OXIDATION AND SCALE CHARACTERIZATION OF SILICON ALLOYED STEELS

Get Complete Project Material File(s) Now! »

Pickling conditions

The majority of plants uses the same pickling line to remove the scale on classical low carbon steel grades and alloyed steel grades. Sometimes, on alloyed grades surface defects can be generated by mixed oxides, due to insufficient time or inadequate conditions of pickling [74].
Pickling conditions vary from one plant to another; the main pickling conditions are listed below:
– Pickling line: it can be continuous, continuous coupled or push-pull.
– Acidic solution: Hydrochloric acid (predominant) or Sulfuric acid (mainly for Packaging steels) are used.
– Corrosion inhibitors: are generally commercial solutions coming from De Leuze, Kebo, Henkel,… suppliers. Their concentrations depend on the type of inhibitor. The maximum concentration is about 1.5 g.L-1 [74].
– Acid concentration: it depends on the tank and line and varies between 10 and 230 g.L-1 for HCl from the first to the last line tank, where injection of fresh acid is provided.
– Iron concentration: it depends on the tank and line and varies between 3 and 400 g.L-1 from the last to the first line tank.
– Line speed: it varies between 50 and 270 m.min-1, depending on the line geometry (length and tank number), steel grade and bath conditions (acid type, concentration and temperature).
– Bath temperature: It is comprised between 70 and 90°C for HCl and u p to 100°C for H2SO4.
Kop et al. elaborated a laboratory pickling cell to simulate pickling conditions at the laboratory. The scale sample is put in a rotating holder simulating the hydrodynamics. The pickling bath with the industrial acid solution and the inhibitor types and concentrations is heated with a thermostatic circulating fluid [75, 76].

Influence of pickling parameters on pickling time

The objective of pickling is to dissolve the entire scale layer in a short time to increase productivity and at the same time, to protect steel from mass loss and high surface roughness due to an excessive over-pickling. Actually, optimal pickling time is defined when all scale is removed with pickling and over-pickling.
Pickling time depends on the metal and scale characteristics and the pickling conditions:
– Scale thickness: it increases with the rolling and coiling temperature and is different at the centre and tail of the sheet due to slower cooling. The required pickling time depends on the scale thickness and chemistry [77-79].
– Scale composition: for low carbon steel scales, hematite and magnetite dissolution is very slow compared to that of wüstite. For example, in a 88 g.L-1 HCl solution at 80°C, the dissolution rates of Fe3O4 and FeO are 0.1 g.m-2.min-1 and 1.6-2.0 g.m-2.min-1 respectively. FeO decomposition into Fe3O4 and Fe can accelerate its dissolution up to a factor 10 [79-81].
– Scale morphology: scale cracking and elongation of 1 to 2 % accelerates pickling with a factor 2 to 3 [82].
– Bath temperature: pickling efficiency increases with bath temperature. Pickling is usually performed at 70-90°C [79, 83].
– Acidic solution: scale is dissolved quickly in highly concentrated and hot HCl or H2SO4 acid solutions. For given concentration and temperature, HCl is more efficient [75, 79, 84].
– Iron cations concentration (Fe2+ and Fe3+): in an HCl medium, the unfavorable effect of Fe2+ does not appear below 80 g.L-1, and remains in every case more moderate than in H2SO4 medium. The ferric ions Fe3+ are known to decrease the pickling time and facilitate the attack reaction of the base metal. But their concentration must be well controlled to avoid over-pickling, mainly for grades which can be pickled rapidly. Generally, the salt concentrations are 50-100 g.L-1 for FeCl2 and 3-5 g.L-1 for FeCl3 at the end of pickling [74, 76, 79].
– Alloying elements concentration: PO43-, Mn2+, Cr3+ coming from steel dissolution increase slightly pickling time.

