If at all possible, you need to work under the supervision of someone skilled in drafting. The basics can be learned from on-line tutorials or from the standard books (French & Vierke, Giesecke, etc), but you will never be a decent drafter until you get an experienced drafter to show you the right way to do things. Oh, and expect to put in at least a thousand hours of drafting to just be a competent newbie. Most university courses, when you can still find them, are taught by people who never worked at drafting/design; here in the US, the better choice is a course from a proprietary school where instructors typically have several years of applied experience.

Come to Vincennes University and get hands on training. ALL of our instructors have earned a living doing what they teach. Many still work on the side in order to stay in touch with industry trends. One class that was added to the industrial drafting degree is a basic machining class. That was due to feed back from companies that higher our grads. They wanted the students to understand what the true meaning between a +/- tolerance of .010" and .001" or even .0001". It has been understood that tight tolerance are often applied unnecessarily costing thousands in manufacturing. Not just in tooling costs but in time to market delays. The experience gained helps drive home that you do not need to act like you are building Swiss watches all the time.

Sounds like a good school. I wasn't aware of it, so thank you for the tip.

The best thing to do is get a copy of AS 1100 or BS 308 and copy the proceedures in them. It will soon become second nature!

Good luck with it!

(you need to work under the supervision of someone skilled in drafting ) ok but you need also a help full program . In my opion the best way to start is to start with 3D drawing program ( more easy than the 2D and make you love drawing and design.so I suugest for you to start with solid work this program will put you in the rigth way.

I'd like to expand this to a point just before Mahmoud's applies:

Spend some time 'at the board', it helps immensely in getting a feel for how a 2-D drawing should look. At times i've seen some excellent 3-D models converted into some pretty confusing drawings.

In the same vein, try to get some time in the machineshop. It helps to have a feel for which operations are improbable; there are often several ways functionally equivalent parts can be drawn, a wise draughtsman will maximise the chances for success on the floor.

Also, always keep thine eyes open, you never know where a useful detail might reveal itself.

These are lifetime recommendations, they're intended improve the depth of your knowledge behind things.

SW is an xlnt CAD program, we use it all the time, & i'd recommend it (not to the exclusion of others)

Mahmoud,

I also love SolidWorks. But it will make terrible drawings unless you have a lot of skill (In my opinion a lot more experience than required for AutoCad). I used to annoy an annoying engineer by deliberately making drawings that were impossible to manufacture; they satisfied every rule and they were generated by the software - just impossible to make.

Others have made some excellent general suggestions. As a machnist of 35 years in the business who has had drawings from almost any aerospace company you can name, I can offer you some specific suggestions. These are things that at the least are irritations to the person manufacturing from engineering drawings, and at the worst can cause very expensive problems.

Please, before you start using true position tolerences, learn what they mean. A true position tol. of .010 does not mean +/-.010. No matter how you calculate it it comes out to +/-.003 for an individual dimension.

I worked with many drawings from General Electric. I considered their drawings to be the best to work with, especially with casting and machining drawings because they always used clearly defined datum planes and tooling points to define those planes, an excellent system.

It is almost always bad to dimension a part from both ends or sides. This can create really bad tolerance stack up problems. The only exception I have found for this is if you have a really long part that has to have something fit to each end and the distance from end to end is not critical.

Please do not try to do the machinist's job for him. For example: Call out a threaded hole with the thread and the tolerance like this- 10-32 UNF 2B. The machinist can then decide what size hole to drill depending on the material, or whether he wants to use a cut tap or a form tap or thread mill. If you call out the tap drill size, you technically lock him into that size which may force him to use a tap which is not right for the material or the part geometry.

Here is one that is just an irritation. Quite often I get drawings that take up as many as four pages. The first thing the machinst or person designing a mold or tooling wants to know is: How big is it. This is information that should be on the first page.

I have many more, but will end with a sermon Ihave preached here before. Once a purchase order has been issued against a drawing, that drawing becomes a contract. Therefore, the drawing must constraine the manufacturer of the product to make what is required with no loopholes. The tolerences should be tight enough to ensure quality and fit, and loose enough so as to not make the manufacturing cost too high.

Hi Garyceng

Please do not try to do the machinist's job for him.

Here is one instance that I have learned to ALWAYS do the machinist's job for him:-

Don't indicate a tolerance of e.g. H7 or g6. It is better to put the tolerance that you desire in full e.g. Ø 50 -0.00 +0.010.

