Hardinge Conquest V 1000 (experiences) and Conquest H-51 and Elmira group? - Page 3
Close
Login to Your Account
Page 3 of 7 FirstFirst 12345 ... LastLast
Results 41 to 60 of 131
  1. #41
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    23,298
    Post Thanks / Like
    Likes (Given)
    3583
    Likes (Received)
    6519

    Default

    On a separate idea:

    You seem to have this whole belt thing about accuracy, while the encoder is on the spindle...

    I have a (well, 2 actually) Hardinge VMC's with Siemens controls. They are belt drive. One has a geared head. The encoders are on the spindles, the motor is belt drive.

    I can tap a class 3 hole, unload the tool and doo whatever else, and then reload that tool and go back in that hole to tap it deeper, and still the class3 GO gauge has a tight fit. This is on parts mounted in a fixture on a 4th axis, but I would not attempt that after I moved the 4th.


    If you expect such high res on a 4th, you had better be checking to see where the encoders on the units that you are looking at are. Are they using the motor? Or is there a wrap-around (linear?) scale inside the unit? (or on the back - in the coolant?)

    And even if it does have a scale, and you set it dead nuts, and then hit the brake - doo you think it will still settle in dead nuts? Or will the brake influence it a bit within the amount of lash?

    With the numbers that you are tossing around - I kant imagine trying to produce on any commodity 4th, and would have the best luck on a C. (IM/HO)


    ------------------

    Think Snow Eh!
    Ox
    Last edited by Ox; 11-19-2015 at 06:11 PM.

  2. Likes Jashley73 liked this post
  3. #42
    Join Date
    Sep 2014
    Country
    UNITED STATES
    State/Province
    California
    Posts
    1,493
    Post Thanks / Like
    Likes (Given)
    922
    Likes (Received)
    1056

    Default

    Eric,

    My two cents.

    KEEP IT SIMPLE.

    The tighter the parts, the more robust and repeatable the process needs to be.

    A two axis lathe and 4 axis mill seems to me to be the best solution for you and your company unless the tolerances are so tight between turned features and milled features that you need something along the lines of 5 sided with turning. Your budget would need to be about double of what seems comfortable to you.


    I have a couple customers who specialize in very tight work.

    One does 4 axis milled work where a 2mm cutter is in the cut for over an hour and must hold a few microns in Z axis. He uses his Okuma MB Japanese machines with scales and super nurbs for the not as close parts but the really tight stuff goes onto his Mitsui Seiki 5 axis machines. The Okuma's just can't keep tight enough in Z at a certain point. Neither can his 27k Brother.

    Another one does a finish turn on a small part (that I can hold about five of in my hand) that needs to have a finish cut holding better than 3 microns between a bore and OD. He is exception at small work and this task takes more than 10 times longer to set up each part than to cut. The machines are in a temp controlled room in the middle of his shop. He uses a mix of Wasino and Takisawa on these critical parts.

    Reading specs about what machines can do is very different than actually making parts to those specifications. The tighter the work, the better the process. If this is new production than the two alternative roads seems to me to be of using simpler machines (while taking great care to transfer and locate from one machine to the other) versus very complex machines that can handle the dynamics of cutting multiple surfaces. I would only choose the later course if I truly was an expert in 5 sided machining. By expert, I mean spent ten or so years machining and programming that style part.

    The simple path can be helped with the great workholding available today. I would look at 3R, Erowa and my favorite Schunk for critical parts that I want to clamp once and then transfer from one machining to another while keeping them in the same work holding.

    Just my opinion.

  4. Likes cameraman, Jashley73, Rstewart liked this post
  5. #43
    Join Date
    Mar 2011
    Location
    NY USA
    Posts
    803
    Post Thanks / Like
    Likes (Given)
    155
    Likes (Received)
    510

    Default

    Quote Originally Posted by 2outof3 View Post
    Eric,

    My two cents.

    KEEP IT SIMPLE.

    The tighter the parts, the more robust and repeatable the process needs to be.

    A two axis lathe and 4 axis mill seems to me to be the best solution for you and your company unless the tolerances are so tight between turned features and milled features that you need something along the lines of 5 sided with turning. Your budget would need to be about double of what seems comfortable to you.


    I have a couple customers who specialize in very tight work.

    One does 4 axis milled work where a 2mm cutter is in the cut for over an hour and must hold a few microns in Z axis. He uses his Okuma MB Japanese machines with scales and super nurbs for the not as close parts but the really tight stuff goes onto his Mitsui Seiki 5 axis machines. The Okuma's just can't keep tight enough in Z at a certain point. Neither can his 27k Brother.

    Another one does a finish turn on a small part (that I can hold about five of in my hand) that needs to have a finish cut holding better than 3 microns between a bore and OD. He is exception at small work and this task takes more than 10 times longer to set up each part than to cut. The machines are in a temp controlled room in the middle of his shop. He uses a mix of Wasino and Takisawa on these critical parts.

