Skip to main content
1-Visitor
March 27, 2014
Question

Is a model good enough?

  • March 27, 2014
  • 21 replies
  • 9539 views

Posted this a few days ago and not sure as to whether or not it went through, so trying one more time. This also piggy backs a little on the "should scale be removed" thread.

    21 replies

    1-Visitor
    March 27, 2014
    Stefan,

    This does dovetail nicely into my initial Thread-bomb. It all depends on how you define your product?
    2-Explorer
    March 27, 2014
    Here are my thoughts - because you asked.



    I agree with your bottom line - "not sure if we're there yet" I would go
    one step further and say "we're not there yet." Yet being the operative
    term.



    The model used to communicate the geometry to the vendor is generally a STP
    file these days, right? IGS becoming less common / popular and DXF still a
    decent choice for 2D geometry (e.g. for laser cut sheets, etc)



    The PDF of the 0-60-100% detailed part is generally used these days for
    over-alls, critical-to-function or inspection dimensions, hole call outs,
    tolerances, material, finish, etc.



    While all this info can now fairly easily be included with the NATIVE 3D
    part, the "dumb solid" has yet to allow that capability. (to my knowledge -
    and if it is capable to put that info into a STP, it's likely my vendors
    will need some training to use it)



    My practice is, for several reasons, to not send out native files to a
    vendor. These reasons include client confidentiality, revision control,
    software access/capability of the vendor, file size, convenience, risk of
    accidental change, etc. One exception is a sheet metal shop I use begs me
    for the native solidworks files - they actually charge me less because it
    saves them engineering time. But then there is the real challenge I have
    with them when they make a change on the fly to make their job easier and
    then need to communicate that to me and I need to either recreate the change
    on my model, or take their modified parts and slip them back into my assy. =
    risk.



    But I envision a time when a new 3D dumb solid format comes along that can
    handle annotations much like a DXF supports dimensions and text. Then, we
    will be ready to submit our designs for quote without a 2D pdf.



    I look forward to that day.

    -Nate


    23-Emerald III
    March 27, 2014
    Maybe a couple months ago there was a discussion on ASME Y14.41 (MODEL BASED DEFINITION).

    6-Contributor
    March 27, 2014
    Stefan,

    The answer to your question "Is a model good enough?" depends on all the
    people internal and external
    who need to touch your company's drawing. Preference aside, you would
    need to implement the tools needed
    so that everyone in the chain has the solution to perform their job,
    design, manufacturing, inspection
    , etc. Model annotations allow the user to generate a "3D" drawing so the
    amount of effort is basically
    the same for engineering if you decide to product fully annotated models.
    Since your drawing is like you said a contract, you need to format the
    model's output to
    address all of legal attributes also. Model based definition has come a
    long way but each company is very
    different and has to address its individual needs.


    Best Regards,
    Jose



    15-Moonstone
    March 27, 2014
    Stefan,

    Is a model enough? Yes, see ASME Y14.41 or the corresponding ISO DIS 16792 both of which are entitled "Digital Product Definition Data Practices". It will lay out everything as far as what is needed to enforce the design intent of the model onto the manufacture of the part.

    I like how you put "Not necessary to manufacture a part, but necessary to ensure that a part was manufactured correctly." because that is one of my main sticking points to jumping on the bandwagon (though not my only one). If your supply chain can support inspection without having to reference a drawing (either you supply or they create from your supplied model) then you are a long way there. Until I can answer that question affirmatively, I will refuse to rule out good quality suppliers for a questionable savings by doing away with drawings completely. I fall in the "in-between" drawing level camp. I have had great success with critical to function (CTF) drawings where dimensions are only included if they are critical and therefore need to be inspected.

    As I was responding, Nate Rollins got in with an excellent response that I will just add to by mentioning the new STEP standard (AP242) from LOTAR International. LOTAR is the Long Term Archiving And Retrieval organization working on the problem of how to make files accessible fifty years from now without losing all the annotation information. The 3D CAD vendors are working on including it in future versions. It will NOT include the history tree directly but will carry over design intent based on annotations.



