How to make the transition to MBD smoother: tips and tricks.

[gabimo]

I have seen this question posted on different site and it is a great interest of mine too:
How to make the transition to MBD smoother: tips and tricks.

Or otherwise stated how to qualify an existing 3D CAD model for the MBD process?
Redraw? Start from scratch?
Measure/Qualify the existing model?

Did anyone has experience in a successful transition to MDB that can be shared? What to do ? What not to do?

I am not only talking about adding the GD&T on the model, adding the notes on the model screen/window, but more on how to make sure the model/assembly is correct and not fudged?

Open type discussion for the experienced users in this field.

 

[drawoh]

gabimot,

Why are you transitioning to MBD? What do you want to accomplish?

--
JHG

 

[gabimo]

To eliminate the drawings (not sure entirely or to be used for convenience only or as a partial definition/inspection). The main idea is to reduce the time spent on creating fully defined drawings (master/ contractual agreement) and transfer the "legal" documentation to 3D CAD model.

 

[drawoh]

gabimot,

How are you going to inspect your parts? What does your fabricator need to make your parts? If they make a drawing, then the time and expense of drafting is not eliminated.

Where I work now, the drawings are partially dimensioned. Anything without a dimension is expected to comply with the default tolerances on the title block, and with the expected manufacturing process. Any dimension and tolerance on a drawing is a feature that must be inspected.

At my last place, MBD was implemented because MBD was the new way of doing things. There was no effort to think through design, documentation, manufacturing and inspection.

Solve problems.

--
JHG

 

[gabimo]

At my last place, MBD was implemented because MBD was the new way of doing things. There was effort to think through design, documentation, manufacturing and inspection.

That's exactly what I am looking for: How did you do it? Any good tips to share? What to avoid? How did you validate the models, specially the legacy models? Every dimensions on the drawing has been comparea with the drawing?
What are the "efforts through design, documentation, etc"?

 

[drawoh]

gabimot,

I have just corrected my typo. "There was no effort..."

Avoid being stupid. Don't implement MBD because it is in fashion. Implement it because it makes your process more efficient and reliable. Make sure it makes your process more efficient and reliable. Look at your total process, not just drafting. We left the legacy models as is. Design and model checking was cursory at best.

--
JHG

 

[gabimo]

drawoh,
Well, while I appreciate your concern and advice, the decision to go to MBD/MDE and model centric design is not coming from engineering. Engineering would be happy to maintain the status-quo, but higher / upper level management (the pressure from the customers, competition, suppliers, etc.) derived this resolution to move to MBD. Strange, right? And I agree with you.
It is not a matter of choice, but the matter of having a job tomorrow.

 

[drawoh]

gabimot,

Why do your customers, competition and suppliers care about your design and documentation method? Ask upper level management why they want MBD, and what they expect you to accomplish. Tell them what the transition will cost, and tell them how productive you will be once it is implemented.

--
JHG

 

[greenimi]

Drawoh and everyone,
What are those horible stories that people are hesitant to make the transition?

 

[gabimo]

If the CAD Analyst- Lead CAD Designer has been hired mainly/specially for this job, then I would say will be equally "stupid" to question the manager (or the manager - manager's) who hired him why the MBD is needed. I think drawoh hijack this thread a bit.

I would appreciate if I get feedback for what I need and not to be asked on "why you need that".

 

[drawoh]

gabimot,

As far as I can tell, MBD at my old job saved some drafting time, drove up costs overall, and reduced quality. MBD is cool, so they implemented it. At my new job, there is aggressive DFM going on. DFM and GD&T are discussed and argued about obsessively. They are very good at it. They talk to vendors. The current process is to partially dimension fabrication drawings as noted above, and send them out as PDFs with a STP file of the CAD[ ]model. There was a coherent plan to communicate with vendors to clearly define ordered parts and get them delivered at minimum cost.

I am not saying that MBD is bad. I am saying that any technology will fail in the hands of idiots. MBD is not a magic wand that instantly makes you 20% more efficient.

If you implement MBD for the sake of implementing MBD, you will fail. You need to design stuff and tell vendors what you expect them to ship you. Talk to your manufacturing people, your purchasing people, your inspectors and your vendors. Identify your problems then solve them. Maybe MBD is the right way.

--
JHG

 

[mkcski]

drawoh:

Very well said!

Certified Sr. GD&T Professional

 

[ewh]

drawoh brings up some excellent points...

If approached with the required background research (communication through design, documentation, manufacturing and inspection, both in-house and with suppliers) it can be done.
Partially dimensioned drawings seem to be the most popular method to date, and this is understandable considering the huge investment that must be made by all parties to achieve full MBD.
Ideally, only minimal dimensioning would appear in MBD. ALL features would be controlled by specific and default feature controls which would be associated to their respective datums, features and surfaces. Limited ± dimensions would be used (hole tolerances for example). From a dimensional management perspective, I know that these files are much easier to check than those that rely on dimensioning alone. The intent is to make the part machine readable to all downstream processes. This takes good communication and the investment of $$$ for software, equipment and training on the part of all parties.

As skeptical as many are that full MBD (code 5 per ASME Y14.100 NON-MANDATORY APPENDIX F) will ever actually occur, it is currently being seriously developed by large aerospace and automotive concerns.

As far as legacy models, if there was proper quality control in their creation for use in drawings, this should not be an issue. The model should correctly represent the part definition and the drawing dimensions should be associated to the model and show the correct values. If not, then this is another kettle of fish to sort out.

"Know the rules well, so you can break them effectively."
-Dalai Lama XIV

 

[CWB1]

I cant comment on the "correct" way to implement MBD as I'm not sure I've ever seen it used effectively. In college 10 years ago I worked for a small firm that used MBD to help reduce lead times from days to hours (usually next-day delivery) for prototype parts procured locally from only a handful of suppliers. IMHO the key was much more than MBD, that only saves a couple hours on a process often bogged down by days of other bureaucracy. Regardless, we simply embedded GD&T and notes into the models then shared our native files with suppliers also running Inventor. We didnt get into 3d pdfs, use MBDs for production documentation, or otherwise get too heavy into the process as I know other companies have been.