--- In STMFC@..., "Arved" <arved_grass@...> wrote:

http://www.finelineprototyping.com/

The ant climbing over the chess set is pretty darn amazing to me. By biggest concern has been the layering effect of most printed materials, such as those from Shapeways vendors. but I don't see any in that green chess set.

That is because, as in most of these sales demos, the part was chosen for attributes that minimize the process problems. Note there are no real angular surfaces, they are either flat, or vertical.

"The (Micro Resolution) process features resolution as fine as 30-40 microns in X and Y, with 25 micron layers. We have built parts for some beta test customers that were measured to be accurate to within 15 microns anywhere on the part. There are limitations still regarding the thickness of the first layer of a feature, but it is about half that of the (High Resolution) build process that represented the previous state-of-the-art." (from the above site)

Is much more really needed?

Yes, I say it is.

40 microns is 0.001574" or, in HO scale, just a bit more than 1/8" (0.1370954 scale inches)

And 25 microns is just under .001", about the thickness an HO scale shingle should be. The question is, if you were building a tankcar dome, would you want it to look like a shingled roof?

Although the price of this kind of quality and fidelity is relatively high, certainly there's a possibility of making masters for relatively inexpensive resin castings.

I have been watching this technology for better than fifteen years now. There is always one little phrase inserted between the beginning of the review that says how easily the part can be modeled in CAD, and the end that tells all the good things that can be done with the resulting part... <after a minor amount of hand finishing>... The problem is, most of our parts are too small for hand finishing, or the hand finishing precludes including small details, like rivets, that are the whole reason for wanting to use this technology. Yeah, you can do a Dreadnaught end, and you can sand the pattern smooth. But then you can't include the rivets, and with no rivets, you might as well have used the solid model to program a CNC mill. In fact, I've done several this way in 1/4" scale, where the rivets are large enough to add one at a time.

Sergent Engineering is already using this technology:

"Sergent Engineering uses high speed 3D printers to produce parts one at a time. This allows us to provide a wide variety of variations of the product that simply would not be financially plausible with other manufacturing technologies due to tooling cost. This has also allowed us to quickly produce some special products for customers' unique circumstances. While some other companies have used similar technology to produce master patterns for soft tooling, Sergent Engineering is probably the only model railroad company to use the technology on a production basis." (from their home page).

Also from the Sergent Engineering home page:

"This technology works great when it works. However when troubles arise, causes of problems are not typically simple to identify and correct. This is where we are today. Stock levels on investment cast products are dwindling and production capabilities for replacements are offline...."

And:

"Update January 2012

While we continue to work to find a solution to the problems in manufacturing the investment cast products, we are also in the process of moving some castings to a die cast process instead. This will lessen the manufacturing load for the equipment and allow the printers to be used only for the really special products. This is not a quick or easy solution, but will be worth it in the long run. "

While I applaud Frank Sergent's putting some money out to explore new technology, I think it proves my point that the technology 'ain't quite there, yet.' Tom Madden worried that he may not see some of this technology mature in his lifetime. I'm ten years younger than Tom, and I have doubts about my lifetime.

Dennis