Another trick I have learned is to separate numbers as numerals from numbers as text:- e.g. Ø 50 PTTN THREE PLACES.

Thinking about it, there actually is quite a bit to learn. I don't agree with the idea of starting in 3D though. It is sometimes tooooooo difficult to produce your 2D manufacturing drawings from a 3D model IF the CAD package you are using takes over too much control of the representation! CAD never works 'straight out of the box'. It took me years to perfect the setup of Auto-CAD but now that I have it only takes minutes!

Always remember that you get paid just as much for rubbing out lines as you do for drawing 'em!

I don't think we are in disagreement here. You are suggesting to give more detail in the tolerance. That is valuable, and not the same as telling the machinist how to do the job, or what tools to do it with. I always appreciate the engineer giving me clear detail on a spec. so I don't have to look it up. I don't appreciate being told what tools to use to get the job done. If the machinist cannot read the drawing and interpret the tolerances if they are listed correctly, and know how to make the part, get another machinist.

And here is one to add to my previous list. Get a drill chart that has all of the sizes both metric and English. Then try to not call out hole sizes that are not on the chart. The machinist can make a hole for you to any size, but if it is a special size, it is going to cost you.

Re: Spec of tools for job.

In our shop we record the results of previous runs as part of a feedback loop. We spec the results of the best previous run (our line consists primarily of standard product; too many folk want to 're-invent the wheel', if there's a 'best way!', how come every machinist we hire has a different way?)

Have also noted that the CNC machinists that most resent 'being told' have the biggest ego/skill ratio! (BTW; these folk have moved on, per the "get another machinist" recommendation [above])

I have to disagree with your assessment that 3D software is a bad place to start. It only takes over if you let it, and in fact, goes a long way towards avoiding mistakes prior to manufacture, using interference detection, finite element analysis, etc. (I use SolidWorks).

Although it is not the most expedient method, when I create drawings from 3D models, I dimension from scratch, as do many of my colleagues.

That way, I walk through the manufacturing process mentally, ensuring that all the right information is there to build it and build it correctly.

Although there is much to learn about drafting convention, I feel that a good understanding of the required (and/or available) manufacturing processes, as well as the thing you are detailing, is at least as important.

I have seen some beautifully elegant designs, perfectly detailed, that were impossible to manufacture with existing methods.

The post was specific to drawing/drafting skills. Not design. It really depends on the company, but generally designers/drafters do not determine the manufacturing process. Certainly if your designing something like a casting your design and details have to be made within the perameters of manufacturing. However generally the drawing is there to communicate the design intent.

Yes design for manufacturing is a consideration but it isn't the intent of the drawing. A machined part can be changed to a cast-machined part or powder metal or a weldment. Many times manufacturing notes can be related back to the designer as a revision.

In a perfect world a design team cooperatively develops a product together. In this case all manufacturing considerations can be take into account.

And we can't forget the other side of things. Some companies do have "drafters" who sole job is to detail new or existing designs developed by designers and engineers. Their purpose is to make sure every feature of the part is detailed and that the drawing is correct to specification. Nothing else.

To answer the post, the most constructive method to develop good drawing/drafting skill is manually.

Although I appreciate your perspective, I think we are talking about two different things - manual vs. CAD drafting, particularly when it comes to drawings derived from solid modelling (3D) programs.

I'm old enough to be familiar with the manual drafting office heirarchy, but in my experience that place is all but dead.

Roles have shifted considerably in the CAD office, as it is virtually impossible to segregate the tasks as they once were.

The structure has been flattened. The capabilities of software and hardware, not to mention budgets, have heightened the expectations placed on designers (quasi-engineer, design and detail), so as to significantly reduce the input required by (and available from, often times only on a consulting basis) engineers , and render "pure" draftsmen virtually obsolete.

I don't disagree with your comment on the trend of a design office. But the question is still specifically toward drafting/drawing. And this is a specific skill required to communicate through a drawing. CAD has nothing to do with it.

The mention of manual drafting is a "start with the basics" suggestion. I'm not necessarily suggesting a person practice their printing like a 1^st grader like I had to. But the remedial "art" of creating a drawing gives a better understanding of what is really required.

On CAD if you want a simple dimension you pick a type, two points and place it. Manually you must start with the extension line. What is the required spacing from the feature? How thick is that line to be? How long is it to be? Should it or should it not cross another dimensions dimension line? Then draw the dimension line. How far from the end of the extension line does the dimension line start? How thick is it? Can it cross an extension line? If you have to cross how do you show it? Then the arrow…….

