© The Minerals, Metals & Materials Society 2018
Boyd R. Davis, Michael S. Moats, Shijie Wang, Dean Gregurek, Joël Kapusta, Thomas P. Battle, Mark E. Schlesinger, Gerardo Raul Alvear Flores, Evgueni Jak, Graeme Goodall, Michael L. Free, Edouard Asselin, Alexandre Chagnes, David Dreisinger, Matthew Jeffrey, Jaeheon Lee, Graeme Miller, Jochen Petersen, Virginia S. T. Ciminelli, Qian Xu, Ronald Molnar, Jeff Adams, Wenying Liu, Niels Verbaan, John Goode, Ian M. London, Gisele Azimi, Alex Forstner, Ronel Kappes and Tarun Bhambhani (eds.)Extraction 2018The Minerals, Metals & Materials Serieshttps://doi.org/10.1007/978-3-319-95022-8_14

A Scientific Roadmap for Refractory Corrosion Testwork

J. Schmidl1  , A. Spanring1, D. Gregurek2 and K. Reinharter2
(1)
RHI Magnesita, Wienerbergstrasse 9, Vienna, 1100, Austria
(2)
RHI Magnesita, TCL, Magnesitstrasse 2, Leoben, 8700, Austria
 

Abstract

In non-ferrous metallurgy the service life of refractory materials typically ranges from several weeks up to three years or more and is strongly dependent on operating conditions. To support our customers with the most viable refractory solution for their needs RHI Magnesita follows a structured approach of thermo-chemical calculations, experimental evaluation from lab scale up to industrial field tests and post-mortem analysis of used refractory materials. This paper will give an overview of the RHI Magnesita refractory selection process and the most recent developments of the experimental setup for the so called HF-IF test. This is a dynamic corrosion test that allows to determine fundamental properties of different refractory material types under process conditions subjected with respective customers process materials. The HF-IF experimental procedure is an excellent tool to simulate refractory wear in industrial processes, diminishing risks associated with plant trials and support decision making to choose the optimal refractory solution for the customer.

Keywords

RefractoryCorrosion testworkHF-ITO test

Introduction

RHI Magnesita provides refractory solutions to almost all metallurgical processes in the nonferrous metals industry. An average sales volume of more than 90.000 t/a allows RHI Magnesita to serve customers of the non-ferrous base metals (Cu, Pb, Zn), ferro-alloys (FeCr, FeNi, FeMn, …) and precious metals industry [1]. Both, primary and secondary smelting operations are facing the challenge to treat especially complex material feed mixes, often at varying feed mix compositions. This changes of input parameters influences not only the interaction with the refractory lining but also the output of the smelting and/or recycling process. Existing results of a process cannot be directly transferred to describe other processes with different input parameters although they might be similar at a first glance. This means that every request needs to be investigated to provide customers with the best available solution. Rather than focusing on trial and error RHI Magnesita follows a well-structured scientific approach that combines metallurgical and refractory expertise to select the best suitable refractory material for a given application.

Typically stresses in the refractory lining can be subdivided into chemical, thermal and mechanical stresses which will affect the service life of the furnace lining and have been published for a number of different applications in literature [2–7]. Therefore several corrosion testing methods such as the induction furnace , short rotary kiln, cup test, etc. have been developed to assess different refractory brand qualities in order to recommend industrial trials at customer plants, leading finally to an improved refractory performance. The latest development in this series of high-temperature testing methods is the so called HF-ITO test which capabilities will be described together with the generic approach for refractory selection in this article.

Problem Formulation

At the very beginning the crucial part in the selection process of a refractory brand quality for an application request is the problem formulation. This part is the specific distinction between “collect all the information you can about a subject” and “let’s do an experiment”.

Due to the large number of input variables and the nature of their relationship, research problems in the field of slag |matte|metal interaction with refractory materials are rather complex. To understand the dimensions of a problem it is necessary to consider focus groups of variables to get an insight into a specific set of questions from the process.

Apart from obvious information like brand quality and service life history it is therefore particularly important to have information about the slag |matte|metal|offgas composition, process temperature and their variations in composition during the expected life time of the refractory lining. The collected information is then used to estimate the phase equilibria using the actual slag |matte|metal composition from industrial samples at process temperature and calculate the oxygen (pO2) and sulphur (pS2) partial pressures of the system. Figure 1 shows the sulphur-oxygen potential diagram for copper production [8].
../images/468727_1_En_14_Chapter/468727_1_En_14_Fig1_HTML.gif
Fig. 1

Sulfur -oxygen potential diagram for the system Cu–Fe–S–O–SiO2 at 1250 °C

This diagram can be understood as kind of a navigation map indicating the common working region that has to be considered during the copper production process, starting from smelting over converting, fire refining up to slag cleaning, Cu-recycling or smelting of complex CuPb containing mattes [9]. It should be emphasised that other impurities will impact the working regions and therefore thermodynamic equilibrium calculations are used to predict the actual pO2/pS2 combination for the process to use this information during corrosion test-work.

Post-mortem Studies

Understanding the refractory wear phenomena through post-mortem studies on used refractory bricks is an essential tool as they provide a precise understanding of the wear parameters influencing the refractory performance on the one hand and provide impetus towards the development of new refractory products.

