Design and no drawing policies 3

[warburton1]

Hello Gents,

I have a new discussion I would to seek opinions on and also to find out from anybody who has experience in this arena.

I am a contract designer working for a large American corporation in Canada. Recently, and based somewhat on my experience on V5 Catia, I was asked to take part in a study program to develop a means to remove engineering drawings from the system. ie. Solid models with associative data would form the back bone of engineering output.

Has anyone had experience dealing with this idea?

I can see many pitfalls and many advantages, but the true cost of this form of engineering appears to be costly (eg. training costs), does anybody have experience of it?

The company in question is demanding everybody (suppliers etc.) must have Catia V5 or an associated viewer. Would this not tie the cost down to Dassaults pricing whims??

Any feedback would be appreciated.

Gary W.

 

[TheTick]

I believe it may be possible to go "paperless", but I don't believe there will ever be an end to 2D "drawings" in one form or another. Drawings are still a very effective means of communicating design.

It is he core function of an engineering department to take knowledge that perhaps few can understand thoroughly and turn it into useful information that can be used by the rest of an enterprise. With respect to models and drawings, engineering serves to take the design in the form of models, physical prototypes, and all other supporting documentation, and communicate the many aspects of that design to many other customers, both internal (purchasing, manufacturing, QC, sales) and external (customers, suppliers).

An annotated 3D model is not sufficient to meet all these needs. This is especially true when dealing with external contacts who are likely not to have compatible technology and training.

In short: write it down, make it clear.

If the ladies don't find you handsome, they should at least find you handy.

 

[mloew]

You might find this interesting:

ASME Y14.41-2003: Digital Product Definition Data Practices
2003 |Order No. N17103 | ISBN: 0-7918-2810-7

http://www.asme.org/codes/pr/y1441.html

3D modeling is the future of design, and ASME Y14.41 sets the rules. Y14.41 extends ASME Y14.5M into the 3D world.

The development of this National Standard was initiated at the request of Industry and the U.S. Government.

The Y14.41 standard establishes requirements, and references documents applicable to the preparation and revision of digital product definition data, referred to as data sets. Y14.41 defines the exceptions and additional requirements to existing ASME standards for using product definition data sets or drawings in 3D digital format. This standard supports two methods of application: model only, and model and drawing in digital format.

The structure of the standard begins with the requirements common to both methods (model only, and model and drawing in digital format), and then branches to the other sections that have differing requirements for each method.

As well, it provides a guide for the many computer aided design (CAD) software packages to develop better modeling and annotation practices for CAD and engineering disciplines.

Developers of this standard include independent consultants, and individuals from General Motors Corporation, Dassault-Systems, The Boeing Company, U.S. Department of the Army - TACOM-ARDEC, Rockwell Collins Inc., Raytheon Company, Thiokol Propulsion, EDS PLM Solutions, General Dynamics Land Systems, Lockheed Martin Aeronautics Company, Ford Motor Company, Caterpillar Inc., DEPCO Inc., Rolls-Royce Corporation, CNH Global NV, Purdue University, and Hutchinson Technology Inc.

Adoption Notice:
ASME Y14.41, Digital Product Definition Data Practices, was adopted on 7 July 2003 for use by the Department of Defense, (DoD). Proposed changes by DoD activities must be submitted to the DoD Adopting Activity: Commander, US Army TACOM-ARDEC, ATTN: AMSTA-AR-QAW-E, Picatinny Arsenal, NJ 07806-5000.

As Featured In:
Automotive Industries - October Issue (

http://www.ai-online.com/issues/article_detail.asp?id=303)

CadCamnet.com



Best regards,

Matthew Ian Loew
"Luck is the residue of design."
Branch Rickey

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[GregLocock]

Our 3D models are the reference, 2D drawings are produced off them when necessary (about 99% of the time, grin).

Be prepared for an enormous database task, in fact database management becomes a core skill for every CAD operator. I'd guess that roughly 2 hours per week per operator is spent on d/b management.

One direct benefit is that every engineer in the company has the ability to pull up a full 3d assembly of the car in its latest form, derived from the 3d solid models, - this is now an essential tool. For instance, if I need suspension mounting point locations I can do it myself, rather than having to ask a CAD guy to pull them out for me. Alternatively if we are designing a new fuel tank we can pull up the complete underfloor area, and put it on a projector during the design team meeting. The assembly engineers can pick up our latest proposals and study how they are going to get the thing thrown together. I don't think we had any idea how valuable this particular aspect would be when we started down this long and dusty road.



Cheers

Greg Locock

 

[genes]

I have to agree with The Tick and offer an additional probably "inappropriate comment". I have seen the proposal of a paperless engineering department many times. Today there are three types of CAD operators, drafters from the board who are just hanging on, designers and engineers that made the transition to CAD and are effective contributors, and our newest catagory, modelers. I will ask, of these three groups of people, who are the ones that can correctly size and tolerance a screw through a clearance hole and into an extruded/cast/threaded hole? A solid model can be provided and probably get some parts made. Does your Quality department have the equipment to verify the actual part back to the model? Does manufacturing have the software to use your CAD model to create fixtures/molds/dies/etc.? Can purchasing read in your CAD BOM? I think you see where I'm going with this. Yes it can be done, but not by the kind of people who read management magazines.

 

[jmw]

I personally think the idea of being able to 3D design has many advantages, not least when used with stereo lithography to produce a model of the component in short order.
2D is a limitation not of our imaginations but technology. I would bet that if Leonardo had computers and 3D modelling he would have jumped at it (accoring to those he believe he was a time traveller trapped in an earlier century, he probably did!).
Paper and pen were the only available tools for a long time and require skill to interpret properly as i can testfity to when i sfirst studied technicla drawing at school. On the other hand, a system that lets you create a 3D rotatable image is far more easily assimilated with less training (for the oberver). I would like to see 3D systems that allow the different components to be properly animated. Most valuable is still the model. Even in the days of 2D this was an vital part of communicating the design to non-draftsmen. Anyone who has visited Greenwich Museum in england will appreciate thatwhen designing warships, invariable a ship model was created, a tremendous labour in the 17th and 18th centuries. Many of these fine models survive and are on display. Drawings can be ambiguous, especially older drawings but the model is unambiguous and a great reference source to historical ship modellers.

 

[Iskit4iam]

The paperless office may be a reality someday, but as long as there is a human between the instructions and the process there will be paper. Even if wireless pad devices, which are the closest thing to paper, become widespread and cheap paper will still be strong.

For the past three years I've been involved with using Inventor and working with an AutoCAD DWG database. I had six designers working with me. Often we had to model existing parts. The process of switching windows to see the current design was very disruptive. Usually, the first thing done was to print the drawing.

In the design mode I found myself constantly refering to multiple information sources. Even mulitiple monitors wasn't as efficient as having paper references.

None of the modeling software I evaluated prior to going with Inventor toleranced in 3D. So I had to refer to the drawing for stack up data.

Paper will be around a long time.

 

[TheTick]

jmw:
respectfully, I must say you are missing the point. I believe the question of the usefulness of 3D CAD was settled sometime in the 20th century. The question at hand is whether it is feasible to entirely eliminate 2D expressions of 3D designs.

If the ladies don't find you handsome, they should at least find you handy.

 

[warburton1]

Thanks for all the responses, its been useful in adding to my own theories.

One nagging problem I foresee is the loss of the sanity check a designer quite often has in producing a 2D drawing. I find it an invaluable tool in resolving problems that may slip through unnoticed (hey, maybe I am just falible afterall).

I also don't like this argument about ambiguity of a drawing, a good designer/draftsman should not leave any ambiguity in his drawing!

Furthermore, when a model is left open for people (notice I say people and not direct NC conversion), then there will be ambiguity in its conversion. Unless of course something is specifically spelled out and isn't that basically what the drawing is for.

That said I am not against "zero drawings", Iskit4iam and Genes made a good point about the human factor, if that is reduced then I think zero drawings are definetely the way ahead, eg NC output from model data etc.

I guess that my be all and all gripe is on the data side. All the CAD software people are too busy locking everybody into their respective packages and not doing enough to make data translation widely available to all. The suppliers I use tend not to use Catia V5 and the cost of the necessary investment to them is the equivalent of their profit margin. I think Step should be made ANSI Y14.41 compliant.

 

[jmw]

TheTick:
I appreciate your point.
The trouble we have is that we carry a lot of baggage with us from our past. I confess that however much time i spend at the computer, and however fast and user friendly they get i still edit best when i print out paper and take a pen to it. But that's because i grew up in a paper era, and its easier for me to do this than learn new ticks, i just don't have the time right now. Someday i will learn to touch type too.
Come on, how many of you guys still use slide rules? Or use 5 hole punch tape to program the computer?
Sooner or later we all do adapt.
The point i was making about 3D is that pretty soon stereo lithography will become the paper print out equivalent. OK, so it may take nine hours for a laser to solidify a resin model in the tank and a day to cure, but that will improve. Given a choice over a 2D print and a 3D model, which would you choose? As has been said, we don't print out the paper because we love it but because many of us can think better that way, but i for one would find a 3D model much to be prefered and pretty soon most of us will lose the knack of reading a sD drawing just as i suspect most of us have forgotten how to use a slide rule (well, at least as well as we did). As that happens then sure, the 2D referrences will become obsolete.
Sorry to say, we'll be upt to our necks in 3D models and it'll take more shredder than Oliver North's shredder to has to get rid of them once we're done with them.

 

[TheTick]

As far as SLA/SLS goes, I work in the metal stamping industry. For stampings, it is not possible to create SLA or SLS models that accurately replicate material properties of formed sheet metal. The only decent way to prototype a stamping is to stamp it with "soft" tooling, still a very expensive and time-consuming process.

This is a bone of contention with many of our plastics-oriented customers, who are accustomed to push-button on-demand replicas of their plastic parts.

 

[Iskit4iam]

For the last 15 years I've managed engineering departments. Anybody not working 3D is making a hugh mistake. Nothing has had the productivity impact that 3D has on my job.

Tick
I agree that baggage is holding us back but I don't think the technology is complete enough either. You can't give a 3D litho to the shop and say "make this." Even if you transfer the model to your CAM system there is some concern over the state of the model. Is it modeled at MMC or nominal or some combination? Until we can bring tolerances into the model we're stuck with drawings.

I only resort to editing 2D when that's easier than generating a model for existing 2D designs. Circumstances forced me to invest several hundred hours in learning the software.

 

[GregLocock]

I think the consensus in this thread is that the solid modelling paradigm is here to stay, and that it does offer some genuine benfits.

In your opinions which aspects of it most fail to live up to the hype?

I've got two to get the ball rolling:

The first I never believed in anyway, but was one of the strongest selling points: associativity of parameteric models through to CAE. In the salesman's dreams this meant that if a part was redimensioned the FEA and so on would all be re-run automatically. Ha. In practice this could not work, except for the most mickey-mouse level of analysis, which is worth about as much as the thought that goes into it.

The second has already been mentoned : tolerancing of assemblies.

I'm surprised people want to hold 3d parts in their hands, I've always been happy working on the screen, even when I see proto parts I don't get excited, that's how it was supposed to look. I've probably got a funny brain.



Cheers

Greg Locock

 

[Iskit4iam]

I vote for tolerancing number one, data management number two. I spend way too much time coordinating model use. If I want one designer to work on the hydraulic system and another to work on cab layout I have to copy the same assembly to each, have them design and assemble the parts and then reassemble the parts in the "good" file.

Never tried to tie FEA to the model but did attempt CAM without much success. Programmer reluctance and softwear inadequacy share the blame.

 

[CanEngJohn]

Warburton1 writes

"I also don't like this argument about ambiguity of a drawing, a good designer/draftsman should not leave any ambiguity in his drawing!"

I know this may sound pessimistic but my biggest issue is that I have not seen a good designer/draftsman produce a print that was without ambiguity in the 10 years I have been in industry. I believe the computers and modelling that came out shortly after I learned Mechanical Drafting in high school (late 80's) has created a generation of "lazy" designers/draftsmen. I fist saw this issue in first year (1990) when a ton of people received basic CAD training without drawing training. Most of the drawings I saw were dimensioned incorrectly and ambiguously.

I find the designers/draftsmen that used pencil and paper (how I was trained) and then retrained for computers are still really good. Problem being there are not a lot of them anymore. I find the ones just starting to come out of school in the last year or two are starting to get it right again. However the generation in between did not get the correct training as they were being pushed through too fast to fill the demands of industry. What worries me is that the ones who are getting it right will probably be corrupted in the near future by the experienced ones who have the bad habits.

Honestly the top issue I see that has to be resolved is that we have a large capacity of designers and draftsmen that need to learn to make unambiguous models before we can move ahead with 3D modelling instead of prints.

Also we use Solid Edge here. We dimension and tolerance our 3D component models. We can use it to see the interactions in the assembly. Very handy feature.