| Abstract: | 隨著時代的改變，各式各樣的飛機被發展出來，像是翼胴合一飛機，雖然它的概念很早就被提出，不過最近開始被更加重視了。翼胴合一飛機的機身和機翼是合為一體的，因此可以提供飛機較高的升力。在本研究之前，學長已經做過了許多關於翼胴合一飛機的空氣動力分析，不過尚未考慮到發動機部分。我們使用現有之飛機外形檔，再利用Pro/E 創造發動機並加在飛機的外形上，接著使用ANSYS 網格技術建造網格，並使用FLUENT 模擬出翼胴合一飛機在巡航時，不同派龍高度和攻角對其升力係數、阻力係數和升阻比的影響。有了這些基本數據後，就可以利用最佳化工具找到最佳的派龍高度及攻角。本文使用之最佳化工具是替代方法，替代方法是由許多不同的模型所組成，其中之一是Kriging model，我們正是用它來尋找最佳答案。為了要知道最佳化的使用方式正不正確，在翼胴合一飛機加發動機尋找最佳派龍高度及攻角的研究前，先預測了翼胴合一飛機在沒有發動機的情況下，使用Kriging model 找飛機巡航時的最佳攻角，在本研究中，我們只考慮升阻比，意即利用已知攻角的升阻比來預測最大升阻比的攻角，當預測的結果符合期望後，才開始飛機加發動機後的研究。在加發動機的案例中，吾人使用了兩種不同的方式去尋找最佳值，一種是先找派龍的最佳高度再找飛機的最佳攻角，另一種是直接預測出兩個值的最佳答案，結果顯示一次預測兩個值的案件比另一個有效率，且有更佳的結果，這也顯示出Kriging model 的優點，並證明吾人已經可以同時預測出兩個以上的最佳值。
In pace with the modern airplane development motivated by fuel efficiency and environmental conservation, many different aircraft configurations and design concepts are created in last two decades to accommodate these challenges. Blended Wing Body aircrafts (BWB) are created for the same reason, and remains to be one of the most promising flight vehicle concepts for future generations to come. But this plane are seldom seen, people are still study its aerodynamic analysis at the beginning stage, moreover the aerodynamic performance of BWB with its engines on. In this research, based on previous works at Tamkang, we construct geometry model first and now this BWB is with engine added on. Then, software ANSYS is implemented to generate different types of mesh, such as structured, unstructured, and hybrid grids. The flow solver routine with proper turbulence model selection is first tested on our previous UAV, M-6 wing, and BWB configurations, and then this simulation routine is also extended to the incompressible take-off speed and 0.85 Mach cruise conditions.
After that, we select the surrogate model to find the BWB optimum angle of attack (AOA) and vertical height for engine positions. The surrogate model is a relatively new method for optimum engineering design, which is especially suited for CFD optimization computation and contains several different modules, and the model we select is the Kriging model. Without spend too much effort on the time consuming CFD simulation for every different AOA and engine positions, it allows us to find the best possible configuration conditions from a mere of about ten properly chosen design of experiment (DOE) cases. This model is verified by first predicting the best AOA value for BWB without engines, and a normalized optimization parameter or objective function is created, which composed of both the lift and drag coefficients. Thus we can predict the optimum AOA for BWB and its engine vertical positions. After the predicting value is achieved, new engine position geometry will be generated according to the surrogate model prediction. Results show that the close agreement between our Kriging model prediction and CFD computation represent a first triumph in the surrogate model implementation, and this could imply tremendous saving in future aerodynamic simulation in the airplane design phases. |
