I-DEAS future, if any 2

[dar654]

I have noticed there are many large companies that still use I-DEAS for FE-modeling/analysis. There are good reasons. The modeler addresses novice, intermediate, and high-end users. Users can use the functionality that meets their individual needs, whether beam, mid-surface, or solid modeling. Modern CAD tools are solid-centric, and the mesh must be solid elements associated with the geometry. I-DEAS does it all, the mesh can be associative, or non-associative. Also, the 3D modeler is extremely powerful relative to competitive products, with his design roots embellished with features useful for analysis (e.g. partition, join partition, to create multiple volumes within the same part -- extremely useful and powerful).

Multiple parts can be displayed on the screen concurrently, rather than one at a time. This has been lost in CAD modelers today, one must put away the current part, create another one separate, put away, then another. I-DEAS allows creation of a complex part, by first constructing multiple sub-component parts on the screen, each precisely located and oriented as the analyst desires. There is no need to create an assembly of parts, and define a multiplicity of relationships to orient each part to another. One can simply sketch-in-place or align the workplane to an existing part face, and then create a new part OR an extrusion to the first part. Competitive FEA tools have primitive 3D modelers, leaving that task to the CAD tools. By having a robust modeler, the analyst can better drive the design, and quickly propose and analyze conceptual changes to optimize structure shapes. More commonly, the analyst asks the designer to provide a new STEP file of all changes; whereas, I-DEAS puts tremendous power into the analyst's hands. Of course, you may often hear, "that's a design job, beneath what an analyst ought do" -- that is wrong.

I-DEAS also offers a simple to use Team Data Manager. This can be used standalone by the user, or set up as a shared database of parts and FE-models. This can simplify replicating an existing FEA model into another model file (and generally breaking TDM associativity during check-in/out for simplicity). Once one has used this function, they would find extremely powerful, and useful.

OK, so why am I rambling onwards. I-DEAS truly needs an upgrade path. I don't mean to NX, with its complexities and data structures (i.e. tree of files in folders to represent a model), limit of one part on the screen at once (requiring assemblies perhaps), or downgrading to FEMAP. Even without major upgrades since I-DEAS 12, the software is still powerful and utilitarian at the same time. There are obvious features that would be nice, such as better meshing to handle bad STEP data to overlook small features, cracks, nearly invisible adjacent vertices. Currently, the user must be given perfect geometry, or the user creates his own perfect 3D geometry (often desirable to rid of tiny design details, and conceptualize for analysis anyway). Other enhancements would include hex-dominant solid meshing, defining more types of contact interfaces between surfaces (fixed, sliding, contact, rotating), disconnecting nodes "within" surface boundaries between adjacent volumes (currently disconnects along boundary, too! not so useful), faster solvers to leverage now common dual-quad processors, etc.

If SDRC (oops, Siemens!) is listening, Siemens should provide an upgrade path for both I-DEAS and FEMAP. Let's call it I-DEAS FX. Retain the I-DEAS Master Modeler and Assembly tools, team data manager, and most the FE tools. Take the best aspects of FEMAP and integrate into I-DEAS FX. FEMAP users could quickly learn the most basic of I-DEAS modeling tools, whether line, surface, or solid primitives to get off the ground, and then learn the easy-to-learn yet powerful solid modeler to advantage. I-DEAS users would finally be able to stay with the product, rather than switch to another FEA tool (Ansys, Hypermesh, Dyna, Comsol, Femap, etc). The large set of organizations still relying on I-DEAS will most likely never switch to NX for their FEA platform. Those who could switch to meet their needs, have already switched. Those left could be recombined into a more powerful I-DEAS/FEMAP product. I-DEAS sites that stopped maintenance would restart maintenance payments. Everyone wins. I'm surprised no one has thought of this, or presented this idea to the user base.

 

[JohnRBaker]

Sorry, the long-term future is NX. While much of the Modeling/Drafting side of I-DEAS was merged into/combined with/replaced by NX functionality, on the CAE side, the final result consists of much more SDRC technology and software fully integrated into the NX architecture, but with the added bonus of now having access to the full suite of NASTRAN technology which UGS had acquired prior to the SDRC buyout. As a result, we truly believe that the 'sum-of-the-parts', across the entire current NX product line is greater than anything either UG or I-DEAS was able to provide before we started the process of bringing both of these products together into a single, albeit UG architecurally-based, 'new' product now marketed as NX.

