Steelmaking
Impact of Solid Particles and Liquid Droplets on Foams – Cold Model and High Temperature Experiments
2022 Volume 62 Issue 1 Pages 104-111
Details
2022 Volume 62 Issue 1 Pages 104-111
In order to obtain a realistic view of the foam in metallurgical slag, high temperature experiments where the foaming heights of FeO–CaO–SiO2–MgO slags containing precipitated MgO∙FeO particles were measured. The foaming height slightly increased when small amounts of particles were present in the slag, but decreased to half height already when approximately 8 vol% particles were present in the liquid phase of the foam. To help the understanding, the foaming heights of silicone oil and food oil containing liquid insoluble droplets and non-reacting particles were also studied at room temperature. In these experiments, insoluble oil droplets were found to stabilize the foam, increasing the foaming height, while the addition of water droplets or solid particles had very little effect on foaming height. In line with the literature, it is believed that the interfacial energy between the droplets or particles and the bulk liquid as well as the interfacial energy between the droplets or particles and gas plays an important role. When the interfacial energy between the different phases becomes too high, the foaming height decreases, while when it’s low enough, the foaming height increases.
Foaming slag plays a profound role in the BOF-converter and EAF steelmaking processes. The foam enhances the kinetics of reactions, isolates the heat in the melt and protects the furnace lining from radiation saving both energy and material costs. On the other hand, too extensive foaming leads to slopping, which can be harmful to the working environment and disturb the process train. Foaming slag is a complex emulsion of gas and liquid and in many cases even solid phases.1) The impact and importance of each phase is although still debated, and only a few fundamental studies have been carried out focusing on a steelmaking perspective.2,3,4,5,6) Instead, many engineering considerations mostly rely on industrial experiences, which often come to the conclusion that the presence of solid particles in general increases the foaming ability of slags.7) It’s important to remember that many parameters have direct or/and indirect impacts on the results of industrial practices with respect to foaming. It would usually be difficult to conclude that the presence of solid particles is the only key parameter resulting in the increase of foaming height in a BOF or EAF. At the same time, it is well known that addition of rigid spheres or insoluble droplets increases the apparent viscosity of a fluid.8,9,10) In fact, even gaseous phases may greatly increase the apparent viscosity of a fluid.11,12) Bigger fraction of the second phase may also make the two phase mixture non-Newtonian.11,12) In combination with the beliefs that the foaming height always increases with increased viscosity, as stated by the foaming index,2,3) it is not strange that one might draw the perhaps incomplete conclusion that the foaming height will increase when additions of solid particles or drops of another fluid are presented to a foam. As a matter of fact, there have been experimental results that disagree with the formerly mentioned literature regarding the foaming ability with increased dynamic viscosity of the fluid.13,14) These results evidently show that the foaming height reaches a maximum with increasing dynamic viscosity of the fluid and decreases with further increase of the viscosity.13,14) In the stainless steel industry, it has been reported that the presence of solid phases such as Magnesiochromite spinel interferes with the foaming ability.15,16,17)
In other research areas within chemistry and physics, where other types of foams have been studied, the subject of emulsions stabilized by solid particles was first studied in the beginning of the 20th century by Ramsden and Pickering.18,19) Later, the phenomenon of an emulsion stabilized by particles was even named as a Pickering emulsion. It is also reported that foams may be ruptured when solids are present in the foam,20,21,22,23,24,25) somehow in great contradiction to the beliefs in the steelmaking society. The lack of knowledge about the effects of particles on foaming slag could mostly be due to the great difficulties in carrying out the high temperature experiments. Hence in this work, a dedicated high temperature experimental procedure is designed to obtain reliable experimental result. Even experiments at room temperature are conducted to facilitate some observations and complement the understanding. Studies at both room temperature and high temperature are conducted mostly focusing on how additions of non-reactive particles and insoluble liquid droplets affect a foam by measuring the change in foaming height.
When determining the total slag compositions of the samples, it is essential to keep the liquids’ composition approximately constant for all samples, while letting the fraction of the solid phase vary. Several slag compositions in the FeO–CaO–SiO2–MgO system at 1823 K were preliminary chosen by thermodynamic calculations using Thermo-calc.26) Based on a big number of calculations, compositions were chosen aiming at different fractions of solid MgO∙FeO phase, and at the same time keeping the liquid composition of the slags nearly constant. Efforts were also made to find samples having approximately the same compositions of the solid phase, since the physical properties of both liquid and MgO∙FeO solid solution must remain the same when comparing the results of the different foaming experiments. The only changing variable should be the fraction of the solid phase (the number of solid particles) in the liquid during foaming. To ensure that the samples met the above-mentioned criteria, a phase study was conducted and samples of 1 gram were prepared as described below. Using samples of small masses was to make sure the phases at the experimental temperature were maintained during quenching. The prepared sample were subjected to microscopic analyses to ascertain that the compositions and the sizes of the precipitated MgO∙FeO particles in all the samples were very similar.
