View Full Version : Vacuum assisted casting in core-sand moulds.

23rd October 2012, 04:44 AM
G'day mates!

Ray has seen the model engine I'm working at.
I think he has also seen my vacuum chamber, but at that moment, I didn't have the time to try it. On a different thread, I promised you that I will report. And to report I have! What a success!

I'll show you some castings first to make you drool. :wink:

Lid for a bearing. At 1, thickness is 2 mm, at 2 it is 1 * 1 mm. Quite good, but doable in oil bound sand (with 95% reject).

Not necessarily to be done this way, but more convenient (small parts are much less effort this way).

And the mind-blowing part. Wall thickness is 1.4 mm.


23rd October 2012, 04:56 AM
And how did I do this?

The moulds are made out of cold box core sand. I glue both parts together and add a standardized (my standard) sprue.

Something like this:

Assembled with sprue
Core box and cores.

These moulds go into this contraption:
The box ...

The lid of the box (of course cast). The screws are M6 and are bleeders to adjust the vacuum.

The moulds are clamped from the rear with this spring loaded quick release crap.

On the box, there is a tube near the bottom. I connect a vane pump from a milking apparatus (no clue how you call that). This sucks 0.5 bar at 700 l/minute.
I switch on the pump, pour molten aluminium into the lids center hole. The vacuum sucks the metal into the mould, wait a few seconds and I have a new casting.


23rd October 2012, 04:59 AM
and I forgot that I made a video:
Vacuum assisted casting (http://www.youtube.com/watch?v=QijGV2G26JI)


24th October 2012, 01:12 PM
Hi Nick,

Thanks for the video, amazing work.
Pretty sure thats a lot more pump than I have. Though at my current rate its not going to matter for a while.
Is the rest of the box wood?
Did you give any thought to preheating the mould?


24th October 2012, 03:00 PM
Hi Nick,

Many thanks for the video and excellent description, very clever. :2tsup: and quite simple when you see it all finished and working. I bet there was a lot of experimenting to get to that stage.

I'm very impressed by the finish quality that you are getting on those parts..

On a different thread, I promised you that I will report. And to report I have! What a success! I'll show you some castings first to make you drool. :wink:

Drool.... :2tsup:

Can't wait to see the finished engine, it's shaping up to be a real gem.


24th October 2012, 06:37 PM
Yes, the box is just made of MDF, that's what I had at hand.
I have cut a slot along the top of the box and put a silicone tube inside as sealant.

I'm not preheating the mould, and until now I see no need to do so. The part in the first post, last picture is almost more than I expected.

I got a few more ideas (plan B, C, ...) that I could try if that one failed. But it didn't.
The ideas are around directing the flow of air inside the cavity and thus to direct the suction.

If you look close to the mind-blowing-casting, you see a small nibble (diam. 1 mm, length 1 mm). Plan B was to drill a 1 mm hole all the way through the mould. The molten metal would have been quickly sucked into that part, flown into the tiny hole and solidified there and reduced the vacuum from that moment on. Then, air flow would be more around the separation plane of the two halves. That was Plan B.

Plan C would have looked like this:
Cut small slots from the outside that follow the cavity, leaving maybe 1 ... 2 mm of sand. Would have looked like cooling fins. Again, that would have allowed me to direct the vacuum. As soon as the melt reaches that region, it would have sealed the mould at that place and suction increased somewhere else.

Plan D would be like this:
At the separation plane, cut channels into the core sand. They would stop a bit before the cavity. I could file them into the core sand, or integrate them into the core box. I could glue strips of paper into the core box's bottom (that makes the separation plane). This would make a shallow channel for the air.

If you didn't understand my bad English, ask and I'll try to explain harder. :)

I bet there was a lot of experimenting to get to that stage.

No, it wasn't. It was quite straight forward. The only thing that got in my way were the experiments with hot chamber and cold chamber die casting. That never worked (except once).
The only problem with the vacuum chamber was, that I had too much vacuum. The melt was sucked into the core sand, and the part looked like it was sintered out of aluminium and sand. And got a different shape than intended.
After the bleeding holes, all went smooth.


24th October 2012, 08:21 PM
Hi Nick,
Didnt mean to imply that your castings needed improvement, as you say they are mind blowing. I was just wondering if heating the moulds was something you had tried along the way and disguarded.


24th October 2012, 10:31 PM
Didnt mean to imply that your castings needed improvement, ...