    Reading specs about what machines can do is very different than actually making parts to those specifications. The tighter the work, the better the process. If this is new production than the two alternative roads seems to me to be of using simpler machines (while taking great care to transfer and locate from one machine to the other) versus very complex machines that can handle the dynamics of cutting multiple surfaces. I would only choose the later course if I truly was an expert in 5 sided machining. By expert, I mean spent ten or so years machining and programming that style part.

    The simple path can be helped with the great workholding available today. I would look at 3R, Erowa and my favorite Schunk for critical parts that I want to clamp once and then transfer from one machining to another while keeping them in the same work holding.

    Just my opinion.

    Yikes, I bet checking those parts is a real PITA.

    I can say that I am very impressed with Hardinge turning. I own super precision lathes including a GT which is the damn energizer bunny of machine tools....millions upon millions of cycles on that machine.

    OP, If you really need to hold the precision you think you do, you need to up the budget, then double it. I'd start with a Hardinge T series, they seemed to be designed from the outset to hold crazy tolerances. They aren't cheap but if you need it, your parts aren't cheap either.

  6. Likes cameraman liked this post
  7. #44
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    23,298
    Post Thanks / Like
    Likes (Given)
    3583
    Likes (Received)
    6519

    Default

    FWIW - I'm not finding the "linear motor" style Hardinge 4th on their website.
    I wonder if they don't make them anymore?
    Maybe just too expensive?
    I know they had heat issues with the prototypes, but I thought they had that taken care of soon after...

    ???


    --------------------

    Think Snow Eh!
    Ox

  8. #45
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    23,298
    Post Thanks / Like
    Likes (Given)
    3583
    Likes (Received)
    6519

    Default

    For a zero backlash rotary axis for the mill, this is what Hardinge was offering: Page 16/17

    http://www.hardingeus.com/usr/pdf/collet/2372e_lr.pdf


    They had the first one at IMTS back in prox 2008, covered in copper cooling lines.
    After that _ they apparently solved the cooling issue.
    At that time they were a Fanuc only thing.

    Now I wonder if they still offer them?


    I have one of the ones listed on page 10.
    Built like a brick shit-house!


    -------------------------

    Think Snow Eh!
    Ox

  9. #46
    Join Date
    Jan 2013
    Location
    Louisville, KY
    Posts
    2,832
    Post Thanks / Like
    Likes (Given)
    6522
    Likes (Received)
    2362

    Default

    Just to be clear - What part tolerances do you actually NEED to maintain? In what material, and at what size?

    One one end of the spectrum, you have people like us that take an entry-level Mazak with only thermal comp, parked near the shipping doors in a non-climate controlled shop, where we can have a 40-degree temperature swing across a day, and quite easily hold +/- 10-microns on the diameter, better than 7-micron circularity, and better than 40 micron runout between diameters, on a particularly cantankerous, semi-hard, 35lbs, and totally asymetric steel forging, with trained-monkey operators. I.E - making very good use of what we have, and doing quite well at a seemingly difficult task.

    On the other end of the spectrum, is a Moore-Nanotech machine in a climate controlled room, inside another climate controlled room, on a thermally isolated and climate controlled concrete pad, with white-lab coat engineers running the the parts which quite easily fit in the palm of one's hand, with unlimited time a budget to boot.


    So...... Where do you fit in? We started this thread talking about budget-conscious quality machines, and have somehow drifted into 1-micron position repeatability... So I'm confused. What's the end-game here?

    Realistically, what kind of geometric tolerances do you actually need to maintain? In what material, and at what size?

  10. Likes ManualEd, dstryr liked this post
  11. #47
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    1,246
    Post Thanks / Like
    Likes (Given)
    1929
    Likes (Received)
    376

    Default

    Quote Originally Posted by 2outof3 View Post
    Eric,

    My two cents.

    KEEP IT SIMPLE.

    The tighter the parts, the more robust and repeatable the process needs to be.

    A two axis lathe and 4 axis mill seems to me to be the best solution for you and your company unless the tolerances are so tight between turned features and milled features that you need something along the lines of 5 sided with turning. Your budget would need to be about double of what seems comfortable to you.


    I have a couple customers who specialize in very tight work.

    One does 4 axis milled work where a 2mm cutter is in the cut for over an hour and must hold a few microns in Z axis. He uses his Okuma MB Japanese machines with scales and super nurbs for the not as close parts but the really tight stuff goes onto his Mitsui Seiki 5 axis machines. The Okuma's just can't keep tight enough in Z at a certain point. Neither can his 27k Brother.

    Another one does a finish turn on a small part (that I can hold about five of in my hand) that needs to have a finish cut holding better than 3 microns between a bore and OD. He is exception at small work and this task takes more than 10 times longer to set up each part than to cut. The machines are in a temp controlled room in the middle of his shop. He uses a mix of Wasino and Takisawa on these critical parts.

