An architecture for a multi-disciplinary integrated turbine rotor system optimizer

The design of gas turbine engine parts such as turbine rotors involves not only designing multiple components but also the incorporation of the knowledge of multiple disciplines and applications to create the ideal rotor for the design conditions. Like a symphony, each discipline has to play a part and work together with other disciplines to create an effective and efficient end product. Traditionally, the design of these components has been separated into the pre-detailed and the detailed design phases. Unfortunately, during the pre-detailed stage of the design, engineers are not afforded enough time to get the perfect rotor thus a delicate balance must then be struck between the fidelity of the results and the time taken to achieve them. In the detailed design however, more emphasis is placed on accuracy of analysis above all else and a final design is achieved. This traditional way of designing can be improved because a suboptimal concept design created in the pre-detailed design step is difficult to correct in the detailed design and attempts to moderate its impact usually come at a steep cost. The use of Multidisciplinary Design Optimization concepts at the pre-detailed design phase (Pre-detailed MDO or PMDO) improves the process by bringing high fidelity knowledge at the pre-detailed stage, allowing for better concepts exiting pre-detailed design. As part of Pratt & Whitney Canada (P&WC) program on Propulsion System Integration and Optimization (PSIO) the author applied concepts of PMDO to create an architecture for the design and optimization of engine parts as well as using that architecture to create an integrated and automated rotor design system (iRSO); a design module that integrated the design of the platform, fixing, disc, and cover plate based on the thermal and mechanical stresses as well as the airfoil based on aerodynamic, thermal and mechanical stresses, and cooling requirements. The architecture allowed more knowledge to be injected into the design process at the early stage of design, allowing the designer to rapidly synthesize a complete rotor and evaluate its attributes over a range of alternative designs. Starting from a set of design conditions, the new architecture allows an engineer to explore many design options, ensuring that a large design space is investigated at the early stages of development with a higher degree of fidelity in all disciplines. This research, due in some part to the fact that it is done in a real world setting (industry), if conducted using the traditional direct research (DR), would have faced obstacle such as: definition of objectives that may not be equally understood and appreciated by the researchers and the end users, the difference in culture and research approaches between the organization and the research institution as well as the importance of academic rigor vs the responsiveness to user requests. These problems were alleviated through the use of Action Design Research (ADR) as it joins the best aspects of traditional DR and the organizational focused Action Research (AR). ADR is helpful as a methodological framework as it recognizes the role of organizational behaviour in shaping the objectives of the research. Research in information systems / technologies (IS/IT) such as this must achieve dual objectives: to build knowledge or a theoretical contribution to the disciple as well as to assist in solving a real world problem in real world settings. Through the creation of an architecture for the design and optimization of engine parts, new methodological knowledge was created on how to best tackle a multidisciplinary design and create optimization capable tools and processes. As an example of the possibilities the system created holds, an optimization of an airfoil was done. The optimization integrated the structural analysis of a 3D airfoil and aerodynamics through CFD. It was able to optimize with the double objective of increasing efficiency and reducing mass and the constraints of peak stress at a certain percentage span range, as per P&WC best practices. The optimization showed great promise in that it did not only give a solution that was the best of both objectives but also provided insight on the degree of dependencies of certain key parameters. For the particular case, it was able to show that the mass could be lowered while increasing efficiencies. It allowed ‘the engineering design process [to] move forward by asking “what if” questions and using the answers to make design changes’ while reducing the time taken and the non-value added work load on the engineer. This also solved the real world problem of mitigating the risk of the traditional two phased approach to design of gas turbine engines. Similar and sometimes higher levels of fidelity were achieved at 20% of the time when using the artifact. Possibility of human error is also mitigated as all the manual transfer of information between the disciplines is now handled autonomously in the system.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.