... and I absolutely didn't understand it that way. And even if you meant it that way, you just can say to me where you think it lacks.
I'm not politically correct, and I don't want to be treated that way.

if heating the moulds was something you had tried along the way and disguarded.
I have a plot of how I think things work. If that plot fails (yes, that happens!), I analyze what was the cause and look for the simplest and most promising cure. Heating was not on these plans, because it looks to me a bit tricky to get repeatable results. I can heat the moulds up to a known temperature, but the handling until pouring has to many variants time-wise that I'll end up with an unpredictable temperature.


24th October 2012, 11:20 PM
And in case someone wants to know where all this crap belongs to and what is the progress of that eternal project ...
I have a thread here (http://www.homemodelenginemachinist.com/f31/man-diesel-engine-1907-dm-2-100-a-19073/). That forum is a bit better suited to my project. So I think you pardon me.


25th October 2012, 02:26 PM
Hi Nick,

Thanks for that link, I've not seen that forum before, but I see a few names that I know and I noticed a link to another forum I'd not seen before either ( I must lead a sheltered life :) ) http://www.alloyavenue.com (http://www.alloyavenue.com/) looks interesting.

When Josh and I visited you for that memorable day's casting and barbeque I came away thoroughly impressed and blown away by the work you were doing, now, I see that I was only seeing a little bit on the surface..

I now know you are completely insane ( in a good way ) :2tsup:

Simply stunning... what a great project..


PS... I'm going to try your vacuum casting technique, I've got a part to make that I think it will work nicely for.

25th October 2012, 06:04 PM
Just for clarification:
I do not think nor claim that I have invented something new. In foundry work, all has already been invented.

I just combined what I had or got easily.


26th October 2012, 01:37 PM
In foundry work, all has already been invented.

Hi Nick,

Yes, but it has to be re-discovered by patience and careful experimentation to re-discover what works. You are much too modest. :)

The books I've read on pattern making have done nothing to make me think it is easy to design a pattern that will cast well.
I suspect that in big foundries producing thousands of consumer goods they have computer modelling software to predict how the metal will flow.

One thing I have learned from watching your work is that I need to do a lot more with cores.

Have you ever thought about die casting?


26th October 2012, 06:49 PM
I suspect that in big foundries producing thousands of consumer goods they have computer modelling software to predict how the metal will flow.

A friend working in a foundry (QC) will tell different stories. The engineers come up with a new part, and the old foundry man says that this will never work. And then, they discover that they have wasted hundred thousands in something that does not work.
I think a lot can be solved by observing and correcting and following known and working patterns (patterns in both meanings). The old-timers had no other chance. They had no simulation software. But still, they made complicated castings.

Have you ever thought about die casting?
In steel moulds? Yes, years ago I played a bit with that. And a few months ago, I (we) have cast zinc pre-shaped parts (were machined all over) in steel moulds. You just have to experiment with the mould temperature and see when it works.
In production, they do it the same way. Heat up the mould, place a burner here and there (by experiment), cast a few parts and throw them away (mould gets the right temperature) and then it works. They run it 7/24, so the temperature stays stable.

It is fun to see what works and push the envelope further with new castings.

You have been talking with Jürgen (from my village, with the BMW spares). I try to make him want a model of the BWM side-car engine. And I'll make an experiment how small I could cast the cooling fins of a cylinder/cylinder head.

I learned (I guess), that these moulds are called "shell cores". Is that right?


27th October 2012, 04:03 AM
Have cast a few more parts with vacuum.
But this one is the smallest one. Won't need smaller ones.

A look at the split separation plane.

Two on a tree. Parts will be sawn off at the red line. Those two "antennas" are artifacts of the split plane. They are 1.5 * 1.2 mm. Now I know I could cast these too. :D
The oval feature (seen on the left part) is 1.5 mm wide.

Core box and shell core.


27th October 2012, 05:59 AM
Hi Nick,

these are looking goooooood. Congratulations!:2tsup:

Best regards from Sindelfingen


Just two remarks

AFAIK / remember "Shell Cores / Sheel Moulds" are made also out of sand in the croning or hot box process. A resin bondend sand is used, this sand is poured on a hot pattern out of metal or into a hot core box. Then the resin bonds the sand because of the heat (melting?) . After a short time the pattern is turned around or the core box is turned over so that the not bonded sand, where the resin did not "melt" falls of /out. This now forms a "shell mould or a shell core" with a relative thin wall. In youtube a video showing shell core making can be find.

Are these parts for a secret bavarian time machine?
The post date is already saturday and here in the office it's still friday.

27th October 2012, 07:13 AM
The post date is already saturday and here in the office it's still friday.

As you know by now, the OZ-people are always ahead of us!

