The meaning of FRTZF without datum reference 2

[SeasonLee]

Hi :

The books I read say the lower segment without datum reference controls the sapcing, the perpendicularity and parallelism are controled by the upper segment. I am confused and I really don't know what is the real meaning of this example.

Thanks in advance for all inputs.

Season

 

[dingy2]

MechNorth:

The holes that are used for lightening the product have control through co-ordinate or possibly polar tolerancing. They do not need control with positional tolerancing since there is no function and relationship. One simply confirms the centre of the holes on X & Y axis using a CMM. The holes locations would probably not be checked on a continuous basis but certainly at sample submission. Their locations are not important.

Let's say that we place a positional tolerance at MMC on the pattern of holes to lighten the product. We would think that their location must be important so off we go and make a hard gauge. Now we have it in the Control Plan checking the hole locations hourly. What a waste of time and money and the costs go up.

I realize that the company that you now represent promote that all holes should have positional tolerance and all surfaces should be reflected in profiles whether or not they are needed. No wonder the drawings that are overloaded with GD&T are scaring the suppliers.

Function and relationship is important. GD&T is a tremendous value used on such features but not all features.


Dave D.

http://www.qmsi.ca

 

[KENAT]

You know, on the issue of unimportant holes, such as a grid of lightening holes, I think there is still an argument for using position.

Just using the circular tol zone increases permitted tolerance variation without changing the effective 'spacing' of the holes.

Now I agree calling up MMC may not be appropriate, in fact LMC may be more appropriate if you're concerned about the edges of holes getting too close to each other (or other features).

I'm also concernd by the very idea of trying to interperet 'what's important' based on just on whether GD&T was used or not, without considering function/how tight the tolerance is/how it fits in process tolerance etc.. I believe it's in part this kind of thinking that can lead to increased prices when we GD&T an old part even though we've relaxed tolerances.

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

 

[caseynick]

GD+T does not increase cost. If anything it decreases cost by increasing clarity of the drawing. It is the amount of tolerance allowed that drives the cost of production and inspection. A .002 tolerance costs more than a .02 tolerance cots more than a .2 tolerance. If a feature is "unimportant" tolerance it accordingly. Don't encourage people to guess what is important and what isn't based on how the drawing is dimensioned.

 

[KENAT]

You don't need to convince me caseynick, tell our vendors and some of our engineers.

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

 

[ringster]

GD and T of itsself doesnot increase costs. However, in you are in an environment where not all users are up to speed on the Standard, there are guaranteed increased costs.

 

[MechNorth]

Actually Dave, weight-reduction holes would be controlled using a surface profile; this establishes a nice boundary zone for the feature and only one inspection setup (no Full-Form @ MMC & 2-Point @ LMC plus probing its true geometric counterpart to find the center nor using a CMM). Whether or not a feature is inspected is not based on whether or not it has a geometric control, it's typically based on the tolerance requirement. If I have a surface profile with a 25mm tolerance on a hole, it's pretty clear that it's a generous tolerance. If the hole has a profile tolerance of 0.25, then it's moderately tight, and may need to be inspected periodically based on manufacturing capabilities. The cost of manufacturing doesn't go up with GD&T, the cost of manufacturing goes up with users' perceptions of GD&T and their ignorance of GD&T.

As part of a significant project which included GD&T implementation and cost reduction as just two facets of the project, I rethought the tolerancing and finish requirements of a very simple piece shaped rather like a heavy-set "L". I grossly opened tolerances and finishes to the point where it should have reduced our costs by at least the targeted 15%. Quelle surprise, vendors quotes ranged from about E750 (Euros) to $350 USD to $85 CDN. The original price was about $225 USD. The European supplier claimed to be experts in GD&T (ASME & ISO), the US supplier claimed ASME expertise as evidenced by working in the aerospace industry, and the Canadian supplier had about 4 hours training but understood enough to adequately study the prints and understand the controls applied (including being able to explain them back to me for confirmation). It was the lack of proper training & understanding that had driven up the costs of the other suppliers. I find that repeatedly in large companies whether on the OEM or supplier side.

