GB2047588A - Reclamation of foundry sand - Google Patents

Abstract

A process for the reclamation of CO2-silicate bonded casting sand includes the step of subjecting the sand after it has been used for casting to treatment with carbon dioxide in the presence of water until substantially all of the reactive soluble alkali metal compounds present in the used sand which are capable of being carbonated are carbonated. This carbonation step improves the removal of the binder by a subsequent mechanical attrition step. In effecting the process, a batch of the used sand may be wetted and subjected to a vacuum, the vacuum then being broken by the introduction of carbon dioxide which may be at a superatmospheric pressure.

Description

SPECIFICATION Improvements relating to the reclamation of foundry sand A common method of binding foundry sand to form a core or mould is the CO2-silicate process.

In this process, said is mixed with a binding agent which is composed principally of an alkali metal silicate. Sodium silicate is the binding agent that is used in practice and this may be considered as comprising a compound formed from the oxides of sodium and silicon with a variable amount of water. Various other additives may be included in the binder to improve its green strength and/or to help in the breaking down of the mould material after casting and knockout from the completed casing after use.

The sand and binding mixture is compacted around or inside a pattern and this compacted mass, or compact is then exposed to carbon dioxide gas which increases the viscosity of the sodium silicate in the binder and produces silica gel and sodium carbonate so binding the sand together to form a mould or core. Molten metal is then cast into the mould and allowed to cool and solidify. The mould or core is then broken up to remove it from the metal casting.

The sand which has been in contact with the molten metal and that which is close to the molten metal is bound very tightly after the casting stage and becomes extremely hard. This extra hardness is caused partly by the dehydration of the silica gel and partly by some fusion taking place between the sand particles and sodium carbonate. As the sodium silicate is present in the form of the oxides of sodium and silicon with a variable amount of water, some unreacted sodium hydroxide is usually present in the binder and this, together with the sodium carbonate and in the presence of sand and heat, fuses the sand to form a vitreous material.

At present, most of the sand used in the CO2-silicate process is discarded after it has been used in the casting of metal. If the used foundry sand is broken down and has further silicate binder added to it, the total percentage of reactive sodium compounds present in the sand and binder mixture as vitrifiable material e.g. sodium oxide or hydroxide, approaches or exceeds 1% as the oxide, by the weight of sand. As this concentration is approached, the tendency of the sand to further vitrification during casting is increased. This leads increasingly to the sand burning onto the metal and/or a breaking up of the sand on the surface of the mould which results in the penetration of the molten metal into the mould and results in a casting with an unacceptable surface finish. Thus, it is not possible to reuse the foundry sand satisfactorily unless the binding agent can somehow be removed.Clearly, the more times the sand is recycled, the greater the build-up of the reactive sodium compounds and hence the worse the problem becomes. Sometimes used sand, after being broken up is used as a backing sand and, in this case, it is used to form the parts of the mould which will not be in direct contact with the molten metal. This means that the mould has to be formed in two separate stages using first virgin sand against the surface of the pattern and then the recycled sand to pack out the remainder of the mould.

Clearly, it is desirable to be able to reuse the foundry sand more than once, since, not only is there the expense and difficulty in obtaining sand of an appropriate grade, but there is also the additional expense of transporting and disposing of the used sand. These are basically two methods which have been used to reclaim CO2-silicate bonded sand, wet processes and dry processes. In a typical wet process the used sand is broken up and then washed to remove most of the binder. The sand then has to be dried before re-use. The water containing the binder also has to be treated before it can be discharged and this, together with the cost of drying the sand tends to make this wet process uneconomic.

The dry processes are based on a mechanical attrition step to mechanically break the bond between the binder and the sand particles and this attrition step is followed by a mechanical separation of the binder from the sand particles. This mechanical technique is partly successful, but at present, it is difficult to remove sufficient of the binder and with it the reactive sodium compounds to compensate for those to be added with the binder when the sand is reused and so prevent build-up of the reactive sodium compounds. Using the existing dry reclamation processes, only a small quantity of this reclaimed sand can be mixed with the virgin sand before a build-up occurs of the reactive sodium compounds and reclamation is therefore generally uneconomic.

