Proceedings of the Combustion Institute最新文献

Assessment of flamelet/progress variable methods for supersonic combustion 超声速燃烧的小火焰/进度变量方法评价

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105798

Alexandra Baumgart , Matthew X. Yao , Guillaume Blanquart

Tabulated chemistry models, including the flamelet/progress variable approach, have been successfully used for a variety of turbulent flame simulations. The progress variable describes the progress of reactions in a system and parameterizes a lookup table of thermochemical variables. This approach reduces the cost of simulations, transporting only one scalar (progress variable) instead of the many species mass fractions required for detailed chemistry. Originally developed for low Mach number flame simulations, recent works have focused on extensions of this approach to compressible flames, supersonic combustion, and detonations, with applications such as scramjet combustors and rotating detonation engines. Unlike low Mach simulations, compressible flow simulations require solving the energy transport equation, which is coupled to the equation of state. This leads to additional modeling challenges regarding the thermodynamics and its impact on the chemistry. The validity of modeling assumptions, for example the relationship between energy and temperature, also varies with the combustion regime. The present work provides a detailed assessment of the existing strategies for chemistry tabulation for compressible/supersonic combustion, including detonations. A priori analysis indicates that approximations which are reasonable for weakly compressible flames may break down for shock-induced combustion. The analysis identifies specific assumptions and approximations that do not hold for detonations, emphasizing that care must be taken when applying tabulated chemistry models outside their intended combustion regimes.

表格化学模型，包括小火焰/进程变量方法，已经成功地用于各种湍流火焰模拟。progress变量描述了系统中反应的过程，并参数化了热化学变量的查找表。这种方法减少了模拟的成本，只传输一个标量（进程变量），而不是详细化学所需的许多物种质量分数。最初是为低马赫数火焰模拟而开发的，最近的工作重点是将这种方法扩展到可压缩火焰、超音速燃烧和爆震，应用于超燃冲压发动机燃烧室和旋转爆震发动机。与低马赫数模拟不同，可压缩流动模拟需要求解能量输运方程，该方程与状态方程耦合。这导致了关于热力学及其对化学的影响的额外建模挑战。模拟假设的有效性，例如能量和温度之间的关系，也随燃烧状态而变化。目前的工作提供了一个详细的评估现有策略的化学制表的可压缩/超音速燃烧，包括爆轰。先验分析表明，对于弱可压缩火焰合理的近似对于激波燃烧可能失效。分析确定了不适用爆轰的特定假设和近似，强调在应用表化化学模型超出其预期燃烧范围时必须小心。

{"title":"Assessment of flamelet/progress variable methods for supersonic combustion","authors":"Alexandra Baumgart , Matthew X. Yao , Guillaume Blanquart","doi":"10.1016/j.proci.2025.105798","DOIUrl":"10.1016/j.proci.2025.105798","url":null,"abstract":"<div><div>Tabulated chemistry models, including the flamelet/progress variable approach, have been successfully used for a variety of turbulent flame simulations. The progress variable describes the progress of reactions in a system and parameterizes a lookup table of thermochemical variables. This approach reduces the cost of simulations, transporting only one scalar (progress variable) instead of the many species mass fractions required for detailed chemistry. Originally developed for low Mach number flame simulations, recent works have focused on extensions of this approach to compressible flames, supersonic combustion, and detonations, with applications such as scramjet combustors and rotating detonation engines. Unlike low Mach simulations, compressible flow simulations require solving the energy transport equation, which is coupled to the equation of state. This leads to additional modeling challenges regarding the thermodynamics and its impact on the chemistry. The validity of modeling assumptions, for example the relationship between energy and temperature, also varies with the combustion regime. The present work provides a detailed assessment of the existing strategies for chemistry tabulation for compressible/supersonic combustion, including detonations. A priori analysis indicates that approximations which are reasonable for weakly compressible flames may break down for shock-induced combustion. The analysis identifies specific assumptions and approximations that do not hold for detonations, emphasizing that care must be taken when applying tabulated chemistry models outside their intended combustion regimes.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105798"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144852906","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

IR-HyChem: Towards modeling the high-T combustion behavior of aviation fuels using infrared spectra IR-HyChem：利用红外光谱模拟航空燃料的高t燃烧行为

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105792

Pujan Biswas, Vivek Boddapati, Andrew R. Klingberg, Alka Panda, Hai Wang, Ronald K. Hanson

A Fourier transform infrared (FTIR) spectra-based approach, namely IR-HyChem, was developed to model the combustion behavior of jet and rocket fuels. Earlier shock-tube experiments employed laser absorption spectroscopy (LAS) to measure the yields of key stable intermediates: CH₄, C₂H₄, and >C₂ alkenes such as C₃H₆, 1-C₄H₈ and i-C₄H₈, during the pyrolysis of neat hydrocarbons across several molecular classes (n-alkanes, lightly branched alkanes and highly branched alkanes). These measurements revealed empirical relations of molecular structure to the yields of these intermediates. The relationships provided important insights into fuel reactivity under high-temperature, combustor-relevant conditions. The IR-HyChem methodology establishes quantitative correlations between spectral features and the yields of these pyrolysis intermediates. Using this framework, IR-HyChem models were demonstrated for two jet fuels (JP-8 and F-24) and a rocket fuel (RP-1), by constraining a subset of stoichiometric parameters in the HyChem lumped reactions, resulting in partially constrained IR-HyChem models. These models were evaluated against ignition delay times (IDTs) measured behind reflected shock waves at elevated pressures, demonstrating strong agreement with experimental data. A Monte Carlo uncertainty analysis revealed that imposing FTIR-based constraints reduced variance in IDT predictions compared to unconstrained models. Furthermore, sensitivity analysis indicated that additional IR spectra-based correlations could improve the accuracy of the IR-HyChem models. Overall, this work demonstrates the utility of FTIR spectra and their potential use as a low-volume tool for developing predictive chemistry models for the combustion of real, multi-component fuels.

