A B S T R A C T

Open-cell ceramic foams are widely used as filters for the purification of molten metals due to their high
permeability and high efficiency for capturing non-metallic inclusions. With the consideration of manufacture
and filtration behaviour, most filters are produced by replica methods. The intrinsic multiscale porous structure
requires the ceramic foams to have sufficient strength. As a consequence, different processing approaches,
material compositions and characterizations are investigated. Meanwhile, to further enhance the filtration
performance, the functionalization of ceramic foam filters is often conducted with surface coating. This review
summarizes the processing and characteristics of ceramic foam filters (CFFs), the nature and formation of nonmetallic
inclusions (NMIs), the filtration mechanism and applications of CFFs. The characterization of mechanical
and filtration performance, as well as the corresponding strengthening and functionalization methods
are also presented.

1.Introduction

Due to their unique features, such as high permeability, high relative strength to density, high specific surface area and high chemical resistance, the cellular ceramics have gained intensive interest in the last decades. Cellular ceramics have been widely used as support for combustion and catalysis, lightweight structures, biomedical components, as well as fluid filtration (e.g. gas, water, melt) [1–3]. According to their morphology, they can be classified as foams, honeycombs, connected rods/fibers, hollow spheres etc. [4]. Molten metal filtration is one of the most important functional applications of ceramic foams. Melt filtration requires high permeability and a high flow rate to prevent the metal from freezing inside the filter porosity. Therefore, open-cell ceramic foams (also named reticulated ceramic foams) with high specific surface area are one of the most popular choices for filtration. Since the first introduction (in 1974) and commercial application (in 1976) for molten aluminum filtration [5], the ceramic foam filters (CFFs) have been adopted subsequently for iron (in 1983) and later for steel, copper alloys, magnesium alloys etc. from the aspect of application or research [6,7]. Considering the service conditions, CFFs generally require the material/structure to offer suitable refractoriness, good thermal shock resistance, chemical resistance to metal and slag, adequate strength at room temperature (to survive the transportation and installation) and at operating temperature (to withstand the thermo-mechanical stresses). As the CFFs are applied in various metallurgical processes and are normally single-use only, the geometry adaptability, easy production and low cost are also important for industrial application. Nowadays, the replica methods patented by Schwartzwalder and Somers as well as by Holland in 1963 are the most widely adopted methods for producing industrial CFFs [8,9]. Following the intensive research interest in cellular materials and structures, some review papers concerning cellular ceramics in general [4,10–13] and specific melt filtration [14–17] have already been published by others. Thereinto,Grauckler and Brockmeyer proposed the feasibility of CFFs for metal melt filtration, the general requirements and the filtration model [14,15]. Oslon highlighted the fundamental

filtration theory and applications of CFFs in aluminum and iron/steel production [16]. Buchilin reviewed the application of CFFs for aluminum melt filtration [17]. Group from TU Bergakademic Freiberg has also published series of research paper in a book presenting the novel techniques and materials of intelligent filter system as seen in Ref. [18]. Compared with previous reviews, this paper delivers a summarization of ceramic foams used for filtration and produced by replica methods, including the general theory of melt filtration, the latest progress in processing, filtration process, functionalization, characteristics/ properties of CFFs as well as their applications. The characterization and improvement of the most important two aspects of CFFs, i.e. mechanical properties and filtration efficiency, are systematically summarized. Additionally, the information of non-metallic inclusions is innovatively reviewed, which is relative important as the reaction among NMIs, molten melts and CFFs is distinct and has great influence on filtration behavior. Due to the large amount of literature in this field, it is not possible to cover all of them and further reading is suggested for specific topics.

Processing and characteristics of CFFs

2.1. Processing of CFFs by the replica method

2.1.1. General processes

Commonly, three methods are mostly used for producing open / closed-cell ceramic foams, including replica method (also called template replication method or Schwartzwalder procedure, produced foams having porosity range of 40–95 %, pore size of 200 μm to 3 mm), direct foaming (porosity range of 40–97 %, pore size of 35 μm-1.2 mm) and sacrificial template technique (porosity range of 20–90 %, pore size of 1–700 μm) [19]. Direct foaming can manufacture both open- and closed cell ceramic foams, while the sacrificial technique mainly forms the closed-cell ceramic foams. However, the main drawback is the insufficient foam stability, when producing large cells and thus impending collapse of individual cells or even the whole cellular structure. Besides the above-mentioned well-known processing methods, there are also exist other advanced methods for fabricating open-cell structure, e.g. so-gel derived method, particle-stabilized foaming method, foams derived by porous geopolymer precursors etc.
However, with regard to the high accessible porosity, the wide range of pore sizes (especially of pore sizes >1.2 mm) and cost/productivity, the replica process is widely adopted for manufacturing CFFs instead of rest methods and so the focus of this work is at the replica process. One recent review paper summarized the current state of open-cell ceramic
foams by replica technique [13]. The templates used in the replica process should have the following basic characteristics: proper cell size distribution, elastic recovery after squeezing, and complete burnout after heat treatment. In the replica process, the polymer foam (commonly polyurethane (PU)) is firstly coated with a ceramic slurry and thermally treated whereby the temperature treatment of the foams comprises drying, burnout of the polyurethane foam and sintering of the ceramic. Low heating rates are selected for the burnout of the polyurethane foam so that the decomposition and diffusion of the decomposition products (of the PU) can take place slowly to prevent excessive pressure building up in the webs, which would cause cracks in the material [19,20]. According to the thermal analysis of PU, the decomposition mainly occurs between 250 and 450 ◦C and is completed at 600 ◦C [21]. The sintering temperature depends on the ceramic material used and the desired material porosity. The PU foam is burnt out and leaves behind a replica of the original foam. Hence, the morphology of the PU foam is decisive for the CFFs. The general processing procedure is shown in Fig. 1. The concave cavities, voids and cracks formed during the thermal decomposition of polymer template negatively affect the mechanical strength of the CFFs. At the same time, the burn-out of PU template releases gases (such as NOx, HCN, CO etc.), which are hazardous for both the environment and humans.

