Tips for engineers from machinists

Ok, let's face it Engineers make some stupid decisions sometimes. So how about we give some advice to our friends on the other side of the print. Some of you may know that I have recently moved into a mechanical engineering position and I can tell you that it is a blessing to the engineering department that I am here. Before I moved to engineering I was a machinist in the shop for this same company. Since I have moved though I have found my jaw on the floor alot of the time with some of the questions I get. Maybe one day this will become a required reading for engineering students(wouldn't that be wonderful).

This I am sure will also be a very good opportunity for some of you to vent about some of the ridiculous callouts you have gotten over the years.

So let me get the ball rolling.

- If you are calling out a tapped hole in 1/4" plate or similar unless it needs to be bottoming make it a through hole.

- Unless a 6-32 tap is necessary go with 8-32

- No I cannot tap a 2-56 hole 2" deep

- Designing a part to use mill finish material? Then don't give me tolerances tighter than raw bar stock actually is if you don't want me facing it.

- A BIG ONE, if that dimension is not important then call it out as such but don't give me crap for how long it takes because you have it dimensioned like it's going into space.

No I cannot drill a .063 hole through a 3" piece of stainless on a lathe. At least not if you expect it to come out at all concentric on the other side.

Work for 5 years on the floor of a job shop.

Here's something Conrad Hoffman posted a while back.

Dear Valued Customer

Ahh this old tale. You guys aren't tired of this yet?

It might shock you guys to know that there are often more variables in a design than how easy it is to machine. I would say the majority of mechanical engineers may never have any interaction with a machine shop as part of their profession. Yet everyone here will demand that they have 10 years of machinist training as part of their degree.

Right now, you could find this same thread regarding welding, heat treating, plastic molding, stamping, plumbing, concrete, rigging, electrical, computer programming, et al.

That is sort of the point ewlsey. From an engineering perspective I feel it is irresponsible to call yourself a designer of something you know nothing about. Don't know about welding? Ask a welder before you make some ridiculous callout. That is my point. As for machining if you are a mechanical engineer and you design machined parts you should at least make some sort of effort to know the basics. I agree I think most guys on here are a bit too harsh in their expectations but there are other things that make me wonder if they could even operate a hand drill. In the CU engineering program there is ONE class called Engineering processes and design and is NOT required. BTW can you every really be tired of bitching about people who make your work harder for no reason?

HAHA thanks Charlie, exactly what I was looking for. I will be printing this and putting it up on the wall in the office.

-If there is features located around a round part, for example a keyway, some setscrew holes, hole pattern on a flange and a hexagon. Please mark on the drawing if the features don't have to be accurately located in relation to each other. It can make a big difference in amount of work depenging of the part.

I see there is three accuracies to be specified:
-Alignment not important (Features don't intersect and it doesn't matter in appearance of the part)
-Align by eyeballing (to make the part look better, or if the features may intersect in some angle)
-Align to common tolerances of the part, or specified more accurate angle tolerance

best tip for anybody inany profession, when you start a new place keep your mouth in a lower gear than your brain....listen to everyone then decide for yourself who's opinions to value. Many will be out to sink you but your best allies are humble pie and intelligent questions. Never tell anyone how good you are...never tell anyone how bad/ dumb they are.

Please leave big tolerances wherever you can, don't use tight tolerances unless they're absolutely required. Tighter the tolerance, the harder to make. The harder to make, the more cost in the part.

Leave allowable radii in corners of turned and/or milled shoulders, specify max and min if possible.

Don't spec out threaded holes deeper that 2xD, and never expect threads right to the bottom of a hole unless it's under 1.5xD (to allow thread milling.)

Understand angular tolerance and don't make it a default +/- 1/2º on really short ones (such as edge break chamfers.)

Do not specify a process on a part print, specify the part geometry only. (Example: Drill .201 x 1" deep, tap 1/4-20 UNC x.75" deep. Just state 1/4-20 UNC 2B x .75" deep.)

I could go on all day...as any of us could.

This guy I had made objects from drawings for seemed to have a real gift for designing things that were a PITA to make his way. Now,he has an "Industrial designer" who drives me nuts by always specifying tolerances in .0001" all the time. And she thinks she is hot shit,too. This,these days,are mostly drawings for tourist junk,too. She seems to have no idea what the hell she is asking me to do with those tolerances.

