WHAT METALLURGIST

MUST KNOW ABOUT

GRAIN SIZE
ASTM E 112
A STEP BY STEP GUIDE for getting
fast and accurate results
using an Image Analysis Software.

PAULO CESAR IBAÑEZ H.

- IMAGE ANALYSIS EXPERT-

A Publication of
Table of
Contents
Introduction

Chapter 1 – Basics - Grain geometry

Chapter 2 – TOP 3 - Measurement methods

Chapter 3 – Method #1- Comparison

Chapter 4 – Method #2- Planimetric

Chapter 5 – Method #3- Interception

Chapter 6 – Automatic measurement with

an Image Analysis Software
Table of
Contents
Chapter 7 – Relative accuracy and
Confidence interval

Chapter 8 – Conclusion

References
 INTRODUCTION
What metallurgist must know about Grain Size.

This e-book has been created to provide a practical guide from

non-experts to highly experts metallurgist when need to measure
grain size according to ASTM standards.

The objective of this guide is to help experienced metallurgists, to

be confident using the new digital tools, and for the new
generations to have a better understanding of the basics.
Let´s say that for experienced metallurgists is the way to be
updated and for younger metallurgists is the way to get all the
experience behind this measurements.

This is what you get with this guide:

• You will understand the principles of measuring grain size.

• You will be able to identify a problem when measuring.
• You will be able to measure the grain size manually, semi-
automatic and full automatic.
• You will be able to explain results.
• You will be able to ensure the quality of your measurements.

So let´s start…
 1
CHAPTER ONE

Uunderstanding
grain size geometry
 Grain size geometry
“Shapes must be space filling.” Plateu 1873

What is a grain?
This is the first question that comes to our mind and the answer is
quite simple:

“A grain is a particle with a specific geometry”

The ideal grain geometries are:

Tetrakaidekahedron with 14 faces, 24 corners and 36 edges and
the Dodecahedron is a polyhedron with 12 faces.

The polyhedrons are geometric bodies which flat faces enclose a

finite volume.

Tetrakaidecahedron: Dodecahedron
14 faces 12 faces
 GRAIN SIZE GEOMETRY
What does it mean ?
It means that a geometry body to be considered as a grain, it must
comply with the following requirements:

• Shapes must be space filling.

• Surfaces must exhibit minimum surface area and
minimum surface tension. Plateu 1873.

As you can see in figure A), there are a group of individual grains,
and because of their geometry, it can be joint to become a bigger
particle.
So a metal structure is formed by grains.

Figure A)
 Why geometry is important?

Geometry is important because the ASTM E 112 standard

provide measuring procedures to determine the average
grain size in metallic materials, and the procedures are
geometric based.

So are independent of the metal or the alloy, because the

measurement procedures could be used to the grain size
estimation for crystal or cell size for non-metallic materials
as well.

This is stated in the introduction of the ASTME112

standard.
 2
CHAPTER TWO

Top 3 Measurement
methods
 TOP 3
Definition of ASTM GRAIN SIZE

In order to measure the grain size, is necessary to have a scale that

represents the grain size.
Imagine giving a result of grain size in microns, or in inches, or
millimeters, is kind of complicated don´t you think?

ASTM defines grain size as: n= 2 G-1

n= number of grains / in2 at 100X
G= ASTM GRAIN SIZE NUMBER

To make it easier, here is a table with a Grain size number and the
number of grains / in2 at 100X

G n G n

1 1 6 32

2 2 7 64

3 4 8 128

4 8 9 256

5 16 10 512
 TOP 3
Definition of ASTM GRAIN SIZE

Now if you are asking, how could I measure it?

Current ASTM E 112 standard describe the use of these 3

measuring methods :

1. Comparison charts.
2. Jeffries Planimetric Grain Size.
3. Heyn / Hilliard / Abrams Intercept method.

Which method should I use?

We will explain in detail how to measure the grain size in the next
chapters, but before measuring you have to be sure if the
equipment available in the lab can help you or not.
Because you need to prepare the sample and then analyze it.

For prepare your sample you will need a cutter, a mounting

press, a grinder and polisher and a fume hood for etching the
sample.

Keep reading…
 TOP 3
Definition of ASTM GRAIN SIZE

To analyze the sample you will need a microscope with

reflected light ; it could be Upright or inverted; but make sure the
microscope has a 10X objective and 10x eyepieces, to have a 100X
magnification.

