|Abstract: ||本論文量測高壓貧油預混氫氣/空氣( = 0.6)球狀紊焰的火焰速度，進一步地探討其可能存在之自我相似性，並與先前研究採用其他燃氣(如甲烷、丙烷、富氫等)之球狀紊焰和本生燈紊焰資料作比較。實驗利用實驗室已建立之高壓雙腔體爆炸設備來進行研究，雙腔體爐分別由一高壓保護腔體(外爐)與一3D高壓十字型燃燒器(內爐)所組成。高壓十字型燃燒器，乃由一大水平圓管與另兩垂直圓管焊接而成(從3個方向看均呈十字型)，在垂直圓管上設有四個釋壓閥使其引燃爆炸時可處於等壓狀態，並在水平圓管兩端各設有一組旋轉風扇與空孔板。當兩風扇反向旋轉時，可於內爐中心區域產生一近似等向性紊流場，其方均根紊流擾動速度(u′)可達8.4 m/s。本實驗使用貧油氫氣( = 0.6)，其Lewis數(Le ≈ 0.58 ≪ 1)，進行一系列不同壓力(p = 1 ~ 5atm)與紊流擾動速度(u′ = 0.8 ~ 4.0m/s)之燃燒實驗，並量測正規化紊流火焰速率[(1/S_L^b)(d/dt)]與火焰紊流雷諾數(ReT,flame = u′/)之關係，其中S_L^b為已燃氣之層流燃燒速度、為火焰平均半徑、t為時間、(≈ SLL)為熱擴散係數、SL為密度校正後之未拉伸層流燃燒速度，而L為層流火焰厚度。我們同時量測正規化紊流燃燒速度ST/SL與紊流強度乘以壓力比值(u′/SL)(p/p0)之關係，其中ST為紊流燃燒速度、p為實驗之初始壓力與p0為一大氣壓。另外，也量測ST/u′與Karlovitz數(Ka)之關係，並與先前本實驗室團隊所得結果，含CH4/Air ( = 0.8, Le ≈ 0.98)、CH4/Air ( = 0.9, Le ≈ 0.997)和C3H8/Air ( = 0.7, Le ≈ 1.62)作比較。結果顯示，本實驗之數據與先前之數據，若進行適當之校正，彼此之間可以吻合，顯示正規化紊流燃燒速度之通式是有可能存在的。|
此外，為了驗證球狀火焰與本生燈火焰在平均傳遞變數¯c=0.5火焰輪廓下，其S_(T,c ̅=0.5)值是否能夠互相吻合，我們以S_(T,c ̅=0.5)/S_L與(u′/SL)(p/p0)以及S_(T,c ̅=0.5)/u′與Ka之關係，分別與Kobayashi et al. (1996、1998、2005)以及Gülder et al. (2000、2009、2014)的本生燈火焰資料作比較。結果顯示，本實驗室之數據分別與Kobayashi et al.以及Gülder et al.之數據，在適當校正後可以吻合，適用於S_(T,c ̅=0.5)/S_L=〖C_1 [(u^′/S_L)(p/p_0)]〗^0.38或S_(T,c ̅=0.5)/u′=C_2 〖Ka〗^(-0.42)之通式，其中C1和C2為實驗常數。但是，值得一提的是，本實驗室團隊先前所獲之合成氣之數據(未報告於本論文)，無法與前述之數據吻合，顯示前述正規化紊流燃燒速度通式並不適用於合成氣燃料，此仍為一待解的問題。
;This thesis measures turbulent flame speeds of expending spherical lean premixed hydrogen/air mixtures (the equivalence ratio = 0.6) under high pressure conditions, further explores possible its self-similar turbulent expending flames, and compares with previous studies using different fuels (such as methane, propane, rich-hydrogen) of both spherical turbulent flames and Bunsen turbulent flames. Experiments are carried out in a high pressure double-chamber explosion facility that is consisted of a huge high-pressure safety vessel (outer chamber) and a 3D high-pressure cruciform burner (inner chamber). The high-pressure 3D cruciform burner is constructed by a large horizontal vessel which is welded by a vertical vessel and a horizontal vessel, forming a cruciform shape when viewed from all three directions. Upon explosion, four pressure release valves installed around the vertical vessel will be opened to release the pressure rise from the inner chamber so that combustion can be kept at constant pressure conditions. Using a pair of the identical frequency-controlled counter-rotating fan and perforated plate equipped at the two ends of the large horizontal vessel, an intense near-isotropic turbulent flow field can be generated in the central uniform region of the cruciform burner, where the root-mean-square turbulent fluctuation velocity (u′) can be up to 8.4 m/s. In this study, lean H2/air mixtures ( = 0.6) with a Lewis number Le ≈ 0.58 ≪ 1 are applied over a wide range of u′ vary from 0.8 m/s to 4.0 m/s. We measure the normalized turbulent flame speed [(1/S_L^b)(d/dt)] as a function of a flame turbulent Reynolds number (ReT,flame = u′/), where S_L^b is the laminar burning velocity at burn side, is the flame mean radius, t is time, (≈ SLL)is thermal diffusivity, SL is the unstretched laminar burning velocity after density correction, and L is the laminar flame thickness. Also, we measure the relationships of the normalized turbulent burning velocity (ST/SL) with (u′/SL)(p/p0), where ST is the turbulent burning velocity, p is the initial pressure of the experiment, and p0 = 1 atm. Moreover, the relationships of ST/u′ with Ka (Karlovitz number) are measured and analyzed with a comparison of previous data, such as CH4/Air ( = 0.8, Le ≈ 0.98), CH4/Air ( = 0.9, Le ≈ 0.997) and C3H8/Air ( = 0.7, Le ≈ 1.62) obtained by our group. It is found the present H2 data can be represented by previous general correlations when appropriate corrections are made.
In order to compare the data between spherical flames and Bunsen flames, we extract ST data at the mean progress variable ¯c=0.5 of the flame contour. It is found that the data of both spherical flames and Bunsen flames [Kobayashi et al. (1996、1998、2005) and Gülder et al. (2000、2009、2014)] can be represented by the relationships of S_(T,c ̅=0.5)/S_L=〖C_1 [(u^′/S_L)(p/p_0)]〗^0.38 or S_(T,c ̅=0.5)/u′=C_2 〖Ka〗^(-0.42), where C1 and C2 are experimental constants. However, we hasten to note that the previous data of syngas obtained by our laboratory team (not shown in this thesis), cannot be represented by the above general correlations. This is still an open question.