No abstract

Increasing demands for cloud-based computing and storage, Internet-of-Things, and machine learning-based applications have necessitated the utilization of more efficient cooling technologies. Direct-to-chip liquid cooling using cold plates has proven to be one of the most efficient methods to dissipate the high heat fluxes of modern high-power CPUs and GPUs. While the published literature has well-documented research on the thermal aspects of direct liquid cooling, a detailed account of transient hydraulic investigation is still missing. In this experiment, a total of four 52U racks with four high-power TTV-servers (Thermal Test Vehicles) in each rack were designed and deployed. Each server consists of eight GPU TTVs and six NV switch heaters. Each of the two racks has a different vendor rack manifold and cooling loop modules (CLM). A 450 kW coolant distribution unit (CDU) is used to supply 25% propylene glycol coolant to these racks. Each rack has its own rack-level flow control valve to maintain the same flow rate. The present study provides an in-depth analysis of hydraulic transients when rack-level flow control valves are used with and without flow control. The operating conditions of the CDU are varied for different parameters, such as a constant flow rate, constant differential pressure, and constant pump speed. Furthermore, hydraulic transient is examined when the cooling loop modules are decommissioned from the rack one by one. The effect of this step-by-step decommissioning is assessed on the CDU operation and other racks. The pressure drop-based control strategy has been developed to maintain the same flow rate in the remaining servers in the rack when some cooling loop modules are decommissioned.

Detailed study of material compatibility of the various electronics packaging materials for immersion cooling is essential to understand their failure modes and reliability. The modulus and thermal expansion are critical material properties for electronics mechanical design. Substrate is a critical component of electronic package and heavily influences failure mechanism and reliability of electronics both at the package and board level. This study mainly focuses on two major challenges. The first part of the study focuses on the impact of thermal aging in dielectric fluid for single-phase immersion cooling on the non-halogenate substrate’s thermo-mechanical properties. The second part of the study is the impact of thermal aging on thermo-mechanical properties of substrate in the air. The non-halogenated low Coefficient of Thermal Expansion (CTE) bismaleimide triazine (BT) resin laminate is used for its ultra-low CTE which in turn reduce the warpage of substrate. Moreover, the substrate has high glass transition temperature and high stiffness suitable for the application which requires high heat resistance. The substrate is aged in ElectroCool EC100 dielectric fluid, and air for 720 hours at three different temperatures: 22°C, 50°C, and 75°C. The complex modulus is characterized before and after aging for both parts and compared.

Computational modeling is often leveraged to design and optimize electronic packages for both performance and reliability purposes. One of the factors that affect the accuracy of computational models is the accuracy of the material properties. Microelectromechanical system sensors, in particular, are usually extremely sensitive to slightest material property changes in the package. Therefore, even small measurement variations in material characterization due to different sample preparation methods or different testing techniques can impact accuracy of computational models that are leveraged for designing or analyzing sensor performance. The challenge in material characterization is even greater for materials that require curing. Die attach polymers, for example, have strict curing profile requirements that are used during the manufacturing process. Such curing conditions are usually hard to duplicate in laboratories, and the samples used for material characterization may not necessarily be representative of the actual component in the final product. In this study, the effect of parameters such as temperature curing profile, application of pressure during curing, and sample preparation technique on temperature-dependent thermomechanical properties of two types of die attach elastomers is investigated. The mechanical properties, including the elastic modulus (E), coefficient of thermal expansion, and glass transition temperature of the die attach material, are measured using a suite of techniques such as dynamic mechanical analysis and thermomechanical analysis. The analysis is performed for a wide temperature range corresponding to typical sensor applications. It is shown that sample preparation and characterization techniques have a considerable impact on the measurements, which results in different MEMS sensor performance predictions through computational modeling.