Any Rules on Implied Symmetry?

[curiousmechanical]

Hello,

This has been bothering me for a while.

Please see the attached .pdf for reference.

I come across a lot of old drawings at work that use a simple dimensioning style that takes advantage of symmetry. I also see this style used in many catalogs.

I like this style because it gives the views a clean look (also fewer dimensions) and ensures that mating parts will match concentrically (which I sometimes require).

However, I have not seen any mention of this dimensioning style in drafting textbooks or ASME Y14.5M-1994 (I only have the old version). Are there any rules about this? Is it considered good or poor practice?

If it is legal, how much off-center can the feature be (say the two holes in my example)? How do you tolerance this sort of thing?

If it is not legal, how do I dimension parts that I want to be concentric? For example, we order a lot of flame cut plates and they come in oversized most of the time. If I dimension a part from the corner, the bolt pattern will be way off-center. This could result in a sloppy appearance (material hanging off two sides on a mated joint).

I've seen others dimension to centerlines, but that seems to be illegal.

"NOTE: The following shall not be used as a dimension line: a center line,..." - ASME Y14.5M-1994

I am really looking forward to your responses.

This will solve a lot of debates at work!

Thank you.

 

[curiousmechanical]

Hi Guys,

Sorry for the delay! I thought this thread lost its steam. I'm glad to see that it's back.

Thank you all for the additional feedback.

KENAT

Unfortunately, I have not had the time to hit the books on GDT, so my understanding is still limited (my time has been completely consumed by Mathcad). Although, thanks to your post, I think I now get the gist of it.

Please see the attached. I would be curious to see the correct way to dimension example: 1 (which uses my current cheap and dirty method). Example: 2 is my poor attempt.

Also, if you have the time, I have two general questions for you experienced GDT guys.

No.1

Since graduation I have been in the world of custom machinery (first at a machine shop and now an OEM). So far, I don't see many people use GDT and I suspect the ones who sometimes do don't use it correctly. So my question is, of the engineers who do make drawings, what percentage of them do you think use GDT? I am trying to get a view out of my small world. I am also trying to find out if this is something I NEED to know. I am under the impression that this stuff is for industries with high-volume production (i.e. automobiles, consumer goods, military, etc.). Is that the type of environment that most engineers work in?

No.2

At my company the engineers make the drawings. However, I have heard that at larger companies, designers make the drawings. So, who needs to know the details of GDT; engineers, designers, or both?

Thanks guys!

 

[ptruitt]

Without knowing the function of the part, I'd say 'keep it simple'. See the attached drawing.

Hopefully, the next generation entering the workplace won't even know plus/minus. Plus/minus is ambiguous. GD&T is concise. If 'plus/minus' works for you, the short-term gain might be sufficient and a reasonable position to take. But we live in an increasingly complex society and, where it is headed, GD&T is absolutely, unequivocally mandatory if this high-level of technology is to be sustained.

 

[KENAT]

It's gonna come back to the functional intent of the design.

If it needs to be 'centered' on the the part, then you really have to use GD&T, I can't think of a robust way to even try and do it with +-. Even if you dimension to a Centerline, you have to locate that centerline from an edge and you're back to square one.

Your Example 1 is incomplete, you only give the size not the relative location, there is no information on how centered the (assumed) cut out is.

Your example 2 is not much better, you still haven't set up centering datums. You've picked the edges as datums, not the 'width' of the part.

You really need more help that it's easy to give in this type of situation. Without knowing function etc. I'm hesitant to say too much but if I wanted a cutout centered on a physical part, I might do it something like the linked. I haven't attempted to exactly match your tolerances and assume there's a block tol or similar somewhere.

http://files.engineering.com/downlo...54e8-46ba-a722-19da302f5223&file=SYMMETRY.tif

If it's ok for it to be located from one corner, then Ptruitt's will kind of work, though it's not clear which corner is which on the physical part which can cause problems. It probably wouldn't be my first choice on how to approach it though.

Q1. I'm in fairly low volume work but we use GD&T as it allows us to actually capture functional intent while keeping tolerances as loose as possible. Not everyone is very good at it though, and I doubt some of our vendors really understand it - we've actually used GD&T on parts so we could better capture the function while actually relaxing tolerances and yet they came back with a higher price! Some shops seem to increase the price based on how many FCF they see - probably because a lot of designers only use GD&T for 'tight tolerances' or 'where it really matters' etc. which is a never ending circle of confusion and lost opportunities.

Q2. We only have one 'Designer' left and he gets given tasks pretty much the same way as 'Engineers'. It was similar at my last place. An Engineer not doing his own drafting doesn't need to be expert in GD&T, however, if applicable he should know enough to be sure the Drafter captured the design intent/requirements properly.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[curiousmechanical]

Thanks for the feedback guys.

Based on your comments, I will keep GDT on my to-do-list. It may not be required at my current company, but it does seem more common than I thought. I have a long career ahead of me and I don't want to become prematurely obsolete.

As for my example, it was just something I whipped up for discussion (no real function). I was just trying to say that what I usually do is example: 1 (I understand this is bad practice). Example: 2 was just my best guess at the proper way.

KENAT

You are correct; I am not yet capable of having an educated discussion on this matter. However, I will hold onto you example and one day I hope to understand it (just a matter of making the time).

Here is the basic gist of where I am coming from:

Most of the parts we make start from a flame cut plate. The tolerances on the profiles of these plates need to be pretty loose (+/- 1/16"). Most of the time, I don't really care about outside dimensions of the part. However, I want things to bolt up on center. Some of these plates come in pretty oversized (ex. + 1/4"). I have dimensioned parts from a corner and the hole pattern (or cutout) comes out looking clearly off-center. Even worse, when you bolt two pieces together, the edges don't line up and it just looks sloppy (touching on two edges and a gap/overhang on the other two). It seems like when I do the symmetric tolerancing (the cheap and dirty style), the machinist will indicate to the center of the piece and makes that his (0, 0). The parts come out how I wanted, but in the back of my mind, I feel like a looser because I know that my dimensioning practice is bad. When I ask around the office, it seems like most people solve this dilemma by giving a reference dimension to the centerline (which I think looks just as bad). I am happy to find out that GDT is the only correct answer. Now I just need to learn it!

 

[KENAT]

So, while it's something I tend to avoid because some of the implications can be a bit complex/confusing, if the pattern of holes is your real functional datum, then it may be that the pattern should be your datum.

ASME Y14.5M-1994 Section 4.5.8 and figure 4-22 goes over this, but for anything more than a 2 hole pattern it can get confusing so I rarely use it, especially when you use the MMC of the pattern in subsequent call outs.

Instead I may choose one of the holes as my secondary datum, and another hole as my tertiary, more like figure 4-8. It doesn't quite as closely follow true function but may be close enough.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[curiousmechanical]

KENAT

Thank you for that additional info!

I really appreciate your help.