Experiments have also been carried out using potassium silicate as the binding agent since this has a higher temperature at which it reacts with the said to vitrify it. However, potassium silicate is much more expensive than sodium silicate, and thus this process is also uneconomic.

In accordance with this invention, the reclamation of CO2-silicate bonded casting sand includes the step of subjecting the sand after it has been used for casting to reaction with carbon dioxide in the presence of water until substantially all of the alkali metal compounds that are present in the used sand and that are capable of being carbonated under the conditions of the reaction, are carbonated.

This carbonation step converts all the reactive alkali metal compounds present in the binder, for example the alkali metal oxide and hydroxide, to the carbonate and/or the bicarbonate. We have found that when the said is then subjected to a conventional attrition and separation stage, much more of the binder is removed from the sand leading to a great reduction in the number of alkali metal compounds present in the sand which in turn means that a very much greater proportion of reclaimed sand can be reused.

In practice sodium silicate is used as the principal constituent in the binder although potassium or other alkali metal silicates may be used.

Preferably the sand which has been used for casting is broken up into individual grain sized particles before the carbonation treatment with the carbon dioxide and, in general, this means that the sand is broken up into particles that will pass through a sieve of mesh 1 mm. This breaking up of the sand is usually carried out by mechanical means such as a rock crusher, followed by separation on a vibratory screen.

The presence of water during the carbonation step accelerates the speed of the carbonation step so enabling the end point to be reached with a shorter treatment time and also shifts the equilibrium to give a greater degree of carbonation at the end point. The sand used may be exposed to carbon dioxide which is saturated with water vapour but preferably, the used sand is subjected to an initial treatment with steam or mixed with water prior to the carbonation step.

Sand has to be dried before the subsequent attrition and separation stage and consequently, it is desirable to introduce as little water as possible into the sand before and during the carbonation stage as is consistent with increasing the speed and shifting the end point of the reaction. We have found that as little water as 3% by weight of the sand leads to a practicable carbonation process but the water may be present in any greater quantity. In practice, a compromise is achieved between reducing the treatment time necessary for the carbonation, attaining the required degree of carbonation, and the time and cost of a subsequent drying step.

The treatment with carbon dioxide may be continued until it is determined that substantially all of the alkali metal compounds present in the used sand have been carbonated. In practice, the treatment time is dependent upon the quantity of binder in the sand being treated so that once the treatment time for that particular binder has been established, the approximate treatment may be determined by considering the concentration of binder that was added to the sand during the mixing stage.

Typically, a concentration of between 2% and 5% W/W of binder based on sodium silicate solution is added to foundry sand before the casting operation and such an addition will normally result in a sodium oxide and hydroxide content of the used foundry sand of the order of 0. 15% by weight as the oxide, for every 1% addition of binder. The minimum amount of carbon dioxide that is needed to carbonate this quantity of sodium oxide and hydroxide is between two thirds and one and a half times the weight of the oxide and hydroxide present measured as the oxide.

The used foundry sand may be subjected to the carbonation treatment in a static bed, a fluidised bed fluidised with carbon dioxide, or tumbled in a drum with carbon dioxide flowing through it. In all of these cases, the carbon dioxide is normally re-cycled and the residence time of the treatment is adjusted until substantially all of the reactive sodium compounds have been carbonated. Clearly, these techniques can be used as part of a continuous process for the reclamation of silicate bonded foundry sand.

Preferably the used foundry sand is treated in a batch process in which a batch of the used sand is subjected to a vacuum and then the vacuum broken by the intorudction of carbon dioxide. Preferably the carbon dioxide is introduced at a superatmospheric pressure so that the batch of used sand is exposed to treatment with carbon dioxide at a superatmospheric pressure.

A superatmospheric pressure of two atmospheres has been found to give satisfactory results.

In general, it is preferred to conduct the carbonation under conditions such as to yield bicarbonate rather than carbonate. An equilibrium is established when approximately three quarters of the reactive compounds have been carbonated to form sodium bicarbonate and about a quarter of the reactive sodium compounds are present as sodium carbonate. When the carbonation of the reactive sodium compounds in the binder is carried out to this stage, we have found that a separation as high as 70% of the binder material can be achieved as compared with a separation of only 16% when untreated used sand is subjected to the same, single attrition stage.

