EXPERIMENTAL MEASUREMENT:
METHODS AND METHODOLOGY
Strategies and Tactics for Measurements in
16.62x
16.621 Experimental Projects Lab I
BASIS FOR MEASUREMENT
? Whatever exists, exists in some amount
? Our goals:
– Measure the appropriate quantity,
– Measure with the appropriate accuracy,
to allow us to assess our hypothesis
? Some questions:
– What data do we need?
– What do we measure?
– How do we measure?
– How well do we need to measure?
– What do we do with the data?
2
ERROR AND UNCERTAINTY IN MEASUREMENT
? This is the subject of a future class, but the idea is so important that
I mention it today as well
? In engineering the word “error”, when used to describe an aspect of
measurement does not necessarily carry the connotation of mistake
or blunder (although it can!)
? Error in a measurement means the inevitable uncertainty that
attends all measurements
? We cannot avoid errors in this sense
? We can ensure that they are as small as reasonably possible and
that we have a reliable estimate of how small they are
[Adapted from Taylor, J. R, “An Introduction to Error Analysis; The Study of
Uncertainties in Physical Measurements”]
3
SOME GENERAL MEASUREMENT CONCEPTS
? The measurement exists in some context
– We want to do this measurement because…….
? The context (and objective) shapes the measurement strategy and
tactics [“Begin with the end in mind”- Steven Covey]
? You generally do not directly measure the quantity you want
– Most measurement devices are systems which translate the
measured quantity (voltage, for example) into the quantity you
want (pressure)
? All measurements have some uncertainty, or error, connected with
them
– It is important to know not only the result of the measurement but
also the fidelity of the result (how big is the uncertainty, how
much can I trust the answer?)
4
ELEMENTS OF A MEASUREMENT SYSTEM
Measured
Medium
Observer
Variable
Manipulation
Element
Primary
Sensing
Element
Variable
Conversion
Element
Data
Transmission
Element
Data
Storage/Playback
Element
Data
Presentation
Element
Sensing
Signal Conditioning
Readout
Adapted from Ernest O. Doebelin, Measurement Systems: Application and Design, McGraw Hill
5
MEASUREMENT ELEMENT DEFINITIONS
– Primary Sensing element
? Retrieves energy from the measured system
? Produces some form of output
? Strain gage, thermometer tip
– Variable conversion element
? Changes data from one physical form to another
? Elongation to resistance, temperature to volume change
– Variable manipulation element
? Performs mathematical operation on data
? Amplifier, filter
Adapted from Ernest O. Doebelin, Measurement Systems: Application and Design, McGraw Hill
6
MEASUREMENT ELEMENT DEFINITIONS
– Data transmission element
? Gets data between measurement elements
? Wire, speedometer cable, satellite downlink system
– Data storage/playback element
? Stores data for later retrieval
? Hard drive, RAM
– Data presentation element
? Gets data to form detectable by human
? Indicators, alarms, analog recording, digital recording
Adapted from Ernest O. Doebelin, Measurement Systems: Application and Design, McGraw Hill
7
OBSERVATIONS ABOUT THE ELEMENTS
? Two primary functions
– Extract data
– Present data
? Two points of intended contact with the “outside world”
? Functional blocks
– Can have more than one of each type
– Not all measurement systems have all functions
– May be in any sequence (except primary sensing and data
presentation)
8
ERROR ANALYSIS
– Error is inherent in experimental process
Uncontrolled Inputs (Co-Factors)
Discrete Error
Controlled
Inputs
(Factors)
l
s
Humans
Experiment
Amb Cond
Inst Drift
Mate
ria
Mach
in
e
Variati
o
n
Outputs
(Responses)
Continuous Error
9
16.622 MEASUREMENT GOALS
Our goals:
? Measure the appropriate quantity (DO THE RIGHT THING)
? Measure with the appropriate accuracy (DO THE THING
RIGHT)
– This is not only strategy (selecting the needed accuracy)
but also implementation (being able to carry out the
measurement)
– To do this, you may have to learn something about
differerent specific measurement methods
10
16.622 MEASUREMENT GOALS?
[ONE SIZE DOES NOT FIT ALL]
Jonker/Yenson (Prototype Rowing Blades)
– 10% change in oar performance
Diedrich/Smith (Vortex Breakdown)
– 10% agreement with predictions of theory
Pigeon/Whitaker (Near Vacuum Hall Thruster)
– Measureable thrust
– YES-NO Result, BUT result must be able to be
demonstrably different than measurement uncertainties
11
AN EXERCISE CONCERNING THE GREITZER
SPAKOVSZKY 16.62X TEAM PROJECT
Hypothesis proposed by our advisor, Murman: The
thermodynamic loss in the flow through a blade row in a jet
engine is less than 2% of the inlet dynamic pressure
Objective: Go to Pratt&Whitney, obtain a blade row, measure
the loss
Success criterion: Establishing the loss through the blade row
to an accuracy such that the hypothesis can be assessed
? What does all this mean for the measurements we need?
