Newbie - how do I find a CNC shop that would deal with individuals? Please help!

[Tkamsker]

Question How precise mußte that part be ?
I Do cnc r&d stuff
I Played around with lost wax (pla) from 3d prints and Alu which Works but in your case some Hand Rework is required.
Probably that Route is in Budget and constrain reach for you ?
If you have a Steel foundry somewhere Close ? I had an request where i 3d printed the stuff which is not machineable and Then i found an.steel company which would give it a Try to cast it and do it in Steel but then the price was too high for that Company ..
If you want to wait till augtum (i build a trunnion table) for my cnc it might be a good test ,... Thomas

Gesendet von meinem SM-N9005 mit Tapatalk

[alexhawker]

[QUOTE=Boris123;1491526]

I think your surfaces can still be represented in NURBS-based (non-mesh) programs. These programs are not only limited to cylindrical/conical/spherical/prismatic geometry - those are just the basic tools. [...] No offense, but odds are if you can't afford DMLS, you may find 5-axis CNC similarly cost-prohibitive.

- - - Updated - - -

Nothing offensive about giving people helpful information. Perhaps I spoke too soon - I was under the impression there were something like fifty DMLS machines in the entire world, and little players like NASA have to wait for months to get an order filled. I'll look into this. Still, every project has a budget, and there's a price point above which this just won't make sense.

I did forget about using B-splines, too. it should be possible to reduce the number of points drastically. Might take me a while to convert the program though.

Thank you for the good advice!

Happy to help! Consider drawing in or including in the RFQ an acceptable radius for internal sharps, or drawing that in directly. That always makes it easier for those quoting your part(s).


Sent from my iPhone using Tapatalk

[hanermo]

It would help a lot if you included details.

Material, price, size !, use of, etc.
Tolerances !
I mean, if several thousand $ is ok, there are any nr of shops that can do it.

Are there undercuts ? Are they important ?
Hard tool steel or brass ?

20 mm diameter or 300 mm ?
Accurate to 0.01 mm or better. os just so it looks good ?
Needs to hold 2000 bars pressure (eg) or just needs to fit in a hole ?

Thats why machine shops dont want to deal with small customers, who dont have a clear idea of what they need and spec it.

Depending on the above variables, the part migh cost under 500 (4 axis hobby shop, low tolerances, decorative) or any nr of thousands and up.

[Tony T]

Boris, Post your stl file up here. Someone may be able to open and change the format for you.

[victorofga]

for following readers...


boris posted everything about this project..

2 in lenght 4.4 in dia

hardened steel...

without he mentioning about tolerances, details can be made with a certain tolerance about 0.05 mm

while the whole model has virtual undercuts, the unrolled model don't have.. however what is ""vertical" wall in the model that will require 5 axis machining due the conical tool..


i think the stl model also could work, using one pattern, toolpathing and indexing..

==============================================

converting from stl to solid with automated process, will result same much little surfaces..

some type of programs can rebuild on a hard way.. programs that working with handheld laserscanners..

i think simplest is rebuilding one pattern and arraying around center..

also, even its a nurbs surface still it will be a large file..

[victorofga]

mightbe a help, if you don't want to deal with rebuilding model..

powershape by delacam makes something you need..

it looks like create of selected areas or similar way nurbs surfaces..



they have forum.. try to post your model, and someone might helps you..

Delcam User Forum ? View forum - Power Solution

beside mastercam powermill knows the most..

[Boris123]

Hello everybody! Again, thank you so much for all the help.

Looks like our biggest problem is not the geometry or file format. We've grossly underestimated the cost of 5+ axis CNC machining.

I should be able to fix the software issues (round off the internal sharp angles, convert the surfaces to b-splines, generate a STEP or IGES file instead of an STL, maybe plug in one of the existing modeler libraries, etc.). It still won't make it feasible, if the cost is too expensive. A $20,000 mousetrap won't sell very well.

We're making a prototype right now, but if it works, we'll need to figure out a way to make more than just two of them.

It's time for us to regroup and think. We're investigating alternative technologies - the spark-erosion that someone mentioned on the forum, waterjetting, etc. I'll post some updates on the thread so people know what's going on.

For example, maybe it's possible to inverse the surface and 3d-print a shaped electrode for a spark-erosion machine. Don't know if it's completely harebrained - we're just starting to look into it.

For those who asked for more information:

* I'd rather not place the STL file on the public forum, sorry. Just don't want to get into possible legal problems. I'm okay with showing it to individuals, but not to search engines.
* I don't think it's possible to convert the file automatically, and it's too hard to convert manually. I believe the best approach would be to change the program that generates it - and I'm comfortable doing that myself.
* The reason it's so hard to convert the STL file, is because what it contains. The surface is made out of about 170,000 individual triangles. STEP or IGES files aren't designed to work with data organized like this.
* Automatic file conversion won't work, because there are too many unordered triangles, that make for too many possible combinations.
* I tried several different file conversion utilities already, they either freeze completely, or work for several hours and produce unusable files (a quarter gigabyte in size).
* The material is hardened steel. Something like H10, with surface hardness 50...55 HRC.
* The tolerances on the cut-out groove are not high - maybe ±0.2 mm for the shape itself, ±0.01 mm for the overall positioning, the inner corners can be rounded with a radius of about 0.5 - 1.0 mm (or more, if needed).