For this reason, it is preferred that when the sand is subjected to a drying step this drying step is carried out at a temperature below that at which the bicarbonate breaks down to form the carbonate. It has been found that drying at a temperature between ambient and 50"C gives the greatest removal of the binder.

However, even when the sodium present is entirely in the form of sodium carbonate there is still a substantial improvement in the separation of binder from the sand. As an example of this, when sand is subjected to a drying treatment at around 1 75 C which results in all the bicarbonate being converted to carbonate a separation of about 25% of the binder from the sand occurs compared with a separation for an untreated sample of about 16% when subjected to the same, single attrition stage.

As the sand is re-cycled and so subjected to repeated attrition stages it suffers mechanical damage, which is believed to smooth and round the individual particles of sand and this impairs their bonding properties. Further, there does appear to be a gradual build-up in the content of the reactive sodium compounds when attempts are made to use a re-cycle rate of 100% for the used sand. It is therefore preferred that the re-use rate of the reclaimed sand is in a range from 60%-80% on a continuous basis. Such a rate of re-use means that individual particles of sand are typically re-used about four times. Such a re-use rate leads to a considerable saving in the cost of obtaining and disposing of the sand for a foundry.

Various examples of the reclamation of CO2-silicate bonded casting sand in accordance with this invention will now be described and contrasted with those used conventionally.

Example 1.

In this example, Redhill sand grade AFS65 which has been bonded with 5% by weight of sodium silicate C112 and used as a mould in which molten steel was cast, was broken down using hammers until it passed through a sieve having a mesh size of 1 mm. C112 silicate is manufactured by l.C.I. and has a silicon oxide to sodium oxide ratio of 2:1. The sand was then split into three separate samples A, B, and C. Sample C was left on one side to act as a control and samples A and B were subjected to a carbonation step in accordance with this invention.

Samples A and B were both treated by being wetted with 5% by weight of water and then carbon dioxide gas was passed through the samples for two hours at a rate of 0.95 m3 per hour. Sample A was spread out into a layer 2.5 cm thick and dried in a fan assisted oven at 175"C for one hour and then allowed to cool in the oven. Sample B was spread out to a depth of 0.5 cm and allowed to dry at roomvtemperature for about two days.

After being dried, all three samples were analysed to determine their relative contents of sodium hydroxide, sodium carbonate and sodium bicarbonate by titration with hydrochloric acid using phenolphthalein and screened methyl orange as indicators.

The results of this analysis showed that a substantial hydroxide content was present in sample C, the untreated sample, but that there was substantially no hydroxide in samples A or B.

Sample A had substantially no bicarbonate which is to be expected since this was dried at 175on which is above the temperature at which sodium bicarbonate breaks down to form sodium carbonate. Sample B had about half as much bicarbonate present as carbonate when both were expressed as percentages by weight as sodium oxide.

Each of the separate samples A, B and C were then subjected to mechanical attrition in which sand was allowed to fall under gravity onto an impellor having a radius of 25cm which was rotated to 2040 rpm. The sand particles were flung by the impellor against a fixed target sited over a collection hopper. A flow of air was arranged transverse to the sand particles falling into the collection hopper to remove fines which in this case, were formed by the binder broken off the sand particles on impact against the target. Finally, the sand collected in the hopper is placed on a fluidised bed and any remaining fines separated by an air flow across the top of the fluidised bed.

Sodium determinations were carried out on small portions of each of the samples A, B and C both before and after attrition and the results of these determinations are shown in Table 1 in which sodium content is given as sodium oxide.

TABLE 1 Sodium Content % Sample Sample Sample as w/w Na2O A B C Before 0.616 0.702 0.676 Attrition 0.658 0.651 0.662 After 0.496 0.397 0.582 Attrition 0.487 0.380 0.539 % removed 22.8 42.5 16.1 Finally, a sieve analysis was carried out on unused Redhill sand AFS 65 and each of the samples A, B and C and the results of these sieve analyses are shown in Table 2.