(2) Sample Preparation and AnalysisFirst, the oxides were prepared. FeO was produced by mixing iron powder (99.9% purity) and Fe2O3 (98% purity) to 51 mole% O in an iron crucible with lid. The mixture was then heated to 1123 K in Argon atmosphere and kept for approximately 62 hours. The method had been proven earlier,27) and obtained FeO was confirmed with XRD analysis. The sintered FeO body was then crushed into small pieces. CaO, SiO2 and MgO were all calcined at 1173 K for 12 hours. 1 gram of each calculated slag composition was then prepared by mixing the oxides for approximately 20 minutes in an agate mortar before adding the powder mixture into small molybdenum crucibles with inner diameter 8 mm and height 30 mm.
The experimental setup used was a vertical tube furnace equipped with Kanthal Super heating elements, schematically presented in Fig. 1. An alumina tube functioned as reaction chamber. The upper part of the tube was connected to a water-cooled cap and the lower part to a ball valve, both made of aluminium. O-rings were used to make the connections gas tight. A hook, sealed by an O-ring, was inserted through the lid of the water-cooled cap. The samples were tied to this hook by Mo-wire. The wire length was chosen to position the samples in the even high temperature zone. A type B-thermocouple, also sealed by an O-ring, was inserted through the top lid as well. The thermocouple was placed so that its tip was in height with the samples. Furthermore, the top lid was connected to a gas outlet hose. The gas inlet was placed on the upper flange of the ball valve. Below the ball valve, an aluminium quenching chamber was connected. The quenching chamber consisted of a 50 cm long extension tube ending in a container of larger diameter, which was filled with 3.5 L mineral oil (Leybonol LVO100) to cool the samples in. The ball valve was kept closed when the furnace was hot to prevent radiation from the alumina tube to the oil as well as evaporation of oil into the reaction chamber. The ball valve was only opened when dropping the samples into the oil, and the long extension tube between the oil container and the ball valve functioned to limit oil splashing into the furnace tube.
Experimental setup for finding suitable slag compositions to use for high temperature foaming experiments.
The crucibles containing the samples were placed in the furnace, and it was checked that the furnace was gas tight. To remove oxygen from the furnace, it was evacuated for 30 min and filled with Ar. This was repeated 3 times. After the last time, an Ar flow of 0.1 L/min was maintained. The samples were first heated to 1923 K with a rate of 1.5 K/min. After 45 minutes of homogenization, the temperature was decreased to 1823 K by 1.5 K/min, allowing precipitation of the desired MgO∙FeO phase. The samples were quickly cooled in oil after 10 minutes of retention time at 1823 K.
The quenched samples were cut horizontally and mounted into epoxy pellets. The pellets were grinded and polished in ethanol and 3 μm diamond paste before examination in both a light optical microscope (LOM) and scanning electron microscope (SEM). The amount of precipitated MgO∙FeO could easily be examined in the LOM. The SEM was employed to examine the composition of each phase using energy dispersive X-ray spectroscopy (EDS). For the EDS analysis, 10 points in the solid phase and 10 areas in the liquid phase were selected. Although this method is not exact, it is accurate enough for the purpose of this paper, i.e. to find and analyze the trends, not to report data.
The precipitated MgO∙FeO was successfully found in both LOM and SEM. An example from the LOM analysis can be seen in Fig. 2. The lighter grey circles are the precipitation formed during the cooling from 1923 K to 1823 K. The size of the particles varies between 20 and 70 μm and it should be mentioned that all the samples have particles in the same size range, while their fractions of particles are different.
Two phase mixture with MgO·FeO(s) and liquid slag.
