Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Precision CNC steel machining

  1. Feb 28, 2007 #1
    I am currently investigating ways to make a mold for hot-embossing. Normally this is done by for instance SU8 on a silicon wafer, or DRIE of a silicon wafer.

    But my design is deeper than a normal wafer, around 5 mm deep.
    Because the devices are actually fairly large (for microsystems) I would be happy having a resolution of only 25 micron (x, y, and z).

    I am wondering what the capability of CNC machines are. The structures aren't too complex, the smallest line required is 90 micron high and 175 micron in diameter, 5mm long.

    Is it possible to have this machined in steel?
  2. jcsd
  3. Feb 28, 2007 #2


    User Avatar
    Science Advisor

    I don't know of any tooling for a typical CNC 3 or 5 axis mills that can do that small of a feature. You may need to look into EDM but I doubt that would give you the surface finish you would need. All of our extremely small features relative term) are done with either EDM or laser. Typical machines I am used to have an accuracy to about .0005-.001" depending on the operator and the feature being machined. Grinding set ups can do much better than that, but can they actually make the features you need? I don't know on that without seeing what it is you are doing.

    Don't they normally do this with CVD?
  4. Mar 1, 2007 #3
    No, normally SU8 is applied with spincoating, or DRIE is used to etch the wafer. Wafers are only 0.5 mm thick. My problem is one of bith large and small sizes. The large features are easily machinable, around 3 mm deep and 6 mm wide. They are however much to deep for any of the micromachining processes normally used. (because wafers don't come in this thickness)

    The features are a set of lines that will create channels when embossed in polymer. The largest is 3by6mm (height-by-width), the smallest 180by90micron.

    EDM and 'laser' is more of a cutting method. So I do not think they would be the tool of choise for these ridges.

    I am confused what you mean with grinding.
  5. Mar 2, 2007 #4


    User Avatar
    Science Advisor

    I see. Grinding is another material removal process, but it is a bulk removal. It is still too large scale of what you are doing.

    This is definitely a specialized area. I'll keep my eyes open while looking around. Sorry I couldn't be of more help.
  6. Mar 5, 2007 #5
    It's an area in between classical engineering and micro-engineering based on IC processes. It is interesting for the field of MEMS which I am in.
    But indeed, at this point in time I do not think there is anything worthwile.
    Thanks for your response anyway.
  7. Nov 27, 2010 #6
    I don't know of any tooling for a typical CNC 3 or 5 axis mills that can do that small of a feature. You may need to look into EDM but I doubt that would give you the surface finish you would need.
    Last edited: Nov 27, 2010
  8. Nov 30, 2010 #7


    User Avatar
    Gold Member

    Microlution makes some 3 and 5 axis mills with 1 micron positioning accuracy: http://microlution-inc.com/. Not sure about the depth though.
  9. Nov 30, 2010 #8
    A decent 3-axis can do that if you only need 25 microns of resolution. The bridgeports in the machine shop at my uni have a position accuracy of 0.001in (25.4um). A good machine will have an accuracy and resolution of 0.0005in (12.5um). Now if you need to start doing circles or complex curves at that size you might start having some issues but strait cuts shouldn't be a problem.

    FYI, prepare to have a hard time finding and paying a lot for tooling.

    EDIT: I just realized that this thread is 3 years old. :grumpy:
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Similar Threads - Precision steel machining Date
How steel fails? Mar 5, 2018
Difference between uncertainty and precision? Sep 9, 2016
Building Precise Machine in Home Oct 24, 2011
Precise to accurate Feb 3, 2011