During relining the post-mortem samples are taken, quite frequently supported by RHI Magnesita experts, and their position in the furnace is documented. Together with the process information this allows a precise description of the wear mechanisms of the spent refractory sample. From the macroscopic appearance typically four zones can be observed

  • accretion layer of slag |matte|metal from the process

  • reaction layer

  • infiltration zone

  • original microstructure of the brick up to cold side

The bricks are then sliced and cut into smaller samples to be prepared for further microscopic examination and analysis of the chemical composition. The basis for the evaluation of the refractory wear is an infiltration diagram (Fig. 2). It shows the mean phase distribution as a function of the distance from the hot face depending on the thermal gradient, thermal profile and is based on chemical and mineralogical, especially microscopic investigation.
../images/468727_1_En_14_Chapter/468727_1_En_14_Fig2_HTML.gif
Fig. 2

Phase distribution of a post-mortem refractory sample from hot to cold face, calculation based on SEM -EDS area analyses

The diagrams are designed to follow the evolution of phase distribution starting from the hot face up to the cold face and can be discussed to determine following aspects of refractory degradation

  • liquid infiltration by slag

While significant content of iron oxide is found together with relatively less silicates this may mean that the slag is not especially aggressive towards magnesia -chromite bricks, since free silica would probably react with the excess iron oxide . The low depth of infiltrated slag also shows that slag may play only a secondary role on the wear of the bricks in the converter.

  • liquid infiltration by matte and copper

It is also noticeable that metallic copper infiltrates deeper into the brick than copper matte: this can be the result of the slightly lower melting temperature of copper (1085 °C) in comparison to Cu2S (1130 °C).

Another observation is the presence of copper oxide (Cu2O) at the end of the infiltrated zone coexisting with metallic copper . Oxides of copper are not observed in the hottest parts of the brick. Because its it is clearly not coming from the slag , the most likely conclusion is that the copper oxide is being formed in situ.

Associated to the fact that metallic copper content increases when matte decreases and infiltrates deeper, it is hereby postulated that matte infiltrates in larger amounts than copper , and that most of the copper present in the structure of the brick (not in the cracks or joints) is formed by oxidation of matte according to following reactions
$$ {\text{Cu}}_{2} {\text{S}} + 1.5\,{\text{O}}_{2} = 2\,{\text{Cu}} + {\text{SO}}_{3} \quad \Delta {\text{V}} = - 48\% $$
(1)
$$ 2{\text{Cu}} + 0.5{\text{O}}_{2} = {\text{Cu}}_{2} {\text{O}}\quad \Delta {\text{V}} = + 62\% $$
(2)
$$ {\text{Cu}}_{2} {\text{S}} + 2{\text{O}}_{2} = {\text{Cu}}_{2} {\text{O}} + {\text{SO}}_{3} \quad \Delta {\text{V}} = - 15\% $$
(3)

If only the direct oxidation of copper sulfide (density 5.8 g/cm3) to copper oxide (density 6.15 g/cm3) is observed according to (Eq. 3) it should result in no loss of structural integrity.

However, if reaction occurs according to the two first steps, whereby there is a large shrinkage caused by formation of metallic copper (density 8.9 g/cm3), followed by a large volume increase as a result of its oxidation .

Theoretically, this volume increase could be accommodated in the voids previously filled by the infiltrated matte (the net volume increase is given by the third equation, meaning that the oxidation of copper generated by oxidation of matte will occupy less volume than the original matte). However, as copper is formed, voids are open to the infiltration of more matte or metal and the formation of Cu2O may increase damage by spalling.

  • redox reaction by sulphate diffusion

The microstructural studies have confirmed the presence of three sulfates from CaO and MgO. The melting temperature decrease in the order of CaSO4 (TM > 1400 °C), CaMg3(SO4)4 (~1200 °C) and MgSO4 (~1135 °C). It is clear from the analysis, that CaSO4 is present in the hottest parts of the bricks and MgSO4 in the coldest. The double sulfate appears, when present, between both sulfates or up to the cold face. This distribution suggests that sulfates might be good tracers to determine the thermal gradient inside the lining.

Refractory Quality Pre-selection

The appropriate tool to determine the chemical suitability of a refractory quality for a given application is a thermochemical calculation using for example thermochemical computational tools such as FactSageTM [10]. This computation method indicates which phases reach thermodynamic stability and, therefore, can be formed by the interactions between refractory and slag . The calculations are performed at isothermal conditions and fixed oxygen partial pressure. A typical representation of the result is shown in Fig. 3.
../images/468727_1_En_14_Chapter/468727_1_En_14_Fig3_HTML.gif
Fig. 3

Typical representation of thermochemical equilibrium calculations in the slag |refractory system carried out with FactSageTM

Starting from a fully liquid slag at the given process temperature and oxygen potential, refractory material is added stepwise and the thermodynamic equilibrium is calculated for each single step. The grey line indicates the ideal behaviour, assuming that no interaction between slag and refractory takes place. The results are represented by illustrating the formation of liquid slag and solid phases at equilibrium condition. For the illustrated system three phases are stable at equilibrium – liquid slag , spinel and monoxide. Dissolution of the refractory material into the slag phase is indicated by an increase of the liquid slag area above the grey line. This mechanism is called active corrosion or direct dissolution in literature. If the refractory |slag system tends to form solid precipitates the liquid slag area at equilibrium is beneath the grey line which is called passive corrosion or indirect dissolution.

