Where Can I Learn to Create Proper Mechanical Drawings for European Standards?

[Dafe]

Hey, I'm studying mechanical engineering and my opinion is that the emphasis on drawing is too low.
Could someone recommend me a book or other material so I can learn to draw proper mechanical drawings on my own?
Preferably for European standard..

Thanks

 

[Danger]

All of my draughting references have been out of print for decades. This might be of some assistance, though: http://pergatory.mit.edu/2.007/Resources/drawings/"

 

[Dafe]

I've read that site and I know most of those drawing rules, but thanks!
It's kind of hard to think of things to draw..Where did you guys learn to draw properly?School or at work?

 

[FredGarvin]

Mostly at school. For practice, take any object you see. Brackets, boxes, etc... Hardware stores are great places to see examples to draw. Start with relatively straight lined objects. Add curves and surfaces after you get proficient at the beginning stuff.

I think that there isn't a whole lot out there these days. You are correct that the subject is sorely lacking now a days. I would recommend looking for old engineering books in used book stores and on line. I have a couple of old ones myself.

 

[Dafe]

I've talked some to my uncle who has been an engineer for decades. He tells stories of how they used to draw all the time, several hours a day. At my school they are like; yea yea, you'll learn it when you need it.
We've had the lousy introductory crashcourse, but it isn't nearly enough.
I'll actually do some drawing this summer, after my exams. I hope you won't mind if I post some of them here for criticism..

As for old engineering books, which are you referring too? The machine-elements type of book or something else?

 

[FredGarvin]

You've got a great attitude towards this. It's nice to see.

Your Uncle is right on. In the days, guys would spend all day at a drafting board. They got very good at it.

The older engineering books had this topic in them many times. I would just say to keep your eyes open.

I almost forgot... these are not European standard, but a free reference is from the US DOE:

Drawings Vol 1:
http://www.hss.energy.gov/NuclearSafety/techstds/standard/hdbk1016/h1016v1.pdf

Drawings Vol 2:
http://www.hss.energy.gov/NuclearSafety/techstds/standard/hdbk1016/h1016v2.pdf

I can't believe people sell these free books on EBay for $5-10!

 

[Danger]

In my case, it was general draughting in grades 9-11. I have a full set-up, including a draughting machine and home-made table, but no place to put it.
Now I'm doing everything either on scrap paper or in Illustrator.

 

[Dafe]

Thanks for the links and the kind words.
People sell used toothbrushes on Ebay :/

I'll have a go at those pdf's, I'll most likely encounter a couple of drawings done in the US anyways :)

Thanks again, got to get back to fluid mechanics (evil)!

Danger: You do full-scale mechanical drawings on scrap paper? Man you must have a steady hand :)

 

[caslav.ilic]

I will certainly agree that nothing beats the beauty of a hand-made draft (by which I mean either literally hand-drawn or using some manual drafting software), but the point of creating such drawings is not as clear today as it used to be once.

In ye old days, it seems to me that the drafts were much more intervened with the core engineer's job, computation. For one engineer to start doing his numbers, another would have to supply him with some drafts too. Then he would make a draft based on his calculations for next guy around the table, etc. In other words, drafts were crucial part of an iterative design process, and that's why they had to draft so much.

Today, much of this can be handled by direct number crunching and automatic 3D model generation. The design team can pass around "piles of numbers" as working inputs to each other, and the visualization for sanity-checks can be automated to great extent. Finally, CAD software can semi-automatically produce workshop drafts (may even not be needed for prototypes, having numerically controlled machines in the workshop).

When I was starting with "serious" drafting assignments for aerospace courses (ca. 1999), I was wondering why they insisted so much on drafting. For example, the first big draft was to make tri-projection of an existing airplane assigned to each of us, on a big A0 sheet. When I inquired as to why we need to do this, let alone physically plot the draft if made on the computer, they replied: because you will need it in the future courses, to take measurements for the calculations and position different subsystems inside. Needless to say, after having it approved and signed, I had never ever unfolded that A0 sheet again.

Later on, for design class, following the instructions on how to draft hull lofts, we were assigned to define few dozen cross sections of our airplanes. It was something like 20-30 hours job on average for us students, depending on whether we'd do it on computer or by hand (though, by that time, rarely anyone would pull out his drafting pens...) I did spend that expected time, but instead of directly drafting, making a little computer code to draft it for me (generate a series of draft instructions for CAD output), after I do the real design work of sufficiently specifying the shape. That enabled me not only to have arbitrary number of cross sections at arbitrary positions at a flick of a number, but also considerable liberty in experimenting with the hull shape, and, most importantly, it was a lot of fun.