And as a side note, NX is NOT limited to using ONLY native OS-based file management systems as we do offer a broad range of fully integrated PDM capabilities with the Teamcenter product suite, again something which was born out of the combining of both UGS and SDRC technology, in this case Iman and Metaphase, into what is now recognized as the most widely used PDM system used by industry today, even by many companies which are using competitive CAD systems (for example, on a global basis, more CATIA data is being managed using Teamcenter than by any combination of Dassault/IBM supplied PDM tools).

Anyway, you really need to look at the current offerings of Siemens PLM. Note that our latest version of NX will be released soon, combining the best of UG and I-DEAS technology across the complete CAD/CAE/CAM sprectrum as well as the 10+ years of combined development effort since the merging of the UGS and SDRC organizations.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.

 

[dar654]

Yes, its real sad that I-DEAS as a standalone modeling and FEA package will be defunct. Siemens needs to listen to major I-DEAS FEA sites, and ask why they have no plans to ever migrate to NX. Most would not care about having a widely used PDM system based on Oracle, and increasing admin requirements and complexity. Keep it simple, and affordable from a total cost standpoint.

OK, FEMAP will continue on. Its geometry modeling capability is rather weak, to say the least. Let's integrate the I-DEAS modeler and best FEM aspects into FEMAP, to provide FEMAP with robust modeling and FEA functionality. Long live FEMAP.

 

[JohnRBaker]

If you really want to have an honest conversation with people who have already made the transition to NX, perhaps you should post your concerns in the Siemens UG/NX forum where those people would be found. This I-deas forum will have few if any former I-deas users, since if they're FORMER users, they're all over in the NX forum.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.

 

[dar654]

Thanks for listening John. I believe the UG/NX forum is for customers on maintenance. The site I'm at stopped maintenance a couple years ago, before which I'd done a thorough evaluation of modeling in NX vs in I-DEAS. Otherwise, I would enjoy conversing with those who have fully migrated. The ability for a site to transition depends on how they have used I-deas in the past, versus how the newer NX product has been implemented. I'm also not on the direct staff, but contracting, so this is the best I can do now. But at least I put my thoughts into the product evangelist! Were FEMAP to be upgraded to the I-DEAS master modeler engine, it would stand apart from other standalone FEA tools -- which in my observations have weak 3D geometry modeling tools. Another problem in the SF Bay Area is who is really using NX for FEA. Like the 2012 election, the ultimate issue is about jobs, jobs, jobs! Perhaps, you can pass on my thoughts to the developers. 30 year I-DEAS FEA engineer.

 

[cadcoke5]

First, I have no experience with I-Deas' FEA or anything more than the basic package. But, it is interesting to me that a number of things that you mention as a benefit of the I-Deas modeler, are things that I found very confusing. I come from a background using SolidWorks, Inventor, and Pro-Engineer. I-Deas seems to have a work flow that is so different from the ones I've worked on in the past, that it took me quite a quite a while to feel confident with I-Deas.

In some ways, I-deas is a stuck in the decade it started. The other modelers (especially SolidWorks, and Inventor) have benefited from being started as a 2nd generation product. I-Deas has legacy issues holding it back to some extent. E.g. much of the interface is not Windows standard, even if it runs on windows.

Of course, learning a new CAD packgage, and expecially learning the logic running behind the scenes, takes a very long time. It is always very difficult to switch CAD packages, and every one has their strengths and flaws. Figuring a work flow that does not crash into the program's flaws takes a lot of trial and error. Perhaps a year or more of effort is needed.

Personally, I am anxious for our company to switch to NX, because I continue to be frustrated with I-Deas.

Joe Dunfee

 

[dar654]

Your are correct. Initially, I-DEAS can be a bit confusing the first week. But the modeler is will relevant, and extremely powerful -- and serves the novice user, intermediate, and high-end user, i.e. FEA analyst. Of course, there are those who feel any modeling is the job of the designer. But with an open-mind to learning, the tools places power in the hands of the analyst, who has the best feel for how the structure can be strengthened, and quickly conceptualize and analyze changes, thereby jumping ahead of the designer (who would follow-up with actual design detailing). I have taken classes Ansys, Patran, Catia, Solidworks, Autocad, and evaluated NX, and find any equivalence with what I-DEAS has offered the user for modeling.