Neither of these people have a machining background.

I got pissed with Ford several years ago when the under frame of my car seat broke. The dumbass had made the thick sheet steel to be bent with a sharp 90º angle and it made a nice clean 45º break right across the corner. I noticed that this stupid problem had been fixed on the next year's Taurus,and a decent radius made instead.

Just curious why the 6-32 vs the 8-32??

or Vice Versa

Dear Machinist,

- Before you pick up the phone to educate me on dimensions, please look at the title block of the drawing... all of it.
- If the drawing looks incomplete, please see sheets #2,#3, etc. Then call me.
- If the drawing still looks incomplete, take it out of the plastic sleeve and UNFOLD IT!
- If you don't like that I have both metric and imperial geometry in my designs, guess what... neither do I!
- If you think my idea of how long it should take to make the part is moronic, maybe you should talk to the shop that did the prototypes (yeah that's right, I farmed them out while you were milking that production job like a dutch dairyman's daughter), wasn't a problem for them.

shall we stoke the fires a bit more?

Directly to gwilson's point about tolerances........

Pay attention to what the default tolerances are in your CAD program, and CHANGE the blinking setting if it is (as with AutoCad) a default of 0.0001" resolution. I suspect most complaints about tolerances come from that stupid default setting.....

However......... A few points about engineering and engineers, particularly NEW engineers.

1) SOME folks go into engineering from an interest in mechanical things/electronics, etc. Those folks usually DO enough that they understand, or at least need not be told more than once..... they generally are good practical engineers. (they are not necessarily less good at math, but it wasn't their driver for choosing a career)

OTHER people go into engineering because they are good at math, and they make particularly GOOD theorists, but often very bad practical engineers. *

2) Many engineering schools pay lip service to the matter of the practicalities of engineering, but in the end, they know that their job is to provide the "engineering" input to the student, he or she is expected to pick up the practicalities "on-the-job". The courses offered are "OK", but the institution is academic, and the professors are drawn from the ranks of the folks who were "good at math"...... not necessarily so good at the practicalities.

3) The whole "on-the-job training" deal is a lot less prevalent than it used to be. The older engineers were likely "downsized out" and cheaper young engineers hired in to replace them.

*
Example #1
I'm an electrical type generally, designing motor drives and switchmode power supplies, etc (and I am also the model shop machinist at work). The manufacturers put out "application notes" that show how to use their fancy new "chips" (integrated circuits), which are intended to give designers a good idea of how to use the parts. Usually they will give an example circuit and show just how that circuit uses all the features of the IC chip.
I had a lot of trouble with one "chip" a few years back... I went according to the app note, but the chip just didn't do what it was supposed to do.... So finally I called up the (rather large) company and eventually got an applications engineer who knew the chip. I asked what was going on, sent him the info, and discussed the application. He went away to look into the matter....
Came to find out, after the apps guy checked around for a while, that while the app note SAID that a particular pin in the IC worked a certain way, IN FACT IT DID NOT..... In actuality, the application note was totally WRONG about how it operated, and their example circuit was totally bogus, IT COULD NOT WORK! The theorist who wrote it up just didn't understand the chip....... and obviously had never actually BUILT the circuit that was published.....

Example #2
The boss hired an outside mechanical engineer to design a frame for a motor-generator test setup..... large no-bearing generator unit under test (we get just the rotor and stator, no end bells, bearings, or shaft), and a gearbox and motor hooked to it a the drive means (with a big VFD). The motor and gearbox weigh about 1200lb, and the generator weighs about 1600 lb (it is a slow-speed generator). Torque about 1600 Newton-meters normal, 7000 N-m peak during a fault.

This guy designed the structure as a welded flat frame made out of 3" x 0.125" wall square tubing, with one piece down each side, plus cross-members. Due to test and assembly issues, he designed the shaft and span between bearings to be extra-long.

Genius-boy also decided that 2.5" keyed TGP shafting was plenty good enough for a 48" span of shaft holding a 350 kilo rotor with about 50 kilos of magnets on the OD.... STRONG magnets..... This when the rotor is made for use with much larger tubing as a shaft.

We had to try it, but we ended up with a very much different design which we had to come up with to save the day after genius-boy's design failed to work due to the shaft and frame deflecting and the rotor basically jamming in the stator. Three electrical types, plus our mechanical engineering intern fixed the problem in 4 days....design, waterjetting, and welding included............... Genius boy took 4 weeks + for the non-working version..

