3D Printing Those Missing Parts

[SimonBrown]

This is not directly an MV related post, but it may be relevant.

I'm restoring a Yamaha TY250 motorbike. A real 80s classic two stroke. Some of the parts are now unicorn grade and the LH crankcase cover is one of these. Made from cast magnesium alloy its a) fragile and b) difficult to repair and c) no longer made.

I happen to run a company that specialises in 3D reconstruction. So I decided to scan, reverse engineer and 3D print one....the first sample for fit and form (not function...read on) has just arrived and I thought I would share the results. Firstly, the scan of the broken part:

Then reverse engineered into a 3D CAD file:

Damaged and printed alongside each other:

Trial fit:

Now this version is plastic and one that wont withstand high temperatures - hence not good for the "function" of fit/form/function - but it does prove the model is accurate and we can now print in carbon filled nylon (under bonnet injection moulded car parts) or in metals such as aluminium, steel or even titanium.

I will report back when the "production" part arrives. In the meantime, maybe this will trigger a few thoughts...

[terryb]

can you print in metals now?

[SimonBrown]

3 minutes ago, terryb said:

can you print in metals now?

This one comes under the "Can-I-Should-I" bracket.

Short answer is yes and a wide range of materials are available.

Longer answer is...how much do you want to spend? And do you really need it in metal?

For me the cost of a metal housing is...expensive when compared with a under-the-bonnet nylon material. So for now I will go with a plastic that is heat and chemical resistant.

[David Herbert]

Or you could 3D print a plastic pattern with a shrinkage allowance and sand cast in aluminium from it or to be posh you could 3D print a wax pattern and lost wax cast in metal from that. It is creating the original CAD file that is difficult and you seem to have that sorted.

David

[SimonBrown]

19 minutes ago, David Herbert said:

Or you could 3D print a plastic pattern with a shrinkage allowance and sand cast in aluminium from it or to be posh you could 3D print a wax pattern and lost wax cast in metal from that. It is creating the original CAD file that is difficult and you seem to have that sorted.

David

Yes to all. The wax pattern printing option is probably the way forward....and once in CAD applying a global scale factor to cope with shrinkage would be easy.

[terryb]

or just billet mill it from the cad?

[SimonBrown]

52 minutes ago, terryb said:

or just billet mill it from the cad?

Another option - yes.

[draganm]

On 11/9/2023 at 7:35 AM, terryb said:

can you print in metals now?

not really, they call it " printing" but it's laser-sintering in a bed of AL oxide powder. The machines are well over a million dollars, it's expensive to do, and the example we got (after a student just HAD to have one), was rather poor. Very grainy/porous texture, prone to oxidation after you machine a clean face, brittle, and you can't weld on it due to the very high oxide content.

     Can't comment on other metals, but the AL was not good IMO. This was 5+ years ago, so not sure if it they improved it?

A competent shop that does a lot of 3D profiling on a CNC mill could probably make that part in 10 to 12 hours and probably wouldn't even require 5-axis mill

[teletech]

10 hours ago, draganm said:

not really, they call it " printing" but it's laser-sintering in a bed of AL oxide powder. The machines are well over a million dollars, it's expensive to do, and the example we got (after a student just HAD to have one), was rather poor. Very grainy/porous texture, prone to oxidation after you machine a clean face, brittle, and you can't weld on it due to the very high oxide content.

     Can't comment on other metals, but the AL was not good IMO. This was 5+ years ago, so not sure if it they improved it?

A competent shop that does a lot of 3D profiling on a CNC mill could probably make that part in 10 to 12 hours and probably wouldn't even require 5-axis mill

SLS has improved miles in the last few years.  At this point they are making flyable commercial aircraft parts, surgical implants, and a host of other end-use parts and I think you can have a decent machine for a hundred grand.  The hundred grand might only get you a bound-powder-fusion machine that prints metal parts you then bake to final form (remember shrinky-dink toys?).

You wouldn't need even a 4-axis CNC, just cut the cavity, flip it over and cut the exterior.  Aluminum should be about 3 hours machine-time with setup if you don't need the contour lines to be too perfect.

All that said, printing a pattern and having one (or several and help others out) cast from that pattern would be the most authentic way to do it and still likely the cheapest... IF you can find someone to pour you some metal without making a big production out of it.  I've had luck getting an artist who works in cast metal make small parts.

[SimonBrown]

Update on progress.

One advantage we have today is a wider range of commercially available and cost effective materials, and this includes the kind use in high volume plastic parts used for under the bonnet parts in mass production.

Left to right; Original cast magnesium part now with two broken lugs and plastic filler. PLA fit/form SLA printed tester to validate model. Finished first-off production part in nylon with shot peen finish.

Apart from shot peen finish this part is straight off the printer.

The material is used in under bonnet components and is thermally stable enough, resistant to oils and is watertight.

A spot of off-road testing will prove this out. Once the motor is back together.

Is the match exact?

No, but as the bike is Japanese I am very comfortable with the Kintsugi method.

Full story here:

Reverse engineering blog