The Bucket Foundry

[WilDun]

So you guys will both be using the 'lost PLA' method of casting? - sand/plaster I suppose?
Now I'm really starting to feel old, but at least I have heard something about that process!

[Peiter]

Hi Wil,

If I can judge it correctly, Polinzei has gone with printing tools (I don't know the right word) to form his sand cores in. Together with the foundry, I decided to go the lost PLA way, where the printed part will be the actual cylinder later on. Wax risers were attached to the PLA cylinder and the mold was then dipped in a liquid, coated with fine ceramic dust, dried, dipped again, coated again, etc. until the layer was thick enough. Next, the PLA and wax was melted away during a final step, after which the cylinder could be cast.

Excuse me if the lay out is messy, I am typing this message on my phone 😊.

[ken seeber]

Hi Wil,

Next, the PLA and wax was melted away during a final step, after which the cylinder could be cast.

Excuse me if the lay out is messy, I am typing this message on my phone ��.

Peiter, thanks for your contributions.
Did you have any issues with the PLA cracking the investment during melt and what temps did you used for melting and burnout?
Also, another question, what shrinkage allowance did you make and/or what was the overall shrinkage, eg from a dimension on the PLA to the same dimension in the cast aluminium.
Your phone message is just fine. Keep it up.

[WilDun]

Hi Wil,
If I can judge it correctly, Polinzei has gone with printing tools (I don't know the right word) to form his sand cores in.

Coreboxes?
I suppose that as the printer/ 'lostPLA' system becomes more popular then coreboxes will be out, but nothing ever goes away completely and I'm sure that some of us will just continue to use that method in conjunction with Co2/ Sodium Silicate moulds and cores - it's maybe a little more 'hands on' and takes more time to do, but cheaper and still fine for a home workshop such as mine where time is not money! - I must say though that it's still very interesting to me.

Ken, I have heard of ceramic shells cracking with expanding wax patterns (expansion before melting), but maybe the PLA doesn't expand as much before melting? / burning out?.
Guess it would also depend on the section thickness - a large piece of wax would probably take a little longer to melt, which means it will have more time to expand before that happens.
I wouldn't know how much expansion to expect from PLA though.

[Flettner]

Coreboxes?
I suppose that as the printer/ 'lostPLA' system becomes more popular then coreboxes will be out, but nothing ever goes away completely and I'm sure that some of us will just continue to use that method in conjunction with Co2/ Sodium Silicate moulds and cores - it's maybe a little more 'hands on' and takes more time to do, but cheaper and still fine for a home workshop such as mine where time is not money! - I must say though that it's still very interesting to me.

Ken, I have heard of ceramic shells cracking with expanding wax patterns (expansion before melting), but maybe the PLA doesn't expand as much before melting? / burning out?.
Guess it would also depend on the section thickness - a large piece of wax would probably take a little longer to melt, which means it will have more time to expand before that happens.
I wouldn't know how much expansion to expect from PLA though.

I realy don't understand why these printers aren't used to build the core boxes, now that make a lot more sense.

[husaberg]

too short...............

[Arifidyan]

Polinzei, cool looking stuff!

Below is my first part on the PLA made cylinder and casting it;
At first I wanted to buy a cylinder kit made by Stage6 or a Malossi speed 7T which all perform reasonably well.
However, in december 2014 I stumbled upon a guy on google who was casting an aluminium piece directly from his PLA printed 3D model (http://3dtopo.com/lostPLA/).
This made me interested in the subject and together with a metal foundry nearby where I live, we tried a sample ourselfs. It turned out that PLA will melt away 100%
without residues, so the method was indeed quite good to cast cylinders!

Together with a good friend we then formed the idea to start designing a cylinder ourself where we could work around the following:
- The original stud size is 56x56 with 7mm studs, this pattern is too small for a 47.6 bore cylinder to shape it correctly at the exhaust and b-port.
- Most aftermarket cylinders have the exhaust channel at a straight angle whereas we have tried to obtain the desired 20-25 degrees (if I recall it correctly).
- Cooling of most cylinders for a Derbi engine is asymmetric on one side of the exhaust channel.