While it is obvious that active corrosion will directly force the refractory degradation this cannot be easily derived from equilibrium calculations for passive corrosion . The formation of new phases is accompanied with volume changes and depending on the actual quantity of this change and the position inside the refractory brick resulting from infiltration of liquid slag into pores. This corrosion mechanism could also lead to mechanical degradation by forming cracks and ultimately spalling off of larger brick parts.

Corrosion Testwork

A variety of different corrosion test methods are available at RHI Magnesita and selected according to the findings of the above mentioned results. Special emphasis shall be given to the so called HF-ITO method. This test is especially designed to generate maximal refractory wear in a relatively short testing time in contrast to pilot scale tests in an induction or a short rotary furnace [11] and enables a good pre-selection of suitable brand qualities. Based on this evaluation a decision can be made to either perform pilot scale tests at the RHI Magnesita Technology Centre or directly run a field trial at the industrial application. The degradation of the refractory in contact with liquid slag can be expressed by following equation
$$ J_{i} = \frac{{D \cdot \Delta c_{i} }}{\delta } $$
(4)
where Ji is the dissolution rate of the refractory material which is assumed to be controlled by diffusion and D is the diffusion coefficient, δ is the diffusion boundary layer thickness and Δci represents the concentration gradient of the actual refractory component in the slag versus the saturation solubility in the slag .
While thermodynamic calculations only represent the phase equilibrium of the slag |refractory interface which accounts for Δci, this test method enables

  • to simulate agitation of the liquid bath, influencing the diffusion boundary layer

  • to assess differences between types of raw materials of the refractory type (sintered, fused, pure/mixed, OXICROM, …) by macro- and microscopic examination of the samples after testing

  • compare the performance of up to four different brick qualities under standardized conditions

Both parameter are essential information in the selection process and the HF-ITO method is not only a well-suited scientific test but also an economically solution as it can deliver insights and results in a short period of time and thus resulting in shorter response time to the customers. The original test set-up is illustrated in Fig. 4.
../images/468727_1_En_14_Chapter/468727_1_En_14_Fig4_HTML.gif
Fig. 4

Original HF-ITO test set-up

The principle test set-up was installed in December 2013 as a stand-alone device, equipped with an offgas unit. The major unit is an induction furnace with a power supply of 40 KW that can be run at a frequency of 5–100 kHz. The furnace can operate at a maximum process temperature of 1700 °C and a rotation speed of the sample holder from 1 to 10 rpm. Up to 3 kg of test material, industrial or synthetic slag |matte|metal samples can be melted in the crucible. The standard crucible material is graphite but, depending on sample composition other crucible materials have to be considered. The sample holder is placed above the crucible to pre-heat the sample finger and then immersed into the liquid bath. Four samples can be investigated in parallel, each of it with a dimension on 20 × 25 ×115 mm. The temperature measurement is taken from the outer sidewall of the crucible using a Metis MQ11 pyrometer. Compromises have to be made to the furnace atmosphere. It is not possible to accurately control the oxygen partial pressure, two options are available—the test can either be performed in air or inert atmosphere using Argon as purging gas. The testing period typically is around 4 h but can be adapted according to the need of the experiment.

To improve significance of the obtained results from the HF-ITO test further developments in both the operation procedure and the test set-up are in progress. Figure 5 shows a sketch of the current HF-ITO test design.
../images/468727_1_En_14_Chapter/468727_1_En_14_Fig5_HTML.gif
Fig. 5

Sketch of the current HF-ITO test design

  • crucible material—the selection of the crucible material is of high importance. When using standard graphite crucibles, interactions of the test material change its composition. Due the small melting volume this could have a high impact on the obtained results. Therefore crucibles are lined with refractory material that ensures minimum interaction with the test material.

  • atmosphere control —in order to be able to control and adjust a certain pO2/pS2 potential the furnace atmosphere can be purged with pre-mixed gas mixtures.

  • Temperature measurement—installation of a thermocouple directly in the melt for precise measurement of the actual temperature

  • sample quenching—after designated time the samples are pulled out of the bath and quenched as fast as possible in order to represent equilibrium/process conditions in the final sample for microscopic investigation.

Conclusion

RHI Magnesita follows a well-structured scientific approach that combines metallurgical and refractory expertise to select a refractory material for a given application. This includes detailed know-how of the metallurgical process conditions, collecting information from post-mortem analysis, thermochemical considerations, planning and execution of corrosion testwork from laboratory scale to pilot scale as well as industrial field trials. Based on this process RHI Magnesita is able to support customers with the most viable refractory lining solution for their processes. The presented recent developments in HF-ITO test set-up show that this corrosion test is a versatile method to evaluate refractory brand qualities under process linked conditions.