Investigation of physical and mechanical properties of mixed gypsum material for slip casting molds

. The issues of improvement and optimization of the properties and structure of porous molds used in the ceramic industry are still relevant. In this work, plaster materials based on the mix of α-and β-calcium sulfate hemihydrate (α-CaSO 4 · 0.5H 2 O, β-CaSO 4 · 0.5H 2 O) were received and their properties were studied. Based on the test results of the strength characteristics, water adsorption and kinetics of formation process, the optimum composition of gypsum binder was proposed for application in technical ceramics production.


Introduction
In ceramic industry, there are many methods of manufacturing different kind of products. However, the most common is slip casting into porous molds. This method is based on the formation of a solid ceramic structure from the suspension due to partial dehydration by porous mold. One of the advantages of slip casting is a possibility to obtain either small or oversized products with complex configuration.
The most suitable material for slip casting molds is a gypsum binder due to low cost, good reflection of model configuration, ability to absorb water from the slurry and adequate strength [1][2][3].
Gypsum binder as modeling material can be presented in the form of α-or β-calcium sulfate hemihydrate (α-H, β-H), sometimes they can be used both [4,5]. The difference between their characteristics explained by the method of obtaining: α-H formed by hydrothermally heating of CaSO4·2H2O, β-Hby dry heating of CaSO4·2H2O [6,7]. The choice of hemihydrate form depends on technology requirements. It is known that gypsum as modeling material must have strong mechanical strength and high water absorption, what is difficult to achieve. Using for modeling manufacture various modifications of hemihydrate it can be possible to control the structure and properties of plaster molds.
There is a lot of research aimed at improving the properties of building materials, but there are not enough reports on developments in the modeling industry. The fact is that the quality of plaster molds affects the quality of products. Therefore, modern high-performance ceramic and porcelain-faience production requires the use of high-quality equipment and technologies, including plaster molds.
Despite all advantages of plaster molds, they also have disadvantages, the main of which is a low service life. The simple way to achieve a high percentage of material's water adsorption is to increase the water-to-plaster ratio. As a result, the mechanical strength of the material decreases, the mold surface wears out quickly and requires frequent replacement. In this regard, improving the quality characteristics of porous plaster molds is an urgent issue.
Therefore, the main objective of this research is to study properties of mixed gypsum binders based on α-and β-hemihydrate and choose the optimal composition for modeling material used in manufacturing of quartz ceramic products for technical purposes.

Raw materials
α-calcium sulfate hemihydrate (α-CaSO4·0.5H2O) has a water requirement at standard consistency of 39 % by weight, its compressive strength two hours after mixing is 16.0 MPa. β-calcium sulfate hemihydrate (β-CaSO4·0.5H2O) has a water requirement at standard consistency of 60 % by weight, its compressive strength two hours after mixing is 5.0 MPa.
Compositions of binders were received by blending them before adding to water.

Water requirement and setting time
Water requirement at standard consistency and the setting times of plasters were determined according to the test procedure of Russian Standard GOST 23789-2018 [8].

Mechanical properties (flexural and compressive strength)
The rectangular specimens were made by pouring the pastes into three iron molds with the size of 40 mm × 40 mm × 160 mm. Flexural and compressive strength test was carried out 2 hrs after mixing the plaster with water according to the procedure of Russian Standard GOST 23789-2018 [8].

Water absorption
To determine water absorption, we used half samples (40 mm × 40 mm × 80 mm) rested after strength testing. They were dried at 50 ± 5°C to a constant weight and then cooled. The dried specimens were weighed, filled with distilled water to half the height for 2 hours, and then completely filled with water and kept in this form for another 2 hours. After the immersion period, the samples were removed from the water, wiped with a damp cloth and weighed.
Water absorption was determined according to the formula (1) specified in Russian Standard GOST 23789-2018: where M1 is a wet weight of the specimen and M is a dry weight of the specimen.

Hydration process
Temperature change during the hydration process of calcium sulfate hemihydrate was measured according to the procedure described by Zhao et al. [9]. Semi-adiabatic system was performed as a homemade container which is laid over by polystyrene foam.