Also make sure that the microscope can allows you to place a
reticle on the eyepieces, or there are some brands and models that
you can install a grain size chart between the light path , so you an
see the grain size image on the eyepiece and then you can compare
it with your sample.
With different type of reticles you can measure manually all the 3
methods.

But nowadays the use of digital cameras for microscopes is more

common, so consider to have an Image Analysis System with your
microscope.
Upright microscope
– Reflected Light

Inverted microscope
– Reflected Light

“ Inverted microscope is mostly used on metallography,

because the sample is placed down face and that “
surface is flatter and also because there are no height
limit as with the vertical microscope.

Pictures courtesy from

Clemex Technologies.
 Top 3
Comparison charts – Planimetric - Interception

Comparison Method
The most simple and easy way to measure the average grain size,
is using the comparison charts.

You only have to look on the eyepiece and compare the image you
see there with the standard charts, that´s it.

You can also use the grain size reticle on the eyepiece to have a
more accurate comparison, because the grain size standard charts
are printed in the reticle, and you can compare it while observing
on the eyepiece.

Is fast and easy, but is only for the average grain size, not for
individual grains.
 Top 3
Comparison charts – Planimetric - Interception

Jeffries Planimetric Grain Size

With this method you will know the Grain Size number from
calculating the average of the grain area.

The measurement is done in 2 steps.

Step #1 is to calculate the average grain area in mm2
You can do that counting the intercepted grains for the reference
circle and counting the entire grains inside the reference circle.
Step #2 Calculate the ASTM Grain Size Number, using the value
of the average area. (We will give more details in the next chapters)
 Top 3
Comparison charts – Planimetric - Interception

Heyn / Hilliard / Abrams Intercept method

With these method, you have 2 ways to measure. Using a straight

line you can calculate the grain size :
#1.- Counting the grains intercepted by the line.

#2-. Counting the grain boundary intersections.

 Top 3
Comparison charts – Planimetric - Interception

Heyn / Hilliard / Abrams Intercept method

Now, If the structure is no equiaxed , or the shape shows some

distortion, you may use 3 concentric circles like these:

But you have to define if you are going to measure the grains
intercepted or the intersections of the grain boundaries.
 3
CHAPTER THREE

Comparison method
 Comparison method
Method #1

As we say before you only have to look on the eyepiece

and compare what you see on the eyepiece VS the
standard charts.

How could you be sure is the right number?

That´s one of the main complications about this method, because
it becomes subjective to the person who is comparing.

And is even more subjective when the grains are smaller and there
are slight differences in size .

Imagine a structure where there are 55% of grains are size 8 and
45% are 7, complicated right?

Because as you can only compare the structure as an average, you

can not know about the percentage of each size of grain, and you
have to made a decision , if you qualify it as 7 or as 8.
 Here is what ASTM E112 says about the comparison
method.

“ Grain size ratings using the comparison chart method

by an individual metallographer will vary within +/- 0.5 G
units.
When a number of individuals rate the same specimen, the
spread in ratings may be as great as 1.5 to 2.5 G units.”

ASTM E 112 - Section 19.8

 4
CHAPTER FOUR

Jeffries Planimetric
method
Jeffries Planimetric method
Method #2

Getting the ASTM grain size number by considering the

grains quantity within an area.

Now you know some of the limitations of the comparison method,

which is easy to apply, but is subjective and affected by the
metallurgist criteria.
Here is a more objective method to know the average grain size.

With this method you can get the grain size number in 2 steps.

Step #1: Calculate the average grain area in mm at 100x

magnification. (N A = Number of grains per mm2)

N A = f ( N1 + ( N2 / 2))
f = Jeffries multiplier ; f = Mag 2/ circle area
N1= Number of grains completely inside the circle.
N2= Number of grains intercepting the circle.
Jeffries Planimetric method
Method #2

Nothing better than a an exercise.

Considering the image below as reference for the exercise:

Where: N1 = 68 and the N2 = 41

The Circle area (A) = 20106.2 mm and Magnification = 100X
Replacing the numbers in the equation we can get this:
NA = f ( N1 + ( N2 / 2))
NA = f (68+(41/2)) NA = f (88.5)
f = Mag 2 / A f= (100 2 / 20106.2)=0.497
NA= 0.497 (88.5) =44.01 mm2 NA= 44.01 mm2
Jeffries Planimetric method
Method #2

Now we know the NA (number of grains per square mm)

We need to know the grain size number. NA= 44.01 mm2

Step # 2
You can know the grain size number in two ways:
1. After calculating the NA number, you can go to the table 6
from ASTM E 112 standard and cross the NA number with the
Grain size number. For this case the G is 2.5.
2. Or you can calculate it, using the following formulas:

A= Grain area in mm2 = 1/NA = 1/ 44.01 = 0.0227 mm2

d = Grain diameter mm= A ½ = 0.0227 ½ = 2.5

G = 2.5

Or you can use this formula:

G= (-3.322 Log A ) – 2.955 = (-3.322 Log (0.0227)) – 2.955 = 2.5

G= 2.5
 5
CHAPTER FIVE

Heyn / Hilliard / Abrams

Intercept method
 Intercept method
Method #3

Here´s another objective method to calculate the Grain

size number.