Your answer may be that the CAD does it for you. But how do you know the CAD is right? How do you know some IT tech who installed the software just didn't guess when he/she set it up. Most CAD allows you to customize these. So is it right?

You can take a course and be told (or read) these requirements but until you practice them you will not learn them.

"the expectations placed on designers (quasi-engineer, design and detail), so as to significantly reduce the input required by (and available from, often times only on a consulting basis) engineers , and render "pure" draftsmen virtually obsolete."

I hate to say it but some people are just draftsmen (or is it draftspeople now?). And that is fine.

I very much agree in your assesment of a designer. In many companies an engineer attends meetings and does paperwork. It's scarey when an engineer starts trying to do a drawing. Most engineers I know fully admit they never took any drafting courses or even a course on how to read a drawing. They wing-it. To any engineers reading this, it's not a shot at you guys/gals, it's just my experience. And many of my engineer friends appreciate the help I've given them over the years with this issue.

My 2 cents...

Your comments show the value of Manufacturing Engineers. Thanks.

I've seen a lot of smart newbies, who had two weeks of AutoCad, insist on always putting down the drill size (and some of the 3D packages do it for you!). The trouble is, they use the same % threads for all materials. I've even seen them do drill sizes for reams where the ream would require removing 0.030" of material if the machinist followed the drawing.

I know it's no longer popular, but engineers (particularly Mech) really need to be trained by drafters and machinists and welders during their first year or two on the job.

Oh, I forgot one further comment on your post. When did universities quit teaching engineers about the Commercial Code? That seems like such a critical area and most of our state universities seem to have totally dropped any mention of it. I know a machine shop down the road that has literally gotten rich off one company's practice of changing drawings after a PO is issued. He always wins.

I think the primary cause of the part change after PO issue would be the 'speed to market' pressure, coupled with a management style that drives, rather than leads.

Nah. This company doesn't do anything fast. Their problem is they don't believe in revision control and their engineers never stop perfecting designs.

Well said. Excellent points. The inquisiters first step seems to have been taken, namely a desire to learn the profession correctly. The next step is to take the excellent advice of garyceng and stay away from the glitzy 3D programs. Start simply, draft in 2d whether on a computer or by hand. Take a part, sketch it in 3D by hand first and develop an appreciation for the "old way" drafting was done. Learn basics from a good book on drafting and hone you skill under the tuteledge of a skilled designer if you can find one. Work for a firm that needs a detailer, work at home at night to perfect your craft. Learn to spell engineer.

Best of luck!

If it were me, I would contact the American Design Drafting Association and ask which school/schools have received the most awards for the particular drafting discipline that you plan to pursue. Attend that institution and plan to work your ass off.

As many before me in this thread have said; the importance of an understanding of basic mechanical/machine principles and practices can not be overstated, Board work is extremely important: it teaches precision and perspective in a way that a computer will never equal. Any graphics program is only as good as its input. Know your geometry, physics, and math. This knowledge will distinguish you from your peers.

Above all else; if you do not understand something, ask a question. If the answer does not make sense............ask again.

Good Luck,

Your return will equal your investment

A good start is ASME Y14.5M. There are additional ASME standards for specific drafting areas but this is the main one. Most places I have bumped into over the years follow this standard. Even the companies that insist on having their own standards generally base theirs on this.

I consider myself lucky. I was trained on the drafting board and my first position out of school was on the drafting board. The value and knowledge of drafting I learned from manual drafting far exceeded any CAD training and applications. Manual drafting is a skill (almost and art) that will be valuable throughout your career. Even in a CAD focused company.

Education in drafting is fine. But doing is learning. Almost any city will have a drafting/design course that will give you the basics. What you learn on the job while working under an experienced draftsperson you can never get in a class.

CAD software is dangerous. Any newbie can sit and apply a bunch of views and slap on a whole bunch of dimensions and notes. And when you print it it looks really cool and neat. But what you end up with is a nightmare to read and sort out. I've seen, too many times, a room full of graduates costing companies a lot of money through poor drawings. Hell at one time I was one of them. HA!

My CAD preference for most general design applications is SolidWorks. I've gone from AutoCAD to Unigraphics and a bit of Catia and ProE. Plus a couple of others I can't even remember. Unless you are into higher end design like surfacing, SW is more than enough.

Just my 2 cents…..

