confusion over basic dimensions

[hobbs101]

I think I'm not understanding something fundamental about basic dimensions. See attached drawing.

I've put in basic dimensions for the holes. I'm happy with that.

I've used basic dimns for the slots. I realise that this probably isn't correct, but I'm treating the slots as 2 holes that are 'joined' together. By rights I should probably use the surface profile callout. Would be interested to hear peoples thoughts on this.

Should I use basic dimensions for the square cutouts?

Why can't I use basic dimensions for the outer profile of the part?

I think what I'm not understanding is the 'features of size' rule. I don't really understand why I use basic dimensions for the position of a hole, but not for the outer profile of the part.

Any help much appreciated. Thanks.

 

[CheckerHater]

You can use basic dimensions for ANYTHING.

Something special about basic dimensions is that they only represent "half" of the specification. They have to be combined with feature control frames to mean something, for example position of the hole.

You can apply profile control to the outside edge of the part and use basic dimensions to define it.

You have to learn more about different controls, used with and/or without basic dimensions to get better understanding of the principle.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[KENAT]

Use of position with slots is entirely appropriate and correct, how you've done it though is not - though I'm more familiar with ASME than ISO.

Put the basic dimensions to the center of the slot, then dimension length & width of slot. Apply position tolerances separately to the width and then length of the slot as these are the 'features of size'. Use 'boundary' if appropriate. There have been previous posts with more info.

Posting guidelines faq731-376

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

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

 

[CheckerHater]

And yes, you can use basic dimensions with square/rectangular features, but you will have to provide 2 FCFs to control position of the feature in 2 directions.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[hobbs101]

Great information. Thanks.

I will rework the drawing and post it soon. In the meantime, if basic tolerances can be used for anything, then when should I be using them and when shouldn't I? I always use them for holes, but hardly ever for external profiles. I'm sure it's not a case of only using them for critical tolerances. What are the rules?

 

[hobbs101]

I've reworked the drawing. Any comments much appreciated. Thanks.

 

[CheckerHater]

When you apply position to square features, drop the diameter symbol and add second control in perpendicular direction just like you did with the slots.

When it comes to "rules", I have to disappoint you: there are no rules. You're the one who must make the decision - that's what you are paid for. Your job is to to produce drawing that will result in working part. Nobody on this forum knows better how your part functions - only you.



"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[hobbs101]

Thanks CheckerHater. I realised I might be getting it wrong with the square features, but what I'm trying to do is to control their position to within a circular tolerance. Can I not do that or do I have to put the second FCF in as you suggest? That of course will control it to within a square tolerance.

You are right, I'm paid to make decisions like that. But the reason I'm on this forum is to improve my knowledge to better make those decisions! It's not necessarily a case of producing a working part. For instance, see the attached. Same part, different dimensioning techniques, same result!! If it's a subtly different result then please tell me, that will go some way to answering my original question.

 

[CheckerHater]

Sorry, I couldn't see the attachment. And yes, it's all about working part. You are the one who knows the function of the part - which features are engaged with other parts of assembly and require precise fit and which ones are less relevant.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[hobbs101]

Here's the attachment.

 

[CheckerHater]

Believe it or not, but results are NOT the same.

The view to the left tells me that point-to-point local measurements of the part should stay within the tolerance, but it doesn't say anything about form or orientation of the edges of the part.
Also dimension tolerances do not control angles.

Essentially without geometric controls your part looks like this:

http://www.hn-metrology.com/gpscauses.htm

BTW, there is more interesting stuff on that website


"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[hobbs101]

Ah, I thought as much!!

So, why don't I use geometric tolerances for ALL dimensions? If dimension tolerances don't control angles how has anything ever been made correctly? Most drawings I've seen (done by colleagues in various jobs) have dimensional tolerances for overall sizes and profiles, with geometric tolerances for hole positions and control of certain surfaces.

Is it a case of, if it's critical use GD&T, if not then hope for the best with dimension tolerances?

