Model based product definition 6

[cgsuresh55]

I would like to learn from the experts in this forum what they think about the new trend of "MBD" or "Model Based Design/Definition". Does it really make sense in the long term?

Suresh

 

[TheTick]

Seems like the world is progressing from miscommunicating with bad drawings to miscommunicating with no drawings.

 

[capnhook]

Don't worry, a large group of seemingly very self-important PhD's will be convening soon on this very subject, sharing meals at a couple of fine luncheons. They will get nearly nothing accomplished, and the world will keep spinning.

Proud Member of the Reality-Based Community..

[green]To the Toolmaker, your nice little cartoon drawing of your glass looks cool, but your solid model sucks. Do you want me to fix it, or are you going to take all week to get it back to me so I can get some work done?[/green]

 

[JNieman]

In my opinion, it's about scope. No one model, no one drawing, will ever be "the definition" for the product that gets it done from engineer/designer to fabricator. Even when everything was done on paper by hand, there were still design-drawings and shop-drawings.

I think the concept of model-based-data is fantastic for a design definition. It's succinct, less prone to over-definition/tolerancing of features, harder to deviate from design-intent, and still provides down-stream users with the information needed to fabricate the products completely.

The downside is learning curve, which exists any time we do something different and will always be present no matter what.

I think it's equally as, or less, prone to communication errors or human error, so I don't see much down side in the concept itself as a communication medium. I don't think it's any less efficient than traditional media, either. Or, at least, it's not significantly less efficient.

_________________________________________
NX8.0, Solidworks 2014, AutoCAD, Enovia V5

 

[cgsuresh55]

Thanks for all responses. JNieman - I agree with your points - MBD does look very good for product definition per se. But how about the downstream entities in the enterprise - the shop-floor, for example? Now you will need completely CNC equipped shops that don't need technicians to read drawings! The whole infrastructure needs to be geared for this new world. Isn't that a bit too ambitious?

Also,would anyone know if any company has yet implemented this philosophy?

 

[capnhook]

Technicians, techmishicians....? Your Machinists are the ones that will lead you to your "new world", cgsuresh55.

Lets make more of THEM.

capn


Proud Member of the Reality-Based Community..

[green]To the Toolmaker, your nice little cartoon drawing of your glass looks cool, but your solid model sucks. Do you want me to fix it, or are you going to take all week to get it back to me so I can get some work done?[/green]

 

[JNieman]

Our CNC machinists make parts using a program derived from the MBD model.
Our shop-floor drawings contain all the MBD dimensions, tolerances, critical features, key characteristics... things that really matter.
Our inspection check lists include all those same items as are on the drawing and any other dimensions we see fit to verify for internal purposes (features used as witness lines for other operations, temporary features for intermediate setups, etc)
Any other dimensions needing to be on the print are added as reference dimensions for the purpose of aiding machinists in their job.

Here's what we don't do for our CNC machinists - put enough dimensions on the drawing to fully define the part, thereby creating a situation where machinists have to, or feel the need to, check every single piddly little unimportant dimension. We create an intelligent and efficient inspection scheme based upon risk management to decrease wasted time.

All else is answered via a simple model viewer with analysis tools, provided for each and every CNC machining cell (we have cells of 2-4 machines).

If we have parts going to entirely manual machinists, or those who program at-the-machine, then it gets a fully defined print. That's the only case for this. Those that program-at-the-machine sometimes don't get a fully defined print either, though, and get a DXF file to program from.


You may or not realize this but the vast majority of the machining world is ALREADY 'geared up' for this new world. Even simple Joe Bobs in the garage making tractor shackles are using HAAS Toolroom lathes and VF mills. Many people are taking their knee mills of-old and buying conversions with programmable controllers. Welcome to the 21st century

_________________________________________
NX8.0, Solidworks 2014, AutoCAD, Enovia V5

 

[drawoh]

cgsuresh55,

About forty years ago, I worked as an inspector in a car parts plant. They made oil strainers and battery cable shields. I did dimensional inspection by sitting the parts on fixtures. I never saw drawings. This is MBD.

