What's new
What's new

The purpose of drawings

The real Leigh

Diamond
Joined
Nov 23, 2005
Location
Maryland
Given that the Asymmetric Tolerance thread has prompted a heated debate, much of which addresses the purpose of drawings,
I thought I would start a new thread and provoke a new and more focused argument. :D

The purpose of a drawing is to document a design. It is NOT to make the machinist's like easier.

How a particular part is made can vary over time, perhaps using processes that differ substantially from those originally envisioned.
The process should not matter. As long as the resulting part meets the specs on the drawing (and therefor works in the product), all is well.

Realize that there are many other people in the manufacturing chain using the drawings, not just machinists.

- Leigh
 
The purpose of drawings is to communicate the design. A good design takes into account functional requirements, manufacturing and inspection methods, and more. If it is neat, clean, and easily understood by all parties then that counts too.

It should not be engineer vs. machinist although it may seem that way sometimes. Good engineers understand manufacturing and good machinists understand or at least respect engineering.
 
Quote: "The purpose of a drawing is to document a design. It is NOT to make the machinist's like easier."

+1. A drawing or print has one purpose in life, and that is to convey information. All I would add is that just like anything else, it's only as good as a person wants it to be....the drawing will only show what the draftsman, designer or engineer places upon it. I was going round & round with a designer about a month ago over a new print with nothing but an isometric view and a footnote **See 3D model for details** :nutter:
 
As I tell our designers, we can make anything you want. The question is how much do you want to spend.

The designer shares in the responsibility of what a design costs to make and which in turn helps determine the success of the design.

Bruce
 
A drawing is a tool that a person that has no ability to make a part uses to obtain what he thinks he needs.

What he really needs is a part that works.

The part needs to work wether the drawing is right or not.

One example is a part that was drawn in the '70s the detail of one part (of an assembly) is a mirror image of what will work.

!st article has to be made backwards but in order to assemble this you need a mirror of the part.

Has worked to my advantage for years.
 
Where you say it is not to make the machinists life easier I would say it is to make the machinists life harder. Why give a .005" tolerance when just by adding another 0 we all struggle a bit more. Do not go into the designers office to point out that that last 0 is unnecessary in this instance because he has a lot more 0's which have a use by date on them
You can blame 90% of this on the CAD programs.

They set a default precision (X.xxx or X.xxxx) for dimensions on the print, and they all come out the same way, even when some don't need to be that precise.

That never (seldom) happened when you had human beings making drawings.

Engineers are taught the importance of using proper tolerances when defining parts. That information is frequently lost in the drafting process.

- Leigh
 
A drawing is a tool that a person that has no ability to make a part uses to obtain what he thinks he needs.
This is an example of the attitude that bothers me most, and is the reason I started this thread.

Engineers are not paid to know how to make parts, although they do get classes on manufacturing methods, and tutelage in that area in the early years in the profession.

Engineers are paid to make designs that work under all conditions, whatever those might be.

Machinists have no basis on which to question the specs on a drawing.
If a spec seems unreasonably tight,or otherwise difficult to make, by all means question it.
But you're in no position to re-design it. You do not have access to all the information needed to do so.

- Leigh
 
One thing I don't understand is how many people here and elsewhere can assume the implied tolerance based on the number of digits after the decimal point. I know of no industry standard that says X.XXX = +/- .001. Yet everyone seems to assume that.

Personally, I like to list all dimensions to 4 places and give specific tolerances were needed. If I was a making the part, I would rather see:

1.9375 +/-.010 than 1.94 with a generic title block tolerance.

In my example, the dimension is listed .0025 larger than nominal. The machinist has to guess that I meant 1-15/16. Not a good method IMO.
 
So than when the engineer is so arogant that he refuses to accept the fact that he is wrong and refuses to change the drawing what is wrong with me fixing his mistake.(somebody has to or the part will never be built)
 
So than when the engineer is so arogant that he refuses to accept the fact that he is wrong and refuses to change the drawing what is wrong with me fixing his mistake.(somebody has to or the part will never be built)

That is totally wrong. You don't know what is right, or what is wrong. How dare you act like an engineer when you have a nat's perspective on what that part needs to do. Drawings/revs are often referenced in legal contracts, producing something that falls outside that drawing opens you up to having it get returned, liability for failure in the field, and so on. Make the fucking part on the print. The engineer will learn really quick that he'll hang himself with bad prints, and that'll take the arrogance right out of 'em.

Now, if the print is shit, no bid or price accordingly, and feed that information back to the engineer. I'm an engineer at a company that makes parts, so I deal with it from both sides... but the print is what I'm legally obligated to adhere to, so right or wrong, that's the part the customer gets.

Edit: I realize that may come like I'm the arrogant engineer you're dealing with, but I'm more than willing to discuss whatever concerns the machinist might have with my print. If I made a mistake, I'm more than happy to fix it, but half the time, I really do need the thing made the way I have it called out, and I try to articulate why that is specified the way it is to the machinist... for the record.
 
If an Engineer does not need to know how to make parts in order to design parts how does he know if the part can be made. I wonder how he gets on with design for manufacture.
Designs go through multiple levels of engineering review before the drawings are ever released for fabrication.
If a junior engineer gets off on a tangent the problem will usually be corrected immediately by his supervisor, and definitely should be caught in review.

Would it be rude of me to ask how old you are and/or what you do for a living? My money says you are not on the shop floor, I could be wrong
Yes, it would be rude, but I'll answer it anyway.
I'm retired now. I started my professional life as a journeyman tool & die maker.
After Nam I went back to school and got a degree in engineering, and finished my career in that field, although still doing prototype machine work.