The problem I have is that I don't fundamentally understand geometric tolerances. This is probably because I've never been formally taught it. I've had to pick it up as I've gone along. I've bought and (part) read books on it, but still it hasn't clicked. And to be honest, I don't know anyone amongst all the engineers/designers that I've worked with that fully understand it either.

 

[3DDave]

Geometric tolerances are easy - there's a perfect form, location, and/or orientation from which the noted allowable deviation is allowed to take place. It's like a paint by numbers except the fabricators is graded on how closely they can stay to the lines.

Tolerances for right angles that are not applied to a geometric tolerance feature control frame datum reference framework are supposed to be controlled by a note or titleblock reference. Mostly they are ignored because most manufacturing equipment is better than most tolerances assigned to the function. By Y14.5 rules, an inspection on a rectangular block of material should come back with 12 angles measured, one for each edge, but I've never seen them mentioned on any inspection report. I saw several projects for TACOM where there were no tolerances given; our engineers, the fabricators, the inspectors, and TACOM sign-off all missed the lack of angle tolerances. When they previously were there, they were much larger than processing equipment would typically produce.

You can buy a copy of Y14.5 -

https://www.asme.org/products/codes-standards/y145-2009-dimensioning-and-tolerancing

for $185 for the '2009 version. For other training material, I'd lean toward Fischer and Krulikowski.

The last example provided uses profile of a line, which means that each section though the depth of the part is constrained, but not each section to the next. It's like a stack of playing cards, each the same, but each one is free to move and misalign in location and orientation to the next. It doesn't show that way in the standard's examples because they included enough datum references to fix the orientation and partially fix the location.

 

[hobbs101]

I have access to Y14.5 and I've got one of the Krulikowski books. I'll take some time to read them properly.

I've still not had an answer to my question - So, why don't I use geometric tolerances for ALL dimensions?

I'm doing another drawing now. It's essentially a picture frame - a large square sheet with a large square cutout. Now, do I treat the cutout in the same way that I've dimensioned my 9.5x9.5 cutout in my last attachment? Or do I dimension it using toleranced dimensions?

Another example. On page 111 of ASME Y14.5-2009, the example drawing has the overall dimensions as toleranced dimensions, not geometric tolerances. Why? From what CheckerHater said this part could be completely skewed. Why isn't there perpendicularity control between datums A, B and C?

I suspect I'm not understanding the concept of 'Feature of Size'.

 

[3DDave]

You need to read the basic rules section that covers why parts are not skewed and decide if that provides enough control. This was also covered in my previous answer.

A part can be entirely controlled with feature control frames, if that's what you find is best. Tolerances applied to dimensions fall under the definition of geometric tolerances; no difference there.

 

[Tunalover]

The OP is not limited to applying positional tolerances to the slot in two separate directions. He can control the size using a surface profile tolerance (no datums reqd) and the position using a non-cylindrical positional tolerance. Treat the slot as an irregular feature of size (which it is) and do it like it says in Figure 8-24 and the body of the standard that refers to that figure.



Tunalover

 

[patdh1028]

CheckerHater's comment about uncontrolled angles was misleading. Angles of surfaces in a feature of size are controlled within the boundaries of a size tolerance zone per Rule #1, they can't just skew freely. The animation on that page isn't necessarily incorrect, but it's absolutely possible to fully define that part using only size tolerances.

 

[CheckerHater]

Padth,

First, there is no Rule 1 in ISO

Second, no, it's impossible to fully define part using only size dimensions/tolerances.

Do your homework:

http://files.engineering.com/getfil...-841a-73c9b1aa69e6&file=ISO_14405-2_A.1-2.png

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[3DDave]

I think there should be a separate forum for ISO than for ASME. The rules are enough different on the fundamentals that this is a cause of a lot of misunderstanding. The OP calling ISO GD&T is confusing as GD&T is an Americanism for FCF controls.

 

[CheckerHater]

And separate forum for users of Metric system (sorry, couldn't resist)

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future