Cold spray process additive manufacturing of heat transfer features in liquid cooled computer components

由于功耗增加和硬件封装尺寸限制，最新一代的计算机组件需要高性能的冷却系统。与传统的空气冷却技术相比，使用基于液体的冷却（LBC）系统用于计算机组件的热管理已经证明了改善热性能的显著前景。尽管有这种潜力，但该技术仍处于利基行业的范围内，因此大多数现有的LBC系统的热性能优化效果较差。IEI目前正在为视频卡和图形处理器（GPU）开发LBC系统。该项目旨在使用冷气体动态喷涂（CGDS）作为一种新型增材制造（AM）工艺，用于生产高性能LBC系统中使用的高性能传热特性。以前的工作已经证明了这种方法的可行性。然而，这些研究也强调了在大批量生产环境中将CGDS视为AM技术时的某些关键挑战。特别地，与原材料废料（粉末材料）相关联的成本可能很高，并且通常存在与工艺可靠性/可重复性相关的问题（由于CGDS针翅制造夹具的部件上的不期望的材料积聚）。这些基本的工程问题可能会严重阻碍该技术的经济可行性，因此它们是当前拟议项目的重点。这些问题的成功解决将有助于证明CGDS工艺作为GPU大批量生产环境中的AM技术的可靠性和经济可行性。这可能会给加拿大工业带来巨大的好处，因为它可以将CGDS作为一种新的可行且具有成本效益的AM技术引入，这反过来又可以在本提案中针对的具体应用之外开辟广泛的新可能性。

The latest generation of computer components requires high performance cooling systems due to increased power consumption and hardware packaging size restrictions. The use of liquid based cooling (LBC) systems for thermal management of computer components has demonstrated significant promise for improved thermal performance in comparison to traditional air cooling techniques. Despite this potential, the technology remains within the confines of a niche industry, and as such most existing LBC systems are poorly optimized for thermal performance. IEIis currently developing a LBC system for video cards and graphics processing units (GPUs). The proposed project aims to use Cold Gas Dynamic Spraying (CGDS) as a novel additive manufacturing (AM) process for the production of high performance heat transfer features to be used in those high performance LBC systems. Previous work has already demonstrated the feasibility of this approach. However, these studies have also highlighted certain key challenges when considering CGDS as an AM technique in a high volume production setting. In particular, costs associated to raw material waste (powder material) can be high and issues related to process reliability/repeatability (due to undesirable material buildup on components of the CGDS pin fin fabrication fixture) are often present. These fundamental engineering issues can significantly hinder the economic viability of the technique, and as such they are the focus of the current proposed project. The successful resolution of these issues would be instrumental in demonstrating the reliability and economic feasibility of the CGDS process as an AM technique in a high volume production setting for GPUs. This could have enormous benefits to Canadian industries as it could introduce CGDS as a new viable and cost-effective AM technique, which could in turn open up a broad range of new possibilities beyond the specific application targeted in this proposal.