With the Heyn / Hilliards / Abrams Intercept method you can

calculate the grain size number in 2 ways.
#1 Counting entire grains intercepted or,
#2 Counting grain boundaries intersected.
You will need to draw a line with a specific length, and you can use
the following formula´s:

If you want to calculate the grain size counting the entire

grains.

NL= N / Lt
Or you can use this formula if you want to calculate the grain size
counting boundary intersections.

PL= P / Lt

Lt= length of the reference line.

 Intercept method
Method #3

Way #1 - Counting intercepted grains

Here is an example.
NL= N / Lt
Lt= Line length

This example has 5 Intercepted grains + 2 half at the ends this give
us a total of 6 grains.
 Intercept method
Method #3

Way #2 - Counting grain boundary intersections

For this example there are 6 grain boundary intersections.

#2 Counting grain boundary intersections
PL= P / Lt
Lt= Line length
 Intercept method
Method #3

So, once we know the N or P value, we need to calculate the NL or

PL. Consider the line length about 2 mm.
For NL = N/Lt = 6 / 2 = 3
For PL = P/Lt= 6 / 2 = 3
For this exercise, the NL and PL value is 3.

To know the grain size number we have to calculate using this

formula:
G= [6.643856 Log10 ( NL)]-3.288 Eq.2 Table 6 ASTM E 112

G= [6.643856 Log10 ( PL)]-3.288 Eq.2 Table 6 ASTM E 112

Or you can calculate it using the Mean Lineal Intercept.

The Mean Lineal Intercept length, which is the average length of a

line segment that crosses a sufficient large number of grains, and
it can be used to calculate the grain size number with this
equation:
G= [-6.643856 Log10 (l )]-3.288 Eq.2 Table 6 ASTM E 112

l = 1/ NL = 1/ PL
 Intercept method
Method #3

Now, in case there are not equiaxed grains, or the grain

shape present some distortion, you can use more straight
lines at different angles.
Or you may use 3 concentric circles as ASTM recommends
with a line length of 500 mm.
 Intercept method
Method #3

Let´s do an exercise with the following data:

Lt = 11.4 mm

N= 41 grains outside circle, 25 on the middle cirlce, 20 in

the inner circle
 Intercept method
Method #3

Applying the formula for NL, this is what we got:

Lt= 11.4 mm
N= 41 + 25 + 20 = 86
NL= 86 / 11.4 = 7.54 mm-1
l= 1/7.54 = 0.133 mm
G= [6.643856 Log10 (0.133 )]-3.288
G= 2.5
We can conclude that the Interception and planimetric
methods are reliable and widely used, but present
difficulties because is very tedious to get the result and it is
more complicated if more area is required.

In order to increase accuracy you need to analyze at least

160 mm2 according to ASTM E 112, and this is a difficult
task to do it manually.

Keep reading, you are about to discover how to use an

Image Analysis System.
 6
CHAPTER SIX

Automatic Measurement
using an Image Analyzer
 Automatic Measurement
Easier, faster and reliable

How could I measure more fields, faster and

with reliable Results?
You got it, using an Image Analysis System.

Below is the list of the “must know” principles for any Image
Analysis System, please take your time to see the videos, we will
use that information for the excercise.
(To see the videos please click on the links)

1. Before starting the analysis.

2. Setting up the microscope.
3. Optimizing Image Quality.
4. How the system Works.
5. How to calibrate the system.
 Automatic Measurement
Easier, faster and reliable

Now you know the principles, you are ready to discover how to
measure Grain Size automatically.

Imagine that you have a nice sample and your system is properly
setup and calibrated.
You only have to select and detect the grain boundary and allow
the system does all the work, this will be done in seconds.
Click here to see it in action.

