Smart questions
Smart answers
Smart people
INTELLIGENT WORK FORUMS
FOR ENGINEERING PROFESSIONALS

Member Login




Remember Me
Forgot Password?
Join Us!

Come Join Us!

Are you an
Engineering professional?
Join Eng-Tips now!
  • Talk With Other Members
  • Be Notified Of Responses
    To Your Posts
  • Keyword Search
  • One-Click Access To Your
    Favorite Forums
  • Automated Signatures
    On Your Posts
  • Best Of All, It's Free!

Join Eng-Tips
*Eng-Tips's functionality depends on members receiving e-mail. By joining you are opting in to receive e-mail.

Donate Today!

Do you enjoy these
technical forums?
Donate Today! Click Here

Posting Guidelines

Promoting, selling, recruiting, coursework and thesis posting is forbidden.
Jobs from Indeed

Link To This Forum!

Partner Button
Add Stickiness To Your Site By Linking To This Professionally Managed Technical Forum.
Just copy and paste the
code below into your site.

tgwow (Mechanical) (OP)
27 Mar 08 17:08
ASME Y14.5m, 1994 Section 2.15.1 Radius Tolerance states:

"A radius symbol R creates a zone defined by two arcs (the minimum and maximum radii).  The part surface must lie within this zone...."

Section 2.15.2 Controlled Radius Tolerance states:

"A controlled radius symbol CR creates a tolerance zone defined by two arcs ( the minimum and maximum radii ) that are tangent to the adjacent surfaces..."

How do you connect the two arcs? Because no matter how you do it, there's a zone that is zero wide.  Wouldn't you need to connect the two arcs with straight edges?  And if you did, wouldn't that be the tolerance zone of CR?

What does the tolerance zone look like for R2+-2?
Helpful Member!  ctopher (Mechanical)
27 Mar 08 17:31
Same definition, but:
R; Flats and reversals allowed
CR; Flats and reversals not allowed

Chris
SolidWorks/PDMWorks 08 2.0
AutoCAD 06/08
ctopher's home (updated 10-07-07)

Helpful Member!  powerhound (Mechanical)
27 Mar 08 23:00
The R2 +/-2 is a silly callout because you don't have to have a radius at all and the part will be to print. It also allows all kinds of crazy geometry on that corner and still you would have a good part. I wouldn't preoccupy myself with this issue because the dimension, as specified, doesn't really make sense. If you give your customer a goofy looking corner that is still to print they are just going to reject it and you'll make the part again to keep him happy. Slap a decent looking radius on that corner and send him on his way. If you want to push the envelope, leave a sharp corner and see what he says...haha.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
28 Mar 08 10:12
ctopher,

How could the tol zones be the same when CR requires tangency and R does not?  The tol zones in the figures in ASME for R and CR appear the same but are they really the same?  Try drawing it for the example of R2.4 +/- .3 from ASME.  Because the arcs aren't require to be tangent, the zone can move anywhere between the tangency point of the minimum radius and the sharp and you still satisfy the definition of R.  (See attached pdf)


powerhound,

This may seem silly but it is fundamental in understanding what R means.  So, is it possible to draw the tol zone of R2 +/-2 using the ASME definition?
fcsuper (Mechanical)
28 Mar 08 11:24
R does have less restrictions than CR.  R used to mean some of what CR now means, but CR was created (I believe) to address the issue of tangency and fairness of curve.  So, yes, the start of a R can have a variety of possibilities, where CR has only one.  A minimum, R can be used to break a sharp edge.  A minimum, CR can be used to establish an actual radius form from the two surfaces.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

powerhound (Mechanical)
28 Mar 08 13:56
tgwow,
  Regarding your drawing; the arcs are required to be tangent per 1.8.5 and 1.8.6 of the standard but not tangent to each other, they must be tangent to the corner that they are supposed to fillet. You would draw the minimum arc tangent to the edges of the corner and then the maximum arc tangent to the edges of the corner. The bottom view of your drawing is correct.

Regarding the "silliness" comment; what I meant was that if a sharp corner is acceptable then why even put a radius there? It's just one more thing that QC will have to inspect. The tolerance zone for a 2 +/-2 will look like a triangle with one of the legs as a 4 radius.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
28 Mar 08 14:45
powerhound,

Section 1.8.5 and 1.8.6 requires that the actual part geometry be tangent to the sides but says nothing about the tolerance zone.  

Therefore, why would you draw the min and max arcs tangent for R?  Additionally, the definition of the R tolerance zone does not state inclusion of the tangent sides as boundaries.

