The Bucket Foundry

[Michael Moore]

I added some links to FB videos from Sunday to Jeff's thread at Alloy Avenue.

Brett, we played it safe and refilled the crucible after each cylinder. The small crucible might have been fine (I don't think Jeff has weighed a complete casting/runner system yet) but we had all day and it didn't take long to get another melt ready.

Jeff picked the two cylinders up from me this afternoon and should be machining them tonight. He'll bring me chunks of runner so I can see how hard they got.

Quenching both cylinders at once had the water vigorously boiling (and slopping several gallons out on the floor) for about 45-50 seconds. This was water that was @ 208F at the start of the quench. 20 minutes later it was still at 198F. We used a big plastic ice chest for the tank. Jeff plumbed an immersion heating element into the side of the chest that brought the 110F tap water to 190F in about 50 minutes.

Having a handle for the rack that kept my hands from being directly over the superheated steam made it a much less dramatic (and painful) process.

cheers,
Michael

[WilDun]

That's excellent work, I would have been happy to have half as good results!

[Lightbulb]

You have done very well indeed and thank you for sharing the weblink and posting here as you have been progressing the project.
Neil Lickfold

[ken seeber]

Sliding cylinder update. Finally got the time to set up the water jacket and cylinder and do the assy machining for the actuation rod and also the water jacket base mounting holes. These have been done on the same PCD as Fletto's standard gauge mounting pattern on his crankcases. This will allow interchangeability from the basic ex kart engine crankcase I will build up and Fletto's setup so it can actually be tested on a bike.
I think the pics tell the story, the small O ring on the rod will provide the sealing around the sliding rod.
As usual, progress is slow ( ) , a fraction of that ever so youthful Fletto.

[F5 Dave]

Wow, I used to think I did adventurous things, mb100/h100 triple ex port full case reed etc in the 90s, but I'm blown into the weeds by the real innovators.

Mind you I knew I was in the minor leagues when some clown showed up to a bucket gp, was it the late 80s? with a uni flow engine in a bike. I never knew such a thing existed.

[Tunisti]

First of all, I've gotta say that this thread is just way too impressive! If I wore a hat, I'd tie it to a weather balloon and let it go for good.

But, I have a problem. When making cores with sodium cilicate there is something terribly wrong. I've calculated the needed amount of CO2 to harden the cores, and they harden quite well with the required gasing. I've also tried several temperatures, concentrations and sands. But when the cores stay over night in the workshop, they have become very brittle and they're falling in pieces even from the smallest touch. I thought that it would be due to relative humidity of the air, so I packed the cores into sealed container with CaCl2 in small cup to remove the moisture. Still they changed just as brittle as before.

So, I'm thinking that I've got some wrong kind of sodium silicate. It's not easy to obtain in Finland and no one tends to know anything about it in here. The liquid is water clear, high in viscosity and doesn't seem to settle very easily via atmospheric CO2. What I've seen in videos, the liquid is brownish and the cores seem to withstad pretty rough handling when assembling the moulds. So, any ideas what's wrong? I've read that you can modify sodium silicate by adding some water to it, but I think it's not going to do the trick.

Can I modify my sodium cilicate with some additives/procedures, or should I search for that brown one? Where could I get it? I know that cores can also be baked, but that's not an option right now... Almost forgot, what is the glue that is used to seal mould halves and glue cores to another cores? Tradenames or chemical formulas could help me a lot.

-Sami

[WilDun]

I found this PDF on Sodium Silicate on my desktop but had to send it this way - the forum doesn't seem to want to handle PDF messages and I couldn't find the link!
They're not my words of course, but here it is anyway! I've been told that Sodium Silicate itself is actually a good glue.

BTW, Does anyone know how to send a PDF to this forum??

EDIT (TZ says in the next post below that it can be downloaded to desktop etc. and read as a PDF so I've deleted the original print).


-------------------------------------------------------------------------

[TZ350]

Here Will, this is your PDF, I was able to download it from your previous post.

[WilDun]

Here Will, this is your PDF, I was able to download it from your previous post.

Thanks TZ,
Still get the same thing as before but everyone will no doubt be able to download it too! - all probably more savvy than me.
Now if I don't understand this, then how the hell will I ever be able to understand the Ryger!

[jhenise]

I did a bit of messing around with this and that when I first started making sand molds and had a lot of problems and ok results.

