Casting system

Abstract

There is provided a tundish with a basin and a baffle module. The baffle module is positioned inside the basin such it is the inlet of the tundish which receives a flow of molten alloy. The baffle module includes walls having a first height that is greater than the level of the molten alloy in the tundish, an opening in one of the walls to allow the molten alloy to exit an interior space defined by the walls, and at least one slag baffle extending from one of the walls to a wall of the basin, the at least one slag baffle having a second height that is smaller than the first height and smaller than the level of the molten alloy in the tundish. The slag baffle is positioned such that the flow of the molten alloy passes through the slag baffle to reach an outlet of the tundish.

Description

CASTING SYSTEMCROSS-REFERENCE TO A RELATED APPLICATION[1] This disclosure claims priority from U.S. provisional application no. 63 / 504770 filed May 29, 2023 which is incorporated herein by reference in its entirety.TECHNICAL FIELD[2] This disclosure relates to casting systems for casting molten alloys, particularly to the tundish of the casting system.BACKGROUND OF THE ART[3] When casting light metal alloys having a density lower than common casting contaminants, the contaminants make their way into the cast product and are detrimental to the mechanical properties of the cast product. The presence of contaminants in a cast product can lead to breaks or fractures. Light metal alloys include aluminum and magnesium. In aluminum alloy casting for example, there is an increasing amount of vanadium impurities present in the molten aluminum produced in the Hall-Heroult reduction cells which is not compatible with the purity standards required by the industry for example to produce high performance electrical conductors. To mitigate that problem, a precise amount of boron is added in the molten metal during alloy preparation in the holding furnace. Boron reacts with vanadium, but also with titanium to form intermetallic precipitates such as VB2, TiB2 and TiVB2. Electrical wires containing such intermetallic precipitates show breaks due to the presence of these intermetallic precipitate inclusions (agglomerates formed during manufacturing and casting of the alloy). Breaks are most common in wire drawing which is a method to reduce sections of wires using dies of smaller and smaller diameter until the required dimension is obtained. One way to reduce the number of intermetallic precipitate inclusions is to use filters; however, this alone is not feasible as filters become clogged and plugged at some point during the continuous casting process. A continuous casting process can last up to several days (without stoppage). Reducing the cast duration in order to try to avoid clogging the filters is not a viable option because that would compromise the productivity of the operation. Accordingly, improvements in the casting process to reduce or eliminate the inclusions of intermetallic precipitates are desired.SUMMARY[4] In one aspect, there is provided a tundish comprising a basin and a baffle module, wherein the baffle module is positioned inside the basin such that the baffle module is the inlet ofthe tundish receiving a flow of molten alloy, and wherein the baffle module comprises walls having a first height that is greater than the level of the molten alloy in the tundish, the walls defining an interior space in which the molten alloy is received, an opening in one of the walls to allow the molten alloy to exit the interior space defined by the walls, and at least one slag baffle extending from one of the walls of the baffle module to a wall of the basin, the at least one slag baffle having a second height that is smaller than the first height and smaller than the level of the molten alloy in the tundish, wherein the slag baffle is positioned such that the flow of the molten alloy passes through the slag baffle to reach an outlet of the tundish. In some embodiments, a bottom wall of the baffle module is fixed to the basin. In some embodiments, the bottom wall of the baffle module is glued or mechanically secured to the basin. In some embodiments, the basin comprises steel, refractory reinforced fibreglass material or refractory concrete. In some embodiments, the baffle module comprises reinforced fibreglass material. In some embodiments, the opening is positioned opposite of the slag baffle. In some embodiments, the opening has a height equal to the first height. In some embodiments, the baffle module is characterized by having continuous walls. In some embodiments, the second height is 60 to 90 % of the level of the molten alloy in the tundish. In some embodiments, the second height is 70 to 85 % of the level of the molten alloy in the tundish. In some embodiments, the second height is 75 to 85 % of the level of the molten alloy in the tundish. In some embodiments, the baffle module positioned in the center of the basin such that a cavity is defined between the walls of the baffle module and interior walls of the basin. In some embodiments, the tundish further comprises a skim dam. In some embodiments, the skim dam is positioned upstream of the slag baffle. In some embodiments, the baffle module has two slag baffles.[5] In one aspect, there is provided a system for casting low density metal alloys comprising the tundish of the present disclosure. In some embodiments, the system further comprises a ladle for containing a volume of molten alloy. In some embodiments, the system further comprises a ladle shroud located downstream of the ladle relative to a flow of the molten alloy and in fluid communication with the ladle. In some embodiments, the tundish is located downstream of the ladle shroud relative to the flow of the molten alloy and in fluid communication with the ladle shroud. In some embodiments, the system further comprises a casting shroud downstream of the outlet of the tundish relative to the flow of the molten alloy and in fluid communication with the outlet of the tundish. In some embodiments, the system further comprises at least one mold in fluid communication with the casting shroud adapted to receive and mold the molten alloy into a desired shape. The casting system can optionally comprise a liquid metal filter. The liquid metal filter can be placed in the ladle shroud, between the ladle shroud and the tundishor in the tundish. The optional liquid metal filter is for example a ceramic filter (e.g. a foam or honeycomb) or a tortuous path that entrains particles.[6] In one aspect, there is provided a system for casting low density metal alloys, the system comprising: a ladle for containing a volume of molten alloy; a ladle shroud located downstream of the ladle relative to a flow of the molten alloy and in fluid communication with the ladle; a tundish as defined in the present disclosure located downstream of the ladle shroud relative to the flow of the molten alloy and in fluid communication with the ladle shroud; a casting shroud downstream of the outlet of the tundish relative to the flow of the molten alloy and in fluid communication with the outlet of the tundish; and at least one mold in fluid communication with the casting shroud adapted to receive and mold the molten alloy into a desired shape.[7] In a further aspect, there is provided a method of casting a low density metal alloy, the method comprising: flowing a molten alloy from a ladle to a tundish, wherein the tundish comprises a basin and a baffle module, wherein the baffle module is positioned inside the basin such that the baffle module is the inlet of the tundish receiving the molten alloy, and wherein the baffle module comprises: walls having a first height that is greater than the level of the molten alloy in the tundish, the walls defining an interior space in which the molten alloy is received, an opening in one of the walls to allow the molten alloy to exit the interior space defined by the walls, and at least one slag baffle extending from one of the walls of the baffle module to a wall of the basin, the at least one slag baffle having a second height that is smaller than the first height and smaller than the level of the molten alloy in the tundish; flowing the molten alloy from the tundish through the slag baffle to a mold to impart a desired shape and obtain a cast alloy. In some embodiments, the method is a continuous casting method for a casting period of at least 5 days. In some embodiments, the method further comprises replacing the baffle module and restarting a casting period of at least 5 days.[8] Many further features and combinations thereof concerning the present improvements will appear to those skilled in the art following a reading of the instant disclosure.DESCRIPTION OF THE DRAWINGS[9] FIG. 1 is a schematic of a casting system according to one embodiment of the present disclosure.