    Reading specs about what machines can do is very different than actually making parts to those specifications. The tighter the work, the better the process. If this is new production than the two alternative roads seems to me to be of using simpler machines (while taking great care to transfer and locate from one machine to the other) versus very complex machines that can handle the dynamics of cutting multiple surfaces. I would only choose the later course if I truly was an expert in 5 sided machining. By expert, I mean spent ten or so years machining and programming that style part.

    The simple path can be helped with the great workholding available today. I would look at 3R, Erowa and my favorite Schunk for critical parts that I want to clamp once and then transfer from one machining to another while keeping them in the same work holding.

    Just my opinion.
    we are definitely on the same page here... So I have designed everything to make my life as easy as possible. So the cameras themselves have a myriad of "set and forget" factory adjustments to zero everything and calibrate it... So that really leaves only one pair of tapers (male and female, 15 degree, 3 inches deep and 3.5 " diameter at widest end ID) So axial alignment and repeatability of "mating surfaces" I.e. take it apart and put it together again (rotationally indexed) should be pretty high of the order of +/- 2 micron axially or better. Every other component has mostly 6DOF in precision adjustments, and adjustments are made using high end auto collimators and digital sensors. So really a Conquest H-51 or better or worse only has to make a finishing cut in a few seconds that is "straight" on the taper. Admittedly that requires the coordinated movements of X and Z but I think its pretty doable especially with good surface finishes. The tapers themselves don't have to reference any other surface in a mechanically precise way, "machined in". The horizontal cuts on a 4" diameter drum just need to be "flat" and each flat made to within 20 arc seconds of each other set 60 degrees apart on a cylindrical surface for example... There is much more complex geometry than that but the complex stuff is of a much lower precision.

    So when I look at test sheets for cut and measured test components I do factor in a lot of wiggle room, error and "Signal" degradation. So for example if a "Super precision" Conquest H-51 claims 0.25 to 0.5 micron circularity... I know that if I have a H-51 on my own turf that even if I have superb climate control and expert "feel" for the machine and cutting conditions, and that I read the machine bed time stores at night and tuck it up in bed, and wait for a transit of Jupiter with the Moon I will still not be able to attain 0.25 to 0.5 micron circularity and 5 micron part accuracy... :-) BUT I know I am in with a fighting chance to get +/- 1.5 micron circularity with a high rejection rate. And with the nature of a taper and indexing I should be able to achieve +/- 2 micron XY axial "mating" of two afore mention cut tapers (given the nature of tapers)... I am probably going to have to cut a few hundred of these components to figure out the various "sweet spots" and tweak the design also.

    I think its pretty doable and I would bet a six pack, that sitting on a shelf in Elmira somewhere is a fixture or fitting that pretty much does what I already need it to do... Or require very minor alteration or adaptation. Or they may be able to manufacture the key component for me Or I just make it myself.

    But I will definitely check out the "I would look at 3R, Erowa and my favorite Schunk for critical parts that I want to clamp once and then transfer from one machining to another while keeping them in the same work holding."

    Thanks so much for that :-)

  12. #48
    Join Date
    Oct 2014
    Country
    CANADA
    State/Province
    British Columbia
    Posts
    1,043
    Post Thanks / Like
    Likes (Given)
    1076
    Likes (Received)
    586

    Default

    So what kind of labs are these cameras used in?

    And whats the budget for your measuring equipment?

  13. Likes Jashley73 liked this post
  14. #49
    Join Date
    Sep 2014
    Country
    UNITED STATES
    State/Province
    California
    Posts
    1,493
    Post Thanks / Like
    Likes (Given)
    922
    Likes (Received)
    1056

    Default

    Quote Originally Posted by cameraman View Post
    we are definitely on the same page here... So I have designed everything to make my life as easy as possible. So the cameras themselves have a myriad of "set and forget" factory adjustments to zero everything and calibrate it... So that really leaves only one pair of tapers (male and female, 15 degree, 3 inches deep and 3.5 " diameter at widest end ID) So axial alignment and repeatability of "mating surfaces" I.e. take it apart and put it together again (rotationally indexed) should be pretty high of the order of +/- 2 micron axially or better. Every other component has mostly 6DOF in precision adjustments, and adjustments are made using high end auto collimators and digital sensors. So really a Conquest H-51 or better or worse only has to make a finishing cut in a few seconds that is "straight" on the taper. Admittedly that requires the coordinated movements of X and Z but I think its pretty doable especially with good surface finishes. The tapers themselves don't have to reference any other surface in a mechanically precise way, "machined in". The horizontal cuts on a 4" diameter drum just need to be "flat" and each flat made to within 20 arc seconds of each other set 60 degrees apart on a cylindrical surface for example... There is much more complex geometry than that but the complex stuff is of a much lower precision.