一种基于傅里叶变换红外（FTIR）光谱的方法，即IR-HyChem，被开发用于模拟喷气和火箭燃料的燃烧行为。早期的激波管实验采用激光吸收光谱（LAS）来测量几种分子类别（正链烷烃、轻支链烷烃和高支链烷烃）的纯烃热解过程中关键稳定中间体：CH4、C2H4和>；C2烯烃，如C3H6、1-C4H8和i-C4H8的产率。这些测量结果揭示了分子结构与中间体产率的经验关系。这些关系为研究高温、燃烧器相关条件下的燃料反应性提供了重要的见解。IR-HyChem方法建立了光谱特征与这些热解中间体产率之间的定量相关性。利用这一框架，通过约束HyChem集中反应中的一部分化学计量参数，对两种喷气燃料（JP-8和F-24）和一种火箭燃料（RP-1）的IR-HyChem模型进行了验证，得到了部分约束的IR-HyChem模型。这些模型与在高压下反射冲击波后测量的点火延迟时间（IDTs）进行了评估，证明了与实验数据的强烈一致性。蒙特卡罗不确定性分析显示，与无约束模型相比，施加基于ftir的约束减少了IDT预测的方差。此外，敏感性分析表明，额外的基于红外光谱的相关性可以提高IR- hychem模型的准确性。总的来说，这项工作证明了FTIR光谱的实用性，以及它们作为一种小体积工具的潜在用途，可用于开发真实多组分燃料燃烧的预测化学模型。

{"title":"IR-HyChem: Towards modeling the high-T combustion behavior of aviation fuels using infrared spectra","authors":"Pujan Biswas, Vivek Boddapati, Andrew R. Klingberg, Alka Panda, Hai Wang, Ronald K. Hanson","doi":"10.1016/j.proci.2025.105792","DOIUrl":"10.1016/j.proci.2025.105792","url":null,"abstract":"<div><div>A Fourier transform infrared (FTIR) spectra-based approach, namely IR-HyChem, was developed to model the combustion behavior of jet and rocket fuels. Earlier shock-tube experiments employed laser absorption spectroscopy (LAS) to measure the yields of key stable intermediates: CH<sub>4</sub>, C<sub>2</sub>H<sub>4</sub>, and >C<sub>2</sub> alkenes such as C<sub>3</sub>H<sub>6</sub>, 1-C<sub>4</sub>H<sub>8</sub> and <em>i</em>-C<sub>4</sub>H<sub>8</sub>, during the pyrolysis of neat hydrocarbons across several molecular classes (<em>n</em>-alkanes, lightly branched alkanes and highly branched alkanes). These measurements revealed empirical relations of molecular structure to the yields of these intermediates. The relationships provided important insights into fuel reactivity under high-temperature, combustor-relevant conditions. The IR-HyChem methodology establishes quantitative correlations between spectral features and the yields of these pyrolysis intermediates. Using this framework, IR-HyChem models were demonstrated for two jet fuels (JP-8 and F-24) and a rocket fuel (RP-1), by constraining a subset of stoichiometric parameters in the HyChem lumped reactions, resulting in partially constrained IR-HyChem models. These models were evaluated against ignition delay times (IDTs) measured behind reflected shock waves at elevated pressures, demonstrating strong agreement with experimental data. A Monte Carlo uncertainty analysis revealed that imposing FTIR-based constraints reduced variance in IDT predictions compared to unconstrained models. Furthermore, sensitivity analysis indicated that additional IR spectra-based correlations could improve the accuracy of the IR-HyChem models. Overall, this work demonstrates the utility of FTIR spectra and their potential use as a low-volume tool for developing predictive chemistry models for the combustion of real, multi-component fuels.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105792"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144772152","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SAGE: A machine learning model for primary particle segmentation in TEM images of soot aggregates SAGE：一种用于烟尘聚集体TEM图像中初级粒子分割的机器学习模型

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105821

Timothy P. Day, Khaled Mosharraf Mukut, Luke Klacik, Ryan O’Donnell, James Wasilewski, Somesh P. Roy

Accurate characterization of the morphology of soot is essential for our understanding and better modeling of the physical and chemical properties of soot. The morphological characteristics of soot are traditionally explored experimentally via Transmission Electron Microscopy (TEM), usually by investigating the images via manual segmentation, which is highly labor intensive. To improve this process, a novel model for the automatic segmentation of primary particles in TEM images of soot is presented in this work. The goal of the model is to identify and isolate each primary particle from a TEM image of a soot aggregate. The model, titled Soot Aggregate Geometry Extraction (SAGE) employs a two-stage training process using a convolutional neural network: an initial training on synthetically-generated TEM images followed by a refinement training by using manually segmented real TEM images. The model was tested against a dataset of real TEM images that included images from sources different from the training data (i.e., different instruments and different researchers). When tested against this real TEM image dataset of soot, SAGE shows good performance with an F

_{1}

score of 67.7%, indicating its ability to correctly identify primary particles while achieving a balanced trade off between missing true particles and detecting false ones. SAGE is able to detect more primary particles with better shape and size alignments with the ground truth data than traditional methods such as circular Hough transform or Euclidean distance mapping methods, leading to a much higher mean Intersection over Union score of 62.2%. Unlike most existing approaches that produce circular segmentations and require image-by-image tuning, SAGE effectively captures irregular particle boundaries without additional adjustments. The particle size distribution obtained from SAGE matches well with the ground truth. The median errors of predictions obtained from SAGE fall below 5% and 1%, respectively, for radius of gyration and fractal dimension of particles.

