It’s Time to Get Over IGES

About five seconds into your average conversation about CAD translation, someone will invariably mention one of the venerable and long-held neutral formats: STEP or IGES. Neutral CAD formats are the great equalizers of CAD interchange; the old stand-bys of an increasingly multi-CAD world. They provide the bridge to migrate geometry from one proprietary CAD system to another. While direct translation or newer direct modeling techniques now often preempt the use of neutral formats, chances are high you'll need to rely on them at one time or another. But a word of caution: while STEP and IGES are often grouped together in the same breath, they are certainly not equivalent. In fact, if formats were horses, it's probably time to send the IGES pony to the glue factory.

scotchmagnetictape

But first, a little about the two formats.

STEP

The Standard for the Exchange of Product model data (STEP) enables the exchange of data between all manner of CAD, CAM, CAE, PDM, and other systems across a variety of industries and applications. STEP is an international standard, ISO 10303 to be exact, and like any other ISO standard, is certainly not free. As you can probably imagine, the standard itself is a complex and labyrinthine bureaucrat's dream, and likely a cure for insomnia. But worry not, we'll give you the short, short version. STEP comes in a variety of flavors designed for specific purposes called Application Protocols (APs), and continues to be extended with new APs. The most relevant APs for CAD are:

  • AP203 Configuration Controlled Design: Intended for the Aerospace industry, AP203 focuses on exchange data for 3D solid geometry of both individual parts as well as assembly structure.
  • AP214 Core Data for Automotive Mechanical Design Processes: Intended for the automotive industry, AP214 is a superset of AP203, additionally providing interchange for colors, layers, Geometric Dimensioning and Tolerancing (GD&T), and certain forms of design intent on top of existing AP203 capabilities.
  • AP203ed2 Configuration Controlled Design (2nd Edition): An update to the original AP203, added much of the capability of AP214 including parametric information and GD&T along other meta content, while also improving interoperability with other AP's like AP214. Be aware however, that many CAD systems still use the older implementation of AP203. Your mileage may vary.
  • AP242 Managed Model Based 3D Engineering: The holy grail of standards, AP242 has been in development for precisely forever and due for publication any day now - but promises to unify the various flavors of AP203 and AP214 into a succinct whole.

IGES

The Initial Graphics Exchange Specification (IGES) started as a US Air Force project specifically for exchange of CAD geometry, which culminated in an ASCII-based exchange format adopted by the American National Standards Institute (ANSI). An aged, monolithic standard originally designed for wireframe and surface data, IGES underwent several revisions to extend its capability. Those revisions finally brought the standard into the realm of solid modeling. The chief advantage of IGES is that it is unencumbered by licensing, i.e. it's cheap as free. Which is probably why it's still everywhere after all these years. But all meaningful development of IGES stopped in 1996 at version 5.3. For perspective, that year Netscape Navigator 3.0 had 80% of the browser market, Windows 95 was having trouble convincing most of us to leave NT 3.X or MS-DOS 6.22, the first Blackberry was still 3 years away, and Phil Collins left Genesis to focus on his solo career. Yeah, that old.

Which brings us to the larger point. It's likely time to leave IGES behind for good. Especially in light of available direct translation, direct modeling, and the continued evolution of STEP, there's little excuse for holding on to IGES. After all:

  • IGES is older than dirt. Regardless of how good it might have been in 1996, software has changed quite a bit since and will continue to change with new capabilities and requirements. For crying out loud, the IGES logo has tape reels in it. TAPE.
  • IGES doesn't support product structure. No assembly models except as a monolithic import. Decimating product structure is generally a bad idea. How long have components been around?
  • IGES is abandoned. No one's going to improve it. Ever.
  • IGES solids capability is weak. Solids came late to IGES development.  Consequently, translation of solids is more prone to geometric inconsistencies, resulting in unsewn/untrimmed surfaces. While IGES is still effective in wireframe translation, who uses wireframe anymore?

Despite these shortcomings, just about every modern CAD system still carries IGES translators, the result of 3 decades of inertia.  So what's the harm in continuing to use IGES? IGES just can't keep up with regard to modern solid geometry, and will invariably cause geometry errors that will cost you real money down the line. Use a modern toolkit instead, be it the various flavors of STEP, refined direct translators, or direct modeling capabilities that can operate on foreign geometry. The next time you talk about neutral CAD formats, leave IGES out of it. Think about it, won't you?

