NEW CORRELATIONS FOR HIGH-PRESSURE TURBINE HOUSING AND SHROUD SEGMENTS

At the preliminary design phase of a gas turbine, time is crucial in capturing new business opportunities. In order to minimize the design time, the concept of Preliminary Multi Disciplinary Optimization (PMDO) was used to create parametric models, geometry and cooling flow correlations towards a new design process for turbine housing and shroud segments. First, dedicated parametric models were created because of their reusability and versatility. Their ease of use compared to non-parameterized models allows more design iterations and reduces set up and design time. A user interface was developed to interact with the parametric models and improve the design time. Second, geometry correlations were created to minimize the number of parameters used in turbine housing and shroud segment design. Third, a correlation study was conducted to minimize the number of engine parameters required in cooling flow predictions. The parametric models, the geometry correlations, and the user interface resulted in a time saving of 50% and an increase in accuracy of 56% compared to the existing design system. For the cooling flow correlations, the number of engine parameters was reduced by a factor of 6 to create a simplified prediction model and hence a faster shroud segment selection process.