Metrology and Inspection

Course Objective
To provide the basic understanding of :
➢ Measurements
➢ Tolerancing
➢ Design and understanding of different
manufacturing methods ( Conventional and
non-conventional)
Feel for manufacturing:
➢ What machine at what situation?
➢ What is possible by existing manufacturing
methods?
➢ Cost effective routes!
➢ Feel for numbers!
Gas Turine Engine Layout Radial Compressor

Combustion Chamber Turbine Wheel

Suggest Processes to make a Turbine blade?

http://www.visitec-em.de/turbine_blade.html
 At 500/550 pound in weight .
The biggest Gravity Die Casting in The World at the Time
http://www.rainwater.demon.co.uk/bertha.htm
Weight 320 Tonnes.
http://www.materialsforengineering.co.uk/engineering-materials-
news/sheffield-produce-worlds-biggest-casting/74805/
 Hole drilled into human hair to
demonstrate positional accuracy
of Kern Nano CNC

Microdrill with diameter of 60 µm

http://www.visitec-em.de/microdrill.html
 Laser Micromachining

Sculpted a bull roughly the size of a single red blood cell (10 µm long by 7 µm
high) using two infrared laser beams
S. Kawata et al., Nature 412, 697 (Aug. 16, 2001).
Laser Focus World October, 2001
 Example
Silicon Guitar

• The world's smallest guitar is 10 micrometers long –

• about the size of a single cell -- with six strings each about 50 nanometers, or 100 atoms, wide. Made by
Cornell University researchers from crystalline silicon, it demonstrates a new technology for a new
generation of electromechanical devices. Photo by D. Carr and H. Craighead, Cornell.The above image
(508 x 327 pixels) is the digital image created by the electron microscope, and is the highest-resolution
version available.

8
 Example: How much will it cost?

Diameter: 100 +/- 0.00001um

Length 300mm.
Roughness 0.001um

million dollars !!
 Designer

Producibility
Mechanical
Engineer
(Manufacturability)

Quality Assembly
Assurance

Producilibility → Success of a Design

(Technical, Economical)
How to ensure what we produce is what we want ?

Inspection
 Manufacturing Processes

Measurement: Fundamental exercise of inspection; Act of measuring or being

measured.
Why measure? Manufactured parts correspond to the specs of the product:
(QA) Confirm functionality, performance, reliability, process capability, etc.
How to measure? Attributes ( Qualitative, go or not-go, Gaging, decisions)
Variables (quantitative, dimensions, useful for analysis and decision)
What to measure? Size/geometry of tools
Size/geometry of a part from a machine tool
Benefits Determines capability of a process
Indicates the need of maintenance
Feedback of manufacturing
 Attribute Vs Dimensional

➢ Qualitative ➢ Quantitative
➢ Fast and economical ➢ Slow and expensive
➢ Pass or fail ➢ Exact dimension is
needed
➢ Mostly for standard and
less severe applications ➢ Useful for highly reliable
(Automobile) applications (Aircrafts)
➢ Useful after process ➢ Needed for
development development
➢ Large production ➢ Small production
volume volume
Standard Measurements SI
What tools do we use for calibrating
measurements and inspection ?
 Length Standards: Gage blocks or slip gages

Standard set of rectangular gage blocks with ±0.000050-in.

Wrung-together gage blocks in a Screw gage blocks wrung together to build

special holder, used with a dial gage up a desired dimension.
to form an accurate comparator.
Standard gages of meter – exist in any workshop ( standard blocks); need to calibrate every
specific period ( various precision).
 Accuracy Vs. Precision
Accuracy: ability to reach the nominal size
Precision: repeatability of result

(Left):Accuracy versus precision. Dots in targets represent location of shots. Cross

(X) represents the location of the average position of all shots.
Allowance and Tolerance
 Allowance and Tolerance
Allowance: intentional desired difference between the dimensions of 2 mating parts
Tolerance: undesirable but permissible deviation from a desired dimension.
Most manufacturing processes →products with dimensions distributed normally (clustering
around the average)

n
 xi
x= ( )
n
i =1 2
 xi − x
n  = i =1
n
When mating parts are designed,
each shaft must be smaller than The manner in which the distributions of the
each hole of a clearance fit. two mating parts interact determines the fit.
UNTL upper natural tolerance limit =  +
LNTL, lower natural tolerance limit =  − 

