How clean is your steel?

– why quantification of inclusions provides

confidence in a long fatigue life
Erik Claesson, Joakim Fagerlund, Lily Kamjou and Patrik Ölund

1
EXECUTIVE OVERVIEW

The use of clean steel is known to offer a Ovako has codified this new approach to
dramatic improvement in the fatigue life of inclusion quantification within the new
critical automotive powertrain components. standard that now enables powertrain
This is due mainly to the precise engineering designers make use of full benefits of
of the inclusions that initiate fatigue failures. clean steel.
Therefore, to have confidence that a clean
steel will perform as expected it is vital to While this white paper is aimed mainly at
quantify both the size and statistical designers, especially those that make
dispersion of the inclusions present. structural calculations, it will also be of
interest to automotive engineers, purchasing
There is however a significant challenge for professionals and anyone involved in
the industry created by present standards specifying steels for demanding applications.
that are effectively ‘obsolete’ – they rely on
outdated methods of quantification that are Four key messages emerge:
incapable of recognizing the benefits of clean •  Inclusions are the main factor that
steel. The net result is that powertrain limit fatigue life (assuming correct
designers are unable to access the potential of design, heat treatment and surface finish)
these materials to optimize their components •  Current international standards do not
in terms of performance, size and weight. enable designers to benefit the full potential
of clean steel
In this white paper Ovako explains how it has •  A combination of test methods are now
addressed this quantification challenge with available to quantify inclusions in modern
a new approach that combines the traditional clean steels
technique of light optical microscopy (LOM) •  Ovako has developed its own publicly
with the modern techniques of scanning available standard that embraces these test
electron microscopy (SEM) and immersed methods so that designers can access the
ultrasonics. This approach provides a advantages of clean steel
comprehensive overview of the size and
distribution of micro and macro inclusions.

2
1 – INTRODUCTION

The drive for higher levels of fuel efficiency fatigue data. This experience shows that the
requires powertrain components to be lighter, presence of unwanted particles in the steel,
stronger and capable of resisting ever greater known as ‘inclusions’ represent a significant
and more complex loads. In many cases it is danger. This is because they act as local stress
fatigue strength that is the most critical factor raisers that multiply the nominal load to
when selecting a powertrain steel com- above the component’s safe fatigue limit.
ponents since fatigue accounts for the
majority of all mechanical service failures. Clean steels, such as Ovako’s BQ and
IQ-Steels, in which the size and distribution
Fatigue occurs if a metal component fails of inclusions is closely controlled can have a
when subjected to repeated loading, even at hugely beneficial effect in improving fatigue
loads well below what it could easily sustain characteristics. The potential improvement
on a single loading. The ‘safe load’ or ‘fatigue in fatigue is illustrated by the rotating
load limit’ is the load at which a component bending fatigue properties for conventional
will survive without failure beyond a certain steel, BQ-Steel and IQ-Steel as shown in Fig 1.
number of load cycles.
Rotating bending
Ovako has drawn on decades of industrial Rotating bending fatigue
experience to develop a large database of MPa
1000

900

800

700

600

500

400
Longitudinal Transverse
300

IQ-Steel BQ-Steel Conventional steel

Fig 1 – Clean steel offers significant potential

to improve the fatigue life of critical power-
train components.

3
The issue is that current international
1200

standards for steel inclusions do not reflect 1000

the recent major advances in steel quality, IQ-Steel Fatigue failure

Fatigue strength (MPa)

particularly when inclusions are small and/ 800

or widely dispersed. BQ-Steel

600
It is not just a case of measuring the size of
inclusions. Because large inclusions are found 400 Conventional steel
so rarely in clean steel it is vital to sample
a sufficiently large volume of material to be 200

confident that the tests reflect a true picture Safe zone

of the probability of them occurring. 0
0 10 20 30 40 50 60 70 80 90 100

It is no exaggeration to say that current steel Defect size (μm)

