The Control of Solidification Kinetics of the Vacuum-cast Thin-wall Nickel-based Superalloys by Changing the Geometrical Characteristics of the Ceramic Mold

This paper provides an analysis of experimental research and results of investment casting process. Temperature field in a ceramic mold is one of the problems during numerical simulation. Reducing the costs of production in precision casting involves the reduction of scraps, which is one of the fundamental problems of the foundry industry. Reducing these costs is associated with optimization of precision casting technology of aircraft engines critical parts, including control of the solidification front in thin-walled castings of nickel super alloys cast in a vacuum. It is achieved by changing the geometrical characteristics of the ceramic mold. The results of the tests were used to optimize the industrial production of aircraft components in Precision Foundry of WSK Rzeszów. Temperature distribution gained in the conducted tests allowed verification and optimization of computer simulations.


Introduction
The important technological problems occurring in the production of nickel-based superalloy casts include: developing conditions for pouring and solidification of casting and, what's involved, determining the effect of all factors on the orientation of the liquid metal solidification front and the cooling of it [1][2][3]. The problem of defects appearance in precision castings, despite the use of advanced technology, is not yet solved. There is no comprehensive study and evaluation of changes in temperature field in the cross-section of the ceramic mold during annealing, pouring and solidification processes of molten metal in the area of thin-walled castings, such as the trailing edge of nickel-based superalloy blades cast in vacuum. These elements are used in lowpressure turbine aircraft engines. Lack of an access to knowledge about the distribution of temperature in the cross section of the ceramic mold during annealing, casting, solidification and cooling of the casting makes it impossible to determine the causes of defects, especially in thin-walled castings. This is due to a reserve of technological knowledge by global foundries which produce components for hot engines. In many precise foundries it is necessary to address these issues in case of changing types of castings as well as in the introduction of new types of materials [4][5][6][7][8].
The results of researches will allow to determine a precise effect of the ceramic layer thickness (the number of layers made) on the ceramic mold temperature distribution during annealing process as well as during the pre-filling process and solidification of nickel-based superalloy super thin-walled cast components. These data are necessary to determine the impact that the change of geometrical characteristics has on the temperature distribution of the ceramic mold. Hence, also on the direction of the solidification front. Obtained results will also be helpful to determine the correct annealing time of the ceramic mold. This knowledge will significantly contribute to the labour reduction, thus, to a substantial reduction of the cost of manufacturing of casting, not just airlines ones, but also other casting machine components manufactured by precise casting.

Methodology
The molds were performed at the precise foundry WSK PZL Rzeszów. Methodology for the molds execution was similar as in the case of production ones keeping all the process parameters so that the molds were made as similar as possible to the production ones in the means of physico-chemical properties. Two model sets for casting wax plates were designed and built with dimensions shown in (Fig. 1). The shape of this cast provides enough flat surface for the precise placement of thermocouples. Four wax models with H-type gating system were combined into a model set. In order to stabilize and strengthen the model set and for better wax melting, 4 vents were applied in a further stage (Fig. 2a).

Fig. 2. a) Wax model set b) ceramic mold -after first layer application
Two wax sets formed the basis for the implementation of multilayer ceramic molds. Ceramic molds were made with the use of alcohol coating. For the first layer CoAl 2 O 4 + ZrSiO 4 were used (Fig.2b). This layer is characterized by good thermal conductivity and durability, and prevents the reaction of molten metal with a ceramic material. Afterwards, the next layers were applied with placing thermocouples: Thermocouple #1 after 2nd layer, thermocouple # 2 after 3rd layer, thermocouple #3 after 4th layer and thermocouple #4 after 7th layers. Two molds fabricated with 9 and 11 layers coated. B-type thermocouples with 0.8 mm wire diameter were used for the measurement. When all the sets of ceramic layers were ready, the assembly was dried for 48 hours at a suitable temperature and humidity. After drying the ceramic mold was placed in high-pressure autoclave to fuse out the wax. Pre-annealing was performed at 600°C for 2 h in order to remove the wax model residue. Then, insulating of the mold and pouring were performed (Fig. 4).

Results and discussion
After the experiment, the temperature change on the crosssection of the mold during annealing, pouring, solidification and cooling was received. Depending on the quantity of the ceramic layers in the mold, temperature distribution was as follows (  The next step was to analyse the distribution of temperature on the cross-section of the ceramic mold during annealing, pouring of molten metal, solidification and cooling (Fig. 9). The aim of the study was:

Summary
As a result of the experiment, the temperature distribution curves of ceramic molds with 9 and 11 layers (shells) were made. These curves allowed to determine the effect of selected geometric features changes on the temperature distribution in the studied ceramic mold system, during annealing, pouring, solidification and cooling processes.
In the molds with 9 ceramic layers after 80 minutes of heating, temperature had been stabilized, however, it had not reached the ambient temperature (1250 o C) throughout the section. In the molds with 11 ceramic layers after 110 minutes of heating, temperature had been stabilized, however, it had not reached the ambient temperature (1250 o C) throughout the section. The temperature difference in the molds while pouring was slightly higher for the mold with 9 ceramic layers. The maximum temperature recorded by the measuring system during pouring was 1305 o C for 11 layers and 1322.5 o C for 9 layers. For the mold with 9 layers a much larger temperature gradient was observed. Mold with 9 layers cools down much faster by handing over the accumulated heat to the surroundings. The results of the tests were used to optimize the industrial production of aircraft components in Precision Foundry of WSK Rzeszów. Temperature distribution gained in the conducted tests allowed verification and optimization of computer simulations. The results will also be used when designing the ceramic molds for new implemented products.