| 摘要: | 本論文實作設計一氫燃燒器，可安全處理燃料電池之陽極尾(氫)氣。我們先使用貧油甲烷預混燃氣，來測試預混噴流燃燒器之性能，以找出燃燒器最佳操作條件，再進行純氫和空氣預混紊流燃燒實驗。在燃燒器噴嘴上游處，設置一弱噴流漩渦產生器(weak jet-swirl generator)，由四支20o向上斜角設計之切邊小噴嘴所構成。可在原燃燒器出口下游處，形成一均勻擴張流場(diverging flow field)。除了可達到穩焰效果外，並將原本生燈拋物線狀之火焰拉伸成一底部平坦之碗狀火焰，因碗狀預混火焰可穩定於其底部之弱噴流漩渦擴張流場，是故此碗狀火焰不易被吹熄或產生回火。實驗探討不同空孔板固體率(solidity ratio, SR = 36%和64%)、噴嘴直徑(Dj = 15 mm和25 mm)、噴嘴長度(Lj = 1~3 Dj)、出口環邊斜角之有無、漩渦參數(swirl number, S ≡方位角方向與軸向之角動量流率比)等重要參數，對燃燒器穩定操作範圍(避免吹熄或回火)之影響，並量測不同燃料熱供給(thermal input)對廢氣排放的影響。使用高速(5000 張/秒)雷射斷層攝影術，擷取碗狀預混漩渦噴流火焰之動態時序資料，並統計平均其中等時序間隔之1000張瞬時二值化後之碗狀紊焰面，以獲得火焰之平均傳遞變數(mean reaction variable, )，其中c = 0為未燃反應物而c = 1為已 燃之生成物。運用高速質點影像測速技術，量測弱漩渦噴流場之連續瞬時速度分佈，以獲得相對應之全場均方根紊流強度(u’)與平均速度( )資訊，進而評估在不同u’值之火焰紊流燃燒速度(turbulent burning velocities, ST)。結果顯示，SR值會影響u’，即u’會隨SR增加而增加，以SR = 64 %可產生u’ ≈ 0.2 為最佳。當Dj = 25 mm時，使用Lj = 2Dj加上出口環邊45o斜角(tapered rim)的噴嘴，有最佳之穩定燃燒範圍，且在當量比? = 0.9和噴嘴出口流量Qj = 53 ~ 75 L/min時，[NOx]會隨Qj增加而略為增加，但均小於13 ppm(以15% O2為基準)。有關氫燃燒方面，目前可安全處理? = 0.3 ~ 0.6之大流量氫氣(Qf = 10 ~ 20 L/min)，在此範圍並無量測到任何NOx。有關ST量測方面，採用類似Bédat & Cheng (1995)的分析方法，選取 = 0.05位置估算ST與u’值，發現此估算方法所得之ST結果有甚大的誤差及不確定性，因為ST深受所選位置及紊焰震盪之影響，有高達1.7倍的誤差。是故，此弱漩渦燃燒器被證實具有極低[NOx]排放之優點，但若將其運用於估算ST值，則必須小心看待所得的結果。 This thesis studies experimentally to design a hydrogen burner that can be used in burning the anodic offgas (H2) ejected from the fuel cell system. We first use lean premixed methane/air mixtures as a fuel to test the performance of the round burner and to find its optimum operation conditions. Thus, pure H2 premixed turbulent combustion experiments can be safely conducted. Four tangential small jets inclined at 20o are equally positioned around the bottom of the nozzle of the burner for generating weak jet-swirl flow field. So a diverging flow field in the downstream of the burner’s nozzle can be formed. Such diverging flow field not only can stretch the original parabolic Bunsen flame(without swirl) to the bowl-like flat flame but also it can stabilize the flame front. Focuses are on some the effects that may influence the limit of the stable operation (flashback and blowoff limits), including solidity ratio (SR = 36% and 64%) of perforated plates, the nozzle diameter (Dj = 15 mm and 25 mm), nozzle length (Lj = 1~3 Dj), the exit rim with 45o tapered or without, and the swirl number (S ≡ the ratio of the azimuthal to the axial momentum flux). Emissions from different thermal inputs are also measured. The bowl like turbulent flame front images obtained via the high-speed (5000 frames/s) laser tomography are processed and averaged to extract the mean reaction variable ( ), where c = 0 and c = 1 are reactants and products, respectively. Furthermore, we apply high-speed particle image velocimetry to measure the time evolution of corresponding instantaneous velocity fields, so that we can obtain root-mean-square turbulent intensities (u’) and mean velocities () of the whole flow field and thus turbulent burning velocities (ST) as a function of u’ may be estimated. Results show that SR affect values of u’ for which u’ increases with increasing SR. The optimal stable combustion range is found when a nozzle with Dj = 25 mm, Lj = 50 mm hvaing 45o tapered rim is used. At ? = 0.9, emissions of [NOx] are found to be very small, all smaller than 13 ppm (corrected to 15% O2). The higher the mean volume flow rate of the nozzle (Qj) ranging from 53 to 75 L/min, the more [NOx] emissions from 2 ppm to 13 ppm. Concerning the hydrogen combustion, we have tested and found the flashback and the blow off limits of the burner when it is operated at ? = 0.3 ~ 0.6 corresponding to H2 fuel flow rate Qf = 10 ~ 20 L/min; no measurable [NOx] emissions are found. About ST measurements, we apply the method proposed by Bédat & Cheng (1995) by choosing = 0.05 for the estimate of values ST and u’. We found that this method in estimating ST has great errors and uncertainties. Finally, it is concluded that this weak-swirl lean premixed jet burner has the advantage of very low [NOx] emissions, but it’s estimated ST data need to view with great cautions. |