standards are effectively obsolete in that they Fig 2 – the fatigue strength of steel trans-
offer no effective guidance for designers in mission components is related directly to the
considering clean steels. Therefore a new defect size.
approach to quantification has been
developed that brings together a range of
methods providing a full and statistically
valid picture of the inclusion population in
a steel sample.
It is particularly in the use of 10 MHz
immersed ultrasonics that Ovako has made
advances in limiting the defects that have the
most influence on the final performance of
the finished component. This is also a
method that can be readily adopted by most
steel producers. SEM of large areas is a
technique that will take more time to become
established as a standard cross-industry
procedure. However, SEM is currently used
as part of Ovako’s in-house process and
product development procedures.
The effective quantification of steel inclusions
now provides the basis for a new standard for
clean steels. While it is primarily intended
for use with Ovako’s own products it is freely
available for public use.
Key points:
•  Clean steels in which the size and distri-
bution of inclusions is closely controlled
can offer the possibility to improve com- Fatigue failure is a major challenge for gear
ponent fatigue life by up to 50% (see fig 2) design – photos with thanks to the Institute
•  Current steel standards are effectively of Machine Elements Gear Research Cen-
obsolete as they do not offer guidance in tre (FZG), Technical University of Munich,
selecting clean steels from an original article in Gear Technology
•  Ovako has published a new standard based magazine.
on the effective quantification of steel
inclusions

4
2 – WHERE DO INCLUSIONS COME FROM?

It is useful to understand how steel inclusions

are formed and how they are categorized.
There are two main types – endogenous and
exogenous:
•  Endogenous ‘micro’ inclusions are
formed by the physical-chemical effects
that occur during the melting and
solidification process. They can be formed
from the oxygen and sulphur remaining
after the deoxidation and desulphurization
process or through reoxidation, see fig. 3a
and 3b.
•  Exogenous ‘macro’ inclusions result from Fig 3a – type D micro inclusion
parts of the slag, refractories, teeming
powder, or sand from a casting mould,
see fig. 3c.
ISO 4967, ASTM E45 and DIN 50602 are the
current standards that apply when assessing
micro inclusions. However, modern clean
steels have very few inclusions above 25 µm,
and the size of the assessed area in standard
ASTM and DIN tests using optical methods
is too small to provide any statistical
confidence.
Blue fracture is currently used to assess
macro inclusions. But invariably, using this
method a clean steel producer will generate Fig 3b – type A MnS (Manganese Sulphide)
only zero ratings for macro inclusions. micro inclusions

Fig 3c – Macro inclusions

5
3 – INCLUSION QUANTIFICATION METHODS

Ovako has focused on reviewing and Conventional steel 15ppm oxygen.

developing methods to create an approach
that can accurately reflect a realistic inclusion
content in clean steel. This is vital for both
improving steel quality and also in predicting
how a component will perform.

3.1 Light Optical Microscopy (LOM)

Light Optical Microscopy is the traditional
technique. It is covered by standards such as
ISO-4967, ASTM A295/E45 and DIN 50602.
The results are evaluated using charts such as
the JK reference scale.

This technique is only suitable for qualifying

Clean steel 7ppm oxygen
inclusions between 2 µm and 15 µm and is
limited to very small sample sizes – typically
the evaluated area is 1200 mm2. LOM does
not provide any data on the chemical com-
position of inclusions and is therefore not a
suitable tool for process development.
Fig 4 illustrates that the small sample size
is a specific issue with LOM.

3.2 Blue fracture testing

Blue fracture testing is an historically
well-established technique used to reveal
macro inclusions larger than 0.5 mm. It is
performed on a bar cross-section area that Fig 4 – A specific issue with LOM is the
has been hardened, fractured and then small sample size as shown in this schematic
tempered blue to increase the visibility of illustration (not to scale). Normal procedure is
defects. to examine 6 samples. But their small size does
not represent the true size and distribution
This technique is used by Ovako only due to of inclusions. In this particular case it is even
customer demand. It is of little relevance possible to obtain the false impression that
in clean steel though, as it is over 30 years standard steel (4a) has fewer and smaller
since an inclusion has been found using this inclusions than a clean steel (4b).
method.

