The use of reclaimed sands in the cold-box core room is discussed, comparing three different reclamation processes. The first is called a thermal process that treats resin-bonded sand in a hot fluidized bed. The second is a wet process, removing bentonite and coal dust from green sand with water and attrition treatments. The third is a mechanical–thermal–mechanical process (pneumatic–hot fluidized bed–pneumatic), operating with a mixture of green sand and resin-bonded sand. The experimental study, in the laboratory and with foundry trials, shows that all processes result in some impurities in the sand that bring some changes in the core properties. Residues from core coatings, from bentonite and from decomposed seacoal, can lower the core strength and the bench life; at the same time, some of the residues can improve hot properties of the core, reducing veining tendency and core deformation. Therefore, some reclaimed sands can be used as special additives in the core room.
The iron foundry industry is considered, by many environmental specialists, a recycling industry, because this industry recycles a big amount of scrap as raw materials (steel and iron scrap, copper scrap, etc.). Despite this benefit, the foundry industry also produces many residues, such as spent sand, slag, dust and refractories, and among them, spent sand is the residue with the highest volume. In an iron foundry of cylinder blocks and heads, the consumption of new sand is typically 1.0–1.2 t sand/t castings. This sand is used mainly for core production, and during the casting process, it is burned and incorporated in the recycled molding sand. The new sand added in the mixer to control the grain size of the green molding sand and is also incorporated in the recycled molding sand. After each casting cycle, almost the same amount of sand must be removed from the foundry and represents the spent sand. Most of the spent sand is either landfilled or sent to other kinds of industries, like the concrete industry. In recent years, the foundry industry is building more and more sand reclamation plants, in order to reuse the treated spent sand again in the core room.
Nowadays, the main core process is the cold box (phenolic urethane resin cured with amine gas), and this process is very sensitive to sand characteristics, such as humidity, clay content and acid demand, and the reclaimed sand has to fulfill these specifications or to be used in special conditions in the core room.
In the present work, reclaimed sands obtained by three industrial reclamation processes are compared, looking for reuse in the core room. The first two processes (thermal and wet) are installed in the south of Brazil and the third process (mechanical–thermal–mechanical) is running in the north of Mexico.
Sand Reclamation Processes
The first process is called a thermal process, treating resin-bonded sand in a hot fluidized bed.1,2 It deals with scrapped cores, core sand from core blower magazines, blow pins, from a phenolic urethane cold-box core room. The goal of this industrial unit is to avoid sending chemically bonded sand to landfill. This sand contains silica sand, some additives, such as chromite sand and bauxite, and also core-coating residues. Figure 1a shows the various operation steps of the process. In a vibratory screen, the cores are disintegrated by attrition and separated from large metallic materials (tie beam, chaplet, screw and nut). The sand passes through a magnetic drum and goes to a fluidized bed furnace, where the resin is burnt at 620 °C. The sand is cooled to 50 °C, and the fines are removed in the exhaust system.
The second process is a wet treatment and deals with spent sand from a green sand molding system, containing silica sand, bentonite, coal dust, residues from thermal modification of these additives and residues of core sand incorporated at the shakeout. Figure 1b shows the process flow of an experimental plant. The sand goes through a rotary screen to remove coarse material. The removal of bentonite and coal dust is promoted by the attrition between the particles and the washing action of the water, in an attritor with two blades with opposite rotation. The fines are removed in a second rotary screen and in a series of hydrocyclones. It follows the drying of the sand in a fluidized bed at 250 °C. The slurry of the process is sent to an effluent treatment plant.3
The third process is a mechanical–thermal–mechanical process, treating a mixture of excess green sand and scrapped cores. In a rotary screen, the cores and sand lumps are disintegrated by attrition and separated from metallic materials. The sand goes to a first pneumatic attritor with four stages, to remove most of the loose fines. It follows the fluidized bed furnace, at 700 °C, to burn the resin and to remove hydroxyls from the bentonite. After being cooled, the sand goes to a second pneumatic attritor, to remove all the fines from the sand grains.4,5 Figure 1c shows the process flow.
Table 1 shows the sands tested in the experimental work. The reclaimed sands TR and WR came from SN base sand, from Araquari—Brazil, used in an iron foundry in the south of Brazil, while reclaimed sand MTMR came from SM base sand, from the USA, used in an iron foundry in the north of Mexico.
The coating used in the core room, based on lamellar minerals, contains 43% SiO2, 34% Al2O3, 6% Fe2O3, 2.7% K2O, 0.8% TiO2, 0.7% CaO, 0.7% MgO and 11% LOI (informed by the coating supplier).
The procedures of the tests with the sands and cores are shown in Table 2. Some tests were not performed with SM sand due to lack of sand at the laboratory.
X-ray examinations of the sands were conducted with a Bruker diffractometer, operating at 40 kV and 40 mA and 0.02°/s.