Wear or lack of process control → shifting

 Clearance?
Range of Fit?
MMC?
LMC?
Bilateral?
Unilateral?
Limits
 How to Specify Tolerances?
ANSI – 8 classes of fits
•Class 1. Loose fit: large allowance. Accuracy is not essential.
•Class 2. Free fit: Liberal allowance. For running fits where speeds are above 600 rpm and
pressures are 600 psi ( 4.1 MPa) or above
•Class 3. Medium Fit: Medium allowance. For running fits below 600 rpm and pressure below
600 psi ( 4.1 MPa) and for sliding fits.
•Class 4. Snug Fit: Zero allowance. No movement under load is intended, and no shaking is
wanted. This is the tightest fit that can be assembled by hand
•Class 5. Wringing fit: zero to negative allowance. Assemblies are selective and not
interchangeable.
•Class 6. Tight fit: slight negative allowance. An interface fit for parts that must not come apart
in service and are not to be disassembled or are to be disassembled only seldom. Light pressure
is required for assembly. Not to be used to with stand other than very light loads.
•Class 7. Medium force fit: an interference fit requiring considerable pressure to assemble;
ordinarily assembled by heating the external member or cooling the internal member to provide
expansion or shrinkage. Used for fastening wheels, crank disks, and the like to shafting. The
tightest fit that should be used on cast iron external members.
•Class8. Heavy force and Shrink fits: considerable negative allowance. Used for permanent
shrink on steel members.
 ISO System of Limits and Fits
▪ Clearance fits
▪ Transition fits/ Location fits/ Assembly fits
▪ Interference fits

What is Hole based or Shaft based

system?
WHY?
ISO System of Limits and Fits

Shaft-basis and hole-basis system for specifying fits in the ISO system
 Geometric Tolerances
Permitted tolerance on shape/geometry/form/position

MMC – Parts are made with the largest amount of material possible
LMC - Parts are made with the least amount of material possible
RFS – Regardless of feature size
Datums- Concept/Feature common for design, manufacturing and inspection

(Left) Geometric tolerancing symbols; (Right Up) Feature control symbols for
part drawings; (Right Down) Example of use of geometric tolerancing
(tolerancing for flatness)
 Inspection methods for measurement
Metrology: measurement laboratory selected according to certain criteria:
• Gage capability ( rule of 10)
Measuring device has to be 10 times more precise than the tolerance measured:
Eg. +/-0.001 → +/-0.0001 → +/-0.00001
•Linearity
Linear working range (Input Vs Output)

•Repeat accuracy
Repeatability of the measurement

•Stability
Retaining calibration over time, no-drift
•Magnification
Amplification of the output portion of the device, bigger dials.
•Resolution
Sensitivity; smallest input value that can be detected or measured
Measurement instruments ( linear)
➢ Ruler (0.5mm)
➢ Combination set
➢ Vernier Caliper(0.01mm)
➢ Micrometer caliper
➢ Optical Comparators (0.001mm)
➢ Laser/ interferometers(0.0001mm)

Combination set
Three styles of calipers in common use today Digital Micrometer for measurements
from 0 to 1in., in 0.0001-in. graduations.
Internal and external
Vernier Caliper
Micrometer
Vision Systems of measurement Coordinate Measuring Machines
(CMM)

Optical Comparator, measuring

the contour on a workpiece.
Coordinate measuring machine
Digital indicators with in/.mm
with inset showing probe and a
conversions add to the utility of
part being measured.
optical comparators.
 Interference bands can be used to measure the size of objects to great accuracy

a a)(bottom left) Calibrating the X-axis linear table displacement of a vertical spindle
Resolution ~ 10nm milling machine; b) ( top right) Schematic of optical setup ; c) ( bottom right)
Schematic of components of a two frequency laser interferometer.
 Angle Measurement

▪ Sine Bar: 1sec of arc

Setup to measure an angle on a part using a sine

bar. The dial indicator is used to determine when
the part surface X is parallel to the surface plate
Gages for Attributes ( mass production)

(Top) Plain plug gage having go member on

one end and not-go member on other other;
(Bottom) Ring gage with stepped go and not-
go member

Go and not-go ring gages for checking a

shaft.
 Surface Roughness Measurement Instruments

Schematic of stylus profile device for measuring

Schematic of surface profile as produced by
surface roughness and surface profile with two
stylus device showing some typical y values
readout devices shown: a meter for AA or rms
with respect to the center line.
values and a strip-chart recorder for surface profile.

