鈴木 隆

Recently, ammonia (NH3), which has a higher energy density than hydrogen, has gained attention for zero-carbon emission goals in the transportation sector. However, in a conventional internal combustion engine (ICE), NH3 combustion mechanism is still under investigation. In this paper, to further expand the knowledge on the adoption of NH3 in ICEs, authors conducted NH3/gasoline co-combustion experiments in a modified, 17.7:1 compression ratio, naturally aspirated spark-assisted CI engine with sub-chamber. The sub-chamber was chosen in order to enhance the combustion speed of NH3. In addition, the sub-chamber was equipped with glow and spark plugs to overcome the high auto-ignition temperature of NH3. Engine performance and NOX emissions were studied under three different intake air temperatures. During the experiments, NH3 content was increased gradually where the engine was run under lean conditions. Although higher NH3 content was achieved compared to our previous work, increasing the intake air temperature resulted in decreased charging efficiency. In addition, corrosion was found on the piston ring after 120 h of operation, negatively affecting the engine performance. Furthermore, NH3/gasoline co-combustion duration was shortened drastically with the influence of the sub-chamber, where the longest combustion duration under the present conditions was found to be 17°CA.

ディーゼル機関のモデルベースト燃焼制御器では，冷却損失を考慮した圧縮ポリトロープ指数の予測にこれまで実験式を用いてきた．著者らは，実験数を軽減するため，新たに物理モデルを開発し，制御器に実装した．実機にて過渡運転性能を評価したところ，実験値と比較して，モデルの予測誤差は0．３１％と評価された．

ディーゼル機関の過渡運転性能向上には，モデルベースト制御によるサイクル毎の燃料噴射量と時期の予測が有効である．本研究では，燃焼室の局所壁温度履歴および局所壁面熱流束履歴を測定し，著者らが構築したオンボード用筒内壁温度推定モデルおよび壁面熱伝達モデルの予測精度を評価したので報告する．

A new equation, which was dependent on physical principles, was developed for the study of heat transfer in CI engines which needs turbulence of gas flows to calculate heat flux. Proposed approach was implemented into a 1-D engine simulation, which was used to determine heat flux between in-cylinder gas and wall. Results from the suggested equation were compared to the previous conventional equations; Morel and Hohenberg, and to the engine experiments. The proposed equation showed better accuracy when compared with the conventional equations due to detailed representation of in-cylinder gas flow by dividing the combustion chamber into 6 different regions.

The diesel engines are superior in terms of power efficiency and fuel economy compared to gasoline engines. In order to optimize the performance of direct injection diesel engine, the effect of various intake pressure (boost pressure) from supercharging direct injection diesel engine was studied at various engine rotation. A single cylinder direct injection diesel engine was used in this experiment. The bore diameter of the engine used was set to 85 mm, the stroke length was set to 96.9 mm, and the compression ratio was set to 16.3. The variation of engine rotation started from 800 rpm to 2 000 rpm with 400 rpm increment. The variation of boost pressure is bounded from 0 kPa boost pressure (naturally aspirated) to the maximum of 60 kPa boost pressure with 20 kPa boost pressure increment. The performance of the engine is evaluated in terms of in-cylinder pressure and heat release rate as the most important performance characteristics of the diesel engine. The in-cylinder pressure and heat release rate of direct injection diesel engine are increased with the elevation of boost pressure at various engine rotation. The raise of engine rotation resulted in the decrease of maximum in-cylinder pressure and heat release rate.

本稿ではMAP作成の手間を自動化し, かつ環境変化に対するロバスト性を有する制御系設計法として, フィードバック誤差学習を用いたモデルベースト制御を提案する. また, 制御器の学習機構として小脳演算モデルコントローラを用いることで計算負荷を低減することができる. そして, 本手法の有効性を実機試験にて検証する.

Overall efficiency of internal combustion engines are heavily depended on intake air temperature which is directly related to the heat transfer inside an intake system. Previously, authors developed an equation by using port model setup to calculate Nusselt number with introduction of Graetz and Strouhal numbers. This study modified the port model equation to improve its accuracy in a real engine experimental setup. Predicted intake air temperature was compared to the measured data with a maximum error of 5.6%. Additionally, 100 K of temperature difference was found between the boost pressure values of 944hPa and 678hPa from 1-D engine simulation results.