A course to teach drafting to engineers who can do CAD 3

[Sparweb]

Some of you will understand what I'm asking right away (or have bemoaned this themselves) but for the rest I will explain what I'm considering and why I'm asking:

There is a difference... in fact there is a very wide gulf... between the ability to select and place the drafting icons on your CAD screen, and the ability to produce a clear, concise, and adequately detailed drawing that a fabricator can use to produce the part or assembly correctly. The distinction is important to me. Many students come out of school knowing the former. Few are taught the latter, formally. I wasn't either, but I was lucky that I started in workplaces with people prepared to teach and they set me on the right track. I think it's my turn to do the same for my younger co-workers. They are currently churning out crap.

I work with a large (~30) team of engineers who, for the most part, have a lot of experience with CAD, mostly CATIA and Inventor (SW and AutoCAD take strong runner-up places) but few of them have any training at all in preparing good drawings. Sure, they know what a hidden line is, but few know when to use it for effect and clarity. What I want to do is to create a drawing class for this group so that they can produce better, more clear drawings for the fabricators that use them. I have just completed a project where the average rate of shop error, query, revision, or change notice rate is roughly 10 per drawing. The time wasted on these frequently exceeded the original drawing time. While I don't believe that the engineering dep't has to take responsibility for all production errors, I have personally seen many errors that could be traced back to confusing instructions on drawings.

I have started preparing to give a course that does the following:[ul]
[li]Teaches good techniques for detail, assembly, and installation drawings[/li]
[li]Shows examples of good drawings and bad, and discusses the reasons[/li]
[li]Gives the engineers strategies to prevent omissions on their drawings[/li]
[li]Prepares the engineers to lay out a drawing package in a logical order before making the first drawing[/li]
[/ul]
...and by the way, I do not want the class to:
[ul]
[li]Teach the minutiae of the company's drawings standard - they can read that for themselves[/li]
[li]Teach where to click the button to make a feature appear - that course already exists[/li]
[li]Hold their hands (or their mice, either) - the point is to make the hand-holding stop[/li]
[li]Point fingers at anyone who has more difficulty than others[/li]
[/ul]

I have been looking for such a course and have yet to find one. The courses I find all seem to be about "how to click the button". So I decided to start designing a course of my own.

I confess that I'm old enough that I can say I took 3 drafting courses, with boards, pencils, rulers and such in school. I think I'm the beneficiary of that, and I'd like to pass it on, just in a way that doesn't force these kids to use a #2HB. There won't be much point in trying this if I start off on the wrong foot giving them the impression that "my way is better" or that something archaic is good for them. I love CAD and I know that CAD can be used to make excellent drawings because I've seen it done.

For those tempted to remind me that management has to care before all of this happens, I agree. I have been laying the groundwork in the minds of my superiors and their superiors for a few months, and "taking the temperature" before I make my case. I believe I am ready to do that, with facts and numbers of dollars saved by the company, or cost if not done.
Also made a few allies.

Does anyone have suggestions for making this course a success?

STF

 

[CheckerHater]

In today's world requirements for the drawings are divided between Y14.100, Y14.24. Y14.5, etc., etc., etc., so they are diluted in homeopathic doses.

But back in time you could find them concentrated in one place (experts on this forum will have no problem identifying the document):

…Engineering drawings and associated lists prepared to this Level shall provide engineering definitions sufficiently complete to enable a competent manufacturer to produce and maintain quality control of item(s) to the degree that physical and performance characteristics interchangeable with those of the original design are obtained without resorting to additional product design effort, additional design data, or recourse to original design activity. These engineering drawings shall:
(a) reflect the end-product
(b) provide the engineering data for the support of quantity production, end
(c) in conjunction with other related reprocurement data shall provide the necessary data to permit competitive procurement… of items substantially identical to the original items…
Engineering drawings… shall include details of unique processes,… when essential to design and manufacture; performance ratings; dimensional and tolerance data; critical manufacturing assembly sequences; input and output characteristics; diagrams; mechanical and electrical connections; physical characteristics including form and finish; details of material identifications; inspection, test and evaluation criteria; necessary calibration information and quality control data

So, tunalover, your good old bosses are probably stuck in good old times

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[MadMango]

I think you should print out a 3-view drawing of a part, maybe with tapped holes and a counterbores/countersinks (or whatever covers 80% of your type parts), but do not dimension it. Have the individuals in your class detail it with pen/pencil (at least the extension/dimension lines, and leaders, no need for numerals yet), and have them returned to you. You bleed all over them, hand them back to the students, and discuss the Pros and Cons.