[0010] FIG. 2 is a schematic of a tundish according to one embodiment of the present disclosure.

[0011] FIG. 3A is a schematic of a baffle module according to one embodiment of the present disclosure.

[0012] FIG. 3B is a schematic of a baffle module according to a further embodiment of the present disclosure.

[0013] FIG. 4 is a photograph showing a sample of aluminum alloy deposit from an casting furnace delineating three samples to be cut: samples A, B, and C.

[0014] FIG. 5 is a photograph showing the cut samples A, B, and C from Fig. 4.

[0015] FIG. 6 is a photograph showing sample A of Fig. 4.

[0016] FIG. 7 is a photograph showing sample B of Fig. 4.

[0017] FIG. 8 is a photograph showing sample C of Fig. 4.

[0018] FIG. 9 is a microscopy image of sample A.

[0019] FIG. 10 is a microscopy image showing TiVB2 and AI4C3.

[0020] FIG. 11 is a microscopy image of the zone shown by the arrow in Sample A of Fig. 9.

[0021] FIG. 12A is a scanning electron microscopy (SEM) image of sample A.

[0022] FIG. 12B is a cartography of Fig. 12A showing the Al.

[0023] FIG. 12C is a cartography of Fig. 12A showing the Ti.

[0024] FIG. 12D is a cartography of Fig. 12A showing the Sr.

[0025] FIG. 12E is a cartography of Fig. 12A showing the V.

[0026] FIG. 12F is a cartography of Fig. 12A showing the Fe.

[0027] FIG. 13 is a microscopy image of Sample B.

[0028] FIG. 14 is a microscopy image of Sample C.

[0029] FIG. 15 is an image of a traditional tundish that was cut in half and divided into zones longitudinal zones A, B, C, D and E and transversal zones 1T, 2T and 3T.

[0030] FIG. 16 is an image of zone A of Fig. 15 which is subdivided into zones I, II, III and IV.

[0031] FIG. 17A is a microscopy image of zone D of Fig. 15.

[0032] FIG. 17B is a microscopy image of zone C of Fig. 15.

[0033] FIG. 17C is a microscopy image of zone B of Fig. 15.

[0034] FIG. 17D is a microscopy image of zone 3T of Fig. 15.

[0035] FIG. 17E is a microscopy image of zone 2T of Fig. 15.

[0036] FIG. 17F is a microscopy image of zone 1 T of Fig. 15.

[0037] FIG. 17G is a microscopy image of zone I inside zone A of Fig. 15.

[0038] FIG. 17H is a microscopy image of zone III inside zone A of Fig. 15.

[0039] FIG. 18A is a microscopy image of an agglomerate from in zone 3T of Fig. 17D.

[0040] FIG. 18B is a microscopy image of an agglomerate from in zone 2T of Fig. 17E.

[0041] FIG. 18C is a microscopy image which shows close up on carbides in Fig. 18B.

[0042] FIG. 18D is a microscopy image which shows a close up on the grain joints in Fig.18B.

[0043] FIG. 19A is a SEM image of a contaminant agglomerate (scale 1 mm).

[0044] FIG. 19B is a cartography of Fig. 19A showing the Al.

[0045] FIG. 19C is a cartography of Fig. 19A showing the Ti.

[0046] FIG. 19D is a cartography of Fig. 19A showing the V.

[0047] FIG. 19E is a cartography of Fig. 19A showing the Fe.

[0048] FIG. 20 is a SEM image of a contaminant agglomerate (scale 10 pm).

[0049] FIG. 21 is a mass spectroscopy graph of the analysis of the agglomerate of Fig. 20.

[0050] FIG. 22A is a microscopy image of a first agglomerate labeled agglomerate 1 .

[0051] FIG. 22B is a mosaic of Fig. 22A.

[0052] FIG. 22C is an image analysis of Fig. 22A.

[0053] FIG. 22D is a microscopy image of a first agglomerate labeled agglomerate 2.

[0054] FIG. 22E is a mosaic of Fig. 22D.

[0055] FIG. 22F is an image analysis of Fig. 22E.

[0056] FIG. 22G is a microscopy image of a first agglomerate labeled agglomerate 3.

[0057] FIG. 22H is a mosaic of Fig. 22G.

[0058] FIG. 22I is an image analysis of Fig. 22H.

[0059] FIG. 23A is a microscopy image of a contaminant agglomerate having 50 % porosity.

[0060] FIG. 23B is a microscopy image of a contaminant agglomerate having 37 % porosity.

[0061] Fig. 24 is a graph showing the percentage of retaining particles in function of the particle diameter.

[0062] FIG. 25A is photograph of a baffle module according to one embodiment of the present disclosure.

[0063] FIG. 25B is a schematic top view of a baffle module according to one embodiment of the present disclosure.

[0064] FIG. 25C is a schematic side view of a baffle module according to one embodiment of the present disclosure.

[0065] FIG. 25D is a photograph of a basin of the tundish after one casting round.

[0066] FIG. 25E is a photograph of the tundish according to one embodiment of the present disclosure after continuous casting for 6 days.

[0067] FIG. 25F is a photograph of a tundish cut in half and the half divided in five section 1 to 5).

[0068] FIG. 25G is a photograph of section 1 of Fig. 25F divided into five zones A-E.

[0069] FIG. 25H is a photograph of zone A of Fig. 25G.

[0070] FIG. 251 is a photograph of zone B of Fig. 25G.

[0071] FIG. 25J is a photograph of zone C of Fig. 25G.

[0072] FIG. 25K is a photograph of zone D of Fig. 25G.

[0073] FIG. 25L is a photograph of zone E of Fig. 25G.

[0074] FIG. 25M is a photograph close up of Fig. 25K.

[0075] FIG. 26A is a microscopy image of the contaminant deposit of Fig. 25F (scale bar 500 pm).

[0076] FIG. 26B is a microscopy image of Fig. 26A (scale bar 250 pm).

[0077] FIG. 26C is a microscopy image of Fig. 26A (scale bar 200 pm).

[0078] FIG. 26D is a microscopy image of zone D of Fig. 25K (scale bar 2000 pm).

[0079] FIG. 26E is a microscopy image of zone D of Fig. 25K (scale bar 500 pm).

