Inspection Dimension (SolidWorks drawings) 6

[evolDiesel]

In SolidWorks and I believe other platforms they have this option to turn a dimension into an inspection dimension (see example below):

The problem is we can't find anything on this in a standard. We specifically follow ASME Y14 and can't find any coverage on this format.

Can anybody shed some light on this standard, where it comes from, and any links documenting it. We are worried that it's local to a CAD program and not a real, observed standard.

Thanks,
Jack

Jack Lapham, CSWP
Engr Sys Admin
Dell M6400 Covet (24 Season 8, Ep 22)
Intel Core 2 Duo T9800, 2.93GHz, 1066MHZ 6M L2 Cache
8.0GB, DDR3-1066 SDRAM, 2 DIMM
1Gb nVIDIA Quadro FX 3700M (8.17.12.5896)
W7x64 | sw-01: 55.92
SolidWorks x64sp4 in PDMWx

 

[evolDiesel]

Jeff,

That story sounds plausible and based on all this user feedback sounds the most likely.

Jack

Jack Lapham, CSWP
Engr Sys Admin
Dell M6400 Covet (24 Season 8, Ep 22)
Intel Core 2 Duo T9800, 2.93GHz, 1066MHZ 6M L2 Cache
8.0GB, DDR3-1066 SDRAM, 2 DIMM
1Gb nVIDIA Quadro FX 3700M (8.17.12.5896)
W7x64 | sw-01: 55.92
SolidWorks x64sp4 in PDMWx

 

[ptruitt]

I may still have a template with the symbol in it for use when I applied pencil to paper. Perhaps the template manufacturer set the standard.

Peter Truitt

 

[fcsuper]

Jeff,
The symbol should be submitted to both the ISO and ASME bodies for consideration. Since you are Mr. InspectionXpert, maybe you can send them letters?



Matt Lorono, CSWP
Lorono's SolidWorks Resources & SolidWorks Legion
Follow me on Twitter

 

[DeanD3W]

Hello,
I'm new to this forum, but work with GD&T and ASME standards a lot.

I don't think encircling a dimension with race-track to designate it as an "inspection dimension", as Solidworks seems to imply should be a practice, is a good idea. With GD&T we don't inspect dimensions unless they're a size tolerance or the odd, or ambiguous, exceptions such as a radius or chamfer tolerance. On a typical drawing the tolerances being inspected will be size tolerances and those specified with feature control frames like Position, Profile, Runout, etc. I don't think anyone would propose to encircle a feature control frame with a race track shape, so I think a symbol placed beside the tolerance would be a better approach. In other words, we generally inspect tolerances, not dimensions, if using GD&T, and the entire concept of an "inspection dimension" is flawed, in my opinion.

I think Solidworks and others should delete the "inspection dimension" from their system.

Once a given tolerance is identified in a workable way as "special" then the next question would be in what way it should be special. Maybe statistical parameters like Cp or Cpk must meet certain criteria or maybe the given tolerance should be verified more frequently than others... This all sounds like measurement plan material to me. If all tolerances on a drawing have an ID number then those numbers can be referenced in a measurement report to designate how or when data should be gathered.

This may lead to a discussion about tolerance numbers on drawings... I think they're necessary if you want a clear way to match data in a report with the associated tolerance on the drawing. ASME Y14.45 will be a new standard for measurement data reporting practices that should be released within the next two years (maybe less). Y14.45 will include tolerance numbers in some way or another. It's obviously not completed yet, but I don't see how we can avoid having index numbers on tolerances, at least on an inspection drawing. Without them, we're left to the inspector adding numbers on a hard copy of a drawing as they go, which isn't the best approach for anything but very small operations.

Dean Watts
D3W Engineering LLC

 

[fcsuper]

Dean said, "With GD&T we don't inspect dimensions unless they're a size tolerance or the odd, or ambiguous, exceptions such as a radius or chamfer tolerance. ...we generally inspect tolerances, not dimensions, if using GD&T, and the entire concept of an "inspection dimension" is flawed..."

Dean, you may be over-thinking this. The fact that each dimension is actually a tolerence zone is a truism that is already understood by those using the process.

Each company has to decide for itself what is inspected on a regular basis. Quality control is a major factor in this. The more regulated the industry, the more control is necessary, and the more clarity required to establish that control.

The rounded box outlines clearly mark linear dimensions for regular inspection. This doesn't mean only these get inspected. It simply means that these much be inspected.


Matt Lorono, CSWP
Lorono's SolidWorks Resources & SolidWorks Legion
Follow me on Twitter

 

[ptruitt]

Matt,

As a designer, I never want to say that anything be inspected at all. I see that as another issue. I present the design, not the process (including inspection), whenever possible. (I do make exceptions.) This gives the greatest flexibility to the organization and reduces the number of drawing changes. Inspection criteria is more likely to change than a robust design. Drawing changes cost money, introduce dangers such as financial risk, legal risk, and possibly safety risk.

