Reclamation of Mixtures of Spent Sands of Inorganic and Organic Type

The results of investigations of the reclamation of spent moulding and core sands, originated from one of the Polish foundry plants, are presented in the paper. Four mixtures consisting of two types of spent sands (spent moulding sand and spent core sand) were subjected to the regeneration process. Each tested mixture consisted of an inorganic type spent moulding sand and of an organic type spent core sand. Proportions of mutual fractions of spent moulding and core sands in mixtures was 70%-30% and was representative for the waste sands from the foundry, from which these sands originated.


Introduction
The regeneration process defined as the whole of processes allowing for the reclamation and reuse of matrices of spent moulding and core sands is still being improved, from the point of view of searching for more efficient methods. This problem becomes specially important when the regeneration process is applied for mixtures of spent sands, which often happens in a foundry practice, where after casting knocking out the spent moulding and core sands are directed for the reclamation process. In such case the reclamation process of mixed sands is considered and its product, applied usually as a fresh matrix substitute in preparations of moulding sands, should be specially controlled with taking into account its granulation characteristic, chemical reaction as well as the amount of spent binding material not removed from the grain surfaces during the reclamation process [1][2].
Since the regeneration of mixed spent sands of inorganic and organic type was investigated it was necessary -in order to assess the amount of spent binding materials remaining on matrix grains -to apply methods characteristic for the reclaims originated from the technology of organic sands and inorganic ones [3].

The research scope
The following mixtures of spent moulding and core sands were subjected to the reclamation process:  SPENT SAND I, consisting in 70% of a spent moulding sand with the GEOPOL binder and in 30% of a spent core sand with the REZOLIT binder,  SPENT SAND II, consisting in 70% of a spent moulding sand with the GEOPOL binder and in 30% of a spent core sand with the ESTROFEN binder,  SPENT SAND III, consisting in 70% of a spent moulding sand with the GEOPOL binder and in 30% of a spent core sand with the AVENOL binder,  SPENT SAND IV, consisting in 70% of a spent moulding sand with WATER-GLASS 145 and in 30% of a spent core sand with the ESTROFEN binder.
The obtained reclaimed materials were subjected to the following investigations:  Determination of the amount of dusts generated due to the reclamation process,  Sieve analysis of the reclaim,  Determination of ignition losses of spent sands and reclaims,  pH values and the acid demand value (ADV),  Na 2 O content in the reclaim,  Tensile strength determination R m u of moulding sands samples (made after storing time of 1.5h, 3h and 24h), in which this reclaim was a matrix component.

Experimental stand
The regeneration process was performed in the prototype vibratory reclaimer REGMAS (Fig. 1). The set of sieves, in a vertical arrangement, was inside the reclaimer body. Under an influence of vibrations caused by the operation of two electric vibrators the spent sand is disintegrated and then goes to the lower part of the reclaimer, where the secondary regeneration occurs. Reclaimed materials together with a binder removed from the grains of spent sands (in a form of dusts) are transported by a spiral trough from the lower part of the device of discharge in the upper direction where it goes into the cascade classifier, in which a separation of the powdery fraction occurs.

The obtained results
The results of measuring the amount of dusts formed due to the reclamation process are presented in Figure 2. It can be noticed, that the highest amount of dusts was formed as the result of the regeneration process of spent sand II, slightly smaller at the regeneration of spent sand I and the lowest at spent sand IV. It is assumed, that the amount of dusts generated during the reclamation process has an essential influence on the purification degree of the reclaimed material [4][5][6], since these dusts contain large amounts of the spent binder rubbed out from grain surfaces. The sieve analysis results of reclaimed materials before their dedusting in the cascade classifier are listed in Table 1. Samples for these tests were taken from the trough transporting reclaimed materials to the classifier. It can be noticed, on the bases of the obtained results, that successive regeneration cycles cause decreasing average grain diameters d L (average logarithmic diameter) , d a (average arithmetic diameter), d g (average geometric diameter) and d h (average harmonic diameter). In case of determining the theoretical specific surface of the sand grains set, it is seen that with the increased operational intensity of the reclaiming device this parameter is increasing. This is caused by increasing amounts of a fine, spent binding material (rubbed out from grain surfaces of the spent moulding sand), which is later removed during the pneumatic classification.
where: U c -ignition loss of the reclaim, S -ignition loss of the spent sand.
The results of index W SR values are listed in Table 3.
The results of testing ignition losses for each spent moulding sand and its reclaims are presented in Table 2 and in Figure 3.

Reclamation cycles
Reclamation of used sand I

Reclamation of used sand III
Reclamation of used sand IV Fig. 3. Ignition losses of the tested mixtures of spent sands and their reclaims The results presented in Figure 4 indicate that the Na 2 O content on the reclaim grains surfaces decreases when the process intensity increases. This process occurs the most intensely in case of the reclamation of spent sand II, the Na 2 O content decreases slightly less intensely in case of spent sand I, while the smallest decrease of this index was observed in case of the regeneration of spent sand III and IV.  Table 4. They indicate that the best results were obtained for the reclaims originated from sands I -III, while the reclaim obtained from sand IV is characterised by slightly worse parameters.
The last element of the presented investigations constituted the determination of strength properties of moulding sands prepared with the reclaimed materials present in the matrix in the given below amounts. Properly prepared samples of moulding sands were subjected to tensile strength tests R m u , which were determined after 1.5h, 3h and 24h of the storing time. The composition of the applied moulding sands was as follows: For the reclaims I -III from the regeneration process of spent sands I -III: -matrix 100 parts by weight mas., -GEOPOL binder 2.4 parts by weight mas., -PRSTAL activator 0.38 parts by weight mas.. For the reclaim IV from the regeneration process of spent sand IV: -matrix 100 parts by weight mas., -water-glass 145 3.3 parts by weight mas., -MACH 3 activator 0.38 parts by weight mas.. Two kinds of matrices were applied: -Matrix 1 -mixture of the appropriate reclaim (50%) and fresh high-silica sand Biała Góra (50%), -Matrix 2 -mixture of the appropriate reclaim (80%) and fresh high-silica sand Biała Góra (20%). The results of tensile strength tests are listed in Table 5.  Table 5). These investigations indicated that moulding sands I -III achieve high strength values, which allows to apply sands with 80% fraction of the reclaim. It is different in case of sand IV, which matrix was the reclaim from spent moulding sand IV. The safest proportion of the applied mixture of the reclaim-high-silica sand is 50-50.

Conclusions
The performed investigations allow to present the following conclusions:  Out of the applied mixtures the best regeneration susceptibility characterises sands I -III, while the worst sand IV. The vibratory reclamation method can be successfully applied for sands I -III. In case of sand IV the spent sand should be probably preheated at temperatures 140-200 o C before the regeneration process. This heating should increase brittleness of the spent binder coating and facilitate its removal from the matrix surface in the reclamation process.  Reclaims obtained from spent sands I -III, can be successfully applied as components of moulding sands matrices in relation: 80% of a reclaim -20% of a fresh sand, while in case of reclaims prepared from spent moulding sand IV the application of the matrix mixture containing 50% of a reclaim and 50% of a fresh high-silica sand seems to be safer.