EDDY

The beginnings of lost-wax casting started more than 6,000 years ago. Safran Tech's advanced turbine blades research platform (Safran PFX,), is one of the few global players to draw on this age-old heritage to manufacture very high-tech objects, producing single-crystal blades with complex shapes, particularly tight geometric tolerances and operating at extreme temperatures. Improving the performance of these blades is a key step towards optimizing turbomachinery efficiency, and thus maintaining French sovereignty in terms of fighter aircrafts. Safran Tech's casting research strategy identifies the “core shifting” issue as a highly strategic one. Indeed, through analysis of dimensional defects observed during the production of hollow blades, PFX become convinced that, during casting, the forces generated by the molten metal on the foundry core (which, provides the cooling circuit for the blade) are sufficient to displace or even to deform the latter, and thus alter the dimensions of the cooling circuit to the point of generating scrap. With the increasing complexity of cooling circuits, dictated by the rising operating temperatures of civil and military engines, future blades will be more sensitive to this phenomenon. It is therefore important for Safran to control the occurrence of this defect. To achieve this, PFX relies on numerical simulation. The platform develops multiphysics tools that are used during the blade pre-sizing phase, with which Safran imagines more robust blade designs, and with which the group hopes to shorten its development cycles. However, the conditions under which shifting occurs are poorly understood. In particular, the hierarchy of physical effects responsible for the hydrodynamic forces exerted by the molten metal on the core has not been established. This fundamental misunderstanding stems from the fact that the possibilities of instrumenting the lost-wax casting process are drastically reduced, due to its very severe implementation conditions (T > 1500°C, confined space, presence of a magnetic field). From an academic point of view, this configuration is far from being a solved problem. On the one hand, it can be likened to a problem involving a deformable solid (the core) penetrating a liquid (molten alloy) whose density is much higher than its own, but it must be added that this solid is held in place by a non-perfect embedding and that the flow around the core takes place in a confined space. On the other hand, the blades are manufactured in clusters of complex geometry featuring numerous symmetries, but whose symmetry of boundary conditions is questionable due to their gravitational feeding by a free jet. Without detailed experimental observation, neither the modeling strategy nor the robustness of the numerical tools developed can be fully established. For PFX, the implementation of development cycles accelerated by digital simulation therefore seems compromised. The EDDY project will take this understanding a step further, using experimental setups that are representative of the process issues at stake. At LMFA, an experimental model of core shifting will be developed, using a liquid metal alloy at room temperature, with complete hydrodynamic similarity to the industrial application. It must be equipped with in situ instrumentation, the development of which will be part of the project's challenges. At IRDL, water-based devices will be used to support this approach with optical measurements on similar cases, and to complement it by addressing the issues of pouring and non-perfectly rigid embedding. Through this strategy, the EDDY project will lead to a detailed understanding of the phenomenology of the fluid-structure interaction problem underlying core offset, identify first-order phenomena, and produce reference observables in academic configuration for validation of the numerical models developed at PFX.