To overcome the drawbacks related to the replica method, some improvements and new technologies have been proposed. To strengthen the foams, several routines can be performed: i) improving the wetting behavior of the polymer template [22], ii) better controlling the pyrolysis process [23,24], iii) improving the properties of the ceramic slurry by additives or adjusting the particle packing [25,26], iv) ameliorating the surface quality by re-coating or filling the strut cavity by infiltration [3,27]. To become more environmentally friendly, less toxic binder systems [26] and templates (polystyrene
mold-scaffold [28] and water-soluble polyvinyl alcohol (PVA) [29]) have been used. Direct and indirect additive manufacturing (AM) have also been applied for the processing of porous ceramics. The former directly builds the porous structure using ceramic powders, colloidal solutions or pastes with various techniques, such as filament deposition, selective laser sintering etc. [30,31]. The latter (also known as hybrid AM) prints polymer cellular structures used as the template for the following replica process of ceramics [32,33]. Although the AM technology is still rarely used for industrial CFFs production because of the cost and productivity factors, its application for highly wrought parts purification and other
special cases is promising. Some attempts have already been made. Selective laser sintering was adopted to produce a cylindrical PU template for fitting the hollow-ware component in bottom-teaming ingot casting [34]. Bock-Seefeld et al. produced carbon-bonded alumina filters with the AM of PVA template by laser sintering and alginate-based slip
coating [29]. With this printed water-soluble template, the reproducible fabrication and precise geometry design are guaranteed for CFFs without the need for thermal decomposition of the template. Direct casting of CFFs is also possible. Wetzig et al. produced carbon-bonded alumina spaghetti filters by alginate-based robo-gel casting, which survived the steel immersion test and offered increased strength as well as sufficient permeability [35].

2.1.2. Slurry and compositions

2.1.2. Slurry and compositions In general, the structural uniformity and pore control are important for the usability of CFFs, which are mainly influenced by the slurry and production processes (coating and firing). Thereinto, the ceramic slurry for coating is a disperse system containing ceramic powders, liquid medium and various additives, such as binders, deflocculants, thickeners, wetting agents and plasticizers. The slip density and rheological behaviour of slurry is a function of solid content and additives. The slurry should be fluid enough to coat the polymeric template entirely, but sufficiently viscous to remain on the surface. A shear thinning slurry fulfils these requirements, while Newtonian fluids do not show a proper coating behavior [36]. A good coating behavior was achieved for foams with lower or higher ppi numbers of 50 using a thixotropic slurry with a viscosity of 8600 mPas (at 20 s􀀀1) or shear-thinning slurry with a viscosity of 130 mPas (at 20 s􀀀1), respectively [37]. Thicker struts are normally produced by the slurry with higher viscosity [38]. Normally, a thinner slurry is adopted for the re-coating process. The shear thinning rheological behavior can be achieved by the adjustment of thickener and solid content. By comparing slurries with different kinds of thickener, the slurry with a shear-thinning index n lower than 0.85 (fitted with the power law fluid model of Ostwald de Waele) showed a good coating performance [39]. The increasing content of stabilizing agent can also enhance the loading of slurries on sponge because of the high viscosity and thixotropy [40]. Furthermore, the preparation of slurries by ball milling involves a fine and homogeneous foam structure, while foams produced by the coating using a high shear mixer possess larger pores as the ball milling crushed the Carbores P grains (carbon binder) [41].
Compared with other additives, the binders are more intensively studied because of their great importance on the performance of CFFs. There are binders used to increase the strength before and after thermal treatment. An example of the latter is a phosphate binder, which is used to reduce the sintering temperature, as one way to control the produc
tion cost and energy consumption. For example, phosphate allows the reduction of the sintering temperature for pure alumina from 1600 ◦C to 1300 ◦C. However, Bergin et al. reported that very low amounts of magnesium in aluminum melts (0.00035 wt pct) could react with the phosphate-bonded alumina filters and cause severe degradation as well as the formation of the toxic phosphine, whose explosion is further
crease with temperature [42]. Hence, phosphate-bonded CFFs are not recommended for magnesium-containing alloys. A further example is the application of carbon binders (such as pitch, tar and phenolic resin) for the preparation of carbon bonded ceramic foam filters. For carbon-bonded filters, the carbon-containing binders form a carbona
ceous matrix between the ceramic particles and present good thermal shock resistance, low wettability by melts and high creep resistance. Normally, synthetic resins, pitch or bitumen are used, which offer good performance but are hazardous since they contain the carcinogenic benzo[a]pyrene. Several alternative binders, such as a modified coal tar pitch with the commercial name of Carbores®P, have been studied. The development of novel Al2O3–C filters based on Carbores®P was studied in Ref. [26]. It has been proved that the application of fine-grained binder and increased fraction of Carbores®P resulted in an increased open porosity and mechanical strength for filter material Al2O3–C [43]. The application of Carbores®P reduces the benzo[a]pyrene to <300 ppm. However, the European REACH regulations impose strict protocols for transporting and processing polycyclic aromatic hydrocarbons (PAHs). To eliminate the benzo[a]pyrene from the system, eco-friendly binders such as lactose and tannin are beginning to be used in the pro
duction of carbon-bonded refractories [44]. Their promising application of carbon-bonded filters for steel melt filtration has been reported in Ref. [45] for the first time. With their lower carbon yield, the mechanical properties of foams are degraded. Although the addition of n-doped silicon or alumina coating can help to a certain extent, further improvement is still necessary for achieving satisfying mechanical strength and thermal shock resistance. Thus, these environmentally friendly binders (in combination with Carbores®P) have been inten
sively investigated [46]. Their pyrolysis process and microstructure evolution characterized showed that an increase of the cluster sizes, decay of disorder and lattice parameters changes happened with tem
perature [47].

2.1.3.Coating processes
For polymer template coating, a dip coating process is always carried out as the first step. The template is soaked with the slurry by dipping the polymer template into the slurry. Mechanical pressure by squeezing is applied during dipping to achieve a complete coating of the template. The coated template is passed between rollers or centrifuged [48] to remove the excess slurry after the impregnation. The optimal distance between the rollers is equal to 20 % of the foam’s thickness [36]. Incomplete removal of the excess slurry may lead to the formation of closed cells, which influences pressure drop, permeability behavior and filtration efficiency. The majority of CFFs are produced by the impreg
nation/rolling process for the first coating, whereas dip/centrifugation or spraying are used for the following layers. Both dip/centrifugation and spraying are applicable for the first and second coating process. The second coating is for reducing the surface defects and adjustment of the foam morphology. By investigating the structure, surface properties and strength of CFFs, different coating routes have been compared and summarized [40,49].
Compared with the impregnation and rolling followed by a cold spray process proposed by Emmel et al. [26], centrifugation is more suitable for the overall adjustment of the coating thickness, homoge
neous material distribution, high number of open cells and foams with increased dimensions [50–53]. Lower mean material pore size was identified for Al2O3–C filters coated by centrifugation (150 nm) compared to that of the spray-coated filters (200–500 nm) [40]. Besides, short-distance cold spraying or a complete dipping and rolling pressing before the centrifugation are beneficial for achieving the desired ho
mogeneity and proper mechanical strength [48]. The authors also pointed out that a low-viscosity alumina slurry with lower solid content is appropriate for centrifugation coating, while the spray coating used a slurry with higher viscosity during the second coating step. Luchini et al. [49] tried to perform an additional thermal treatment between the centrifugation steps for the first and second coating. The formation of macro-cracks was prevented due to the additional thermal treatment, while a high number of micro-cracks developed and lowered the strength. The effect of the coating process (spray or dip/centrifugation) and sintering step (one or two times) was compared for spinel-coated alumina foams. Better mechanical strength and uniformity of foam structures were obtained by combination of centrifugation and two-time sintering [48].
In addition, the application of a flame spraying coating step offers a different approach [34,54,55]. During flame spraying, the molten coating material is sprayed onto the carbon bonded alumina foams and solidifies with a high cooling rate. Due to the impact, droplets of ma
terial deform into flat lamellae and a dense layer is formed. For example, flame-sprayed alumina coatings on Al2O3–C filters have a lower porosity compared to cold sprayed coatings, which inhibits the formation of gas bubbles during contact with a steel melt. On the other hand, the cold sprayed alumina coating has high reactivity towards the steel melt [54]. Electrophoretic deposition can also be used to coat electrically conductive foams with stable suspensions with sufficient surface charge. Moritz et al. examined the feasibility of coating carbon-bonded foams with alumina using electrophoretic deposition [56]. When properly designated and implemented, a channel-like alumina coating can be successfully formed. Voigt et al. [57] later improved the process pa
rameters in order to limit the amount and size of cracks in the sintered coating.

