Let me start with a confession: I’m an imposter in the metrology community. My background is in the design engineering community—well-meaning folks who don’t quite grasp the complexities of the manufacturing process and the measurement systems that support it.
Measurement people know the effect of arbitrary decisions in the design process. We designers put 0.005-in. tolerances on a part because it’s as easy as a mouse click in CAD. Any engineer should understand that most materials expand when heated and contract when cooled. Poll design engineers and ask what temperature their CAD models were when they applied the dimensions. Many will stare at you with a blank look.
My ignorance, however, provides a unique perspective on some aspects of the measurement business. For example, can someone explain the logic of applying linear temperature compensation to large-volume coordinate metrology of nonhomogeneous assemblies? I’m talking about airplane wings, car bodies, windmill blades, and other large assemblies made out of various materials fastened together, carrying the inherent manufacturing tolerances of their components. I have discussed the issue with experts from three national measurement institutes, and still can’t make any sense of it.
I understand the principle of applying compensation to measurement data to account for the effects of varying temperatures of the workpiece and measurement system. Most materials behave predictably when the component is subject to a temperature change. A scale bar made from a homogeneous material will predictably expand with increasing temperate at a ratio defined by the coefficient of thermal expansion for the material. Because a scale bar’s entire function is to define a traceable link to a reference standard, applying temperature compensation of the scale bar is both logical and necessary.
Given the same concept, let us consider a collection of parts made mostly from aluminum, with some carbon fiber composites, titanium, and steel bits. They are assembled using two handfuls of fasteners and some adhesive. Now let’s go out into a building that is at 80°F and collect some surface points from the assembled structure using our impressive CAM toolkit. Compare that measured data to an engineering intent expressed by one of the before-mentioned design engineers via a geometric dimensioning and tolerancing (GD&T) drawing. You have explained to us designers what is stated in paragraph (l) of the ASME Y14.5-2009, Section 1.4—“Fundamental rules”: “Unless otherwise specified, all dimensions and tolerances are applicable at 20°C (68°F) in accordance with ANSI/ASME B89.6.2. Compensation may be made for measurements made at other temperatures.” Therefore, some would argue that before this comparison can take place, you must apply temperature compensation to the measurement data. Measurement software can help us apply a compensation for aluminum of 0.0000232 because it’s as simple as a click of our mouse. Of course, coordinate data collected from a free-form surface on a multimaterial structure expands and contracts linearly with a temperature change.
If you didn’t catch the sarcasm in the last statement, let me be more blunt: As measurement professionals, we too can be lulled into blissful ignorance by the automation provided by measurement tools. The bright side is that these issues can be mitigated. The more experienced measurement professionals among us were brought up on first-principle measurement tools. They have the background necessary to question the answers provided by the automated tool and make sense of disparate data. Unfortunately, many of us started in measurement with the software giving us all the answers, and thus we have little or no first-principle knowledge to fall back on when the answers don’t make sense. To fill the gap, this community must find ways to transfer knowledge from the experienced experts to the next generation of practitioners.
The Coordinate Metrology Society (CMS) is committed to the growing 3D measurement workforce and continues to make progress toward a certification for industrial metrology practitioners. An Academic Committee was formed last year to represent the universities and community colleges that support our industry through research, coursework, or industrial outreach. This year, the society hosted its first Metrology Careers Fair on Thurs., July 19, 2012, at 10 a.m.–11 a.m., at the 2012 Coordinate Metrology Systems Conference (CMSC), which was held this week, July 16–19, 2012, at the The Roosevelt Hotel in New Orleans. The Careers Fair is designed to generate interest in the field of metrology and to engage and educate young engineers, measurement novices and experts, and industrial professionals looking for a new career path.
The CMS has also hosted a series of measurement studies at its annual conference. The 2012 CMSC measurement study will focus on the importance of practical testing. The results will be analyzed by the CMS Certification Committee and used to support the first industry-recognized, level-one personnel certification in portable 3D metrology, scheduled for completion by the end of 2012. The CMS is also conducting pilot certification examinations as a final step toward a metrology certification program.
These efforts are a good foundation for knowledge sharing within the community. By instituting the same questioning culture in our measurement community that we request from design engineers, we can have a significant effect on our ability to make efficient use of automation without losing our ability to think our way out of a problem.
Talion Edwards is an associate technical fellow in 3D imaging and reverse engineering and is the principle investigator for the Boeing Research and Technology reverse engineering development task. He also serves as co-chairman of the Boeing Enterprise 3D Imaging Community of Practice, for which he coordinates a yearly workshop held in conjunction with the Coordinate Measurement Systems Conference (CMSC). Edwards was the CMSC chairperson during 2011. He continues to serve on the executive committee of CMSC and is leading the effort to establish a central certification for industrial metrologists.
Comments
Nice article
First of all, let me say it's a nice and interesting article, especially liking the first few paragraphs :-)
As for your question, I am no experts, but going KISS leads me to think this:
1) There is two tempratures to compensate for.
- One is of the measuring defice itself and one is for the product. A metal ruler for example will have a different scale size at 100C then it has on 20C, so the actual value needs to be calculated back to the 20C. This should go for every measurement tool
- Secondly is the workpiece itself. If like you suggested, it's a homogenious temprature and material, or a serial connected chain of materials with known tempratures, the calculation back from your measured size, known materials and tempratures, it can still be easily calculated, as the behavior is still linear.
- When like you suggest, it is a combination of materials in a complex shape, one would think it becomes more complex to measure. However, if the design is know, the calculation is still possible, because each individuel element will behave linear, but because of the combination, the deformation will not be in a linear direction, but freeform, which will required 3D compensation in the measurement data.
Only in case the design of a mix of complex shape is not known, you cannot calculate and compensate... which is logical, as you won't know the coefficent for the material in the first place.
My Conclusion => Temprature compensation for the measuring tool makes sence, as to ensure that the same absolute size is found with a measurement. The distance from paint A to point B in a space.
Where this point A and B is on the actual product, will change, due to the temprature. That compensation is mathmatical and can only be done once one knows the design of the part. If the design of the part is known however, one should be able to compensate for any and all changed, regardless of the mix of materials used.
If you disagree with any of the above, I'd love to hear it and why.
Temperature compensation in composite assemblies
If you are making just one or two, temperature soak them to 68F/20C.
If you are making lots of them, the assembly variation in the final dimension will probably be more significant than the thermal variation...
Clear and insightful