"Art without engineering is dreaming; Engineering without art is calculating."

Have you read faq731-376 to make the best use of these Forums?

 

[JNieman]

Just to throw a wrench into this ordeal...

OP does not mention any manner of industry standards, nor whether or not the library of ASME drawing standards applies.

The question is a bit general, and we /are/ talking about entry level education, here, just a step beyond learning how to create drawings, and into some best practices for arranging/organizing/communicating with drawings. I understand ASME Y14.xxx covers this a lot. And thoroughly. However, not everyone cares. Engineers also don't all farm out their work. Many of us are attached to the shop making the products, as well, and as such, -will- want to include some process information. A good example is match-drilling or line-boring components after assembly, rather than just calling out what they need to be and letting the manufacturer figure out connecting the points.

All I'm saying is; don't mire the discussion down into the weeds of picking nits in ASME standards if the question never even encompassed them. I think a focus on 'ideas' is better than the particulars of ASME code. I think whatever applies should be spoon-fed by their checker/mentor.

I think the idea of being able to arrange an assembly intelligently is important. Whether you start at the highest level; finished product, and work your way backwards to the smallest detail/component, or another way (company standard, I assume) it's important as it helps everyone visualize the manufacturing process. It also forces one to think about features being made at a detail-level and at assembly-level. For the particularly sharp young engineer, it can also cause them to ask questions about how parts are made, maybe even noticing opportunities for improving manufacturability by rearranging things. I think that goes a long way.

The idea of creating enough drawing views to -completely- define the part is a skill sometimes needing development. Sometimes it's easy to forget to fully define something because it's your design and of course you know what that slot/hole/boss/feature looks like because you modeled it. The 101 courses usually just focus on "here's how to create a view, project a view, section a view, whatever" but never the "why". The OP seems to focus on moving from "how" to "why" imo, and maybe that question would be a guiding force. Go through all the "How do I" steps in creating a drawing, and ask yourself if there are any good "why" questions to write down for the newbs.

 

[berkshire]

Tuna lover,
I do not think my statement is way off the mark
As you say yourself "The drawing says WHAT to make not HOW to make it."
The drawing represents the final product , If a how to is required , then there are process specifications that should accompany the drawings.

You just mentioned : I just came from a company who required that, for a purchased custom transformer, for example, all electrical requirements be stated on the drawing with no other separate requirements specs called out in the notes. The E-sized drawing had the first three sheets filled with notes and tables laying out the electrical requirements. It would have been MUCH easier to have a separate A-sized spec called out in a note that spelled out the electrical requirements while leaving the rest of the drawing to describe the form and fit. This way the drawing could have been one, maybe two C-sized sheets. It is much easier to make revisions with a word processor than with a CAD system. And the person doing the revisions doesn't have to be a well-paid engineer or designer (a typist can work from a redline provided by the EE). At this company, how drawings were prepared was controlled by top management, not Engineering. That was pretty obvious by looking at their custom component drawings!
I would much sooner see a separate process specification, but I can also see why a company would want all of the required information in one in-removable document.
B.E.

You are judged not by what you know, but by what you can do.

 

[Tunalover]

berkshire-
Some will say if you send the drawings out for quote to a shop that pays their people slave wages but offers better prices, then if you don't define those processes in gory detail (that are well-understood by all the other shops with higher prices and better educated people) then you'll get what you pay for: you'll be paying a fortune to revise drawings to suit that slave wage shop until the cows come home while you get crappy parts and spend half your time on the phone explaining the drawings. You need to use shops that understand that only lesser-known processes, not all processes, need to be spelled out on the drawing.

Tunalover

 

[berkshire]

Tuna lover ,
Please re read your post , the short answer is not to send your parts there. And don't let your purchasing department do it either.
B.E.

You are judged not by what you know, but by what you can do.

 

[Mike M]

I like what Kneat said.

The advantage of starting with pencil drafting is visualization. My community-college introduction to CAD not many years ago emphasized this. CAD gives you the views for free, but the pencil doesn't. Some exercise to challenge and encourage orthographic visualization seems important.