TABLE 2 Sieve Unused Micron Redhill Sample Sample Sample Size AFS 65 A B C 699 0.6 0.3 0.2 0.2 500 2.9 3.4 3.0 3.3 353 11.4 14.3 11.4 11.7 251 38.9 35.2 31.0 31.6 211 20.0 18.5 19.0 18.1 152 22.6 22.8 27.6 27.2 104 2.9 4.1 5.7 5.6 76 0.6 1.3 2.1 2.1 53 0.1 0.1 0.1 0.3 Tray - - TOTAL 100.0 100.0 100.1 100.1 The results of these tables clearly show that the increase in the carbonation of the samples as a result of subjecting them to a carbonation step in accordance with this invention, as contrasted with sample C, increase the percentage of the sodium content of samples A and B which was removed on the attrition step as compared to sample C.Finally, table 2 shows that the physical sizes of the sand particles after reclamation are, on average, slightly smaller as can be expected when sand is subjected to an attrition operation, but is still comparable with the sieve analysis of unused sand and also that the analysis on samples A, B and C are generally similar.

Example 2 In this example, a Redhill sand of grade AFS 110 bonded with 23% of a proprietary silicate binder Carsil 7, (Registered Trade Mark) which had been used to form a mould for an aluminium casting was broken down using hammers until all the grains produced passed through a 1 mm sieve. A portion of this material was kept on one side to act as a control sample D, whilst the remainder of the material was split into five further samples, E, F, G, H and I.Sample E was wetted with i of water by weight and exposed to an excess of carbon dioxide flowing through it for half an hour whilst sample F was wetted with 1 % by weight of water and exposed to an excess of carbon dioxide flowing through it for up to one hour and sample G was wetted with 5% of water by weight and exposed to an excess of carbon dioxide flowing through it for two hours. Sample H was not wetted at all but was exposed to an excess of carbon dioxide flowing through it for two hours and sample I was dried at about 60"C for two hours and then not wetted at all but exposed to an excess of carbon dioxide flowing through it for two hours. Table 3 shows the chemical constitution of the samples D. E. F. G.H and I with the hydroxide, carbonate and bicarbonate proportions expressed as percentages by weight as sodium oxide.

TABLE 3 Total Sample Hydroxide Carbonate Bicarbonate Alkali D 0.06 0.26 NIL 0.32 E 0.00 0.14 0.16 0.31 F 0.00 0.11 0.19 0.30 G 0.00 0.21 0.10 0.30 H 0.02 0.14 0.12 0.28 0.05 0.27 NIL 0.31 From this table it is clear that even the minimum carbonation treatment using half a percent by weight of water and treatment with the carbon dioxide for half an hour results in all of the hydroxide being converted into carbonate or bicarbonate. Upon attrition, sample D and sample I behaved similarly to sample C in Example 1 and samples E. F. G and H gave enhanced percentages of sodium removal when compared with sample D, with sample F which has the greatest concentration of bicarbonate giving the greatest percentage removal of sodium.Sample I which was dried before the carbonation treatment, showed only a very slight change in its chemical composition as a result of a carbonation treatment and indicated that sample H contained some water although none was added and it is this already present water which is assisting in the carbonation step.

Example 3 To ensure reproducibility the "used" sand used for this and subsequent examples was prepared by the following laboratory method. Firstly 6kg of Chelford 50 sand were mixed with 3% on a weight for weight basis of F.G. 11 2 silicate marketed by Joseph Crosfield a Sons Limited of Warrington, using a Hobart food mixer. This mixture was transferred to an 8" plastics flower pot and compacted. A single vent hole was made in the centre of the compacted sand to within one inch of the bottom of the flowerpot. The compacted sand was then gassed with carbon dioxide at a flow rate of 4 standard cubic feet per minute (0.12 standard cubic metres per minute) for 45 seconds using a gassing cup. The hardened sand mass was then broken into manageable lumps and reduced to grain size using the Hobart mincer.The resulting sand was then dedusted in a fluidised bed and this procedure produced a reasonably consistent used sand which had a soda content of between 0.35 and 0.39% weight for weight expressed as sodium oxide. Samples of the used sand were treated in a batch process by adding a quantity of water to the used sand and then subjecting this wet used sand to a vacuum followed by breaking the vacuum with carbon dioxide at atmospheric or superatmospheric pressure. The vacuum that was drawn over the samples was between 71 and 73 cm of mercury and, after this treatment the samples were dried overnight at ambient temperature.Portions of each of the samples were tested to determine their alkalinities both before and after the vacuum and carbonation treatment and the samples were then subjected to an attrition step similar to that used in the first example and the percentage soda removals were determined after the attrition step. Six different samples were treated as follows.

