[turbinepowered]

Re: (chippievw)

Quote, originally posted by chippievw »

All manufacturers Im currently aware of use the sand method to make their engine blocks and heads, this incls Bmw, Vw, and Ferrari. F1 also uses sand. Manifolds are also made using sand, as you can see from my photoabove and the rundown on it. Im sure you could use foam for other stuff and Im not dissing it, BUT you do need a template in foam for each pour that would either have to be rapid proto-typed, cnc'd, on cut by hand. With sand you only need one.

I wasn't dissing you for taking the sand route, and hey, I'm all for watching people take the harder, higher road and come out ahead.
I just was thinking that for a one-off part you'd have less total investment in a lost-foam plug than you would a complex set of sand plugs and molds.
I've done a a smidge of casting, nothing fancy, just some knobs. Was awesome though, I'm tempted to build another furnace and melt away some cracked alloy rims to make some custom cast brackets.

 

[chippievw]

Damn, sorry I dident mean to come across that way with my reply, sorry! Im just following the route which I know, and have studied. Lost foam becomes isent simple either by any means, as you still have to form the shapes needed but in foam, and since you have to form them going the timber pattern route too, at least the prints/patterns are then re-usable.
And your right, the sight of molten alloy is amazing, i thnk you should get back at it again!
Brian.

 

[turbinepowered]

Don't worry about it, man. I'm kinda tired, I might have taken that a bit personally.
Props to ya, I'm out for sleep! http://****************.com/smile/emthup.gif

 

[ABA Scirocco]

Re: (chippievw)

FWIW, General Motors is currently using lost foam casting for some of their cylinder heads in fact, the cylinder head in my Dad's 03 Cavalier looks like it's made of expanded polystyrene that's been painted to resemble aluminium. I believe their foam patterns are multi-piece laminations with each piece of foam being separately formed in a permanent metal tooling die and the pieces bonded together before being coated with a ceramic slurry and cast in dry sand. I'm not entirely sure why GM uses this method of production, in most respects, it's not any simpler than the conventional sand casting methods. I'm sure there must be some cost benefit but, it certainly wouldn't make sense for anything other than large production run.

 

[chippievw]

Re: (ABA Scirocco)

Hmmmmm ok, sounds good. theres two types of this type casting, one is like lost wax, the wax part shape is sprayed with a slurry mix, once hardened the wax is then melted out of the ceramic shell and it is then filled with metal.
The other method using foam, the part is made from foam, it is then filled/packed into sand box, molten metal is poured in, the metal melts the foam on contact and fills the shape it once occupied.
Another use for foam is it is sometimes used to make patterns instead of timber, it is carved/cut to part shape negative, it is the sprayed with phenolic balls suspended in epoxy to create a hard shell on the foam part, the sand can then be struck off this now stable foam print like you would above with the timber pattern.
Hope this helps with other process http://****************.com/smile/emthup.gif


Modified by chippievw at 9:27 AM 11-28-2009

 

[ABA Scirocco]

Re: (chippievw)

There are at least three different variants of the lost foam casting process that I can think of, the common thread connecting them is that the molten metal comes into direct contact with a polystyrene foam pattern. The foam pattern is vapourized instantly on contact resulting in a void that's immediately taken up by the molten metal. The particular lost foam process used by GM with the ceramic slurry is the one that results in the best dimensional integrity and best surface finish so that the raw casting requires less machining to finish than could be achieved by any other non-permanent mold casting process, save lost wax.
The only way I could see lost foam casting being viable for short run or amateur casting of a water cooled cylinder head is if one had ready access to a CNC mill so that you could accurately machine a foam pattern, in several sections which could then be glued together to form the finished pattern.
The other use of foam you refer to, I would consider to be conventional sand casting just using an unconventionally produced reusable pattern.



Modified by ABA Scirocco at 2:05 PM 11-28-2009

 

[chippievw]

So onto the making the pattern for the flange.
Im going to make the pattern double so I'll get two flanges per pour, I could make it four but Im going to stick with two for the smaller flasks for the time being.
Here is a rough cross-section sketch of how the finished pattern will look once turned.
Remember this is for forming the shape of the main mould(s) and will be split after turning and both halves fixed either side of the parting plate.