This worked. The plant made thousands of parts per day. It took seconds for me to inspect the parts. The inspection tools probably were expensive, but given the production levels and time saved, the process worked.

What is MBD supposed to accomplish? Any drafting time saved by MBD goes away if the fabricator takes the model and completes the drawing. They might do this. They have to inspect things somehow. I benefit from making detailed fabrication drawings of the stuff I design. Aside from providing control over the fabricated parts, the close look at each part required for me to do the drawing, is my final functional check on everything.

MBD is a idea if you are doing rapid prototyping.

--
JHG

 

[verymadmac]

Good luck finding software to read the definition data file in 50 years, or computers capable of running the software.

there is no single definition system that is all things to all aspects if a program, the trick is to find the best system for each individual program, which seems to be rarely accomplished.

 

[KENAT]

cgsuresh55 - take a look around at the previous posts on this subject.

thread1103-239768
thread730-282775
thread730-263441
thread1103-262305
thread730-256663
thread730-251417
thread1103-182896
thread730-184173
thread1103-182500
thread730-184173
thread730-221206
thread730-77046
thread730-282775
thread1103-385943
thread1103-373786
thread1103-349546
thread1103-308088
thread1103-276931

I'm still of the perception that full MBD can be made to work in tightly integrated supply chains but outside of that are more troublesome.

Also based on my reading of Y4.41 I still have trouble seeing how a model is as easy to 'read' the tolerances etc. from as a good 2D drawing.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[btrueblood]

It's easier to point at a piece of paper (like a drawing) when arguing over whether a part meets its requirements or not. Even when I use a CAD model to define a shape/geometry, I still have a flat 2D drawing that specifically calls out that model file and any tolerances that pertain to the geometry within. All of my model files are in a neutral format (.step or similar) and have nothing but geometry in them. Most vendors are pretty happy with the system, and we get good parts.

 

[HGMorgan]

MBD is not a new trend.
I was part of a team at one of the largest aerospace companies in the world, 10 years ago, and the team mission was how to implement MBD while implementing CATIA V5.
MBD is not a bad idea, but the V5 process for doing MBD is poor.
V5 has MBD tools, but they are not strong, and we had to use work a rounds to output an MBD model into Enovia.
They ended up doing MBD for a while, now they have gone back to drawings, and are revising some of the old MBD in order to output drawings.
Several issues of their MBD process struck me as funny:
(1) The process they used for MBD was outputting Power Point slides of each MBD view. This really cracked me up because the drafting module of V5 is great and very easy to use.
(2) They had originally said that they would use the model viewer software so shop could review models for MBD. Computers, software, training and update issues instead of a drawing that anybody could use?

MBD will be the way to go one of these days in the future, that time is not now.


Harold G. Morgan
CATIA, QA, CNC & CMM Programmer

 

[JNieman]

I have to use Enovia to work with CATIA V5 MBD. Can confirm; it is inelegant and cumbersome. I'm reasonably sure I know which aerospace company you speak of as I've also laughed at the supplied powerpoint slides. Other portions of this company also supplies NX-native models and those are much easier to work with and require minimal work to get 2D drawings from, when I need to. Some of the time we're able to export a JT file and simply work from that (computer at each manufacturing cell).

_________________________________________
NX8.0, Solidworks 2014, AutoCAD, Enovia V5

 

[KENAT]

HGMorgan, funny I lost faith in pure MBD about 7 or so years ago when some folks on here from aero/defense were saying how they were all MBD etc. but then made mention of having to send jpeg screen shots etc. with the models.

Hybrid i.e. model and partial drawing has some really good applications etc. but pure model I'm still highly skeptical of any time soon.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[HGMorgan]

Oh, I forgot, we called it
MBD (Manufacturing Builds Drawing)

Harold G. Morgan
CATIA, QA, CNC & CMM Programmer

 

[drawoh]

HGMorgan,

Good one!

I will be remembering that one.

--
JHG