Development of Ultra-High Speed Detectors to Study the Physics of Turbulence

湍流中的粒子跟踪需要非常高的时间(约100,000赫兹)和空间分辨率成像探测系统(512 x 512像素)。这样的系统以非常高的速率生成数据，需要传输、存储和分析。此外，对许多颗粒的三维跟踪需要由四个摄像机同时进行立体成像。传统的探测器技术无法处理由此产生的大量数据(每个摄像头每秒26 GB的数据)，这些数据需要流入内存并存储以供进一步处理。在这里，我们将开发新一代超高速高分辨率粒子跟踪仪器，用于跟踪大约300分钟的示踪粒子在三维中的运动，时间分辨率为100,000赫兹。这一新系统将比任何现有的空间分辨率相当的传统成像系统快两个数量级。该仪器将基于将在康奈尔大学开发的四个超高速、高分辨率像素阵列探测器(PAD)的立体成像。每个Pad都有1024x512个“智能”像素。独特设计的PAD探测器、数据采集系统和数据分析软件的开发将为高雷诺数湍流的拉格朗日分析带来技术突破。充分发展的湍流无处不在。几乎所有的工程流动和自然产生的流动(如大气、海洋和天体物理流动)都涉及高雷诺数湍流。湍流的输送和混合特性最终影响到我们的日常生活。在从湍流燃烧器或内燃机中的混合和化学反应到大气中污染物(或生物制剂)的传输，甚至是将牛奶“简单”混合到咖啡杯中的系统中，对湍流的理解是必不可少的。显然，研究湍流的许多方面最好的方法是跟踪流体颗粒的运动。到目前为止，在高雷诺数湍流中跟踪粒子的困难使得许多长期的预测不可能得到验证。这一项目是解决这一问题的重要一步。即将开发的探测和分析系统将以每秒10万张照片的速度同时跟踪300个粒子。(一台标准的电视摄像机每秒拍摄30张照片。)这种高速将通过使用像素阵列技术来实现，其中每个像素都具有局部智能。这类数据将使一些重要应用的理论进步成为可能，例如标量混合和云形成。此外，这项技术一旦开发出来，可能会影响到比目前无法想象的更广泛的领域。例如，对颗粒的跟踪是许多环境问题(例如污染物或生物剂扩散)的核心。成功的焊盘制造可能会产生比这项研究所需的更多的传感器芯片。其中许多芯片将提供给更大的社区。

Particle tracking in turbulent flows requires very high temporal (approx. 100,000Hz) and spatial resolution imaging detection systems (512 x 512 pixels). Such a system generates data at very high rates, which need to be transferred, stored, and analyzed. In addition three-dimensional tracking of many particles requires the simultaneous stereoscopic imaging by four cameras. Traditional detector technology cannot handle the resulting onslaught of data ( 26 gigabytes of data per second per camera) that would need to be streamed into memory and stored for further processing. Here, we will develop the next generation, ultra-high speed, high-resolution particle tracking instrumentation for following the motion of approximately 300 minute tracer particles in three dimensions with a temporal resolution of 100,000Hz. This new system will be two orders of magnitude faster than any existing conventional imaging system of comparable spatial resolution. The instrument will be based on stereoscopic imaging with four ultra-high-speed, high-resolution pixel array detectors (PAD) that will be developed at Cornell. Each PAD has 1024x512 "intelligent" pixels. The development of the uniquely designed PAD detectors, the data acquisition system, and the data analysis software will result in a technological breakthrough for the Lagrangian analysis of high Reynolds number turbulent flows. Fully developed turbulence is ubiquitous. Virtually all engineering and naturally occurring flows (e.g., atmospheric, oceanographic, and astrophysical flows ) involve high Reynolds number turbulence. The transport and mixing properties of turbulence ultimately impact our daily lives. In systems from the mixing and chemical reactions in a turbulent burner or internal combustion engine to the transport of pollutants (or bioagents) in the atmosphere, or even the "simple" mixing of milk into a coffee cup, the understanding of turbulence is essential. It is clear that many aspects of turbulence are best studied by following the motion of fluid particles. The difficulty of tracking particles in high Reynolds number turbulence has thus far made it impossible to test many long-standing predictions. This project represents a major step to confront this issue. The detection and analysis system to be developed will simultaneously follow 300 particles at a rate of 100,000 pictures per second. (A standard TV camera takes 30 pictures per second.) This high speed will be achieved by using pixel array technology in which each pixel has local intelligence. Data of this kind will enable theoretical advances in a number of important applications such as scalar mixing and cloud formation. Moreover, the technology, once developed, is likely to impact broader fields than can not be envisioned at this time. For example, tracking of particles is at the heart of a number of environmental problems (e.g., contaminant or bioagent dispersion). A successful PAD fabrication will likely result in more sensor chips than needed for this study. Many of these chips would be made available to the larger community.