    So when I look at test sheets for cut and measured test components I do factor in a lot of wiggle room, error and "Signal" degradation. So for example if a "Super precision" Conquest H-51 claims 0.25 to 0.5 micron circularity... I know that if I have a H-51 on my own turf that even if I have superb climate control and expert "feel" for the machine and cutting conditions, and that I read the machine bed time stores at night and tuck it up in bed, and wait for a transit of Jupiter with the Moon I will still not be able to attain 0.25 to 0.5 micron circularity and 5 micron part accuracy... :-) BUT I know I am in with a fighting chance to get +/- 1.5 micron circularity with a high rejection rate. And with the nature of a taper and indexing I should be able to achieve +/- 2 micron XY axial "mating" of two afore mention cut tapers (given the nature of tapers)... I am probably going to have to cut a few hundred of these components to figure out the various "sweet spots" and tweak the design also.

    I think its pretty doable and I would bet a six pack, that sitting on a shelf in Elmira somewhere is a fixture or fitting that pretty much does what I already need it to do... Or require very minor alteration or adaptation. Or they may be able to manufacture the key component for me Or I just make it myself.

    But I will definitely check out the "I would look at 3R, Erowa and my favorite Schunk for critical parts that I want to clamp once and then transfer from one machining to another while keeping them in the same work holding."

    Thanks so much for that :-)
    I am going to come off as a bit of a dick but what makes you think you can produce a process to hold these tolerances. Do you have the decade plus of direct machining experience to not just manage the system but to deal with the issues you will be facing?

    I don't know a dozen shops I could point you to that are capable of not just holding those kind of tolerances day in and day out but in building a robust checking system.

    I have been in a lot of shops. Like thousands.

    Not saying it can't be done but it is going to be an epic ride through hell.

  15. Likes cameraman, dstryr, MK Proto liked this post
  16. #50
    Join Date
    Jan 2013
    Location
    Louisville, KY
    Posts
    2,832
    Post Thanks / Like
    Likes (Given)
    6522
    Likes (Received)
    2362

    Default

    Quote Originally Posted by cameraman View Post
    So that really leaves only one pair of tapers (male and female, 15 degree, 3 inches deep and 3.5 " diameter at widest end ID)

    <...snip...>

    So really a Conquest H-51 or better or worse only has to make a finishing cut in a few seconds that is "straight" on the taper. Admittedly that requires the coordinated movements of X and Z but I think its pretty doable especially with good surface finishes. The tapers themselves don't have to reference any other surface in a mechanically precise way, "machined in".


    I will still not be able to attain 0.25 to 0.5 micron circularity and 5 micron part accuracy... :-) BUT I know I am in with a fighting chance to get +/- 1.5 micron circularity with a high rejection rate. And with the nature of a taper and indexing I should be able to achieve +/- 2 micron XY axial "mating" of two afore mention cut tapers (given the nature of tapers)... I am probably going to have to cut a few hundred of these components to figure out the various "sweet spots" and tweak the design also.
    I'm sure you already know this, but when cutting any non-straight forms and shooting for good surface finish, you can't rely on wiper-inserts, and you can't do it quickly... If you're serious about holding that <1 micron RA surface finish, you're going to have to look at alternative methods of part finishing - I'd suggest roller and diamond burnishing would be a good place to start. Maybe even hydrostatic pressure-polish burnishers if you need to maintain an accurate/predictable face-to-taper-to-diameter Relationship. Roller burnish tools are very expensive, I'm not sure about diamond-burnish tools, and hydrostatic systems will run you about $40,000 if the need dictates... Otherwise, your turning feed will be so slow, that you'll run the risk of stringer-chips marring the surface, and then it's bye-bye finish...

    Also, you're looking for +/- 2 micron axial repeatability on the 15* taper - I assume you mean 15* included angle? If you're serious about that +/- 2 microns, have you given any thought into the force used to mate the two tapers? It won't take much force at all to expand the female taper, and "drive" the male deeper than your 4-micron limit.

    And then, how are you going to inspect all this?

    Like 2outof3 was saying... There's a lot of other process variables that just the machine tool won't really take care of...

  17. #51
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    23,298
    Post Thanks / Like
    Likes (Given)
    3583
    Likes (Received)
    6519

    Default

    Jash:

    Where did he say that he was expecting one micron S/F?
    Or are you just using that as a default Ra for the size tol?
    (I just don't recall him mentioning a surface finish)

    Isn't one micron a 40 on the inch scale?
    That would seem rough for that tol eh?
    Or is it a 4? Or .4?
    I know my inch scale, and I know the conversion for the metric in decimal form, but not sure on the text form.
    (You doo know that I have trouble with the writen language right? )



    Cam:

    I have no idea how you would ever qualify size that closely on any taper?
    Doo you?
    30" shadow setting near the lathe, run a part, check taper, run another?
    But on a 4" part, and 30" shadow doesn't offer much magnification, so .... ???

    How much machining experience doo you have?
    Doo you understand the scale of what you are stating?
    Just b/c your CAD can draw it, and you maybe see it on a D scale print, doesn't really mean that it is producible.
    On the other side - you did mention X scrap rate...

    I don't think that you have answered the Q's about material to be used?
    When Hardinge makes these test cuts, it is out of solid brass.
    What material are you hoping to cut that big, and hold that small of a tol as well as circ and such when done?