My last full-fime machinist job was about ten years ago, after I had semi-retired, running a Bridgeport in a job shop, doing whatever came in.
I was their go-to miller for tight-tolerance jobs.
I've worked in tight-tolerance (aerospace) environments and on commercial projects. I've always done small-quantity work; never involved in production.

I still have a small shop (Bridgeport/Monarch 10ee/Nichols horizontal) and do one-off prototype and repair parts.
I also have access to a huge shop (surface grinder with a 24-foot bed, etc).

So I'm sensitive to the issues on both sides of the question under discussion.

- Leigh
 
That is totally wrong. You don't know what is right, or what is wrong. How dare you act like an engineer when you have a nat's perspective on what that part needs to do. Drawings/revs are often referenced in legal contracts, producing something that falls outside that drawing opens you up to having it get returned, liability for failure in the field, and so on. Make the fucking part on the print. The engineer will learn really quick that he'll hang himself with bad prints, and that'll take the arrogance right out of 'em.

Now, if the print is shit, no bid or price accordingly, and feed that information back to the engineer. I'm an engineer at a company that makes parts, so I deal with it from both sides... but the print is what I'm legally obligated to adhere to, so right or wrong, that's the part the customer gets.

Edit: I realize that may come like I'm the arrogant engineer you're dealing with, but I'm more than willing to discuss whatever concerns the machinist might have with my print. If I made a mistake, I'm more than happy to fix it, but half the time, I really do need the thing made the way I have it called out, and I try to articulate why that is specified the way it is to the machinist... for the record.

So than I have been totally wrong for 30 years?


And by the way reality always trumps an engineer
 
One thing I don't understand is how many people here and elsewhere can assume the implied tolerance based on the number of digits after the decimal point. I know of no industry standard that says X.XXX = +/- .001. Yet everyone seems to assume that.
There is a spec (MIL I think) that calls out the precision for various numbers of digits. It's taught in drafting class.
Some vellums come with those pre-printed in the title block.

Personally, I like to list all dimensions to 4 places and give specific tolerances were needed. If I was a making the part, I would rather see:
1.9375 +/-.010 than 1.94 with a generic title block tolerance.
In my example, the dimension is listed .0025 larger than nominal. The machinist has to guess that I meant 1-15/16. Not a good method IMO.
I agree, and do the same thing.

I once got caught on a fractional dimension. I needed 15/16 (0.9375), and clearly labeled the drawing as such. I got 1 5/16 (1.3125). Oh well.

- Leigh
 
So than when the engineer is so arogant that he refuses to accept the fact that he is wrong and refuses to change the drawing what is wrong with me fixing his mistake.(somebody has to or the part will never be built)
That's a management issue, and should be resolved at that level.

You have no business "fixing" the drawing, because you don't know where the problem lies.
It might be the result of transferring a dimension from another drawing, and that one is wrong.
Very few drawings exist in isolation. They're usually one of a group, and errors must be corrected at the group level.

-Leigh
 
If you're making the part in a way that is not allowed on the print, then yes... you're wrong. period. Now, perhaps what the engineer REALLY wanted is the part you're infact making... if that's the case, send him a marked up print, tell him this is the part you're actually making, and we need to reconcile the print to what you're actually making. Refuse to work thier parts until they do.

Or, make the part the way his print specifies, charge accordingly, and refuse returns on the basis that you made the part per his print. Anything else, you're holding a live grenade with the pin out.
 
That's a management issue, and should be resolved at that level.

You have no business "fixing" the drawing, because you don't know where the problem lies.
It might be the result of transferring a dimension from another drawing, and that one is wrong.
Very few drawings exist in isolation. They're usually one of a group, and errors must be corrected at the group level.

-Leigh

I'd say the machinist has no business making any part but what the drawing says... but the machinist isn't out of line to say "hey, your drawing tells me to do this, but it really seems like you want to do that... if you want that, change the drawing!"
 
I was their go-to miller for tight-tolerance jobs.
I've worked in tight-tolerance (aerospace) environments and on commercial projects. I've always done small-quantity work; never involved in production.

Tight tolerance + Bridgeport = :confused:


Engineering is an iterative process. Any engineer that thinks he can kick his drawings down to the shop never hear a complaint is dreaming.

Any machinist that thinks the engineer is going to open all of the tolerances to +/-.050 and every part will be a 3x4x6 block of 6061 easily held in a Kurt vise is also dreaming.

The back and forth interaction between production and design is how good products are made. Neither side is more right. A perfect design that cannot be built is not a perfect design. However, many engineers know the difference between a genuine concern, and a lazy machine shop.
 
The purpose of a drawing is to document a design. - Leigh

Slight disagreement. In my industry, drawings are only generated to convey QA information. All parts are machined to 3d datasets. And we've begun to design parts with no drawings whatsoever.

20 years ago, engineers were much more machine sympathetic. But our suppliers now tell us their high speed CNC machines can easily hold +/-.010" on any feature and can cut flanges at any angle to the base. Lately, even hemstitiching/kellering is no longer avoided. I get the sense there is no cost benefit to making a part that is easy to machine. The machine shops charge us the same high price regardless. Often times, they bid on parts before seeing them and almost always before any tolerance data is conveyed.
 
always remember that in any strong business the profit comes from what is hard for you and impossible for others.

this could mean either designing the part better or making the part better.
 








 
Back
Top