 


buyers-guide-for-pdmThe Buyer's Guide for PDM 

Every class of technology undergoes an era of innovation and disruption. For PDM systems, we’re in such an era today. Lifecycle Insights' Principal Analyst Chad Jackson put together the perfect buyer's guide to help you weigh your options.

guide to CAD file management

 

 

 

  • John Chawner

    Sorry, but I couldn’t disagree more. IGES has undeservedly turned into the whipping boy of CAD data exchange. IGES isn’t getting any younger but since when has being old – unto itself – been a bad thing? (I actually remember listening to FM radio play of Phil Collins’ first solo album on the date of its release. Please consult Wikipedia for ‘FM radio.’) IGES is tainted by the fact that a lot of CAD is written to IGES files in a way that explicitly violates the spec. That’s a problem with the CAD software, not the IGES standard.

    Arguing IGES v STEP borders on vi versus emacs, boxers versus briefs, Simpsons versus Family Guy, Pepsi versus Coke, ham versus turkey at Xmas, and whether or not peanut butter should be stored in the refrigerator.

    The point being that IGES and STEP (and native files and tessellated approximations) can both play a role in CAD data exchange. One being good does not make the other bad. There are times with both will work and there are times when one may have an advantage over the other depending on your purposes and expectations.

    (BTW the answers are IGES, vi, briefs, Family Guy, Coke, ham, refrigerate.)

    • Nick Hulslander

      So you’re telling us that you’re replying to this message from Netscape on a Windows NT 3.X computer? Just because they’re old doesn’t mean anything at all, right?

      The point being that age, development and relevancy definitely go hand in hand and there is no denying it. Just because something is ancient and still works doesn’t make it better for the cutting edge CAD systems of TODAY.

      • John Chawner

        Hi Nick:

        “depending on your purposes and expectations.”

        IGES is a standard that, as Mr. McCullough points out, hasn’t been fiddled with in quite some time meaning that there’s no “versioning” issues. Most CAD systems write it and most CAE systems read it. There’s a large number of existing files out there that can be used. And if people just took a little bit of effort to follow the spec, IGES files could be an even more robust geometry transfer mechanism than it is today.

        Is IGES perfect? No. Is it the best choice for everyone? No. Is it the best choice for every purpose? No.

        This is not a zero sum game. Just because STEP exists doesn’t mean IGES shouldn’t. It is not a matter of obsolescence as your browser and OS examples imply.

        I just object to demonizing IGES. It isn’t bad. It is what it is and when used correctly can be effective and robust.

    • Matthew

      From where I sit IGES’ bad rap is well deserved. Buggy surface models are a waste of designer time.
      I agree that age alone is not a good argument against an interchange format, but poor performance and lack of support for solid models is. At the same time, while IGES is not and will not be my go-to interchange format, I certainly don’t want that capability removed from my CAD system. Another fall-back option is always a good thing.

  • John McCullough

    I enjoyed the humor about how old IGES is, but like John Chawner I disagree with your conclusion. IGES might still actually be a superior format to STEP for long term data archival. The fact that it has been stable for a long time and no one is continuing to mess with it is the very reason why. It is unfortunate that it has been mostly used for surfaces only.
    I’m sure many CAD/CAM vendors that never really finished their IGES translators hope that you are right and the format will somehow die before they need to put any more work into it. The more likely future is that it’s use will continue to fade for another 20 years at the rate it has been for the last 20. The last diehard IGES users will be the military contractors who are required to keep project data accessible for 30 years.

    • Paul Smith

      Is it possible the very reason IGES is valuable is that hasn’t changed for such a long time. Isn’t that a testimony to it’s inherent strength. After all, geometry hasn’t ‘aged’ .
      Whereas STEP is ‘in development’ and that is it;s problem! I send a AP214 file to a client who tries to open it is platform that only imports 203. Ouch… or his translator has issues and it doesn’t work with either 203 or 214 but we cannot decide where the problem lies. Is it me or him or both?
      So I send an IGES and he can progress … that works for both of us.

      If IGES isn’t evolving is great – we can trust it. And what is a solid anyway! I bet you have NEVER sent one via computer. Cheers

  • Genero Novus

    As a “user” of various modellers and CAD systems over the years I never had the view that a choice of two or more generic CAD formats is a disadvantage.
    Using STP, SAT and IGS as final data formats ( … yes, all three) for data backup AND as final product for the client is a win/win/win scenario.
    It appears that an either/or logic is being adopted when considering generic CAD formats, which I find baffling. Unless I have the authority to create these generic formats and modify them – I prefer to just use them to their collective advantage … read and/or logic. Disadvantages of the proposed and/or approach could be:
    storage space,
    file numbering and
    time overhead for creation.
    Well, cost for data storage space is deflating faster then Genesis without Phil Collins,
    file numbering is solved by using
    FileName.STP
    FileName.SAT
    FileName.IGS
    (where ‘FileName’ is identicat throughout)
    and using a cable tie to put them together into FileName.ZIP or such.
    The time overhead for multiple file format creation in my experience is seconds (!) per assembly in most cases. I can deal with that.
    The IGES format, BECAUSE it is now static and limited lends itself – like the very limited and static Latin language for DEFINITION that is not time and version sensitive and therefore does not need constant updating and maintenance.