[0080] FIG. 26F is a microscopy image of zone D of Fig. 25K (scale bar 250 pm).

[0081] FIG. 26G is a microscopy image of zone D of Fig. 25K (scale bar 200 pm).

[0082] FIG. 27A is a SEM image of an agglomerate with two zones marked with a square and a triangle.

[0083] FIG. 27B is a mass spectroscopy spectra of the square zone of Fig. 27A.

[0084] FIG. 27C is a mass spectroscopy spectra of the triangle zone of Fig. 27A.

[0085] FIG. 28A is a SEM image of a contaminant agglomerate (scale 250 pm).

[0086] FIG. 28B is a cartography of Fig. 28A showing the Al.

[0087] FIG. 28C is a cartography of Fig. 28A showing the Ti.

[0088] FIG. 28D is a cartography of Fig. 28A showing the V.

[0089] FIG. 28E is a cartography of Fig. 28A showing the B.

[0090] FIG. 29A is a photograph of tundish cut in half and the half divided in three pieces #1 , #2 and #3.

[0091] FIG. 29B is a photograph of the piece #1 of Fig. 29A from which a section 1-A was cut out from the deposit.

[0092] FIG. 29C is a photograph of the piece #2 of Fig. 29A from which two sections 2-A and2-B were cut out from the deposit.

[0093] FIG. 29D is a photograph of the piece #3 of Fig. 29A from which three sections 3-A,3-B and 3-C were cut out from the deposit.

[0094] FIG. 30A is a photograph of a cross section of section 1-A of Fig. 29B.

[0095] FIG. 30B is a photograph of a cross section of section 2-A of Fig. 29C.

[0096] FIG. 30C is a photograph of a cross section of section 2-B of Fig. 29C.

[0097] FIG. 30D is a photograph of a cross section of section 3-A of Fig. 29D.

[0098] FIG. 30E is a photograph of a cross section of section 3-B of Fig. 29C.

[0099] FIG. 30F is a photograph of a cross section of section 3-C of Fig. 29C.

[0100] FIG. 31 A is a microscopy image of section 1-A (scale bar 2000 pm).

[0101] FIG. 31 B is a microscopy image of section 1-A (scale bar 500 pm).

[0102] FIG. 31 C is a microscopy image of section 1-A (scale bar 250 pm).

[0103] FIG. 31 D is a microscopy image of section 1-A (scale bar 200 pm).

[0104] FIG. 31 E is a microscopy image of section 1-A (scale bar 50 pm).

[0105] FIG. 32A is a microscopy image of section 2-A (scale bar 2000 pm).

[0106] FIG. 32B is a microscopy image of section 2-A (scale bar 500 pm).

[0107] FIG. 32C is a microscopy image of section 2-A (scale bar 250 pm).

[0108] FIG. 32D is a microscopy image of section 2-A (scale bar 200 pm).

[0109] FIG. 32E is a microscopy image of section 2-A (scale bar 50 pm).

[0110] FIG. 33A is a microscopy image of section 2-B (scale bar 2000 pm)

[0111] FIG. 33B is a microscopy image of section 2-B (scale bar 500 pm),

[0112] FIG. 33C is a microscopy image of section 2-B (scale bar 250 pm),

[0113] FIG. 33D is a microscopy image of section 2-B (scale bar 200 pm),

[0114] FIG. 33E is a microscopy image of section 2-B (scale bar 50 pm).

[0115] FIG. 34A is a microscopy image of section 3-A (scale bar 2000 pm)

[0116] FIG. 34B is a microscopy image of section 3-A (scale bar 500 pm),

[0117] FIG. 34C is a microscopy image of section 3-A (scale bar 250 pm).

[0118] FIG. 34D is a microscopy image of section 3-A (scale bar 200 pm),

[0119] FIG. 34E is a microscopy image of section 3-A (scale bar 50 pm).

[0120] FIG. 35A is a microscopy image of section 3-B (scale bar 2000 pm)

[0121] FIG. 35B is a microscopy image of section 3-B (scale bar 500 pm),

[0122] FIG. 35C is a microscopy image of section 3-B (scale bar 250 pm),

[0123] FIG. 35D is a microscopy image of section 3-B (scale bar 200 pm).

[0124] FIG. 35E is a microscopy image of section 3-B (scale bar 50 pm).

[0125] FIG. 36A is a microscopy image of section 3-C (scale bar 2000 pm)

[0126] FIG. 36B is a microscopy image of section 3-C (scale bar 500 pm),

[0127] FIG. 36C is a microscopy image of section 3-C (scale bar 250 pm),

[0128] FIG. 36D is a microscopy image of section 3-C (scale bar 200 pm).

[0129] FIG. 36E is a microscopy image of section 3-C (scale bar 50 pm).

[0130] FIG. 37A is a microscopy image showing a front cross section of a wire break of sample A of Example 5.

[0131] FIG. 37B is a microscopy image showing a side cross section of a wire break of sample A of Example 5.

[0132] FIG. 37C is a close up of Fig. 37B.

[0133] FIG. 38A is a microscopy image showing a front cross section of a wire break of sample B of Example 5.

[0134] FIG. 38B is a microscopy image showing a side cross section of a wire break of sample B of Example 5.

[0135] FIG. 39A is a microscopy image showing a front cross section of a wire break of sample C of Example 5.

[0136] FIG. 39B is a microscopy image showing a side cross section of a wire break of sample C of Example 5.

[0137] FIG. 39C is a close up of Fig. 39B.

[0138] FIG. 39D is an energy dispersive x-ray (EDS) spectra showing the composition of Sample C of Example 5.

[0139] FIG. 40A is a microscopy image showing a front cross section of a wire break of sample D of Example 5.

[0140] FIG. 40B is a microscopy image showing a side cross section of a wire break of sample D of Example 5.

[0141] FIG. 40C is a close up of Fig. 40B.

[0142] FIG. 41 A is a microscopy image showing a front cross section of a wire break of sample E of Example 5.

[0143] FIG. 41 B is a microscopy image showing a side cross section of a wire break of sample E of Example 5.

[0144] FIG. 41 C is a close up of Fig. 41 B.

[0145] FIG. 42A is a microscopy image showing a front cross section of a wire break of sample F of Example 5.

[0146] FIG. 42B is a microscopy image showing a side cross section of a wire break of sample F of Example 5.