SO i decided that since this foundry stuff is tricky enough as is, I would switch to a standardized system of chemicals that were sold to work together, and ever since i have had almost no problems on the chemical front!

I get everything through a local distributor here is the USA from a company called HA International. I have contacted HA directly and it is hard to get them to part with anything less then tanks and drums, but if you find a supplier they may be willing to siphon off buckets of this and that here and there.

http://www.ha-international.com/

For sand binder I use their sodium silicate formula: Chem Bond 4905 resin and 230 catalyst
http://www.ha-international.com/tds/321517.pdf

4905 can be found elsewhere too:
http://shop.clay-planet.com/chembond...on-bucket.aspx

You mix the catalyst into the sand then the resin, it creates a chemical reaction that forms CO2 in situ so you don't need gas and get perfect hardness throughout the mold every time.
I use AFS 90 silica sand, they call it "Oklahoma 90" over here. I story teh molds in a room heated to ~80 oF to keep dry a few days before casting of also seal cores in a container with CaCl2 or CaSO4 (dririte)

For mold release i use NixSTix NB
http://www.ha-international.com/tds/362307.pdf

To glue cores and molds together I use Core Weld NB
http://www.ha-international.com/tds/361635.pdf

And for smooth finish on cores and molds I use ArcoLite 408N zircon core wash
http://www.ha-international.com/tds/312225.pdf

I am not saying these are the only good products out there, but they work well in my hands. The point is if you find a foundry chemical supplier and but a set of things designed to work together they work well together.

[WilDun]

I did a bit of messing around with this and that when I first started making sand molds and had a lot of problems and ok results.

SO i decided that since this foundry stuff is tricky enough as is, I would switch to a standardized system of chemicals that were sold to work together, and ever since i have had almost no problems on the chemical front!

I am not saying these are the only good products out there, but they work well in my hands. The point is if you find a foundry chemical supplier and but a set of things designed to work together they work well together.

Thanks for the info.

[ken seeber]

Jeff,
Some time ago, we did a cylinder prototype for submission to the Australian Karting Association. This was offered as an alternative to the original cast cylinder for the Yamaha KT100J engine, a popular lower performance engine. One of the issues for this engine was the wide performance variation due to the casting variations of both the cast in, cast iron liner with cast ports and its matching to the cylinder jacket ports and passages. We were offering a cast cylinder with much more accuracy in conjunction with a shrunk-in CNC machined liner.

One of the key things for us was to ensure we had a good match between the ports of machined jacket and the machined ports of the liner. The key variable being the position of the cast ports in the finned jacket. So we concluded that the most accurate way was, when doing the initial machining op, to use the ports and cast bore as the initial datums. We felt this was much better than trying to hold the jacket, for example, by the fins or other external features.

So, we made up a vertical central post fixture that featured an upper and lower level of 3 balls each. These were expanded outwards by tapered plugs within the fixture, these being drawn together with an M10 screw, with a spring between the plugs. This centralised the bore. To set the height, we inserted a piece thru the exh port and this had a dowell that entered a hole on the fixture post, this piece could rotate slightly such that the roof of the exh port sat down on the 2 small balls. For indexing, the small piece had a screw head on one side and the jacket was rotated up to this piece so the side of the exh port registered against the screw head. As the jacket was the noprmal way up, the first machining operation was to do the cylinder head sealing face and the stud holes, creating a datum location for all the subsequent operations.

The rubber bands were a poor man’s way to stop the balls falling out. A sort of negative castration I guess.

[WilDun]

Ken, did the new cylinders improve the performance? - interesting looking machine guess it did the job well?
I often had to do a lot of jigs and fixtures for CNC machines etc. and it gave me the chance to have a free licence to think up weird and wonderful stuff and more or less do my own thing.
I just refused to operate the CNC machines, couldn't stand the thought of being tied down to production - missed out on learning to fully understand the CNC stuff though!

Was back in my workshop most of the day today so I must be coming right.

What's the latest on the sliding cylinder?

[tjbw]

Originally posted by ken seeber

...

Thanks for sharing that Ken, I bet it saved a lot of set-up time.

[tjbw]

...

missed out on learning to fully understand the CNC stuff though!

Was back in my workshop most of the day today so I must be coming right.

...

Wil, you understand how to machine, so already know what's required for CNC, it just requires a list of instructions and co-ordinates etc.

Good to hear that you're on the mend.