The online tutorials are great, and one can actually follow them to get basic understanding. Ultimately, a large organization needs at least one proponent who understands the modeler very well to fully leverage the product for analysis needs. In fact, I'm trying to help users at work to get them beyond the basics, and into real 3D modeling, and leveraging the history tree for easy system model updates of component parts. The key is to create component parts in-place, i.e. precisely located relative to one another and each with its own short history of boolean operations. They can then be combined to create a bushy history tree via join, join partition (so powerful!), add (e.g. for contact surfaces), or whatever suits the current analysis task. To alter the model, one could modify the component part separately, then display the system model and component part at the same time. Click on the History Tree icon and select an edge of the prior part within the system model, then display the full history tree for the system model. The feature you selected for the component will display its text feature name on its leaf. Select the top node of that feature on the three (not the last leaf itself), Modify, Replace, select an edge of the modified component part, Update. Voila, the system model has been quickly updated by snapping in the modified component part -- without having to directly cut/carve the system model. Nearly all my modeling these days uses this approach. Features in the history tree are labeled as component parts are created (as is customary for a good CAD design engineer) to they all don't just say e.g. "Extrude 542". To this end, my modeling is efficient, as I seldom create a complicated model for a single validation run. Nearly always, there are design changes that result that require updates to component parts. This approach lends to quick re-analysis.

For very advanced users, one can extract a component from within the system model history tree, e.g. say you lost the original part but can identify the branch in the history tree for such part, then one can use Conatruct, Feature Copy, ... to extract the part and its features. I used to think this was designer only stuff, but now find it useful. And the list goes on.

Likewise, I used to wish there was a way to turn off the team data manager, but eventually saw the light and learned to leverage the simple database for modeling purposes. Its takes time and practice, no classes needed.

Now if you always get perfect STEP parts from the design engineer, then I-deas might not be the best choice.

 

[SDETERS]

I have been an I-deas designer for about 10 years or so. We are switching over to NX from I-Deas. I think I may have one and half years on NX. I have not found one thing that I-Deas did that NX can not do. (beside the difference in filleting) I love the I-deas history tree as dar654 has pointed out. For NX using multiple bodies "bushy tree in i-deas" in the part navigator one can use Feature groups, manage solid bodies and accomplish close to what one can in I-deas. In NX I feel the wave link (associative copy in I-Deas) is way more robust. Our transistion from the I-Deas TDM to teamcenter and finally NX has not been smooth at all. But we are finally getting there. A guy in our company had a good idea and mentioned Siemens should have taken solidedge and replaced it with I-Deas. I liked that idea. I-deas is a good software and the drafting package was very nice. NX 2d I am still struggling with but finally getting there. There are things I want to do but struggle doing. For example sketching in lines, attaching then adding Datums ETC. to the line itself, on revolved features. I think NX has some really powerfull tools. Soft Blend I used a couple weeks ago and wow that tool worked great. I enjoy some better "drafting" tools in NX. Draft from tangent faces for example. SO I see NX having some more powerfull tools than I-deas and I think when the updates and more input from I-Deas users, I think NX will be more powerfull efficient modeler than I-Deas. It is just the huge cost of getting away from the TDM which Rocked. I have no experience on the FEA. Side of things

Thanks

 

[dar654]

I would probably agree that for part modeling, NX is more robust than I-DEAS. But for modeling oriented towards finite-element analysis there can fundamental differences (depending on the organ.) that can make migration difficult. Team Center does not help that either, and TDM is more robust than most FEA types have even ventured. There was one feature in NX that was really nifty when importing geometry from another CAD tool. One could select geometry on the part, and modify its dimensions or even delete. I suspect NX was creating a "feature" on-the-fly on top of the Orphan part, and then allowing the user to alter or delete. I would have loved to have this capability in I-DEAS, as too often the STEP data has excess detail for FEA and one would like to simplify without having to revert to the original CAD tool to hide such geometry. Can anyone name this NX capability, i.g. like a virtual feature?