[0147] FIG. 42C is a close up of Fig. 42B.DETAILED DESCRIPTION

[0148] It is provided a casting system and a method of casting a light metal alloy using the disclosed casting system. Casting systems generally contain a casting furnace, a ladle, a ladle shroud, a tundish, a casting shroud and a mould. The casting furnace is used to melt metal at high temperature to obtain a molten metal. The molten metal is then received into a receptacle for holding the molten metal called a ladle. After undergoing any applicable ladle treatments, such as alloying, degassing, and arriving at the correct temperature, the molten metal flows to a tundish through the ladle shroud connecting the ladle and the tundish. The tundish acts as a reservoir of molten metal to feed the casting machine which has at least one mould. The tundish acts as a buffer of hot metal and can smooth out flow of metal thereby regulating metal feed to the moulds.

[0149] A light metal alloy as used herein can be defined as a metal alloy having a low density, for example a density of less than 3.5 g / cm3. In some embodiments, the light metal alloy is an aluminum alloy or a magnesium alloy. Low density metal alloys such as aluminum and magnesium are susceptible to the accumulation of intermetallic precipitates having a higher density than them. These intermetallic precipitates become contaminants that make their way into the cast product and reduce the purity of the cast product. They can even hamperthe mechanical properties of the cast product. Although the present disclosure makes reference henceforth to aluminum alloys, the methods and systems described herein also apply to other light metal alloys such as magnesium.

[0150] The amount of intermetallic precipitates formed during the casting of molten aluminum, for example, is directly proportional to the amount of boron, vanadium and titanium present in the molten aluminum. These intermetallic compounds have a density which is much higher than that of the molten aluminum. Therefore, a large proportion will settle at the bottom of the furnace. Unfortunately, a significant number of particles smaller than 10 microns are too small to settle and will therefore remain in suspension even if there is no mixing in the furnace for more than an hour.

[0151] Once the furnace is tilted to begin casting, small TiVB2 particles are entrained in the launder and tundish. After several hours of casting, these small particles begin to agglomerate at various locations. Once formed, the clusters of TiVB2 inclusions continue to grow by agglomerating further small particles and can reach sizes of over 1 mm in diameter. The large clusters can detach from its substrate, and despite their high density and size, they can freely reach the casting wheel due to the fast and turbulent metal flow in the tundish. The large TiVB2 clusters can therefore be found in the cast products such as rods or wires.

[0152] It was determined (see Examples below) that the agglomeration of the inclusions takes place preferentially in the tundish. Accordingly, improving the retention of the agglomerates in the tundish would reduce or eliminate the leaking of contaminants into the cast products. If the agglomeration of contaminants is addressed upstream of the tundish then only a small portion of the contaminant accumulations will be reduced and there will be no significant difference in the quantity of contaminants present in the cast product. This is one of the reasons the present disclosure is focused on the tundish to successfully reduce the contaminants in the cast product at the stage where the most significant accumulation occurs. The terms “contaminants” or “inclusions” as used herein, in some embodiments, mean the accumulation of particles such as TiB2, SrO, VB2, TiVB2, AI2O3, MgAI2O4, NaCI, CaCI2, MgCI2, and / or KCI.

[0153] The TiVB2 clusters are the weakest link in cast aluminum products, specifically rods or wires, and therefore the inclusion of TiVB2 in the cast product should be minimized and preferably eliminated. The risk of failure during wire drawing or conform extrusion is greatly increased when these clusters are present. The failure probability is directly proportional to the inclusion size an inversely proportional to the wire size. When using traditional casting systems it was found that the average size of the clusters found in the wire breaks can be well above 500 microns. Poor rod quality can cause a large number of wire breaks, resulting in numerous customer complaints.

[0154] To solve the aforementioned issue, an improved tundish was developed using one or preferably multiple baffles to capture larger clusters of inclusions before they can reach the casting wheel and thus the cast product. The baffle retention system relies on gravity to capture the contaminant particles (that have a higher density than the molten alloy) and is located in the tundish but before the casting spout and the casting wheel, in orderto minimize the risk of clusters forming downstream of the tundish.

[0155] Fig. 1 illustrates a casting system 1 having a ladle 10, a ladle shroud 20, a tundish 30, a casting shroud 40 and a mould 50. The ladle shroud 20 fluidly connects the ladle 10 containingmolten aluminum metal with the tundish 30. Accordingly, the tundish 30 is located downstream of the ladle 10 relative to the flow of the molten aluminum metal within the tundish 30. Generally, the tundish 30 is located below the ladle 10 (i.e., at a lower elevation relative to the ladle) such that gravity acts on the molten aluminum metal and drives the flow of the molten aluminum metal from the ladle 10 to the tundish 30. The casting shroud 40 fluidly connects the tundish to a mould 50. It will be appreciated that in some embodiments multiple moulds may be used in a casting system.

[0156] The tundish 30, illustrated in Fig. 2, has a basin 31 and a baffle module 32. The baffle module 32 is positioned inside the basin 31 such that the baffle module 32 is the inlet of the tundish 30 receiving the molten aluminum alloy from the ladle shroud 20. The baffle module 32 has walls 32a having a first height that is greater than the level of the molten aluminum alloy in the tundish 30. The baffle module 32 is contained within the basin 31 i.e. within the walls of the basin 31 a. In preferred embodiments, there is a space separating the walls of the basin 31 a and the walls of the baffle module 32a as illustrated in Fig. 2. The walls 32a define an interior space 32c in which the molten aluminum alloy is received. In preferred embodiments, the baffle module 32 has a floor 32d to facilitate the replacement of the baffle module 32 while retaining the deposit of contaminants on the floor 32d. The floor 32d can be fixed to the basin floor 31 b for example with a suitable glue or the floor 32d can be fixed mechanically to the basin floor 31 b. In other embodiments, the walls 32a are connected to the floor 31 b of the basin 31 (and the floor 32d of the baffle module is absent). In such embodiments, the replacement of the baffle module 31 independently of the basin 32 becomes more difficult. Indeed, casting is a continuous process (e.g. 5-7 days) after which the baffle module 32 can be replaced without replacing the basin 31. Accordingly, the presence of the floor 32d which makes the baffle module an independent module allows for the advantage of replacing the baffle module without replacing the basin of the tundish. The reusing of the basin 31 allows for a reduction in material and waste as well as a reduction in production cost.