Four suitable slag compositions were chosen based on the phase study. The total compositions, fractions of solid phase and phase compositions of these samples, are listed in Table 1. The slag compositions are given in weight percentage. As can be seen in the table, one composition was chosen to be fully liquid as reference. The volume fraction of the precipitated MgO∙FeO phase was evaluated using grid method on photographs taken from the LOM. The liquid composition has been normalized with respect to FeO, CaO, SiO2 and MgO. However, it should be mentioned that the samples also contained approximately 10 wt% MoO2. The presence of MoO2 was due to the oxidation of the crucible by FeO in the sample, which explains the change in slag composition from total to liquid composition in Slag 0. Unfortunately, it was almost impossible to find a crucible being inert to the slag. However, dissolution of MoO2 would have little effect on the discussion as long as the amount of MoO2 is approximately the same in all samples. The solid phase did not contain MoO2. The MoO2 content of the liquid phase according to EDS is given in Table 2. It is noted that all samples contain similar levels of MoO2 in the liquid phase, independent of the fraction of solid phase.
| Slag no. | FeO [wt%] | CaO [wt%] | SiO2 [wt%] | MgO [wt%] | MgO∙FeO [vol%] | FeO in MgO∙FeO [wt%] | FeO in liquid [wt%] | CaO in liquid [wt%] | SiO2 in liquid [wt%] | MgO in liquid [wt%] |
|---|---|---|---|---|---|---|---|---|---|---|
| 0 | 36 | 27 | 27 | 10 | – | – | 29 | 33 | 28 | 10 |
| 1 | 40.3 | 22.7 | 22.7 | 14.3 | 1.6 | 52 | 31 | 31 | 25 | 13 |
| 2 | 42.3 | 21 | 21 | 15.7 | 8.7 | 51 | 33 | 30 | 25 | 12 |
| 3 | 43.5 | 20 | 20 | 16.5 | 12.5 | 53 | 33 | 30 | 25 | 12 |
| Slag no. | MgO∙FeO [vol%] | MgO [wt%] | SiO2 [wt%] | CaO [wt%] | FeO [wt%] | MoO2 [wt%] |
|---|---|---|---|---|---|---|
| 1 | 1.6 | 12 | 22 | 28 | 28 | 10 |
| 2 | 8.7 | 11 | 22 | 28 | 29 | 10 |
| 3 | 12.5 | 11 | 22 | 27 | 30 | 11 |
The observed and calculated fractions of solid phase follow the same trend, as seen in Table 3 where these are compared. Hence, the present slag compositions allow for a relative study on the impact of solid particles on slag foaming. The observed and calculated solid fractions do not agree completely. This could be due to uncertainties in the calculations and the dissolution of Mo. For this reason, the discussion is based on the experimentally obtained solid fraction.
| Slag no. | Calculated [vol%] | Observed [vol%] |
|---|---|---|
| 0 | 0 | 0 |
| 1 | 10.6 | 1.6 |
| 2 | 15.1 | 8.7 |
| 3 | 17.9 | 12.5 |
For the high temperature foaming experiments, the four slag compositions described in Table 1 were prepared in the same manner as earlier described during the phase study, but with bigger mass. 70 grams of a slag (powder-mixture) were put into a Mo crucible and placed in a vertical tube furnace equipped with Kanthal Super heating elements. See schematic drawing in Fig. 3.
Experimental setup of high temperature experiments.
An alumina tube acted as reaction chamber. The Mo crucible containing the slag was placed in a molybdenum holder attached to a lifting system. While the furnace was heated to 1923 K, the slag was kept in the cooling chamber. The lifting system then lowered down the sample to the hot zone. After 45 minutes in the hot zone, the sample temperature was decreased to 1823 K with a rate of 1.5 K/min, allowing precipitation of MgO∙FeO. The sample was kept at 1823 K for 10 minutes before 1 gram of Ø2 mm pig iron containing 3.9 wt% C was released into the slag through a molybdenum rod, as shown in the focused part of Fig. 3. The carbon from the pig iron reduced the FeO in the slag, generating CO gas which in turn led to foaming. The slag was left for 1 minute before it was raised to the cooling chamber. Within this minute, a foam was generated, the carbon was used up, and the foam collapsed. The foam although left a distinguished mark on the crucible wall, enabling measurement of the maximum foaming height of the slag using a ruler.
Parts of the collapsed slag were then collected and mounted into epoxy pellets. The pellets were carefully grinded and polished with ethanol and 3 μm polishing paste before examination in both LOM and SEM. The purpose was to verify the presence of solid precipitation in all slags.