As for what features to inspect, that's subject to the tolerances and manufacturing capabilities. Similarly, how you measure something is subject to the tolerances, fequency of inspection, and inspection capabilities (hardware, software & operator). If you occasionally needed to check the clearance holes, you could do an open setup fairly quickly; far quicker than a CMM can be programmed and establish the datums. I don't like bureaucracy, but I do like documentation, and an Inspection Protocol is something that I regularly recommend to clients. It ties in the symbology on the drawing that's used to indicate inspection requirements, the equipment to be used, the sampling methods / rates, etc. Developed by engineering, manufacturing and quality together as a team, you can effectively establish your inspection requirements. That way there's no guessing or assumptions as to what needs to be inspected.

I occasionally get a student who feels that you don't need to control everything back to the datums because they aren't critical features. I lead the class through rectangular and wedge shaped tolerance zones, the impossibility of verifying a single center or single radius of a non-FOS, the typically conflicting conventional & geometric controls, and ask how much time their employer wants them to spend in court defending an inadequately documented design. IF FOR NO OTHER REASON, the avoidance of misinterpretation with resulting litigation should be enough to compel an engineer to completely control their drawings.

I recognize that from an inspection perspective, that seems pointless, but from an engineering perspective it is absolutely essential. From an inspection perspective, you need to know what features on the drawing need to be inspected and with what frequency, but the presence of a geometric control DOES NOT establish a requirement for inspection in & of itself, it only communicates the designer's intent for that feature.

What tends to be forgotten is that GD&T is an ENGINEERING COMMUNICATION TOOL; it tells everyone else involved in the production life of the item exactly what the designer intended in an unambiguous, consistent (subject to the designer's abilities of course) and legally defensible format basesd on a defensible standard. It DOES NOT tell anyone how to manufacture the part, nor how to inspect it.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[MechNorth]

Oh, and Dave, I enjoy our debates too!

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[dingy2]

MechNorth:

I was training in an automotive plant recently and their customer forced them to place in the notes section of their drawings that all surfaces must have a profile of a surface tolerance of a value or some sort. I also saw on the same drawing a feature control frame with a profile tolerance and exactly the same value. It, obviously, was redundant. I asked why they double dimensioned this feature.

I was told that the other features didn't really require the profile tolerance (no design requirement) while the feature shown with the feature control frame actually did. It WAS IMPORTANT to its function and relationship and the design intent of the part.

Should the Designer have reflected all surfaces with a profile tolerance when it wasn't needed or no design intent? It can and is quite legal to show co-ordinate tolerances in addition to GD&T. Not all dimensions must have a feature control frame and there is no where in the present standard that dictates this.

Design intent - absolutely. In my over 20 years of training in the subject, I would never ever suggest that ALL dimensions have a feature control frame. It is a valuable tool to reflect the design intent where applicable.

Don't kill the fly on the wall with a 12 gauge shotgun.



Dave D.

http://www.qmsi.ca

 

[MechNorth]

Dave,

Again, someone is making an assumption that a geometric control automatically requires verification on the final article ... please show me where it says that in Y14.5M-1994. Without a CMM, how would you be verifying features that aren't located by a geometric control? Please explain to me how you would find the center of a hole with the old +/- location tolerances and no datum reference frame. Applying the inverse of your logic, if the hole's position is controlled by conventional tolerances, then you'd never inspect it? So, if it is grossly out of position, how would you know?

Individual / job function perspectives determine how we see things. What seems irrelevant to manufacturing and inspection may be of low importance to the designer, but still of importance none the less. The weight reduction holes, for example, may seem irrelevant as to the exact location for manufacturing and inspection because they have somewhat loose tolerances, however a designer may have done FEA and other analyses and established that stress flow is optimized by the configuration being moderately restricted, or stress concentrations may be negatively impacted by a more significant shift of the geometries. An engineer's perspective cannot just include what manufacturing and inspection sees, but also must include consideration of legal implications of an incompletely defined product.