准确表征烟灰的形态是必不可少的，我们的理解和更好的建模烟灰的物理和化学性质。传统的实验方法是通过透射电子显微镜（TEM）来研究煤烟的形态特征，通常是通过人工分割来研究图像，这是一种高度劳动密集型的方法。为了改进这一过程，本文提出了一种新的烟尘TEM图像中初级粒子的自动分割模型。该模型的目标是从煤烟集合体的TEM图像中识别和分离每个初级颗粒。该模型名为Soot Aggregate Geometry Extraction (SAGE)，采用卷积神经网络的两阶段训练过程：首先对合成的TEM图像进行初始训练，然后使用手动分割的真实TEM图像进行细化训练。该模型在真实TEM图像数据集上进行了测试，该数据集包括来自不同训练数据来源的图像（即不同的仪器和不同的研究人员）。当对真实的煤烟TEM图像数据集进行测试时，SAGE显示出良好的性能，F1得分为67.7%，表明它能够正确识别初级颗粒，同时在缺失真颗粒和检测假颗粒之间实现平衡。与传统方法（如圆形霍夫变换或欧几里得距离映射方法）相比，SAGE能够检测到更多具有更好形状和尺寸与地面真实数据对齐的初级粒子，从而导致更高的平均Intersection over Union得分为62.2%。与大多数产生圆形分割并需要逐图调整的现有方法不同，SAGE有效地捕获不规则粒子边界，而无需额外调整。SAGE得到的粒度分布与地面真实值吻合较好。SAGE对颗粒的旋转半径和分形维数的预测中值误差分别在5%和1%以下。

{"title":"SAGE: A machine learning model for primary particle segmentation in TEM images of soot aggregates","authors":"Timothy P. Day, Khaled Mosharraf Mukut, Luke Klacik, Ryan O’Donnell, James Wasilewski, Somesh P. Roy","doi":"10.1016/j.proci.2025.105821","DOIUrl":"10.1016/j.proci.2025.105821","url":null,"abstract":"<div><div>Accurate characterization of the morphology of soot is essential for our understanding and better modeling of the physical and chemical properties of soot. The morphological characteristics of soot are traditionally explored experimentally via Transmission Electron Microscopy (TEM), usually by investigating the images via manual segmentation, which is highly labor intensive. To improve this process, a novel model for the automatic segmentation of primary particles in TEM images of soot is presented in this work. The goal of the model is to identify and isolate each primary particle from a TEM image of a soot aggregate. The model, titled Soot Aggregate Geometry Extraction (SAGE) employs a two-stage training process using a convolutional neural network: an initial training on synthetically-generated TEM images followed by a refinement training by using manually segmented real TEM images. The model was tested against a dataset of real TEM images that included images from sources different from the training data (i.e., different instruments and different researchers). When tested against this real TEM image dataset of soot, SAGE shows good performance with an F<span><math><msub><mrow></mrow><mrow><mn>1</mn></mrow></msub></math></span> score of 67.7%, indicating its ability to correctly identify primary particles while achieving a balanced trade off between missing true particles and detecting false ones. SAGE is able to detect more primary particles with better shape and size alignments with the ground truth data than traditional methods such as circular Hough transform or Euclidean distance mapping methods, leading to a much higher mean Intersection over Union score of 62.2%. Unlike most existing approaches that produce circular segmentations and require image-by-image tuning, SAGE effectively captures irregular particle boundaries without additional adjustments. The particle size distribution obtained from SAGE matches well with the ground truth. The median errors of predictions obtained from SAGE fall below 5% and 1%, respectively, for radius of gyration and fractal dimension of particles.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105821"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145044042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Thermal decomposition kinetics of plastic mixtures (PE/PVC and PE/PP) based on chemical reaction neural networks 基于化学反应神经网络的塑料混合物（PE/PVC和PE/PP）热分解动力学

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105874

Wei Sun , Xinzhe Chen , Dongping Chen

Polyethylene (PE), polypropylene (PP), polyvinyl chloride (PVC), and their mixtures are widely used in packaging, electrical, and construction fields. Thermal decomposition is one of the primary methods for their recycling. In this study, the thermal decomposition kinetics of three plastics: PE, PP and PVC, along with two mixtures, e.g., PE/PVC and PE/PP, were investigated using a chemical reaction neural network (CRNN). Three models with four species and two reactions (4-2 model) are developed for PE, PP, and PVC decomposition. The experimental thermogravimetric (TG) curves can be well reproduced. The corresponding kinetic models for plastic mixtures are also proposed by integrating the kinetic models of single components. The results indicate that the predicted TG curves of PE/PVC mixtures align closely with the experimental data, confirming the absence of coupling effects between PE and PVC decomposition. However, the model for PE/PP mixtures fails to accurately predict the thermal decomposition process with a noticeable underprediction of the initial decomposition temperature. A careful analysis highlights the strong coupling effect between PE and PP decomposition upon heating, and a simple combination of kinetic models for single components cannot fully reveal the thermal decomposition mechanisms of the PE/PP mixtures. This work opens up a new modelling approach to build the kinetic models for plastics and evaluate the potential coupling effect in the practical decomposition of plastic mixtures.

聚乙烯（PE）、聚丙烯（PP）、聚氯乙烯（PVC）及其混合物广泛应用于包装、电气、建筑等领域。热分解是其回收利用的主要方法之一。本研究利用化学反应神经网络（CRNN）研究了PE、PP和PVC三种塑料以及PE/PVC和PE/PP两种混合物的热分解动力学。建立了PE、PP、PVC分解的四种两反应模型（4-2模型）。实验热重（TG）曲线可以很好地再现。通过对单组分动力学模型的积分，提出了相应的塑性混合料动力学模型。结果表明，预测的PE/PVC混合物的热重曲线与实验数据吻合较好，证实了PE和PVC分解之间不存在耦合效应。然而，PE/PP混合物的模型不能准确预测热分解过程，对初始分解温度的预测明显不足。仔细分析表明，PE和PP在加热时分解之间存在很强的耦合效应，单一组分的动力学模型的简单组合并不能充分揭示PE/PP混合物的热分解机理。本工作为建立塑料的动力学模型和评估塑料混合物实际分解中潜在的耦合效应开辟了一种新的建模方法。