Sample 1 Reference sample. No carbonation treatment at all Sample 2 Treatment with 1 % of water and then treatment with a flow of carbon dioxide gas at 20 standard cubic feet per hour (0.57 cubic metres per hour) for five minutes Sample 3 Addition of 1 % by weight of water, evacuate, and introducle carbon dioxide at a pressure of 5 p.s.i.g. for five minutes Sample 4 Add 1 % water by weight, evacuate, subject to carbon dioxide at a pressure of 30 p.s.i.g. for five minutes Sample 5 Add 1 % by weight of water, evacuate, subject to carbon dioxide at 30 p.s.i.g. for five minutes, re-evacuate and then break the vacuum with air, and Sample 6 Add 1 % by weight of water, evacuate, subject to carbon dioxide at 30 p.s.i.g. for fifteen minutes.

The results of these tests are shown in Table 4.

TABLE 4 Sample % Soda Removal 1 2 54.4 3 66.3 4 57.3 5 62.5 6 60.7 The apparent increase in soda content shown by sample 1 gives an indication of the likely error in the results of these tests.

Example 4 A series of tests generally similar to those carried out in Example 3 were carried out using a softer vacuum of the order of 45-50 cm of mercury and in these tests the following samples were prepared.

Sample 7 Add 1 % by weight of water, evacuate, and subject to carbon dioxide at 30 p.s.i.g. for five minutes Sample 8 Add 1 % of water, evacuate, subject to carbon dioxide at 30 p.s.i.g. for 1 minute, and Sample 9 Add 0.5% by weight of water, evacuate, subject to carbon dioxide for two minutes at 30 p.s.i.g.

These samples were then subjected to attrition and tested as before. The results of these tests are shown in Table 5.

TABLE 5 Sample % Soda Removal 7 44.7 8 34.3 9 63.8 These results, whilst somewhat random, appear to show that better soda removal can be obtained by subjecting the used sand to a higher vacuum.

Example 5 Further experiments were carried out to optimise the treatment conditions for the process in accordance with this invention when it includes an evacuation step in the carbonation procedure. During these optimisation tests, a vacuum of between 71 and 73 cm of mercury was used. The general procedure for these examples follows those for Example 3 and Example 4 and the four samples that were prepared were prepared as follows.

Sample 10 Add 1 % by weight of water, evaucate, and subject to carbon dioxide at 30 p.s.i.g. for five minutes Sample 11 Add 1 % by weight of water, evacuate, subject to carbon dioxide at 30 p.s.i.g.

for 1 minute, Sample 12 Add 0.5% by weight of water, evacuate, subject to carbon dioxide at 30 p.s.i.g.

for two minutes.

Sample 1 3 Add no water, evacuate, subject to carbon dioxide at 30 p.s.i.g. for five minutes.

These samples were then subjected to attrition and tested as before. The results of these tests are shown in Table 6.

TABLE 6 Sample % Soda Removal 10 70.0 11 63.5 12 70.7 13 8.8 These results show that the attrition efficiency is reasonably insensitive to changes in the water content and in the duration of carbonation treatment with the exception that some water must be present and it appears from these results that an addition of 3% by weight of water and a carbonation step of 2 minutes at 30 p.s.i.g. is likely to produce satisfactory results.

Example 6 Some experiments were carried out to investigate the rebonding characteristics of oncereclaimed used sand. The sand was reclaimed by the treatment outlined in the Example 5 as the optimum treatment i.e. the used sand was mixed with 3% by weight of water, evacuated and then exposed to carbon dioxide at 30 p.s.i.g. for two minutes. After this carbonation treatment, the sand was subjected to attrition and dedusted. The used sand was then mixed with a proportion of virgin sand and test pieces for compression strength determinations were prepared using the mixture of reclaimed and virgin sand together with further additions of silicate and water. The silicate used throughout was type FG112 silicate manufactured by Joseph Crosfield a Sons.The test pieces for compression strength determinations were prepared and evaluated using standard test equipment in which two inch (5.08 cm) diameter test pieces were prepared and gassed at a carbon dioxide flow rate of 4 litres per minute for a time of 30 seconds. The compression strengths were determined after standing times of 0, 1 and 24 hours.