Stefan is my casting mentor.
To my big fortune, he came to my very first casting party. We were all clueless. All we had was a furnace, sand, a few patterns and a lot of ignorance.
At that Saturday, he showed us all the basics.

He was a mould and pattern maker by trade, so I couldn't find someone better.

So I'm very thankful to him. I don't know where I would be today with his support.


27th October 2012, 03:27 PM
Hi Nick,
Are these parts for a secret bavarian time machine?
The post date is already saturday and here in the office it's still friday.

Hi Stefan,

Shhh... that project is classified top secret..

By the way, welcome to the forum :2tsup:

They are 1.5 * 1.2 mm. Now I know I could cast these too. :D

Nice work.. I think I can do better than that... :) I have an idea.. more later


28th October 2012, 01:50 PM
Hi Nick,

For casting small parts with fine features, I'm thinking of a simplified die casting process. Basically the idea is as follows. It's the opposite of vacuum :)

1. The die is made of multiple parts 2-3-4 whatever it takes.. So you can get the finished part out :)

2. The plunger and reservoir is be made to screw into to the pattern, so as to be re-usable

The die and the reservoir are assembled as one part, the plunger is used to pressurize the molten aluminium into the pattern ( a bit like a linotype machine I guess )

Load up the reservoir with a preformed aluminium slug in the pattern in the electric furnace, that way I can get precise temperature control of the die and reservoir.

I don't have to worry about the aluminium solidifying in the pattern, since it's the same temperature as the reservoir.

When the aluminium reaches pouring temperature, remove the assembly from the furnace and operate the plunger until the aluminium reaches the riser, and then block off the riser and put a load on the plunger to maintain pressure while it cools.

Since the molten aluminium inside the pattern is under pressure ( not too much ) it should flow into all the corners and fine features of the die.

If that's not making sense, perhaps I could should draw a diagram.


28th October 2012, 07:11 PM
remove the assembly from the furnace and operate the plunger

So that is a hot chamber.
The best I could get was 4 strokes of the plunger, then it seized. BTDT.

Here is a video. It is not public, because it failed and I gave up that path.
Hot Chamber (http://www.youtube.com/watch?v=OZQHNeL8S60)

I don't even have a video of my cold chamber die casting machine. That one also failed. But better, with more efford. :D


29th October 2012, 05:38 PM
Hi Nick,

Thanks for that video, I am thinking of a similar process, but with some differences to your approach,

first, steel dies, and it's a one shot process, not multiple parts, second the reservoir screws into the die, and third difference, I'm thinking to heat the whole assembly in the oven..

The parts I'm wanting to make are similar to commercial diecast boxes for a small custom electronics project, except for the small number I want to do, the commercial tooling cost would be way to expensive..


PS.. I should have known you would have already tried it.. :)

29th October 2012, 06:39 PM
I have some buts ...

The plunger WILL seize. You will have to bore it out. A hammer is not enough. I bored out four, then I got bored.
How will you remove the dross when you assembled the whole thing and then melt the aluminium?
Steel dies are not that hot. Something like 200 °C or so.

To some extent, my cold chamber die casting worked better. The problem was, that the amount I wanted to cast was too small. As soon as I filled the chamber, the aluminium solidified (almost) and it didn't get through the bore into the die anymore (except once).
That plunger worked much better (and I found a trick to make it seal), but I used weights to push down the plunger. More than 10 kg weren't enough. I think that requires hydraulics or a long lever. I have a book about die casting. If you want, I can look for the pressures involved.
Hot chamber with ali is very uncommon. The reason is, that the ali fuses with steel. I tried CI and SS, none worked for too long. Maybe chrome plating helps?

I'm really happy that I gave up that pressure die casting route. And I'm very confident that the vacuum assisted casting will work with steel dies too.


29th October 2012, 10:13 PM
been watching this , and have been left with out words ,


i had to say something , as a newbie to casting , thank you for posting
cheers ken

at 999 posts where do i post my 1000 , wood turning ,. instrument making , CNC , or here ?

30th October 2012, 03:09 PM

Hi SawDustSniffer,

I remember when I cast some bronze parts, and was feeling pretty proud of the efforts, when a friend pointed out that the Greeks and Romans were doing this stuff 2000 years ago...
And as Nick said earlier, this stuff isn't new, people have been doing this stuff for a long time..

We are drifting off the vacuum casting topic, so I won't hijack this thread any further with die casting, I'll start a new thread for that.

PS.. I think I'm almost catching up with the Bronze Age, 3200BC to 500BC, still haven't got to the iron age :D