UDK 669.14.018.252:620.18:543.428.2	ISSN 1580-2949
Original scientific article/Izvirni znanstveni članek	MTAEC9, 45(1)55(2011)
CHARACTERIZATION OF THE INCLUSIONS IN SPRING STEEL USING LIGHT MICROSCOPY AND SCANNING ELECTRON MICROSCOPY
KARAKTERIZACIJA VKLJUČKOV V VZMETNIH JEKLIH S SVETLOBNO IN VRSTIČNO ELEKTRONSKO MIKROSKOPIJO
Arsim Bytyqi, Nuša Pukšic, Monika Jenko, Matjaž Godec
Institute of Metals and Technology, Lepi pot 11, 1000 Ljubljana, Slovenia Štore Steel Company, d. o. o., Železarska cesta 3, 3220 Štore, Slovenia arsim.bytyqi@imt.si
Prejem rokopisa - received: 2010-04-08; sprejem za objavo - accepted for publication: 2011-01-11
The quality of high-grade steel depends mainly on the size, distribution and composition of the inclusions in the steel. A good-quality product can be obtained by controlling the size, distribution and chemical composition of the inclusions. This study was limited to non-metallic inclusions, mainly compounds of metallic elements and carbon, nitrogen, oxygen or sulfur. The inclusions in spring steel were investigated with light microscopy and scanning electron microscopy. Generalized extreme value theory was applied to the size distribution of the inclusions. The types of inclusions were evaluated and classified in the majority as having a complex non-metallic composition. The experimental results show the possibility to define permissible limits for non-metallic inclusions with a minimal negative influence on the steel's properties.
Keywords: non-metallic inclusions, SEM, EDS, spring steel, GEV
Glavni dejavnik, od katerega je odvisna visoka kakovost jekla, je velikost, razporeditev in sestava nekovinskih vključkov. Z nadzorom velikosti, porazdelitve in kemijske sestave vključkov lahko dobimo kakovosten proizvod. Preiskava je bila omejena na nekovinske vključke, ki so v glavnem spojine kovinskih elementov z ogljikom, dušikom, kisikom ali žveplom. V prispevku sta karakterizacija nekovinskih vključkov in čistost vzmetnega jekla opredeljeni s preiskavami v svetlobnem mikroskopu in v vrstičnem elektronskem mikroskopu. Porazdelitev velikosti vključkov je bila opredeljena s posplošeno teorijo ekstremnih vrednosti. Iz rezultatov je bilo mogoče oceniti naravo vključkov, in sicer je bila večina vključkov kompleksnih, zraščenih z enofaznimi vključki. Rezultati eksperimentalnih postopkov kažejo na možnost določitve dopustne meje nekovinskih vključkov z minimalnim negativnim vplivom na lastnosti jekla.
Ključne besede: nekovinski vključki, SEM, EDS, vzmetno jeklo, posplošena teorija ekstremnih vrednosti
and a lot of effort was expended in recent years to 1 INTRODUCTION	predict the limits for inclusion size or mechanical
properties by extrapolating the data5. Non-metallic inclusions are formed during the steel's However, despite the major advances in inclusion
production process and have a great impact on the control, there is still no rapid and accurate method for
strength, plasticity, fracture toughness, fatigue strength determining the type,
size and number of inclusions
and other properties1. Producers can achieve better present in a steel sample. The scope of possible changes
control of the processing and improve the quality of pro- i^ the production and characterization processes requires ducts by a characterization of the individual inclusions. scientific studies.
The size and distribution of inclusions are particularly
important, because large macro-inclusions are the most
harmful for the mechanical properties of steels2. How-	2 EXPERIMENTAL ever, large inclusions are difficult to inspect because of
their low occurrence rate. Many experimental results High-purity steel was produced in Štore Steel, d. o. o.,
show that the largest inclusions are the most probable	in compliance with the EN 51CrV4 standard. The
fracture origins in a given volume of material3. There-	chemical composition is given in Table 1. Five samples
fore, it is important that steelmakers learn to better	with dimensions of 20 mm x 28 mm x 10 mm were cut
control the inclusions' characteristics. The characteri-	from the slab according to the ISO 4967 standard.
zation of non-metallic inclusions in steel in terms of	Samples were prepared with standard metallographic
number, size, shape and chemical composition is thus an	techniques, then examined by light microscopy and by
essential requirement for metallurgical process techno-	scanning electron microscopy (SEM). The advantage of
logy. It is necessary to measure the size and spatial	optical microscopy is the ability to examine a large area
distributions of the inclusions for a correlation with the	in a short time, but the information we are able to gather
mechanical properties4. Different statistical models can	about inclusions is limited. Imaging at higher magni-
be used to characterize the size distribution of inclusions	fication and microchemical analyses were made using
SEM with an energy-dispersive x-ray spectroscopy (EDS) capability.
