For a non-robotics (and non-commercial) project, I have an object that I intend to have 3D scanned and converted into a CAD file of some type, ending with the object being 3D printed somehow.
A company offers fairly low-cost scanning services. I figure I can get the object scanned, import it into Autodesk Inventor, clean the scan up* and send it out to get printed.
My question is into which file format the object should be scanned. A 3D mesh model is less costly than an IGES or STEP file, but I am not sure of the difference between the two, especially in terms of my ability to import it into Inventor and then work with the file to clean up surface imperfections.
Then I have questions about 3D printing: This is a prototype, so it needs to be inexpensive as a proof of concept, but eventually these may actually be produced (Just a dozen or two per year). Ideally these should be chrome-plated plastic (I assume vacuum vapor deposition for “plating”, other ideas?).
Any thoughts on the best method/material for the prototype (low cost), and then for production (can be ‘plated’)?
*I have seen the results of raw scans, and they always need cleaning, plus the object is not a perfect example with some surface imperfections already.
The object is like a radio trim plate, about 7-1/2 x 2-1/2 x 1/4" thick, wall thickness generally 1/16" to 1/8" with no fine details, and mostly air as it is a frame, not a solid rectangle…if that helps.
I have been working with Rapid Prototyping and related technology
for many years. Here is how I would approach your problem.
Get the point cloud from the scan into the CAD system you want to use. Use the points to determine the locations of critical features. I have limited experience with scanning but what little I do have indicates you are best off to have a clean model. Hmm not explaining this well. You want to build a CAD model of your part, but only use the point cloud as a reference to see if you are close to where you want to be. Especially if you are using Inventor, whose surfacing capabilities are pretty poor. I do not think you will be able to do much editing of a scanned surface with it.
The clean model can then be exported as a .stl file (your 3D mesh) virtually all CAD programs can do a .stl, though some will have you do a “save as” and others (like Inventor) have you “Export”. Look both places. You might have to set some parameters like the distance .stl data can vary from the actual surface. If this number is too big you get obvious facets to the surface, if too small you get a huge file.
A good plater should be able to plate the plastic. If they can get copper to stick, which they can usually do in solution, then they can get any other of the common plating metals to stick to the copper. The metal cladding will also make the plastic much stiffer, which is something to think about if you are planning any snap in features.
There are different materials/processes that can be used for the part itself. Stereolithography is the oldest and probably has the finest detail capability. It takes place in a big vat of UV curing resin. Objet is a similar process but they have managed to get rid of the vat. Both have a variety of resins with different degrees of flexibility.
Another Process is Selective Laser Sintering (SLS) which builds in a bed of nylon powder. SLS has a naturally rough finish that grabs plating very well, we use it that way for airplane parts.
Another process is Fused Deposition Modeling or Stratisys. This process uses extruded thermoplastic to make the parts. This can be either PLA or ABS plastic.
Any process is likely to have to be sanded or otherwise prepped to get a smooth surface. I have personally used plating on all but FDM for various jobs. Any competent plating house should be able to handle that part of things. I can give you a referal here in CA but that might not help you much.
So I start with the best object I can get - not one that had obvious defects - and get that scanned into a point cloud (I guess this is an OBJ file?).
I then cloud to either verify my drawing -or- help guide me as I draw. (I can pay them to “reverse-engineer” the object, essentially doing the drawing for me, but that would be several hundred dollars)
Whatever’s printed, I need to finish it to the surface characteristics I want - mirror finish in this case.
Then it goes to the plater. How thick is 3-step plating? I guess in the .001 or .002 range? I’m sure the plater can tell me exactly. Vacuum deposition is just a few ten-thousandths, I wonder about pricing for that.
Do I have it kinds sorta right?
Do you ever print items that end up with a chrome mirror finish?
Looks like you have the process figured out. I do not recall the plating thickness we used, but I think it was around 0.0005 for the final layer. I have never done “mirror” but we did do “shiny” and I have seen plated parts that were mirror finish at a vendor. (cup holders for a Lear, or was it a Gulfstream?) Unfortunately that vendor folded shortly after that.
One of my favorite projects was producing a custom chess set as a retirement present for our division president. The chess pieces were different airplanes made by Northrop and Grumman as this individual had been part of both companies. Of course the Northrop side was “gold” and the other side was “silver”. The plater offered to use the real stuff but we decided we were spending enough on the job already. The pieces were made with the Objet process to get the better surface finish. We also did a set in SLS to see how the rougher finish looked and as spares . Both were stunning but the Objet definitly looked better. I am not sure just how the finishing was done as it was done by a vendor. The final polish may have been done by the plater.
Most of our other plating work was on stuff that would be buried in an airplane anyway so we really did not care too much about finish.
Our most recent division president recently announced he is retiring at the end of January, I am waiting for a phone call…