lost pla/pattern castings ?

[picclock]

I'm thinking of building a minnie traction engine on a budget, like NormanV's thread. The price of castings is pretty high so I thought I might investigate 3d printing the parts then have a go at casting them. I have quite a lot of scrap aluminium, brass and copper which could be used. Possibly bronze using very old non magnetic 1 and 2p pieces (pre 1993 ?)

Most of the parts just really need an impression in the plaster(?) rather than the full double sided casting process. I figured I would subject the plaster to vacuum to get rid of the bubbles then remove the pla using some form of mould release.

For the rear wheels I thought I would try an alloy of aluminium with 5% copper as Wikipedia indicates its not a bad idea. 

Any thoughts or advice on this much appreciated as I have yet to cast anything in metal.

Best Regards

picclock

[Joules]

Wet sand the PLA parts, use sanding sealer on them and wet sand again.  For release agent you can use Vaseline.

[vtsteam]

I suggest trying a small simple part first with your chosen method, cast in straight aluminum, without 5% copper.

You can melt that in even an open wood fire pit, after it produces glowing coals at the bottom. Put your crucible in there with the coals, and no need for draught.

It will melt, it will only take a few minutes longer than a furnace lined with $200 worth of refractory, plus gas burner, plus draught blower, and all the techno-wonder gear, etc.

This experiment will tell you everything about what needs improving or revision in your chosen molding method, with little expense or waste. Gradually work those details out before investing in anything else, like a furnace.

Furnaces do nothing but heat metal. Molding process is everything -- and that has to be learned through experience.

[vtsteam]

The title says "lost PLA" which implies melting out the pattern material, either in the heat of the pour metal, or by mold baking, rather than releasing a permanent pattern (eg. with Vaseline).

I would think that baking would be the preferred "loss" method for a 3D printed pattern, as there is a larger volume of flammable material to get rid of otherwise.

Generally foam patterns are used for the metal replacement method, because they have a much smaller amount of flammable material. Though even then, cooler metals like zinc alloys can have problems completely eliminating pattern material before solidification.

In any case, as I said, try something very simple to start with, and see what happens -- this will guide you as you perfect your preferred method.

[picclock]

@Joules
Why? My 3d prints come out really smooth. Vaseline for release agent I can understand. I thought that I would make some impressions in plaster of paris then when its set off remove the pattern and fill the hole with Aluminium or whatever.

@vtsteam
I already have refractory bricks, the kiln type, which I have used to silver solder steel parts. Once tried using an aluminium support and it melted like a candle  . Since then have always used steel parts for clamps/support. No problem melting ally or I suspect copper/Brass. Real problem is not overheating things, because although I have an infrared gun type thermometer it only goes to 650C.

Will try with a test piece and see what transpires. Maybe mold the chimney support in Aluminium then copper just to see what gives.

Would really like to cast the rear wheels as if it can be done done with an aluminium/copper alloy I should be able to incorporate the treads and not suffer from excessive wear or oxidation.

Thanks for the helpful comments.

Best Regards

picclock 

[Joules]

Prints can still be porous and you have the fine layer lines that can also trap the plaster, don’t assume prints to be water tight is my experience.

LOL, I read the title lost PLA casting, and then the description of using the PLA parts for patterns, so some contradictions here.

[picclock]

Hi Joules

Thats coz I don't have a clue what I'm doing . Take your point about porosity but doesn't seem to be present on my prints. Layers seem well fused. Main problem I have had with 3d printing is warping of the plastic. Not sure how to sort that :-(

I think most of the bits would be pattern cast (see, i'm learning), but one or two would be best cast properly, likely with risers and green sand, or plaster of paris mold in a sand filled bucket for stability. For those baking the mold upside down to purge the plastic from it seems a good idea. But I really have no clue .. .

Just trying to get it right in my head before I decide to go ahead.

Could be a good project for the winter.

Best Regards

picclock

[Joules]

I am just finishing up some patterns for sand casting.  Generally a solid parts warps more than a shell print.  Not shell as in single layer, but in this case 3mm thick walls.  The two parts are printed flat then welded together, dowels locate the two halves.  The edges get filled and smoothed off.  My example below has primer for the first rub down to see where the voids are as the print is quite course at 0.15mm layer.

[vintageandclassicrepairs]

Hi All,
picclock,
I have a friend who makes sculpture's from bronze castings using the "lost wax" method
I know very little about casting but one thing I have learned from talking with him is that
plaster of paris has to be baked into basically porcelain before molten metal is poured into
the mould
If molten metal is poured into air dried POP it will boil any remaining water instantly and the POP
will shatter 

John

[vtsteam]

Okay, picclock good title change.

You'll probably need to incorporate draft in any conventional pattern to release it from plaster, and a mold release as Joules pointed out, and even then, some pieces can be a real pain to release.

