Tolerance, Bilateral vs Limit - Page 7
Close
Login to Your Account
Page 7 of 7 FirstFirst ... 567
Results 121 to 131 of 131
  1. #121
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Florida
    Posts
    4,061
    Post Thanks / Like
    Likes (Given)
    1611
    Likes (Received)
    1885

    Default

    OK, it seems what I am seeing/interpreting here is some people are using the model as the 'bible' to program and make their parts, and to me that isn't right, unless your customer specifically says " the model is bible- pull al dims and features from it +/-.xxx"

    I've never been handed a solid model and been told to program to it with a general tolerance. I have -

    1) been given both, understanding print is bible
    2) been given a solid and LDD (limited dimension drawing) and used the model for everything not dimensioned to a standard tolerance of +/-.xxx
    3) not given a solid model and program from a paper print

    As far as Barbter comment "You could pull the tool up with length/wear comp but that isn't right for the next time the prog runs..."

    Ummm... that is kind of what we are expected to do as machinist, isn't it? Comp tools/offsets to make the part right? For example, model or not, if a dimension is +/-.001", personally I would comp the tool +.002", run it, measure, then adjust comp to get it to size - maybe it ends up being +.0003", or -.0002"... I would not expect that offset to be exactly the same the next time I ran the program (cutter wear, tool holder runout, etc). I would expect the next guy to "dial it in" as I did.

    Also, for example, in Mastercam if you measure a tool very good (cmm, micro-view etc), you can have it add the correct amount to get a c'sink to the right size. IE- you need a .312" diameter (at top c'sink) for a fhcs, you click on the button and tell it you want a finish diameter of .312, and the tool has a .050" flat on the tip, it calculates what depth to go to achieve a .312" diameter. It works really well, but, I always put a note in the program to check the c'sink with the correct fhcs for size. I expect them to do that, not just say "the program went to this depth blah blah blah I thought it was right ".

  2. #122
    Join Date
    Oct 2007
    Country
    SPAIN
    Posts
    3,647
    Post Thanks / Like
    Likes (Given)
    2070
    Likes (Received)
    1348

    Default

    Quote Originally Posted by Mike1974 View Post

    As far as Barbter comment "You could pull the tool up with length/wear comp but that isn't right for the next time the prog runs..."

    Ummm... that is kind of what we are expected to do as machinist, isn't it?
    The point i was trying to make is....say you have a complex ish part - lots of features (fuel manifold or the like).
    There's a base/pocket bottom which is modelled on min size.
    Print calls +/-0.1mm but it's modelled on bottom limit.
    Your cutter that finishes this in the same operation finishes a lot of other features too which are modelled and 'grammed to (model) mid limit.

    So we're dealing with just the one feature, buried within a toolpath, within the tool call.
    We could manually edit and throw a G10 line in there to pull the tool up 0.1mm for this one feature, then a following G10 line at the start of next feature to drop it back down to match the tool length register in the control.
    But that aint right...

    Yes i know it's what we do, but the ball ache is to do it correctly - modify the CAD/Part file, post it out, verify again, prove out again etc etc...
    That's all i was trying to say

  3. Likes toolsteel, TeachMePlease liked this post
  4. #123
    Join Date
    Nov 2012
    Country
    UNITED STATES
    State/Province
    Wisconsin
    Posts
    2,108
    Post Thanks / Like
    Likes (Given)
    3458
    Likes (Received)
    2649

    Default

    I prefer limit tolerance. I dislike -,- 0r +,+ tolerances.

    I program from models quite a bit. I always check a few key dimensions. It drives me nuts when the model isnt made to mean dims. Call out whatever sort of tolerance you need on the print....but model the way it should be made as perfect. I always wonder why we pay someone to model parts....and then pay me to change those models to be able to make the parts. Seems we could save time and $ by just modeling it for ease of manufacturing.

  5. Likes mhajicek liked this post
  6. #124
    Join Date
    Nov 2012
    Country
    UNITED STATES
    State/Province
    Wisconsin
    Posts
    2,108
    Post Thanks / Like
    Likes (Given)
    3458
    Likes (Received)
    2649

    Default

    Quote Originally Posted by barbter View Post
    The point i was trying to make is....say you have a complex ish part - lots of features (fuel manifold or the like).
    There's a base/pocket bottom which is modelled on min size.
    Print calls +/-0.1mm but it's modelled on bottom limit.
    Your cutter that finishes this in the same operation finishes a lot of other features too which are modelled and 'grammed to (model) mid limit.

    So we're dealing with just the one feature, buried within a toolpath, within the tool call.
    We could manually edit and throw a G10 line in there to pull the tool up 0.1mm for this one feature, then a following G10 line at the start of next feature to drop it back down to match the tool length register in the control.
    But that aint right...

