A new Actuator Line (ACL) method was implemented into the TAU-Code for use in CFD simulations of propellers and applications thereof. Compared to the existing ACL implementation, the new method uses a simplified but computationally more efficient approach: for the angle of attack calculation a temporal average of the inflow is used instead of a spatial average, thus eliminating additional core communication. As a result, the ability to capture unsteady inflow is limited, but installation effects which influence the unsteady propeller response can be captured. In order to validate the new method, simulations were carried out for both positive and negative thrust conditions. The integral coefficients are closely aligned to the ACD results and match the fully resolved counterpart to within about 6%, the larger deviations being found for the negative thrust conditions. The unsteady slipstream matches that of the fully resolved simulations well, in both positive and negative thrust conditions. The areas of locally accelerated or decelerated flow, the blade tip vortices and the blade wake are all predicted with similar shape and intensity. Compared to the existing ACL implementation, the computation cost could be slightly reduced (by less than 2%) with similar accuracy of the results.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

A Simplified Actuator Line Implementation for the DLR TAU-Code

  • Mário Firnhaber Beckers,
  • Michael Schollenberger,
  • Thorsten Lutz

摘要

A new Actuator Line (ACL) method was implemented into the TAU-Code for use in CFD simulations of propellers and applications thereof. Compared to the existing ACL implementation, the new method uses a simplified but computationally more efficient approach: for the angle of attack calculation a temporal average of the inflow is used instead of a spatial average, thus eliminating additional core communication. As a result, the ability to capture unsteady inflow is limited, but installation effects which influence the unsteady propeller response can be captured. In order to validate the new method, simulations were carried out for both positive and negative thrust conditions. The integral coefficients are closely aligned to the ACD results and match the fully resolved counterpart to within about 6%, the larger deviations being found for the negative thrust conditions. The unsteady slipstream matches that of the fully resolved simulations well, in both positive and negative thrust conditions. The areas of locally accelerated or decelerated flow, the blade tip vortices and the blade wake are all predicted with similar shape and intensity. Compared to the existing ACL implementation, the computation cost could be slightly reduced (by less than 2%) with similar accuracy of the results.