So, onto the making the timber blank, I chose an off-cut of walnut I had because it was handy and near the size I need.
Any type timber can be used for this, but, It has to be one that cuts pretty good and turns easy on the lathe, soft wood for example is less able to hold fine detail.

I measured off-cut thickness and set the rip fence to double that and cut two pieces.
Chop chop.

Cut.

Now, onto the glue up.
Remember, a piece of paper gets glued between them in order to re-split them after turning, both halves mount either side of parting plate.

Normal water based Pva, nothing fancy.

Lay paper down on top of one, and drop the other onto it.

Like so.

Ill be able to mount that in the lathe once it has fully dried tomorrow.
Back to the flange for a second, a closer look shows the parting line from factory, which in this case will also be the position of paper/centre line on my pattern blank.

Ill be turning up the blank to make the double flange pattern once its dry.

 

[ABA Scirocco]

I can't wait to see more. I really should get myself a wood turning lathe and a jointer.

 

[turbinepowered]

Re: (ABA Scirocco)

The few times I've seen a complex shape rendered in foam for lost-foam sandcasting, the shape was assembled like you mentioned earlier, from thin sheets cut into appropriate cross-sections then glued together with a polystyrene glue.
If you had stamping dies you could probably assemble a foam plug pretty quickly, honestly.

 

[chippievw]

Indeed, but you would need one every time you wanted that particular part. With timber patterns, the time goes into making the first one accurately, then, hundreds of sand moulds can be striked off it.
Thats my view anyway, plus, I aint got a stamping machine lol!

 

[chippievw]

Well, it dried!
The ends are trimmed square.

Next, two pieces of 6mm Mdf are screwed either end, these have two purposes, they hold the two halves together in-case they come apart on the lathe, ruining your smile for Christmas, and they also provide positive location for the lathe drive and tailstock centre, if you were to drive them into blank, they would be hitting dead on the glue line and without a doubt would split the blank.

With the centres are marked a rough max diameter circle marked.

The blank is planed down to match this circle roughly, it saves knocking the corners off it on the lathe.

The centres are driven home before mounting.

Tail centre.

Mounted on my home made lathe! Actually, a lot of people spend thousands on wood turning lathes, why I dont know, all a lathe has to do is turn the blank, and have a solid tool rest, the rest is upto the guy holding the chisels.

A couple of seconds with the 25mm skew and its down to round, or round enough for the time being.

Basic sizes are transferred off the flange plus a little more for the shrink rule.

The marks are then made heavier by turning the lathe by hand to mark all round the blank.

You can now see all the particular lines where the various diameters change.

When turning a blank like this that has a narrow central diameter all turning or as much as possible is done before the centre diameter is reduced, that way the blank stays solid and whip is minimised.
In the next few photos you will see how the various sections are taken to near their final diameter.

Here Im turning the larger spigot that the core will hang on.

And the portion that the hose will clamp on.

Moving onto the spigot area in the centre of pattern. Note, always working towards headstock(left), that way keeping high cross-section to the left of me, ensuring that the blank can always transmit the torque to the area Im cutting without failing.

A screw is inserted at this point once centre is turned, that way clamping the two halves together incase of glue line failure, you do not want this blank coming at you.

Moving left always and onto the other flange pattern.

A quick check on all important dimensions is done and the part is then burnished, this involves holding its own shavings against it in order to shine the part, it works well.

 

[chippievw]

Removed from the lathe the 6mm Mdf plates can be removed either end, and the screw in the centre. A quick datum mark is make before parts are split so you know which way they came apart.

Screws removed its given a quick tap, this splits the paper glue line.

The waste either end is trimmed off to provide spigots of equal length.
A very rough drawing now shows what shape the core cross-section needs to look like.

And left down on the parting plate to give you an idea.

You can now see the cross-section transformation from the drawing I did earlier, to the part in real life.