As for making a good drawing, I ask myself if every feature is called out for size and location in three dimensions, and ensure tolerances are indicated and appropriate.

Also explain third-angle projection. Right view on the right, top view on top, etc, and aligned.

 

[JNieman]

Maybe it's my youth but I don't see a single advantage to pencil-on-paper these days.

Anything you teach on a board can be taught in any 2D CAD package like Draftsight/AutoCAD, etc. Seems to me that it's a waste of precious time and resources to cripple someone down to working on paper for the sake of someone's nostalgia.

Yes, I started on the board. I've done enough pencil and ink drawing.

Any high school or college that wastes a student's time with that crap is doing a disservice to the precious few (and in the latter case, preciously expensive) hours the student has to invest in expanding their knowledge.

 

[CheckerHater]

Pencil and paper exercise your brain. In fact I would make Descriptive geometry prerequisite for drafting class. It gives you understanding of what is happening on your screen.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[3DDave]

Pencil and paper require use of the part of the brain responsible for coordinating places with movement. The brain is forced to accurately simulate the planning steps prior to execution in order to get the desired results. Pencil and paper also allows for the rapid generation of imperfect versions of geometry, which is the opposite of the effect of CAD tools, for which such geometry is tedious to create.

 

[JNieman]

Watching Jeopardy and doing Sudoku/Crossword puzzles exercises your brain, too. You don't have to pay $3,000 per class to do that, though. (based on 28 credit hours over a $30,000 year) Or pay $400 for a Sudoku textbook. Or (more specific to the OP) when you're paying someone $45-80k/yr as an entry level engineer having to learn this basic stuff, and a mentor making >$100k/yr reviewing their work... seems like "expedience" is a concern. Giving them tools that reduce their efficiency significantly is costly and I guess I simply disagree on the benefits.

I just don't see the disconnect that occurs by deriving the numbers and typing them into a command line or variable box, versus having them derive the numbers and then measure it with a scale or protractor. Then you're teaching them habits and trains of thought that you have to break, anyways, when you want them to start making good parametric design models.

Descriptive geometry should have been covered in high school geometry or later trigonometry classes that are assumed to be covered already, imo.

ETA: Also, OP specifically said this isn't "the basics" but rather best practices for professional drawing production. In fact, to quote:

I confess that I'm old enough that I can say I took 3 drafting courses, with boards, pencils, rulers and such in school. I think I'm the beneficiary of that, and I'd like to pass it on, just in a way that doesn't force these kids to use a #2HB.]

SparWeb - since you have a team environment, maybe you could assign people to work in pairs checking eachothers drawings. You note that errors stem from instructions/notes. Pair employees up based upon complimentary strengths/weaknesses as best as can be done, and ask the checker to review it and ask themselves "Can I work from this drawing without question?" and any gaps in knowledge or requirements for clarification should be discussed and satisfied. It's not a perfect solution, as you still have a small factor of "the blind leading the blind" but it should hopefully mitigate most.

 

[3DDave]

Watching Jeopardy, et al, doesn't involve muscles with simulation of geometry. What you miss is that many students never really learn the subjects they are taught, as exemplified by the need for the standard to have a formula for the projected tolerance zone on a non-projected tolerance and, more amusing, having to distinguish a fixed fastener from a floating fastener in a calculation. If people had a general understanding of math and geometry these would be seen as irritatingly pedantic.

There's no need to measure with a scale or use a protractor in a class about making understandable drawings. It is, however, a time waster to spend time with the CAD interface trying to determine just how to go about creating some aspect of the drawing. I've seen too many users stop all the action for 'what do you click, can you show that again?'

 

[JNieman]

Apologies if my post edits came after you wrote your reply.

 

[JNieman]

If I were to go for simulation and practice of spatial awareness and imagination, I'd much rather they build something with their own hands. Go out on the floor and have a toolmaker, machinist, assemblyperson walk them through what they do, how they do it, and why this particular part of the process sucks (there's always -something- that sucks)

Additionally, the OP specifically stated this is not a 'course' to tell them what to click. It's not a "what's wrong with engineers today!" course.