The samples were examined with optical microscopy in order to determine the quantity and size of the inclusions. A total of 554 images of polished samples at 100x magnification with a total area of 150 mm2 were acquired. Image processing and measurements of the inclusion cross-section areas were made with the analySIS software.
The SEM was used in the secondary-electrons imaging and backscattered-electrons imaging modes to analyze the inclusions; details of the composition of the inclusions were investigated by EDS mapping and point analyses.
The light microscopy and SEM analyses (JEOL JSM-6500F) were performed at the Institute of Metals and Technology in Ljubljana.
Table 1: Chemical composition of the spring steel sample No. 49115 in w/%
Tabela 1: Kemijska sestava vzorca vzmetnega jekla {t. 49115 v masnih deležih, w/%
Element	C	Si	Mn	Cr	V	S	Al
Composition w/%	0.52	0.35	0.96	0.93	0.12	0.007	0.010
G{z) = exp
(1)
where is the probability that the maximum inclusion is no larger than the size z, and a and X are the scale and location parameters. It was first applied to the inclusion sizes in steels by Murakami and coworkers. The distribution was later generalized with the addition of another parameter
( r /(z - X) ^
G(z) = exp-1+^^-^	(2)
^ L V a yJ j
which reduces to the Gumbel distribution (Eq. 1) for ^ = 0.
A standard inspection area So = 0.27 mm2 was defined and 554 such areas were examined. The cross-section area of the largest inclusion in each inspection area was measured and a square root of the area z =# was calculated. Values of the distribution parameters were then calculated using the maximum-likelihood method. The hypothesis ^ = 0 was not statistically supported and thus abandoned. The parameter values of the generalized extreme-value distribution are listed in Table 2.
Table 2: Estimated parameters with standard errors from the GEV method for the experimental steel
Tabela 2: Ocene parametrov s standardnimi napakami za generalizi-rano metodo ekstremnih vrednosti
Parameter	Value	Std. error
a	1.96	0.06
X	6.37	0.09
1	-0.0089	0.0018
3 RESULTS AND DISCUSSION 3.1 Light microscopy
Since thorough inspection is difficult and time-consuming, statistical methods have been developed for predicting the characteristic maximum inclusion size in a large volume of steel by extrapolating from data gathered on small samples. In this study, Generalized Extreme Value (GEV) theory was applied to the results and a prediction for the characteristic inclusion size was calculated.
The basic concept of extreme value theory is that when a number of data points following a basic distribution are collected on multiple samples, the maxima and minima of each of these sets also follow a certain distribution. The distribution function was given by Gumbel6,7:
-(z - X) ^ - exp-
To estimate the size of the maximum inclusion in the volume of spring steel V; the return level T is defined as follows:
T =V/Vo	(3)
where Vo = h ■ So is the standard inspection volume, as defined by Murakami et al., where h = ^ ^areamax . /N.
The estimate for the characteristic size of the maximum inclusion is then calculated from [refs]:
1-? A
(4)
=X - a (1-r- ln
i V1-L-
1-1 ^^
T
J y
There exists an estimation upper limit of the inclusion size when ^ <0. The characteristic sizes of the maximum inclusion in different volumes of spring steel are shown in Figure 1. The estimation's upper limit is 224 pm. The predicted result of the inclusion size can be used in a database for a reference in the steel-making process.
Figure 1: Estimated characteristic sizes of the maximum inclusions in different volumes. Confidence intervals were calculated using the maximum-likelihood method for a 95% confidence level. Slika 1: Ocenjene karakteristične velikosti največjih vključkov za različne volumne jekla. Intervali zaupanja so bili izračunani po metodi največje verjetnosti za 95-odstotno stopnjo zaupanja.
3.2 SEM analysis
Scanning electron microscopy of the investigated spring-steel samples revealed different types of inclusions. Figure 2 shows a typical complex inclusion. Three analyses were performed on this inclusion to reveal the detailed composition, mainly of calcium sulfide and aluminum oxide, Table 2.