2.2.Characterization of CFFs

Of particular interest are structural, mechanical, thermo-mechanical, thermal properties as well as chemical and phase composition for CFFs. This chapter would focus on characterization methods which supports particular properties of interest as above [11,58,59].
2.2.1.Structural properties
Since the macro and microstructure have a significant influence the properties of CFFs, a detailed description of the foam characteristics is needed. There are three types of pores: functional pores, material pores in the struts and strut cavities formed by the template burn out, as seen in Fig. 2. To characterize the foam structure, different kinds of porosities and densities can be determined: i) foam density is the bulk density of the foam and is calculated by dividing the weight by the foam contour volume; ii) true density is the material density without any pores and can be measured by helium pycnometer on pulverized foams; iii) relative density is the ratio between foam density and true density, that can be used as an indicator for the foam’s total porosity; iv) strut bulk density and open porosity of the struts are commonly determined by the Archimedes method. Mercury intrusion porosimetry provides another way to measure the bulk density of the struts, open porosity of the struts and pore size distribution (of the strut cavities and the material pores).

The influencing parameters on the measured results for intrusion and extrusion mercury porosimetry of ceramic foams have been summarized by Voigt et al. [60]. Optical methods, such as light microscopy and scanning electron microscopy, are often used for measuring the strut tapering, strut diameter, cell window size, fractured struts and so on. The functional porosity of the filter, functional pore size and strut diameter are also quantified with microfocus X-ray computed tomography.

2.2.2. Mechanical properties of ceramic foam filters The basic requirements for CFFs are high filtration efficiency as well as sufficient thermo-mechanical and mechanical resistance. Mechanical and thermo-mechanical stresses mainly occur during the transportation/ handling, during the first impact of the molten metal and during priming. Current chapter focuses on the mechanical properties of CFFs as well as on their characterization, which are governed by the intrinsic properties of the ceramic material, the local (e.g. struts strength) and global structural properties (e.g. cell size, relative density and defect size). The fine-grained filter materials are similar to regular ceramic materials, whose mechanical properties are well understood and plenty of investigation methods are available [61–64]. Meanwhile, the strut strength can be measured by the rupture of an individual strut within a foam under bending [65]. The measurement, which is also useful to determine the influence of critical strut defects, is available for foams with large cell size and long struts. For foams with small cell size, the compressive strength of struts can be calculated based on the strength of the foam as well as its relative density and geometrical parameters (e.g. length and strut thickness) [66]. The correlation between the fractography of markings and the strength of various materials can also be used to determine the strength of individual struts [67]. In this section, the mechanical properties of fine-grained filter materials and struts will be less extended, but focus on the mechanical properties of CFFs. Due to the open-porous structure of CFFs, the classical elasticity theory is not valid for ceramic foams. Numerous studies have discussed the dependency of mechanical properties on the foam’s structural characteristics. Hagiwara et al. [68] found that hollow struts are correlated to a higher specific stiffness, expressed by the dynamic elastic modulus. The work of Brezny et al. on vitreous carbon [69] suggested no variation of the elastic modulus with the cell size, while Dam et al. showed an increased elastic modulus with the cell size [70]. Grabenhorst et al. studied the influence of sample size and macro pore size using different dynamic (impulse excitation technique, ultrasonic wave velocity method) and static (three-point bending test) techniques for Young’s modulus determination [71]. Generally, ceramic foams present
a brittle mechanical behavior and their failure begins after the elastic limit [66]. The load is transmitted as discrete forces and the cell edges elastically deform until fracture. The discrete strut cracks coalesce into macrocracks and cause structural damage of ceramic foams. This progress brings about three stages in the stress-strain curve under compression: an increase at a low strain level (linear elastic deformation), a plateau with constant stress (cell edge yield or fracture) and a fast increase until the end of the test (densification due to crushing). As there is no standard available for the measurement of the mechanical strength of ceramic foams, widely differing methods and parameters are used. In the literature and industry, compressive strength (or so-called crushing strength) and, in some cases, flexural strength can be found. In the area of compressive strength, the measurement parameters found in literature vary greatly [26,70,72,73]. Voigt et al. [74] showed that the measuring parameter (ceramic foam size, size of loading plate, usage of compliant pads and homogeneity of foam) have a major influence on the compressive strength determined and therefore a comparison of compressive strengths determined with different measurement parameters is not possible. For a reliable measurement, the testing specimen should contain a minimum of eight cells in total to achieve a representative volume [75]. The linked evaluation of microtomographic scans by in-situ X-ray CT and load-displacement curves by compressive test showed the 4-dimensional damage process of typical CFFs [76]. The foams exhibited elastic behavior at the early loading stage and then individual struts and edges fractured locally with increased loading until a complete structural damage was achieved. The mode I and mixed-mode fracture of foams were also characterized by bending test, direct uniaxial tensile test, Brazilian test and others. Neumann et al. compared the fracture statistics of open-cell alumina ceramics under compression and bending [77].