First, I started making silicone molds of all sorts of cylinders I could get my hands on, a friend's 22 hp 50cc, an Aprilia RSA, a Honda 2007 a-kit, a KTM 125, etc..
(in the picture below are only a few of the silicones, I have enough to fill my desk )
Attachment 316263
With these molds, I went to a Fablab where they had a reasonable 3D scanner and I scanned the molds with an accuracy of (If I recall it correctly..) 0.1 mm.
Next, I sat together with my friend and given our appetite and enthusiasm to discover things ourselfs (in some people's eyes stubbornness), we decided to go with the Honda port lay-out as shown in the images.
Attachment 316260Attachment 316261Attachment 316262
The Honda cylinder had the A-port its edge nearest to the B-port almost perpendicular to the bore and aimed at the other side. We decided to go with a more common direction of aiming the A-port further towards the C.
The axial angles are set to about 22 deg. for the A-port, 8 deg. for the B port and 50 deg. for the C-port.

After designing the cylinder in CAD, it was time to print it in PLA. As we had to remove / clean the white shell after casting, we decided to remove the "roof" above the exhaust port such that we could thoroughly clean the cylinder.
Attachment 316265
Inherent to the FDM process used by our Ultimaker 2, overhangs and roofs larger than 50 or 60 degrees require support material to deposite the material during printing. Therefore, we had to design the cylinder in such a way that it was printable and all the support material was removable.

I'll elaborate on making the design printable and the actual casting next time, hopefully you guys like it!

cheers,

Peter

This awesome. Did you make crankcase too?

[WilDun]

I realy don't understand why these printers aren't used to build the core boxes, now that make a lot more sense.

Yes, I would agree that it is a good idea too, but whatever you make with a 3D printer will not make a perfect pattern (at least going by some results I've seen from printers) - I guess that will all be resolved eventually, but possibly by some very expensive printers! In the meantime some prints will really need to be tidied up and plastic isn't always the easiest stuff to tidy up (from my experiences)!

But don't get me wrong, I am also a big fan of this method, except that I'll run out of time before it can reach a workshop like mine!

[polinizei]

My Zortrax M200 prints very good results. (Better than Ultimaker 2)
It prints up to 0,09mm layers.

Its possible to finshing the printed parts with acetone, to geht a shiny gloss.
https://www.youtube.com/watch?v=h2lm6FuaAWI


I like the layer finish for every day parts.

[F5 Dave]

Printer at work costs almost as much to lease as my first house outright. It's not bad. And working as we speak, sadly out of my reach.

[polinizei]

There are lots of printer sevices in the web. So there is no need for your own printer for beginning. But CAD know how is a others thing.

[WilDun]

All sounds very interesting - watched the video on smoothing with acetone and also the one on using epoxy for smoothing but I guess the epoxy would affect the burnout.
I wonder how accurate the printers are for dimensional accuracy, especially after smoothing?

Now you guys will have to see it right through to the final product!
I probably won't be trying a printed pattern or corebox but I will look forward to watching your progress!

But ...... for me, it's back to making wooden patterns.

[polinizei]

I'm envious to all that wooden stuff...
Right now its easier for me, to find time for cad modelling and let the printer do the work...
If bury in the workshop would be an option, I would prefer it.

[Peiter]

The burnout temperature was raised slowly to prevent the ceramic shell from cracking. Eventually, the burn out was at about 500 degrees for 6 hours or so, according to the foundry .

Smoothing the PLA print is indeed still an issue I'm adressing, as the surface was quite rough in the final aluminium cast part. I have tried a chemical on the PLA but I didn't like it as it softened the part and with epoxy I'm affraid it won't burnout correctly. On the next iteration of my cilinder, I'm going to sandblast the parts softly with glass beads to smoothen the side surfaces of the print.

Polinzei, seeing that you smooth your parts with Acetone, I suppose you are printing with ABS instead of PLA? From what I have heared, ABS wouldn't be suitable to burn-out. You did interest me however to rethink the current lost PLA method and perhaps go ahead and design the core boxes in ABS similar to what you have done .

[Frits Overmars]

The burnout temperature was raised slowly to prevent the ceramic shell from cracking. Eventually, the burn out was at about 500 degrees...

Peiter, in a perfect world those 500° would be Kelvin, but that would be too much to hope for and maybe too cold to burn everything out. So are you talking Celsius (Centigrade is accepted too) or are you referring to this guy Fahrenheit?