    On the other hand, you didn't ask us if your part was producible in the first place.
    I like a challenge, but ....
    Maybe one of Hardinge Bros, grinder lines may be a better fit?






    --------------------------

    Think Snow Eh!
    Ox

  18. Likes cameraman liked this post
  19. #52
    Join Date
    Jan 2013
    Location
    Louisville, KY
    Posts
    2,832
    Post Thanks / Like
    Likes (Given)
    6522
    Likes (Received)
    2362

    Default

    Post 33 he makes reference to .2 micron RA finishes. Doesn't explicitly mention that's the tolerance he's after, but he's made reference to needing high precision + surface finish. I'm kinda reading between the lines some, but my guess is, he wants something in the 1-micron RA / 8-16-microinch range...

  20. Likes cameraman liked this post
  21. #53
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    23,298
    Post Thanks / Like
    Likes (Given)
    3583
    Likes (Received)
    6519

    Default

    Well surface finishes in that range should be default with tols that small.
    I don't see that even being an issue.


    ------------------------

    Think Snow Eh!
    Ox

  22. #54
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    4,748
    Post Thanks / Like
    Likes (Given)
    1067
    Likes (Received)
    3264

    Default

    Interesting thread. I'm curious about distortion from holding the parts (even with EDM-style holding), stress relief of the material used, and perhaps most important - how will the parts hold up to day-to-day use? If these camera parts are changed out in the wild (and not in a clean room), there's almost certain to be dirt on the surfaces, and eventually edges will be dinged.

    How's the real world robustness of these components?

  23. Likes cameraman liked this post
  24. #55
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    1,246
    Post Thanks / Like
    Likes (Given)
    1929
    Likes (Received)
    376

    Default

    Quote Originally Posted by 2outof3 View Post
    I am going to come off as a bit of a dick but what makes you think you can produce a process to hold these tolerances. Do you have the decade plus of direct machining experience to not just manage the system but to deal with the issues you will be facing?

    I don't know a dozen shops I could point you to that are capable of not just holding those kind of tolerances day in and day out but in building a robust checking system.

    I have been in a lot of shops. Like thousands.

    Not saying it can't be done but it is going to be an epic ride through hell.

    He, he...

    Or shall I say mwahhahah...;-)


    Not to be a "Dick" back at you... But I absolutely live for stuff like that... I have PH.D. from Cambridge university and almost 20 international patents and many papers published over the years in different fields and have headed up many very challenging research projects in very new fields and areas... So to do these things you have to be someone that likes taking risks and doing things that have never been done before and push the frontiers of human knowledge and capability to a truly world class standard for things that DON"T exist in books or are taught anywhere... I.e. you have to be very independent, bloody minded, have a ton drive, and ton of insight but also be very tenacious/never give up and really push, push , push and push. I'm one of these weirdos that is very good at putting theory into practice and generally for anything that requires a mechanical feel or very high standards of finish and precision I have a real knack for. Normally I do things more on the nano scale so I have a good feel for things that are on the micro scale.

    All my life people have told me.. "Ohh you can't do that...!!" or this or whatever... OR "That's impossible" or "it can't be done Blah blah blah ..." And every time I have shown them to be complete wrong... :-) [Not that it's my motivation or intention to prove anyone wrong but it just works out that way; I actually have a very good track record for making very difficult and new things work. I have absolutely no doubt that we will become world experts on exactly what it is we do... But It will take about ten years to start to hone and develop our own secret sauce and associated techniques... And I have no doubt that over time we will develop in house capabilities that will be of the standard and quality that you describe... In our own way.

    As I said it's embarrassing how few of these systems we have to build initially and make good $. So it is going to be very much a cottage industry for a long while so we don't have to think in terms of gigantic production quality/automation issues as everything will very hand build in spite of CAD/CAM and CNC machines. Just to put this into perspective the precision parts we are talking about in this thread only account for not even 2% of the whole system and parts assemblies and systems that we need to make. The rest is much more generic and straight forward.


    2outof3 writes "I don't know a dozen shops I could point you to that are capable of not just holding those kind of tolerances day in and day out but in building a robust checking system. "

    This is what I love about being able to be hands about the design and initial prototype machining work and QC and testing... So there are several "dark" arts at play here that have nothing to do with the camera components directly themselves. So a lot of design work goes into making various reference masters and assembly jigs... So much more creative and interesting engineering goes into making the various assembly jigs to align everything than the actual cameras and lens components themselves. So I "sneak" into the design of the components a lot of special reference surface and hidden features that enable exactly the kind of tests you are talking about... Also there are some little known methods and techniques that have been developed over about 100 years (especially in the fields of photogrammetry/photogrammetric engineering and survey camera design) that enable very sophisticated procedures for calibration and checking/testing of parts to whole systems.

    Overall I am not completely stupid... so If I am having trouble with my 2% of high precision critical components then we go to grinding and find specialists in my area... But having those elements in house makes it much easier to follow the chain , where as components made "out of house" can be more time consuming as much more time is required to check the integrity of externally manufactured components... (I think that's the kind of scenario you are getting at) So that is one good argument for developing in house capability.).