 

[Sparweb]

JNieman,
I think you've got the spirit I was looking for. It's not about the pencils, paper, or criticism of anybody's education. That stuff is too basic for me to tackle in a group setting. I would be turned off myself by a pencil drawing class, so I can't expect my colleagues to enjoy it. I also don't think exercises for their mental graphics processors is the direction I want to go. Right now, I am immersed in some extremely detailed work in CAD, and I have my own OCD cranked up to the max - not a condition I should be in if I'm to patiently teach a drafting course.

Whether done by pencil or by mouse, the text in the notes and the 3-views on the sheet may still be the same for the same part for the same designer. In a perfect world, but it isn't. For example, the pencil forces simplification, while CAD permits over-complication. Here, again, the emphasis is not on how the pencil makes a drawing clear, but how the CAD reduces clarity by presenting too much information. Therefore my emphasis, if the subject comes up, is how to make a complex CAD model show clearly on a drawing. The model may contain 100,000 parts, but that doesn't mean that the drawing generated from the model must have all of them. The trick is to show engineers how to make the drawing show only the relevant parts. Maybe there are only 10 of them. They have to DO SOMETHING for this to happen. Most of them don't know what. Many of them don't even realize it's possible. When those guys create a CAD drawing from a 100k-part model, they show all 100k parts on it. Imagine you're on the shop floor and get a drawing which shows every part of the aircraft structure, and points at your work which involves only a handful of them.

This is just one example where our group is creating drawings that are very difficult to use.
I'm selecting areas where improvement will be the most effective. There are so many:
[ul]
[li]levels of detail, reducing the clutter for clarity[/li]
[li]point of view, to orient the detail view the same way as the installer's head is oriented[/li]
[li]detail scale, if the part is so small that it can barely be seen, an additional level of detail is required[/li]
[li]effective use of views (sections are not break-outs, details are not sections, and 3-views are not sections)[/li]
[li]reference points must have some relationship to the installed parts[/li]
[li]write the notes in order of operation, especially when processes are described in the notes[/li]
[li]re-read the notes, out loud; help for ESL coworkers[/li]
[li]clean up, check the drawing for yourself before submitting[/li]
[/ul]

The list above is just a hit list for INSTALLATION drawings. I am also generating lists of low-hanging fruit for assembly, and detail drawings.
I could include other types of drawings (schematics, etc.) but I should speak to my own strengths.
There are other subjects which span across notions of install/assemble/detail that I also want to cover.

STF

 

[3DDave]

Since manufacturing floor instructions seem important to your goal, it would make more sense to acquire software that is designed for that job, rather than try to force software that isn't, into that position**. Engineering CAD is great for building up models, but it is terrible for slicing and dicing according to arbitrary rules. Process planning software is made to reflect alternate interpretations of sequencing. Many include positioning of the operator and any tools required to ensure the operator can see the item and can reach both the item and the tool.

I am concerned that there aren't a significant list of read-only suitable notes. New notes should be rare, especially if you have 'ESL' workers.

"Many of them don't even realize it's possible." Sounds like poor hiring practices. I get that some people are lazy and don't want to do work, but if they aren't educated enough to know how, then they should not be there. I made a list of 50 tasks for self-evaluations, asking for each: Never heard of, Have Heard of, Have used, Have used frequently, Have taught others to use. as categories so that adequate training planning could be managed. The CAD vendors could also help by keeping statistics for users about what functions they used and how often, and compared against a list of all functions; best would be how often the process was aborted, indicating the user missed a step, but no one is really interested in improving, just go-along, get-along.

What is interesting is that most of what you want was automatic with hand-drawn drawings. The cost to draw 100,000 parts would be too high to bother with and most detail would be left off for time constraints. And no one wanted to re-letter a lengthy set of notes. Now it costs time and effort to remove it. Since this will slow everyone down, what carrot will be offered and what justification will be required to increase the drawing budget?



**One place I worked used a desktop publishing tool for this; about 90% of the instruction maintenance time was spent overcoming limitations in the software because they wanted to use it like Excel with cells and such even though it's a text-flow program. It was chosen because the person in charge of floor instructions is, get this, an art major. The instructions had significant errors - calling out non-existent items and missing items that should be placed. The shop had a continuous stream of error flags and it took a long time to fix anything because the process was so tedious. One would think an FDA compliant shop would be more careful, but not so. Their care was in concealing these errors to avoid an audit.