Kinetics of formation process
To study kinetics of ceramic product formation experimental molds were made. Schematic representation of the device is given in Fig. 1.   Fig. 1. Schematic representation of the device to study formation kinetics: 1pouring capacity; 2sample forming in cavity, 3moisture adsorbing layer of the mold; 4gypsum mold.
The process of sample formation was controlled through the sight glass covering one side of mold. Changes in thickness of layers (2) and (3) over time were recorded using graph paper. Fig. 2 shows the evolution of strength metrics depends on the mixture composition. It is clear that an increase for β-modification of CaSO4·0.5H2O in plaster mixture leads to a decrease in the mechanical strength of hydrated material. Therefore, if modeling plaster must be strong enough (compressive strength more than 10 MPa), then the mixed binder should not include β-H in an amount of more than 60 % by weight.

Flexural strength
Compressive strength Water adsorption In addition, another important property in modeling industry is the setting time. It is fact, that too fast setting of gypsum paste undesirable because of negative impact on the technological parameters of mold manufacturing. For example, fast setting makes difficult pouring of gypsum paste into the big sized frame of the form and it leads to differences in its properties or defects like cracks or shells. Therefore, when choosing a modeling plaster it is important to study this characteristic. Fig. 3 shows the change in setting time at a standard consistency depending on the composition of the mixture. Here, also, because of characteristics of β-hemihydrate modification, the paste sets faster with an increase in its amount. In addition, as shown in Fig.  4, hardening process of gypsum paste starts earlier a large amount of β-H is present in the composition. This fact is explained by higher water requirement of β-H, which consists of flaky particles made up of small crystals and has a high specific surface area [10,11].
Given this, for a successful pour operation it is advisable to use a gypsum composition with a mass fraction of β-hemihydrate of not more than 40%.  The results of determination of water absorption are shown in Fig. 2. Change in water adsorption of samples with different ratio of α-H and β-H made at standard consistency shows that β-H provides higher water adsorption.
The graphs (Fig. 2) shows, that the presence in a gypsum binder the β-modification of CaSO4·0.5H2O in the amount less than 40 % by weight leads to a sharp decrease the water adsorption. A composition with 40 % by weight of β-hemihydrate provides a satisfactory level of both characteristics of strength and water adsorption.
In ceramics industry an important parameter of products casting is a rate of formation. Therefore, it was necessary to make a mold from the chosen gypsum composition and to study the kinetics of formation. Gypsum molds of different composition were received. For testing, aqueous slurry based on quartz glass was used. Test results are given in Fig. 5. It is established that within the first 30-40 min of formation the rate of capillary water adsorption of molds is high, and then its gradual delay is observed. It is caused by formation of a dense layer of ceramic material in initial minutes on a plaster mold surface, which slows down the adsorption from the slurry.
The experimental plaster mold made from the composition of α-and β-hemihydrates due to β-modification existence has more developed porous structure in comparison with mold from α-hemihydrate. The rate of ceramic material formation in the initial period of time increases by 15-30 % here.
It should be noted that the rate of capillary water adsorption of mold made from the gypsum composition is more equable in time, which is a positive result of experiment. Equable suction of water from the slurry provides equable rate of product formation.
The next step of this research work is to improve the strength characteristic of mixed plaster by addition of plasticizers. Slip casting molds made from modified mixed plaster are suggested to have a high level of strength and service life.

Conclusions
The quality and properties of porous molds are directly depends on the composition and properties of the gypsum binder used. Each modification in the composition of a binder has its own characteristics and affects its properties entirely. In this work the properties of gypsum binders consists of the α-and β-calcium sulfate hemihydrate in various mass ratio were investigated. In case of applying a mixed plaster for casting molds it seems possible to use advantageous properties of each phase: α-H provides a high mechanical strength and β-H gives a good level of water adsorption. In relation to the technology of manufacturing oversized molds the optimal compound of gypsum binder is suggested to consist 60 % by weight of α-hemihydrate and 40 % by weight of β-hemihydrate. This composition instead of monophase binder makes it possible to achieve the two properties of high strength and water adsorption of material together. Moreover, it is possible to improve the strength characteristics of blended material by using a different kind of additives, for example, superplasticizers.