Top) Microtopographer, a stylus profile device

used to measure and depict surface roughness
and character (surface profile); Bottom)
Typical surface-roughness profiles.
 Above left) Terminology used in specifying and measuring surface quality;
Above Middle) Symbols used on drawing by part designers, with definitions of symbols;
Above Bottom) lay symbols;
Right) lay symbols applied on drawings
Comparison of surface roughness produced by common production processes
 Everything is Connected!

➢ Tolerance, Roughness and process are

interconnected.
➢ Selection of one determines the other.
➢ Eg: +/-0.5mm, Ra=0.1um → no sense
➢ +/- 0.01mm, Ra=25um → no sense
Nondestructive Inspection and Testing

Destructive Non-Destructive
Quantitative results Qualitative results
Do not require Skilled interpretation of
interpretation results

Restricted for costly Low material cost

measurements
High cost equipment Low cost equipment
Consistent results More subjective results
 Visual Inspection
Liquid (Dye)-Penetrant Inspection (LPI or DPI)

To be done prior to surface finishing or

surface modifying operations such as:

shot peeing, polishing, etc.

 Visual Inspection
Magnetic Particle Inspection
➢ Only for Ferromagnetic Materials
(Fe, Ni, Co Alloys)
➢ Surface and sub surface flaws
➢ Orientation dependent,
Perpendicular to the lines
➢ Parallel defects will not be
detected
 Ultra-Sonic Inspection: Deep defects

Left) Ultrasonic inspection of a flat plate with a single transducer;

Right) Plot of sound intensity or transducer voltage versus time showing the initial pulse and echoes from the
bottom surface and intervening defect.

A B

A) Dual transducer ultrasonic inspection in the pulse-echo mode;

B) dual transducers in through-transmission configuration.

Ultra-Sonic Inspection: Deep defects

www.triquestndt.com/conventional/ultrasonic-testing
Radiography: Xray, Gamma Ray, Neutron
Beam

Full-size radiograph of
the Liberty Bell. The
photo reveals the
famous crack, as well as
the iron spider installed
in 1915 to support the
clapper and the steel
beam and supports
which were set into the
yoke in 1929.
Radiography: Xray, Gamma Ray, Neutron Beam

Preparing Gas line for X-rays Stress Crack in Casting

➢ X-ray and Gamma ray radiation

sources better for denser
materials

➢ Neutron beam radiation better for

plastics, rubber and hydrogenous
materials.

Shrinkage Crack in Casting

 Eddy Current Testing

➢ Conductive specimen
➢ Relation of the magnetizing coil,
magnetizing current, and induced
eddy currents. This dynamic magnetic
field induces the eddy currents and
the changes in the eddy currents Eddy currents are constrained to travel
produce a secondary magnetic field within the conductive material, but the
which interacts with the sensor coil or magnitude and path of the currents
probe will be affected by defects and
changes in material properties.
Eddy Current Testing
➢ Sensor detects changes in impedance and
phase angle of coil caused be defects in test
specimen
Factors to consider:
➢ Magnetic permeability of test material
➢ Frequency of AC current used to energize coil
➢ Coil geometry
 Other Non-Destructive Test Methods

➢ Acoustic emission monitoring (Active Flaws)

➢ Leak testing (Bubble test, pressure drop test, advanced)
➢ Thermal testing (Composites, electronic devices)
➢ Strain sensing (Brittle or photoelastic coating, strain gage)
➢ Advanced optical methods (Hologram, hologramic interferometry)
➢ Resistivity methods
➢ Xray Computed Topography (CT scan)
➢ Acoustic holography (Ultrasonic Sound)
Process Capability and Quality Control
Example of process
(Casting)
 How a process can get out of control?
1. Wear
2. Inherent errors
3. Heating
4. Vibrations
5. Fault measurement/settings

Solutions:
Shift the job
Relax the specs
Sort the products
Improve precision:
switch the cutting tool switch the work holding device
change the material overhaul the process
find (eliminate cause of variability)
 Inspection and Quality Control (QC):
How much one should inspect
100% selected items none
Cost

Ongoing manufacturing Destructive tests

Too large production →too expensive
→ Statistical Process Control

Histogram of a process to compare measured sizes with designed Basic design of the X’ chart, R chart and  chart used
specifications in SPC.
Process Capability and Quality Control
Any process exhibits a level of inherent variability →

inherent capability

Process Capability : Ability of the process to yield consistently

the desired output.

How to measure/quantify the Process Capability (PC)?