Surface finish is pretty critical with plaster -- more so than with greensand (and even that benefits from the highest finish possible).

Copper containing metals need a LOT higher temperature than aluminum and behave differently as well. Hence the suggestion to start simply w/ Al .

If I remember correctly traction engines have diagonal treads on the rear wheels, and I can't imagine pulling a pattern intact out of plaster with that configuration, if treads are included. Lost wax (general term here) methods would work.

Be aware that if you try to pour to top off a simple mold without sprue and gates lots of difficulties can arise. Shrink cavities for one. Or the opposite -- an overpour and surface run-off with lots of excess to cut off and machine back. It doesn't save time or effort to machine a big bunch of lumps compared to cutting off a gate or two where the rest of the casting is close to finished in proportions because it was fully enclosed in a mold..

Or both at the same time: cavities and lumps -- the most likely scenario. I find the best patterns to direct pour are long sticks of something I intend to cut into lengths for some job. Sort of like making your own extrusions. I think it makes sense there. Or pouring round or square section shaft stock. Anything a lot deeper than wide with a spoilage allowance at the top makes sense -- and the nice thing is, you often don't need draft because it's easy to rap a shape like that.

Anyway....my experience in that type of casting.

[vtsteam]

ps. picclock, if you want, I can move this thread to the "Metal Stuff" sub-forum -- it will more likely be seen there by people like Ironman, who has visited occasionally in the past. And others into casting here. Up to you...

[vtsteam]

Just a little bit more of my experience -- the only lost wax type casting I've done is lost foam, with metal pour burn-out. I got some nice results with patterns I would not have been able to mold easily otherwise. But there were some drawbacks that I didn't like, and for the most part I have returned to making conventional wooden patterns, except for very special circumstances.

My main issue with lost foam is that it takes a lot of time to make a pattern and its mould casing. For me, that represents a couple days. And if the pour doesn't go well, you have to make a new pattern and mould again.

With a conventional pattern, (and I can make those in a matter of hours, usually), If I get a bad pour, I can pour again using the same pattern within an hour or less. Or make multiple parts with one pattern.

A wooden pattern also becomes part of a library of patterns I can use to make other things. I've re-used many of my patterns for different projects. Or replaced worn parts with a copy. So I like all of that.

[Joules]

For casting the wheels you could try two part moulds, If it was me I would exaggerate the band for the treads and machine them afterwards.  In fact you probably want to think quite bit about including extra material for machining and shrinkage.  The oxide painted pattern above has domed ends on the bosses, these are 5 degree taper at the edges, same with the lugs for mounting.  Yeah, don't forget draft and this will exacerbate the layer problem, hence my comment about wet sanding, filling and sanding again to get a smooth finish.

I still have a lot of sealing and sanding to do, this won't be ready for Friday 

[vtsteam]

You'll never cast diagonal treads on traction engine wheels with spokes or holed rims in a two part mold in plaster (or greensand for that matter) with a conventional pattern. Lost wax only.

[Joules]

My bad, I should have said rims, not wheels and it would be three part mould to get the gap in the rims.  Steel treads on ally rims makes far more sence.

[vtsteam]

Well unless I miss my guess here, Joules, I think the whole idea picclock wants to try at least is to take advantage of the 3D printer's promise to make complete complex shapes with details and undercuts and then cast that in plaster, for high level rendition, but if possible reduce the amount of additional material (greensand, sprue, flasks etc) and so be able to pour into an open mold. And printed traction engine wheels with treads is what he's interested in trying, for example.

Now this sounds like a tall order, and I don't own a 3D printer and not likely ever to, and am not even a lost wax/material casting fan, BUT..........

I do think what he wants is feasible. With maybe some slight modifications.
One is that it must be a baked out lost pattern process.
Second is that it be a two part plaster mold.
Third is that it not quite be an open top mold but a shallow sprue, carved like a short funnel into the plaster. And maybe some vents.

In short, think of it similar to the way a bronze statue is cast into investment. (Mattinker, where are you?)

I think that might work to do what he wants.

Is it something I'd want to do?

Nope. I like wooden patterns, greensand, and machining.

But I'm certainly willing to cheer someone else on if that's their ambition.

[S. Heslop]

Myfordboy has a video on this that youtube won't stop recommending me.

I'm not 100% on the guy - from what I know he's done alot to spread a bunch of myths in the casting world. But the part in the thumbnail is an interesting shape. Shame he doesn't show what its for.

He makes it look like a fairly simple process with good results though.


I think the interesting idea with investment casting is for doing large batches of parts, assembling a tree and casting a whole bunch in one go. Maybe there's scope for making silicone moulds from the 3d printed (and ideally sealed and sanded smooth) objects for investment wax as a middle step.

[S. Heslop]

Looking at that part again, it's clearly part of a 3d printer extruder.