– What would the final “slam dunk viewgraph” look like?
? We want to measure “loss”. How do we do this?
? How well do we need to do it?
12
16.050 TO THE RESCUE
? Entropy = loss (half of what I learned in 16.050)
? How do I measure entropy?
– Who in the department has an entropy meter?
? What do I need to measure?
? Suppose my measurements have an “uncertainty” of + 1% of the
dynamic pressure
– What would be a success?
? Suppose my measurements have an “uncertainty” of + 0.1% of
the dynamic pressure
– What would be a success?
? Suppose my measurements have an “uncertainty” of + 10% of
the dynamic pressure
– Can I succeed?
13
A TURN TO YOUR PARTNER EXERCISE
? For your project:
– List the quantities you need to know to assess yourhypothesis
– Define why these quantities are necessary
? Are these the actual quantities that you will measure? (Is measuring
the voltage output of a pressure transducer the same as measuring
the pressure directly?)
– If not, what are the quantities that you will be measuring?
– If not, what are the ideas, principles, etc. that you will use to
connect the two
? How well does the measurement have to be made?
14
A TURN TO YOUR PARTNER EXERCISE
? Pick a variable that is important for your experiment
? Describe the:
– sensors
– Instrumentation
– measurement chain
you will need to obtain the data.
15
UNDERSTANDING MEASUREMENT DEVICES
Devices are specified by stating things such as…
– Accuracy and Precision
– Static Sensitivity
– Zero Drift and Sensitivity Drift
– Linearity
– Resolution
– Threshold
– Hysteresis (dead band)
– Readability
– Span
– Dynamic Performance
16
ACCURACY AND PRECISION
Accuracy of a measurement is a measure of how close the result of the
experimen is to the true value
Accuracy is a measure of the correctness of the experiment
Precision is a measure of how well the result was determined, without
reference to its agreement with the true value
Bevington, P. R. and Robinson, D. K, 1992, “Data Reduction and Error
Analysis for the Physical Sciences”
17
MEASUREMENT ACCURACY AND PRECISION
Accurate?___ Accurate?___
Precise? ___ Precise? ___
Accurate?___ Accurate?___
Precise? ___ Precise? ___
19
CALIBRATION
? Once a measurement device is selected, it must be
calibrated
– Calibration – Comparison of instrument’s reading to a
calibration standard
– Calibration standard created from a measurement
? Inherent error
– Rule of thumb: Calibration standard 10x accuracy of the
instrument being calibrated
? Basic issue is how do we know that what we record has any
relation to what we wish to measure?
20
SENSITIVITY
? Quantity of output for each unit of input
– For 1 unit of input:
? 10 units of output = more sensitivity
? 1 unit of output = less sensitivity
? Sensitivity is not accuracy!
21
STILL MORE……….
– Resolution
?
– Threshold
?
Smallest smallest
detectable change
Smallest detectable value
Output
– Hysteresis (dead band)
Input
? Loop resulting from
complete calibration
using monotonically
changing values
Input
Output
22
MEASUREMENT METHODS
? Ask your advisor
? Ask lab staff
? Texts or articles
– e.g., Beckwith, T. G., Mechanical Measurements, Addison
Wesley
– e.g., Ernest O. Doebelin, Measurement Systems: Application
and Design, McGraw Hill
23
MEASUREMENT METHODS: MIT Video Tapes
? Introduction to Measurement
? Calibration, Accuracy and Error
? Mass, Force, Strain, Torque and Pressure
Measurement
? Measuring Dynamic Variables
? Fluid Quantity and Flow
? Infrared Temperature Measurement
? Contact Temperature Measurement
Available in Aero-Astro Library for Viewing
24
SOME GENERAL MEASUREMENT CONCEPTS
? The measurement exists in some context
– We want to do this measurement because…….
? The context (and objective) shapes the measurement strategy and
tactics [“Begin with the end in mind”]
? You generally do not directly measure the quantity you want
– Most measurement devices are systems which translate the
quantity you measure (voltage, for example) into the quantity you
want (pressure)
? All measurements have some uncertainty, or error, connected with
them
– It is important to know not only the result of the measurement but
also the fidelity of the result (how big is the uncertainty, how
much can I trust the answer?)
25