Again, thank you so much for your kind help!

Boris

[murrayclair]

Your most cost effective solution will be to find an "investment casting" foundry that has a rapid prototyping (3D Printer) these ones build the part in wax. Then this wax model is used to cast the part. It shouldn't be more than a few hundred bucks. You may then need a machine shop to skim the faces and diameter etc after casting.

Sent from my GT-I9305T using Tapatalk

[hjl4]

I agree with with murrayclair. Lost wax process is very accurate and inexpensive.

[Boris123]

Your most cost effective solution will be to find an "investment casting" foundry that has a rapid prototyping (3D Printer) these ones build the part in wax. Then this wax model is used to cast the part. It shouldn't be more than a few hundred bucks. You may then need a machine shop to skim the faces and diameter etc after casting.

Sent from my GT-I9305T using Tapatalk

Thank you so much for you suggestion murryclair and hjl4!

We're trying to make a tool-quality part out of hardened steel. I was under the impression that casting isn't normally used for that - for example, it's impossible to cast a drill bit that would drill iron. Am I wrong about that?

Thank you!

Boris

[Boris123]

A quick update: I found a way to change my algorithm so that it generates bicubic patches, wraps them into NURBS, and dumps into an IGES file. Each patch should replace many hundreds, if not thousands, of triangles - so I'm hoping the file will be reasonably sized. Rewriting and debugging the program should take me a few days. Still, it may not solve the main issue of the CNC machining being (likely) too expensive. I'll keep at it.

Thank you so much for all the information and ideas!

Boris

[...]
converting from stl to solid with automated process, will result same much little surfaces..

some type of programs can rebuild on a hard way.. programs that working with handheld laserscanners..

i think simplest is rebuilding one pattern and arraying around center..

also, even its a nurbs surface still it will be a large file..

[suker]

Hi, I come from China,this is my first time be here, of course I doing some work that related with cnc. I want to learn something and make friends in here.
Suker



Sent from my RM-939_apac_prc_200 using Tapatalk

[Boris123]

Hi, I come from China,this is my first time be here, of course I doing some work that related with cnc. I want to learn something and make friends in here.
Suker



Sent from my RM-939_apac_prc_200 using Tapatalk

Hello Suker! I'm a newbie here myself. I find people here are very helpful and friendly, but you have to listen to them (read) very carefully. I found people respond best when you ask specific, well-defined questions. I try to do as much research as I can before asking for help.

There is a separate thread specially dedicated for introductions, here:
http://www.cnczone.com/forums/genera...d-tell-us.html

Hope this helps, and good luck!

Boris

[murrayclair]

You can investment cast tool steels and other heat treatable alloys. Most investment casting foundries will cast you 4140 or 4340 ang you can heat treat as well as nitride.

Sent from my GT-I9305T using Tapatalk

[suker]

Thanks a lot,actually I have a machining shop, I am looking for some different idea about cnc machining.




Sent from my RM-939_apac_prc_200 using Tapatalk

[Boris123]

You can investment cast tool steels and other heat treatable alloys. Most investment casting foundries will cast you 4140 or 4340 ang you can heat treat as well as nitride.

Sent from my GT-I9305T using Tapatalk

I just learned about centrifugal casting. That sounds like it might work, except from what I've seen, this technology is not well suited for prototyping/small runs...

[victorofga]

boris youre very talented..




if you need more of these works, then I seriously thinking you need to invest into a mini mill..

have you looked on ebay? or alibaba marketplace..

reason I saying it, what I thinking from beginning, you have your own invetions.. sending out in shops you never know where it will landing..

with own equipment you don't need sending out blueprints..

sinter technology leaves a surface that might need to grind..

casting, unless its a low meltimgpoint stuff it will leave a very rounded edges, not ot mention it wont flow into chambers like you need details..

with an own machine you can mill a copper edm electrode, what actually used for toolmaking

for those tiny details in hardened steeltool,, you don't have too much option..

[murrayclair]

Centrifugal will not give you what you want. You need Investment casting or lost wax. They can 3D print a wax from your .stl file. Then they dip the wax in liquid ceramic. This is done several times to build a thick ceramic shell around the wax printed from your .stl. the ceramic is hardened and the wax melted out. Now molten metal is poured into the ceramic shell. You will get all of your sharp internals at the tolerance you need.

Sent from my GT-I9305T using Tapatalk

[suker]

Hi could you show me the 3d file,maybe I can help you.

Sent from my RM-939_apac_prc_200 using Tapatalk

[hjl4]

casting, unless its a low meltimgpoint stuff it will leave a very rounded edges, not ot mention it wont flow into chambers like you need details..

I disagree. Yes you will have a radius on the sharp edges, but within 0.05mm or better, because the mold is preheated before casting.
The ceramic has dual function as it forms the shape and insulates the casting so the liquid metal has time to fill all the little crevises.

Even an EDM will inevitably leave a small radius on the edges as the copper electrode wears down.