Figure 2: SEM image of a complex, non-metallic inclusion taken by secondary electrons
Slika 2: Kompleksen nekovinski vklju~kek, posnet s sekundarnimi elektroni v vrsti~nem elektronskem mikroskopu
Aluminum oxide inclusions are the result of aluminum de-oxidation. The great advantage of steel de-oxidation with aluminum and the use of calcium to modify the form of alumina inclusions is that they reduce the dissolved oxygen activity in liquid steel, resulting in a final product with a large index of cleanliness and a smaller tendency for porosity formation8.
Different inclusions were observed with scanning electron microscopy in sample 2. This chemical composition was similar, but the size was different. According to the EDS analysis the content of calcium sulfide (CaS) was greater than the content of aluminum oxide (Al2O3). As can be seen in Figure 3, aluminum oxide (Al2O3) in the inclusion center is surrounded by calcium sulfide (CaS). The distribution of the elements (Table 3) indicates that sulphur is bound in the manganese sulphide (MnS).
The analysis of the non-metallic inclusions in the steel 100 Cr6 indicated the presence of complex, oxide-sulfide and sulphide inclusions with different shape characteristics, while, a higher content of the sulphide inclusions is related to the high content of sulphur9.
A complex inclusion can be seen in Figure 4. The detailed analysis of the inclusion supported the finding and showed that the inclusion consisted of calcium sulfide and aluminum oxide. The acquired SEM
Figure 3: (a) SEM image of a complex, non-metallic inclusion and the corresponding elemental distribution for (b) O Ka1, (c) Mn Ka1, (d) Al Ka1, (e) S Ka1, (f) CaKa1.
Slika 3: (a) Nekovinski vklju~ek, posnet s sekundarnimi elektroni v vrsti~nem elektronskem mikroskopu in ploskovni posnetki elementov (b) O Ka1, (c) Mn Ka1, (d) Al Ka1, (e) S Ka1, (f) Ca Ka1
Table 2: Semi-quantitative chemical analysis of the inclusion in the studied spring steel in w/%
Tabela 2: Semikvantitativna kemijska analiza nekovinskega vkliu~ka preiskovanega vzmetnega jekla v masnih deležih, w/%
Tnlpmpntc
O
Mg
A1
S!
S
K
Ca
Mn
Fe
Spectrum 1
4 8
1 53
98 1
39 6
3 33
7 80
Spectrum 9
5 48
1 53
36 1
41 1
3 95
6 33
Spectrum 3
48 7
9 49
34 8
9 19
1 99
1 99
4 35
4 91
Table 3: Semi-quantitative chemical analysis of the inclusion in w/%
Tabela 3: Semikvantitativna kemijska analiza nekovinskega vklju~ka v masnih deležih, w/%
PlpmpntQ
O
Mg
Al
Ca
Mn
Fe
Spectrum
4 90
9 06
9 14
33 8
5 59
47 1
5 07
i;
Figure 4: (a) SEM image of a complex, non-metallic inclusion and the corresponding elemental distribution for (b) S K«1, (c) Ca K«1, (d) Mg Ka1, (e) Al Ka1,(f) O Ka1
Slika 4: (a) Nekovinskih vklju~ki posneti v vrsti~nem mikroskopu in ploskovni posnetki elementov (b) S K«1, (c) Ca K«1, (d) Mg K«1, (e) Al K«1, (f) O K«1
Table 4: Semi-quantitative chemical analysis of the inclusion in the studied spring steel in w/ Tabela 4: Semi-kvantitativna kemi~na analiza nekovinskega vklju~ka v masnih deležih, w/%
Elements
Al
Ca
Mn
Fe
Cr
Cu
Spectrum
0.81
26.4
0.37
48.2
22.4
0.64
Figure 5: (a) SEM image of a complex, non-metallic inclusion and the corresponding elemental distribution for (b) S K«i, (c) Mn K«i, (d) Al Kai
Slika 5: (a) Nekovinski vklju~ek, posnet s sekundarnimi elektroni v vrsti~nem elektronskem mikroskopu in ploskovni posnetki elementov (b) S K«!, (c) Mn K«!, (d) Al K«,
Table 5: Semi-quantitative chemical analysis of the inclusion in Figure 4 in w/% Tabela 5: Semikvantitativna kemiiska analiza nekovinskega vkliu~ka s slike 4 (w/%)
Elements
V
Cr
Mn
Fe
Total
Spectrum 3
0.48
26.4
! 5!
