Any thinking about supporting the new 3MF format to replace STL for 3D printing?

Hi Bob, I haven't really heard from anyone yet that it would help them out.

Do you have a situation where using 3MF would help you get something done?

- Michael

It seems AMF was dumped in favor of 3MF.

3MF does not include curved triangles in the definition. This seems to have been a stumbling block for AMF, since apparently no developers supported curved triangles in the applications which offered it as an import/export format.

Good Morning, Michael

I continue to search for a solution to 3D printing my patterns for foundry casting. .STL or IGES is the preferred format for these printers. Unfortunately, the quality is just not there yet to eliminate post processing of my patterns, primarily because of the step pattern left behind by the slicing software and the limitations of the printer. The consortium that is preparing the format claims that it eliminates a number of problems inherent with .STL. I am always looking for solutions to the post processing necessary for my patterns. I don't know how the slicing software deals with calculations that may reduce or increase the volume of any given pattern. If 3MF provides for a smoother and more accurate output, then I'm for it. I have submitted one of my patterns to Shapeways to be printed on their EOS SLS printer to see if that might be a possible solution. Anyway, I was just enquiring. Thanks, Michael

Hi Bob, well the 3MF file is still going to contain inside of it the same triangulated mesh data as STL format does.

Since it will have the same geometry there isn't any way that 3MF files are going to be smoother than STL files.

The new functionality in 3MF is more about supporting things like components and name labels and various organizational things like that as far as I can tell, and support for embedded thumbnail images and colored faces. There isn't anything different from STL format in the area of smoothness or actual fundamental geometry content.

So supporting 3MF would not make any differences in the particular area that you're interested in, at least as far as I know.

- Michael

Hi Bob,

The problem is definitely 3d printing machine related not file format, all FDM (Fuse Deposition Modelling) machines will always render a layered finish to your models some finer than others, machines with a heated chamber tend to produce a better finish however more expensive and if a smooth finish is required then further hand work is necessary.

The model you'll receive from Shapeways SLS machine will be a much smoother finish, sort of a medium to fine sandblast, however models with shallow curved surface will still end up with steps because it is a layer by layer of a certain thickness process, so again, if a totally smooth surface is required hand work will be necessary.

-

Thank You, DannyT...

You are correct, I will still be dealing with step patterns, however, the thinner the layer the easier to post process the results and the more accurate in maintaining the cubic volume of the supplied file. My .STL files are provided within 4 decimal places of cubic volume. My best and most accurate results came from a ZCorp SLA epoxy hardened output, but, the output was too brittle for direct foundry molds. I continue to search for a study that provides accuracy results from 3D printers from high end to desktop printers. Warping continues to surface (pun intended) even from high end printers, when printing ABS. SLA is not a viable option. I've had SLA patterns warp in transit to the foundry. I have a prototype scheduled for delivery today from Shapeways that I am hoping will solve my problem to my satisfaction. The real proof of the pudding will be the cast weight in ductile iron from the provided split pattern.

Normally, I have to send two files representing the output as a split pattern for pattern plate mounting. I use dowel pin alignment for registration on the pattern plate, so, each split bottom has holes exactly placed to match a template held by the foundry. I add dowels to each split for alignment during finishing and then added and glued to match a 3/4" pattern plate. With the Shapeways printer, I was able to send both splits as a single file separated by including the dowels between splits. I was also able to create a thin shim that floats between the two splits. When I receive the output, I can cut the dowels in half and free the shim to be used by the foundry if additional weight is required to produce an output weighing 2 pounds 8 ounces when cast. This is an added bonus insuring that everything is printed at the same time and not susceptible to volume differences when splits are printed separately. At some point I will have to have the printing repeated so the foundry can cast two molds at one time creating a pair.

Sorry to be so long winded, but, it's accuracy that's paramount for my requirements. The thicker the step pattern the more difficult to determine what is needed to produce an accurate result. I paint all of my output with Acrylic Lacquer Primer and fill with a filler for follow up sanding to a smooth surface. The key...do I sand to the valleys or sand to the hill tops. Any method to improve accurate volume is my requirement. The preliminary discussion implied that 3MF might help.

Regards, Bob

Thank You, Michael...

If that is the case, then it won't help me much. I was under the impression that the format would deal with triangles a bit differently, i.e., not necessarily straight lines. From your description, I'll search for something else. If the Shapeways prototype comes out as smooth as the old ZCorp SLA epoxy hardened finish, I'll be happy. Unfortunately, it was too brittle for foundry casting. I should receive the 3D printed output today via USPS. I'm anxious to see the results. Regards...Bob

Hi Bob,

> I was under the impression that the format would deal with triangles a bit
> differently, i.e., not necessarily straight lines.

Can you remember where you saw that information?

I've read through the 3MF spec from here:
http://3mf.io/wp-content/uploads/2015/04/3MFcoreSpec_1.0.1.pdf

and the only mention of geometry that I've found in there is about triangles made up of 3 straight sides.


One of the things which I found confusing about 3MF initially was their main web page shows a sub-d smoothed model with quads in it on as their background graphic, so I read through the spec expecting to find something related to that in it but there was nothing mentioned. So there's a bit of a disconnect there between their fancy background graphic and what the actual spec allows for.

- Michael

The now obsolete AMF format did have support for curved triangles.

I don't think it would have improved the actual 3d print though. It would have made the files smaller and improved the time for data exchange.