6
3.3 Scanning Electron Microscopy
(SEM)
In contrast to LOM, scanning electron
microscopy, see fig 7, is capable of assessing
large areas – typically 5,000 mm2 and
provides rich data on inclusion chemistry,
morphology and size. The chemistry of
inclusions is vital for process development,
while morphology and size is vital for product
development. This quantification method is
used for inclusions between 2 µm and 25 µm.
Fig 5 – SEM equipment at Ovako.

3.4 Immersed ultrasonic testing

Fully automated ultrasound testing methods
used by Ovako to test for larger inclusions
have produced impressive results, see fig 5.

To test for inclusions above 120 µm, a single

sample of 500,000 mm3 steel, milled plane
parallel, and immersed in a water tank is
scanned with a 10 MHz probe, see Fig 6. This
is the equivalent of 16,000 blue fracture tests.
This test does not produce information about
the chemical composition of the inclusions,
but it is an important tool for process
development.
Fig 6 – immersed ultrasonic testing at
Ovako
To test for smaller inclusions, it is possible
to increase the ultrasonic probe frequency to
SEM UST/RBF UST 10 MHz
15, 25, 50 or even 80 MHz. However, as the
frequency and resolution is increased the size 100

of the sampled volume will decrease.

10-5

3.5 A combination of techniques

Probebility

10-10
creates the full picture
Three techniques – LOM, SEM and ultrasonics 10-15
– are combined to obtain a complete over-
view of the total inclusion content that feeds 10-20
directly into the refinement of our production
processes for new, cleaner steels, see Fig 7. 10-25
0 20 40 60 80 100 120
Equivalent diameter in micron
It should be noted that to obtain a full picture LOM
of the relationship between inclusion Fig 7 – summary of inclusion quantification
population and fatigue properties Ovako methods.
recommend that rotating bending fatigue
testing (RBF) should be carried out on
appropriate samples.

7
4 – CLOSING THE STANDARDS GAP

Reliable quantification of inclusions has Ovako ultrasonic standard

made it possible to develop a new generation The Ovako internal test procedure involves
of clean steels. However, current design the testing of steel billets. Through experien-
standards do not take into consideration the ce, Ovako has concluded that the ‘worst part’,
benefits of these materials in terms of fatigue of the ingot is at the very bottom. Therefore,
properties. It is therefore difficult for power- sampling is made on material originating
train designers to select clean steel. from this area.

The result of this effective gap between the A central part of the billet is prepared by
available standards and the capabilities of milling. The samples are scanned in an
clean steels is that significant opportunities immersion ultrasonic tank with a focused
are being missed to optimize components in 10 MHz transducer. The equipment is
terms of their performance, size and weight. calibrated with known defects and calibrated
so that the smallest feature that will be
Customers sometimes request the blue detected is as a defect corresponding to
fracture testing procedure described in ISO a 0.12 mm FBH (flat-bottom-hole).
3763. In Ovako’s experience, 10 MHz
immersion ultrasonic testing is a much more The minimum detected feature size and
powerful method of generating information tested mass (or volume) are important testing
regarding macro inclusions. This has resulted parameters. Immersion ultrasonic testing
in the development of an in-house ultrasonic offers both a higher detectability and allows
testing procedure that Ovako offer to replace testing of a more significant volume of
blue fracture testing. material, as shown in Table 1.

ISO 3763 Blue fracture Ovako 10MHz UST

Length ≥ 1.0 mm
Minimum feature detected FBH* ≥ 0.120mm
Thickness ≥ 0.1 mm
Coverage Surface Volume
Approximately Approximately
Amount of material investigated1)
2000 mm2 1054 cm3
Number of tested specimen 2 3

Table 1 - Detectability and amount of tested material

* FBH –Flat bottom hole.

8
To illustrate the improved detectability, blue
fracture samples were manufactured from
an ultrasonic test piece that, when scanned,
showed a high number of imperfections,
see Fig 8. The scanned sample had a large
number of defects exceeding 0.2 mm FBH
(see next section) due to the large number of
imperfections. Yet when blue fracture testing
was carried out no indication of any defect
could be found on the fracture surface, as
shown in Fig 8.
Fig 9 - Ultrasonic scan showing the
Ultrasonic testing produces an output positions where blue fracture specimens
like that shown in Fig 10. The different were selected.
amplitude sizes correspond to defect sizes.
The class >100 % Full screen height (FSH)
will correspond to an artificial defect
exceeding 0.2 mm FBH.