48.2
22 5
100 0
examination revealed that the sulfur and calcium content are lower in the light area of the inclusion than in the dark area. Correspondingly, the oxygen and aluminum contents are higher in the dark area compared with the light area. This indicates that the inclusion is composed of two different parts: the light part mainly consists of aluminum oxide (Al2O3), while the dark areas mainly consist of calcium sulfide (CaS) Table 4. The characteristic shape of the inclusion is globular and the size is 4 um. It is Quoted in references that inclusions with an
irregular shape and sharp edges cause larger stress concentrations around the inclusions than inclusions with a smooth shape, even if there is not a significant difference in the size10.
Manganese sulfide was found either as a singular inclusion or as an outer shell on an oxide. Figure 5 shows an SEM image of a sulfide type of inclusion consisting of mass fractions 48.2 % of manganese and 26.4 % of sulfur Table 5. These inclusions are narrow and long because of their high ductility at the steel
S
rolling temperature, and then deform and are elongated as the steel is rolled11. Because of their elongated shape, MnS inclusions may affect the transverse toughness of the steel12. Apart from the chemical composition, the size of the inclusions is crucial for the steel's properties. The results show that the critical inclusion size in the spring steel 51CrV4 was 0.14mm13.
4	CONCLUSION
The nonmetallic inclusions in spring steel were investigated in the laboratory, and the inclusion size distribution and compositions were determined using optical microscopy according the international standard ISO 4967 and with SEM combined with EDS, respectively.
The more specific conclusions from this work are as follows. The majority of inclusions found in steel contain manganese sulfide (MnS), calcium sulfide (CaS) and aluminum oxide (Al2O3) or a mixture of these. A comparison between the inclusion compositions in the spring steel showed that they are very similar in terms of chemical composition. The SEV method can be effectively used to estimate the size of the maximum inclusion in different volumes of the spring steel. However, a reasonable estimation can be derived from the combined analysis of two methods, i.e., the metallographic and statistically methods. The inclusion characteristics are not sufficiently investigated for many of the commercially available spring steels and should be amended.
5	REFERENCES
1 Gao Hong-bin., Jia Yun-hai., Li Mei-ling., Yuan Liang-jing, WANG Hai-zhou. Research on original position statistic distribution analysis
model of aluminum inclusion grain size in steel. Central Iron&Steel Research, 2009
2Lifeng Zhang., Brian G. Thomas. Inclusions in continuous casting of steel. Mech. Engr. Buldg. (2003), 138-183
3	A. Ahmad, J. Purbolaksono, Z. Yahya., Estimating inclusion size in WE43-T6 magnesium alloys based on Gumble extreme values. Materials Science and Engineering. A 513-514 (2009), 319-324
4	Shehata M. T. Boyd J. D. Measurement of Spatial Distribution of Inclusions. Inclusions and their Influence in Material Behavior, 4 (1998), 123-131
5	Atkinson H. V., Sh. G. Characterization of Inclusions in clean steels; A review including the statistics of extremes methods. Progress in Materials Science, 48 (2003) 5, 457-520
6	J. M. Zhang, J. F. Zhang, Z. G Yang, G. Y. Li, G. Yao, S. X. Li, W. J. Hui, Y. Q. Wenig, Estimation of maximum inclusion size and fatigue strength in high strength ADF1steel. 394 (2005), 126-131
7	C. W. Anderson, G. Shi, H. V. Atkinson, C. M. Sellars, J. R. Yates, Interrelationship between statistical methods for estimating the size of maximum inclusion in clean steels. Acta Materiala. 51 (2003), 2331-2343
8	Jose Carlos, S. Pires, Amauri Garcia. Modification of oxide inclusions present in aluminum-killed low carbon steel by addition of calcium. Metalurgia & Materialis., 57 (2004), 1-5
9	A. Gigovic, M. Oruc, I. Vitez, B. Vujicic. Analyse and research of nonmetallic inclusions for steel 100Cr6. Metalurgija, 48 (2009), 29-32
10Pekko Juvonen. Effectc of non-metallic inclusions on Fatigue properties of Calcium Treated Steels. Helsinki University of Technology, PhD Thesis, (2004), 14
11L. M. Cabalin, M. P. Mateo, J. J. Laserna. Large area mapping of non-metallic inclusions in stainless steel by an automated system based on laser ablation. Elsevier. Spectrochimica Acta Part B, 4 (2004), 567-575
12	Claude Lincourt, Madhavarao Krishnadev. Effect of inclusion morphology on the fracture toughness of high strength steels. Mater. Technol. 27 (2009), 1687-1698
13	Miha Kovacic, Sandra Sencic. Critical Inclusion Size in Spring Steel and genetic programming. Materials and Geoenvironment, 57 (2010), 17-23