A small bit of work on the flange area next and that will be the main pattern done.
For anyone thinking this was a huge job turning this, its approx 40min lathe time, so, not too bad.
(more to follow)

 

[turbinepowered]

Re: (chippievw)

Quote, originally posted by chippievw »

Indeed, but you would need one every time you wanted that particular part. With timber patterns, the time goes into making the first one accurately, then, hundreds of sand moulds can be striked off it. Thats my view anyway, plus, I aint got a stamping machine lol!

And you need to strike a set of sand plugs from your timber part each time you want to cast yours, too.

They're approximately analogous, I suppose it's all a matter of which you want to deal with or have the tools to deal with.
Sweet lathing job, too! That's going to look good when it comes out. Why'd you leave the flange end of the water necks so big?

 

[chippievw]

So now the with all the turning done the flange outline can be cut, I marked out the flange shape on a piece of card and transferred it to pattern.

Its positioned keeping the parting line in mind and the draft angles, if positioned wrong it would cause the sand to become locked on splitting the mould and ruin the moulds also.

Like so, the line is pretty faint..

The two halves are screwed together again for the cutting operation.

Quick run around with the coping saw and they are ready for a quick sanding.

Done

And un-screwed again.

Thats that pretty much done, bar mounting them on the parting plate.
Next up, making the core, thats pretty handy.
(more to follow)


Modified by chippievw at 1:59 PM 11-29-2009

 

[turbinepowered]

I need lathe skills. *Sigh*

 

[cnbrown]

Re: (turbinepowered)

Best thread on the internet

 

[chippievw]

Wouldent say that now cnbrown, but hell, Im enjoying it!!

Onto the core pattern,
Much the same process again, blank of correct size, centres marked, centres driven home, corners knocked off with planer.

Next, the main points are marked out on the now round blank.
Not forgetting the extra material either end for the ''hanger spigots''.

Taking shape, care is needed at curve area to provide sufficient strength and wall thickness on the finished part.

Nearly done, diameters are checked and the ends narrowed down with the parting tool to make cutting off waste easier.

When held up against main pattern you can now see how the core will look inside the mould.

If you look at the plastic flange two dimples can be seen inside, these are clearance for the bolt heads holding flange to head, I have to make two hollows yet in core pattern to ensure there is sufficient material here after the pour for when I counter-bore the bolt holes. I could make the same two dimples in the main pattern doing away for the need to counter-bore, but Ill be boring the flange holes in a jig anyway so its only a matter of using a stepped counter-bore bit. That way, the flange holes are bored, and the clearance around them is also done at the same time using the stepped bit. You'll see that later anyway, its very simple.

Some of you may be wondering where the groove for the ''o'' ring is, I chose to omit it for a reason, on some heads pitting can occur at the flange area making sealing a problem when using an ''o'' ring, this way, I can face the flanges flat and use a gasket or sealer, or I can machine the groove with a simple mill jig. That way I have options. If you wanted to mould the groove you could very easily by inserting a removable core on that end, but doing it this way I wont need to.
Next up, transferring the core mould to a core box. The core box is basically a box with the shape of the core I just turned inside it. You then pack sand into this box tightly, the box is separated and you then have the core shape replicated, but in sand and ready to place inside mould.
(more to follow)

 

[chippievw]

The core box is very simple, its exactly that, a box to transfer timber core pattern, to a plaster of Paris negative core shape. Once this is done the sand can be rammed inside it from the end, the plaster mould opened, and the sand core removed.
A rough box is made, the only real size that matters is the length, pins are used to hold it together, these should not be long, just enough to hold sides together, that way making taking the casing off plaster easy.

Next, pins are driven into the core pattern, making sure they are central, the heads are cut off after, bringing them to approx 10mm long.

It hangs central on the mould half, slots are cut to position it.
At this point I decided that I wouldn't bother making the indents in the core pattern for the bolt relief's, Instead it'll be way easier to just skim off a bit in that area with a small trowel before it gets placed in the mould.
You'll see that later anyway.
A plaster mix is made up, this is moulders plaster I had for repair to crown mouldings, but Im sure any plaster or repair compound will do.

The core box and timber core pattern are given a wipe of release agent, I have used light grease for concrete moulds in the past so its not that important what you use, Im sure car/household wax would do too.
The bottom mould is filled almost to the top.