Ovako has used this practical experience to

create a standard based on setting a limit to
the number of defects found in divided by the
investigated volume. Physically, this relates to
the maximum number of defects larger than
0.2 mm per unit volume. The data in
Fig 9 is then further processed according
to the method in table 2.

Fig 10 – Ultrasonic testing output from a

medium carbon steel.

The number of defects

2+8+5 15
larger than 0.2 mm FBH

Total inspected mass in kg 4.9+4.7+4.7 14.3 kg

number of defects larger

15/14.3 1.05 #/kg
than 0.2 mm FBH/ kg

number of defects larger 1.05 x 7.8

Fig 8 – Blue fracture test showing no than 0.2 mm FBH /dm3 (density)
8.2 #/dm3
indication of any inclusion
Table 2 - Calculation of number of defects
larger than 0.2 mm FBH /dm3

9
The standard sets the proposed limits shown
Quality < 0.4 % C ≥ 0.4 % C Comment
in Table 3.
Guaranteed
The test conditions are: values based
BQ < 60 (UST) < 30 (UST)
•  Billet 80 to 250 mm round or square on statistically
testing
•  Samples from bottom part of the ingot
(minimum 1.2 weight % crop) IQ < 10 (UST) < 5 (UST) Tested values

•  Average of minimum 3 samples Table 3 - Proposed limits for various quality

•  Ovako testing procedure OFL047 classes and carbon contents.
•  Additional evaluation as described in the
example above i.e. number of defects
larger than 0.2 mm FBH /dm3

The standard makes it possible to set design

parameters that correlate with the properties
of clean steel – typically this could be an im-
provement of some 30% in the fatigue limit.

The standard has already been applied with

considerable success for demanding com-
ponents in bearing and diesel injection
applications. It is now being applied for
powertrain components.

Ovako has developed the standard

along the same lines as established inter-
national standards. So while it is primarily
intended for our own use Ovako is making it
freely available for customers to take to any
steel supplier.

10
5 – PRACTICAL EXAMPLE

In order to examine the effect of steel clean-

Steel A Steel B
liness on the fatigue properties, two different
steel bars (70 mm in diameter) with a major Steel grade 18CrNiMo7 18CrNiMo7
difference in steel cleanliness were investigated.
Casting Process Continuous cast Ingot Cast
The steel grade was a carburizing steel
(18CrNiMo7). Area Reduction ~10 ~65

Table 4 – Investigated Steels

Assessment of steel cleanliness was per-
formed by three different methods: micro
inclusion rating by ASTM-E45; ultrasonic
evaluation; scanning electron microscope.

Ultrasonic and SEM-evaluation revealed a

major difference in cleanliness between the
two steels, whereas the traditional micro
inclusion rating method did not reveal any
significant difference. The results from ultra-
sonic evaluation can be seen in Fig 11.

Fatigue samples were manufactured from

Fig 11 – Ultrasonic C-Scan (10MHz), Steel A
the two investigated steels. These samples
(left) and Steel B(right)
were prepared in a transverse direction to the
rolling direction of the bar, as this is the most
MPa
critical direction (inclusions are elongated in 900
the rolling direction of the steel bar).
800
The results of the fatigue testing gave a
700
fatigue limit of approx. 540 MPa for Steel A
and 800 MPa for Steel B, see Fig 12. 600

500

400

300

200

100

0
Steel A Steel B

Fig 12 – fatigue limits from the two

investigated steels

11
6 – SUMMARY

Non-metallic inclusions are the critical factor Ovako has responded to this quantification
that determine the fatigue life of steel. The challenge by developing a new approach
use of modern production techniques has based primarily on 10 MHz ultrasonic testing
resulted in a new generation of clean steels in to correctly identify the size and nature of
which the size and distribution of inclusions inclusions.
is closely controlled.
This method of quantification is codified in
Using clean steels for powertrain components Ovako’s standard that now enables power-
can offer a significantly enhanced fatigue life train designers to utilize the advanced fatigue
– up to 50% in some cases. properties of clean steel to optimize their
components.
Currently, powertrain designers are not able
to fully exploit the advantages offered by
clean steels as today’s international standards
do not provide the opportunity to specify
them. The main reason for this is that the
techniques outlined in current standards that
have been applied historically for conventio-
nal steels are not sufficient to quantify the
much smaller and more dispersed inclusions
in clean steel.