As for the use of chained/ordinate tolerances, they are not precluded in the current (1994) standard, however they are not recommended either. "Section 2.1.1.1 Positional Tolerancing Method Preferably, tolerances on dimensions that locate features of size are specified by the positional tolerancing method described in Section 5. In certain cases, such as locating irregular-shaped features, the profile tolerancing method described in Section 6 may be used.". I also couldn't locate a single example in the standard where conventional tolerancing was illustrated for position dimensions for a feature of size. Regardless, it is a rather shortsighted view to exclude all other relevant standards, particularly those that will dominate the automotive sector very soon...i.e. Y14.41. In ASME Y14.41-2003, Table 8-1, the accepted use of +/- tolerances shall be restricted to:
1) Fillets, Rounds & Chamfers
2) Reliefs, Step Surfaces
3) Countersinks
4) Oblique Surfaces
5) Entry Depth and Spotface
6) Remaining Thickness
7) Notches, Flats, and Pin Heights.
As Y14.5 & Y14.41 are supposed to be companion standards, don't be surprised if the forthcoming revision of Y14.5 reflects the same restrictions of use for conventional tolerancing.
Now, if you look at the Note in Section 8.2 of Y14.41, it indicates that the list is not exhaustive, but rather indicative to similar and other valid applications. While it doesn't explicitly preclude the applicability to position dimensions, such application would be contrary to the indicated application methods of such controls which are restricted to attachment to size callouts, directed leader to the feature (surface), or on an extension line from the feature. Would you have such a tolerance applied to a centerline which isn't even present on the model?

Industry, automotive and aerospace in particular, is migrating slowly but inevitably away from design documentation by drawings, and towards the use of CAD models for all aspects of manufacturing and verification. GD&T users can choose to fight it (& lose) or embrace it and move forward. I'm already seeing a slowly growing trend in small & mid-sized companies using GD&T toward minimally-dimensioned and dimensionless drawings by requiring the use of the CAD models.

You can't stop a swarm of wasps with a pair of chopsticks.

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com

 

[dingy2]

MechNorth:

All dimensions require verification whether through the customer or through your own company. That is not described in the ASME standard and the standard just does not get into confirmation methods at all. This is usually performed on samples and then manufacturing also perform ongoing verification on characteristics (features) deemed important to the function and relationship of the part.

I would suggest, as per 1.9.1 of the standard, that co-ordinate tolerancing of holes reference that datum reference planes so the datum structure must be present.

So many years ago, I measured parts for a living, both with Portage Layout Equipment (manual CMM) and on the surface plate. The holes in a pattern were reflected in co-ordinate tolerancing. We actually reviewed the machining of the part or its mating surfaces to develop the 3, 2 and 1 point set up on the datums. This, of course, was prior to GD&T.

One would measure the actual hole size and then calculate the dimension to the edge of the hole. From the secondary datum, measure to the side of a hole and reflect the actual calculated dimension. We would also measure from the tertiary datum in the same manner.

I will let you have the last kick at this one if you like.

Dave D.

http://www.qmsi.ca

 

[MechNorth]

Tks Dave. For that last prod (I won't kick you), I'd like to redirect you back to Section 1.9.1 Rectangular Coordinate Dimensioning, not Tolerancing.

In fact, Section 1.9 deals with dimensioning of the Location of Features, not Tolerancing of the Location of Features. You seem to be reading more into that title than is indicated in the text. Indeed, coordinate dimensioning whether chain, ordinate, or polar, are all perfectly acceptable and are indicated in the standard.

Chapter 2 deals with Tolerancing and Related Principles, and Section 2.1.1.1 again deals with Positional Tolerancing Method. In that section, it stipulates two ways to position an enclosed boundary whether a feature of size or an irregular shape; position, and profile of a surface. At no time does it mention or allude to controlling the locations of such enclosed boundaries by means of conventional tolerances.

One of the basic tenets of GD&T is that you are controlling features rather than dimensions. You cannot verify a basic dimension. You can verify the controls applied to the feature whose theoretically perfect location is established by basic dimensions which are tied back to the datum features whether directly or though chaining.

For more fun, I started another thread on this topic ... full vs partial applicatiion of GD&T. See ya there, Dave!

Jim Sykes, P.Eng, GDTP-S
Profile Services

http://www.profileservices.ca

TecEase, Inc.

http://www.tec-ease.com