[0157] The baffle module 32 is illustrated in Fig. 3A. The baffle module has walls 32a to entrain molten aluminum alloy in the interior space defined therein. The walls 32a which have the first height which is higher than the level of molten aluminum alloy, force the flow of the molten aluminum alloy in one or more directions inside the tundish. In the embodiment illustrated in Fig. 3A, the baffle module 32 has an opening 32b in one of the walls to allow the molten aluminum alloy to exit the interior space. Although multiple openings 32b can be included in the walls 32a of the baffle module 32, a single opening 32b is preferred to prolong the entrainment of molten aluminum alloy in the interior space and to be able to better direct the flow of the molten aluminum alloy to go through a slag baffle 32e. To prolong the path of the molten aluminum alloy, in preferredembodiments, the opening 32b is positioned opposite of the slag baffle 32e or opposite of the outlet 31 c of the tundish 30 (as illustrated in Figs. 2 and 3A). To allow sufficient flow across the opening 32b, the opening can define an empty space that has a height equal to the first height. This is preferably a complete opening as illustrated in Figs. 2 and 3A. The efficiency of retaining contaminants is generally not increased by increasing the height of the opening. However, there is a possibility of having an accumulation of contaminants that generate undesirable turbulence if the opening has a height too high. Accordingly, a complete opening is preferred to avoid the potential creation of turbulence.

[0158] The slag baffle 32e extends from one of the walls 32a of the baffle module to a wall31 a of the basin. The slag baffle 32e has a second height that is smaller than the first height and smaller than the level of the molten aluminum alloy in the tundish. The second height is sufficiently elevated such that the contaminant inclusions that have a higher density than the molten aluminum alloy are retained behind the baffle (with respect to the flow of molten aluminum alloy). The second height must also not be too elevated as it should allow a flow of the molten aluminum alloy through. Accordingly, in some embodiments, the second height is 60 to 90 %, 65 to 90 %, 70 to 90 %, 75 to 90 %, 60 to 85 %, 60 to 80 %, 60 to 70 %, 65 to 85 %, 65 to 80 %, 70 to 90 %, 70 to 85 %, 75 to 90 %, 75 to 85 %, or about 80 % of the level of the molten aluminum alloy in the tundish. Still, in some embodiments, the second height can be defined as having a height that is 35 to 70 %, 35 to 65 %, 35 to 60 %, 35 to 55 %, 40 to 70 %, 40 to 65 %, 40 to 60 %, 40 to 55 %, 45 to 70 %, 45 to 65 %, 45 to 60 %, or 45 to 55 %that of the first height.

[0159] In Figs. 2 and 3A, a single baffle module 32 is illustrated and is positioned in the center of the basin. However, other embodiments are contemplated in the present disclosure including a different positioning of the baffle module 32 or the inclusion of multiple baffle modules 32. In one example, baffle modules 32 can be nested inside one another. In that nested embodiment, the interior baffle modules 32 will have their slag baffles 32e extending to a more eccentric wall 32a of another slag baffle 32e (instead of extending to the wall 31 a of the basin 31). Although two slag baffles 32e are illustrated in Figs. 2 and 3A, there can be a single slag baffle 32e for a given baffle module 32 as long as the entire flow of the molten aluminum alloy passes through that single slag baffle 32e. When multiple slag baffles 32e are present, the flow of the molten aluminum alloy as a whole must have passed by at least one slag baffle 32e (e.g. 50% of the flow through one slag baffle and the other 50% through a second slag baffle). In addition, a single baffle module32 can contain more than two slag baffles 32e. In some embodiments, multiple slag baffles 32e can be positioned in series with respect to the flow of the molten aluminum alloy to increasingly retain more of the high density contaminants. When including multiple slag baffles 32e in seriesthe slag baffles 32e may have a similar height or may have increasing heights. In Fig. 3B, a further embodiment is shown where the baffle module 32 has two successive openings 32b which lead to a longer retention of molten alloy in the baffle module 32. The longer retention time leads to a lower molten metal velocity which improves the settling of particles in the baffle module.

[0160] In preferred embodiments, the baffle module has continuous walls and generally the tundish only contains walls that are continuous (i.e. not porous). Filters and other porous or meshlike structures are not necessary in the tundish because they can clog the flow of the molten aluminum alloy.

[0161] In some embodiments, the casting system further comprises a liquid metal filter. The liquid metal filter is placed upstream of the slag baffle 32e, for example in the ladle shroud 20 or in the tundish 30. The liquid metal filter is for example a ceramic foam filter or a honeycomb ceramic filter, or can be a tortuous path for the liquid metal flow that retains particles. The term “tortuous path” as used herein means a winding path with multiple consecutive turns or bends to increase the distance flowed by the liquid metal over a given length between two points. The liquid metal filter retains smaller contaminant particles (e.g. less than 350 pm). One issue with filters is that the contaminant particles will combine at the filter and form larger particles (e.g. 350 pm - 3 mm) which can be released from the filter. These larger contaminant particles are however dealt with by the slag baffle 32e and do not make their way into the cast alloy product. Accordingly, there is a synergy in the use of a liquid metal filter with the slag baffle module of the present disclosure. Indeed, with the baffle module, the larger the contaminant particles the easier they deposit and are retained by the slag baffles because of their increased size and weight. In some embodiments, the liquid metal filter is a ceramic filter such as a silicon carbide based ceramic. In some embodiments, the filter can be characterized by an average pore per linear inch (PPI) of 1 - 100 PPI, 5 to 80 PPI or 5 to 60 PPI. Filters are an optional inclusion in the present casting system. However, when included, the filters do not require human intervention. This is because, as previously explained, the leaking phenomena of larger particles is not a problem thanks to the baffle module. Therefore, contrary to traditional tundish systems, a filter can be employed without worrying about clogging and particle leakage.

[0162] In some embodiments, a skim dam is included in the tundish to prevent any oxide film or dross floating at the surface of molten aluminum alloy to be entrained downstream of the tundish. The skim dam can be located in a portion of the tundish far enough from the casting spout to avoid creating turbulence. In some embodiments, the tundish of the present disclosure consists of the basin, the one or more baffle modules, an inlet spout, a dip tube, a casting spoutand the skim dam. The inlet spout and the dip tube control the amount of molten metal that enters the tundish. At the outlet of the tundish, the casting spout delivers the molten metal to the casting wheel.

[0163] In some embodiments, the tundish uses a series of baffles to locally change the molten metal flow. The molten metal enters the tundish at a very high speed from the top by escaping through the opening between a spout and a dip tube. With a traditional tundish that does not have a baffle module, the metal flow is turbulent and random, which prevents the settling of even very large clusters of contaminants (e.g. TiVB2). The openings and slag baffles used herein are designed to limit the turbulent flow in the center portion of the tundish. When the molten metal leaves the interior space of the baffle module, the flow is more laminar, thus promoting the settling of contaminant inclusions. Accordingly, one advantage of the baffle module is that it forces the molten metal to travel a longer distance before exiting the tundish. This longer distance is beneficial to further promote particle settling and promotes a laminar or less turbulent flow. Another advantage of the baffle module in the tundish is to restrict the molten alloy flow from the bottom forcing only the top layer of liquid out of the tundish by the use of the slag baffle. By the time the metal reaches the slag baffle, the large clusters have already settled at the bottom of the tundish and the remaining contaminant clusters are settled upstream of the slag baffle. The slag baffle thus is an efficient barrier that prevents the contaminant clusters from exiting the tundish and reaching the casting wheel and moulds.