But, how does the system calculate the Grain Size

number?
It uses the Mean Lineal Intercept equation:
G= [-6.643856 Log10 (l )]-3.288 Eq.2 Table 6 ASTM E 112
And the Average grain size per object
G = (3.321928 log10 NA) -2-954 Eq.1 Table 6 ASTM E 112
 Automatic Measurement
Easier, faster and reliable

The nex question is, What about the structures like the
picture below? Does the system Works the same way?

No, it does not work in the same way, because the binarization is
not accurate with the structure above.
For cases like this, you can use the Intercept method with the
digital tools of the System. Click here to see it in action.
 Automatic Measurement
Easier, faster and reliable

Probably you are asking yourself, how do I know that the

measurement is correct?

This is a very good question, let me show you something really

signifcant from ASTM standard:

ASTM E 112 point 19.8:

“Grain size ratings using the comparison chart method
by an individual metallographer will vary within +/- 0.5 G units.
When a number of individuals rate the same specimen, the
spread in ratings may be as great as 1.5 to 2.5 G units.”

ASTM E 112 point 19.14 :

“An individual metallographer can usually repeat planimetric
or intercept grain size measurements within +/- 0.1 G
units. When a number of metallographers measure the same
specimen, the spread of grain sizes is usually well within +/- 0.5
G units.”
 Automatic Measurement
Easier, faster and reliable

Measurement errors are always present, a metallographer can

make a little mistake while comparing, or counting manually as
well as the Image Analysis systems.

But, how could we know the error is aceptable?

The answer, Calculating the Relative Accuracy.

According to ASTM E 112 in the point 15.6:

“If the % RA is considered to be too high for the intended
application, more fields should be measured and the calculations
in 15.1-15.5 should be repeated.”

As a general rule, a 10 % RA (or lower) is

considered to be acceptable precision for most
purposes.
 Automatic Measurement
Easier, faster and reliable

You need to process enough data to have a robust analysis, that it

means that you need to measure more than 1 field of your sample.

How many fields?

To calculate the Confidence interval, is necessary at least 5 fields.

With 20 fields, you can get a fare enough accuracy.

If you want to reach the highest accuracy, you will need 60 fields.
Not easy to do it manually.
 7
CHAPTER SEVEN

Relative accuracy and

Confidence interval
 % R.A. and C.I.
Relative Accuracy and Confidence Interval

To calculate the relative accuracy, first you have to calculate the

Confidence Interval.
Here are 2 ways to calculate the CI
It could be based on the fields (n) or in the number of grain áreas
or intercept lengths. (N)
95% 𝐶𝐼=𝑡.𝑠÷√𝑛
95% 𝐶𝐼=𝑡.𝑠÷√N
Where:
n= the number of fields measured (for field measurements)
N= the number of grain áreas or intercept lengths for individual
measurements
t= Table 4 provide the values for t as function of n or N
s= standard deviation from grain measurements.
 % R.A. and C.I.
Relative Accuracy and Confidence Interval

After calculating the CI, you can now calculate the % of R.A.

%RA= ( 95% CI ÷ X) * 100

X= the mean value measured (NA,NL, or PL or I)

Making this manually is not easy, it takes more time to do it if

you want to deliver a professional report.

Imagine how useful is that the

%RA and CI is being calculated
automatically by the software.

Highly convenient, don´t you think?

This way you can focus on prepare
your report and explain the results.
 8
CHAPTER EIGHT

CONCLUSIONS
 CONCLUSIONS
Relative accuracy and Confidence Interval

After all this chapters, you can now be sure what is necessary to
do, to perform a profesional measurement for grain size.

The use of the image analysis systems makes your work easier, but
is necessary you understand how it Works and how to adjust it to
get objective Results.

Also, with this guide you know now , how many fields are required
to analyze in order to get a 10% of RA or less.

So regardless the type of material, this rules apply and even in

some cases, where the structure is not easy to process digitally;
with the use of the image analysis you can get objective
information only by measuring more fields.

We expect you have enjoyed this e-book, we try to sintetize a huge

ammount of information about this topic in some pages that can
be useful and practical.

Disclaimer:
This document can not be used for legal or comercial disputes, and any posible injury, damage to personel
or facilities due to the interpretation of this information ,is under user responsability only.
 REFERENCES

Grain Size Measurement _ George Vander Voort

ASTM E 112-13 Standard
ASTM E 1382-13 Standard

Special thanks
We want to thank to George Vander Voort, for sharing
his knowledge for so many years.

His deep knowledge and experience has help us to build this e-

book.
 Feedback

Do you like this e-book?

If you found this material, useful and
practical, please give us your
comments about it.

My comments are…

www.ingintegral.com