I know this isn't a 3D printing forum, but I thought I'd try my luck.
We need to generate a scale model of a yacht and for some reason they want to go the 3D printing route as opposed to CNC. The model is about twice the maximum printed size.
I had this flash of inspiration and said "can't we model up the yacht using Inventors plastic tools, and then snap it together like a model you (used to) buy in a box? If i was careful with my tolerances it could work I think.
My main question is, does anyone know what form my pieces would have to take? Would they have to have a flat side that will orientate with the printers base and be built up from there.
Would they need supporting ribs for large thin planes?
depends what your printing with... sintering machines, undercuts are usually suported by the unsinted powder, cheap coil fed machines, (£800 on ebay!!! crazy) wont support undercuts.
I would forget inventor plastic options and just slice the model up sensibly so you have flat edges to glue together. There are a huge number of machines that produce very different results, finishes and material properties so make sure you get some samples first.
speak to the people building the model, they will tell you about wall thicknesses etc
good luck and post a pic when your done
Thanks Stephen, that makes complete sense to me.
I was also thinking... maybe keeping certain sections loose could be a benefit. i.e. taking the deck off to see the transverse bulkheads in position etc. Maybe I'm getting a little too ambitious here. But a set of numbered bulkheads in position could really help the guys when they are busy with the build.
I have had Inventor models printed with good success. "Shapeways" (Google them) has been my sole supplier, so I can't comment on other vendors, but my results have been quite good.
They have several material options available including a few metals. Your best bet would probably be their White, Strong & Flexible (a nylon-12 sintered powder) or for fine detail their Frosted Ultra Detail (a laser cured liquid acrylic), though each will have limitations based on size and accuracy. I have also used their Alumide (aluminum filled nylon-12), which is a bit harder but also a bit more brittle, to make mating threaded parts. Allowances have to be made for suitable clearances if parts are to be inserted into each other, snap fit, etc., but simple abutment assembly shouldn't be too much of a problem for simple features. I have also made palm-sized involute bevel gear sets in WSF that have come out quite well. A set of scale pipe sections with shiplap style joints between them were so accurately printed in FUD that they were literally a snug slip fit on assembly. There are some significant limitations on size, but for desktop/briefcase size models you're probably in the ballpark. Watch out for minimum wall thickness limitations for each material. (Generally they're in the 1mm range or a little less for the plastics, 3mm or so for metals.) Price is generally per volume of material consumed by the print plus a small fixed charge per model, though it is possible with some materials to have multiple (different) parts per model file. Check their Material Portfolio section for a full run-down. Feel free to ask me (or get on the Shapeways forum yourself) if you have any additional questions.
I will try to get back here tomorrow with more information - but the biggest problem with scale models is part thickness.
Do a Derived Component with the (desired) Scale option of your thinnest part. How thick is the scaled part?
Thin parts become essentially nothing when scaled for printing.
On your design I would expect a lot of thin parts and significant scale down to accommodate printer build envelope.
Thank-you both for your replies, the majority of my designs so far have gone to hand fabrication, sheet metal, laser cutting etc. And I am looking for opportunities in printing and CNC. Although it's almost disappointing to see your part come back- and it looks identical to what's on your screen.
I was wondering about surface thickness JD, maybe better to derive as a surface and thicken manually?
@BarryZA wrote:I was wondering about surface thickness JD, maybe better to derive as a surface and thicken manually?
I don't know what this means?
Are you saying to override the thickness (from what a true scale would result)?
Yes, we have done that.