{"title":"Thermal decomposition kinetics of plastic mixtures (PE/PVC and PE/PP) based on chemical reaction neural networks","authors":"Wei Sun , Xinzhe Chen , Dongping Chen","doi":"10.1016/j.proci.2025.105874","DOIUrl":"10.1016/j.proci.2025.105874","url":null,"abstract":"<div><div>Polyethylene (PE), polypropylene (PP), polyvinyl chloride (PVC), and their mixtures are widely used in packaging, electrical, and construction fields. Thermal decomposition is one of the primary methods for their recycling. In this study, the thermal decomposition kinetics of three plastics: PE, PP and PVC, along with two mixtures, e.g., PE/PVC and PE/PP, were investigated using a chemical reaction neural network (CRNN). Three models with four species and two reactions (4-2 model) are developed for PE, PP, and PVC decomposition. The experimental thermogravimetric (TG) curves can be well reproduced. The corresponding kinetic models for plastic mixtures are also proposed by integrating the kinetic models of single components. The results indicate that the predicted TG curves of PE/PVC mixtures align closely with the experimental data, confirming the absence of coupling effects between PE and PVC decomposition. However, the model for PE/PP mixtures fails to accurately predict the thermal decomposition process with a noticeable underprediction of the initial decomposition temperature. A careful analysis highlights the strong coupling effect between PE and PP decomposition upon heating, and a simple combination of kinetic models for single components cannot fully reveal the thermal decomposition mechanisms of the PE/PP mixtures. This work opens up a new modelling approach to build the kinetic models for plastics and evaluate the potential coupling effect in the practical decomposition of plastic mixtures.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105874"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145216961","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Thermochemical state analysis of DMMP on non-premixed boundary layer flames DMMP在非预混边界层火焰中的热化学状态分析

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105898

Raúl Felipe Corrales Flores , Federica Ferraro , Arne Scholtissek

Boundary layer flames (BLFs), established near flammable “active” walls in fire scenarios, are fueled by gaseous volatiles released during the thermal degradation of wall materials. Their suppression is essential for fire safety and often relies on the use of flame retardants. This study investigates the inhibition effectiveness of dimethyl methylphosphonate (DMMP), a phosphorous-based flame retardant, by analyzing its impact on non-premixed flames in a counterflow configuration. The counterflow flame is a suitable reference configuration since it offers a controlled environment for resolving the relevant transport and chemical effects, while also allowing a direct comparison with experimental data from the literature. Using methane as a reference fuel, the numerical framework is validated for undoped and DMMP-doped flames, and then used to examine how strain rate, oxidizer temperature, and injection location (fuel or oxidizer side) influence flame inhibition. To connect with near-wall conditions, boundary conditions from a non-premixed BLF generated in a side wall quenching (SWQ) setup, are also applied in the counterflow simulations. The results show that DMMP promotes combustion at low strain rates and high flame temperatures, but acts as an inhibitor at higher strain rates and lower flame temperatures. The injection location strongly influences inhibition efficiency: due to transport limitations, in a methane non-premixed flame at atmospheric conditions nearly 100 times more DMMP must be issued from the fuel-side compared to oxidizer-side injection to reach a comparable flame inhibition. Furthermore, lower oxidizer temperatures enhance inhibition by increasing DMMP penetration into the reaction zone. Since flame retardants are typically released with the fuel in real fires, these findings hint towards challenges and opportunities for achieving effective suppression.

边界层火焰（blf）在火灾场景中建立在易燃的“活性”墙壁附近，由墙壁材料热降解过程中释放的气态挥发物推动。它们的抑制对消防安全至关重要，通常依赖于阻燃剂的使用。本研究通过分析磷基阻燃剂二甲基膦酸二甲酯（DMMP）在逆流配置下对非预混火焰的影响，研究了其抑制效果。逆流火焰是一个合适的参考配置，因为它为解决相关的传输和化学效应提供了一个受控的环境，同时也允许与文献中的实验数据进行直接比较。以甲烷作为参考燃料，验证了未掺杂和dmmp掺杂火焰的数值框架，然后用于研究应变速率、氧化剂温度和喷射位置（燃料或氧化剂一侧）对火焰抑制的影响。为了与近壁条件联系起来，在侧壁淬火（SWQ）装置中产生的非预混BLF的边界条件也被应用于逆流模拟中。结果表明，DMMP在低应变速率和高火焰温度下促进燃烧，但在高应变速率和低火焰温度下起抑制作用。喷射位置对抑制效率有很大影响：由于输送限制，在大气条件下的甲烷非预混火焰中，与氧化剂侧喷射相比，燃料侧喷射的DMMP必须高出近100倍，才能达到相当的抑焰效果。此外，较低的氧化剂温度通过增加DMMP渗透到反应区来增强抑制作用。由于在实际火灾中，阻燃剂通常与燃料一起释放，因此这些发现暗示了实现有效抑制的挑战和机遇。

{"title":"Thermochemical state analysis of DMMP on non-premixed boundary layer flames","authors":"Raúl Felipe Corrales Flores , Federica Ferraro , Arne Scholtissek","doi":"10.1016/j.proci.2025.105898","DOIUrl":"10.1016/j.proci.2025.105898","url":null,"abstract":"<div><div>Boundary layer flames (BLFs), established near flammable “active” walls in fire scenarios, are fueled by gaseous volatiles released during the thermal degradation of wall materials. Their suppression is essential for fire safety and often relies on the use of flame retardants. This study investigates the inhibition effectiveness of dimethyl methylphosphonate (DMMP), a phosphorous-based flame retardant, by analyzing its impact on non-premixed flames in a counterflow configuration. The counterflow flame is a suitable reference configuration since it offers a controlled environment for resolving the relevant transport and chemical effects, while also allowing a direct comparison with experimental data from the literature. Using methane as a reference fuel, the numerical framework is validated for undoped and DMMP-doped flames, and then used to examine how strain rate, oxidizer temperature, and injection location (fuel or oxidizer side) influence flame inhibition. To connect with near-wall conditions, boundary conditions from a non-premixed BLF generated in a side wall quenching (SWQ) setup, are also applied in the counterflow simulations. The results show that DMMP promotes combustion at low strain rates and high flame temperatures, but acts as an inhibitor at higher strain rates and lower flame temperatures. The injection location strongly influences inhibition efficiency: due to transport limitations, in a methane non-premixed flame at atmospheric conditions nearly 100 times more DMMP must be issued from the fuel-side compared to oxidizer-side injection to reach a comparable flame inhibition. Furthermore, lower oxidizer temperatures enhance inhibition by increasing DMMP penetration into the reaction zone. Since flame retardants are typically released with the fuel in real fires, these findings hint towards challenges and opportunities for achieving effective suppression.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105898"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145262470","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Three-dimensional detonation structures and effects of thermal confinement in a linear channel 线性通道内三维爆轰结构及热约束效应