2.2. Room Temperature ExperimentsTo facilitate observations of particle distribution in the foam, cold model studies were conducted. Silicone oil or food oil was employed as bulk liquid and poured into a Plexiglas vessel with an inner diameter of 90 mm and a height of 400 mm. A schematic description can be found in Fig. 4. A silica filter with pore size 16–40 μm was glued 100 mm from the bottom. Argon gas was introduced from the bottom of the vessel into a gas chamber at 1.0 l/min, controlled with a flow meter of model Bronkhorst mass flow meter (model F-201CV-1K0-AAD-33-V, calibrated for 1 ln/min Ar, 2 bar (g)/0 bar (g) at 293 K) through a Bronkhorst High-Tech B.V. E-7000 Flow-bus controller. The gas chamber allows a uniform Argon flow through the silica filter into the oil, generating a spherical foam.28)
Schematic description of experimental setup for room temperature experiments.
The material properties used in the cold model experiments are listed in Table 4. Two types of solid particles were studied. Nylon has similar density as the liquid phase, whereas copper is denser. Since the foaming slag in the BOF is a slag-metal-gas emulsion, it is interesting to study the behavior of the denser particle in the foam.
| Density [kg/m3] | Viscosity [Pa∙s] | Contact angle of silicone oil [°] | Contact angle of food oil [°] | Size [mm] | |
|---|---|---|---|---|---|
| Silicone oil (l) | 964 | 0.1 | – | – | – |
| Food oil (l) | 927 | 0.050 | – | – | 0.05–0.2 |
| Water (l) | 1000 | 0.001 | – | – | 3–5 |
| Mineral oil (l) | 883 | 0.220 | – | – | 0.05–0.2 |
| Nylon (s) | 1140 | – | <5 | <5 | < 0.35 |
| Copper (s) | 8960 | – | <5 | <5 | < 0.05 |
The viscosities of all liquids were measured using a Brookfield LVDV-II+Pro Viscometer to which suitable spindle sizes were chosen for each liquid. The contact angle was estimated by ocular examination of a liquid drop on top of a plate of stated solid material. As can be seen in the table, both types of oils wet on both Nylon and Copper.
Lastly, the size of the droplets and particles were determined. The sizes of the water drops could be measured during the experiments using a regular ruler, the droplet size of the oils was measured in a LOM after mixing the liquids, see example in Fig. 5, and the particle sizes of the solid materials were given by supplier.
Droplets of mineral oil in Silicone oil.
All additions were introduced into an already foaming silicone oil or food oil. A reference height of the pure foams could thereby be measured using a ruler before additions. A new foaming height was measured when the additions had been introduced to the foam.
The MgO∙FeO phase was successfully precipitated during the high temperature foaming experiments. An example of foaming slag 4 along with a comparison from the phase study of slag 4 are shown in Figs. 6(a) and 6(b). Solid particles due to precipitation, a liquid matrix and gas phase are all presented.
(a) Slag 4 from the phase study including solid MgO·FeO and liquid matrix, and (b) slag 4 from the foaming experiments including solid MgO·FeO, liquid matrix and gaseous phases.
In Fig. 6(a), it is shown that the slag from the phase study contains solid MgO∙FeO with a particles size distribution between 20 and 70 μm, and a liquid matrix. Comparing the collapsed foaming slag 4 in Fig. 6(b), it is shown that the matrix is full of other phases. The reason is because the cooling rate when freezing the slag is much less efficient in this experiment. Partly, the slag sample is much bigger (70 grams compared to 1 gram) leaving a bigger mass to cool. Also, even after the foam has collapsed, the slag contains many micro gas phases which decrease the heat transfer in the slag and thereby decrease the cooling rate. The precipitated MgO∙FeO phase that was found was although very similar to the one found in the phase study, with a size distribution between 20 and 70 μm. The authors thereby draw the conclusion that only the MgO∙FeO was solid during the foaming experiment, and that the other phases were formed during the in-efficient cooling of the foam.
The resulting maximum foaming heights of all slag composition are shown in Fig. 7. The foaming height was found to increase with small amounts of precipitates. On the other hand, even less than 0.1 volume fraction of the solid MgO∙FeO is enough to decrease the foaming height by half, leaving a splashing porridge-like slag.
Foaming height vs volume fraction of precipitation.
The foaming height of 100 grams of silicone oil was measured to 25 mm, while the foaming height of 100 grams of food oil was measured to 50 mm. The change in foaming height was then measured after additions. Note that a layer of liquid without foaming is observed under the foam in all cases though the layer thicknesses were relatively small, only a few millimeters. In the case of additions of solid particles, the stated mass fraction is the mass fraction of solids in the whole system, not only in the foam. The distribution of particles between the liquid and foaming phase is impossible to measure. It is evident that the mass fraction of solids in the foam increases with increased mass fraction in the system by observing the color of the foams. This aspect is illustrated in Fig. 8, where in Fig. 8(a) the mass fraction of nylon is 0.016 and in Fig. 8(b) the mass fraction is 0.05.