    Rob Reifsnyder
    Mechanical Design Engineer/ Producibility Engineer / Components Engineer / Pro/E SME / Pro/E Librarian
    [LM_Logo_Tag_RGB_NoR_r06]
    1-Visitor
    March 27, 2014

    A fully detailed drawing is a narrative from the designer or engineer to the inspector about what constitutes an acceptable part. Drawings are easily reproduced, easily annotated, and easily distributed.



    A CAD model, so far, has none of these features, except within its own ecosystem. Even then it can be unclear because of the many ways in which the software fails to record accurately what the designer or engineer required.

    1-Visitor
    March 27, 2014
    A well-made point. I would counter that a generally accepted and adopted model viewer and display convention would complete that narrative (See ASME Y14.41-2003) as well without the need for secondary files (.drw’s)
    13-Aquamarine
    March 27, 2014
    In response to what Robert said:
    "LOTAR is the Long Term Archiving And Retrieval organization working on the problem of how to make files accessible fifty years from now without losing all the annotation information."

    There is a lot more to this problem than the file type. The long term viability of the storage media is the real issue. The fragility of digital media is the real long term problem. If you need long term storage of your design. Then I think you should be creating fully detailed drawings. Not everyone needs this level.

    Remember punch cards? Back in the early 90's we had a lot of critical historical design data in card decks. At the time we only had one working card deck reader. We spent quite a number of hours getting that data read into our system at the time.

    A few years ago I needed to make some changes to a design that had been done in ProE in the early 90's. The files had not been brought forward in our system because at the time they had been archived on a ¼" tar tape. I located the tape, but no one had a working cartridge reader.

    Last year I was out in one of the Labs and took down the serial number and vendor for a glove box that was built in the 70's. I called up the vendor to see if they had any drawings. The guy looked in his computer system and said no, we don't have any drawing for that. I hung up. A few minutes later he called back and said, "guess what, one of our old guys suggested I look in the flat files, and we found the velum's, I make some blue prints and send them to you."

    The moral of the story, is if you're looking for real long term viability of your design. Paper is still the one of the top formats. Even newsprint which has a horrible life span compared to velum, will last over a hundred years easily. I have family photos I've scanned that are easily over a hundred years old.
    I think the gold standard for longevity is still Microfiche.

    History is littered with data formats that are no longer viable.
    VHS tape
    Beta
    Removable hard drive platters
    8" floppy disks
    5 ¼" floppy disks
    3 ½" floppy disks

    I predict CD's and DVD's will be next.

    David Haigh
    1-Visitor
    March 27, 2014

    It would - were one to exist. I think there is no chance that a good one will ever exist** as that would guarantee unambiguous interchangeability of models between CAD systems. There is no motive for CAD software makers to make interchange that easy.


    Some can use interchange, but the ones I've heard successful are the result of communications and documents outside the model that provide critical information the model does not contain. Tolerances, materials, datums, surface finish and lay. These are items that are not well represented in interchange models.


    '14.43 does not supplya description of a model viewer or sufficient display conventionsto be useful. Independent companies can't build software that complies with '14.43 and works interchangeably. The XKCD page on standards comes to mind.


    ** My lifetime limitation and all that. The future is a very long time.x




    In Reply to Bob Lohbauer:


    A well-made point. I would counter that a generally accepted and adopted model viewer and display convention would complete that narrative (See ASME Y14.41-2003) as well without the need for secondary files (.drw’s)
    12-Amethyst
    March 27, 2014
    Just to add a little bit more to the discussion ...

    One of the big problems I saw in a facility trying to use 14.41 was the ease of looking over a model compared to a drawing. Even with users proficient in Pro (engineering staff) having them query each feature individually to get dimensions, GD&T, notes, etc. was a nightmare. As far along as Pro has come since that experience (WF3) it still strikes me as a rather cumbersome process compared to just having everything right there on a drawing.

    In my personal experience there are a lot of people that have mentioned they just cannot check over a design as thoroughly on screen as they can with a paper copy right in front of them. Even when they pull up the drawing on screen more stuff in need of correction gets missed than with a printed-out drawing.

    Any time you need to have that information in the field, rather than in the office, you'll run into additional problems. Not just making sure a user is present that knows how to manipulate the software, but also just getting the information to the site, either due to file size or on occasion the remoteness not allowing for any kind of interconnectivity. (Think of stuff assembled in the field, like mining trucks and such.)