So, unless the OP is hooked up on the beauty of hand-made drafts :), I'd suggest spending more time on getting to know CAD software (3D modelling), and especially make good use of any parametric and programmable capabilities it may have, for real design work. Remember: If recomputation/remodelling/redrafting iteration depends on a single input, then having to do anything except providing that single input, is, in the long run, waste of resources.

--
Chusslove Illich (Часлав Илић)

 

[Dafe]

Great post caslav.
At my school we spent more time learning ProEngineer than Autocad and general drawing, but there was nothing said about parametric or programming capabilities. We only modeled a lot, did some analysis and animation.

I've actually never witnessed that kind of semi-automatic design process. Do you have some good resources on it?

 

[caslav.ilic]

At my school we spent more time learning ProEngineer than Autocad and general drawing [...]

That is excellent. AutoCAD really is not the mechanical engineer's tool for the job (though the last time I used it was in 2002, so I don't know if it went some way in the meantime).

[...] but there was nothing said about parametric or programming capabilities. We only modeled a lot, did some analysis and animation.

I had some limited experience with ProEngineer, and it indeed does have these capabilities, and I did make some use of its parametric facilities. The idea is simple.

Imagine you have a metal rod with two holes in it. If you just want to draft it, you put these holes in any way so that they fit where they should. If the rod is 100 cm long, you don't care if you draft the first hole so that it's 20 cm from the rod's left end and second hole 5 cm to the left of the first, or first hole 80 cm from the right end and second hole 75 cm from the right end -- the final effect is the same.

However, as a designer, you may know that the rod length will scale according to some external influence, but that the first hole should always be 20 cm from its left end and second hole always 5 cm further left. Therefore, you draft holes with such positioning, and then when you change the rod length, the holes will remain positioned according to the design intention. That's basically parametric drafting -- where you not only represent an object, but also build up relations between its elements according to design intentions, so that future changes are much easier.

In this sense, each ProE dimension is parametric and should be carefully chosen, and properly linked to other elements either directly (e.g. "make the hole axis normal to the rod axis") or through parameter tables (eg. "let the holes' diameter always be a third of the rod diameter").

I've actually never witnessed that kind of semi-automatic design process. Do you have some good resources on it?

I don't have any resources as in published stuff, only personal experience. But this is expected, as it is difficult to write something about such processes. They are both natural and heterogenous, very much site dependent: you have a bunch of tools that you need to use, and you wonder how to chain them as smoothly as possible. Both to speed things up, and to avoid stupid errors.

For this you need to be aware of what each particular tool can do in terms of external input/output (for using them in the tool chain) and internal state updates (for automating computation/modelling within the tool itself). Aside from that, it is advantageous to get to know a general purpose scripting language (e.g. Python), which can be used as a glue in the chain (e.g. to convert one tool's output to another tool's input), as well as to do some parts of the computation itself.

(For the rod example above, for example you may wish to know how to make ProE take the rod length from some text file, and write out its updated model in certain format, with only a single command line call -- didn't get to do this, don't ask me :)

Furthermore, heavy-duty computational software (e.g. for fluid dynamics and structural analysis) is frequently run on computing clusters, which run some sort of Unix/Linux operating system. The computational tools are then typically used in command line, non-interactive environment. So, if use of such tools is anticipated, it would be good to gain some knowledge of Unix shell and shell scripting (the above mentioned Python also comes in handy here).

A typical chain would be: from simplified computational model, to CAD model update, to physics simulation, to results extraction into a human-presentable form (and possibly all wrapped in an optimization cycle).

--
Chusslove Illich (Часлав Илић)

 

[Dafe]

I believe I have a lot of learning ahead of me :)
It's kind of hard to know what to spend time on besides the actual "schoolstuff", because I still don't know where I'll end up after school.

As practical and time saving as technology can be, it's still is a pain in the ass!
I actually like modeling, scripting and the such, let's hope I'll get to do that at work :)

Thanks for taking the time to post!

 

[mechlady]

Just from my experience as a mechanical engineer, I think you should know more about how to modify drawings than how to create one.

 

[jaap de vries]

My experience is that kids learn how to make the most elaborate drawings using pro-E, solid works or AutoCad, but they are then incapable of turning that into a workable machine drawing. I say there should be more emphasis on implementing the ASME Y14.7 so students can actually communicate with machine shops and get an idea of what is possible to make (within budget)