MBG said:

Just curious why the 6-32 vs the 8-32??

Click to expand...

#6-32 the thread depth is relatively deep on the diameter (compared to 8-32, 10-32) so the cross section is thinner, weaker, more likely to break,

Hmm - then you guys should get rid of your inspection department. Don't need to be checked because everything you do is perfect. Presumably too, those newbie machinists that start at the bottom make no mistakes and are just as qualified as the experts?

But why then all the stories here about work falling out of lathes.............................

I know a place where GD&T, sections and 3D renderings have made hidden lines unnecessary on drawings. Exactly how that works has not been explained to me, but company policy seems to be no hidden lines (you know- those dashed ones---------). I also lament the loss of different weight lines. Modern CAD has given us the ability to make the best prints that have ever been made, but somehow I think skilled draftsmen with T-squares and plastic triangles sometimes did a better job of communicating what was needed.

JST said:

Example #2
The boss hired an outside mechanical engineer to design a frame for a motor-generator test setup..... large no-bearing generator unit under test (we get just the rotor and stator, no end bells, bearings, or shaft), and a gearbox and motor hooked to it a the drive means (with a big VFD). The motor and gearbox weigh about 1200lb, and the generator weighs about 1600 lb (it is a slow-speed generator). Torque about 1600 Newton-meters normal, 7000 N-m peak during a fault.

This guy designed the structure as a welded flat frame made out of 3" x 0.125" wall square tubing, with one piece down each side, plus cross-members. Due to test and assembly issues, he designed the shaft and span between bearings to be extra-long.

Genius-boy also decided that 2.5" keyed TGP shafting was plenty good enough for a 48" span of shaft holding a 350 kilo rotor with about 50 kilos of magnets on the OD.... STRONG magnets..... This when the rotor is made for use with much larger tubing as a shaft.

We had to try it, but we ended up with a very much different design which we had to come up with to save the day after genius-boy's design failed to work due to the shaft and frame deflecting and the rotor basically jamming in the stator. Three electrical types, plus our mechanical engineering intern fixed the problem in 4 days....design, waterjetting, and welding included............... Genius boy took 4 weeks + for the non-working version..

Click to expand...

I have a buddy who grew up on a farm that took one and a half years of engineering. He said there were geniuses in that class that could calculate anything. Figure out the math like it was nothing. However, when it came to doing actual designs, they were lost. Couldn't visualize things, like which way a chain drive would turn, etc.

Hi

philmeik said:

.... Since I have moved though I have found my jaw on the floor alot of the time with some of the questions I get....

Click to expand...

The fact that your engineers are asking questions demonstrates their willingness to take advantage of a new source of information (you) and to learn.
Dumb people think they know everything. Smart people know how little they know.

Are the questions from smart engineers who know that they don't know stuff?

FYI : For my first engineering degree, the University required 400 hours practical time to be completed. I spent my time in a training workshop and a ship repair facility. I really enjoyed that time but over a 25 year career I haven't used that experience in a professional capacity. Only a minority of engineers have any direct or indirect contact with a machine workshop. My University did not offer any courses on manufacturing methods. The lesson for young engineers is to ask questions to learn more and to avoid mistakes.


Dazz

philmeik said:

Ok, let's face it Engineers make some stupid decisions sometimes. So how about we give some advice to our friends on the other side of the print. Some of you may know that I have recently moved into a mechanical engineering position and I can tell you that it is a blessing to the engineering department that I am here. Since I have moved though I have found my jaw on the floor alot of the time with some of the questions I get.

Click to expand...

Sounds like your company needs to hire an actual engineer and not give away titles. I hope for your career's sake you learn a bit of humility and how to ASK questions, not answer them.

FYI - EVERYBODY makes stupid decisions upon occasion. Some of us just have a lot less time to waste and better things to do than run our mouths.

philmeik said:

Don't know about welding? Ask a welder before you make some ridiculous callout.

Click to expand...

No thanks. That is how piss poor designs start - by referring to an "expert" that is anything but. Given the choice of asking a welder who likely knows a few processes vs asking a weld engineer, I would ask the weld engineer who specializes in their design and should have a much broader knowledge base.