3D PMI What is the future and or are 2d drawings going away like the drafting board? 5

[SDETERS]

As I mentioned in a different thread, are 2D drawings a dying art>? What is the future of 2D drawings, vs. 3D documented models? We are still using 2D drawings, but we are controlling only critical dimensions on the 2D drawing. We are looking into annotated 3D models or 3D PMI. As Cad programs keep improving to translate between different packages, CAM CMM, are 2D drawings going to be like the old drawing boards?

 

[Burunduk]

I guess that vendor didn't read this:
article

 

[ctopher]

I haven't seen that article. Thanks for sharing.

Chris, CSWP
SolidWorks
ctophers home

 

[3DDave]

In addition to software version control, there is my reluctance to send intelligent CAD files out into the wild. With 3D STEP files, dumb is good. 2D PDF copies of drawings tell vendors what I need them to know.

There have been efforts to secure files, even PDFs, against modification or dispersal. It remains a problem that has no great solution. Even STEP + 2D PDF can get to a 3rd party factory and them pumping out copies in a few days. It will be interesting to see how long it is before Shawn Woods (YouTube mousetrap enthusiast) sees his latest invention duplicated and sold via the same distributor he uses (one named after a piranha filled river).

MBD requires a more sophisticated checker to ensure the model is made to match the expected nominal for the machinists to account for their particular process variations so the finished item will be within requirements. Without a drawing to compare to that particular job is very difficult.

The paper doesn't include automated variation analysis. It also skips the automated tools that can be used to ensure all feature dimensions appear on the drawing. Finally, electronic paper is cheap - the example of a blind hole would be removed were there a view from the opposite side.

 

[looslib]

That article is good. It was written in 2016 and noted deficiencies in MBD. With the release of STEP AP242 and ASME Y14.41.1 since then, have the CAD vendors done anything to improve the use of MBD.
We use Creo here and I know PTC is pushing the software to be better at producing MBD models. The problem is that we operate in a highly secure environment and CAD models do NOT leave our network. Even taking designs internally for SLA is a complicated matter when it comes to transferring the information between systems. I do not see us pursuing MBD anytime soon.

"Wildfires are dangerous, hard to control, and economically catastrophic."

Ben Loosli

 

[3DDave]

I see the limitation is the sophistication level of the downstream users and the willingness to spend money on the software to support it.

Given the low acceptance of variation simulation and low willingness or ability to close the loop from inspection results back to design engineering, I don't know how far it can reach.

From what I recall NIST was initially working on was a lights out factor for the DoD - the DoD was looking to cut procurement problems and having to stockpile spare parts and other inventory by instead stocking raw materials and a pile of CNC machines. I suppose some of those raw materials would be used to make casting patterns to be used in the casting area. It raises the question as to how such a factory would manage variation and if the designs entering it would require considering that variation; it would not be productive to have a factory automating the production of scrap and there would be no ability to negotiate design changes to accept out of tolerance items. This is what STEP was intended for.

 

[Wuzhee]

As long as CAD can't create visible, real threads, knurls, etc, then MBD will never be a complete thing.
I was always tought that 2D is a thing of the past and everyone uses MDB nowadays, although my CNC class was mostly making tooling jobs on paper by hand and 2D with pencil and ruler...
Never seen a contractor or my workplaces in 6 years who used only 3D data.

 

[greenimi]

I guess that vendor didn't read this:
article

Burunduk and all,
If you have more articles and studies as the one you linked above, please post them here (or in a different thread), as I would like to educate myself in this area of expertise. So far, I have very limited experience and mostly negative around and in the MBD/ MBE area (basically nothing works as intended and it is very hard to sell it....) , but continuous improvementis is one of my priorities.
Thank you

 

[Burunduk]

greenimi,
in this link is not a study or article, but it's some webinar video on geometric tolerancing, and at the very end of it, starting from minute 38, there is a very brief question answering on MBD, with some issues & possibilities mentioned. So, I think that might be of use to you, considering you want to explore MBD more. Another reason to pay attention to this, is that I have reason to believe (but am not certain) that the person asking the questions is one of the major contributors in this forum .

 

[3DDave]

That's just taking 2D snapshots of model views and storing the 2D views in the PDF. 3D PDF is unsuitable for manufacturing.

 

[SDETERS]

The supplier is using the 3D model to make the tooling, or machining too correct? The 3D PDF with the PMI in my mind is a rotatable drawing at this point. The embedded PMI in the 3D model seems useless for down stream operations like CMM programing today, due to translation and program limitations.

 

[greenimi]

Thank you Burunduk,
I've been watching the video and the applicable MBD section. It is informative to a certain extent.
I am more interested to find ammunition on how to sell MBD to higher management and to the average engineers. So, far I am very disappointed about what I’ve seen (and learned) in general about the MBD current capabilities.
Again, my interest is to find pertinent professional studies (from reputable associations and/or companies) about native CAD models, CAD translators (to STEP, JT, 3D PDF, etc), inspection and metrology (CMM interaction with MBD annotated models), manufacturing CNC interaction with CAD derivatives to generate the CNC path (or mold design and creation). Something that people tried and studied and found out that works. Again, what I discovered (with my limited knowledge) is that not many of “the features”, goals and MBD/ MBE ideas described above work.
The video is good, but I acquired from it limited knowledge. Not much essential staff learned from it.