 

[SDETERS]

Synchronous Modeling. Being able to move and changes faces on imported parts. Really Really nice.. I-Deas will not get this I do believe Good Thread Dar654

 

[dar654]

SDETERS, Yes, when I saw Synchronous Modeling in 2009, I thought was just so cool! and highly useful! Too bad it seems NX will not be implemented here. BTW, does NX meshing allow hex-dominant meshing with the latest release? I'd seen that in 2009 in competitive analysis software, and wish we had that too.

 

[Dward]

Just seen this thread - I'm with dar654 on this one.

I've been using 3D cad for 25 years now, and Ideas for the last 17 years, since MS3. We also have Solidworks here, so I have first-hand knowledge of the benefits that Ideas can bring to my design work. I don't do any FEA, almost all my design work is on complex plastic injection mouldings. Things like multiple parts on screen, create in place, offset surface, extract part/feature, and - most important of all - a history tree with a proper 2D structure to it - these are the things that make Ideas special and unique. I'm sure NX can do some of these things, but I haven't yet seen in it the ability to design structure into the history tree, in a way that makes design and subsequent modification very straightforward.

It's probably difficult to argue your case with somebody who calls themselves an "evangelist" - you sort of feel their mind is already made up, in a way that cannot be changed. I still believe in the old saying "the customer is always right", especially when that customer is continuing to pay large sums of money in maintenance each year. If Ideas is dead on its feet, why are they still charging money for it? I would dearly like to think that dar654's plan of floating Ideas as a separate system will happen, but I don't think the people that matter really understand what they have, and what they may be preparing to throw away.

Dick Ward
Bristol UK

 

[JohnRBaker]

For the record, my title was NOT my idea but rather was given to me by my boss at the time ;-)

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.

 

[dar654]

I have used I-DEAS for FEA the last 30+ years to CAEDS v1 on mainframes, then UNIX, and now Windows -- and have been an I-DEAS "evangelist" since Master Series was released -- even supported I-DEAS for large design team for several years. I would hope once all I-DEAS customers with any hope of converting to NX have migrated, that Siemens would wake up and find a way to continue the I-DEAS technology as a standalone analysis product. I-DEAS stands alone from all other FEA tools with its superb geometry modeling tools, i.e. easy-to-learn, powerful, data management, and part handling. I-DEAS meets the needs for basic users, intermediate, and high-end analysts. Here's a sample "bushy" history tree for a complex FEA model that I created from scratch (nothing complex, just a bunch of boolean operations, and so much better than a 500 step straight-line history tree):

Granted, few I-DEAS analysts have learned or been taught to create 3D models in this fashion, with structure and multiple leafs and sub-trees for ease of modifying components within the "system" model. No need for assembly, or Assembly FEM. Simply construct each component part at the precise location to mate up with adjacent components, or copy such parts and move to another location, and then boolean join, join-partition, add to create desired interface between volumes. e.g. "add" to create coincident surfaces where contact is desired for stress analysis, or later modify to join-partition to create adjacent volumes with continuous mesh across the interface for a modal analysis. As each component part is created, assign a Name, so that the component can be easily modified for later design changes, and then use the History Tree to "Replace" the feature within the tree. As with any good CAD engineer would do, try to name each leaf in the tree for easy identification -- particularly the bottom leaf in the tree, the more named leafs the better. After all, how often does the initial CAD design meet stringent stress or dynamic objectives? Rather, components are sequentially optimized in shape, until objectives are met. Using the tree to slide in a proposed component change shortens the overall design and analysis cycle.

My current role allows me to provide guidance to peers, to make them more efficient and not be satisfied with using basic modeling function, but to leverage the I-DEAS 3D modeler and history tree. Hopefully, the entire team will take advantage of this approach to reduce overall time to optimize the design. A little training and encouragement can go a long ways. Of course, we need to keep the product alive.

So in a way, I am an "I-DEAS EVANGELIST", with all of its connotations. I-DEAS modeling is fun compared to other standalone analysis tools; we get paid to have fun on-the-job. Perhaps, John Baker was an I-DEAS evangelist at one time, but his title might be reworked to I-DEAS->NX evangelist. In the meantime, perhaps Siemens will see the light, and find a way to keep the I-DEAS technology alive and even grow it.

As in the Big Yellow Taxi, "Don't it always seem to go That you don't know what you've got Till it's gone ..."