Thermal expansion of the sands was measured at a Netzsch dilatometer (DIL 402C), with loose sands, in a sample of Ø 6.5 × 9 mm, with a heating rate of 5 °C/min. It was not applied any compression on the sample; in this way, it was measured only expansion of the sand (from alpha to beta quartz), not contraction (from beta quartz to tridymite).
Acid demand value, a critical characteristic of the sand used for the core box process, was measured at two pH conditions (pH 2 and pH 6). In general, the higher the ADV of the sand, the lower the time the sand will be usable, because the resin starts to react sooner than normal.6,8
The hot distortion test was developed by BCIRA.9 A core in the shape of a bar (115 × 25 × 6 mm), is heated up to 890 °C and submitted to a load of 0.3 N in one end of the bar (Figure 2). The deformation of the core bar was measured, during the heating.10
The tensile strength of core sand was measured with the dog bone specimen, with 1.5% phenolic urethane resin (50/50% of each part) cured with amine gas.
Foundry tests were conducted with the step-cone casting11 for evaluation of veining and penetration tendency and with the core deformation test casting,12 as shown in Figures 3 and 4. Veining and penetration are quantitatively evaluated for each step of the cone, considering the metallostatic pressure on each step, according to a procedure developed by Giese and Thiel.13 Core deformation was measured on the casting, as the maximum deformation of the core, in the middle of the casting, as shown in Figure 4.12
Results and Discussion
Impurities were identified in all sands, mostly in the reclaimed ones. Table 1 summarizes the results. The reclaimed sands coming from green sand molding contain residuals of bentonite (active or dead), and in the case of the wet reclaimed sand (WR), without any calcination step in the reclamation process, also coal dust and coke were identified. Figure 5 shows an example of a sand grain (WR sand) with a layer of bentonite and coal dust. Some bentonite particles in MTMR sand exhibit chlorine, probably from the underground water of Mexico, used in the mulling or cooling processes. TR sand, coming from residues of the core room, presents some core-coating particles as impurities, identified with EDS microanalysis by the high content of K and Al.
X-ray diffraction of all sands presents alpha quartz as the main phase (besides the impurities already mentioned). Figure 6 shows the results for MTMR sand. As the temperatures of the thermal reclamation processes (620–700 °C) are lower than the temperature of beta quartz to tridymite transformation (867 °C), it was not expected that the reclamation processes could stabilize the sand. However, WR and MTMR sands contain recirculated molding sand that was exposed many times to iron pouring temperatures, and the results show that this exposure does not stabilize a significant amount of the sand. Unfortunately, the reclaimed sands will suffer again the high expansion on heating through 573 °C, due to alpha–beta transformation, as new sand. This was confirmed by measuring the linear expansion characteristics on heating in the dilatometer (Figure 7).
Table 3 shows the properties of the investigated sands. In general, reclaimed sands display a reduction of fines, thus reducing the grain fineness number (GFN) and the specific area of the sand. The permeability of the reclaimed sands is also increased due to the reduction of the fines content.
The wet reclaimed sand shows high loss of ignition (LOI) and high AFS clay contents, caused by the presence of residuals of clay and coal dust.
The sands from the USA, base sand (SM) and reclaimed (MTMR), display high values of ADV (acid demand value), in both pH 2- and pH 6-tested conditions (Table 3).
No significant difference in the results of flowing time and bulk density was observed, so it is not expected any difficulty on the control of the mixing and blowing processes using reclaimed sands, either pure or mixed with new sand.
Figure 8 displays the results of tensile tests of cores, in three different aging conditions (immediate, after 24 h and after 24 h under humid conditions). One can see the typical increase in strength after 24 h, due to the completion of the curing reaction of the resin. With WR sand, the increase with 24-h aging is small (compared with the results of other sands), similar to that observed by LaFay et al.16 also with a sand from a wet reclamation process. This is attributed to impurities in the WR sand (bentonite and coal dust residues). To investigate this in detail, a sample of the WR sand was submitted to a calcination heat treatment; Figure 9 shows the results obtained. One can see the increase in the strength after aging for 24 h, following the same behavior presented by the other sands (Figure 8).
In Figure 10, one can see the results of the bench life tests. In general, the bench life is lower using reclaimed sands, particularly with TR and MTMR sand. The effect of contaminations of base sand on the bench life, associated with increased ADV, is well known.6,8 This could explain in part the lower bench life results with SM and MTMR sands. Although TR sand did not present high ADV (Table 2), the fine fraction of this sand was investigated. This sand contains residues from core coatings, which could be transformed in fine particles during the reclamation process and are not completely removed in the final steps. It was found, in the fraction under 100 mesh (< 0.15 mm) of TR sand, an ADV (pH2) of 15.1 ml (compared to the average ADV pH 2 of 4.3 ml for this sand), The high ADV of the fines could possibly be connected with the reduced bench life of this sand.