A direct uniaxial tensile test was applied for open-cell alumina-mullite foams by fixing the specimens in position with an epoxy bond layer [72]. The acoustic emission combined with video observation further proved the brittle failure behavior of rigid ceramic foams. Due to the difficulties related to the damage at loading points and continuous stress transmission, the Brazilian test for cellular materials cannot always show macroscopic damage. With the application of a curved loading jaw, compliant pads and an adjusted loading routine, Dai et al. successfully performed the Brazilian test for open-cell alumina foams and correlated the biaxial strength with the foam cell size and toughness of foam material [78]. Mechanical testing at higher temperatures is more complicated than testing at room temperature, so investigations are limited. Goretta et al.
[79] investigated the compression strength of Al2O3 foams between 800 ◦C and 1500 ◦C and found an increase of strength until 1200 ◦C followed by a degradation. Wu et al. conducted high-temperature compressive tests for Al2O3–C foams with/without coating using an inductive heating system and found that the high-temperature compressive strength was generally higher than the cold crushing strength because of the plasticization processes at elevated temperature
[80]. At the same time, the applied alumina and carbon-containing alumina coatings improved the filter’s strength. An increase of high-temperature compressive strength was also observed for Al2O3–C filters with a coating based on carbon nanotubes and alumina nanosheets/ spheres [81]. Bergin et al. [82] tested the compression strength of commercially available CFFs (Al2O3 with different binders) at room temperature and 730 ◦C, whereby no separation between the influence of the high temperature and thermal shock was conducted. Furthermore, a method for measuring the compressive strength of CFFs submerged in molten aluminum was developed [82]. The high temperature reduced the strength in the range of 9.2–58.6 % compared with the one measured at room temperature and the immersion into melts further decreased the strength up to 69.4 %. This study confirmed the vulnerability of CFFs in contact with melts and gave a new insight into the mechanical performance at real service conditions. An illustration of the experimental set-up is given in Fig. 3. For the prediction of mechanical properties, the unit cell developed by Gibson et al. [83] is often used. With the aid of the unit cell, specific formulas for the modulus of elasticity, compressive strength, fracture toughness and thermal shock resistance as a function of the relative density can be derived. A distinction is made between foams with open and closed pores and according to the type of material used (brittle, elastic, ductile). The formulas calculate the properties of the foam by relating the relative density to the properties of the corresponding dense material [83]. However, the formulas only apply to ceramic foams with a relative density <0.1. Additionally, the above-mentioned analytical formulas are not universal. With the consideration of the macroscopic functional pores as pre-existing critical defects for reticulate CFFs, an Irwin equation bases can simply the correlation between the strength and foam structure. 2.2.3. Thermo-mechanical properties of ceramic foam filters During their application for molten metal filtration, CFFs are exposed to high temperatures and sudden thermal shock. High thermomechanical stresses during priming as well as creep can cause failure of the CFF followed by insufficient filtration or even additional contamination of the metal melt. Thus, the high-temperature mechanical properties of CFFs are of great importance for evaluating their survivability and durability. Thermal shock resistance gives information
about the ability to survive thermal stresses whereby the ability depends on a large number of parameters, such as material properties, geometry, complexity of the samples as well as heat transfer conditions [84]. Under cold shock conditions, the surface experiences a transient tensile stress while a compressive zone is developed at the center of the plate. During the cold shock, a crack initiates at the surface. Under hot shock conditions, the center of the plate is under tensile stress and is prone to cracking; spalling of the surface layers due to large compressive stresses is also possible [84]. CFFs are a special case as the struts are typically smaller than 1 mm, so the temperature gradient within the sample is small even though the heating and cooling processes are fast. Regarding CFFs, the impingement test is an often described and industrially applied test to investigate the thermal shock resistance [39, 85–87]. The impingement test gives just a qualitative information whether the filter survives a filtration process, but does not provide any quantitative information on the consequences of the thermal shock. An often-used possibility is to test the thermal shock resistance by quenching in water or compressed air. Orenstein et al. [88] used water and oil to test open-cell Al2O3 foams with varying functional pore size and found an increase of the thermal shock resistance with increasing cell size but weak correlation with the relative density. Vedula et al.[89] developed a damage parameter which was directly proportional to the residual elastic modulus measured at silica bonded silicon carbide, alumina, alumina-zirconia and zirconia–alumina foams. Gumban et al.

[90] tested the influence of different testing parameters (descending/ ascending thermal shock, preheating, temperature, quenching medium and so on) on the thermal shock resistance of Al2O3 foams. Furthermore, the authors compared the experimental data of 5 different CFF materials with data calculated by the thermal shock resistance model of Gibson et al. [91]. Regarding the creep of ceramic foams, there is only little information available. Goretta et al. [79] performed creep tests on open-cell alumina ceramics above 1200 ◦C and found that the creep deformation was linearly increase at the lower temperature range, while the strain rate increased with the formation of creep cracks. A further study investigated the creep of ZrO2 foams. Gibson et al. [83] proposed a model for estimating the creep of ceramics as follows:

Here, ˙ε and ˙εos are the strain rate related to the strut material and strut, respectively, while ˙ σ and ˙ σos are the corresponding stresses and n is the strut creep parameter. Compared with other ceramic foams, the time and temperature dependent creep behaviour and refractory under load are relative important for porous ceramics used as melt filters, which served at relatively high temperature. More relevant research should be concerned and performed.
2.2.4. Strengthening of ceramic foam filters
During the template pyrolysis/burnout and shrinkage related to the solidification and sintering, the strut cavities, microcracks and flaws are formed. Besides a slow heating schedule, further improvement and strengthening of CFFs are desired. Recoating and infiltration of the hollow struts have been explored for improving the mechanical strength.
Recoating of CFFs means the application of additional coating layers with two possible goals.

– Closure of strut cavities, microcracks or flaws (formed during burnout) for an improve of the mechanical properties

-Application of coating layers with other chemistry as skeleton foam to achieve new or improved functionality (i.e. improved filtration performance) – functionalization

The recoating process can be the repetition of slurry impregnation and drying processes [92], recoating of the foam after sintering [50], a chemical vapor deposition [93], electrolytic deposition [56] and others. The results of a recoating are typically with larger strut diameter, minimized surface flaws, decreased permeability, higher strength and corrosion resistance. Functional coatings in the area of metal melt filtration have been developed to improve the filtration performance. The influence of recoating of CFFs on filtration is discussed in the metal melt filtration chapter. This recoating goes typically along with changes in the mechanical properties. Emmel et al. [94] compared Al2O3–C foams coated with different ceramic materials: a decrease of compressive strength in the order of alumina, spinel and mullite was observed, in relation to their sintering activity. In general, the shrinkage of the coating materials around struts after the sintering together with the thickening of struts contributes to better mechanical strength. For example, both alumina and carbon-containing alumina had a positive effect on the filter’s high-temperature strength [80]. Filters coated with multi-wall carbon nanotubes-based slurries showed lower strength than uncoated ones due to the partial oxidation during the second heat treatment under reducing atmosphere after the coating procedure [95]. However, a higher binder content in the coatings delivered a slightly higher mechanical strength. Additionally, Ranglack-Klemm et al. tested the high-temperature performance of Al2O3–C coated with a coating based on carbon nanotubes with alumina nanosheets or alumina nanospheres [81]. With the
application of these functional coatings, the compressive strength