Peter Truitt

 

[DeanD3W]

Matt,
You said "Dean, you may be over-thinking this. The fact that each dimension is actually a tolerence zone is a truism that is already understood by those using the process."

I don't mean this in an unfriendly way Matt, but I think that you're possibly under-thinking this. I am very much one of those using the process, and I don't agree with your statement. Nothing about basic dimensions creates a tolerance zone, or in your words "is actually a tolerance zone". The tolerance zone is created by the position, profile of a surface, etc call-out defined in a feature control frame, which then requires the aid of basic dimension(s) to obtain the shape/orientation/location of that tolerance zone. A basic dimension cannot be measured. The location, orientation, size, and/or form of a particular component (axis, surface, derived median line, etc) of a feature is determined and whether that component is within its tolerance zone is evaluated. That's not to say that providing data to describe where the controlled feature component is, relative to where it is supposed to be, isn't worthwhile and very common. Regarding those values, I've run across quite a few people that like to say that they're measuring and reporting the value for a basic dimension, but what they're doing is much better described as providing the location components for a feature with a position tolerance applied, or the surface deviations for all or selected points for a feature with a profile of a surface tolerance applied (surface deviations, either "+ material" or "- material" being the profile component reporting method that works the same for a feature of any shape, angle or location). A basic dimension is nothing more than a theoretically exact location, or orientation, or shape (radius of curvature for a profile zone on a curved feature, for instance). To say you can measure the value of a basic dimension would require setting aside its definition.

Do you mean that you prefer to say that the value of a basic dimension can be measured and reported, or are you referring to using directly tolerance dimensions to locate or orient a feature? I took your comment to mean the former, but if you mean the latter, then I'll still disagree, but with a different argument against .

I don't intend any of this to be adversarial... Good discussions with good people are great and worthwhile expenditures of time IMHO. I'll do my best to be one of those good people while posting on what seems to be a very worthwhile forum.

Best Regards,
Dean

http://www.d3w-engineering.com

 

[fcsuper]

ptruitt,
The same could be said for using a quality control system where mfg engineering makes no statements about inspection. It just depends on the system in place. In an ISO environment, it is normal to have drawings contain some level of information to streamline inspection documentation. The last thing some companies want is their inspectors making decisions where they are unqualified about engineering matters. You want to talk about risk? Unqualified people making critical and undocumented decisions is far more risky than revising a formally controlled drawing once in awhile.

DeanD3W,
Not to seem too friendly, but it's simply. People are using these with great effectiveness. If someone is focused on the BASIC dim itself and not the FCF, then they are using GD&T incorrectly, and that is not applicable to this discussion (unless you want to make the case that reading GD&T incorrectly should be factored in to how inspections are conducted). Besides that, I've never seen an outline applied to a BASIC dim, nor a FCF. It's simply a shorthand that is applied to linear dims.

Matt Lorono, CSWP
Lorono's SolidWorks Resources & SolidWorks Legion
Follow me on Twitter

 

[DeanD3W]

OK Matt, I understand that you like the "inspection dimension" race-track symbol for directly toleranced dimensions. If GD&T is being used properly then there really aren't too many directly toleranced dimensions to encircle... Only size tolerances and maybe a few others like radius or chamfer tolerances. If there's going to be a way to designate inspection being required for a given tolerance then that method should work well for all tolerances, including those specified with feature control frames. A symbol placed beside the tolerance is simply a better method because it works equally well for directly toleranced dimension or for feature control frames.

I'll be standing by my position that the "inspection dimension" race-track that is placed around a dimension should be deleted from CAD systems. It's a poor practice that is based upon a +/- world, rather than upon a proper approach with GD&T.

Dean

http://www.d3w-engineering.com

 

[dgallup]

Obviously, the standards industry has a left big hole to fill. There seems to be no international standard for designating the important/significant/critical/inspection dimensions on a drawing. Even the ISO TS16949 quality standard fails in this regard as they say the symbols to be used and their meanings depend on the customer. That is an obvious cop out on standardization.

I know the purists will insist that all dimensions are important & they all need to be met. While I agree with that in principal, the reality is no one is going to pay for the inspection of each & every dimension, even on things like safety critical jet engine parts. (See recent failures on RR engined AirBus planes).

I agree the racetrack symbol is flawed as it can not be applied to GD&T. ANSI should take the lead & come up with a standard for special characteristic symbols including but not necessarily limited to:
Safety/Regulatory
Fit/Form/Function
SPC
Inspection

 

[ptruitt]

Designers should be very cognizant of quality issues. Many organizations have the design group sign off on the overall quality program and part-specific plans. But quality plans are most efficient when they are flexible. So the documented process capability of an established supplier might justify dock-to-stock. But the quality of a new, alternate, supplier might need more scrutiny. Although there are always exceptions, keeping process information off of design documents is usually the right thing to do, in my opinion. That includes the inspection process.