Color of foam after addition of (a) 0.016 and (b) 0.05 mass fraction nylon. (Online version in color.)
The results of the foaming height experiments are presented in Fig. 9. Note that the mass fraction of solids in Fig. 9(a) is an apparent mass fraction, and not the actual mass fraction of solids in the foam, as explained above. In view that the present work focuses mostly on the trend of foaming as a function of the amount of second phases, the results are considered satisfactory.
Foaming height of silicone oil or food oil with additions of (a) solids and (b) insoluble liquids.
The results in Fig. 9(a) show that the foaming height does not increase when the solid particles are added to the foams. Instead, within the mass fraction interval studied in the experiment, the foaming height did either not change at all, or decreased slightly. The same results could be observed when water was added to the foam.
On the other hand, when the mineral oil or food oil was added to the silicone oil foam, as shown in Fig. 9(b), the foaming height increased. Only approximately 0.02 mass% of food oil were enough to increase the foaming height to a maximum of 68 mm, almost 3 times as high as foaming silicone oil. At the same time, approximately 0.07 mass% of the mineral oil was necessary to reach a foaming height of 135 mm, more than 5 times as high as the pure foaming silicone oil.
3.2.2. Behavior of AdditionsDuring the room temperature experiments, it was seen that all additions except water were evenly distributed in the foam, following the movement of the main liquid of the foam. The uniform distribution was even observed in the case of Cu addition, even though copper had a density of 8.96 g/cm3, while silicone oil had a density of 0.964 g/cm3. However, at a certain degree of additions, the solids started to accumulate in the bottom of the vessel. Water accumulated to bigger droplets and floated around in the bottom part of the foaming silicone oil, without disturbing the foaming silicone oil too much.
To facilitate the discussion, the high temperature experiments will first be discussed, followed by the room temperature experiments where a comparison will be made as well.
The method used for foaming the FeO–CaO–SiO2–MgO slag including a precipitated MgO∙FeO phase was successful. The combination of using Thermo-Calc26) to predict the compositions and then adjusting the compositions with a phase study was effective for obtaining slag samples with different particle fractions. The precipitated phase was found to be FeO∙MgO, see Fig. 2. The same solid particles were also found on the top part of the foaming slag, as shown in Fig. 6. The presence of the particles of MgO∙FeO on the top region of the foam indicated evidently the experiments were conducted with the intended two-phase mixtures.
As mentioned in the experimental part, about 10 wt% molybdenum oxide was dissolved into the slag phase from the crucible. Since all samples contained approximately equal amount of Mo in the liquid phase (Table 2), the effect of Mo on the interfacial tension between the gaseous, liquid and solid phases would be similar in all samples. Since the purpose of this study was to study the trends, the presence of Mo in the sample would have minor effect on the present discussion.
Two high temperature experiments were conducted under identical experimental conditions (both used Slag 1, 1.6 vol% MgO∙FeO) to examine the reliability of the results. As can be seen in Fig. 7, the reproducibility of the high temperature foaming experiments is evidently brought out by these two experiments, which show very similar foaming height approximately 170 mm.
The experimental result shown in Fig. 7 reveals that when small amounts of MgO∙FeO particles are present in the liquid phase, an increase in foaming height is to be expected. This is in accordance with the literature, referring to both laboratory experiments and industrial observation.13,15) On the other hand, when the amount of solids exceeds a certain limit, approximately 8 vol% in the present experimental setup, the slag foam starts to act sticky. In the present experiments, splash stains of the slag could be found on the inside wall of the crucible, rather than a mark from a foaming slag.
It should be mentioned that solid particles might decrease the reaction rate to certain extent, hence the gas generation, due to the occupation of contact area between the liquid slag and the metal. However, the situation is similar in both the experimental setup and the industrial process. Even though the kinetic conditions differ between the laboratory and the industry, the relative change should be comparable.
The results from the foaming silicone oil and food oil with addition of solid particles in Fig. 9(a) show that the foaming heights in the present systems slightly decrease or remain stable with increasing addition.