    Yes i know it's what we do, but the ball ache is to do it correctly - modify the CAD/Part file, post it out, verify again, prove out again etc etc...
    That's all i was trying to say
    I think when we are creating models, programs, fixtures etc. some ppl get off on doing things in a manner that is technically correct (often times there is more than one way that is correct) but not neccessarily the easiest.
    The goal from design to manufacture (my opinion) should be to make parts that serve the intended purpose adequately in the easiest possible way. Not to find obscure ways to throw a wrench in the gears so that we can look down our noses and say..."well...if you knew what you were doing this wouldnt have happened"
    This is one reason many ppl prefer programmers and designers who have come from the shop floor.....these ppl should be looking toward making the job of manufacturing easier....not a damn quiz.
    If as a programmer my programs are to the mean dimension every time my guys should be able to zero their wear offsets going from one job to the next and be confident they are within .001" or .002" .....that is how money is made. not telling them...."well, thats your job.....to run a part .....then decide to move X .0042" and Z-.0037.". It is....and we can make it easier or harder as programmers, designers etc. Thats just asinine when the tools to program to mean are right there in front of you. If the first one is bad.....you have some other issue....not my program.
    my .02

  7. Likes mhajicek, barbter liked this post
  8. #125
    Join Date
    Nov 2013
    Country
    UNITED STATES
    State/Province
    Utah
    Posts
    4,674
    Post Thanks / Like
    Likes (Given)
    1238
    Likes (Received)
    2618

    Default

    Quote Originally Posted by Mike1974 View Post
    OK, it seems what I am seeing/interpreting here is some people are using the model as the 'bible' to program and make their parts, and to me that isn't right, unless your customer specifically says " the model is bible- pull al dims and features from it +/-.xxx"
    But I'm not religious, so where does that leave me?

  9. #126
    Join Date
    Feb 2014
    Location
    Sunny South West Florida, USA
    Posts
    2,944
    Post Thanks / Like
    Likes (Given)
    10830
    Likes (Received)
    3323

    Default

    Quote Originally Posted by litlerob1 View Post
    But I'm not religious, so where does that leave me?
    Free to go fishing on Sunday mornings.

  10. Likes mountie liked this post
  11. #127
    Join Date
    Nov 2014
    Country
    UNITED STATES
    State/Province
    Florida
    Posts
    4,061
    Post Thanks / Like
    Likes (Given)
    1611
    Likes (Received)
    1885

    Default

    Quote Originally Posted by litlerob1 View Post
    But I'm not religious, so where does that leave me?
    It leaves you going to hell if you don't change your ways!
    Obviously you get to make the rules then, duh...

    J/K I don't believe any of that religious stuff.

  12. #128
    Join Date
    Sep 2015
    Country
    UNITED STATES
    State/Province
    Tennessee
    Posts
    535
    Post Thanks / Like
    Likes (Given)
    257
    Likes (Received)
    421

    Default

    Quote Originally Posted by JCByrd24 View Post
    I’m getting trolled right?? I never said it’s exclusive to size. I used voltage as an example. You insisted that size was not mentioned in the article. It was, a lot.

    ASME Y14.5 and common sense say the small limit should be before the large when shown on one line. You don’t say I’ll arrive to work between 7 and 6:45 do you?
    What the hell are you talking about, please IM me your business name so I don't accidently ask you to make something for us. You're comparing a variable (voltage) to a known controllable dimension?

    Is this how your shop works?

  13. #129
    Join Date
    Sep 2015
    Country
    UNITED STATES
    State/Province
    Tennessee
    Posts
    535
    Post Thanks / Like
    Likes (Given)
    257
    Likes (Received)
    421

    Default

    Quote Originally Posted by CarbideBob View Post
    And it is so wrong IMO on one sided.
    If you ran the old boards or the new CAD for me and tried this ....... Call Marvin here in Michigan.
    There are standards that go way back into paper.

    I do hate one sided tolerances as the question is always to hug the side or go for mid.
    If you have size control at 10-20% of the entire range (which I would hope all can) what is a better part? What should be the target?
    Let's say that you know you can do +/- .001 all day long and are very sure of that and have all the stupid SPC and CP bullstuff to back it up.
    The print is 3.000 +0.000/-0.010, so you have miles to play and ship good parts. Where do you run and why?
    Should the person making the model make this decision for me?
    Bob
    The person making the model has asked you to adhere, I would never ask something like that of you. I was a master tool and die before I became an engineer, I know the limits, I know the clearances, I know the fits. If I ask you to make something to the model, and I give you the tolerances I expect you to make it to my wishes. Anything else is unsatisfactory.

  14. #130
    Join Date
    Sep 2015
    Country
    UNITED STATES
    State/Province
    Tennessee
    Posts
    535
    Post Thanks / Like
    Likes (Given)
    257
    Likes (Received)
    421

    Default

    Quote Originally Posted by mhajicek View Post
    Modelling something at 4.9976" does not take one iota more computer resources than modelling it at 5.0000". The CAD software is internally calculating everything to umpteen decimal places already regardless.
    who tf is sending you 1 GB step files? or parasolid? I call bullshit.

    I could model Georgia's new new Mercedes Dome and keep it under 2GB compressed.

  15. #131
    Join Date
    Sep 2015
    Country
    UNITED STATES
    State/Province
    Tennessee
    Posts
    535
    Post Thanks / Like
    Likes (Given)
    257
    Likes (Received)
    421

    Default

    Quote Originally Posted by SeymourDumore View Post
    Teach, you are not a jerk, but you are actually wrong in this particular case.

    You see, when the job hits the floor, not only is it already processed, but an opsheet is created for the setup guys and the operators alike.
    If the incoming B/P called for a 4.000 -.001/-.003 dia shaft, when the opsheet is created it will have only a 3.998 +/-.001 on it.
    It will not have ANY mention of ANYTHING being 4.00 whatever, because it is completely irrelevant to anything we do on the floor.

    So, - in our shop - we aren't changing the definition of the word Nominal, rather redefining the actual Nominal dimension.
    IOW, your nominal 4.00 dia becomes 3.998, which is nominal for us.
    Then your shop flow need serious attention, as well as your documentation.

    And Teach is not wrong, your "op sheet" is destined for scrap. You have communication interference, you admit you have difference between two prints.

    Also, you don't understand what nominal means.


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
  •