Peter Truitt

 

[evolDiesel]

Fellas -

I just want to chime in and thank you for a fantastic discussion.

Jack

Jack Lapham, CSWP
Engr Sys Admin
Dell M6400 Covet (24 Season 8, Ep 22)
Intel Core 2 Duo T9800, 2.93GHz, 1066MHZ 6M L2 Cache
8.0GB, DDR3-1066 SDRAM, 2 DIMM
1Gb nVIDIA Quadro FX 3700M (8.17.12.5896)
W7x64 | sw-01: 55.92
SolidWorks x64sp4 in PDMWx

 

[fcsuper]

ptruitt,

If this was 2001, I would've agreed with you. However, I've since been through the bureaucratic hell that results from forcing the process itself to carry the burden of determining minimal inspection requirements. I could (and maybe should) write a whole book on the topic.

DeanD3W,

I think you and I agree on the concept, but that you aren't OK with this particular solution. An alternative is the use of a flagnote (which might be too easy for an inspector to miss on a drawing with many different flagnotes) or creating one's own symbol supported by a general note (which I *have* already done for visual inspection points).

Matt Lorono, CSWP
Lorono's SolidWorks Resources & SolidWorks Legion
Follow me on Twitter

 

[ptruitt]

Frank Watts is author of "Engineering Documentation Control Handbook". His third edition (2008) provides fine advice that I usually agree with. Although people really need to read this book cover-to-cover, in my opinion, page 151 gets to the ownership of process documentation.

Peter Truitt

 

[ptruitt]

I should also recommend reading the topic "Process Design Specifications" at the bottom of page 43. I would caution folks that it is difficult to see where the author is coming from and where he is going without reading the book cover-to-cover. Matt is right in that it takes books to cover this topic. Fortunately, Frank wrote an excellent book, already.

Peter Truitt

 

[KENAT]

I'm afraid I'm with Pruitt on this. At the aerospace place I used to work, it was up to the quality dept to come up with a 'Quality Plan' part of which was determining which dimensions/tolerances got inspected with what regularity etc. They took into consideration various factors when doing this and if appropriate annotated drawings were included in the plan.

While the standard does set some precedent by mentioning indicating statistically toleranced dims/tolerances, in general I feel it falls under the fundamental rule of not normally specifying process on the drawing.

This goes for the related topic of 'critical dimensions' too.

Posting guidelines faq731-376

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

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

 

[ptruitt]

I see the ST callout as being goal-oriented, not process oriented. If historical inspection documentation demonstrates the process is capable for that part, the QC folks might be entirely justified in blessing a dock-to-stock program for that part. The racetrack demands an action, but falls far short of stipulating what that action might entail. To make the racetrack highly efficacious, the drawing notes would need to reference a static quality program plus it would need to reference a part-specific plan at a specific QC plan revision level. This is inefficient, of course. From a purist viewpoint, it forces the part to be reviewed any time a change is made to the overall quality plan.

Peter Truitt

 

[Nella95]

I know it seems awfully silly, but I'm with Dean in that the 'racetrack' symbol should not be put on the print in order to make the print. (I believe I'm under-thinking this, by the way.) Maybe an inspection print should be made separate from a working print? I don't know - I've just seen problems in the past when so-called inspection dimensions are followed yet other non-inspected dimensions aren't thus interfering with the primary function of the part.

 

[fcsuper]

Nella,

What is starting to happen is that the responsibility for inspection is moving out of the realm of Incoming Inspection and into the area of the vendor themselves, who then provide the inspetion data and a certificate of compliance or comformance or whatever. In other words, it is communicating information to the vendor that they need in order to fulfill their requirements with the product they make. I guess an even similar way to say it is that the requirement to inspect becomes part of the specification. From that perspective, it belongs on the drawing just the same as any other specification.

Matt Lorono, CSWP
Lorono's SolidWorks Resources & SolidWorks Legion
Follow me on Twitter

 

[DeanD3W]

My issue with the "inspection dimension" racetrack frame is not due to concern about inspection information on a drawing... I think it could be termed concurrent engineering, so a good thing, when design, manufacturing, and inspection discuss measurement practices, then specify the agreed upon approach on a drawing (in some organizations that info must go on a separate inspection drawing, but in others the additional overhead of maintaining both a design drawing and inspection drawing is unacceptable).

My issue with the racetrack frame is that it only works for directly toleranced dimensions and if inspection plan information is to be included on a drawing the method needs to be one that works equally well for a tolerance specified by a feature control frame. That method would likely be a symbol placed by the tolerance spec.

By the way, my "under thinking" comment in an earlier post was only used in direct response to what I take as a slightly less than super polite comment about me "over thinking" this topic. I would not normally make such a comment. If my comment is taken out of context then it would seem out of line, so I hope no one does take it out of context .

Dean Watts

http://www.d3w-engineering.com