Effective Quality Control of Steel and Iron Products with Combustion Analysis
时间：2015-05-15 阅读：7892
-
提供商
弗尔德（上海）仪器设备有限公司 -
资料大小
482.9KB -
资料图片
-
下载次数
303次 -
资料类型
pdf -
浏览次数
7892次
The element concentration of carbon (C), sulfur (S), hydrogen (H), nitrogen (N),
and oxygen (O) in iron products, such as steel, have a significant influence on
material properties like ductility, brittleness, or hardness. Hence, reliable
determination of the C, H, N, S, O contents is a routine quality control task in
steel and iron production. This article explores the application fields and
advantages of elemental analyzers, which are also known as combustion
analyzers, in the steel industry.
Different analytical methods
A great variety of different analytical methods are used in steel and iron
production1. Disregarding analyses of physical parameters, such as hardness or
tensile strength, and focusing solely on chemical analyses, the different
techniques can be categorized, for example, with regard to the measurement
method applied (AAS, OES, photometry, mass spectrometry, etc.). A common
differentiation is made between wet chemical techniques which require sample
digestion, e.g. by ICP-OES, and direct analysis methods like spark spectrometry
or combustion analysis. Another possible classification could be made for total
element analysis as opposed to surface or layer analysis. The following article
mainly focuses on the differences between spark spectrometry and combustion
analysis as these are the most common and established techniques for
measuring concentrations of C, H, N, S, O. Moreover, both methods are hardly
mentioned in general textbooks about analytical chemistry2, despite the fact that
they are widely used in laboratories and production.
Spectrometric methods such as spark spectrometry are theoretically suitable for
analysis of an unlimited number of elements of the periodical system.
Combustion analyzers, however, are specialized in the quantification of carbon,
hydrogen, nitrogen, sulfur and oxygen. The limitation to these elements has
some advantages. It is possible to measure solid samples which, due to their
geometry (powders, drillings, foils, pins), composition (coke, oils), or the
1
Handbuch für das Eisenhüttenlaboratorium, 2009, Band 1
2
e. g. Cammann (Instrumenle Analytik, 2001); Kellner, Mermet (Analytical Chemistry, 1997); Skoog
(Instrumenle Analytik, 2013 element to be determined (e.g. hydrogen) are not suitable for analysis with a
method like spark spectrometry.
However, there is no universal combustion analyzer available in the market
which can measure all elements. A further differentiation with regard to the
chemical nature of the sample is required. Thus, samples are divided into
organic materials with high carbon content like coal or coke, and inorganic
materials like steel, iron or copper. For organic samples a further differentiation
is made between determination of C,S and C,H,N,O,S;
respectively for inorganic materials between N,O,H and
H analysis. These different requirements are covered by
different analyzers. Table 1 gives an overview of
analysis systems, reaction gases, temperatures and
required additives. What all elemental analyzers have in
common is a combination of sample disintegration by
combustion and measurement of the released gases in
infrared and thermal conductivity cells (e.g. in ELTRA’s
CS-800, fig. 1).
A specialty of combustion analyzers is the possibility of fractional analysis which
is not - or only to a limited extent - available with spectrometric methods.
Fractional analysis means determination of the elemental concentration
according to the chemical and/or physical bonding in the material. TOC analysis
(total organic carbon), fractional hydrogen analysis or determination of surface
carbon have become an established part of routine analyses (see table 2).
Measuring the surface carbon content with elemental analyzers is compley
different from layer analysis with spectroscopic methods such as GD-OES.
Combustion analyzers exclusively detect the surface carbon which originates
from oil or process water, for example. Bound carbon, e.g. from carbonitriding,
is not accessible with this technique. Technically possible is a differentiation
between varieties of bound oxygen (e.g. from iron or lode stuff); however, due
to a number of influencing factors it is hardly suitable for routine analysis.
between varieties of bound oxygen (e.g. from iron or lode stuff); however, due
to a number of influencing factors it is hardly suitable for routine analysis.
Specifications of combustion analyzers
The determination of C, H, N, S, O is stipulated in various standards (see table
3). Table 3 shows a representative excerpt; depending on the region, further
ISO or ASTM standards (e.g. ASTM E1019) may apply. The standards define the
allowed measuring range for a particular element, permissible calibration
materials and their use and, where required, procedures of sampling and sample
preparation. Technical requirements for analyzers, however, are hardly specified
by the standards. For oxygen and nitrogen measurement, the standard only
mentions general laboratory instruments without further specification. For
carbon and sulfur analysis, however, standardized components of C/S analyzers,
such as gas purification, dust trap or induction furnace are listed. Hence, all
common C/S and O/N/H analyzers fulfill the requirements of the standard.
Despite the fact that spark spectrometers and combustion analyzers measure
the same matrix and ascertain similar values, process steps such as sample
preparation, calibration, measurement procedure and measuring range show
significant differences which are described in the next paragraph.
Sample preparation for combustion analysis is quick and easy. All that is
required is a representative analysis sample in a quantity which can be
accommodated by the crucible used. Typical sample weights are between 250
mg and 1000 mg. Sample geometry (drillings, powders, wires, etc.) is not
important for elemental analyzers. For O/N/H analysis it must be ensured that
the sample is purged from surrounding atmosphere inside the sample drop
mechanism. This is very easy for compact individual samples; powders or
drillings, however, should either be analyzed in a special loading mechanism or
by using air-tight tin or nickel capsules. Surface contaminations can be removed
with the help of an organic solvent like acetone.