12
7 – REFERENCES/FURTHER READING

By Patrik Ölund By Erik Claesson, Lily Kamjou and Patrik

Ovako Technical Report 1/2016. The IQ-pro- Ölund
cess – the Ovako isotropic quality process. Ovako White Paper. Clean steel – living up to
power density challenges (2016).
By Patrik Ölund
Replacing blue fracture testing according Gear Technology Magazine – August 2015 –
to ISO 3763 with Ovako 10MHz immersion Page 58 – Article by FZG http://www.gear-
ultrasonic testing. technology.com/issues/0815x/gt0815.pdf

Standards for gear life calculations ISO 6336

and ISO/DTS 19042-1.

Appendix:
Detection of non-metallic inclusions in steels
with high cleanliness demands such as case-
or through hardening bearing steels by the
ultrasonic method, can be downloaded at
Ovako.com.

13
8 – ABOUT THE AUTHORS

Erik Claesson Lily Kamjou

Erik Claesson is head of the group function Lily Kamjou is a Senior Specialist in Ovako’s
Industry Solutions Development with main Industry Solutions Development department.
responsibility to increase end-user value. In her current role, she focuses on application
The scope is global and a special task force development specializing in the powertrain
including design and manufacturing com- area. Kamjou joined Ovako in 2008 and
petences has been added to the steel product is based at the company’s headquarters in
performance perpective. Claesson joined Stockholm, Sweden. She has held a variety of
Ovako in 1998 and has held various manage- positions in the automotive sector including
ment positions in manufacturing. Since 2010 working with the highly demanding market
he has heading group responsibilities aiming for diesel injection systems. Kamjou has a
towards increased value creation. Claesson master’s degree in materials engineering
has a master’s degree in materials technology from the Royal Institute of Technology
and metal forming from the Royal Institute (KTH, Stockholm) and a bachelor’s degree
of Technology (KTH, Stockholm). in social science from Stockholm University.

Joakim Fagerlund Patrik Ölund

Joakim Fagerlund is a Senior R&D Engineer Patrik Ölund is head of group research and
in Ovako’s R&D department. In his current development at Ovako. Educated at The
role, he focuses on development related to Royal Institute of Technology (KTH,
steel cleanliness and fatigue. He has been Stockholm, Sweden (1985-1990), he worked
with Ovako for more than 10 years and is at the Swedish Institute for Metals
based at the company’s headquarters in Research (1990-1995) doing research relating
Stockholm, Sweden. He has a master’s to inclusions, fatigue and heat treatment.
degree in materials physics from In 1995 he joined Ovako in the research
KTH Royal Institute of Technology. department, which he now heads. Ölund
was the winner of the Kami Prize 2013,
presented to a distinguished scientist whose
research has become the basis of a technical
development within the Swedish steel and
metal industry.

14
15
 CO-61 GB 17:11 © Ovako 2017

Disclaimer
The information in this document is for illustrative purposes only. The data and examples are
only general recommendations and not a warranty or a guarantee. The suitability of a product
for a specific application can be confirmed only by Ovako once given the actual conditions. The
purchaser of an Ovako product has the responsibility to ascertain and control the applicability of
the products before using them.
Continuous development may necessitate changes in technical data without notice. This document
is only valid for Ovako material. Other material, covering the same international specifications, does
not necessarily comply with the properties presented in this document.

Ovako 20171107
©
Ovako AB
SE-111 87 Stockholm, Sweden
www.ovako.com Phone: +46 (0)8 622 13 00