    The other thing I would add is that I tend to initially over engineer things and exceed the requirement quite substantially and then later on see "Artfully" what elements can be downgraded while preserving the integrity of other more critical aspects of a system. So for example even if I can't get my fittings to work at the mechanical precision and repeatability I would like I still incorporate a belt and braces approach/redundant systems. So The lens and lens cone assemblies contain projecting laser fiducials that register themselves to the large sensors... So via software the internal orientation of the system is known or findable (a sort of "In Camera" digital calibration)... SO even if things drift we have a plethora of techniques to geometrically rein everything in "Real time". But through judicious use of high precision components I can reduce some of that complexity for "compensatory" dynamic (on camera) systems... Interestingly a lot of the cameras designs share a lot with the design of machine tools, for thermal compensation, symmetric design, linear encoders (temperature sensors) and so on. Even if these cameras were built really badly we have one all mighty transformation matrix and associated mathematical techniques at final calibration that makes it all come right (mathematically) as these are designed for 3d measurement, and even if things move about there are some crafty in0field calibration techniques that can be employed.

    To answer your question I have been working with camera engineering and 3d imaging metrology problems for over 25 years and have had the privilege of working side by side cheek by jowel with some really excellent machinists... This is not my first "Rodeo". So worst comes to worst I have access to some very serious engineering outfits in the space and nuclear industries in the U.K. and the U.S. so I am not worried... To be honest I don't think we will need to call on their services, but we have that as a final fall back position if we get into real trouble.

    No matter what we make it so... That's what we do.

    2outof3 writes [B]"I am going to come off as a bit of a dick but what makes you think you can produce a process to hold these tolerances. Do you have the decade plus of direct machining experience to not just manage the system but to deal with the issues you will be facing?"[/B]

    I LOVE IT when people say something like that as I think "YESSSSS!!!" as that tells me I am on the right track... 100% (makes me pleased and excited)... That in some ways is the "Oxygen" that drives what we do.

    Thanks for your comments and made me realize that in house capability and development of that expertise what we specifically do is indeed the way to go... So that will take a decade or more but in the mean time we will roll out ground breaking systems as the cameras are only "Eyes" for the software we have been developing for over a decade. The software is where its really at :-)

    Cheers,

    Eric

  25. #56
    Join Date
    Aug 2002
    Location
    West Unity, Ohio
    Posts
    23,298
    Post Thanks / Like
    Likes (Given)
    3583
    Likes (Received)
    6519

    Default

    Well, in that case, I would doo everything possible to keep the real tricky stuff in-house, if for no other reason than for anyone possibly leaking info.

    Sounds good and good luck!
    We were just to see if you really knew where you were headed.
    Most anyone else wouldn't be suited up for such a battle.
    You obviously are expecting an "ends" coming to justify the means.


    But still - for the record, IM/HO - you have the ability of C and A axis turned around.


    -------------------

    Think Snow Eh!
    Ox

  26. #57
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    1,246
    Post Thanks / Like
    Likes (Given)
    1929
    Likes (Received)
    376

    Default

    Quote Originally Posted by Milland View Post
    Interesting thread. I'm curious about distortion from holding the parts (even with EDM-style holding), stress relief of the material used, and perhaps most important - how will the parts hold up to day-to-day use? If these camera parts are changed out in the wild (and not in a clean room), there's almost certain to be dirt on the surfaces, and eventually edges will be dinged.

    How's the real world robustness of these components?

    I think you hit the nail on the head... My first degree was in materials science,

    milland writes " stress relief of the material used, and perhaps most important - how will the parts hold up to day-to-day use"

    For me that will be interesting and fun and I get to do some real "science" as a lot of that is not in books or papers...

    The cameras we are building are probably not like any camera you may have seen before so a lot of the critical components are very "chunky" "Heavy" and robust.

    Milland writes "If these camera parts are changed out in the wild (and not in a clean room), there's almost certain to be dirt on the surfaces, and eventually edges will be dinged."

    So generally these are not lenses and cameras in the traditional sense... So we actually make the removal of the lenses from the sensor kind of a pain in the arse to dissuade people from constantly switching out lenses... The lenses and sensors and everything in between are perfectly matched for the 3d in-field imaging and surveying... BUT having said that lense "cones" can be swapped out and we plan to have special conic wipers (much like those used in the machine tool industry) to wipe out large bore tapers and also design in "dirt traps" for wiping procedures in the various edges and grooves in the various components. Again there are laser fiducials in the camera to track drift of alignment of the lens to sensor orientation and position and there are also in-field calibration techniques that can take up any "field induced" wonkiness... Finally every camera (much like a machine tool) is issued with a calibration certificate so it is expected that after a year of continuous use that the cameras be Fedex'd back to us to be cleaned/serviced and recalibrated and issued with a new certificate. There are a lot of tools that come with the camera to attach and detach various elements so that its not so easy for someone that is "Kack handed" to mash these things too badly. So even in a "battle field" scenario or on a remote archeological site good results can be attained even if these systems are eating a lot of dirt and dust. A lot of things can be retroactively calibrated digitally many months or years after field recording.