increased at 1100 ◦C. The authors also mentioned the insufficient adhesion of the coatings, which limited the CFFs for a short-term filtration service. Filters coated with an oxide layer had a higher compressive strength compared to the reference ones, due to compression stresses on the struts generated by the sintering shrinkage as well as the thickening of struts.
A higher increase of the mechanical properties is reached with a filling of the strut cavities by an infiltration process. For example, researchers infiltrated SiC ceramic foams with silicon and the mechanical strength was significantly increased [3,96]. In addition, suspensions of colloidal silica, alumina, refractory oxides and aqueous sol of aluminum or zirconium hydroxides can also be used [97,98]. Salvini et al. produced alumina foams and added a vacuum infiltration by alumina and SiO2 colloidal suspensions with different solid contents, particle sizes and infiltration times. The best result for strength and strut diameter was achieved with 40 wt% SiO2 for 1 min infiltration time [99]. Both vacuum and ambient pressure infiltration techniques have been developed [27,100]. Additionally, the reduction of the residual stress on the struts by recoating and infiltration of materials with different thermal expansion coefficients compared to the substrate material eliminated the microcracks on the struts, which is beneficial for the mechanical strength and thermal shock resistance [101]. However, as filters are single-use products, the infiltration process is much less used for CFFs compared to recoating methods due to its higher cost and more complex and time-consuming procedure.
The use of additives during the processing of the CFFs can also improve the mechanical strength. A positive effect from the introduction of metallic or inorganic fibers on the mechanical properties and heat resistance has been reported [16]. Liang et al. added fly ash into SiC reticulated ceramics, whose strength increased thanks to the formation of mullite [102]. Meanwhile, the creation of microporosity in the struts increased the contact angle with the molten metal and the CFFs were believed to have a better filtration efficiency [103]. Nano-engineered Al2O3–C filters with carbon nanotubes and alumina nanosheets additives formed Al3CON during heat treatment and featured better mechanical properties [104].

3.Non-metallic inclusions (NMIs)
Besides the CFFs themselves, the nature of NMIs and their interaction with specified melts and filters are another dominating factor for filtration efficiency. Thus, the effect of NMIs, typical NMIs in different metals and characterization of NMIs for proving purification efficiency are summarized in this chapter before heading to filtration of CFFs. Inclusions basically constitute a cloud of phases dispersed in the metal matrix, introducing discontinuities. In some alloys, very fine cipitates, which interact with the metal matrix at the scale of tions or even at atomic scales, are deliberately created by heat treatment after production (precipitation hardening) to improve the yield strength [105]. The technology referred to as “oxide metallurgy”’ aims to use inclusions as nucleation sites for the iron matrix in desirable form and structure. Additionally, some secondary inclusions can prevent grain growth in steel, with positive effects on the mechanical properties [106]. Nevertheless, non-metallic inclusions have for the most part a significant negative influence and should be avoided.

Non-metallic inclusions can be classified in terms of their origin: Exogenous inclusions enter the melt from the outside, for example through erosion of the refractory material [107,108]. Endogenous in clusions are formed directly in the melt, such as the formation of alumina by the reaction of the aluminum with atmospheric oxygen and/or oxygen dissolved in the aluminum. In the area of steel nous inclusions are denoted as either “primary” inclusions – generated after the deoxidization process -, “secondary” and “tertiary” inclusions – formed during cooling and solidification -, or “quaternary” inclusions – generated below the solidus temperature [108]. Another distinguishing feature is the inclusion form. The non-metallic inclusions can form individual particles, agglomerates or films.
3.1.Typical inclusions in the various metals
The chemistry of the non-metallic inclusions depends on the used metal and the used alloying elements but there are commonalities. Non- metallic inclusion made of alumina (Al2O3) can be found in steel, aluminum and aluminum-containing copper alloy melts and exist as particles, agglomerates and films. The following tables show an over
view of typical inclusions found in steel (Table 1), iron (Table 2), aluminum (Table 3), copper (Table 4) and magnesium (Table 5).
3.2.Effect of inclusions on the processing and properties of metals
The presence of non-metallic inclusions lowers the melt flowability [137,138], the machinability and workability [132], mechanical properties [132,139–142], corrosion resistance [107], as well as the surface quality of metal products. Furthermore, a higher amount of casting porosity caused by non-metallic inclusions was overserved by Tian et al. [143].
The damage effect of the inclusions depends on the type of inclusion form (agglomerate/particle/film), the chemistry, the size of the inclusion and the cast metal. In the area of steel casting, Zhang et al. [107] defined tolerable maximum inclusion sizes varying for different steel products, such as for wire, ball bearings etc., which generally should be smaller than 20 μm. Brown et al. [132] mentioned fatigue strength reduction factors caused by surface imperfections in ductile iron. According to Hedjazi et al. [142], the tensile strength and elongation at rupture of aluminum decrease as the number of inclusions per area increases. Moreover, the decrease in tensile strength and elongation at fracture of an Al–Cu–Mg alloy is larger for macro-particle inclusions (>50 μm) than for micro-particle inclusions (<50 μm) for the same area of inclusions. In most cases, whether a non-metallic inclusion is mental to steel properties or not depends on its size: sometimes a strophic defect is caused by just a single large inclusion in a whole steel heat. The particle sizes range from a few micrometers for primarily endogenous inclusions, up to a few millimeters for their agglomerates or exogenous particles. Though the large inclusions are far outnumbered by the small ones, their total volume fraction may be larger. The size of the inclusions allows for further differentiation and gives also dications about the moment these particles were generated. A tion was proposed between macroscopic inclusion (>20 μm, mainly exogenous inclusions or clusters of small inclusions), microscopic
clusions (1–20 μm, primary, secondary or tertiary inclusions) and sub-microscopic inclusions (<1 μm: quaternary inclusions) [144]. It has been demonstrated that the frequency of macroscopic inclusions decreases exponentially with increasing the inclusion diameter. Taking 1% C–Cr-bearing steel as example, the influence of series inclusion types on stress-raising properties has been summarized in Ref. [145]. Thereinto, the calcium aluminates, spinels and silicates are detrimental for fatigue properties, while the rest inclusions (i.e. sulphides, iron oxides etc.) are non-detrimental or even possibly beneficial.

Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References
[1]Wang L, An L, Zhao J, et al. High-strength porous alumina ceramics prepared from stable wet foams. Journal of Advanced Ceramics 2021;10:852–9.
[2]Wen Q, Qu F, Yu Z, et al. Si-based polymer-derived ceramics for energy conversion and storage. Journal of Advanced Ceramics 2022;11:197–246.
[3]Liang X, Wan F, Li Y, et al. The enhanced thermal shock resistance and combustion efficiency of SiC reticulated porous ceramics via the construction of multi-layer coating. Ceram Int 2023;49:9097–103.
[4]Colombo P. Conventional and novel processing methods for cellular ceramics. Philosophical Transaction of the Royal Society A 2006;364:109–24.
[5]Yarwood JC, Dore JE, Preuss RK. Swiss aluminium limited. US Patent; 1976. US3962081.
[6]Bakke P, Engh TA, Bathen E, et al. Magnesium filtration with ceramic foam filters and subsequent quantitative microscopy of the filters. Material and Manufacturing Process 1994;9:111–38.
[7]Sahoo M, Barry JR, Kleinschmidt K. Use of ceramic foam filters in the brass and bronze foundry. AFS Transactions 1981;89:611–20.
[8]Schwartzwalder K, Somers AV. Method of making porous ceramic articles. US patent; 1963. US3090094A.
[9]Holland IJ. Method of making a porous shape of sintered refractory material. US Patent; 1963. US3097930.
[10]Bhaduri SB. Science and technology of ceramic foams. Adv Perform Mater 1994; 1:205–20.
[11]Fey T, Betke U, Rannabauer S, et al. Reticulated replica ceramic foams: processing, functionalization, and characterization. Adv Eng Mater 2017;19:

[12]Ohji T, Fukushima M. Macro-porous ceramics: processing and properties. Int Mater Rev 2012;57:115–31.
[13]Haase D, Fussel A, Adler J, Petasch U. 60 years of open-celled ceramics based on replica technique: applications, obstacles, and opportunities. Adv Eng Mater 2024:2301804.
[14]Gauckler LJ, Waeber MM, Conti C, et al. Ceramic foam for molten metal filtration. Journal of metals 1985;37:47–50.
[15]Brockmeyer JW, Aubrey LS. Application of ceramic foam filters in molten metal filtration. Application of Refractories: Ceram Eng Sci Proc 1987;8:63–74.
[16]Olson RA, Martins LCB. Cellular ceramics in metal filtration. Adv Eng Mater 2005;7:187–92.
[17]Buchilin NV, Maksimov VG, Babashov VG. Ceramic filters for aluminum melt. Glass Ceram 2015;72:246–52.
[18]Aneziris C, Biermann H. Multifunctional ceramic filter systems for metal melt Filtration,Towards zero-defect materials, Springer Cham, doi.org/10.1007/978- 3-031-40930-1.
[19]Studart AR, Gonzenbach UT, Tervoort E, et al. Processing routes to macroporous ceramics: a review. J Am Ceram Soc 2006;89:1771–89.
[20]Dressler M, Reinsch S, Schadrack R, et al. Burnout behavior of ceramic coated open cell polyurethane (PU) sponges. J Eur Ceram Soc 2009;29:3333–9.
[21]Zierath B, Greil P, Stumpf M, et al. Enforcing of mechanical properties of alumina foams. Advances in Ceramic Armor, Bioceramics, and Porous Materials: Ceram Eng Sci Proc 2017;37(4):149–62.
[22]Jun IK, Koh YH, Song JH, et al. Improved compressive strength of reticulated porous zirconia using carbon coated polymeric sponge as novel template. Mater Lett 2006;60:2507–10.
[23]Han YS, Li JB, Wei QM, et al. The effect of sintering temperatures on alumina foam strength. Ceram Int 2002;28:755–9.
[24]Schmidt A, Salomon A, Dudczig S, et al. Functionalized carbon-bonded filters with an open porous alumina coating: impact of time on interactions and steel cleanliness. Adv Eng Mater 2017;19:1700170.
[25]Paiva AEM, Sepulveda P, Pandolfelli VC. Processing and thermomechanical evaluation of fibre-reinforced alumina filters. J Mater Sci 1999;34:2641–9.
[26]Emmel M, Aneziris CG. Development of novel carbon bonded filter compositions for steel melt filtration. Ceram Int 2012;38:5165–73.
[27]Vogt UF, Gorbar M, Dimopoulos-Eggenschwiler P, et al. Improving the properties of ceramic foams by a vacuum infiltration process. J Eur Ceram Soc 2010;30: 3005–11.
[28]Taslicukur Z, Balaban C, Kuskonmaz N, et al. Production of ceramic foam filters for molten metal filtration using expanded polystyrene. J Eur Ceram Soc 2007;27: 637–40.
[29]Bock-Seefeld B, Wetzig T, Hub´alkov´a J, et al. Fabrication of carbon-bonded alumina filters by additive-manufactured, water-soluble polyvinyl alcohol filter templates and alginate-based slips. Adv Eng Mater 2022;24:2100655.
[30]Butscher A, Bohner M, Hofmann S, et al. Structural and material approaches to bone tissue engineering in powder-based three-dimensional printing. Acta Biomater 2011;7:907–20.
[31]Maurath J, Willenbacher N. 3D printing of open-porous cellular ceramics with high specific strength. J Eur Ceram Soc 2017;37:4833–42.
[32]Wilson CE, de Bruijn JD, Van Blitterswijk CA, et al. Design and fabrication of standardized hydroxyapatite scaffolds with a defined macro-architecture by rapid prototyping for bone-tissue-engineering research. J Biomed Mater Res 2004;68: 123–32.
[33]Pelanconi M, Rezaei E, Ortona A. Cellular ceramic architectures produced by hybrid additive manufacturing: a review on the evolution of their design. J Ceram Soc Jpn 2020;128:595–604.
[34]Wetzig T, Neumann M, Schwarz M, et al. Rapid prototyping of carbon-bonded alumina filters with flame-sprayed alumina coating for bottom-teeming steel ingot casting. Adv Eng Mater 2022;24:2100777.
[35]Wetzig T, Schmidt A, Dudczig S, et al. Carbon-bonded alumina spaghetti filters by alginate-based robo gel casting. Adv Eng Mater 2020;22:1900657.
[36]Zhu X, Jiang D, Tan S. The control of slurry rheology in the processing of reticulated porous ceramics. Mater Res Bull 2002;37:541–53.
[37]Tulliani JM, Montanaro L, Borello C. Open-pore mullite foams for diesel particulate filtration. Fourth ECerS. Proceeding. 4 th European Ceramic Society Conference 1995;4:493–500.
[38]G´omez SY, Alvarez OA, Escobar JA, et al. Relationship between rheological behavior and final structure of Al2O3 and YSZ foams produced by replica. Adv Mater Sci Eng 2012;1:549. 508.
[39]Voigt C, Aneziris CG, Hub´alkov´a J. Rheological characterization of slurries for the preparation of alumina foams via replica technique. J Am Ceram Soc 2015;98: 1460–3.
[40]Jiang D, Zhu X. Preparation and properties of silicon carbide reticulated porous ceramics. Key Eng Mater 2003;247:19–26.
[41]Schramm A, Voigt C, Hub´alkov´a J, et al. Influence of the manufacturing technique on the macro-and microstructure of reticulated carbon-bonded alumina foams. Adv Eng Mater 2020;22:1900525.
[42]Bergin A, Voigt C, Fritzsch R, et al. Experimental study on the chemical stability of phosphate-bonded Al2O3-based ceramic foam filters. Metall Mater Trans B 2021;52:2008–25.
[43]R¨oder C, Weißbach T, Himcinschi C, et al. Raman spectroscopic characterization of novel carbon-bonded filter compositions for steel melt filtration. J Raman Spectrosc 2014;45:128–32.
[44]Biermann C. Entwicklung eines neuen umweltfreundlichen Bindemittel-Systems für die Feuerfestindustrie auf pharmazeutischer und lebensmittelchemischer Basis. Technische Universit¨at Bergakademie Freiberg; 2016. Ph.D. thesis.