Effect of Silicon on pickling mechanism

After oxidation in the hot strip mill, silicon oxides (silica and fayalite) concentrate at the steel/scale interface in many forms; silica thin layer, fayalite grains in wüstite, fayalite infiltrations and silica internal oxidation grains in steel substrate.
The upper iron oxides layer dissolves easily but the low solubility of silicon oxides present at steel/scale interface increases the pickling time. In the fayalite infiltrations and high internal oxidation cases, a long over-pickling step is necessary to suppress fayalite and oxides at the grain boundaries [80, 89, 103].
According to [104] a relatively long pickling time is necessary to suppress completely the scale of a 3 wt.% Si steel, whatever the nature of acidic solution. In HCl bath, the addition of nitric acid induces a more reactive bath towards the metallic surface. The addition of HF acid increases the pickling efficiency without increasing the metal dissolution, due to its efficiency to dissolve silicon oxides.
Chattopadhyay et al. [105] studied the effect of silicon contents from 0.06 to 2 wt % in automotive steels on pickling time in 5, 8, 11 and 14% HCl solution at 82°C simulating the industrial process. The different steel grades have the same scale thickness but do not clean at the same pickling rate. This is due to differences in scale composition. Fayalite Fe2.44Si0.55O4 appears for silicon content higher than 0.35 wt. % and has a negative effect on pickling. Pickling time increases with silicon content (Figure I. 16).
The role of silicon and aluminum contents on pickling time of a low alloyed hot rolled steel sheet has been investigated by Y.N. Chang [106]. The increase of silicon content from 0.5 to 2 wt.% induces an increase of the amount of FeO required combining with SiO2 to form Fe2SiO4 at high temperature. The Fe3O4 transformation was reduced at room temperature. The pickling rate was thus slightly promoted as mentioned by Chattopadhyay et al. [105]. However, an addition of 0.4 wt. %Al was detrimental to pickling due to the formation of segregated FeAl2O4 at the scale/base metal interface.

Electrochemical measurements

Electrochemical measurements during pickling and over-pickling were done in an experimental setup adapted to steel samples covered with scale. It consists of an electrochemical cell connected to a potensiostat and frequency response analyzer.
Different electrochemical methods were programmed to extrapolate the different aspects and contribution of electrochemical processes in pickling and over-pickling mechanism.

Electrochemical set-up

The electrochemical cell used for pickling and over-pickling experiments is composed of three electrodes immerged in an acid bath and connected to a potentiostat:
– Reference electrode: has a constant potential in the same range of temperature. All potential values are given versus this potential. As pickling and over-pickling experiments in this study are conducted in hydrochloric acid (HCl) solutions, the saturated calomel electrode (SCE) with KCl solution was used to avoid contamination of the acid solution by foreign ions (K+ is inert and Cl- is already present in the acid solution). A saturated KCl extension tube is used to protect the reference electrode during high temperature dissolutions.
– Auxiliary electrode: is a platinum grid with a surface high enough to avoid current saturation. The potential is applied between the reference electrode and the working electrode. The current is measured in the circuit made of the WE and the auxiliary electrode.
– Working electrode: A rotating electrode with a three cylindrical pieces sample holder was elaborated (Figure II. 4). The outer piece is open partially at the bottom to expose the sample surface to the solution. Acid infiltration into the sample holder is prevented with a rubber circular joint between the outer piece bottom and the sample. The intermediate piece is screwed on the bottom piece pressing the sample on the rubber joint. The central piece made of stainless steel is screwed on the intermediate piece ensuring electrical contact between the back of the sample and the head of the rotating electrode.
To fit in the holder, the sample must have a disc form with 25 mm diameter and a maximum thickness of 5 mm.

Scale characterization after pickling and over-pickling

After pickling or over-pickling tests in the electrochemical cell, the rotating electrode is extracted from the acid solution, the sample is removed from the rotating electrode sample holder, cleaned with ethanol to remove the acid solution drop and remaining scale and put under vacuum to protect the surface from dust and atmospheric corrosion. In the ICP-AES cell, the sample is removed from its lodge and follows the same treatment as for the one in the electrochemical cell.
The sample surface was observed with a LEICA® S440 SEM or Zeiss® Ultra 55 SEM-FEG after a low pressure carbon deposition treatment to facilitate electrical charges evacuation. A global first micro-graph is taken to identify the phases on the surface corresponding to different pickling levels; they are distinguished with their brightness in the SEM backscattered electrons mode. Then zooms on each zone are performed to observe the form and roughness of the surface with the secondary electrons mode and identify its composition with EDS.
For over-pickling tests done on polished steel samples at different temperatures and applied potentials, an optical microscope and a Zygo® NewView® interferometric microscope observations were performed on over-pickled surfaces after increasing immersion times. Sample surface is cleaned with ethanol after each test and observed immediately in parts of the surface to evaluate its roughness and morphology. The surface roughness data were treated with a MoutainsMap® proTM® software. The results were presented as roughness cartographies or parameters. Two roughness parameters were chosen:
– Sq: the standard deviation of the height distribution of the surface.
– Sz: the average difference between the highest height peaks and lowest valleys.
The evolution of these two parameters and the roughness cartographies during over-pickling were analysed to study the distribution and amplitude of roughness.