By definition, wouldn't the tol zone for R2 +/-2 be just the max arc (since the min arc of zero just disappears) located somewhere in the corner (not required to be tangent)?
KENAT (Mechanical)
28 Mar 08 14:59
If I had it on the drawing I'd have thought "R4 MAX" would be a more logical callout.

Depending on the application the 'silliness' point about why even break the edge may have merit.

KENAT, probably the least qualified checker you'll ever meet...

tgwow (Mechanical) (OP)
28 Mar 08 15:10
Kenat,

We're kind of going on a tangent but for parts which are molded I've seen requests from suppliers who would want the CAD model modeled at nominal which would mean in order to "show" the dimension from the model on the drawing for "R4 MAX" and keep it parametric, "R2 +/-2" would be prefered over "R4 MAX".

The alternative is to model at R2 then fudge a dimension "R4 MAX" on the drawing.
KENAT (Mechanical)
28 Mar 08 15:46
Yeah a bit off tangent, more related to thread1103-212575: unilateral tolerancing but is an example of where wanting to go straight from the model makes it difficult to use certain standard drawing/tolerance conventions.

I'm always tempted to tell those suppliers that if the drawing is the master then the model is for reference only and maybe they'll actually need to do some work and go in and modify things.  

The trade off between the possible extra cost of this compared to the impact of worrying about the kind of issues modelling strictly nominal (as in mid range with unilateral tolerances) cam cause need to be weighed.

KENAT, probably the least qualified checker you'll ever meet...

Helpful Member!  dgallup (Automotive)
28 Mar 08 16:33
I mentioned this in the unilateral tolerancing thread but it seemed to go over everyones head.  With Pro/E & I assume other CAD programs you can have an R4 max. dimension (fully parametric model driving dimension) which is set to regenerate the geometry at the middle value (in this case 2) so the drawing calls out R4 max. and the model measures R2.  The drafters, design engineers, manufacturing engineers, CAM jockies & mold makers are all happy.  This is not a fudged in dimension.  

The same technique can be used for asymmetric tolerances like shaft & clearance holes where the tolerances are both positive or negative.  The dimension can read 1.000+.005/+.001 and the model will measure 1.003.  You can also set the dimensions to regenerate at the maximum or minimum value to do tolerance stackups. The state of the dimension (max, nominal, middle or min) is indicated by color.
tgwow (Mechanical) (OP)
28 Mar 08 16:51
Thanks dgallup, that's a nice tip!  I tried it and it worked.  Will need to train everyone so they know what's going on though.
powerhound (Mechanical)
28 Mar 08 16:57
Fig 2-18 shows an obvious separation of the tangencies of the 2 arcs that make up the tolerance zone. Although the points of tangency are not specified, an extension of principle based on 1.8.5 and 1.8.6 is totally appropriate in this case. There are lots of implications that can be made that are not specifically stated in the standard and this is one of them. The tangency rule applies not only to actual features but to tolerance zones as well.

There is really nothing more I can add to clarify this any further. I hope this is good enough.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
28 Mar 08 18:06
powerhound,

Thank you  for your responses.  

Extending the principle based on 1.8.5 and 1.8.6 does seem appropriate and reasonable as an implication.  However, I still think my original premise is still true due to the deficient definition in the standard.  The tolerance zone for R is ambiguous as defined in the standard.  Therefore, the only practical use of R that I can see is to indicate a break edge.
fcsuper (Mechanical)
28 Mar 08 19:37
tgwow,

Yeah, I can see other uses for R as well, but one of the bigger uses is to break edges.

I've noticed that a lot of people do not understand radius symbols.  I had one boss years ago who questioned me every time I used SR (spherical radius) per the standard.  The conversation used to go like this:

Boss:"You have a typo here on your drawing."
Me:"That's not a typo.  It is a spherical radius."
Boss:"What's a spherical radius?"
Me:"I'm dimensioning this spherical radius with the spherical radius symbol."
Boss:"I've never seen that before" ::looks at me like I'm making things up::
Me:"It's the standard since 1994."

We had this conversation twice for separate revisions the same drawing.  Then one day, we was changing the drawing for some other reason and then removed the S from SR on the SR callout.

I didn't throw it in his face.  I just simply added the S back on the next revision.  He stopped bugging me about it after that.  However, this likely drew out so long because we didn't have a copy of ASME Y14.5M in house at the time.  (as this for hijacking a thread? smile  )

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

KENAT (Mechanical)
31 Mar 08 12:02
dgallup, I recall you bringing it up but I don’t think all CAD systems can do it quite like that.  I don't think ours can although I haven't tried to do it for quite a while.  However does seem a good idea.

