precision shaft tolerances 1

[cervantes]

Hi

i am a little bit confused about tolerances given for induction hardened and hard chrome plated linear shaft
(see attachment)

let's analyze diameter 4: for ISO h7 tolerance, default circularity at level 0.005 - ok for me
cylindricity (parallelism) at level 0.01 - ok for me
but straightness - 0.3/1000 = 0,0003 ????

straightness must be usually lower than circularity - but is it ever possible to machine such a good way?....
or maybe I miscalculated something here?

I reviewed also Bosch Rexroth specifications, they are using similar descriptions and values

but why not to use runout as it is a mix of circularity of straightness?

M

 

[pmarc]

Why do you say the straightness tolerance is 0,0003?
From what I see, it is 0,3 mm for each 1 meter of the shaft.

 

[cervantes]

but then there is no way to keep h7 tolerance
if the profile may vary from 0 to 0.3 then it exceeds the shaft tolerance - am I right?

(I am from Europe, Rule#1 is not used here)

 

[cervantes]

maybe this might be helpful

this is how I understand it

 

[pmarc]

It can exceed h7 tolerance because it is ISO-based and because Rule #1 is not in charge.

Picture a perfectly round 1000 mm long dia. 4 shaft bent like a banana within straightness tolerance limits. It will meet size tolerance, but will have (and be allowed to have) 0,3 straightness error of its generating lines. For this shape, one of the two opposite generating lines is convex, the other one concave within 0,3. The generating lines, however, cannot be both convex or concave 0,3 at the same time.

 

[cervantes]

you are talking about this what I presented in attachment?
but then it rather applies to straightness of axis, not profile?

 

[cervantes]

With regards to my last question - just have noticed that cylindricity helps to control it, I got it.

 

[pmarc]

There is a bigger problem with the table - it says that allowable straightness error 0,3/1000 is bigger than cylindricity error 0,1/1000. I do not see how it is possible, either with or without Rule #1 in charge.

Unless they do not really mean cylindricity, but rather, as written, parallelism of opposed generating lines, which is something allowed in ISO.

 

[cervantes]

see attachment, other companies specifies parallelism
also you are right cause cylindricity is a sum of straightness, circularity and parallelism

 

[powerhound]

Where in the world did this last attachment come from? Is this a supported use of parallelism in ISO?

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[cervantes]

it is tsken from manufacturer brochure....

how would you describe it? is straightness and circulsrity enough?

 

[cervantes]

naturally parallelism requires a datum, but what are you thinking about a situation when it is obvious what was the intention of the drawer however it includes formal mistake?

 

[pmarc]

powerhound,

This is not supported use of parallelism in ISO. A similar (but not identical) practice was allowed in the first edition of ISO 1101:1983, but got withdrawn in the second edition issued in 2004. The method used in the brochure is different from the withdrawn because both leader lines terminate with arrowheads, whereas technically one of them should terminate with a triangle (like a triangle of a datum feature symbol).

From my experience, a general problem with many brochures like these is that they are very often not up to date with current standards or attempt to define their own meaning of universally accepted definitions/symbols or are simply messed up. The ones used in this thread are very good examples. In first attachment, for instance, a cylindricity symbol was used to define parallelism of generating lines, which obviously is not the only function of this symbol according to the standards. The same brochure specifies in the table that the maximum straightness error of dia. 4 shaft t3 is 0.30 mm/m, yet in the straightness tolerance frame shown on the drawing they repeated t3/1000 instead of just showing t3. That is simply a redundancy.

 

[powerhound]

Thanks for the info, pmarc.

cervantes, assuming that any given situation is obvious to everyone is exactly why standards exist. How does this errant parallelism callout work? Keep in mind that datums are flat and straight, regardless of the condition of the datum feature. Also keep in miind that there is no datum feature referenced. So how would this actually be checked?

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[cervantes]

ok agree

another kind of question: I am intensively studying GD&T using all available sources from a longer time
when I think that I have completely figured out some issue I am always finding something (like the pages from brochures) which ruins what I know or which creates a lack of faith of my knowledge

Generally I am assuming that a big company, which creates a brochure, is doing this perfectly. If there is something wrong, like on the attachments, the first impression is that this kind of manufacturer shall not make this kind of mistake so that's why something is wrong with my knowledge.

 

[KENAT]

cervantes, actually the GD&T I see in typical brochures or vendor data sheets is abominable.

In fact it's rare to get good 'T' and you're lucky to get coherent 'GD'.

Posting guidelines faq731-376

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

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

 

[cervantes]

hi

I started to have some doubts.

Runout is a sum of circularityy and coaxiality.
And total runout is a sum of cylindricity and coaxiality.

But let's analyze runout (1st option)

This is one-step shaft.

Assuming that circularity is perfect (no deviation), then coaxiality (created by deviation of straightness at highest point, see cross-section) at the attached picture is 0.3 (which shall give run-out value also at level 0.3).

Is my way of thinking correct?

If no run-out is given here and default for eg. 2768-2-H is 0,1 then is this detail ok or not?.... or does 0.3mm per 3000m of straightness overrides default runout tolerance?

 

[pmarc]

cervantes,

First off, in this case forget about using general circular runout tolerance as defined in ISO 2768-2.
The standard explicitly says that: "For general tolerance on circular run-out, the bearing surfaces shall be taken as the datum if they are designated as such. Otherwise, for circular radial run-out, the longer of the two features shall be taken as the datum; if the features are of equal nominal length, either may be taken as the datum."
In your case (1) designated bearing surfaces do not exist, and (2) you do not have two features to be able to determine runout of one relative to the other.

Secondly, in the attached picture, assuming that both machine centers are located exactly in the geometrical centers of the circles at both ends, the actual value of circular runout error at the highest section relative to the axis established by the machine centers is not 0.3, but 0.6. The 'value' dimension equals 0.3. Why you used machine centers is a different story ;-]

 

[cervantes]

thanks, naturally agree about 0.6

so "default" run-out applies only for at least two-step shafts

for single-step shafts case like this, after specyfing datum on bearing surface, straighntess 0.1mm/1000mm could be more or less exchanged to run-out at 0.2mm?... it is hard to imagine this because in the fact I have never seen a drawing of a single step shaft with specified runout, just quick checking now google pictures after typing runout and also all examples refers to at least two-step shafts

 

[mkcski]

From my limited understanding of "confusing" myriad of interconnected ISO GPS system standards, this entire discussion would be avoided if Y14.5 and NOT ISO standards we specified on the drawing. Anyone agree?