Anyone able to advise on patternmaking for alloy castings

[silverbullet]

I am having to make my own special bellhousing https://club8090.co.uk/forum/viewtopic.p ... 7#p8132447" onclick="window.open(this.href);return false;
and whilst I do have some experience of dealing with a foundry, I dont have a lot when it comes to more than very basic patternmaking!
I appreciate the need for release angles and I do know what core boxes look like
Once I have a master made from various materials (alloy, wood, plastic, filler etc) and with allowances for 1:75 shrinkage and machining, I guess that I will need to seal the surface and apply some sort of release agent before taking molds to make the necessary cores?
This is the bit of which I am less than certain, I have never made a mold of anything beyond pouring plaster of paris into animal tracks!

I had patterning quoted for a SA bellhousing a few years back and that would have been around £2k for a 7-part set of patterns (because of all the undercuts)

[itchyfeet]

How about having a crude casting shape done then doing all the complex shapes by cnc mill.
Then you could pay a casting company to make the pattern of the crude shape rather than learing a whole new craft.

[silverbullet]

Even by simplifying the form as much as I can, the problem is the overhang for the transaxle and the pocket for the starter, this results in the need for "plugs" that need to be located accurately.
If it were a simple "bucket" bellhousing then it would be simple one to cast direct from a master.

Going to the moon again...

[Shaundon]

Might be talking nonsense but I've thought about 3d printing a shape in abs then encasing in sand or plaster before pouring molten aluminium in to burn away original shape leaving the aluminium cast to break out when cool. A bit like lost foam casting. Of course I've never tried this and have no idea if it would work.

[itchyfeet]

Might be talking nonsense but I've thought about 3d printing a shape in abs then encasing in sand or plaster before pouring molten aluminium in to burn away original shape leaving the aluminium cast to break out when cool. A bit like lost foam casting. Of course I've never tried this and have no idea if it would work.

That would not work but 3d printing a pattern is a good idea, I'd have thought that would work for sand casting.

[bigherb]

Traditionally the patterns are made out of close grained hardwood sealed with a spirit sanding sealer other types of finish may be necessary if the sand is being gassed. Mylands in Norwood used to make the spirit pattern finishes, as pattern making has died off they don't list them anymore but still they still do sanding sealers otherwise John Burns (Burnco) in Birmingham are the major pattern makers supplies. If any release agent is required the foundry would apply it. Shrinkage allowance can be variable depending on the casting and the LM type of alloy, best to talk to the foundry first. Nick Townsend at Essex Replicas foundry in Charlton might be able to help you.
These is the pattern for Beetle rear hub pullers I had made just clear shelac coating, cast in steel at Rye foundry sadly now gone.


Or pretty colours these are for wheel spacers cast in alloy.

[silverbullet]

I would go down that route if I had the 3d software & modelling skills.
Not sute if a polystyrene maquette pattern would be dimensionally stable on a job like this? I have been told that from such a virtual model an STL(?) data file could then be used to make the vasting patterns directly.
I suppose that a master could also be scanned and turned into a vector data file.
I could in theory make one from a solid blank, but carving one down from a 300 x 400 x 150 alloy slab would take a bit of time and no small amount of machine programming and manual work too.
If I had confidence that I could sell a few wbx/type 1-to-UN1 bellhousings to bug or rail racers then I would wade in and get it patterned. I suspect that its just too much of a niche product.

[Shaundon]

For free modelling software try out http://www.tinkercad.com. Runs in the browser so no need to install and really easy to use. With the supplied tutorial, you can be modeling in 10 mins. Allows you to download .stl files for 3d printing or cnc. Simple but basic so may not be sophisticated enough for your needs but I've seen folk do pretty good stuff with it.

[silverbullet]

Thanks for the link. I will check it out.

[silverbullet]

Had a bit of a play with tinkercad but it seems a bit cumbersome and manipulating the standard shapes is a bit hit & miss. I cant for the life of me work out how to truncate forms, seems geared more to making simple widgets and stuff for minecraft rather than complex objects.
I think that I will stick to working with plywood and U-pol!

[CJH]

How complex a shape is it that you want to produce? If it's something that you can describe with a few 2D sketches and key dimensions then I may be able to draw it in AutoCAD - no promises mind. AutoCAD can output an STL (lithography) file that is used by 3D printers, but it would probably also be a good starting point for making sand moulds. I saw a YouTube video of a sand mould being 3D printed and then used for conventional casting. Designing a sand mould that can be printed in parts and then assembled to form a complete mould, with suitable channels for pouring the metal, is obviously a complex and specialist process, but I can imagine that an AutoCAD drawing file might be a good starting point.

I think that 3D printing a full-sized bellhousing would be expensive, but checking the design via a reduced scale print might be more reasonable.

[silverbullet]

In brief, yes its a bellhousing. I'll pm you a few details Chris but its not really a simple shape!
I used to have Autocad LT but the licence is long expired and I am rusty as an old nail.

[CJH]

I had a play about in Autocad, and I found the 'loft' command that will plot a surface through a set of cross-sections. So apart from all the twiddly bits (bosses, flanges, strengthening ribs etc) the problem of drawing the basic shape boils down to defining the shape at various distances from the end face, and doing this for both the inner and outer faces. As a proof of concept I took some rough dimensions from a bellhousing that I have here, combined with some rough dimensions that Ian gave for the UN1, and made this quick model of the outer face.



Obviously there are practical issues to take into account. For a start, with this shape it seems unlikely that the bolts to mount this to the gearbox would be accessible at the bottom. And of course it's quite likely that I've misinterpreted Ian's dimensions. And then there are all the casting constraints to consider......

[itchyfeet]

This is the sort of job 3d modeling is good at, how on earth people used to do this stuff before 3d CAD I don't know, I suspect alot of trial and error and foam/clay models was involved. I gave up full time drafting before 3d was established but I'm supposed to be getting a solid works seat at work and so I'm going to have to learn, my drafting currently is limited to 2d Autocad but I have 3 draftees in the office to learn from.

To do it properly you would of course need to model the engine, gearbox, starter....

The dfaftees at work draw the part then discuss with the foundry about how it will be cast, wall thickness, rads, flow etc then incorporate these details.

[silverbullet]

Chris the basic form looks about right to me.
Yes there are a few bolt access/assembly matters to resolve but I seem to have identified all of them and have solutions to hand. At least all the UN1 bolts comfortably miss the VW four (on a polar/angular pitching btw) with no risk of overlaps or breakouts.