一新型垂直式風機性能定性分析

Abstract

本研究係發展一新型垂直式風機，此風機具有升力型及阻力型風機的特性，並可將數個風機葉輪堆疊以得到較大的輸出功率。為瞭解其氣動特性及風場結構，主要利用計算流力的方法做數值模擬，此對應進行並簡易的實驗方法量測，以驗證數值分析的可靠度。數值方法為有限體積法，為處理風場葉片的旋轉，採用多重參考座標系統，並假設在”擬暫態”下，將風機固定在數個預先設定的旋轉角度位置，以非穩態的方式計算，如此可簡化全暫態計算的複雜性。 結果顯示此垂直式風機的扭矩與風機轉速成線性遞減的關係，而此扭矩與風速成二次式關係，並與風機直徑成三次關係式，將這些參數經過無因次化後，可以得到扭矩係數與風機葉尖速度比間呈線性遞減的關係，進而可推導出功率係數與葉尖速度比成二次式關係，並可找出最大功率係數及最佳的葉尖速度比值。上述結果是在固定旋轉位置的擬暫態假設下得到，在全暫態的計算(風機隨時間連續轉動)中，上述無因次參數間的關係仍然適用，兩相比較，顯示此簡化模擬在定性上有相當的可信度，但在定量上兩者有相當程度的差異，而全暫態計算所得的扭矩係數及功率係數高出許多。另外實驗所得數據介於兩者之間，由此可知擬暫態分析因流體慣性被忽略而低估扭力輸出，由於實驗設備過於簡陋，及過程有許多不嚴謹之處，造成實驗值的不準確。

A new type of vertical-axis wind turbine is developed in this project. This new design combines the merits of the lift type and drag type of wind turbines. It also possesses a feature by which several turbine wheels can be stacked together to get higher output power. The method of computational fluid dynamics is mainly used to investigate the flow structure and aerodynamic characteristics, also conducted is the experimental work to validate the numerical simulation. The numerical scheme is based on a finite volume method. The multiple reference frames (MRF) is adopted to tackle the rotation of the turbine wheel. To simplify the complicated problems, the quasi-unsteady state is assumed so that the wind turbine is fixed at a number of specified angular positions and unsteady computations are undertaken. It is obtained from the simulation that the resulted torque is related to the angular speed of the wind turbine in a linear decreasing fashion, a quadratic function of the wind velocity and a cubic function of the diameter of the wind turbine. After non-dimensionalization, the resulting moment coefficient becomes a linearly decreasing function of the tip speed ratio. It can be derived from this relationship to show that the power coefficient is a quadratic function of the tip speed ratio and, furthermore, to find the maximum power coefficient and the optimum tip speed ratio. It is also shown that by using the fully unsteady model, in which the wind turbine continuously rotates in the simulation, the above correlations are still valid. However, comparing with the quasi-unsteady calculations, the resulting moment coefficient and power coefficient are much higher. The experimental results lies in between of the fully unsteady and quasi-unsteady predictions. For the quasi-unsteady on calculations, the effect of inertia is not property accounted for. As for experiments, the experimental rigs are too rough and there are a lot of uncertainties about the measurements, cause the experimental inaccuracy.