As to “What does the tolerance zone look like for R2+-2?” based on 2-18 isn’t it something like my attachment, or am I missing something?

KENAT, probably the least qualified checker you'll ever meet...

powerhound (Mechanical)
31 Mar 08 12:41
YES!!!! This is exactly what I was trying to describe when I said "The tolerance zone for a 2 +/-2 will look like a triangle with one of the legs as a 4 radius." . The contour of the corner can fall anywhere within this zone and still be good if specified as R. A specification of CR will fix this. Whatever the radius is, it has to be smooth with no flat or reversals.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
31 Mar 08 12:51
Kenat,

The problem with your sketch is that the definition doesn't say use MIN R, tangent lines, and MAX. It just says to use MIN R and MAX R.  Additionally, the MAX R arc doesn't have to be tangent to the sides so you can move the MAX R arc towards the corner and tilt it.

Even if we make the assumption that you can use the tangent sides, the MAX R arc doesn't have to be tangent to the sides so you can move the MAX R arc all over the place as long as it intersects the sides.  Therefore, the tolerance zone for R is not well defined at all.

KENAT (Mechanical)
31 Mar 08 12:58
tgwow, I'd say 1.8.5 is in effect, no doubt about it and as such my sketch which matches fig 2-18 is correct.

KENAT, probably the least qualified checker you'll ever meet...

powerhound (Mechanical)
31 Mar 08 13:33
KENAT,
 Agreed. Arcs are to be tangent. That is clearly stated in the standard. See Fig. 1-29 which is referenced to in 1.8.5.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
31 Mar 08 13:48
If you examine figure 2-18 carefully, there is no flat (tangent sides) where there should be.  It's quite evident that you can see the jagged edges of curvature.  Try modeling those dimensions and you can see the flats are quite clearly defined.  The flats are large enough to see.  Thus, even the figure is ambiguous.  I'd say that with close inspection figure 2-18 does NOT include the tangent edges.

KENAT (Mechanical)
31 Mar 08 14:00
1.8.5 Is in effect for any drawings made to ASME Y14.5 unless you have an explicit exclusion.

1.1.4 Is pretty much a get out of jail free card for figures in the standard not being quite complete etc.

That said there do appear to be small flats on 2-18 though due to the line weight of the part contour it's a bit difficult to make out.

It seems like you're trying to play devils advocate and push the 'letter of the law' to the limit.  Is there some reason for this?

That said, the fact it stresses tangent for CR but not for R does I suppose make you start to wonder.

KENAT, probably the least qualified checker you'll ever meet...

tgwow (Mechanical) (OP)
31 Mar 08 14:45
Kenat,

Remember that CR used to be called Tangent Radius from 1982 so yeah, why?

I think the definitions of the tolerance zones for R, CR and the figures are not well defined enough to say that the tangent sides are included or not.  There should be no ambiguity even when pushed to the limit.  I shouldn't be able to make the case for not including the tangent sides but I think I can.
aardvarkdw (Mechanical)
31 Mar 08 14:52
I think that what figure 2-18 is trying to convey is that the profile of your radius can fall anywhere within the limits of the tolerance boundary (which incidentally is shown tangent to the edges of the part). The radius is shown exaggeratedly jagged to illustrate the point shown by the boundary lines.

I bet if you took out your caliper and measured the figure the dimensions wouldn't even be correct... see section 1.1.4

David

KENAT (Mechanical)
31 Mar 08 15:00
tgwow, never worked to the 1982 version, so no I don't recallwinky smile.

I don't think there is ambiguity, however there have been areas of the standard discussed before that I find unambiguous but other argue are unclear.  From experience I doubt either of us will convince the other.

KENAT, probably the least qualified checker you'll ever meet...

powerhound (Mechanical)
31 Mar 08 15:55
tgwow,
  As far as I'm concerned you're not making a case at all. I say that with no disrespect intended. The figures and paragraphs in the standard make it clear what the correct interpretation is and refusal to subscribe to what the standard says does not constitute "making a case" just like making the same point over and over again does not constitute pushing something "to the limit." For lack of a better or nicer term this is just plain being stubborn.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
1 Apr 08 9:38
powerhound,

Thank you for your response.  And I do subscribe to what the standard says.  
 