 

[JohnRBaker]

Sorry, but I've been using UG/NX since August 1977, first as a customer and later as a employee of the organization(s) which developed and supported the software. To learn more about where UG/NX came from, just select the last item in my signature below...

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.

 

[dar654]

John, Apologies! I totally misread Product Evangelist as I-DEAS Evangelist. Perhaps, I stand alone as the latter! BTW, the UG/NX Museum is great!!! Really takes one back into the history of UG and I-DEAS CAD software and hardware, and the relics of time.

Naturally, I really liked the SDRC-Chronicle.pdf and familiar faces of decades past. Is Bob Haubrock now involved in NX? I recall we had to shake down a pre-release new version of CAEDS in the late 1980s. Bob stayed late, and gave us a personal tutorial on the new product as a hands-on manager of the FEA team at the time; good guy!

I-DEAS Evangelist

 

[JohnRBaker]

Yes, Bob Haubrock is very much involved with Siemens PLM. He leads the 'Product Management & Business Development' group (consisting of some 70 employees, one of whom just happens to be yours truly) which is part of the 'Product Engineering Software' organization (responsible for the development of both NX and Ideas) headed by another person whom you may be familiar with, Jim Rusk, part of the CAE effort at SDRC prior to the buyout and who, since that time, had been running our Engineering Analysis and Simulation organization until he was recently given responsibility for the development of virtually ALL of the non-Manufacturing/non-PDM software products currently offered by Siemens PLM Software Inc.

John R. Baker, P.E.
Product 'Evangelist'
Product Engineering Software
Siemens PLM Software Inc.
Industry Sector
Cypress, CA
Siemens PLM:
UG/NX Museum:

To an Engineer, the glass is twice as big as it needs to be.

 

[dar654]

John, in that case, please give a friendly hello to Bob from Darryl Chong. My entire CAD/CAE support department was forcibly exited from IBM San Jose in 2004, i.e. reward for great work. Luckily, I was able to quickly switch to nearby United Defense, now BAE Systems. They still used I-DEAS, which suited me fine. Lots of I-DEAS modeling effort and skill required. But NX was/is not in the cards. You likely know other UG/I-DEAS folks at Cypress CA: PO Mehta (SoCal services mgr) and Marilyn Tomlin.

 

[Helrezer]

I just found this thread. It is interesting. I have used Ideas before it was Ideas, and it is great software for the FEA user. However, the reality is that Ideas is going away.

Your company may not chose to upgrade to NX, but NX is the best CAE software in the market today.
The design in context is easy with NX, it does not require an assembly as dar654 says. And the CAE modeling allows design in context with the latest version. It is a very powerful capability. Also, getting STEP files that have poor geometry, is easy to fix with NX and Synchronous Technology. And with the WAVE association, the user can import a new version of the STEP file and keep full associtivity! It is a very impressive capability.

I have worked in the FEA software industry for 40 years, I have used cards, text editors and many GUIs. I am thankful for the graphic pre/post capabilities we have today. And I know before I retire that what we have today will look old fashioned. I figure that I need to learn new software every 5 years, even when using the same product.

I have moved several large companies from Ideas to NX CAE. NX is a new way to do analysis. But NX can accomplish FEA faster and more accurately then Ideas. The user must learn a new workflow, but the capability is there. Most large companies will do an evaluation of the currently available CAE products when leaving Ideas. And I think that is a reasonable business decision. However, I have only had ONE company not select NX as the best in class.

DAR654 says that he evaluated NX several years ago. Well, I would say you have to look at what is available today and compare with Ideas. There has been no upgrades or changes to Ideas but we have put thousands of man hours into NX. And into FEMAP. FEMAP is not going away, but it is also not planning on becoming CAD software.

Marilyn Tomlin
Siemens PLM Software

 

[dar654]

Thanks Marilyn, My brief evaluation was just that, given time constraints before I lost access to NX. NX has surely improved well beyond NX6. I recently started work again, and I-DEAS is again my primarily modeling tool as the company did not upgrade to NX since I left in 2009. Conversion to NX will take dollars and commitment, and the company would need to perform a detailed evaluation of NX for their processes. For FEA, I have no doubt that NX offers the best geometry modeling capability among current CAE tools, and most CAE software rely on importing geometry from CAD. Of course, I would love to see I-DEAS have a new leash on life, as its too good for the bit bucket.