    Really good points/questions...

    That's what I love about PM forum is that it is such an excellent "mine" of information and "tribal knowledge" as 2outof3 would say, that is not in books... And that's why it's good sometimes to throw things out here to give it a really good kicking... That's what it all about, get it right hook or by crook and I listen to everybody and everything.

  27. #58
    Join Date
    Jul 2006
    Location
    Hillsboro, New Hampshire
    Posts
    4,748
    Post Thanks / Like
    Likes (Given)
    1067
    Likes (Received)
    3264

    Default

    Like I said - interesting thread. I've been involved with space sciences for decades, including much work on the fabrication end of UV and optical cameras; X-ray spectrometers and detectors, and the support equipment to fab them. I've almost never had to work to the tolerances you're describing, but it's true that these were mostly "one shot" (launch and forget) devices which had a finite life plan and no human interaction except via programming.

    The way you're describing the elements you're designing I'd have to agree with the belt and suspenders comments, but it sounds like you've added glue and staples to the mix, and maybe some paperclips. Technology is heading so firmly to "analyze and correct computationally", and you're incorporating much of that, so why the emphasis on mechanical values?

    Not that I object that much - as a mech-tech guy, a challenge is always good...

  28. #59
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    1,246
    Post Thanks / Like
    Likes (Given)
    1929
    Likes (Received)
    376

    Default

    Quote Originally Posted by Ox View Post
    Jash:

    Where did he say that he was expecting one micron S/F?
    Or are you just using that as a default Ra for the size tol?
    (I just don't recall him mentioning a surface finish)

    Isn't one micron a 40 on the inch scale?
    That would seem rough for that tol eh?
    Or is it a 4? Or .4?
    I know my inch scale, and I know the conversion for the metric in decimal form, but not sure on the text form.
    (You doo know that I have trouble with the writen language right? )



    Cam:

    I have no idea how you would ever qualify size that closely on any taper?
    Doo you?
    30" shadow setting near the lathe, run a part, check taper, run another?
    But on a 4" part, and 30" shadow doesn't offer much magnification, so .... ???

    How much machining experience doo you have?
    Doo you understand the scale of what you are stating?
    Just b/c your CAD can draw it, and you maybe see it on a D scale print, doesn't really mean that it is producible.
    On the other side - you did mention X scrap rate...

    I don't think that you have answered the Q's about material to be used?
    When Hardinge makes these test cuts, it is out of solid brass.
    What material are you hoping to cut that big, and hold that small of a tol as well as circ and such when done?








    On the other hand, you didn't ask us if your part was producible in the first place.
    I like a challenge, but ....
    Maybe one of Hardinge Bros, grinder lines may be a better fit?






    --------------------------

    Think Snow Eh!
    Ox
    Overall I think it's pretty doable... I have many plan B's and fall back positions and different ways to skin this cat... It's really more about how one goes about things rather than what per se... I.e. the overall approach and attitude to solving the problem or problems...

    It will be interesting to explore the "Physics" of short tapers. If you think about in the machine tool industry you have components such as a Big Cat 40 tool holder that claims 1 micron repeatability with TWO contacting surfaces (it's really three surfaces if you think about it)... I only need two contacting surfaces that have a high repeatability of +/- 2 micron (or in that neighborhood) in the XY plane not Z. A lot of good tool holders can repeat in their spindles to about +/- 2 micron every day and in environment where there is a lot of coolant, dust, particles and grease... I think that's why the Elimira/Hardinge group are going to be especially useful in giving me a really good "Kick" in the right direction as I consider them the world experts on the dynamics of precision short tapers... So thanks to this thread I have an initial "Foot in the door" with those guys to solve this particular slightly tricky problem... AND they can advise as to what the best machine tools and systems or balance between in house versus out sourced capability I should be aiming for.

    Definitely as I progress and make a million mistakes along the way I will certainly share some of the results and findings I/we make. So personally I am really interested in the true Y axis capability of various turning centers to meet the initial requirement over 4th axis on "Mill" or more manual/careful static set ups. I would predict that we split the lines with the cameras (a bit like) Hardinge where we have HP (High precision systems) or SP "Super precision" or we have to call it something else due to trademarks. So for Sp or UP (ultra precision) those associated components may have to be finished using various grinding procedure instead (hence different designs and material). We charge differently and accordingly. Right now grinding is expensive and VERY expensive (and difficult) to implement in house BUT in the next ten years I would really like to see grinding take place on our own shop floor one day... Baby steps first and to shake out where the real technical problems lie. That's why hard turning or T series Hardinge turning centers are also of interest to me and we see how this can all be made to fit within the over scheme and business and technical objectives... I don't think that any of this is beyond the wit of man or impossible but good to get experts involved so that saves a lot of messing about and wasted time for things that don't or won't work. That's the other thing I like about Hardinge is that they are in the U.S. and fairly accessible. They are an old established US company under new and very dynamic and broad scale management and thinking now. So it's easy for me to hop on a plane and spend a few days with these guys. By contrast a lot of folks extol the virtues of Japanese machine tool builders and their knowledge and expertise and associated know how, but frankly its not so easy for me to go to Nakamura etc. and say "can you guys help me with this???"... Even though I have Japanese patents and so on it's not a realistic proposition unless I want to completely destroy the budget! :-) I am very happy to run with U.S. based systems and "Know how" and potentially I think or am hoping that they will be a good long term fit and partner as such. We will find out ;-)