Let's put things in some context here.
(1) Section 1.8 talks about how to dimension diameters and radius and what the graphics should look like.
(2) Section 1.8.5 says to draw rounded corners using an R dimension and yeah draw it tangent to the sides.
(3) Figure 2-18 conveniently shows a corner.  So let's apply 1.8.5. Okay, I can agree with you that Figure 2-18 should include  the tangent sides for a corner.  

What if the the round is not tangent but perpendicular?  What if it's not even a corner?  (See attached shapes)
What do you do for those case?
powerhound (Mechanical)
1 Apr 08 11:05
tgwow,
  Your new drawing is a totally different animal. We should compare apples to apples. 1.8.5 does not apply to this drawing, 6.5.1 does. 1.8.5 only applies to filleted corners. If it's not a filleted corner then profile of a surface or line should be considered. See Fig. 6-13. It has something close to one of your illustrations. Profile tolerances are what should be used for all of the views you've shown. Everyone got hung up on 1.8.5 because of the first drawing you submitted.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
1 Apr 08 12:37
Yeah, they're different animals but I'm concern about the "kingdom of animals".  I'm concern with a generalization of the whole "kingdom of animals".  It's like saying "animals should have legs."  The End.  So where do the legs go?  Enough of that:)

So the Section 2.15 Radius, where it describes the tolerance zone for R, should apply only to filleted corners?
I thought it was for anywhere you use R+/-.  

Again, the tolerance zone of R is ambiguous since R+/- used anywhere else but corners is not well defined.
 

powerhound (Mechanical)
1 Apr 08 13:17
Now that we're talking about something different than what you originally posted I will address this in a different way.
If we are only talking about filleted corners (as in your first drawing), then the arcs that constitute the tolerance zone must be tangent to the lines that are creating the corner to be filleted. If we are talking about radii that are creating an actual part contour (as in your second drawing), then the tolerance zone is going to be created by the location of the arc, the tolerance of the location of the arc, and the tolerance of the arc itself. In coordinate tolerancing this can be a nightmare but with GD&T the tolerance zone can be fully defined. When you locate the centerpoint or tangent points of the arc and apply a profile tolerance, the zone is clearly defined. You can't just apply an R in this case without specifying where the R is.

Take another look at Fig. 6-12 and 6-13 and tell me if you see anything ambiguous about the way they're defined and if you do, please let me know what it is you're looking at. The R82 and R80 use the tangency rule for "location" as this profile is mathematically defined.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
1 Apr 08 15:10
powerhound,

I'm completely at ease with Figs 6-12 and 6-13.  The tol zones are well defined.

However, I'm not completely sure of how to even draw the tol zones for the simple shapes (for any or some chosen dim scheme with R+/-) I drew.  I'm getting the sense that R+/- is not be well defined enough to the point that you'll have to resort to actual profile tolerancing to make it meaningful (excluding corners).  

fcsuper (Mechanical)
1 Apr 08 15:44
tgwow,

I don't necessarily agree with your accessment of R.  However, if you want to get some traction by bouncing your POV off of individiuals associated with the ASME, join the ASME Y14-5 Yahoo! Group here:  http://tech.groups.yahoo.com/group/Y14-5_User_Group

Good luck.

Matt Lorono
CAD Engineer/ECN Analyst
Silicon Valley, CA
Lorono's SolidWorks Resources
Co-moderator of Solidworks Yahoo! Group
and Mechnical.Engineering Yahoo! Group

powerhound (Mechanical)
1 Apr 08 16:40
tgwow,
From your last post I am starting to think this just may be an issue of not fully understanding GD&T and what the standard really says. Your drawings are completely definable with GD&T with no ambiguity. Please know that I am not putting you down or anything like that but your assessment is unfounded. There will frequently be cases where you have to refer to more than one section to get a full understand of a concept and there are frequently extensions of principle that have to be applied because the standard can't address every single scenario that may present itself. Your drawings do not fall into a category of being difficult or tricky to dimension yet you present them as being so. They are elementary parts that are easily and unambiguously toleranced.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
1 Apr 08 17:33
Yes, I'm definitely not an expert on the standard.  That's why I'm here.  I understand that I can define my drawings with no ambiguity with GD&T.  

My question is "In general can R be used meaningfully without GD&T?"  

So help me out. Pick any one of the three shapes, add your +/- dim scheme and show me the tolerance zone for R.  The shapes are simple so this should be easy.
KENAT (Mechanical)
1 Apr 08 17:46
I think that's the point.

Defining many shapes well with just +- dimensions can actually be difficult.  