Guy

I agree with you 100% about ASME Y14.5M, and also about the CAD. CAD is something you should learn after you know how to draw, and CAM is something you learn after you become a machinist. That is not to say that concurrent education is not valuable, but that there is an appropriate order of progression. Expecting the CAD program to automatically dimension a part for you is a road to disaster. A guy I used to work with said "If this was easy, everybody would be doing it". No matter what area you get into, mechanical, machining, electronics, architecture, etc. somewhere along the line you must gain experience and learn to think. This guy is on the right track, and asking the right questions.

Take a course on drafting. You can also get cert as GD&T professional. Do a search on Google and you'll located someone near you for the training. Or contact any community college.

Everyone is making excellent points here. We have all seen CADD-generated disasters that were pretty to view, but pure hell to make and/or pay for. And we have seen those that cause the machinist to spend more time reading about the values of certain callouts than was required to make the part.

I too have noticed that with younger engineers much emphasis seems to be placed on becoming well-versed with the CADD programmes with virtually no solid basics on the expected part accuracy with regards to the manufacturing method.

I trained first as a toolmaker, and later as an engineer. CADD was in its infancy at that time and we still stood at the board to make drawings. The blueprints we read in the toolroom guided us to what made a drawing successful, and this was no doubt valuable.

It does remind me of the times that people have commented on my CADD skills, and I always play it down with the true statement "the CAD is the very easy part-it is a good and practical design that is tough." It is truely so easy to draw with CADD compared to the board days, but I digress. It seems that so many think that just because they possess a CADD certificate that they are a finished draftsman. I apologize for the short rant here.

I am dealing with that now, as I am re-doing and archiving a lot of old CADD drawings of helical reducer transmission components and assemblies. These were reverse-engineered by (I guess) a young intern about twenty or so years ago, and whatever dimension his vernier caliper or micrometer displayed became the gospel for that part without any regard for manufacturing errors or tolerance ranges. Many of these have multiple origin points as well for critical features, and the sketches must be "assembled" to determine where the errors exist.

And here's a good one; when have you seen the helix angle callout expressed in degrees-minutes-seconds decimal parts of a second? I kid you not. I remember the TI engineering calculators were quite new in those days and I guess the guy thought a gear-cutting shop could "split seconds" when it came to helix angle accuracy.

None of us say these things to discourage you in any way. We mention such things as advice for how a part drawing should look and what information it should contain. Remember the man who must provide the cost estimate and the man who must machine that part. The easier you make their jobs results in a lower cost to you, and perhaps a finished part of more repeatable accuracy.

Best Regards,

Ing. Robert Forbus

Well said! Every one of us still has a lot to learn.

New CNC control capability/ spindle speed/ machine capacity has made formerly exotic materials & geometries almost commonplace. But, if one becomes too enamored w/ "trick appearance" manufacturing costs may suffer. (or, the specal look may be just what it takes to stand out in ones market)

One thing to remember: Never get too full of one's-self! The complete package consists of: Design/Engineering, Manufacturing/Fabrication, + Sales/Marketing; It's a 3-legged stool, if all aren't carrying their load, we all fall down.

(a recomendation: an evening @ your local [pub] can work wonders! [!])

RE: REVERSE ENGINEERING

i tend to nominalise dimensions, then do my own tolerances, based on my assumptions of what is proper for the application.

Feedback please, on this technique.

Guru, I very much practice the same method. To me it just makes sense. If I am working on a metric drawing I try to stay with whole millimeters, and on inch drawings I try to stay on basic graduations and not go below 1/16" if possible. With all of that said I attempt to render nominal dimensions/values and then apply the tolerances, remaining mindful of stackup issues if there are enough dimensioned features and/or parts involved.

Best Regards,

Ing. Robert Forbus

I don't disagree, but I guess it depends on the application.

This would be way too tight when designing a house and way too loose when designing a turbine. LOL

For me, reverse engineering would begin as you have stated and then a determination of the tolerance and stack-up studies that would be required for the application.

I worked at (unnamed) military design facility on drivelines. There was a multi-step machining process that was working all in one direction. Big machine and Big applications $$$. Things just were not fitting. To me it was simple stack-up but the proof is in the engineering. So I reverse engineered the whole thing and developed a current and proposed tolerance study. Pages and pages that eventually had to be worked into a presentation.

I made some Sr design engineers look pretty bad (not intentionally) and even though I solved the issue and saved big $$ it did not do well for my career at that company. LOL