[picclock]

@vtsteam
" picclock wants to try at least is to take advantage of the 3D printer's promise to make complete complex shapes with details and undercuts and then cast that in plaster, for high level rendition,".

Spot on. I was hoping to produce the rear wheels like this using a 5% copper aluminium alloy to make the part fairly hard wearing and corrosion resistant. Definitely using lost pla if possible, as I can print this easily, possibly 2 at a time with the front wheels inside. Not sure how it will work out but worth a try.

The thing that impressed me about the myfordboy video is the extreme detail resolution which came out on the casting. You can see the 3d print lines and patterning. Two things i found a bit puzzling, when he burnt out the mold there was no smoke or mess - made me think he had done it then just placed it back in the kiln for the video. Other thing was that there was no vent pipe for the metal. Always thought it was essential yet the casting looks pretty near as good as it could be.

The other parts I thought I would have a go at pattern casting as it looks easier but I'll see how I get on.

Thanks for all the info

Best Regards

picclock

[RotarySMP]

Have you checked out Alloyavenue.com? There are a number of people doing great castings with lost PLA.

I have done a couple of lost PLA castings in aluminium. so I am only a beginner at it.
I didn't have access to real investment casting shell material like Myfordboy used. I built up a shell made from 2/3 fine sand with 1/3 plaster of paris. It is difficult to get the first layer to stick. I read afterwards that using the glycol which is sold as dish washer to stop water pearling aids first layer adhesion.

The shell was built up with a brush to about 1/4" think over a couple days and then dried thoroughly.

I burnt out the PLA and fired the shell over a about two hours in my furnace with the propane burner running at low pressure. Using a 10% infill printed part, there really is not that much PLA to burn out, and it burnt quite cleanly. I was worried about ash in the corners of the mould, so tried to blow out one of the moulds with compressed air, and destroyed the shell. This is the real disadvatage of POP. By the time the PLA is burned out, you have probably exceeded the temp which dehydrates it, reducing the strength near zero. Seem like ash staying in the mould is not really an issue with PLA.

Myfordboy did an interesting twist with that casting. He left the large flat side open, with no shell, and used a normal green sand on that side, so the mould would have largely vented through the sand.

My test piece was a spoon, and had a minimum thickness of only about 1.5mm. Quite a tough test for a first attempt casting, but it worked suprising well. I can't find the photo I took sorry.

The entire work is pretty time consuming, and the risk of failure is (with POP) pretty high, but if it works, the resulting castings can have extremely high tolerances, which has been typical of investment casting methods for the last couple of thousand years.
Mark

[PK]

So I'm a bit confused, what I've seen in this thread doesn't match my own experiences with lost <insert lost thing here> casting.
My attempts fell over at the burn out stage.  Heating the whole thing caused a thin shell to crack as the wax expanded, if that didn't happen then molten wax was forced into the refractory coating, if I used a big block of plaster then I could never get enough water out.
In the video's above the guy is using some magic powder that doesn't seem to need endless hours of baking to dry out and everything just works... 
Is this real?
Where can I get this magic powder?
PK

[Joules]

BOLIVIA

[vtsteam]

There are specialist plaster investments (good choice of term) and specialist waxes, etc. Then there are also specialized techniques developed long ago in bronze statuary founding. Mattinker where are you?

The wheel keeps getting reinvented. People forget, and then struggle to understand what was once known.

[awemawson]

When I did a bit of lost wax casting years ago I got a sack of the high temperature version of Plaster of Paris but it was a pain to use - not sticking etc

In the end I reverted to using bone dry silver sand and Sodium Silicate cured with CO2. Wax was burnt out in a pottery kiln and I poured the metal while the mould was still glowing from the kiln.

I satisfactorily poured aluminium, brass, copper and iron - all that was about 12 years ago, but the door stop we still use today is a cast iron recumbent lion, modeled by my daughter in art classes at school in clay. I tool a Vynamould impression of it, poured wax into the cavity, cast the wax in sand and sodium silicate, burnt it out in the kiln, melted the iron in my induction furnace and that's the result. Poor thing had a hole drilled in him for a rod and knob to save me bending down !

[picclock]

Hi

From just checking out the Alloyavenue.com site and one or two others it all seems quite doable and will be a very useful addition to my workshop.

I looked at an Australian video of a can forge and was amazed at what he achieved melting 1.7Kgs of cast iron in 20mins :-

So I thought I might up my aspirations and try to build a forge which could do this. The only part I cannot seem to source is the high temperature reflective coating which is applied directly to the ceramic wool. He uses zirconw. Searching the internet shows 3 possible alternates, Plistix, Metrikote and ITC-100. However these do not seem to be available in the UK, or indeed europe. Dont know if its one of these eu rule things.. .

If you have a furnace, what do you use as the high temperature liner ? 

At my current state of ignorance help needed.

Best regards

picclock