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105873

Zhaoxin Ren , Jac Clarke

This study employs three-dimensional (3D) numerical simulations to investigate the detonation wave propagation in an unwrapped annular combustor configuration, focusing on thermal confinement effects on detonation structures and blast dynamics. The compressible Navier-Stokes equations are solved for stoichiometric kerosene-air mixtures under three distinct wall boundary conditions: (1) adiabatic (uncooled), (2) isothermal at 300 K (representing actively cooled walls), and (3) hybrid adiabatic-isothermal configurations. Results reveal that wall temperature critically governs detonation morphology: adiabatic boundaries produce regular cellular structures via ‘multi-kernel’ formation (intersections of four transverse waves), while cooled walls (300 K) generate stripe-like ‘line-kernel’ (formed through two-wave intersections), accompanied by double-wave structures, increased pressure fluctuations, and unburned fuel pockets. The hybrid case demonstrates asymmetric detonation development, with stable propagation on the adiabatic side contrasting with elongated cells and intensified wave-wall interactions on the cooled side. Quantitative analysis shows that cooled boundaries reduce the detonation wave height compared to adiabatic cases and promote irregular cell sizes due to suppressed boundary layer reactions. These findings present the first systematic evidence of 3D thermal confinement effects on RDW dynamics, revealing a critical trade-off in combustor design: while lower wall temperatures enhance material durability, they compromise combustion efficiency through increased flow unsteadiness and incomplete fuel consumption. The study advances the fundamental understanding of detonation physics in practical thermal gradients and provides actionable insights for optimizing cooling strategies in rotating detonation engines.

本研究采用三维数值模拟的方法研究了未包裹的环形燃烧室结构中爆轰波的传播，重点研究了热约束对爆轰结构和爆炸动力学的影响。在三种不同的壁面边界条件下，求解了煤油-空气混合物的可压缩Navier-Stokes方程：(1)绝热（未冷却），(2)300k等温（代表主动冷却壁面），以及(3)绝热-等温混合构型。结果表明，壁面温度对爆爆形态有重要影响：绝热边界通过“多核”形成（四个横波的交叉点）产生规则的细胞结构，而冷却壁面（300 K）产生条纹状的“线核”（通过两波交叉点形成），伴随着双波结构、压力波动增加和未燃烧的燃料袋。混合情况表现出不对称爆轰发展，绝热侧的传播稳定，而冷却侧的波壁相互作用加剧。定量分析表明，与绝热情况相比，冷却边界降低了爆震波高度，并且由于抑制了边界层反应，导致胞体尺寸不规则。这些发现首次提供了3D热约束对RDW动力学影响的系统证据，揭示了燃烧室设计中的一个关键权衡：虽然较低的壁温提高了材料的耐久性，但它们会增加流动不稳定和不完全燃料消耗，从而损害燃烧效率。该研究促进了对实际热梯度爆轰物理的基本理解，并为优化旋转爆轰发动机的冷却策略提供了可行的见解。

{"title":"Three-dimensional detonation structures and effects of thermal confinement in a linear channel","authors":"Zhaoxin Ren , Jac Clarke","doi":"10.1016/j.proci.2025.105873","DOIUrl":"10.1016/j.proci.2025.105873","url":null,"abstract":"<div><div>This study employs three-dimensional (3D) numerical simulations to investigate the detonation wave propagation in an unwrapped annular combustor configuration, focusing on thermal confinement effects on detonation structures and blast dynamics. The compressible Navier-Stokes equations are solved for stoichiometric kerosene-air mixtures under three distinct wall boundary conditions: (1) adiabatic (uncooled), (2) isothermal at 300 K (representing actively cooled walls), and (3) hybrid adiabatic-isothermal configurations. Results reveal that wall temperature critically governs detonation morphology: adiabatic boundaries produce regular cellular structures via ‘multi-kernel’ formation (intersections of four transverse waves), while cooled walls (300 K) generate stripe-like ‘line-kernel’ (formed through two-wave intersections), accompanied by double-wave structures, increased pressure fluctuations, and unburned fuel pockets. The hybrid case demonstrates asymmetric detonation development, with stable propagation on the adiabatic side contrasting with elongated cells and intensified wave-wall interactions on the cooled side. Quantitative analysis shows that cooled boundaries reduce the detonation wave height compared to adiabatic cases and promote irregular cell sizes due to suppressed boundary layer reactions. These findings present the first systematic evidence of 3D thermal confinement effects on RDW dynamics, revealing a critical trade-off in combustor design: while lower wall temperatures enhance material durability, they compromise combustion efficiency through increased flow unsteadiness and incomplete fuel consumption. The study advances the fundamental understanding of detonation physics in practical thermal gradients and provides actionable insights for optimizing cooling strategies in rotating detonation engines.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105873"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145216291","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Numerical and experimental comparison of H2/air flame–shock interaction H2/空气火焰-激波相互作用的数值与实验比较

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105847

Emilie Yhuel , Anthony Roque Ccacya , Guillaume Ribert , Pascale Domingo , Nabiha Chaumeix

A high-fidelity three-dimensional numerical simulation is performed to replicate the experimental shock tube setup of the ICARE laboratory, where a hydrogen–air flame–shock interaction (FSI) is studied. Following the experimental procedure, a lean flame (

ϕ = 0.8

) is ignited at the closed end of a rectangular channel before a shock wave, traveling at a Mach number of

M_{s} = 1.9

, is triggered from the opposite side. The FSI occurs when the shock wave encounters the laminar flame in the visualization window, where experimental schlieren images have been captured. To reproduce this experiment, the fully compressible Navier–Stokes equations are solved using the San Diego mechanism, which includes nine reacting species (excluding nitrogen oxides) and 23 kinetic reactions. Species diffusion is modeled using the Hirschfelder–Curtiss model combined with thermal diffusion (Soret effect). Additionally, gravity is accounted for in the simulation.

The three major observations of the experiment were well captured by the numerical simulations through a comparison of experimental and numerical schlieren images: the laminar flame propagation and its interaction with adjacent walls; the first FSI, which leads to the formation of Richtmyer–Meshkov instabilities (RMI); and finally, the second FSI, occurring when the reflected shock wave travels back towards the entrance, generating reactive boundary layers and multiple shock interactions within the funnel of fresh gases produced by the RMI.

本文采用高保真三维数值模拟方法，模拟了ICARE实验室的激波管实验装置，研究了氢-空气火焰-激波相互作用（FSI）。按照实验程序，在矩形通道的封闭端点燃一个细焰（φ =0.8），然后从另一侧触发一个以马赫数Ms=1.9行进的激波。当激波在可视化窗口中遇到层流火焰时，FSI发生，实验纹影图像已被捕获。为了重现这个实验，我们使用圣地亚哥机制求解了完全可压缩的Navier-Stokes方程，该机制包括9种反应物质（不包括氮氧化物）和23种动力学反应。物种扩散采用Hirschfelder-Curtiss模型结合热扩散（Soret效应）进行建模。此外，在模拟中考虑了重力。通过实验和数值纹影图像的对比，数值模拟很好地捕捉到了实验的三个主要观察结果：层流火焰的传播及其与相邻壁面的相互作用；第一个FSI导致richmyer - meshkov不稳定性（RMI）的形成；最后是第二次FSI，发生在反射的激波向入口传播时，在RMI产生的新鲜气体漏斗内产生反应边界层和多重激波相互作用。

{"title":"Numerical and experimental comparison of H2/air flame–shock interaction","authors":"Emilie Yhuel , Anthony Roque Ccacya , Guillaume Ribert , Pascale Domingo , Nabiha Chaumeix","doi":"10.1016/j.proci.2025.105847","DOIUrl":"10.1016/j.proci.2025.105847","url":null,"abstract":"<div><div>A high-fidelity three-dimensional numerical simulation is performed to replicate the experimental shock tube setup of the ICARE laboratory, where a hydrogen–air flame–shock interaction (FSI) is studied. Following the experimental procedure, a lean flame (<span><math><mrow><mi>ϕ</mi><mo>=</mo><mn>0</mn><mo>.</mo><mn>8</mn></mrow></math></span>) is ignited at the closed end of a rectangular channel before a shock wave, traveling at a Mach number of <span><math><mrow><msub><mrow><mi>M</mi></mrow><mrow><mi>s</mi></mrow></msub><mo>=</mo><mn>1</mn><mo>.</mo><mn>9</mn></mrow></math></span>, is triggered from the opposite side. The FSI occurs when the shock wave encounters the laminar flame in the visualization window, where experimental schlieren images have been captured. To reproduce this experiment, the fully compressible Navier–Stokes equations are solved using the San Diego mechanism, which includes nine reacting species (excluding nitrogen oxides) and 23 kinetic reactions. Species diffusion is modeled using the Hirschfelder–Curtiss model combined with thermal diffusion (Soret effect). Additionally, gravity is accounted for in the simulation.</div><div>The three major observations of the experiment were well captured by the numerical simulations through a comparison of experimental and numerical schlieren images: the laminar flame propagation and its interaction with adjacent walls; the first FSI, which leads to the formation of Richtmyer–Meshkov instabilities (RMI); and finally, the second FSI, occurring when the reflected shock wave travels back towards the entrance, generating reactive boundary layers and multiple shock interactions within the funnel of fresh gases produced by the RMI.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105847"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145216292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

On the transition of four flames types of auto-ignited iso-octane droplet cloud 自燃异辛烷液滴云四种火焰类型的过渡研究

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105926

Zixuan Ding , Hengyi Zhou , Haiyu Song , Yu Cheng Liu

This study is motivated by the multi-stage ignition behavior of isooctane droplet clouds suggested by its chemical kinetics studies. A series of 1D numerical simulations and theoretical analyses were conducted to investigate auto-ignition phenomena. Four distinct flame structures, i.e. simple, two-stage, bilayer, and complicated, were identified, corresponding to two ignition modes: cool ignition alone and cool ignition followed by hot ignition. The resulting regime diagram in the T_a–G_ig space exhibits an inverted S-shaped boundary separating the presence and absence of hot ignition, indicating that increasing ambient temperature does not always promote ignition near the flammability limit of isooctane droplet cloud. To explain this non-monotonic behavior, we proposed two Damköhler numbers. In the 700 ∼ 900 K range, hot ignition is triggered by fuel accumulation near the cool flame due to faster chemistry than droplet vaporization, described by a droplet-scale Damköhler number (Da_d). In the 1000 ∼ 1500 K range, radical buildup from the convergence of cool and warm flame initiates hot ignition, which is governed by a cloud-scale Damköhler number (Da_c) comparing reaction and diffusion timescales. Additionally, in non-igniting cases, the cloud radius was observed to decrease nearly linearly, despite each droplet inside the cloud following the d²-law. This led to the development of a conduction-driven vaporization model for droplet cloud, enabling accurate prediction of cloud lifetime in hot environments.

异辛烷液滴云的化学动力学研究表明，异辛烷液滴云具有多级点火行为。对自燃现象进行了一系列一维数值模拟和理论分析。识别出简单、两级、双层和复杂四种不同的火焰结构，对应着单独冷点火和冷点火后热点火两种点火方式。Ta-Gig空间的状态图呈现出一个倒s形边界，将热着火的存在和不存在分开，这表明环境温度的升高并不总是促进异辛烷液滴云可燃性极限附近的着火。为了解释这种非单调行为，我们提出了两个Damköhler数字。在700 ~ 900 K范围内，由于比液滴汽化更快的化学反应，在冷火焰附近的燃料积累引发热点火，用液滴尺度Damköhler数（Dad）来描述。在1000 ~ 1500 K范围内，由冷热火焰汇聚而成的自由基形成引发热点火，这是由比较反应和扩散时间尺度的云尺度Damköhler数（Dac）控制的。此外，在不点燃的情况下，观察到云半径几乎线性减小，尽管云内的每个液滴都遵循d²定律。这导致了液滴云的电导驱动汽化模型的发展，能够准确预测热环境下的云寿命。