Low carbon steel scale properties

In the oxidation conditions of this study, the resulting non alloyed steel (Si00) scale is homogenous and essentially composed of a wüstite la yer covered by a magnetite thinner layer (Figure III. 5a). Below 570°C wüstite is unstable: it transforms into iron and magnetite eutectoid. However, this transformation enables very low cooling rates. Considering a post oxidation cooling speed of about 40 °C min -1 in all experiments of this study, the wüstite transformation can be assumed for low part of scale volume, which is not clearly visible on SEM even at high magnification [19, 58].
Figure III. 5: (a) Cross-section SEM micrograph of Si00 after oxidation at 850°C in 15% humid air during 15 minutes with (b) Raman spectra of its scale phases: hematite external surface, magnetite layer and wüstite layer.
Si00 scale layers and their Raman spectra are presented in Figure III. 5b. On the scale top surface, Raman analysis reveals the presence of a hematite layer with intense peaks at 220 cm-1 and 290 cm-1 and 410 cm-1, as well as small peaks at 500 and 610 cm-1. This layer is very thin and not visible on SEM micrograph of the sample cross-section. Raman characterizations performed on cross-section do not distinguish clearly magnetite from wüstite. The Raman spectrum provided on external layer corresponds well to the spectrum of magnetite (Fe3O4) with characteristic peaks at 310 cm-1, 540 cm-1 and 670 cm-1. Wüstite phase (FeO) is cubic and should not be visible on Raman spectra. However, a peak at 660-670 cm-1 was observed during Raman analyses on the internal layer [105, 119]. This peak could be attributed to the presence of vacancies in the wüstite microstructure (wüstite Fe 1-xO with relatively large x variation according to the phase diagram [19]) or partial wüstite transformation during cooling into Fe + Fe3O4 eutectoid (presence of magnetite areas).
EDS analyses gives the chemical composition of the low carbon steel scale layers. The external and internal layers compositions are close to Fe3O4 and FeO respectively (Table III. 2).

Table of contents :

