Hello chrisd,

You milled an object based upon tool paths created from a million plus point mesh. That is another thing entirely than using the exact points of the mesh itself for the tool path and trying to optimize that route. I'm not sure it's ignorance on the machinists part, they probably know your requested process is not optimal, so rather than have an unhappy customer they'd rather pass on the work, it's a reasonable expectation.

OSTexo, I appreciate your input on this as well as everyone else.

What I would have expected from the services I have talked to is "We can't do it because ..." with some specifics instead of leaving me to speculate about what the problem is. If they can't qualify the problem to the customer, is that not ignorance?

I'm not overly concerned with a high surface finish or accuracy, and I have clarified this to the services I have communicated with. So despite not needing these things, it appears that most services don't have an easy way to dumb down the process. The optimization and spline interpolation that is maintaining a higher accuracy and surface finish also makes for less wear & tear on the machine.

I am aware that the Jewelry industry, who I would not consider hobbyists, produce work from STL files by milling. In many cases they mill wax although, that has largely gone the way of 3d printing. They also machine small molds for non-investment casting.

There are some "micro" milling machines with a work volume of less than 3", and perhaps part of the difficulty in finding someone to do the work is relevant to the size I require.

I need a minimum of 3" XY, up to perhaps 12" -18" max XY and 2" in Z. I only need 3 axis machine (no undercuts).

Once I get some software to convert polys to nurbs, I'll have another go at this.

Hello chrisd,

You're not a customer, you're a prospect (big difference). You're asking them to do something they are not willing to do, perhaps due to your insistence of having them follow a flawed procedure which they know will produce an unsatisfactory result. A company is under no obligation to educate you on why they aren't willing to do something, but I gather by the content of some of your posts that you aren't making any friends by minimizing the operators experience and abilities to properly assess your project, and that certainly doesn't make them ignorant. Even with my very limited knowledge of mold production and injection molding, my limited experience tells me you're barking up the wrong tree when it comes to the software, process and type of equipment you think can be used to create those sort of parts.

"You're asking them to do something they are not willing to do, perhaps due to your insistence of having them follow a flawed procedure which they know will produce an unsatisfactory result. "

No, they need to tell me why they can't do it. It is the responsibility of the service provider to explain what service they can provide (or not). I can't tell them why they can't do it because I don't know.

Logically, only they can tell me why they can't do it.

"Even with my very limited knowledge of mold production and injection molding, my limited experience tells me you're barking up the wrong tree when it comes to the software, process and type of equipment you think can be used to create those sort of parts."

If you don't have any experience with what is being used in the Jewelry industry, then you would not understand that there is a big difference in what is an acceptable 3d file format.

I suspect this comes down to the fact that they are using much smaller CNC machines, and are not pushing around as much mass of the machine to do the milling.

Hello chrisd,

Again, the shop is under no obligation to educate you why they are unwilling to provide a product or service. I suspect that it has something to do with your insistence that they are too ignorant of their own shops capabilities. My knowledge is limited relative to others I know that view these boards that make a living in the mold design industry. That's not to say that I don't have mold or CNC experience, I just don't pretend to know more than others on the subject, and I've found that operators are happy to provide alternatives if you give them room to move. It sounds like you haven't done a thorough evaluation of the available technology to provide you with a proper part to acceptable tolerance using an efficient and relatively affordable workflow. You can bang in a nail with just about any tool given enough time and effort, but hammers do a better job than most for that task.

"Again, the shop is under no obligation to educate you why they are unwilling to provide a product or service."

If the service provider cannot or will not explain their service, that is simply poor customer service, and IS ignorant if the service provider believes that will improve the experience.

"It sounds like you haven't done a thorough evaluation of the available technology to provide you with a proper part to acceptable tolerance using an efficient and relatively affordable workflow."

Again, not true. A machine shop tells me they cannot machine without a nurbs based file. Another machine shop tells me they can machine from an STL, but the size I need is too large for their equipment. I did not fabricate this information, it is true.

If you can draw an intelligent conclusion from this info, then you are smarter than I am. That is why I say I need more information.

At this point it makes sense to deal with the service that can handle the size, and give them the file format they require to do the work.

And I have been polite with them. What you are hearing is the frustration at this point.

I am doing my job as an informed customer, seeking knowledge about a products or services. If a customer does not get that information directly from the service provider, then it indicates ignorance on the part of the service provider either about their service, or about good customer relations. If they do not want to educate the customer about their product, what then should the customer do?

Hello chrisd,

Honestly, you really don't need to convince me of anything. If you're being treated the same way by multiple shops you have to at some point step back and reevaluate your questions and requests especially if you're getting similar responses. I'll admit that it takes patience to deal with the challenges associated with fabrication markets, but that time can also afford you some benefit by exploring your manufacturing options.

Wow Max, that looks amazing! Are the generated surfaces actual NURBS surfaces? Or just the wireframe as curves?

>>Are the generated surfaces actual NURBS surfaces? Or just the wireframe as curves?
At this moment it's just a subdivided mesh. All lines is polylines, not even curves. But it's a just a first step. I've found the algorithm which can produce exact seamless nurbs-surface based on subdivided mesh.
So, stay tuned :)

Very impressive !
eric

Cool!
So what will be the export ? If you export in OBJ : object will be segmented again by the MOi mailer ?

Max, do you have a link to the algorithm?

I came across this: http://www.cl.cam.ac.uk/techreports/UCAM-CL-TR-773.pdf

Thank you,
- Brian

Hi Brian!

I've read a lot of documents about this. The most interesting is:
http://www.dgp.toronto.edu/people/stam/reality/Research/pdf/cc.pdf
http://research.microsoft.com/en-us/um/people/cloop/qteg.pdf
http://www.eecs.berkeley.edu/~sohamum/files/cs284_f12_project_report.pdf

Hi Max, the difficulty comes with the "extraordinary vertices" - vertices where other than 4 edges are coming off of it.

The subdivision surface behaves somewhat differently in those areas and those are the problematic parts to put NURBS surfaces on.

- Michael

Hi Michael.
I am not confident of success, but I'll try to implement Jos Stam's method of calculating the surface exactly.
----

[added] open mesh support

Hi Max,

> I am not confident of success, but I'll try to implement Jos Stam's
> method of calculating the surface exactly.

Ok, but evaluating the subdivision surface exactly does not really solve the problem of conversion of the surface into a different form...

I think what you will end up finding is that regular areas of the mesh will convert quite nicely but that there will be difficulties in the areas of extraordinary vertices.

But I'm looking forward to seeing what you come up with anyway! :)

- Michael