Sample 14 Reference Sample. 100% virgin Chelford 50 sand plus 3% w/w silicate Sample 1 5 60% reclaimed sand plus 40% virgin Chelford 50 sand plus 3% w/w silicate Sample 1 6 60% reclaimed sand plus 40% virgin Chelford 50 sand plus 3% weight for weight silicate plus 0.5% weight of weight water Sample 1 7 60% reclaimed sand plus 40% virgin Chelford 50 sand plus 3% weight for weight silicate plus 1% weight for weight water Sample 18 60% reclaimed sand plus 40% virgin Chelford 50 sand plus 2.5% weight for weight silicate plus 0.5% w/w water Sample 1 9 80% reclaimed sand plus 20% virgin Chelford 50 sand plus 3% weight for weight silicate Sample 20 80% reclaimed sand plus 20% virgin Chelford 50 sand plus 3% weight for weight silicate plus 0.5% weight for weight water Sample 21 80% reclaimed sand plus 20% virgin Chelford 50 sand plus 3% weight for weight silicate plus 1 % weight for weight water Sample 22 80% reclaimed sand plus 20% virgin Chelford 50 sand plus 2.5% weight for weight silicate plus 0.5% weight for weight water.

Sample 23 60% used and not subjected to the carbonation and reclaiming procedures plus 40% virgin Chelford 50 sand plus 3% weight for weight silicate plus 0.5% weight for weight water.

The results of these tests are shown in Table 7.

TABLE 7 Strength Ibs/sq. inch Sample No 14 15 16 17 18 19 20 21 22 23 Standing Time (Hours) 0 79 89 78 89 95 84 127 63 150 134 209 > 280 157 109 138 106 258 59 179 268 24 361 425 280 245 161 120 302 593 110 From this table it can be seen that both 60% and 80% reclaim levels give results which are reasonably similar to the strength development attained with virgin sand and that the use of 3% w/w of silicate with 0.5% water addition gave results which were particularly similar to those shown for virgin sand. Whilst the strength of the sample 23 was adequate, the build up of soda in the resulting re-used sand would not make it suitable for use as a mould.

Claims (11)

1. A process for the reclamation of CO2-silicate bonded casting sand including the step of subjecting the sand after it has been used for casting, to reaction with carbon dioxide in the presence of water until substantially all of the alkali metal compounds that are present in the used sand and that are capable of being carbonated under the conditions of the reaction are carbonated.

2. A process according to claim 1, in which the sand which has been used for casting is broken up into individual grain sized particles before the carbonation treatment with the carbon dioxide.

3. A process according to claim 1 or 2, in which the water present during the carbonation step is present in an amount from :% to 2% by weight.

4. A process according to any one of the preceding claims, carried out as a batch process, a batch of the used sand being subjected to a vacuum and then the vacuum being broken by the introduction of carbon dioxide.

5. A process according to claim 4, in which the carbon dioxide is introduced at a superatmospheric pressure so that the batch of used sand is first evacuated and then exposed to treatment with carbon dioxide at a super-atmospheric pressure.

6. A process according to claim 4 or 5, in which the batch of used sand is carbonated by carbon dioxide at a pressure of substantially two atmospheres gauge.

7. A process according to any one of the preceding claims, in which after the carbonation, the used sand is subjected to a drying step which is carried out at a temperature below that at which the bicarbonate breaks down to form the carbonate.

8. A process according to any one of the preceding claims, in which the sand is subjected to a drying step carried out at a temperature in a range from ambient to 50"C.

9. A process according to any one of the preceding claims, also including as a final step subjecting the sand to mechanical attrition followed by a separation stage to remove spent and carbonated binder from the used sand.

1 0. A process according to claim 9, in which the mechanical attrition step is carried out by flinging the sand particles from an impellor onto a target.

11. A process according to claim 9 in which the mechanical attrition step is carried out in a vibratory crusher in which the sand particles are subjected to vigorous agitation.

1 2. A process according to claim 1, substantially as described.