GENERAL INTRODUCTION
I. LITERATURE REVIEW
I. 2. Oxidation and pickling in steelmaking process
I. 2. 1. Overview of steelmaking process
I. 2. 2. The hot strip mill
I. 2. 3. The pickling baths
I. 3. Oxidation of steel
I. 3. 1. Steel substrate properties
I. 3. 1. 1. Low carbon steel
I. 3. 1. 2. Silicon alloyed steel
I. 3. 2. Oxidation conditions
I. 3. 2. 1. Oxidation in the Hot Strip Mill
I. 3. 2. 2. Oxidation in a pilot furnace
I. 3. 3. Oxidation of low carbon steel
I. 3. 3. 1. Oxidation kinetics
I. 3. 3. 2. LCS oxidation mechanism
I. 3. 4. Oxidation of silicon alloyed steels
I. 3. 4. 1. Influence of silicon on the kinetics of scale growth
I. 3. 4. 2. Influence of silicon on oxidation mechanism
I. 4. Scale and metal oxides properties
I. 4. 1. Hot strip mill (HSM) scales
I. 4. 1. 1. Low carbon steel (LCS) scales
I. 4. 1. 2. Silicon alloyed steel (SiAS) scales
I. 4. 2. Model scales
I. 4. 2. 1. LCS model scales
I. 4. 2. 2. Iron oxides properties
I. 4. 2. 3. SiAS model scales
I. 5. Pickling and over-pickling
I. 5. 1. Pickling of hot mild steels
I. 5. 1. 1. Pickling conditions
I. 5. 1. 2. Influence of pickling parameters on pickling time
I. 5. 2. Pickling mechanism of low carbon steels
I. 5. 2. 1. Pickling steps of LCS
I. 5. 2. 2. Pickling reactions of LCS
I. 5. 2. 3. Iron oxides dissolution
I. 5. 3. Effect of Silicon on pickling mechanism
I. 5. Conclusions
II. 1. Elaboration of steel grades samples
II. 1. 1 Steel chemical composition
II. 2. 2. Steel grains size
II. 2. Oxidation of steels
II. 2. 1. Oxidation experiments
II. 2. 2. Oxidation methods
II. 3. Scale Characterization after oxidation
II. 3. 1. Optical observations
II. 3. 2. Spectrometric methods
II. 4. Scale dissolution characterization
II. 4. 1. Electrochemical measurements
II. 4. 1. 1. Electrochemical set-up
II. 4. 1. 2. Electrochemical methods
II. 4. 2. ICP-AES setup and method
II. 5. Scale characterization after pickling and over-pickling
III. OXIDATION AND SCALE CHARACTERIZATION OF SILICON ALLOYED STEELS
III. 1. Introduction
III. 2. Steel substrate properties
III. 3. Oxidation mechanisms
III. 3. 1. Low carbon steel oxidation
III. 3. 1. 1. Oxidation kinetics
III. 3. 1. 2. Oxidation mechanism
III. 3. 2. Effect of Silicon content on oxidation
III. 3. 2. 1. Oxidation kinetics
III. 3. 2. 2. Oxidation mechanism
III. 4. Scale characterization
III. 4. 1. Low carbon steel scale properties
III. 4. 2. Silicon steel scales properties
III. 5. Conclusions
IV. PICKLING AND OVER-PICKLING MECHANISMS OF LOW CARBON STEEL.
IV. 1. Introduction
IV. 2. Model scale characteristics before pickling
IV. 3. Electrochemical dissolution of scale
IV. 3. 1. Corrosion potential Ecorr
IV. 3. 2. Corrosion current Icorr
IV. 3. 2. 1. Tafel curves
IV. 3. 2. 2. Corrosion current evolution during scale dissolution
IV. 3. 3. Electrochemical Impedance Spectroscopy EIS
IV. 3. 3. 1. EIS diagrams
IV. 3. 3. 2. Evolution of charge transfer resistance R
IV. 3. 3. 3. Evolution of CPE parameters
IV. 3. 3. 4. Evolution of effective capacitance
IV. 4. Total dissolution of scale
IV. 5. Pickling and over-pickling mechanisms
IV. 5. 1. Pickling thermodynamics
IV. 5. 2. Pickling steps and reactions
IV. 5. 3. Over-pickling reactions
IV. 6. Influence of some parameters on picking and O-P mechanism of LCS scales
IV. 6. 1. Scale composition and morphology
IV. 6. 1. 1. Influence of hematite on pickling
IV. 6. 1. 2. Pickling of an industrial scale
IV. 6. 2. Influence of acid concentration
IV. 6. 3. Influence of pickling bath temperature
IV. 6. 4. Influence of a cathodic applied potential
IV. 7. Steel surface after pickling and over-pickling
IV. 7. 1.Steel surface after insufficient pickling
IV. 7. 2 Steel surface after long over-pickling:
IV. 8. Conclusions
V. PICKLING AND OVER-PICKLING MECHANISMS OF HIGH SILICON ALLOYED STEEL GRADES
V. 1. Introduction
V. 2. Model scale characteristics before pickling
V. 3. Electrochemistry of scale dissolution
V. 3. 1. Corrosion potential Ecorr
V. 3. 2. Corrosion current density Icorr
V. 3. 2. 1. Current-potential curves
V. 3. 2. 1. Evolution of the corrosion current density
V. 4. Total dissolution of scale
V. 4. 1. Total dissolution rate (TDR)
V. 4. 2. Electrochemical contribution in scale dissolution
V. 5. Pickling and over-pickling (O-P) mechanism
V. 5. 1. Pickling steps and reactions
V. 5. 2. Over-pickling steps and reactions
V. 6. Influence of some parameters on picking and O-P mechanism
V. 6. 1. Scale morphology and composition
V. 6. 1. 1. Influence of fayalite morphology
V. 6. 1. 2. Behaviour of industrial scale
V. 6. 2. Influence of pickling bath temperature
V. 6. 3. Influence of acid concentration
V. 7. Steel surface after pickling and over-pickling
V. 7. 1. Steel surface after insufficient pickling
V. 7. 2. Steel surface after over-pickling
V. 8. Conclusions
GENERAL CONCLUSION
ANNEXES
REFERENCES

GET THE COMPLETE PROJECT