As mentioned I think above, in the case of a rad you typically have both location and size tolerances to be concerned with.  Using +- you can end up with either excessively tight tolerances generally to meet the requirement in one area or you end up with excessively loose tolerances if you're not careful.

This is one of the reasons GD&T was developed.

Plus some of your shapes may have more than one possible dimension scheme, depending on end function which is what should drive dimensioning & tolerancing.

KENAT, probably the least qualified checker you'll ever meet...

powerhound (Mechanical)
1 Apr 08 18:10
Now you've added another factor into the mix. This is the first mention of dimensioning WITHOUT GD&T that has been made. I'm at home now so I don't have access to CAD but I can kind of describe something.
On all these parts it's as simple as dimensioning the center of the radius then specifying the radius. This is coordinate tolerancing and the tolerances can really stack up and give you something you don't really want. Not only do you have the tolerance of the radius to deal with, you have the location of the radius to deal with too. You can eliminate this with GD&T by making all the dimensions pertaining to the radius basic.  You can also locate the radius with a toleranced dimension and make the radius itself basic and apply a profile tolerance if you want to allow the radius to wander but keep the profile of it controlled.
As I said, I don't see an issue with dimensioning these drawings at all. Specifically what issues are you seeing?
Also see 1.8.2.1 for the rules pertaining to tangency and radii centers and 1.8.6 for the section that would pertain to dimensioning drawing C.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

Xplicator (Mechanical)
1 Apr 08 19:54
I'm really having a hard time seeing the value in this discussion.  
tgwow,
I don't exactly know what it is that you are trying to accomplish here. The shortcoming of the use of the radii "R" callout has been beat to death long before this thread was started.  I believe the standards committee is addressing this or at least acknowledging it in the (November 2007) draft revision of the standard.  If you have suggestions that are part of a solution rather than continuing to point out your discovery that is already well known.  Than pony up the $85 and purchase the rumored next draft and submit your constructive comments for review.  

In the mean time use common sense, if the radii has a function that the "R" callout alone does not control and could compromises the integrity of your design, than find another way within the standard to communicate your intent and its function.  

This is my last post on this subject/thread unless there is some value to be salvaged from it.
tgwow (Mechanical) (OP)
1 Apr 08 22:47
My point is this.

If R+/- is so deficient then it has no place in any drawing except for breaking edges.

I see R+/- used all the over the place without GD&T.
KENAT (Mechanical)
1 Apr 08 23:49
I wouldn't limit it to quite just breaking edges but yes, when using +- it's a bit limited.

To some extent the same goes with chamfers and other faces at other than 90°.

As checker I see all types of stuff on drawings that is open to interpretation, R +/- is often the least of my worries.

I'd say the majority of the people here creating drawings have very little knowledge/understanding or even awareness of these types of issues, the fact that I with limited experience and no real formal training am one of (in top 3) most advanced GD&T users/tolerance analyists in the place speaks volumes.

KENAT, probably the least qualified checker you'll ever meet...

powerhound (Mechanical)
2 Apr 08 0:59
Xplicator,

  While there may be no value in this discussion for you, I sort of hope there is value in it for tgwow. I know there are issues with the use of R alone when specifying the rounded ends of slots and obrounds but what is the major issue with using R to specify and tolerance a radius? How would you do it otherwise? Remember the OP was not on the use of R for the ends of a slot or obround nor is it what I have addressed in any shape, form, or fashion, it was for the shape of the tolerance zone for a dimensioned radius.

tgwow,

  If this is dragging on and dulling your senses and is proving to be of no value, I apologize.

Powerhound, GDTP T-0419
Production Supervisor
Inventor 2008
Mastercam X2
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

tgwow (Mechanical) (OP)
2 Apr 08 12:40
Xplicator,

On the contrary, this has been extremely valuable.
My major issue is this:
If it's difficult to show the tolerance zone of R+/- then wouldn't it be difficult or even next to impossible to verify that your parts are good?  Thus, using R without GD&T is no good in practice.  

GD&T should be used 99.99% of the time instead of just R+/-

And yes, the original OP which was a starting point has been addressed (Thank you all).

However, to have a better understanding I needed to close all the related issues.

Reply To This Thread

Posting in the Eng-Tips forums is a member-only feature.

Click Here to join Eng-Tips and talk with other members!

Close Box

Join Eng-Tips® Today!

Join your peers on the Internet's largest technical engineering professional community.
It's easy to join and it's free.

Here's Why Members Love Eng-Tips Forums:

Register now while it's still free!

Already a member? Close this window and log in.

Join Us             Close