The core pattern is placed in and tapped down onto sprigs.

Any excess is taken off with the mini trowel.

The plan is once the bottom layer is hard enough to coat with mould release, its given a coat, that is, all the exposed plaster surface and again the core pattern.
The top mould box is then filled up level with plaster and allowed to dry FULLY.
Once dry the sides are removed and the mould split, the mould faces are coated with sealer and it is then ready to use in order to create the sand core.

Because Ive never done this before I haven't a clue how long its going to take to set, but I do know that too much heat indoors will dry plaster too quick and it may crack. Not good!
So, ill leave that aside for the time being and go on to talk about the sand, mixing it, mulling it, and doing a practice mould mock-up to check workability, I need to see just how good it is and how fine a detail it can hold, and, whats its like to work with in general!

(more to follow)

 

[chippievw]

There are a few different types of sand used for sand casting, the main ones Ill be using in this thread and the head thread are>
Green casting sand.
Oil bonded sand.
And Co2 curable sand.

+The green sand is made up of fine sand, bentonite, and water, its an ok sand and holds detail well.
The good thing about this sand mix is its pretty easy make, safe, fairly clean to handle and its also cheap to make.
The bad thing about it is it dries out, this can be a problem if there is a delay for some reason and the mould is left for a while, it can dry out inside and start to loose its detail as it starts to crumble.
+Oil bonded sand is made up of sand, bentone, oil and a catalyst, this is a very good sand mix for fine detail.
The good thing about this mix is it doesn't require frequent re-wetting like the green sand does, as in it doesn't dry out.
It also doesn't steam like green sand does, and therefore requires less venting to release gases around the hot new part.
Bad thing is, if you could call it a bad thing is that its a bit dirtier for handling due to the oil and is also more expensive.

+Co2 curable sand is a mix of sand and sodium silicate, this is a very useful mix as once Co2 gas contacts the sand mix it sets hard. It can be used for complex cores, and the main moulds themselves, it also has the added advantage of NO moisture content as no wetting or binding agent is needed since it is totally chemical. I will be using this alot on in the head thread.
Onto the Green type casting sand, Im going to see how I get on making the flange moulds from this, Im still waiting on my Co2 and oil bond sands so I might as well give it a shot first.
Below Im going to mix up the sand recipe, and mull it ready for use.
You can see the size difference between the sand used for casting and normal building sand, normal sand on the right. Casting sand is as fine as salt.

To get the correct amount I filled one flask twice, this gives me the correct amount of sand needed to fill both upper and lower flasks. I ended up with 12kg in total.

To this 12kg approx 1.75kg of powdered bentonite is added.

Mixed in.

Now, this is when the fun starts and the water gets added, approx 2-3litres.

The next few steps are called mulling, mulling is basically working the living daylights out of the mix with your hands, I only have one book on casting wrote in approx 1945 concentrating on train and ship building and even with pig Iron great emphasis is shown towards this step. The more mulling the better it gets. What your doing is coating each sand particle with clay particles(bentonite) making them sticky, and in turn stick to each other, and hold their shape in the mould(Hopefully!)
Machine mullers are recommended for correct distribution of clay particles, but I have enough things for making so Ill mix it by hand as much as I can for a while.

You can see it taking shape as it fluffs up after even a short while doing it by hand, it can already hold thin shapes and is pretty strong!

I let it rest for a while before giving it another mix. Its everything I thought it would be even at this early stage.

(more to follow)

 

[BrothersinArms]

Re: (chippievw)

I don't know if it applies here but something I always do when making a multi-part plaster molds is to use a soda cap to put a small cylinder shape into the drying plaster. Soda caps are nice because you push them in maybe 1/3 the way and with a slight angle to the walls, you won't run into any overhangs. Anyway, when you pore the second layer it's helps to lock them all in line properly.
as for your sand,,, I did ALOT of glass casting (my undergrad is in glass blowing and design) and we used a olivine sand mixed with betinite. I don't know if it would apply to your project as well or not, but just some more info.


Modified by BrothersinArms at 3:22 PM 12-1-2009