 

[Sparweb]

Hi Dave,
It isn't poor hiring practices, it's poor CAD selection. My industry (aerospace) is mostly using CATIA, with Solidworks pulling second place. Guess what? We use Autodesk.
There is some logic - not just thrift - in having one package which comes complete with AutoCAD, Inventor, and the Vault for document control in one package. Selecting 3 slimmer packages from 3 competing software vendors would be a lot more difficult and likely lead to integration problems. More than we already have. Anyway, proficient CATIA users come to our doors all the time, and as we bring them in, we gently break the bad news and hold their hands as they "translate" their skills into the inferior software. Since this allows some level of introduction to all sorts of other things in our company's systems, not just CAD, I don't mind doing this. In the context of a drawing course, it isn't a worthwhile subject since it's more about the pushing of buttons than the presentation on paper.

Read-only notes: Yes I remember the days when I could do that. Now we have process specifications for just about every routine fabrication procedure, which normally keeps notes to a minimum.

My trouble with notes is not the processes, it's the show-me-don't-tell-me problem. Junior users of CAD simply don't know how to depict some tasks with drawings, so they rely on lengthy explanations. In many cases, I think it's just laziness, but there is also a frustration factor. A multi-layer fiberglass lamination will show the stack up of layers if you put a cross-section on the drawing. Do it carelessly in CAD, and you only get a black mash because the laminates are thinner than the ink lines when it's printed. Disaster. Confusion. Despair. Nothing can possibly be done about this. A workaround would take hours. So instead I find a paragraph explaining that lamination 1 is bonded to lamination 2 using adhesive such-and-such... and it just goes on and on. When this frustrated designer has given up on making a drawing, and is also ESL, this is how we get into a minefield.

"What is interesting is that most of what you want was automatic with hand-drawn drawings." Exaaaaaclty....
This is for all those guys who have never had those big chests of drawings to shuffle through, therefore haven't even seen many hand-drawn drawings, let alone pick up a pencil to do it.

OK, I hope that covers the pencil/CAD comparison that the subject deserves. There won't be board drawing in any class I give, but it is a useful comparison when discussing the goal on the printed result.

Thanks again!

STF

 

[DraftingMan]

SparWeb,

After reading this post over the past week, seems to me the biggest problem your company has is that it has no drafters.

Engineers are not taught drafting. They will very easily create a multi-page document explaining the lamination process.

Drafter are taught drafting. They will take said document and create a clear graphical representation of this information, if they know the industry well.

Engineers make too much money to not be doing engineering.

Most of the industries in the world are trying to cut cost so they have decided drafter are not needed because the CAD marketing has said "Making drawings is so easy, just click a button and boom here is your drawing."

"What is interesting is that most of what you want was automatic with hand-drawn drawings." Exaaaaaclty....

Why is this, because each engineer, at one time, had multiple drafters working for them on their project and they both understood the industry they worked in.

John H. Dunten, CD
Certified Drafter

 

[JNieman]

It isn't poor hiring practices, it's poor CAD selection. My industry (aerospace) is mostly using CATIA, with Solidworks pulling second place. Guess what? We use Autodesk.

My sincerest condolences.

After spending a lot of time using Autocad at a structural engineering firm, doing industrial facility designs, pushing 3D bounds, upgrading to Pland3D... and still feeling a "bit behind"... I skip right past job offerings that mention only "AutoCAD" in the software references. Life's too short to spend your days trying to use AutoCAD to publish your ideas.

Drafting is near-obsolete in aerospace, imo. At the least, they should be engineering-affluent Designers who can do some basic engineering tasks under the umbrella of an engineer / engineering manager. Simple drafters are not really practical in a industry that constantly seeks to banish any "checker" responsibilities. The nature of modern design software also makes it rather unintuitive to use a drafter since the engineer should be done with the vast majority of the drafting in one form of another, just by creating the model as they want it.

Not to crack open a can of worms, but as Model Based Definitions becomes more common and their processes refined, it'll blur even more...

 

[ewh]

Not to crack open a can of worms...

Yes, it is coming. I am currently working for a global aerospace company at a location that, after years of hybrid drawing/MBD files, is going totally MBD - no drawings, all information in the model file.
While it is a far different environment than that of the board drawings I learned on, much of the same skillset is still required. The goal is still to define the component as clearly as possible. There are still some bumps in the road getting there, however, but it is coming.

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