Spark spectrometry requires samples with a planar surface, a certain thickness
to prevent the spark from penetrating the sample, and electric conductivity.
Contaminations of the surface can be prevented by „presparking“. A suitable
spectrometric method for analyzing nonconductive samples is, for example, high
frequency glow discharge.
Measurement procedure and calibration
With spark spectrometry all elements are detected simultaneously. The sample is
fed to the spectrometer, “presparked” if required, and finally the spark is used to
measure the intensity of the emitted radiation. A necessary prerequisite for
spectroscopic measurement of oxygen and nitrogen is the use of a spectrometer
with a suitable wave length (e.g. 130-780 nm)3. Calibration is usually carried out
by the manufacturer; the user only needs to make a drift correction using
certified reference material which contains the desired concentration of the
elements to be determined. The calibration usually matches a narrow
concentration range with each matrix (e.g. pig iron, different steel alloys, pure
iron, etc.) requiring an individual measuring method and calibration. The
drawback of this procedure is its dependence on the required reference
materials.
3 Data obtained from manufacturers such as OBLF, Thermo Scientific
The measurement of C, H, N, S, O concentrations with combustion analyzers is
separated into C/S analysis and N/O/H analysis. Pure chemical substances or
pure gases are suitable calibration materials for both methods. For calibration
with gas a defined volume of, for example, CO2 is introduced and directly guided
to the measurement unit without entering the furnace. Both methods allow for a
daily calibration update, or the existing calibration can be updated with a daily
factor. In contrast to spark spectrometry there is no strict matrix dependency.
The user can easily exchange the used standards against others. For C/S
analysis the sample is mixed with additives such as tungsten or iron, introduced
into the induction furnace, combusted in an oxygen stream, and finally the
resulting reaction gases CO2 and SO2 are detected in the infrared cells. ELTRA’s
CS-2000 offers a combination of inductive combustion with a resistance furnace
(fig. 2).
For O/N/H analysis in an electrode furnace the sample is placed in a sample drop
mechanism where it is purged from ambient atmosphere and dropped into a hot
graphite crucible. The sample melts, elemental hydrogen and nitrogen are
released and the oxygen contained in the sample reacts with the graphite
crucible. The method for quantification of the released gases depends on the
instrument manufacturer. The standards don’t mention any limitations regarding
the method. A typical C/S analysis takes approximay 45 seconds, O/N analysis
about 2 to 3 minutes. This is due to the integrated outgassing of the graphite
crucible to reduce possible contaminations. Whereas a spark spectrometer
processes a signal which is constant over time, combustion analysis produces transient signals (see fig. 3) which are integrated through the software. This
does not have a negative influence on the precision and correctness of the
measured values.
Whereas different methods of C, N, S, O analysis produce very similar results,
this is not the case with hydrogen analysis. The highest values are usually
obtained by fusion in an electrode furnace. The applied temperatures of 3,000 °C
melt the sample and the contained hydrogen is compley released. ELTRA
offers analyzers like the ONH-2000 for this application or the H-500 analyzer for
hot extraction at max. 1,000°C. Analyzing residual hydrogen at ambient
temperature is also possible with ELTRA instruments.
Measuring range and accuracy of results
The measuring range of a spark spectrometer depends on the sample material
(iron, steel, aluminum), calibration standards and spectrometric parameters like
optics, performance, etc., and is therefore difficult to define in general.
Combustion analyzers, however, provide the same measuring range for a great
variety of sample materials. An induction furnace determines, for example, a
carbon concentration of 7% regardless whether the material is iron, titan,
marble, or ceramic. The measuring range is usually defined for a nominal sample
weight of 1,000 mg. By adjusting the weight it is possible to measure
significantly higher concentrations. The lower detection limit is defined by the
signal-to-noise ratio - comparable to spectrometric methods - but can be
optimized by increasing the sample weight, pre-cleaning the crucibles and
additives, or catalytic carrier gas purification. A general comparison of detection
limits and accuracy of the two techniques is difficult as factors like sample
homogeneity and equipment configuration need to be taken into account. Table
4 shows a basic comparison. The repeatability defined in standard ISO 15350 is
hardly comparable with the standard deviation generally ascribed to
spectrometers but it reflects a tendency. Spectrometers indeed seem to measure
homogeneous samples more accuray whereas combustion analysis provides
excellent repeatability even for different analyzers and users.
Conclusion
When selecting a suitable method for elemental analysis, economic as well as
analytic aspects need to be considered. From an economic point of view spark
spectrometers, for example, are ideally suited for rapid analysis in the steel and
iron production process. Thanks to the reliable determination of the classic
elements C, N, S and O they offer possible cost savings for routine operations in
the lab. However, this advantage needs to be weighed against the high
investment costs, particularly for automated operations.
From an analytical point of view, combustion analyzers are high-performance
instruments which offer additional benefits to the user. They allow for reliable
measurement of samples with carbon inclusions, provide results over a wide
measuring range as well as the possibility for fractional analysis. Combustion
analysis is extremely reliable due to the fact that the results can be traced back
to primary substances such as chemicals or pure gas. The instruments require
low investment costs and have a long operating life which makes them attractive
also from an economic point of view.
Author:
Dr. Andre Klostermeier
Product Manager
Eltra GmbH, 42781 Haan, Germany
Phone: +49 (0) 21 04/23 33-400
: a.klostermeier@eltra.com