Gears

[timo_gross]

Not having tried any "real" casting so far. So it is all a more or less uneducated question from the "keyboad foundry".

If the gears can be printed in PLA with the lowest possible infill, can the metal be cast directly, leaving the PLA pattern inside the sand?
Like a lost foam casting?
Would Cutting the gear out of foam with a hot wire, be an option?
That would take time because a new pattern for each part is necessary, but it would save you from getting the pattern out of the mold?
When ramming refractory the supervisor showed me to loosely wrap a plastic foil around the pattern, then it slides out more easy.
Based on this I thought how about print the theeth in "vasemode" then put another inside pattern for the spokes in?
a) leave the thin wall teeth inside the sand b) try to remove it carefully to the inside, as it will be flexible.

Greetings Timo

 

[GreenTwin]

The "lost foam" process is used in industry, often to create engine blocks in both aluminum and iron.
It uses a special foam and process, that ensures that the foam basically evaporates without causing any casting defects.
Generally these foam patterns are dipped multiple times in a ceramic slurry.

Several people online are copying the "lost foam" process using a hardware store foam that is different than what is used in the industrial "lost foam" process, and so often serious defects can often be seen in their castings, with unpredictable results, since some hardware store variety foams do not vaporize completely. Sometimes the foam patterns are just surrounded by loose sand.

There is a "lost-pla" process, and it is used similar to the above process, ie: the PLA pattern is placed in loose sand, or perhaps a few coats of something, and the molten metal is poured in on top of it.
Again, I have seen mixed results with this because the PLA does not necessarily vaporize completely.

Then there is the "lost-wax / lost pla / lost foam" process where a pattern is dipped multiple times in a ceramic slurry, with sand dusted on top between layers.
If the wax/pla/foam is burned out completely in a kiln, then the results can be anywhere from good to excellent.
This process lends itself to multi-part "trees", where many parts can be cast at the same time.

The hybrid system I have come up with is to use 3D printed pattern with very little internal support, ram it in a resin-bound mold, and then heat the mold to make the pattern soft enough to be carefully removed without damaging the mold.
I think this method will work well with bound sand, and with patterns that do not have any draft angle, such as on a gear.

Then there is the green sand gear casting method being used by sawyermassey, and this is a bit of a challenge, due to trying to pull the pattern from the sand without breaking the mold.
I think trying to use greensand with a pattern that has little or no draft angle will be challenging, but not necessarily impossible.

Lots of methods.
Take your pick, and lets get casting.

.

 

[GreenTwin]

Close .......not there yet

I have read that the horn sprue from the old days does little to control the velocity of the metal, and thus lots of splashing and air aspiration during the pour.
I have not used a horn sprue, so just guessing, but that is what I have heard.
I have also heard that the flow velocity should be controlled by the gates; at least that is one successful method.

Just tossing out ideas for consideration.
.

 

[Steamchick]

I favour lost wax - but have never tried it. What I have done, is take a part (e.g. a flywheel) and dip in molten wax and quickly remove it. The wax build was very thin. Then after making the mold, this pattern was heated so the wax melted, when the part came out easily. (NO draft). I heard this from someone who had made a mold but wanted a core from a gear, which then had to be removed from the casting so he had an internal gear. as cast. There was a more complicated process somewhere using Plaster of Paris, but I don't fully remember what he did... Sorry I can't explain this.
But the idea of coating a pattern in Wax to give a couple of thou clearance when the wax is melted, seems a good idea to me?
Please ignore if this is wrong?
K2

 

[ddmckee54]

What I have done, is take a part (e.g. a flywheel) and dip in molten wax and quickly remove it. The wax build was very thin. Then after making the mold, this pattern was heated so the wax melted, when the part came out easily. (NO draft).

K2:

What was used to make the mold, bound sand, Plaster of Paris, casting investment? I could see where this method would work if your original part would allow you to make a 2 part mold. But with a complex shape I'm just not seeing it without many parts to the mold.

Don

 

[Steamchick]

Plaster of paris mold. First half cured, then cling film covered the joint and second half cast.
K2

 

[GreenTwin]

I favour lost wax - but have never tried it. What I have done, is take a part (e.g. a flywheel) and dip in molten wax and quickly remove it. The wax build was very thin. Then after making the mold, this pattern was heated so the wax melted, when the part came out easily. (NO draft). I heard this from someone who had made a mold but wanted a core from a gear, which then had to be removed from the casting so he had an internal gear. as cast. There was a more complicated process somewhere using Plaster of Paris, but I don't fully remember what he did... Sorry I can't explain this.
But the idea of coating a pattern in Wax to give a couple of thou clearance when the wax is melted, seems a good idea to me?
Please ignore if this is wrong?
K2

There is shrinkage to be considered, and so if an original gear was used as a pattern, its cast size would be a bit smaller (perhaps too small).
I have heard of some dipping a part in wax to build up the exterior to compensate for the shrinkage.

If a new pattern was dipped in wax, and the wax allowed to become hard, then that could potentially allow the pattern to slide out of the sand if you could heat the pattern a bit.
Some green sand does not like to be heated very much, such as oil-based greensand (Petrobond tm).

For bound sand, the wax layer may work well, since you can heat a resin-bound mold (or a sodium-silicate bound mold) quite a bit without damaging it. The interior of the mold would need to be lightly flamed to drive off any residual wax, else it will bubble during the pour.

For small gears, the wax would probably eliminate too much detail.
For larger gears, I suspect the wax coating would work well with bound sand.
I am not sure if it would work with greensand as well. Would depend on the exact technique.

.

 

[sawyer massey]

Messy but all there ...37 lb crazy
Will move the feeders next time .....only 9 more to go lol
Still hot in the picture......not that sandy now

 

[GreenTwin]

Nice work !!!

.

 

[sawyer massey]

Way cleaner less flash .........and all IST melt iron ....by me anyway lol

 

[awake]

Looking good. Is that 1ST = first melt iron? (I tried looking up IST and couldn't figure out what it was ...)

 

[GreenTwin]

Looking quite good.
I think you are gettting there, or have actually gotten there.

.

 

[sawyer massey]

Looking good. Is that 1ST = first melt iron? (I tried looking up IST and couldn't figure out what it was ...)

Yes it is first melt iron...lol

 

[sawyer massey]

Differential pattern .....almost finished

 

[sawyer massey]

Well hope it machines well
Differential for the Waterloo
38 lb