[45]Himcinschi C, Biermann C, Storti E, et al. Innovative carbon-bonded filters based on a new environmental-friendly binder system for steel melt filtration. J Eur Ceram Soc 2018;38:5580–9.
[46]Wu X, Wetzig T, Aneziris CG, et al. Compression behavior of carbon-bonded alumina spaghetti filters at room and high temperatures. Adv Eng Mater 2022;24: 2100613.
[47]Brehm S, Kraus J, Himcinschi C, et al. A Raman spectroscopic study of the pyrolysis of lactose and tannins. J Raman Spectrosc 2022;53:1361–70.
[48]Voigt C, J¨ackel E, Aneziris CG, et al. Investigations of reticulated porous alumina foam ceramics based on different coating techniques with the aid of μCT and statistical characteristics. Ceram Int 2013;39:2415–22.
[49]Luchini B, Hub´alkov´a J, Wetzig T, et al. Carbon-bonded alumina foam filters produced by centrifugation: a route towards improved homogeneity. Ceram Int 2018;44:13832–40.
[50]Pu X, Liu X, Qiu F, et al. Novel method to optimize the structure of reticulated porous ceramics. J Am Ceram Soc 2004;87:1392–4.
[51]Voigt C, J¨ackel E, Aneziris CG, et al. Spinel coating on alumina foam ceramics for aluminum filtration. Adv Eng Mater 2013;15:1197–205.
[52]Zhang H, Suszynski WJ, Agrawal KV, et al. Coating of open cell foams. Ind Eng Chem Res 2012;51:9250–9.
[53]Luchini B, Storti E, Wetzig T, et al. Mechanical and physical characterization of Al2O3-C foam filters produced by distinct processing routes: the importance of the ceramic strut morphology. J Eur Ceram Soc 2019;39:2760–9.
[54]Gehre P, Schmidt A, Dudczig S, et al. Interaction of slip-and flame-spray coated carbon-bonded alumina filters with steel melts. J Am Ceram Soc 2018;101: 3222–33.
[55]Herdering A, Abendroth M, Gehre P, et al. Additive manufactured polyamide foams with periodic grid as templates for the production of functional coated carbon-bonded alumina foam filters. Ceram Int 2019;45:153–9.
[56]Moritz K, Dietze C, Voigt C, et al. Porous alumina coatings on carbon-bonded foam filters by electrophoretic deposition. Ceram Int 2019;45:10701–6.
[57]Voigt C, Endler D, Dietze C, et al. Influence of processing parameters on the surface quality of electrophoretically deposited alumina coatings on foam ceramics. J Eur Ceram Soc 2022;42:5874–84.
[58]Scheffler M, Colombo P. Cellular ceramics: structure, manufacturing, properties and applications. Wiley-VCH Verlag; 2005.
[59]Voigt C, Hub´alkov´a J, Bergin A, et al. Overview of the possibilities and limitations of the characterization of ceramic foam filters for metal melt filtration. Light Metals 2021:785–93.
[60]Voigt C, Hub´alkov´a J, Ditscherlein L, et al. Characterization of reticulated ceramic foams with mercury intrusion porosimetry and mercury probe atomic force microscopy. Ceram Int 2018;44:22963–75.
[61]Firouzeh ST, Abendroth M, Kiefer B. Application of miniaturized Brazilian disc tests for the determination of high-temperature strength of ceramic filter materials. Procedia Struct Integr 2022;42:1069–73.
[62]Zielke H, Abendroth M, Kuna M, et al. Determining the fracture toughness of ceramic filter materials using the miniaturized chevron-notched beam method at high temperature. Ceram Int 2018;44:13986–93.
[63]Solarek J, Bachmann C, Klemm Y, et al. High-temperature compression deformation behavior of fine-grained carbon-bonded alumina. J Am Ceram Soc 2016;99:1390–7.
[64]Soltysiak S, Abendroth M, Kuna M, et al. Strength of fine grained carbon-bonded alumina (Al2O3-C) materials obtained by means of the small punch test. Ceram Int 2014;40:9555–61.
[65]
Brezny R, Green DJ, Dam CQ. Evaluation of strut strength in open-cell ceramics. J Am Ceram Soc 1989;72:885–9.
[66]Maiti SK, Ashby MF, Gibson LJ. Fracture toughness of brittle cellular solids. Scripta Metall 1984;18:213–7.
[67]Brezny R, Green DJ. Factors controlling the fracture resistance of brittle cellular materials. J Am Ceram Soc 1991;74(5):1061–5.
[68]Hagiwara H, Green DJ. Elastic behavior of open-cell alumina. J Am Ceram Soc 1987;70:811–5.
[69]Brezny R, Green DJ. The effect of cell size on the mechanical behavior of cellular materials. Acta Metall Mater 1990;38:2517–26.
[70]Dam CQ, Brezny R, Green DJ. Compressive behavior and deformation-mode map of an open cell aluminaa. J Mater Res 1990;5:163–71.
[71]Grabenhorst J, Luchini B, Fruhstorfer J, et al. Influence of the measurement method and sample dimensions on the Young’s modulus of open porous alumina foam structures. Ceram Int 2019;45:5987–95.
[72]Brezny R, Green DJ. Uniaxial strength behavior of brittle cellular materials. J Am Ceram Soc 1993;76:2185–92.
[73]Meille S, Lombardi M, Chevalier J, et al. Mechanical properties of porous ceramics in compression: on the transition between elastic, brittle, and cellular behavior. J Eur Ceram Soc 2012;32:3959–67.
[74]Voigt C, Storm J, Abendroth M, et al. The influence of the measurement parameters on the crushing strength of reticulated ceramic foams. J Mater Res 2013;28:2288–99.
[75]Winter W. Strukturph¨anomene bei der Biegung mit Stoffmodellen der kontinuierlichen Sch¨adigung. Z Angew Math Mech 1994;74:T192–5.
[76]Hub´alkov´a J, Voigt C, Schmidt A, et al. Comparative phenomenological study of fracture behavior of ceramic and glass foams under compressive stress using in situ X-ray microtomography. Adv Eng Mater 2017;19:1700286.
[77]Neumann M, Hub´alkov´a J, Voigt C, et al. On the fracture statistics of open-porous alumina foam structures. J Eur Ceram Soc 2022;42:2331–40.