 Process Capability and Quality Control

A. Inspection * to find defects

* to prevent defects

B. Analysis * the nature of the process based on a

sufficient amount of data
The basic principle of quality control –
quality improvement without cost increase
Steps to take:
1) Design an experiment:
machine, settings, speeds,
cutting rates, feed, material…
2) Define the inspection
method
3) Define the sample size
4) Separate parameters from
noise
5) Take measurements
 Process Capability Index, Cp
Does the process meet the specifications?

toleracnce− spread USL − LSL C p  1.33, acceptable

Cp = =
6 ' 6 '

Is the process centered (not shifted)?

estimated mean − nom. x − xnom
D= =
1 1
tolerance spread (USL − LSL)
2 2
1
X - Nominal + 3  USL − LSL For Acceptance
2

Measure of the capability?

C pk =

min USL − X , LSL − X 
Capability index
3 '
Process Capability Index

What PC studies tell us about the process.

Process Capability Index
Statistical Process Control
 Quality Control Charts:

Statistical quality control charts

for 12.700mm (0.5000-in.)-
diameter pins. Note the trend in
the X’ curve before the tool
change, indicating that the mean
was shifting due to tool wear.

UCL- upper control limit x + 3 x= x' + A2 R

LCL-lower control limit x − 3 x= x' − A2 R

Step1:
Select the sample size, n=5
Inspect the dimensions
Estimate X ,  and R
(Sample mean, SD, range)

Step2:
Do the step on k=14 samples.
Estimate X ,  and R
(grand average, SD, Range)
Samples of Chart Patterns

Examples of X’ control charts

Assume that you are incharge of the process development
to develop a process
Aim: 16h11 dia 15.890/16.000

Nominal Dia= Average(15.89, 16)=15.945

Measured values

Sample1 Sample2 Sample3 Sample4 Sample5

15.89 15.9 15.96 15.9 15.895
15.9 15.93 15.92 15.93 15.935
15.98 15.97 15.98 15.97 15.95
16 15.97 15.91 15.94 15.92
Sample
Average X 15.9425 15.9425 15.9425 15.935 15.925
Sample
Range R 0.11 0.07 0.07 0.07 0.055
sigma,  0.055603 0.034034 0.03304 0.028868 0.02345
X = X 15.9375
3 x = A2 R = 3
R
= 0.73 * 0.075 = 0.0548
5 d2 n

where n = 4
d2 and A2 are given in Fig.12.13

R=
 R
5 0.075

 =
 
5 0.034999

'  0.035
 = == 0.0438 c2 = 0.8 for n = 4
c2 0.8

UCLx = X + 3 x = 15.938 + 0.055 = 15.993

LCLx = X − 3 x = 15.938 − 0.055 = 15.883

UCL R = D 4 R = 2 .28 * 0 .075 = 0 .171

LCL R = D3 R = 0 * 0 .075 = 0
 X Bar Variation UCLx
LCLx
16.02 Actual Mean
Nominal
16 LSL
USL
15.98

15.96
X Bar

15.94

15.92

15.9

15.88

15.86
0 1 2 3 4 5 6
Sample #

Is the process ACCEPTABLE?

 R Variation

0.18

0.16

0.14 LCLR
Grand R
0.12 Sample R
LCLR
0.1
R

0.08

0.06

0.04

0.02

0
0 1 2 3 4 5 6
Sample #
Sampling Errors
Seven Tools for Lean Production

➢ Histogram
➢ Run Charts
➢ Check Sheets
➢ Cause and Effect Diagrams
➢ Scatter Diagrams
➢ Pareto Diagrams
➢ Process Flow Charts
Histograms
Run Chart
Check Sheets
Cause and Effect Diagram
Scatter Diagrams
Pareto Diagrams
Process Flow Chart
 Six Sigma Model

USL-LSL 12

x  6 '
 68.26%
 2 95.46%
 3 99.73% 7 out of 10,000
 6 99.9999 3.4 ppm
 Taguchi Method
Create an orthogonal array of experiments to determine the
dominant inputs to variability of a process.

The use of Taguchi methods can reduce the inherent process variability as shown in the upper
figure; factors A, B, C, and D versus process variable Y shown in lower figure
 Taguchi Methods
1. Quality is defined in relation to total loss to consumer
2. Society requires constant improvement for business growth
3. Quality improvements require continuous reduction in
variability … increase precision
4. Quality and cost of manufactured goods are determined by
engineering design and manufacturing system
5. Variability in product/process performance can be reduced by
exploiting non-linear effect of process parameters
6. Statistically planned (Taguchi) experiments can be used to
determine parameters that reduce performance variations
7. Design and improvements of product/process can make them
more robust, i.e, less sensitive to uncontrollable
variations…noise!