Basic Concepts Terminology We need a measurement system What does this have 1 Sensor and Transducer Stage 2 Signal Conditioning Stage 3 Output Stage Transducer f T Conditioning Ruler Optical Ultrasound Capacitive Proximity Amplifier Filter What do we get out Output Oscilloscope Camera A D Converter Best Guess for T Nanomechanical Array Measurement System Chemical Biological Detection A lot of optics but the net result is that nanobeam displacement is converted to an intensity change at the CCD and recorded We get a displacement map of the array at each frequency Sensor Nanobeams Transducer Optical System Interferometer Signal Conditioning light level Output CCD Camera Computer Output Before we make measurements we need a plan 1 Parameter Design Plan What is the objective be as specific as possible What are the variables that will affect my results How can I control these variables 2 System and Tolerance Design Plan What is your measurement technique Does it meet the objective in terms of tolerance limits for error 3 Data Reduction Design Plan How to analyze present and use the result Steps required for the result to meet the objective Independent Variables Concentration Cure time Cure temperature Controlled to some extent strength You are trying to make a new high strength light weight composite material by putting carbon nanotubes into a polymer matrix You wish to determine the yield strength as a function of nanotube concentration concentration Dependent Variables Yield strength Extraneous Variables Are not controlled during measurement Relative humidity Particle dispersion operator error Electrical noise interference Room vibration Need to control important variables Extraneous variables can lead to false trends in data and can make it extremely difficult to interpret Adds NOISE or INTERFERENCE to data Measurement accuracy accounts for 0 10F What s going on here Interference adds false trends to the data it seems to make sense but is biased in some way that is hard to understand Sometimes interference can be minimized through randomization What does this mean y f x1 x2 z1 z2 Dependent variable independent variables extraneous variables Choose experimental conditions under which the extraneous variables no longer exhibit a trend Multiple operators Measurements in random instants in time Measurements in random order We often need to specify a given value to a customer Our composites have an effective elastic modulus in the 001 direction of 5 0 0 1 GPa 95 How did we make this measurement 1 A single coupon from a composite slab Repetition can help to reduce the error get a better estimate for the true value Simply measure again Replication repeat the entire experimental procedure Measure on multiple coupons from sample Measure on multiple samples Quantify differences associated with Batch to batch Experimental protocol Position
View Full Document
Unlocking...