    Cheers,

    Eric

  29. #60
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Colorado
    Posts
    1,246
    Post Thanks / Like
    Likes (Given)
    1929
    Likes (Received)
    376

    Default

    Quote Originally Posted by Milland View Post
    Like I said - interesting thread. I've been involved with space sciences for decades, including much work on the fabrication end of UV and optical cameras; X-ray spectrometers and detectors, and the support equipment to fab them. I've almost never had to work to the tolerances you're describing, but it's true that these were mostly "one shot" (launch and forget) devices which had a finite life plan and no human interaction except via programming.

    The way you're describing the elements you're designing I'd have to agree with the belt and suspenders comments, but it sounds like you've added glue and staples to the mix, and maybe some paperclips. Technology is heading so firmly to "analyze and correct computationally", and you're incorporating much of that, so why the emphasis on mechanical values?

    Not that I object that much - as a mech-tech guy, a challenge is always good...

    You read my mind I almost wrote, "with the belt and suspenders comments, but it sounds like you've added glue and staples to the mix, and maybe some paperclips" ... That's very true... So a lot of what I am doing is based on traditional machine vision and photogrammetric technology both terrestrial and airborne and "space borne"... but pushing things further on the terrestrial front. So a lot of this comes out of a new type of VR model called a coherently stereo textured model (CSTM) (that I have IP for) that are able to render extremely efficiently (400 x more efficient) than standard VR models, but nothing is lost from the original stereograms and they are completely spatially accurate. That way the user doesn't have to drown in millions/billions of 3d point cloud data points. I originally designed the systems to record historic building and archaeological and historic sites to sub millimeter accuracy over VERY large areas across a site very efficiently ... Much more efficiently than laser scanning for example. The models and immersive VR real time renderings are completely life like and do not look "computery" at all and allow one to traverse immense data sets very efficiently. Very usefully conservation tool, and we built a lot of 3d annotation tools into the software. I was generally frustrated at how slow and clunky the standard work flow is for high end field stereo photogrammetric recording was/is and realized that if we built a few more precision and mechantronic elements into the systems that the whole work flow can be very much faster and simpler in the field and back in the office. So over the past 20 years I have reverse engineered a lot of old school surveying cameras from the likes of Wild, and Zeiss from the film era and have been trying to bring the best of "Old school" camera engineering into the digital age. So to the user there is much less of a supplemental surveying burden with control points and less of a burden in the office to be able to perform tricky least squares "bundle" adjustments for very large systems of equations... (automated procedures are too "permissive" and get things wrong and more hands on systems require a lot of mathematical "Feel" to get right) So by physically building in field tighter tolerances for known position and orientation data (internal and external) that gives the user an hell of lot more to run with much more easily right off the bat.[But if you are very good at photgrammetric math and associated software this should allow engineering photogrammetrists to achieve far higher precision and accuracies than before.] Honestly if these cameras existed off the shelf we would buy them but they don't so we have to build them... We are also exploring sort of downgraded "pro sumer" or more general commercial lower res lower precision systems that we could make in larger numbers eventually ... One step at a time... They key difference between your regular photogrammetric camera and what we aim to roll out is that our cameras are "focusable" rather than having one fixed principal distance or range (or hyperfocal distance), so that means with one or two cameras you can record in astounding detail in 3D anything from the size of a bug to a mountain with one instrument. That has real advantages over conventional laser scanning where many different expensive instruments would have to be used to accommodate that range scale and detail. [Good value].

    BTW 20 years ago I got spend some time at JPL and met Charles Illachi (sp) and hang out with the synthetic aperture Radar group... At the time I was with the Getty and we were looking at ground penetrating radar to help explore the possibility of the discovery of new Mayan sites.. I,e, penetrate the tree canopy and some ground surface features. They flew some useful passes in air craft low level before their systems became space borne.

    Sounds like you have built/engineered some really cool stuff... It must be a nice feeling that you have things that you have your name on that are in "space" exploring the universe... Pretty damn cool.. But with space based stuff you have extremes of temperature simultaneously (shadow versus direct sunlight) to contend with and constant changes... As well as mysterious vibrations that can start to creep in (such as the Hubble space telescope) THAT is really tricky, makes what I am doing like child's play by comparison to nail that all down.

  30. Likes TeachMePlease liked this post

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •