1. Introduction

Portable power source with higher energy density has been the initial target of intense investigations on micro combustors. Recent studies on combustion phenomena in micro scale lead to the findings that both wall thermal and chemical boundary conditions play important roles in flame behaviors.

Miesse et al. [1] and Kim et al. [2] investigated the effect of wall material/surface property on quenching distance in parallel plates with external heating. They found that inert material and surface are effective against chemical quenching, and chemical quenching instead of thermal quenching dominates for high wall temperature. However, there are several suspicious issues in their measurements. (1) Bulk material is used, and the wall temperature is not directly measured. (2) The wall surface of their sample plates is rough. Radicals are likely to diffuse into the valleys on the surface, thus the combustion process can be changed by the radical destructions. (3) The thermal conductivity of their sample plates is different for different wall materials, thus the wall thermal condition is changed when wall materials are different.

Several research efforts [1, 3] have been made on flame behavior in millimeter-scale tubes/channels. An interesting phenomenon, flame oscillation, is observed, which is a periodic process of flame ignition and quenching. However, most of the previous investigations are phenomenological, and are carried out in a limited parameter space. The lack of optical access in opaque burners and the curvature of tube burners make the optical measurements of the flame difficult. Systematic investigation with more quantitative information of the flame shall provide more profound discussions.

The present study provides experimental investigations of wall thermal and chemical effects on steady and oscillating flames in micro channels. Major contents include quenching distance measurement for the investigation of wall thermal and chemical effects, systematic investigation of flame behavior in micro channels, and time-resolved/phase-locked optical measurements of oscillating flame.

2. Quenching distance in parallel plates

To examine the wall chemical effects, quenching distance is measured with flame in parallel quartz plates coated with ~100nm-thick metallic films. By doing so, (1) wall chemical effects can be checked without changing the thermal wall boundary condition; (2) both the quartz plate and metallic coated plate have very smooth surface, so that the combustion process is not changed by the surface roughness. The plates are externally heated with IR lamp heaters. The wall temperature is monitored by thermocouple plugged in the plate, and carefully controlled. The measurement point is only 1mm away from the test surface to minimize the difference from exact surface temperature.

The quenching distance measured for quartz plates decreases with increasing wall temperature, which agrees with the conventional thermal theory, and quantitatively agrees well with results of Miesse et al., showing that the thermal effect is dominant. On the other hand, quenching distance for metallic surface is slightly larger than that for quartz plates, but also decreases with increasing wall temperature. Therefore, weak chemical effect exists for both low and high wall temperature, but thermal effect is dominant in the temperature range investigated. It has been demonstrated that flame can propagate into 0.6 mm gap, and it is still possible in thinner channels as long as the wall heat loss is compensated.

3. Flame in micro combustors

Based on the quenching distance results in parallel plates, quartz is chosen as the construction material of micro combustors. Plannar quartz combustors having channel heights of 0.7/1.0/1.5 mm have been developed for quantitative investigations of steady and oscillating flames. The present combustor enables optical access for the investigation of quenching mechanisms in micro channels with laser imaging techniques. Microscale steady and oscillating flames are confirmed through CH*/OH* chemiluminescence. Premixed CH4/air flame is established in the combustion chamber by precise control of heat input to the wall, and is examined for channel height of 0.7, 1.0 and 1.5 mm.

Both steady flame and oscillating flame are observed, and the oscillating flame is in the fuel-rich side of the extinction limits map. Oscillating flame ignites at the exit where the fuel/air mixture is diluted by the surrounding air into a flammable concentration. There is a narrow quenching region between the steady flame and oscillating flame regions, where strong hysteresis exists due to the heat transfer from the flame to the wall. It has been found that the extinction limits region becomes narrower for thinner channel, and broader for higher wall temperature, which shows the wall thermal effect. Quenching distance is dependant on the wall temperature, and can be smaller than 0.7 mm for wall temperature over 800℃ in planar quartz channels, which shows good agreement with the measurement in parallel plate configuration. Flame structure in the micro combustor is investigated with scanning micro OH-PLIF imaging techniques. OH-PLIF images of micro flame show a concave-shape flame front, which is the result of the non-uniformity of wall-temperature distribution. Flame temperature is measured with SiO2-coated thermocouple, and higher flame temperature is found for higher wall temperature. OH 2-line PLIF method for measurement of flame temperature is applied to the steady flame, and the result agrees well with the thermocouple measurement.

4. Flame oscillation

Detailed process of flame oscillation is captured with time-resolved chemiluminescence imaging technique. Chemiluminescence images of an oscillating flame reveal that this is a periodic process of flame ignition at the exit, and propagation upstream, and then quenching due to heat loss to the wall, and finally recharge to the exit for the start of another cycle. A phenomenological formula has been derived for the estimation of oscillation frequency. Oscillation frequency measured in the present experiment is in the range of 30-500 Hz. The oscillation frequency is then investigated over the parameter space of wall temperature, mixture velocity and channel height. The experiment results are explained well by the formula. One emphasis of this part is on the reasoning of flame quenching by gradual heat loss from the flame to adjacent wall during the propagation. Phase-locked OH-PLIF has been developed for 3-D structure of the flame, and flame temperature measurement with the 2-line method. The flame quenches first near the sidewall where temperature is lower. The propagation becomes slower as the flame moves closer to the quenching position, as shown by both the time-resolved chemiluminescence and phase-locked OH-PLIF images. The thermal quenching hypothesis is supported by phase-locked OH 2-line PLIF measurement showing the gradual decrease of flame temperature and analysis of time resolved OH* chemiluminescence images showing the decrease of transient propagating velocity. Finally, flame instabilities associated with premixed flames in the micro combustor have been covered.

5. Conclusions

Quenching distance measurement for different surface material under well-defined wall boundary conditions has been conducted in the configuration of flame in parallel plates. The inert surface gives slightly smaller quenching distance than the metallic surface for both low and high wall temperature, where weak chemical effect exists. Quenching distance for both the inert surface and metallic surface decreases with the increase of wall temperature, showing that the thermal effect. It is concluded in the present study that, unlike the claims of dominant chemical effect for high wall temperature made by Miesse et al. and Kim et al., thermal effect is the overall dominant factor within the wall temperature range and length scale investigated.

Micro quartz combustors with channel height of 0.7/1.0/1.5 mm have been developed for quantitative optical measurements of steady and oscillating flames with Time-resolved chemiluminescence and phase-locked two-line LIF measurement systems. The measurements provide quantitative information of the flame such as the flame temperature, flame speed and flame structure, which are not well discussed previously [1, 3]. The flame temperature, flame speed, flame structure, extinction limits region and flame stability are found greatly affected by the wall temperature. Flame temperature is increased by increasing the wall temperature. Thus, the extinction limits region becomes broader, and quenching distance can be smaller. The quenching distance for the quartz channel can be smaller than 0.7 mm when the wall temperature is over 800℃ . The extinction limits region becomes narrower with a thinner channel due to stronger wall effect.

Oscillating flame, of which frequency is in the range of 30-500Hz, is found mainly distributed in the fuel-rich side in the flammability map. There is a small quenching area between the steady flame and oscillating flame regions, which is due to strong hysteresis due to the thermal coupling between the flame and the wall. Oscillating flame is a periodic process of flame ignition at the exit, and propagation upstream, and then quenching by heat loss to the wall, and finally recharge to the exit for the next ignition. The oscillation frequency can be analyzed with a phenomenological formula considering the thermal effect. It is noted that the quenching of an oscillating flame is caused by the heat loss to the wall, as also pointed out in previous studies [3]. The present investigation supported the hypothesis by quantitative measurements showing the decrease of flame temperature during propagation and the deceleration of flame propagation.

本論文は,「マイクロ流路内予混合火炎における熱・化学的壁面効果に関する研究」と題し,5章より成っている.近年,リチウムイオン電池に替わる携帯電源として,燃料の化学的エネルギーから発電を行うマイクロ発電システムが注目されている.直接メタノール型燃料電池が代表的なデバイスの1つであるが,単位体積当たりの発電密度が高く,使用できる燃料に限定されないことから,燃焼プロセスを経る熱光発電,熱電システムの研究開発も進められている.本論文では,超小型燃焼器で重要となるマイクロ気相燃焼の消炎機構について実験的に検討している.マイクロ燃焼は通常の消炎距離よりも狭い間隙での燃焼現象と定義できるが,従来,熱的および化学的消炎機構の存在が報告されているものの,熱的境界条件を明確に規定した場合の知見は乏しい.本論文では,平行平板間および矩形チャネル燃焼器内での燃焼形態,消炎機構について,光学計測な定量計測により明らかにすることを目的としている.

第1章は,「序論」であり,従来の関連研究を概観し,本研究の目的を述べている.様々なマイクロエネルギー源の研究開発状況,特に本研究の動機付けとなったマイクロ熱光発電システムの概念とマイクロ燃焼器に対する要求使用が述べられている.また,マイクロ気相燃焼についての従来の研究について,壁温をはじめとする熱的消炎の効果,壁面材料に依存する化学的消炎,FREIを代表とする振動的な火炎について述べられている.そして,平行平板間および矩形チャネル燃焼器内での燃焼形態,消炎機構について,光学計測な定量計測により明らかにする,という本論文の目的が述べられている.

第2章は,「平行平板間の消炎距離」と題され,温度境界条件を同一とし,表面の材質のみを変化させた場合の消炎距離の計測について述べられている.溶融石英板の裏側に黒色石英を接合したプレート二枚を対向させ,裏側から赤外線加熱することにより精密の温度制御を行い,拡散火炎が空隙を通過するか否かを基に消炎距離を定義して,壁面温度を変化させて系統的な測定を行った.その結果,石英表面では,従来の研究と同様に,壁温上昇と共に消炎距離が減少すること,石英表面にCr薄膜を形成したときには,石英表面よりも消炎距離が増大するが,その影響は温度の影響よりも小さいことを示し,熱的消炎が支配的であることを明らかにしている.

第3章は,「マイクロ燃焼器内の火炎」と題され,石英製のマイクロ燃焼器内での燃焼形態について述べられている.溶融石英で製作したチャネル高さ0.7mm,1mm,1.5mmの燃焼器外側に黒色石英を接合し,裏側から赤外線加熱することにより精密の温度制御を行って,内部に形成される火炎の計測を行っている.まず,高速度ICCDカメラによるOHラジカル自発光の観察により,静止火炎および振動火炎の可視化が行われた.そして,壁温,チャネル高さを系統的に変化させた場合の燃焼可能範囲のマッピングを行って,壁温の上昇,チャネル高さの増大に伴って,可燃範囲が増大することを明らかにした.また,火炎温度が壁温に大きく依存すること,レーザー誘起蛍光法(LIF)により光学スライスされたマイクロ火炎の構造に壁温の影響が大きいことが述べられている.また,過濃側では,燃焼器出口で着火後,上流へ進行,消炎,燃料のリチャージ後再着火する,振動火炎が発生することが述べられている.

第4章は,「振動火炎」と題され,過濃側で発生する振動火炎について詳細な計測を行った結果を述べている.まず,振動火炎の動きに位相ロックして計測することができるLIFシステムの開発について述べられている.そして,振動火炎が上流側に進行するとともに,火炎速度が減速すること,そして壁温の低い側壁近傍から消炎が始まることが述べられている.また,OH2ライン法を用いた振動火炎の温度計測結果について述べられ,着火後,一旦火炎温度は上昇するものの,進行とともに低下し,消炎に至ることが明らかにしている.また,この温度計測の結果は,火炎速度の加減速と良く整合しており,振動火炎についても熱的消炎が支配的であることを明らかにしている.さらに,振動火炎の周波数の簡易的なモデルが提案され,実験結果を表現できることを示している.

第5章は結論であり,本論文で得られた成果がまとめられている.

以上要するに,本論文では,マイクロ気相燃焼での消炎機構について検討を行い,熱的境界条件を精密に制御した条件下において,静止火炎,振動火炎のいずれも,熱的消炎が支配的であることを明らかにしている.また,従来,可視化画像でのみ議論されていたマイクロ燃焼現象に対して,本研究では位相ロック・レーザー誘起蛍光法によるラジカル濃度測定,温度測定を初めて導入し,定量的かつ系統的なデータを得ることを可能とした.本研究により構築された計測手法は,今後マイクロ燃焼の現象解明の極めて有効なツールとなり得る.従って,本論文は,マイクロ燃焼のメカニズムに関する新たな知見を加えるもので,熱流体工学における学術的価値とともに工業的な利用価値が極めて高く,機械工学の上で寄与するところが大きい.

よって本論文は博士(工学)の学位請求論文として合格と認められる。