An Empirical Prediction Method for Secondary Losses in Turbines: Part II — A New Secondary Loss Correlation

Abstract

A new empirical prediction method for design and off-design secondary losses in turbines has been developed. The empirical prediction method is based on a new loss breakdown scheme, and as discussed in Part I, the secondary loss definition in this new scheme differs from that in the conventional one. Therefore, a new secondary loss correlation for design and off-design incidence values has been developed. It is based on a database of linear cascade measurements from the present authors’ experiments (Benner [1]) as well as cases available in the open literature. The new correlation is based on correlating parameters that are similar to those used in existing correlations. This paper also focusses on providing physical insights into the relationship between these parameters and the loss generation mechanisms in the endwall region. To demonstrate the improvements achieved with the new prediction method, the measured cascade data are compared to predictions from the most recent design and off-design secondary loss correlations (Kacker and Okapuu [2], Moustapha et al. [3] using the conventional loss breakdown. The Kacker & Okapuu correlation is based on rotating-rig and engine data, and a scaling factor is needed to make their correlation predictions apply to the linear cascade environment. This suggests that there are additional and significant losses in the engine that are not present in the linear cascade environment.