[0164] In the Example 4 below, a tundish with a baffle module was successfully tested on a rod caster for aluminum alloys. The amount of turbulence was not an issue because the molten metal surface was stable visually, and the molten metal level measured by the laser in the casting spout was not affected by the presence of the slag baffles. The tundish lifespan achieved was longer than the targeted duration of the casts which was 7 days. The metal solidified in the modified tundish after several days in operation and the industrial process was analyzed in metallography to reveal its content.

[0165] The two part tundish described herein (i.e. basin and baffle module) can be conveniently reused for several casts and only the baffle module needs to be replaced. Before the cast, the base of the baffle module is fixed preferably by glue on the tundish to secure its position throughout the cast. After the cast, the metal solidified in the tundish and the baffle module can be removed with the block of solid metal without damaging the tundish. When the tundish is cooled, another baffle module can be fixed the basin for the next continuous cast cycle. It was found that the basin can be reused several times before it needs to be replaced. Forexample, the basin can be used for at least 5 full duration castings (e.g. 5-7 days per casting) whereas the baffle module is replaced in between every casting operation.EXAMPLE 1

[0166] It was investigated whether TiVB2 is formed and accumulates at the casting furnace stage when melting the aluminum alloy. The molten metal in the casting furnace was maintained at a temperature of 720 - 750 °C throughout the cast. The casting speed was kept in the range of 12 - 13 tons per hour. The molten metal temperature in the tundish was monitored and maintained at about 690-710 °C. These conditions were used throughout the examples herein unless specifically stated otherwise. A refractory barrier with a height of about 65% of the molten metal level in the launder ahead of the tundish was used in this experiment. The operators made visual inspection of the tundish to ensure there was no accumulation of dross at the surface of the molten aluminum and also to make sure that the level of turbulence was acceptable to continue the cast. When the cast ended, the metal close to the barrier was left undisturbed to solidify and form a deposit. The aluminum block was cut and polished to investigate the presence of TiVB2 clusters and nothing was found. This finding is very important because it shows that any filtration device located upstream of the tundish would not be effective at preventing the TiVB2 cluster at getting into the cast bar.

[0167] A sample of the remaining solidified molten metal from the casting furnace was cut in half and is shown in Fig. 4 . Fig. 4 shows the direction of the flow with arrows. The sample was divided into three samples marked as A, B and C on Fig. 4 and referred to herein as samples A, B and C. Samples A, B and C are also shown in Fig. 5 where the macrostructure is shown with arrows denoting zones richer in Sr phases, carbides, and Fe phases. Figs. 6, 7, and 8 demonstrate that samples A, B, and C respectively contained no AI4C3 et TiVB2 agglomerates. The arrows also show zones richer in Sr phases, carbides, and Fe phases. Fig. 9 shows a microscopy image of sample A which was polished which does not show any TiVB2. Fig. 10 shows the microstructure of TiVB2. Fig. 11 is a close up microscopy image at the arrow shown in Fig. 6. Fig. 12A is a scanning electron microscopy (SEM) of sample A in the same region as Fig. 11 . Fig. 12B is a SEM image showing the Ka1 emission line for Al. Fig. 12C is a SEM image showing the Ka1 emission line for Ti. Fig. 12D is a SEM image showing the La1 emission line for Sr. Fig. 12E is a SEM image showing the Ka1 emission line for V. Fig. 12F is a SEM image showing the Ka1 emission line for Fe. This experiment to capture the metal upstream of the tundish was designed to identify the origin of the contaminants (or clusters) that are trapped in the tundish. It was found that very few inclusions are present in the casting furnace and their composition does not matchthose found in the tundish. This was specifically demonstrated by the presence of Sr and Fe which were absent in inclusions trapped in the tundish. The conclusion of this experiment is that after a full duration cast, no TiVB2 clusters were identified in the casting furnace. It can then be assumed that the clusters form at the surface of the spout or the dip tube or both and they are released in a unpredictable manner during the cast. This assumption is also supported by the fact that a sudden downward movement of the dip tubes was observed during the cast to keep the molten level constant. This is associated with a release of clusters. The cleaning of the spout / dip tube aperture create more space for the molten metal to travel so the dip tube need to close (go down) to compensate otherwise the molten metal level would raise over the targeted range. Indeed, Figs. 13 and 14 show microscopy images of samples B and C respectively which were polished and these did not show any TiVB2.

[0168] In order to reduce the TiVB2 contaminants, based on the above results, the casting furnace was found to not be a suitable target for improvement since the contaminants were not found in the casting furnace and therefore are formed downstream of the casting furnace.EXAMPLE 2

[0169] It was investigated whether the agglomeration takes place in the tundish. A traditional tundish made of just a basin was used to cast aluminum alloy. The casting was performed continuously for a period of 7 days with a total mass of 2000 tons of aluminum alloy cast. A cake of 1 1 .78 mm was formed after casting. The tundish was cut in half and then the cake formed was separated into four longitudinal zones: A, B, C, and D and three transversal zones 1T, 2T, and 3T as can be seen in Fig. 15. Zone A was further divided into four zones I, II, III and IV as shown in Fig. 16. The delineated zones were analyzed by microscopy imaging, Figs. 17A, 17B, 17C, 17D, 17E, 17F, 17G, and 17H respectively show samples D, C, B, 3T, 2T, 1 T, I, and III. Fig. 18A shows contaminant agglomerates from zone 3T and Fig. 18B shows contaminant agglomerates from zone 2T. Figs. 18C and 18D are close-ups of Fig. 18B showing the presence of carbides and the structure at the joints of the grains respectively. Fig. 18C was shown as an example showing the presence of carbides in the agglomerates but all the agglomerates were found to contain carbides. Moreover, Fig. 18D is an example showing the accumulation of contaminants at the joints of the grains as this was observed throughout the metal cake. Indeed, these were even observed in zone III close to the surface. The highest contaminant agglomerates identified was at 46.57 mm (Fig. 17G) and the sedimented dense area had a thickness of 35 mm (Fig. 17G).

[0170] An agglomerate was analyzed by scanning electron microscopy (Fig. 19A) and cartography (Figs. 19B-19E). Fig. 19B shows Al, Fig. 19C shows Ti, Fig. 19D shows V, and Fig. 19E shows B. A further close up of the agglomerate by SEM is shown in Fig. 20 and the mass spectrometry analysis is shown in Fig. 21. The presence of Ti, V and B can be seen in the mass spectrometry analysis as well as the cartography.