Ip et al. suggest that particles may act as elastic separator between the bubbles, preventing them from coalescing, stabilizing the foam,29) while Kaptay state that particles only stabilize the foam if the interfacial energy between the three phases is low enough.30) The interfacial force acting on a particle in a liquid and gas interface may be explained by Eq. (1),
| (1) |
| (2) |
It should be mentioned that Eq. (1) is only applicable for solid spherical particles in a foam. For a specific particle size and liquid surface tension, the interfacial force is maximum when the contact angle, θ, is 0° and minimum at 90°. Theoretically, the optimum contact angle to favor foaming has been derived from Eq. (1) and found to be between 50 and 90°.30) Although, depending on the particle size, surface tension and the structure that the particles distribute in the foam, acceptable (critical) contact angles for stabilizing a foam can be found in the interval between 20 and 180°.30) In the present study, copper and nylon are distributed in the two oils as Kaptay explain as either loosely packed single layer (LP1), loosely packed double layer of clustered particles (LP2C) or closely packed double+ layers (CP2+), depending on the mass fraction of the addition. Kaptay proposes that particles have a stabilizing effect on the foam if the θ is below 90° for LP1, below 129° for LP2C and below 180° for CP2+. Since both the oils wet easily on both copper and nylon, the contact angle is close to 0°, and evidently lower than the theoretically minimum critical contact angle of 20° as stated by the literature.30) This could explain why the particles do not stabilize the foam.
In a steelmaking perspective, the fact that different particles may either stabilize or rupture a foam could explain the controversy in the reported findings.6,7,13,14,15) So far, some experiments and industrial trials find that particles help to stabilize the foam, while others find that they may rupture the foam. While Eq. (1) could be of great help in understand the foaming behaviors of slag with solid particles, the lack of the interfacial tension data makes the evaluation almost impossible. It is reasonable to expect that the particles precipitating from the liquid slag, e.g. FeO∙MgO, would behave very differently from the additives, e.g. lime particles. While the former would have very low θ value since it is a phase of precipitation, the latter situation is more complicated. As pointed out by Sichen et al.,31) when a chemical reaction is taking place, the term of interfacial tension would become much less important in the wetting process. The prevailing of the chemical reaction would likely facilitate wetting. Nevertheless, study of contact angle between oxides relevant to BOF and EAF and the liquid slags are essential for a better understanding of the slag foaming.
In contrast to the effect of solid particles added to the foams, food oil and mineral oil greatly stabilize the foam of silicone oil and increase its foaming height. The emulsion of silicone oil and mineral oil in Fig. 5 indicates that the mineral oil is distributed evenly in the silicone oil as small droplets of <50 µm. It is therefore assumed that the insoluble droplets are distributed as loosely packed in the foam. The present study also include water. The high interfacial energy between silicon oil and water forces the droplets to accumulate into Ø3–5 mm drops. In accordance with the literature,25,30) it is evident when the drops are too big, they do not interfere with the foaming process.
It is interesting to note that the Copper particles in the present study did not instantly fall to the bath but were up to a point evenly distributed in the foam, following the movement of the foam. Since the density difference between silicone oil and copper is even bigger compared to steel and slag, one may assume that the metal droplets in an emulsion would also follow the movement of the slag foam instead of falling towards the metal bath. This is in accordance to a previous study, where it was shown that when Ø2 mm iron particles were put into a sugar solution foam, the iron particles are almost stationary in the foam.32) Nevertheless, the effect of the residence time of the particles in the liquid on foaming need further careful investigation.
Both room temperature experiments and high temperature experiments have been conducted to study the effect of solid particles or liquid droplets on slag foaming. The results at room temperature revealed that additions of some liquids stabilized the foam and increased its foaming height, while other did not affect the foaming height. When solid particles were added to the room temperature foams, the foaming height slightly decreased or did not change. In the high temperature experiments, the highest foaming height was measured when 1.6 vol% solids were present in the slag, but already at approximately 8 vol% solids, the slag height decreased significantly, and the slag appeared sticky. It appears that the interfacial energies between the gaseous, liquid and addition phases play a profound role. According to the literature, for solid particles, the optimal contact angle between the three phases to stabilize the foam is between 50 and 90°. Although, depending on the packing structure of the third phase in the foam, the critical contact angle can vary between 20 and 180°. In the present study, all solids were wetted by the liquid phase, resulting in a contact angle of less than 5°. This could explain the why the foaming height did not increase.
This work has been conducted as part of the HYBRIT research project RP1. We gratefully acknowledge financial support from the Swedish Energy Agency. HYBRIT (Hydrogen Breakthrough Ironmaking Technology) is a joint initiative of the three companies SSAB, LKAB and Vattenfall with the aim of developing the world’s first fossil-free ore-based steelmaking route.
