European Institute For Aviation Training and Accreditation (EIATA), Universidad Rey Juan Carlos, Fuenlabrada, 28943 Madrid, Spain.
Gas Turbine Lab, Massachusetts Institute of Technology, 70 Vassar Street, Cambridge, MA 02139, USA.
Sensors (Basel). 2023 Mar 3;23(5):2803. doi: 10.3390/s23052803.
Power plants, electric generators, high-frequency controllers, battery storage, and control units are essential in current transportation and energy distribution networks. To improve the performance and guarantee the endurance of such systems, it is critical to control their operational temperature within certain regimes. Under standard working conditions, those elements become heat sources either during their entire operational envelope or during given phases of it. Consequently, in order to maintain a reasonable working temperature, active cooling is required. The refrigeration may consist of the activation of internal cooling systems relying on fluid circulation or air suction and circulation pulled from the environment. However, in both scenarios pulling surrounding air or making use of coolant pumps increases the power demand. The augmented power demand has a direct impact on the power plant or electric generator autonomy, while instigating higher power demand and substandard performance from the power electronics and batteries' compounds. In this manuscript, we present a methodology to efficiently estimate the heat flux load generated by internal heat sources. By accurately and inexpensively computing the heat flux, it is possible to identify the coolant requirements to optimize the use of the available resources. Based on local thermal measurements fed into a Kriging interpolator, we can accurately compute the heat flux minimizing the number of sensors required. Considering the need for effective thermal load description toward efficient cooling scheduling. This manuscript presents a procedure based on temperature distribution reconstruction via a Kriging interpolator to monitor the surface temperature using a minimal number of sensors. The sensors are allocated by means of a global optimization that minimizes the reconstruction error. The surface temperature distribution is then fed into a heat conduction solver that processes the heat flux of the proposed casing, providing an affordable and efficient way of controlling the thermal load. Conjugate URANS simulations are used to simulate the performance of an aluminum casing and demonstrate the effectiveness of the proposed method.
发电站、发电机、高频控制器、电池存储和控制单元是当前交通和能源分配网络中的重要组成部分。为了提高这些系统的性能并保证其耐久性，必须将其工作温度控制在一定范围内。在标准工作条件下，这些元件在其整个工作范围内或特定阶段都会成为热源。因此，为了保持合理的工作温度，需要进行主动冷却。制冷可以包括激活依靠流体循环或从环境中抽吸和循环空气的内部冷却系统。然而，在这两种情况下，抽吸周围空气或使用冷却剂泵都会增加功率需求。增加的功率需求直接影响发电站或发电机的自主性，同时也会导致电力电子设备和电池组件的功率需求增加和性能下降。在本文中，我们提出了一种方法来有效地估计内部热源产生的热通量负载。通过准确且经济地计算热通量，可以确定冷却剂的要求，以优化可用资源的利用。基于局部热测量值输入克里金插值器，可以准确计算热通量，同时最大限度地减少所需传感器的数量。考虑到需要对有效的热负荷进行描述，以实现有效的冷却调度。本文提出了一种基于温度分布重建的程序，通过克里金插值器使用最少数量的传感器来监测表面温度。传感器的分配采用全局优化方法，以最小化重建误差。然后将表面温度分布输入到热传导求解器中，该求解器处理所提出的外壳的热通量，提供了一种经济高效的控制热负荷的方法。共轭 URANS 模拟用于模拟铝制外壳的性能，并证明了所提出方法的有效性。