[0171] Three agglomerates were selected for further analysis, agglomerate 1 (Fig. 22A), agglomerate 2 (Fig. 22D), and agglomerate 3 (Fig. 22G). The 5x5 mosaic images taken at 1000x magnification showing the carbides for the aggregates are presented in Figs. 22B, 22E and 22H respectively for agglomerates 1 , 2 and 3. These images were then treated and shown in Figs. 22C, 22F, and 22I. The images were used to determine quantities of TiVB2, Al, and AI4C3 (Table 1).Table 1 . Composition of agglomerates 1 , 2 and 3

[0172] Agglomerates 1-3 are non-porous, so two other agglomerates (Figs. 23A and 23B) having respectively a porosity of 50% and 37% were also analyzed and were found to have a composition similar to that of agglomerates 1-2. The porosity was analyzed with Keyence™ 7000 on the totality of the agglomerate.EXAMPLE 3

[0173] The actual tundish design was developed using a water model. Water can be used to simulate the flow of molten aluminum with great accuracy because its dynamic viscosity is very close to that of molten aluminum. This is a commonly accepted practice by people working in the field. The behaviour of TiVB2 clusters in molten metal could be reproduced using silicon oxide particles of the same size. The relative density of silicon oxide and water is very close to the relative density of TiVB2 clusters in molten aluminum.

[0174] The filtration efficiency using the water model could be assessed by gradually injecting a known mass of SiC>2 particles with a specific granulometry range. The particle mass fraction trapped in the system was determined for each granulometric range to characterize every tested design (Fig. 24). The amount of turbulence was also a criterion for selecting the best design.

[0175] Numerical simulations were also used to evaluate the effect of the tundish design. First, the results of these simulations were compared to the results obtained with the water model. The results were sufficiently close to allow optimization to be pursued with numerical simulations.EXAMPLE 4

[0176] A series of 1xxxx, 4xxx and 8xxx type aluminum alloys were cast continuously for 6 days in a 90 x 50 mm mold coated with resin. The casting system used included the baffle module (Figs. 25A-25C) the dimensions of which are shown in mm in Figs. 25B-25C. The baffle module was mounted on the basin of the tundish (Fig. 25D). The resulting tundish after 6 days of continuous casting is shown in Fig. 25E. The stability of the molten metal level was measured in the casting spout. The use of the present tundish design did not create unstable molten metal level which confirmed the design is adequate. The casting speed was even increased up to 15 tons per hour at the end of the cast to demonstrate that it could operate at this speed without any issue. The tundish is not a bottleneck in terms of casting speed and in terms of cast duration (capacity to store the trapped particles).

[0177] In a first experiment, the tundish was cut into five pieces labeled 1 to 5 as shown in Fig. 25F. In piece 1 , the metal accumulated was divided in five zones: A to E for further analysis (Fig. 25G). The cake height varied between 10 and 20 mm in zones A, B and C which are inside the baffle module. The combined cake inclusion in zones D and E is 100 mm.

[0178] Each of zones A to E were cut out and separated, they are respectively shown in Figs. 25H-25L. The cake deposit is accentuated with a dotted line on the figures. Fig. 25M is a close up of Fig. 25K (zone D). Zones D and E are outside of the baffle module and zone E is where the molten aluminum alloy exits the tundish. As marked on Fig. 25L the contaminant deposit does not reach the top or exit of the tundish and therefore the contaminants remain entrained in the tundish demonstrating that the baffle module was successful in retaining the contaminants in the tundish. This is in clear contrast to zone III in Example 2 which contained contaminants and is the exit point of the tundish which is comparable to the studied zone E in the present Example. Therefore, the baffle module of Figs. 25A-25D was shown to avoid the presence of contaminants in the castproduct as opposed to the casting system of Example 2 which does not include a baffle module and was performed with a traditional tundish.

[0179] Microscopy image of the deposit (i.e. below the dotted line in Figs. 25H-25L) were taken and presented in Figs. 26A-26G. Figs. 26A-26C show the structure at the joints of the grains which was similar in all zones A-D. Figs. 26D-26G show the structure of zone D with increasing magnification. Agglomerates having a size of up to 1 .3 mm were identified.

[0180] A contaminant agglomerate was selected and imaged by SEM (Fig. 27A). Two regions of the agglomerate marked with a square and a triangle on Fig. 27A were selected for mass spectroscopy analysis. Fig. 27B shows the results for the square region and Fig. 27C shows the results for the triangle region. In both cases, aluminum was found to be the highest concentration element. Compared to Fig. 21 for a traditional tundish, the content of Ti, V, B and C is significantly reduced. Further analysis was performed on zone D by cartography (Figs. 28A-28E) showing the presence of Al, Ti, V and B. It was observed that the composition of the contaminants is mainly TiVB2 with a presence of AI4C3.

[0181] In a second experiment, the tundish was cut in half and then in three pieces #1 , #2, and #3 (Fig. 29A). A portion of the deposit in piece #1 labeled 1 -A was cut out and is upstream of the slag baffle (Fig. 29B). Two portions of the deposit in piece #2 were cut out and labeled 2-A and 2-B (Fig. 29C). 2-A is upstream of 2-B and they are located downstream of the opening of the slag module and upstream of the slag baffle in a region between the wall of the basin and the wall of the slag baffle. Three portions of the deposit in piece #3 were cut out and labeled 3-A, 3- B, and 3-C (Fig. 29D). All three of these pieces were inside the walls of the slag baffle module (in other words upstream of the opening). It was observed that the cake deposit reduced thickness from #3 to #2 by 4 mm (from 9 mm to 5 mm). Figs. 30A, 30B, 30C, 30D, 30E, and 30F show cross sections of 1-A, 2-A, 2-B, 3-A, 3-B, and 3-C respectively. These sections were analyzed by microscopy as shown in Figs. 31 A-31 E, Figs. 32A-32E, Figs. 33A-33E, Figs. 34A-34E, Figs. 35A- 35E, and Figs. 36A-36E for 1-A, 2-A, 2-B, 3-A, 3-B, and 3-C respectively. Contaminant agglomerates were identified in sections 1-A, 2-B, and 3-B with sizes varying between 650 to 6395 pm. The last baffle at section 1-A was able to retain the agglomerates that reached that point and no further contaminant agglomerates passed beyond section 1 -A (Fig. 30A). Indeed, there was no agglomerates found in piece #3 outside of the slag baffle (see adjacent to 3-C on Fig. 29D). Carbide contaminants were only found inside the agglomerates.EXAMPLE 5

