too soon.

With a 3-part resin binder system, you have a fixed amount of resin and hardener, based on the weight of the sand, and you can vary the set time by varying the amount of catalyst.
I really like the variable set time since you can set it for perhaps 11 minutes, and that gives you just enough time to get things rammed and leveled off.

Some of the 2-part binders have a fixed set time, and the set time can only be varied by changing the catalyst to a faster one, which is rather inconvenient since the suppliers don't seem to want to sell small quantities of catalyst.

I like to be able to make a mold about every 20 minutes at the most, and a little faster sometimes if the mold can be rammed quickly.

You never rap a pattern with bound sand, and that is what gives the castings such good dimensional accuracy.
When machining steam engine parts, the parts center up in the lathe very accurately, and you only need to take a light skim to clean up the surfaces.
The parts come out of bound sand very concentric, and you don't get that wobble and out-of-roundness distortion that you can introduce by rapping the pattern in the sand.

You can see resin sand molds being made in this video starting at 2:10

 

If you did that pattern in sodium silicate, I am nearly certain it would pull. If you don’t have it, I’d try polyester resin with catalyst or epoxy resin and hardener either one around 2% by weight resin mix to clean dry sand. I have not used either poly or epoxy but have read of good results.

I think the the resins in epox or poly would provide some gas generation to improve finish. So coal might not be needed.

Denis

 

I finally have a pattern printing with 2* draft!! It's on my printer (that is still a work in process) ... so we'll see how it comes out.

 

I tried sodium silicate sand. Made my first one too small



Made a 6" ring. I'm cheap. I want the silicate to off gas into green sand.



Cured it, slipped the form up and put the second pattern half on.



A little sand, a little CO2



They came apart OK.



It took a hundred or so light taps to get the patterns out. A lot of gentle taps with no movement.



Assembled



Six pounds if I cast it solid. A feeder reservoir under it and a riser above bigger than the teeth is the plan to avoid shrinkage in the finished part. Bottom fed.

Comments?

 

That looks really good.

.

 

Whoooohoooo! Looks great! Pour some hot iron in that sucker!

Denis

 

Did you weaken the mold with Sugar? I can’t see any coal. If there is no sugar the sand gets mighty hard and sticks amazingly tightly.

I often cook the silicate mold at 230 deg F for 2 hrs mto strengthen the mold prior to pouring. Inserting the mold into green sand for riser and gates etc should work well.

If this is a silicate only (no coal and no sugar), this is a great chance to now do another mold with coal and sand. Pour them both so you can see the difference. Terrific learning opportunity.

Denis

 

Thanks, Pat.

Too late tonight. But if the weather cooperates tomorrow I'm going to destroy it.

I've been using sugar in my pipes and it works great. I decided since I can get to this I can chip it out if need be. Probably should have put sugar in it. I didn't know you could put coal in sodium silicate. Thanks.

I have cooked silicate sand and made it harder and have cooked it and made it more crumbly. I don't know why the difference. I really wanted this hard so I could ram sand around it. I only have N grade sodium silicate and can't seem to find the good stuff to be shipped. If anyone has a source please let me know. I cooked a piece of my first failed silicate sand tonight and made it crumbly. I've been using 4% but went to 6% sodium silicate for this last batch.

I'll dig some fine coal out and try that. 5% by weight?

 

After supper I went back out to build a cheek for the flask I want to use and realized I need no cheek. Just leave the silicate mold together and pour something in it. Gear teeth down all in the cope with a5/8" thick oversize disc for a full contact gate. Drag:



Rammed the drag and flipped it, set the silicate mold on the big gate then put a riser on top (riser diameter is bigger than the gear hub) but all I had already had a big hole in it so I put a round piece on it. The riser comes out of the cope.



Rammed delicately around the silicate mold. The disc gate was made in a hurry from wafer board. With 4 degrees draft it still pulled a ton of sand with it. Oh well, I wanted a big gate.



Only have a little sand over the silicate mold. Here's the view of the cope from the bottom. I should be able to watch the iron fill the mold through the giant short riser.



The teeth are a little misaligned. 3D printed, the model must not be as accurate as I thought. Lineup holes were 3D printed too, and I know they are snapped to grid.

Ready to close.



I dug a ramp from the gate in the cope to the disc gate under the silicate mold. I think 10 pounds of metal will be plenty.

Plan is for the disc gate to act as a feeder from the bottom, and open riser as a feeder from the top.

 

I see a lot of gas holes in the mold. It is possible you are over gassing to end up with crumbly sand. I think you could use six holes around the gear and gently blow CO2 for a five or 10 second application each hole. I often use a 3x3” thin flat pc of thin plywood with a through hole place on the sand and insert the gas tube through it. I think the wood helps spread out the gas from the hole. If the pattern will tolerate it, putting the pattern still in the sand in the oven around 200 for a couple hours may strengthen the sand prior to pulling but after gassing.

A “thousand” little taps is a good plan for pulling. Just a little traction and tap, tap, tap....

Yes, 5% coal.

I wish you could get some RU. With all the refractory supply resources in your area, I am surprised you can’t get it. With RU you know what you are dealing with.

Denis

 

What is the purpose of the riser isolated off to the side formed by the yellow 1” dowel?

Denis

 

I don't have CO2. I put the mold in a box with a cup of baking soda and pour vinegar into the cup to flood the box. I get 1/2" to 3/4" penetration in sand depending on how hard it is rammed. This is a lot more holes than I usually use.

Like an idiot I haven't checked with a local refractory company. Now it's the weekend and I'll forget Monday.

The yellow tapered dowel is a surge trap.

And it has a foam plug in a pouring basin, I just didn't get a closed up picture.

 

For a complex shape like the gear with lots of nooks and crannies and some depth, propylene carbonate can be a nice alternative to gas. See Tom Cobett’s article on silicate usage. Slower than CO2 but complete kick with no risk of overgassing.

http://jarod.eells.us/projects/FoundryEquipment/sodium-silicate.html

Denis