{"title":"On the transition of four flames types of auto-ignited iso-octane droplet cloud","authors":"Zixuan Ding , Hengyi Zhou , Haiyu Song , Yu Cheng Liu","doi":"10.1016/j.proci.2025.105926","DOIUrl":"10.1016/j.proci.2025.105926","url":null,"abstract":"<div><div>This study is motivated by the multi-stage ignition behavior of isooctane droplet clouds suggested by its chemical kinetics studies. A series of 1D numerical simulations and theoretical analyses were conducted to investigate auto-ignition phenomena. Four distinct flame structures, i.e. simple, two-stage, bilayer, and complicated, were identified, corresponding to two ignition modes: cool ignition alone and cool ignition followed by hot ignition. The resulting regime diagram in the T<sub>a</sub>–G<sub>ig</sub> space exhibits an inverted S-shaped boundary separating the presence and absence of hot ignition, indicating that increasing ambient temperature does not always promote ignition near the flammability limit of isooctane droplet cloud. To explain this non-monotonic behavior, we proposed two Damköhler numbers. In the 700 ∼ 900 K range, hot ignition is triggered by fuel accumulation near the cool flame due to faster chemistry than droplet vaporization, described by a droplet-scale Damköhler number (Da<sub>d</sub>). In the 1000 ∼ 1500 K range, radical buildup from the convergence of cool and warm flame initiates hot ignition, which is governed by a cloud-scale Damköhler number (Da<sub>c</sub>) comparing reaction and diffusion timescales. Additionally, in non-igniting cases, the cloud radius was observed to decrease nearly linearly, despite each droplet inside the cloud following the d²-law. This led to the development of a conduction-driven vaporization model for droplet cloud, enabling accurate prediction of cloud lifetime in hot environments.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105926"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145319803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Three-dimensional analysis of hydrogen fuel effects in multi-tube combustor 多管燃烧室氢燃料效应的三维分析

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105790

Shuzhi Zhang , Vansh Sharma , Venkat Raman , Tristan T. Shahin , Alexander J. Hodge , Rohan M. Gejji , Robert P. Lucht , Carson D. Slabaugh

Flame characteristics, including mixing and stabilization in a multi-tube micromixer (MTM) combustor operating with hydrogen fuel are analyzed using high-fidelity simulations. In this flow configuration, a bundle of tubes issue fuel-air mixture into a combustion chamber, leading to multiple flame fronts, which may interact in an unsteady manner. Highly-resolved simulations enabled by adaptive mesh refinement with detailed kinetics are used. The results are first validated against experimental data for both a methane–hydrogen blend and pure hydrogen fuel, showing very good agreement with experimental image data. A detailed analysis of the hydrogen case reveals that global flame structures display distinct shapes across various flow streams. Upstream jet interactions induce mixture stratification along the tubes that is amplified with geometry-induced flow strain in the main chamber. These composition and fluid transients result in unsteady shear layer reactions of varying intensities that promote the recurrent formation of flame pockets and fuel filaments. Although transient flow effects are pronounced near the tube exit, an elongated primary reaction zone is observed for the central tube, the reaction zone being approximately 30% wider in the transverse direction where neighboring tubes are further apart. In addition, streamlines and velocity quiver plots highlight flow asymmetries that interact with heat release in primary flame reaction zones for near-wall tubes, while the central region maintains a predominantly uniform flow, resulting in securely anchored flames during adjacent flame oscillations. The study emphasizes that collective flame dynamics, rather than isolated local behavior, is key to achieving stable operation in such tubular bundled burners.

采用高保真仿真方法分析了氢燃料下多管微混合器（MTM）燃烧室内火焰的混合和稳定特性。在这种流动结构中，一束管道将燃料-空气混合物送入燃烧室，导致多个火焰前缘，这些火焰前缘可能以不稳定的方式相互作用。高分辨率的模拟启用自适应网格细化与详细的动力学使用。首先用甲烷-氢混合燃料和纯氢燃料的实验数据对结果进行了验证，结果与实验图像数据吻合良好。对氢气情况的详细分析表明，全球火焰结构在不同的流中显示出不同的形状。上游射流的相互作用引起了沿管道的混合分层，这种分层被主室中几何诱导的流动应变放大。这些成分和流体瞬态导致不同强度的不稳定剪切层反应，促进火焰袋和燃料丝的反复形成。虽然在管道出口处有明显的瞬态流动效应，但在中心管道处观察到一个拉长的主反应区，在相邻管道相距较远的横向上，反应区宽约30%。此外，流线和速度颤振图突出了近壁管的主要火焰反应区与热量释放相互作用的流动不对称性，而中心区域保持主要均匀的流动，从而在相邻火焰振荡期间安全锚定火焰。该研究强调，集体火焰动力学，而不是孤立的局部行为，是实现这种管状束式燃烧器稳定运行的关键。

{"title":"Three-dimensional analysis of hydrogen fuel effects in multi-tube combustor","authors":"Shuzhi Zhang , Vansh Sharma , Venkat Raman , Tristan T. Shahin , Alexander J. Hodge , Rohan M. Gejji , Robert P. Lucht , Carson D. Slabaugh","doi":"10.1016/j.proci.2025.105790","DOIUrl":"10.1016/j.proci.2025.105790","url":null,"abstract":"<div><div>Flame characteristics, including mixing and stabilization in a multi-tube micromixer (MTM) combustor operating with hydrogen fuel are analyzed using high-fidelity simulations. In this flow configuration, a bundle of tubes issue fuel-air mixture into a combustion chamber, leading to multiple flame fronts, which may interact in an unsteady manner. Highly-resolved simulations enabled by adaptive mesh refinement with detailed kinetics are used. The results are first validated against experimental data for both a methane–hydrogen blend and pure hydrogen fuel, showing very good agreement with experimental image data. A detailed analysis of the hydrogen case reveals that global flame structures display distinct shapes across various flow streams. Upstream jet interactions induce mixture stratification along the tubes that is amplified with geometry-induced flow strain in the main chamber. These composition and fluid transients result in unsteady shear layer reactions of varying intensities that promote the recurrent formation of flame pockets and fuel filaments. Although transient flow effects are pronounced near the tube exit, an elongated primary reaction zone is observed for the central tube, the reaction zone being approximately 30% wider in the transverse direction where neighboring tubes are further apart. In addition, streamlines and velocity quiver plots highlight flow asymmetries that interact with heat release in primary flame reaction zones for near-wall tubes, while the central region maintains a predominantly uniform flow, resulting in securely anchored flames during adjacent flame oscillations. The study emphasizes that collective flame dynamics, rather than isolated local behavior, is key to achieving stable operation in such tubular bundled burners.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105790"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144780513","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Clustering-based data-driven multi-fidelity reduced order modeling of ammonia combustion 基于聚类的数据驱动多保真度氨燃烧降阶建模