 

[3DDave]

I too can Google:

https://www.capvidia.com/blog/mbd-model-based-definition-guide

 

[AMontembeault]

I can only speak anecdotally, but the department I work for has had some pretty good success with transitioning to a hybrid of 3D PDF and traditional 2D drawings, with an eye towards eventually going to just 3D PDF's in the future.

For us, MBA doesn't take any more time than traditional drawing - sometimes less, actually. Downstream of the "drawing" however, there is considerable time savings. I can drop an annotated model into tolerance analysis software and automatically generate all the tolerances for the analysis - previously, I had to manually enter them one by one. Hours of work dropped down to mere minutes. Our CMM operators can also dump the annotated models into their CMM programs and generate an inspection program much faster than previously (in one case, 8 hours of manual programing was reduced to an hour of "cleanup").

A number of our suppliers were only using the drawings for inspection purposes anyway and preferred to work with our CAD models to generate their CNC programs. We buy a considerable amount of CNC mandrel bent tubing, for example.

Finally, we've found that for parts with complex curvatures or difficult compound angles, working with 3D models was a superior option, because we didn't have to worry about potentially confusing 2D projected dimensions, or having to manually calculate the projected tolerances.

That being said, there is still some areas of our designs that 3D PDFs haven't really caught up to yet, but that's probably more a limitation of our preferred CAD software than 3D PDF's in general.

 

[SDETERS]

What type of tolerance and CAD software are you using? The tolerance stack software for us is a possible option, and we are looking into this also.

 

[AMontembeault]

What type of tolerance and CAD software are you using? The tolerance stack software for us is a possible option, and we are looking into this also.

Autodesk Inventor, which can be equipped with a 1-D tolerance analysis tool that largely relies on model based definition for WC and RSS outputs, but we mostly use 3DCS Multi-CAD for monte carlo based 3-D analysis.

 

[greenimi]

https://www.capvidia.com/blog/mbd-model-based-defi...

3DDave,
That is what I was talking about when I said "It is only Propaganda, Advocacy, sugar coating and more importantly "BUY OUR SOFTWARE"

My question to you: how to justify, in front of the decision makers managers, that MBD works? Try it, right? Really?
It is like the now famous saying "Speaker ......... on Health Care: "We have to pass the bill so that you can find out what is in it.""
I don't think in private bussiness that concept works (might work in other institutions)

 

[3DDave]

greenimi - you cannot justify it without proposing a solution and that solution requires software and that software requires a software supplier. If you don't want to deal with a software supplier because they sell software, then you are going on the path to failure as your first step.

That presentation is the same presentation You will have to make to management to get them to buy your solution to decision makers.

You will need to show what you want to do and how you want to do it and if you aren't in the position to create a full stack of software you will need to show how the software solution you propose meets your requirements and what compromises may need to be made, such as ditching current suppliers or retraining workers.

Having been to a large number of selection proposals the one I linked to seemed to be direct about the subject. It's mainly not propaganda and is the least sales pitch I've seen from a supplier.

If you have a template for what would satisfy you, then post that and I can help fill in the blanks.

---

That famous statement is taken out of context. It was a lamentation that the only way to see what the Republican closed-door committees were doing was to pass the bill. She was calling out liars.

Most important, the contents of the Affordable Care Act had been publicly available and publicly debated for months when Pelosi made her remarks in March 2010. The bill, in its original form, was passed by the House of Representatives in October 2009, and in the Senate that December. Although the bill was unusually long (the act runs to 906 pages in the legislative record, with many more pages of regulations) its contents had been subjected to intensive debate and scrutiny in both houses of Congress.

That process was fundamentally different to the secrecy surrounding the Republican-sponsored American Health Care Act, when Pelosi tweeted on 20 June 2017 that "Americans deserve to know what's in the [Republican healthcare] bill." At that time, the architects of the legislation had not published any of its contents. (A draft of the bill was published on 22 June 2017, two days after Pelosi's tweet).
Sources

 

[Burunduk]

Try it, right? Really?

Yes, try it. If you can initiate a pilot project and document all savings or even future potential savings (after overcoming trial run issues) in delivery time and manufacturing cost. Management won't buy it unless you show them some figures based on your own company's experience.

As far as technical issues that may arise go:
For importing models with PMI to most inspection software packages, a correct* conversion of your CAD files to STEP files of the 242 protocol should be the better than other solutions. * - adjust the conversion settings to include the 3D annotations/PMI data.

The main obstacle you may encounter is the difficulty of the involved personnel to give up the familiar in favor of the efficient.

 

[greenimi]

For importing models with PMI to most inspection software packages, a correct* conversion of your CAD files to STEP files of the 242 protocol should be the better than other solutions. * - adjust the conversion settings to include the 3D annotations/PMI data.

Burunduk,
Let's be technical here. Two main problems with your suggested solution
1.) The conversion to AP242 is not complete, correct and fully accomplished (missing surfaces, gaps, missing PMI's, missing association between PMI and their respective features)
2.) AP242 importer beside being an adds-on (read cost $$) is not able to "translate" everything and manual intervention (for over 60% of the features) is needed and required. No automatic / machine readable annotations from AP242 is currently available or working properly.

So, could you, please, offer me some rebutle arguments how can I sell it to management?

 

[Burunduk]

On the technical side I suggested you the method that worked reasonably (and better than other ways) - for me.
If you tried the same and it was that bad, contact the technical support of the software packages being used in your company.

"No automatic / machine readable annotations from AP242 is currently available or working properly" - this is different from my experience.