In various applications with recirculation, liquid droplets can be immersed in gases that may have a wide range of possible compositions, from pure air to hot combustion products. In order to gain fundamental understanding of the behaviour of individual droplets in vitiated air, numerical simulations of kerosene single droplet evaporation, autoignition, and combustion in conditions relevant to gas turbines have been performed. The droplet autoignition behaviour has been analysed in both physical and mixture fraction space for a wide range of vitiated air compositions and initial droplet diameters. Results show that the autoignition time delay decreases with increasing level of dilution with hot combustion products and decreasing initial droplet diameter. Chemistry is confined up to a radius of almost 10 initial droplet diameters and the location of autoignition is influenced by both the initial droplet diameter and the level of dilution. The time evolution of species in the gaseous phase after autoignition shows similar trends for all the diameters and dilution levels investigated here with the peak of temperature and OH mass fraction moving towards the droplet surface as a consequence of the balance between fuel production and consumption. In mixture fraction space, the location of the peaks of temperature and OH mass fraction after autoignition do not change in time whereas other intermediate species such as CH 2 O and pyrolysis products still exhibit a quite variable behaviour. The long-time flame structure has been compared with gaseous laminar counterflow simulations and, although qualitatively similar, the flame structure in the two configurations has differences with implications for flamelet combustion models used in spray combustion. The droplet evaporation, autoignition, and combustion behaviour has been summarized through a regime diagram showing the evaporation and autoignition time delays as a function of both initial droplet diameter and vitiated air dilution. This allows the identification of different states in the droplet combustion scenario and the introduction of critical values of dilution and initial droplet diameter beyond which single droplet rather than cloud combustion can occur, which can be exploited in the design of lean burn gas turbine combustion systems. In conclusion, the approach presented here can be easily extended to other conditions and fuels allowing the generation of regime diagrams for various operating conditions.

42 ENGINEERING↗

DOE OSTI2017

Scalability of Several Asynchronous Many-Task Models for In Situ Statistical Analysis.

This report is a sequel to [PB16], in which we provided a first progress report on research and development towards a scalable, asynchronous many-task, in situ statistical analysis engine using the Legion runtime system. This earlier work included a prototype implementation of a proposed solution, using a proxy mini-application as a surrogate for a full-scale scientific simulation code. The first scalability studies were conducted with the above on modestly-sized experimental clusters. In contrast, in the current work we have integrated our in situ analysis engines with a full-size scientific application (S3D, using the Legion-SPMD model), and have conducted nu- merical tests on the largest computational platform currently available for DOE science ap- plications. We also provide details regarding the design and development of a light-weight asynchronous collectives library. We describe how this library is utilized within our SPMD- Legion S3D workflow, and compare the data aggregation technique deployed herein to the approach taken within our previous work.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2017

A Unified Data-Driven Approach for Programming In Situ Analysis and Visualization: An Interim Report of Sandia Sub-Team Contributions

As we look ahead to next generation high performance computing platforms, the placement and movement of data is becoming the key-limiting factor on both performance and energy efficiency. Furthermore, the increased quantities of data the systems are capable of generating, in conjunction with the insufficient rate of improvements in the supporting I/0 infrastructure, is forcing applications away from the off-line post-processing of data towards techniques based on in ,situ analysis and visualization. Together, these challenges are shaping how we will both design and develop effective, performant and energy-efficient software. In particular, the challenges highlight the need for data and data-centric operations to be fundamental in the reasoning about, and optimization of, scientific workflows on extreme-scale architectures.

97 MATHEMATICS AND COMPUTING↗