The results of hot distortion test are presented in Figure 11. The highest deformation occurred in about 10 s for all sands. Reclaimed sands display lower results of hot deformation, in particular WR sand.
Results from casting trials with step-cone casting and with core deformation casting are presented in Figures 12, 13 and 14. In general, the castings produced with reclaimed sands display reduced veining intensity, lower penetration and lower core deformation, compared to the base sand. The WR sand in particular shows good results concerning veining and core deformation. This is attributed to the presence of an oolitic layer over the sand grains, formed from dead clay, acting as deformation cushion and accommodating in this way the expansion of the silica transformation (Table 4). The removal of fines by the reclamation process, maintaining the oolitic layer over the sand grains, produces special silica sand for iron foundry application. This sand can be considered an anti-expansion additive for core making.
Figures 15, 16 and 17 show additional correlations between the results of laboratory tests and foundry trials. It was found that, in this concern, the deformation measured on the hot distortion test correlates well with the results of veining tendency (Figure 15), penetration tendency (Figure 16) and deformation of the core (Figure 17). Correlations of casting defects with the linear thermal expansion, measured by the dilatometer, were not so good. The conclusion is that it is possible to have a good estimation of the behavior of core sand in the foundry, concerning core deformation, veining tendency and penetration tendency, from the measurement of hot deformation in hot distortion test, a laboratory test.
From the results of the present study, the following conclusions can be drawn:
The reclamation processes studied in the present work can produce sands with properties suitable to be used in the core room. Depending on the reclamation process, the main concerns are strength of the core and bench life. All the studied reclamation processes reduced the fines content of the sand, increasing in this way the GFN and reducing the specific surface of the sand.
All the base sands and the reclaimed sands present alpha quartz as the main phase, and they will present again the high expansion of the alpha–beta quartz transformation. The high expansion characteristic of the sand can be reduced for the reclaimed sands by some oolitic layer over the grains.
All the reclaimed sands show veining tendency, penetration tendency and core deformation lower than the base sand, in the casting trials. The wet reclaimed sand presents the best behavior concerning these foundry defects.
The behavior of core sand concerning core deformation, veining and penetration tendency can be predicted by a laboratory test, the hot distortion test, measuring the maximum hot deformation in this test.
R.L. Langham, Reclamation by the fluidised thermal process, in Reclamation and Recycling of Molding Sands. BCIRA Seminar, (Stratford-upon-Avon, 1988), pp. 1–10
F. Peixoto, W.L. Guesser, Reutilização de areia regenerada termicamente, in Brazilian Foundry Congr CONAF ABIFA, (São Paulo, 2003)
E.C. Carline, G. Fabris, I. Masiero, M.C. Moreira, W.L. Guesser, Processo de tratamento de areia excedente de fundição para uso em macharia e moldagem. PI 1000043-7 (INPI Brasil, 2010)
D. Silsby, Sand reclamation overview, in AFS Sand Casting Conference, (Indianapolis, 2014)
M. Granlund, D.V. Silsbi, AFS Trans. 116, 309–328 (2008)
J. Werling, AFS Trans. 110, 02–30 (2002)
American Foundry Society, Mold and Core Testing Handbook, 2nd edn. (Des Plaines, IL, 1989)
W. Tilch, M. Martin, Giessereiforschung 58(3), 18–31 (2006)
BCIRA, Hot distortion test for chemically bonded sands, in BCIRA Broadsheet 177, 1977, pp 1–2
S. McIntyre, S.M. Strobl, Foundry M & T, 1988, pp 22–26
J. Thiel, AFS Trans. 119, 369–378 (2011)
G.C. Fontaine, K.B. Horton, AFS Trans. 98, 06–16 (1990)
S.R. Giese, J. Thiel, AFS Trans. 115, 07.146 (2007)
E.C. Carline, Estudo sobre a expansão térmica linear das areias de fundição e sua influência sobre as propriedades a quente dos machos (Master Degree, UDESC, Joinville, Brasil, 2011)
E.C. Silva, Influência das propriedades das areias regeneradas para aplicação em processo caixa fria com resina fenólico uretânica (Master Degree, UDESC, Joinville, Brasil, 2019)
V. LaFay, C. Grefhorst, J. Tibbs, J. Thiel, AFS Trans. 123, 125–134 (2015)
The experiments reported in this paper were conducted in the partnership with Tupy Foundry and UDESC (Santa Catarina State University).
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This paper is an invited submission to IJMC selected from presentations at the 2nd Carl Loper 2019 Cast Iron Symposium held September 30 to October 1, 2019, in Bilbao, Spain.
About this article
Cite this article
Silva, E.C., Masiero, I. & Guesser, W.L. Comparing Sands From Different Reclamation Processes for Use in the Core Room of Cylinder Heads and Cylinder Blocks Production. Inter Metalcast 14, 706–716 (2020). https://doi.org/10.1007/s40962-019-00400-6
- sand reclamation
- base sand
- cold box
- metal penetration
- core deformation