TIR per inch 2

[TrailMaker004]

I haven't seen this appear on a dwg in a while, but I know from previous experience there is a way to describe the maximum amount of measured T.I.R. in a given length. For instance: | // | .002 (per inch) | A |, meaning the measured feature must be parallel to datum surface A within .002" per each inch of measured (indicated) length. Problem for me is I can't remember the exact syntax of how to express this on the dwg.

 

[KENAT]

By TIR do you mean 'Run Out' per ASME Y14.5M-1994?

From a quick look I don't see anything in ASME Y14.5M-1994 section 6.7 about applying on a unit basis.

Quick look through other parts of section 6 only appears to show straightness & flatness applied on a unit basis 6.4.1.1.4 & 6.4.2.1.1. Syntax is something like:

[-|Ø0.5 / 25]
[/=/|0.05 / 25 X 25]

It does give some warnings about the use of this in 6.4.1.1.4.

Both examples have an upper 'loose' tolerance for the full length and then a tighter 'unit basis' control in a second line of the composite fcf.

Posting guidelines faq731-376

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

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

 

[pmarc]

ASME Y14.5-2009 also mentions only about flatness & straightness applied on a unit basis.

My opinion is that only these two controls can be applied with this concept at all. I can imagine a lot of difficulties in correct interpretation of such concept when it is applied to a geometrical tolerance that has got any datum reference (e.g. parallelism, runout or whatever).

Maybe a reasonable solution would be applying parallelism control to a surface and then to refine it by flatness or straightness applied on a unit basis?

 

[TrailMaker004]

Thanks everyone for your time and feedback.

By TIR I mean total indicator reading, or full indicator movement (FIM). Generally for parallel or perpendicular reqm'ts, this is interpreted as the maximum observed indicator reading over the actual physical size or length of the feature. When using a CMM w/PC-DMIS, the software will default to assume the length/size of the feature is the maximum distance between the probe hits. Sometimes because of the shape or configuration of the part, it is not possible to place the hits at the extreme ends of the feature. In those cases there is valid concern that the additional unmeasured length of feature could put it over the tolerance limit. The software gives one the option of inputing a reference length over which the specified TIR must be held. The software will then extrapolate the TIR value over the specified reference length and output that value. BTW, in this case I am speaking of parallel or perpendicluar conditions of a line (axis), not a plane.

My current application is that we want the axis of a cylindrical feature to be maintained parallel within a specified TIR to a datum plane. Since we cannot access the entire length of that cylinder (for measuring) our solution (in order to meet the needs of the application) is to extrapolate the measured TIR to a specific unit length (one inch). I have seen this format used many times in the past, but it's been a while and I do not remember the specific syntax used. My best recollection is something like this: | // | .002 in/inches | A | (or something like that).

 

[dgallup]

You should have something that looks like:

|//|.002/1|A|

 

[KENAT]

My implicit point was that 'TIR' as a term isn't in the spec, FIM is mentioned in 6.7.1.1.1 but the term isn't a control itself.

So the actual control as defined in ASME Y14.5M-1994 that you appear to be wanting to apply is called RUNOUT.

You may think me being pedantic, but given your question is about getting the proper syntax, I think it's relevant.

I do not see the concept of Runout on a unit basis in ASME Y14.5M-1994 so getting the syntax right from that point of view is almost irrelevant. Maybe it's in a newer version of the spec or in the iso, I don't know.

I would think the best you can do is copy the syntax used for straightness, which I gave above, just remember the differences for inch V metric. However, given it's not defined in the standard what you may get is debatable.

Posting guidelines faq731-376

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

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

 

[dgallup]

Kenat - You are pedantically correct as always, but from the OP later posts it becomes apparent that he is not talking about run out at all but parallelism. Parallelism per unit length is appropriate.

 

[KENAT]

Maybe, but he still seemed to be talking about a cylinder, in a way that didn't make parallelism spring to mind for me.

Also, I don't see parallelism on a unit basis in the standard either, or am I over looking it?

Posting guidelines faq731-376

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

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

 

[pmarc]

Kenat, you are not overlooking anything. Parallelism or runout on a unit basis are not specified in Y14.5 std., neither in 1994 nor 2009 version.

 

[dgallup]

Don't have it handy at the moment. My CAD system (not implying it is correct) will let me create a geometric tolerance per unit for straightness, flatness, profile of a surface, perpendicularity & parallelism.

 

[dgallup]

I should have mentioned we use ISO standards.

 

[TrailMaker004]

Thanks again for the help guys.

In this instance it is parallelism here I am talking about. The measured feature is cylindrical, but the relative datum surface is a plane. Runout would be appropriate between two cylindrical feaures, or a cylindrical feature and a plane if the cylinder were the datum feature and the plane the measured feature.

I apoligize for the confusion, but having entered the industry in the 70's when everything was still done with surface plates and indicators (and cmm's had not been invented) all orientation and form deviation requirements (whether parallelism, perpendicularity, flatness, roudness, or whatever) were thought of in terms of TIR or FIM. My head still tends to default to that terminology. I understand that those terms may not be used in the std, but conceptually the principles remain the same.

 

[dgallup]

FWIW, ISO 1101-1983 clearly shows a parallelism per unit length in Figure 42. Section 9.1 states:

If the tolerance is applied to a restricted length, lying anywhere, the value of this length shall be added after the tolerance value and separated from it by an oblique stroke.

 

[pmarc]

I have ISO 1101-2004 version in front of me and the only tolerance that is shown on a unit basis is straightness (para. 12.1, fig. 44). It looks exactly the same as in ASME Y14.5M-1994 fig. 6-4 or Y14.5-2009 fig. 5-4. There is nothing about parallelism. I wonder why did ISO withdraw the example of parallelism per unit length in ISO 1101-2004 version?