[78]Dai Y, Hub´alkov´a J, Voigt C, et al. Tensile strength and damage of open-cell ceramic foams under cylindrical splitting test. Journal of Advanced Ceramics 2023;12(11):2003–16.
[79]Goretta KC, Brezny R, Dam CQ, et al. High temperature mechanical behavior of porous open-cell Al2O3. Mater Sci Eng, A 1990;124:151–8.
[80]Wu X, Ranglack-Klemm Y, Hub´alkov´a J, et al. Impact of high temperature on the compression behavior of carbon-bonded alumina filters with functionalized coatings. Ceram Int 2021;47:3920–7.
[81]Ranglack-Klemm Y, Storti E, Aneziris CG, et al. Influence of carbon nanotubes- based coatings on the high-temperature compression strength of Al2O3-C foam filter structures. Adv Eng Mater 2020;22:1900423.
[82]Bergin A, Voigt C, Fritzsch R, et al. Investigation of mechanical and thermo- mechanical strength of ceramic foam filters. Ceram Int 2023;49:8022–31.
[83]Gibson LJ, Ashby MF. Cellular solids, structure and properties. Pergamon Press; 1986. p. 120–68.
[84]Schneider GA. Thermal shock criteria for ceramics. Ceram Int 1991;17:325–33.
[85]Aneziris CG, Ansorge A, Jaunich H. New approaches of carbon bonded foam filters for filtration of large castings. Ceramic Forum International Special Refractories for Clean Steel Technology 2008;85:E100–3.
[86]BDG-Richtlinie VDG. Merkblatt P100. Keramische Filter in Schaumstruktur- Schaumkeramikfilter für Eisen-und Stahlguss 2012.
[87]Emmel M, Aneziris CG, Schmidt G, et al. Influence of the chemistry of surface functionalized ceramic foam filters on the filtration of alumina inclusions in steel melts. Adv Eng Mater 2013;15:1188–96.
[88]Orenstein RM, Green DJ. Thermal shock behavior of open-cell ceramic foams. J Am Ceram Soc 1992;75:1899–905.
[89]Vedula VR, Green DJ, Hellman JR. Thermal shock resistance of ceramic foams. J Am Ceram Soc 1999;82:649–56.
[90]Gumban CG, Hub´alkov´a J, Neumann M, et al. Thermal shock resistance of industrial foam filters-a comparative study. Open Ceramics 2023;16:100463.
[91]Gibson LJ, Ashby MF, Harley BA. Cellular materials in nature and medicine. Cambridge University Press; 2010.
[92]Yoshihisa K, Masashi F. Ceramic foam. British Patent 1986;2:337.
[93]Chollon G, Delettrez S, Langlais F. Chemical vapour infiltration and mechanical properties of carbon open-cell foams. Carbon 2014;66:18–30.
[94]Emmel M, Aneziris CG. Functionalization of carbon-bonded alumina filters through the application of active oxide coatings for steel melt filtration. J Mater Res 2013;28:2234–42.
[95]Storti E, Emmel M, Dudczig S, et al. Development of multi-walled carbon nanotubes-based coatings on carbon-bonded alumina filters for steel melt filtration. J Eur Ceram Soc 2015;35:1569–80.
[96]Adler J, Teichgr¨aber M, Standke G. DE19621638A1. 1997.
[97]Celades CR, Diez V. Process for the preparation of a reinforced ceramic foam. European Patent 0369098 1990.
[98]Crooke PR. Ceramic foam. British Patent 1982;2:777.
[99]Salvini VR, Luchini B, Aneziris CG, et al. Innovation in ceramic foam filters manufacturing process. Int J Appl Ceram Technol 2019;16:378–88.
[100]Rannabauer S, S¨offker GM, Scheunemann M, et al. Increased mechanical stability and thermal conductivity of alumina reticulated porous ceramics (RPC) by nanoparticle infiltration processing. Adv Eng Mater 2017;19:1700211.
[101]Liang X, Li Y, Sang S, et al. Enhanced mechanical properties of SiC reticulated porous ceramics via adjustment of residual stress within the strut. Int J Appl Ceram Technol 2018;15:28–35.
[102]Liang X, Li Y, Yan W, et al. Preparation of SiC reticulated porous ceramics with high strength and increased efficient filtration via fly ash addition. J Eur Ceram Soc 2021;41:2290–6.
[103]Peng W, Chen Z, Yan W, et al. Microstructure and properties of ceramic filter containing porous MgO coating and its filtration of Al2O3 inclusions in molten steel. Ceram Int 2024;50:218–29.
[104]Aneziris CG, Emmel M, Stolle A. Multifunctional carbon bonded filters for metal melt filtration. Advances in Bioceramics and Porous Ceramics V 2012:123–38.
[105]Callister WD, Rethwisch DG. Materials science and engineering: an introduction. Wiley; 2013.
[106]Grong Ø, Kolbeinsen L, van Der Eijk C, et al. Microstructure control of steels through dispersoid metallurgy using novel grain refining alloys. ISIJ Int 2006;46: 824–31.
[107]Zhang L, Thomas BG. Inclusions in continuous casting of steel. XXIV National Steelmaking Symposium, Morelia, Mich, Mexico 2003;26:28.
[108]Ovtchinnikov S. Impact of inclusions on toughness. PhD thesis Freiberg 2002.
[109]Kaushik P, Pielet H, Hongbin YIN. Inclusion characterization: a tool for measurement of steel cleanliness and process control. Iron Steel Technol 2009;6: 82–99.
[110]Yamada K, Watanabe T, Fukuda K, et al. Emoval of non-metallic inclusion by ceramic filter. Transactions of the Iron and Steel Institute of Japan 1987;27: 873–7.
[111]Burty M, Peeters L, Perrin E, et al. 5th European continuous casting conf. Nice, France: Proc.; 2005. p. 462–7.
[112]Yang W, Zhang Y, Zhang L, et al. Population evolution of oxide inclusions in Ti- stabilized ultra-low carbon steels after deoxidation. J Iron Steel Res Int 2015;22: 1069–77.
[113]Thunman M, Sichen D. Origins of non-metallic inclusions and their chemical development during ladle treatment. Steel Res Int 2008;79:124–32.
[114]Park JH, Todoroki H. Control of MgO⋅Al2O3 spinel inclusions in stainless steels. ISIJ Int 2010;50:1333–46