IF 5.2 2区工程技术 Q2 ENERGY & FUELS

Proceedings of the Combustion Institute

Pub Date : 2025-01-01 DOI: 10.1016/j.proci.2025.105881

Aysu Özden , Riccardo Malpica Galassi , Francesco Contino , Alessandro Parente

This study presents a multi-fidelity reduced-order modeling (MF-ROM) framework that uses unsupervised clustering to identify and separately model distinct combustion regimes within the training data. This regime-specific approach allows enhancing computational efficiency while maintaining high predictive accuracy. The proposed MF-ROM framework leverages Proper Orthogonal Decomposition (POD) for dimensionality reduction, manifold alignment for optimal data fusion, and Co-Kriging regression to incorporate both high-fidelity (HiFi) and low-fidelity (LoFi) datasets effectively. First, global clustering is applied to segment the design space into combustion regimes, significantly improving MF-ROM accuracy while reducing the number of required HiFi simulations. Additionally, localized clustering is explored within specific subsets of the design space, demonstrating further refinement in predictive performance. Ammonia combustion is selected as the benchmark case because it is carbon-free and a promising candidate for the energy transition. Moreover, its chemical characteristics make it particularly suitable for the clustering-based MF-ROM approach, as they facilitate the generation of a very diverse training dataset. Results show that the clustering-based MF-ROM achieves the same accuracy as the model without clustering with significantly fewer HiFi simulations, leading to a sixfold reduction in computational cost while maintaining predictive reliability. Moreover, local clustering enhances interpolation capabilities, particularly in regions where combustion characteristics exhibit strong variability. A comparative analysis of temperature and NO emissions confirms that the clustering-driven approach improves both accuracy and efficiency, which can lead to an increase in prediction accuracy up to 50%. The methodology offers a scalable and adaptable approach for optimizing MF-ROMs in reacting flows, supporting the development of low-emission combustion technologies.

本研究提出了一个多保真度降阶建模（MF-ROM）框架，该框架使用无监督聚类来识别和单独建模训练数据中的不同燃烧状态。这种特定于体制的方法可以在保持高预测准确性的同时提高计算效率。所提出的MF-ROM框架利用适当的正交分解（POD）进行降维，流形排列进行最佳数据融合，以及Co-Kriging回归来有效地合并高保真度（HiFi）和低保真度（LoFi）数据集。首先，应用全局聚类将设计空间划分为燃烧状态，显著提高了MF-ROM精度，同时减少了所需的HiFi模拟次数。此外，在设计空间的特定子集中探索了局部聚类，展示了预测性能的进一步改进。选择氨燃烧作为基准案例，因为它是无碳的，是一种很有前途的能源转换候选者。此外，它的化学特性使其特别适合基于聚类的MF-ROM方法，因为它们有助于生成非常多样化的训练数据集。结果表明，基于聚类的MF-ROM在保持预测可靠性的同时，计算成本降低了6倍，与没有聚类的模型具有相同的精度，HiFi模拟次数明显减少。此外，局部聚类增强了插值能力，特别是在燃烧特性表现出强烈变异性的区域。对温度和NO排放的对比分析证实，聚类驱动方法提高了精度和效率，可将预测精度提高50%。该方法为优化反应流中的mf - rom提供了一种可扩展和适应性强的方法，支持低排放燃烧技术的发展。

{"title":"Clustering-based data-driven multi-fidelity reduced order modeling of ammonia combustion","authors":"Aysu Özden , Riccardo Malpica Galassi , Francesco Contino , Alessandro Parente","doi":"10.1016/j.proci.2025.105881","DOIUrl":"10.1016/j.proci.2025.105881","url":null,"abstract":"<div><div>This study presents a multi-fidelity reduced-order modeling (MF-ROM) framework that uses unsupervised clustering to identify and separately model distinct combustion regimes within the training data. This regime-specific approach allows enhancing computational efficiency while maintaining high predictive accuracy. The proposed MF-ROM framework leverages Proper Orthogonal Decomposition (POD) for dimensionality reduction, manifold alignment for optimal data fusion, and Co-Kriging regression to incorporate both high-fidelity (HiFi) and low-fidelity (LoFi) datasets effectively. First, global clustering is applied to segment the design space into combustion regimes, significantly improving MF-ROM accuracy while reducing the number of required HiFi simulations. Additionally, localized clustering is explored within specific subsets of the design space, demonstrating further refinement in predictive performance. Ammonia combustion is selected as the benchmark case because it is carbon-free and a promising candidate for the energy transition. Moreover, its chemical characteristics make it particularly suitable for the clustering-based MF-ROM approach, as they facilitate the generation of a very diverse training dataset. Results show that the clustering-based MF-ROM achieves the same accuracy as the model without clustering with significantly fewer HiFi simulations, leading to a sixfold reduction in computational cost while maintaining predictive reliability. Moreover, local clustering enhances interpolation capabilities, particularly in regions where combustion characteristics exhibit strong variability. A comparative analysis of temperature and NO emissions confirms that the clustering-driven approach improves both accuracy and efficiency, which can lead to an increase in prediction accuracy up to 50%. The methodology offers a scalable and adaptable approach for optimizing MF-ROMs in reacting flows, supporting the development of low-emission combustion technologies.</div></div>","PeriodicalId":408,"journal":{"name":"Proceedings of the Combustion Institute","volume":"41 ","pages":"Article 105881"},"PeriodicalIF":5.2,"publicationDate":"2025-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145262471","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0