Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Fluid jet for polishing and generating via ductile grinding

  • Please log in to reply
13 replies to this topic

#1 BGRE

BGRE

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1538
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 13 January 2018 - 12:56 AM

Anyone considered using fluid jet entrained abrasives for curve generating and/or polishing?
Thesis on subject:

https://repository.t...ection=research

Other:
http://www.aspe.net/...ICQL/FAHNLE.PDF

Nothing particularly exotic is required and fluid pressures are relatively low (~5 bar).
Standard abrasives (silicon carbide, alumina etc) and polishing compounds (Cerium oxide) are used.
  • brave_ulysses likes this

#2 gregj888

gregj888

    Surveyor 1

  • -----
  • Posts: 1728
  • Joined: 26 Mar 2006
  • Loc: Oregon

Posted 13 January 2018 - 01:01 AM

Can only imagine how you would control it over a large surface...

 

Look up some of the stress lap and ion polishing information.  Think you would have to do something similar.



#3 BGRE

BGRE

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1538
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 13 January 2018 - 01:13 AM

Yes CNC control is required and one needs to take into account the erosion profile of the nozzke,
However its cheaper than stressed lap and ion beam polishing or PACE (plasma assisted chemical etching)
The thesis shows one setup using a modified hobby lathe. It eliminates subsurface damage caused by conventional grinding.

#4 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 8426
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 13 January 2018 - 02:21 AM

It eliminates subsurface damage caused by conventional grinding.

 

So does the use of high psi load while grinding, at least to a large degree.


  • ed_turco and Augustus like this

#5 steveastrouk

steveastrouk

    Viking 1

  • -----
  • Vendors
  • Posts: 823
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 13 January 2018 - 11:08 AM

Nothing particularly exotic is required and fluid pressures are relatively low (~5 bar).
Standard abrasives (silicon carbide, alumina etc) and polishing compounds (Cerium oxide) are used.

Apart from CNC to drive the process



#6 BGRE

BGRE

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1538
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 13 January 2018 - 02:26 PM

It has been used for figuring Fabry Perot etalon plates to 1/100 wave or better.
The CNC mechanical complexity is of the same order of that used to print plastic parts.

#7 mark cowan

mark cowan

    Vendor (Veritas Optics)

  • *****
  • Vendors
  • Posts: 8426
  • Joined: 03 Jun 2005
  • Loc: salem, OR

Posted 13 January 2018 - 04:15 PM

I don't doubt it works.  Water jet cutting certainly works.  But this would be at lower pressure.



#8 BGRE

BGRE

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1538
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 13 January 2018 - 04:28 PM

Yes 5 bar or so instead of 3000 bar or more for waterjet cutting.

#9 steveastrouk

steveastrouk

    Viking 1

  • -----
  • Vendors
  • Posts: 823
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 14 January 2018 - 08:36 PM

The CNC mechanical complexity is of the same order of that used to print plastic parts.

And the coding ?



#10 BGRE

BGRE

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1538
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 14 January 2018 - 08:47 PM

The coding will likely have the added complication of variable dwell time at each point.
Calculating the dwell time time from an interferometric error map should be straightforward once the tool is calibrated.
Unless tool action can be tuned off (eg by lifting the tool) then uniform removal over the entire surface will be required to allow the tool to work on isolated error "islands".
Tool calibration should be tracked continuously by using the difference between before and after error maps for each run. Combined with a model for the tool operation and the dwell time map there should be sufficient data to track aany chanes in tool parameters.

#11 steveastrouk

steveastrouk

    Viking 1

  • -----
  • Vendors
  • Posts: 823
  • Joined: 01 Aug 2013
  • Loc: State College, Pa.

Posted 14 January 2018 - 09:34 PM

Its a fascinating approach, and quite amenable to amateur scale optical engineering, once the software has been developed, but who will bell this particular cat ?



#12 mconnelley

mconnelley

    Viking 1

  • -----
  • Posts: 620
  • Joined: 14 Mar 2006
  • Loc: Hilo, HI

Posted 23 January 2018 - 04:27 AM

Hello:

  

  I've been thinking of how to program an algorithm to determine the optimal path for a fluid jet given a particular hit map, and didn't get that far.  But then I realized that what you really want to do is to modulate the erosion of the surface, and there are a few ways to do this.

 

   One is by dwell time.  But rather than some complicated path, you could have the fluid nozzle do a simple raster scan and make the speed of the nozzle inversely proportional to the height of the error on the mirror (i.e. higher the error, slower the speed).  This should then be a fairly simple programming task for the CNC machine.  

 

   Another way to modulate erosion is by controlling fluid pressure by controlling the pump.  So the fluid nozzle can just do a simple raster scan over the mirror surface at constant speed, and you control the pump speed to be proportional to the height of the error on the mirror surface.  If you have a big pump, it may not be able to respond very quickly to commands to change speed, so a slow raster scan would minimize a lag between the nozzle getting to a particular spot and the pump getting to the right pressure.  

 

  I just came up with these two ideas, and haven't tried them yet.  I'm not sure which one I'd prefer to try first.  

Cheers

Mike



#13 BGRE

BGRE

    Surveyor 1

  • *****
  • topic starter
  • Posts: 1538
  • Joined: 21 Mar 2016
  • Loc: New Zealand

Posted 23 January 2018 - 05:53 AM

I'd try the 1st option first as its simpler mechanically. 

 

With The Jules Verne style nozzle you can also modulate the standoff distance.



#14 555aaa

555aaa

    Viking 1

  • *****
  • Posts: 833
  • Joined: 09 Aug 2016

Posted 23 January 2018 - 09:05 AM

Sounds like a good application for making Schmidt corrector plates.


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics