Degradation of Drawing Standards 13

[chancey]

Ok I am not old old school because I didn’t start on the board but has anyone else noticed the degradation of drawing standards? I think this really boils down to the use of CAD. Now that most packages have become easier to use (unlike the early versions of Pro E & MDT) I think more people are creating drawings. Things I see a lot of are missing hidden lines (I understand clarity reasons), lack of centerlines, dimensioning with disregard to intent (like not coming from datums), lack of tolerance consideration (like 3place decimals throughout), dimensioning to hidden lines, poor overall dimensioning & view layout, over crowding, etc. And god forbid they try to use GD & T.

I must just be getting older and grumpier…

 

[KENAT]

MadMango,

Sadly not here. Our manufacturing mgr tracked ECOs on two projects. One very large one which was almost fully checked. Once mid size one with no checking.

The large one had significantly less ECO's than the mid size one.

He used this to convince the old ops VP & help us get in some contract checkers to look at a couple of programs that had gone thru without checking and bring them up to par to allow us to go to other vendors etc.

We have a new VP of operations who our engineering director believes has been convinced to abandon the idea of having good drawings in favour of throwing any old crap out to make it go faster.

Don't let logic & evidence disrupt things.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

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

 

[drawoh]

KENAT,

We have a new VP of operations who our engineering director believes has been convinced to abandon the idea of having good drawings in favour of throwing any old crap out to make it go faster.

This is the fundamental problem. If management is not determined to have good drawings and good product, they will not back up the design checkers and inspectors. Crap will continue to go out the door.

JHG

 

[KENAT]

The Engineering director has no background in mechanical as far as I know and hadn't even heard of a checker before he started here. His theory is 'scrap & rework', churn out stuff as fast as possible and fix it if it doesn't work.

What baffles me is they do it to be 'faster' and yet it surely takes longer to correct problems when they occur (and they do occur) than having it done right in the first place.

You save maybe a week or two early in the project for the sake of a month later or possibly worse if the problem isn't found till full production.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

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

 

[MechEng2005]

I think the other issue is that, in some industries/applications, the crap that goes out the door works 90% of the time (or more). Also, management and sales (both equally driving the direction and policies of the company where I work) don't recognize the "pretty crap" from the quality (though most will catch the "ugly crap", along with the "ugly quality"). If we have components that need to be located precisely, then I will make the drawing to reflect this. However, if it gets in the hands of management/sales, instead of dimensioning for quality, they want it easy to see what something is going to look like. The purpose of the drawing is different for engineering/fabricators/sales/etc... I think this is the issue that needs to be resolved. Perhaps more technical training for the people working with engineers. I also agree with the more drafting training for everybody.

-- MechEng2005

 

[ajack1]

I too can go back to the days of working on the board and remember having my drawing ripped to pieces for not being pretty enough and compared to some of the works of art they certainly were not, but boy were those guys slow.

That is the first thing that has changed, man hours are now a large cost compared to the past, virtually anything that can be automated is, companies will spend a lot of money on equipment to reduce actual man hours. Spending hours just to make a drawing look pretty does not make economic sense.

Secondly CAD has taken away drawing skills and to a degree the need to check everything. Once I have model something up I do not have to calculate where things are I can pull off any part, cut a section at any point it is all there for me, the only “skill” is in how I dimension it but every dimension will be exactly what it is. So surely I should be expected to make fewer mistakes?

Production techniques also play a big part, so many parts now in automotive are so complicated in shape you could not dimension them even if you wanted to, but that does not matter as a CNC will machine directly to the model, if I can machine a part with 25,000 surfaces on it do I really need to dimension where a hole is?

I am perhaps only speaking about automotive here but I doubt it. Products today are so superior in terms of fit, reliability, performance, safety, speed to market and relative cost, if the problem is as great as this thread implies how can that be? When I go and buy a car I car about all those things, I am not sure I care much how pretty the drawing of the A pillar is.

How many companies prosper who have rows of draughtsman standing at boards creating things of great beauty with a row of checkers looking over their work?

 

[KENAT]

ajack. The thing I spend the most time checking is the tolerancing. I don't think you mention that in your post.

Just because you have a model doesn't mean you have the tolerancing.

You mention fit in your penultimate paragraph but no mention that part of that comes from the tolerancing.

I don't care how 'pretty' a drawing is. The main effort of my drawing checking is making sure it's complete (we use some hybrid 3D so for those drawings there are few if any actual dimensions to check) & unambiguous.

I really only care about 'format' type issues if it affects clarity of intent & could make the drawing ambiguous.

Most of the OPs issues fall into the category of those that could cause lack of clarity or give the wrong requirement.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

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

 

[ewh]

As usual, I agree with KENAT here. Model Based Definition is being used more than ever, and that greatly relieves the problem of poor drawings. It does not, however, relieve the problems of poor tolerancing. Even though the end user can interogate and program to the model, he has to know what tolerances to apply, and which features are affected. Lettering and line quality don't matter here, but proper tolerancing (in some manner) is still very important.

"The ambassador and the general were briefing me on the - the vast majority of Iraqis want to live in a peaceful, free world. And we will find these people and we will bring them to justice." - [small]George Bush, Washington DC, 27 October, 2003[/small]

 

[KENAT]

For anyone enjoying all this 'it's not like it was in the olden days' talk some related threads. The second one looks like some of the good content was trimmed out for some reason

thread730-219181 thread732-222977 thread730-221206

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

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

 

[MikeHalloran]

I'm capable of making a very nice drawing, in pencil.
Not to MIL standards, mind you, but at least internally coherent and accurate, clear and legible, maybe even pretty.

All of the CAD systems that I've used get in the way. They make it easier to change a drawing, and to make an accurate drawing, but legible and pretty are... well, the programmers who write this crap don't understand the words, so it's not a high priority.

I've seem some halfway decent SW drawings (but still instantly recognizable as SW drawings) in our archives, produced when our Engineering Department had about 8 people doing about the same volume of projects. Those people are all gone. They've been replaced by 3 people: me (engineer/ manager<conscripted>/designer/drafter), one designer/drafter, and one blithering idiot. (Don't ask.)

There isn't time to do pretty, even if somehow I could induce SW to be helpful with it, or find the time and the money to take an actual lesson. Yes, I know how productive SW is; I spent all day today resolving a parametric blowup having to do with toolbox part configurations on the wrong server, and then fighting the too-smart magical drawing generator, generating the same sheet over and over until I got it... well, acceptably awful.

At our last meeting, a customer took the opportunity to disparage the drafter for crossing an extension line with a leader. He did not know that _I_ am said drafter. He did not realize that his other choice, other than an ugly drawing, was _no_ drawing in time for his damn meeting. My tongue still hurts where I bit it.

...

Actually, when I put on my Manager hat, I might have given him an ugly drawing intentionally, even if I had the staff to make good ones. Said customer spent so much time complaining about how ugly our drawings were, and not so slyly intimating that he used to be a designer, that he completely missed the pieces that were missing, unidentified, and/or wrong.



Mike Halloran
Pembroke Pines, FL, USA

 

[ajack1]

KENAT, I do not doubt the importance of tolerancing but as MBD becomes more and more common dimensioning becomes less so or not even an issue, as does checking as all “sizes” are generated from the model.

I think Mike hits the nail on the head, where he says about someone picking up on the ugly drawing but completely missing things that are wrong or missing, that to me is far more important.

If as Mike says above 3 people can now do the work that 8 did before and as I said at least in automotive things can be done quicker, and produce a better finished product are we really going down the wrong road?

 

[MikeHalloran]

>>> I <<< don't think that 3 people can do the work of 8.

The people who hired me apparently do.


I can find no basis in fact or experience for their beliefs.



Mike Halloran
Pembroke Pines, FL, USA

 

[ewh]

...as does checking as all "sizes" are generated from the model

And you just take the word of the designer that he actually modeled the correct sizes? I have come across too many model/drawing files whrere the dimension on the drawing is the dimension required, but is not what is modeled. You still have to check the model thoroughly, just different items than you would on a drawing.

"The ambassador and the general were briefing me on the - the vast majority of Iraqis want to live in a peaceful, free world. And we will find these people and we will bring them to justice." - [small]George Bush, Washington DC, 27 October, 2003[/small]

 

[ajack1]

I would see that as something different ewh, to have a different dimension on a drawing to what it is modelled at needs the operator to fudge it (at least on any CAD system I am aware of) that is not a mistake it is someone doing something that is wrong.

Having said that I guess that opens up the whole can of worms that is unilateral v bilateral.

 

[JMarv]

And you just take the word of the designer that he actually modeled the correct sizes?

Yes, in the auto industry, that is how it is done. Drawings normally state that the model is the "master." If there is a discrepancy between the model and the drawing, the model is correct and the drawing is wrong.

Joe
SW Office 2008 SP5.0
P4 3.0Ghz 3GB
ATI FireGL X1

 

[KENAT]

ajack, I must be missing something. Checking to see a part is fully dimensioned is pretty quick, this is what going to MBD does arguably remove the need for. Great, that saves a few minutes per drawing or item for checking - although significantly simplifies/speeds the initial design/drafting.

What takes significantly more time is checking the tolerancing and the correct communication of that tolerancing and other requirements not captured by the nominal geometry. I don't see MBD saving much time here, in fact in some cases I can imagine it being more timeconsuming.

From what I've seen of other posters in this forum it seems automotive is an area that takes tolerancing fairly seriously. I appears many other sectors struggle more with this.

I get your point ajack, at the end of the day the drawing the drawing or if you like 'design documentation' (so as to cover models etc) only has to be good enough.

The problem seems to be defining 'good enough'. Is good enough a half assed sketch or thrown together model with little to no consideration to tolerancing or other 'not nominal geometry' issues?

Or is 'good enough' design documentation that will probably get what you want from your usual vendor/in house shop but doesn't explicitly define all requirements or really capture design intent for the benefit of future users.

Or is 'good enough' design documentation that will allow you to source the part from any competant vendor (in relevant field) with a high degree of confidence, will stand up in a court of law as a contract document and captures enough design intent not just to ensure function/fit with the largest possible tolerances but also to allow future users to understand and make changes to it.

While it may very a little by circumstances, I'd usually vote for the latter.

Am I mistaken?

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

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

 

[duk748]

i just love this forum!! - just this week i had a detailer make the comment to me after i had checked a drawing of his & bled all over it "i have worked in this business for 30+years & i still get drawings back that look like this!"
not to be a smart a__ i just looked at him & wondered how he had made it this far in 30+years & was still not turnning out decent drawings?? - i am a board taught designer/drafter since i was 18 & i am now 48 & am still amazed at the lack of standards & engineering knowledge that comes thru our company - sad sign of the times

 

[aardvarkdw]

I am trained as a draftsman but I am now working as an NC Programmer. I have seen first hand where MBD does not simplify things. On several occasions I have had to redraw a part because the engineer did not model the part correctly (read: in a manner that makes using the part practical). For this, I have to use the DRAWING to get my tolerance information so that I can redo their poor work. If the drawing is incoherent (and it usually is because they have had no training and no incentive to make a good drawing. "The model is right isn't it?"), I have to waste my time getting the correct information. Furthermore, I have to explain to the CNC operator why they have no dimensions with which to do the bench inspection that the engineer has required. Instead they have to find someone that knows how to use the cad software that can create a sketch of the dimensions that they are supposed to inspect.

The point of my rant is that just because we are going to MBD does not negate the need for drawings and good drafting/communication skills. If not just to describe the part, at least to communicate to the end user what our intent is.

David

 

[ewh]

Drawings normally state that the model is the "master." If there is a discrepancy between the model and the drawing, the model is correct and the drawing is wrong

That is how we do our parts also. Lofted shapes do not lend themselves to dimensioning. The master model takes precedence. But when hole sizes for example are modeled at the upper size limit or larger, but are correct on the drawing, scrap is going to be produced unless the model is also checked. Whoever created the drawing will get his hand slapped, but without a model check this would not have been caught. That is why our models are checked more thoroughly than our drawings.

Please, let's not get into unilateral tolerancing here;-)

"The ambassador and the general were briefing me on the - the vast majority of Iraqis want to live in a peaceful, free world. And we will find these people and we will bring them to justice." - [small]George Bush, Washington DC, 27 October, 2003[/small]

 

[CheckerRon]

In my aerospace checking function, many of the bad dimensioning and tolerancing issues I encounter in detail parts and fit of related parts are the result of poor modelling. Often due to bad choice of datums at the outset of the modelling function. As a checker, I only catch these after the drawing is made and submitted, and there is a lot of gripping about how time consuming it is to go back and fix the model to get it right (in Pro/E wildfire3). If you don't know how to draw, you create crummy models too!!

"Duk748: I've met several designers like your person. We call it "1 year of experience 30 times".

 

[MotorVib]

I have always had a problem with how schools teach drafting. Would we ask a writer to write a book with out learing to read and just showing them all the cool things that Word has? Dut that is what we do with all of the drafters. Hear is this program you know how to use it now make a drawing. This hit close to home for me my dad told me many years ago that reading and making drawing are what engineers do. This tells the world what we want them to do.

As a side note my company is making all the Mechinal engineers and drafter take a SolidWork test. And from what I have seen on it I could take the test and pass it and still not be able to make a drawing.

Chris

"In this house, we obey the laws of thermodynamics." Homer Simpson