    I guess that’s why medical, pharmaceutical and even the legal professions have chosen a dead and limited language for definition?
    Did I mention the Catholic Church?

    The other very practical use of the IGES format in my experience is the limited/simple attribute. It lends itself for 3rd party PURGING of overly complex and unclean assemblies and designs. They shouldn’t exist – we all agree – but they do, especially in a global ‘cross corporate/cross subcontractor/cross CAD platform’ design and data environment.
    Why simplifying CAD data files? … Because the (complex history) original will always be there, but 80% of value adding can often be found in 20% of CAD data.

    I am thankful for STP, SAT, IGES, Latin … the Catholic Church and even the Protestant Church – because I can choose and use all to add value to my work as I see fit.

    Ave.

    • Matthew

      I guess that all makes sense as output formats for the endstage of a project. I think this piece was more focused on the file formats a designer will input into your project at which point it definitely IS an either-or choice. If I have the choice of importing a purchased part as an IGES and getting buggy surface models or importing a STEP and getting an error-free solid body I’ll take the STEP every time (after Native CAD, of course). IGES is an inferior format for data import because it doesn’t have mature solid body support (though at times I’ll take any extra option as a fallback).
      I do like that when I export an IGES at the end of the project it does wipe away all assembly structure, since we usually don’t want our customers to have everything (and hand it over to a cheaper competitor for version 2.0). This can also easily be accomplished in most CAD systems by saving the assembly as a part before export, which usually has additional options for which components to include for simplifying the model.

  • Sean

    I don’t see how the counter argument can even be made here… IGES isn’t being supported moving forward in time. That pretty much ends the conversation on its future as a file format. Unless someone decides to keep supporting it, than it is over for IGES. It may still have some utility today. Much like Windows XP. You “can” still use it. However, in a few years you wont want to. Are you still using Windows 2000? Can you list one file format that has perpetuated without continuing support and development? Some of the earliest file formats, such as .DWG and .DXF persist because they are continuously supported. Things that are archived are done in the format that the either the originals were created in or in the most modern universal format available. It has always been that way. Nobody is going back and translating new things into Sanskrit and putting them on papyrus scrolls for archiving. This is not a manual tool like a hammer or chisel that retains usefulness though time. It is software. It is either evolving or dying. Period. IGES no longer has the ability to evolve.

  • Eddie Smith

    I often import CATIA native files to Solidworks. When we transfer files via STEP we typically get “shattered” models and runaway surfaces. IGES almost always works properly in these situations.

    IGES is old as dirt, but it works better for me than STEP. PS – I am way too young to listen to Genesis.

  • The one thing that’s very difficult about STEP is that it truly tries to define a genuinely massive scope, however I believe it absolutely should. I’m surprised by the IGES love in the comments–but I do understand the point put forth.

    The projects underway in STEP cover just about everything from mechanical design, to composites, to electrical harness definitions. I for one am very excited indeed!

  • joe

    STEP although it is great and always my choice for solid models it is worthless for translation of 2-d drawings. IGES may not be perfect but neither is .DXF or .DWG so until a better solution comes around IGES is here to stay. There is still a need to be able to share 2-D drawings between companies .Just because something is new does not mean it should be used this seems to be a concept some of the younger crowd can not grasp. Technology is great but only when used in the proper context to apply it just because it exists is just not practical in all cases.

  • digiformer

    I was prefering STP for a long time, since once we had really bad experiences with IGES. But i think this was due to user errors in the export-import process and also bad modeling.
    Now IGES is having a come back for me since it does not recompute the surfaces. In some cases STP changed geometry came out with messed up trim edges or rebuilt 90 deg revolves to 360deg revolves which fail to trim.

    IGES allways transfers the geometry information i put in. And the solid modelers have no problem to import solids from the IGES data.
    Sure it would be nice to get the layer names transferred.

    With SolidWorks its often even better then STEP, since the way STP describes geometry it sometimes messes up small fillets to edges and stuff.

    On the other hand IGES is such a simple to understand (text-) file format. I remember back in the days, when sending huge files over the net was not so easy, i repaired broken IGES files by simply editing the textfiles and saved a lot of time, becouse we almost instantly could work with them.

  • magno_grail

    Wrinting a program to import geometry from a QuadraChek CMM to GibbsCAM. I would have used DXB but Autodesk does not give the file format so I am using IGES. 3D Points, lines and circles are all that is needed so no useless mood lighting is required. STEP has ISO copywrite and not free.