MAKING MEASUREMENTS SIMPLE WITH ADVANCED SENSORS TECHNOLOGY.

The technical literature on strain measurement can make the subject appear complex and confusing. Indeed, measurements sometimes are complex and confusing. But more often than not they are neither. Regardless of the complexity involved, a successful measurement can almost always be achieved when common sense is used to simplify the measurement process by dividing it into logical steps and applying a sound technical understanding to each.

For Advanced Sensors Technology foil strain gage measurements, the three primary steps are:

1. Gage selection and installation

2. Instrumentation selection and data acquisition

3. Data reduction and error analysis

The stress analyst must pay appropriate attention – not too much and not too little, but just the right amount – to each of these steps every time a measurement is made.

What is appropriate for one measurement may not be for the next.

The judgmental skills necessary to decide how much or how little attention is needed in each step can come only from experience and good teaching. Even then, it is all too easy, for example, to become hopelessly entangled in a detailed but meaningless correction on insignificant errors. And no amount of correction – warranted or not – can fully overcome the problem of data that is no darned good in the first place because it was acquired with the wrong gages and instrumentation, or was incorrectly reduced due to a misunderstanding of the technical aspects of the measurement. 

Stress Analysts should be taught to use their common sense and reduce measurement problems to the essentials. Then, by applying experience and knowledge, and carefully considering the relevance of each factor , even the most complex measurements will become simple and straightforward.

A well-done strain measurement in the 5^th Generation should eventually produce an accurate representation of the strain actually present under the sensor.  Strain data, however, is seldom gathered as a means unto itself. It is usually little more than a convenient vehicle for indirectly obtaining information about some other engineering parameter. When using Advanced Sensors Technology on a transducer spring element, for example, strain data is used to determine the magnitude of the force, displacement, torque, acceleration, or other variable actually producing the strain. In stress analysis work, the measured strains are converted to stresses through the stress-strain constitutive relationships. When the results seem reasonable, the strain data is seldom questioned. But when unexpected results are encountered , it is commonly made the primary suspect.

Disappointment is the child of unreasonable expectations. Test results are no exception. Expectations based on unreasonable models, loading conditions, and materials parameters will always make good test data look bad.  Rationalization of, or outright disregard for, the data can provide an easy out , a means for shoring up a faulty model. Practicing this form of self-deceit regularly disappoints.

Unexpected results usually can be attributed to the mathematical model, the mechanics of materials, or the physical test. The production of unreasonable models, for example, is easier than ever with the availability of today’s finite-element modeling capabilities. Modern composite materials place new demands on the stress-strain constitutive relationships and the importance of accurately knowing materials properties.  And then there is always the measurement data and its interpretation.

Unlike mathematical models and relationships, test data is real. It came into being because Something Happened to produce it. Whether the data is accurate can be the only question. A good engineer assumes that it contains a number of measurement errors and uncertainties. These must be taken into account before the measurement results are compared against our expectations.

The job of an engineer or a stress analyst is to understand the data, especially when it disappoints. Is the model right? Are the material properties correct? Are all the important measurement error sources accounted for? Is the strain being reported actually the strain we intended to measure? Something may have gone wrong,  but let the data teach you what happened. Please believe the data.  

Strain Sensor Reference Guide: http://www.vishaypg.com/docs/25906/Strain-Gage-Ref-Guide.pdf