[0182] A casting of a 1xxx alloy (grain refiner 5% Ti 1 % B) was performed with the casting system of the present disclosure (see Fig. 2) for a duration of 11 months. The use of the baffle module was interrupted for 6 days during the 5thmonth for two casting runs. During the entire operation of the 11 months of casting wires having a diameter of 1 - 3 mm, the only time a breakage was observed was during the 6 days that the baffle module was not used. This clearly indicates the effectiveness of the baffle module at retaining contaminant clusters and producing cast alloy products with improved mechanical properties. The wires that were broken were subjected to analysis to confirm that the breakage occurred because of the presence of contaminant clusters. Five samples were analyzed from the 6 day period without the baffle module and they are presented in Table 2.Table 2. Samples extruded without the baffle module that lead to breaks

[0183] Figs. 37A-37C show microscopy images of the break of Sample A, showing an off- centre break. Figs. 38A-38B show microscopy images of the break of Sample B, a fish-mouth break. Figs. 39A-39C show microscopy images of the break of Sample C, showing an off-centre break. Fig. 39D shows the chemical composition spectra of sample C confirming the presence of (Ti,V)B2-Al4C3 clusters. Figs. 40A-40C show microscopy images of the break of Sample D, showing an off-centre break. Figs. 41A-41 C show microscopy images of the break of Sample E, showing an off-centre break. Figs. 42A-42C show microscopy images of the break of Sample F, showing a fish-mouth type break.

[0184] While the present disclosure has been described with particular reference to the illustrated embodiment, it will be understood that numerous modifications thereto will appear to those skilled in the art.

[0185] While the present disclosure has been described in connection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any variations, uses, or adaptations, including such departures from the present disclosure as come within known or customary practice within the art and as may be applied to the essential features hereinbefore set forth, and as follows in the scope of the appended claims.

Claims

WHAT IS CLAIMED IS:1 . A tundish comprising a basin and a baffle module, wherein the baffle module is positioned inside the basin such that the baffle module is the inlet of the tundish receiving a flow of molten alloy, and wherein the baffle module comprises walls having a first height that is greater than the level of the molten alloy in the tundish, the walls defining an interior space in which the molten alloy is received, an opening in one of the walls to allow the molten alloy to exit the interior space defined by the walls, and at least one slag baffle extending from one of the walls of the baffle module to a wall of the basin, the at least one slag baffle having a second height that is smaller than the first height and smaller than the level of the molten alloy in the tundish, wherein the at least one slag baffle is positioned such that the flow of the molten alloy passes through the at least one slag baffle to reach an outlet of the tundish.

2. The tundish of claim 1 , wherein a bottom wall of the baffle module is fixed to the basin.

3. The tundish of claim 2, wherein the bottom wall of the baffle module is glued or mechanically secured to the basin.

4. The tundish of any one of claims 1 to 3, wherein the basin comprises steel, refractory reinforced fibreglass material or refractory concrete.

5. The tundish of any one of claims 1 to 4, wherein the baffle module comprises reinforced fibreglass material.

6. The tundish of any one of claims 1 to 5, wherein the opening is positioned opposite of the slag baffle.

7. The tundish of any one of claims 1 to 6, wherein the opening defines an empty space equal to the first height.

8. The tundish of any one of claims 1 to 7, wherein the baffle module is characterized by having continuous walls.

9. The tundish of any one of claims 1 to 8, wherein the second height is 60 to 90 % of the level of the molten alloy in the tundish.

10. The tundish of any one of claims 1 to 9, wherein the second height is 70 to 85 % of the level of the molten alloy in the tundish.1 1 . The tundish of any one of claims 1 to 10, wherein the second height is 75 to 85 % of the level of the molten alloy in the tundish.

12. The tundish of any one of claims 1 to 11 , wherein the baffle module positioned in the center of the basin such that a cavity is defined between the walls of the baffle module and interior walls of the basin.

13. The tundish of any one of claims 1 to 12, wherein the tundish further comprises a skim dam.

14. The tundish of claim 13, wherein the skim dam is positioned upstream of the slag baffle.

15. The tundish of any one of claims 1 to 14, wherein the baffle module has two slag baffles.

16. A system for casting low density metal alloys comprising the tundish as defined in any one claims 1 to 15.

17. The system of claim 16, further comprising a ladle for containing a volume of molten alloy.

18. The system of claim 17, further comprising a ladle shroud located downstream of the ladle relative to a flow of the molten alloy and in fluid communication with the ladle.

19. The system of claim 18, wherein the tundish is located downstream of the ladle shroud relative to the flow of the molten alloy and in fluid communication with the ladle shroud.

20. The system of claim 19, further comprising a casting shroud downstream of the outlet of the tundish relative to the flow of the molten alloy and in fluid communication with the outlet of the tundish.21 . The system of claim 20, further comprising at least one mould in fluid communication with the casting shroud adapted to receive and mould the molten alloy into a desired shape.

22. The system of any one of claims 16 to 21 , further comprising a liquid metal filter.

23. The system of claim 22, wherein the liquid metal filter is a ceramic filter or a tortuous path.

24. A system for casting low density metal alloys, the system comprising:a ladle for containing a volume of molten alloy; a ladle shroud located downstream of the ladle relative to a flow of the molten alloy and in fluid communication with the ladle; a tundish as defined in any one of claims 1 to 15 located downstream of the ladle shroud relative to the flow of the molten alloy and in fluid communication with the ladle shroud; a casting shroud downstream of the outlet of the tundish relative to the flow of the molten alloy and in fluid communication with the outlet of the tundish; and at least one mould in fluid communication with the casting shroud adapted to receive and mould the molten alloy into a desired shape.

25. A method of casting a low density metal alloy, the method comprising: flowing a molten alloy from a ladle to a tundish, wherein the tundish comprises a basin and a baffle module, wherein the baffle module is positioned inside the basin such that the baffle module is the inlet of the tundish receiving the molten alloy, and wherein the baffle module comprises walls having a first height that is greater than the level of the molten alloy in the tundish, the walls defining an interior space in which the molten alloy is received, an opening in one of the walls to allow the molten alloy to exit the interior space defined by the walls, and at least one slag baffle extending from one of the walls of the baffle module to a wall of the basin, the at least one slag baffle having a second height that is smaller than the first height and smaller than the level of the molten alloy in the tundish; flowing the molten alloy from the tundish through the slag baffle to a mould to impart a desired shape and obtain a cast alloy product.

26. The method of claim 25, wherein the method is a continuous casting method for a casting period of at least 5 days.

27. The method of claim 26, further comprising replacing the baffle module and restarting a casting period of at least 5 days.

Images