清水贵如油，作为有限资源，水资源在过去数十年间经常被浪费，而目前的生产成本则日益提高。在全球许多地区，清洁的水源（尤其是地下水）正在以惊人的速度耗尽。随着地球人口的增长，这一情况只会恶化，并且可能因气候变化而加剧。

根据《经济学人》期刊有关水资源的最新专门报告，到 2050 年，生活在长期缺水的国家或地区的人员比例将从 21 世纪的 8% 剧增到 45%，也就是说将有 40 亿人缺水。

     尽管每个在校学生都知道地球主要由水覆盖，但超过 97% 的都是盐水，需要昂贵的淡化过程才能转化为饮用水。非盐水大约占 3%，其中 70% 是南北极的冰山。因此，除了海洋生命，地球上的所有动物都必须依赖不足 1% 的地球可用水生存。

     在世界上缺水的地区，以及人口高密度地区，缺乏清水已成为痢疾等疾病的主要来源，而这又会导致营养不良。而且儿童营养不良很可能导致终身健康问题，影响人们的劳动能力。据估计，在某些国家或地区痢疾的长期影响可能会损失 4-5% 的 GDP。

     因此，大量科学家正在将他们的注意力转到如何从被污染的水或盐水中生成清洁的饮用水。水净化通常包括几个步骤，基于物理（沙过滤）、化学（氯化）或者甚至是生物（处理池）等方法。

因为计算时间约与流速的平方成反比，因此清华大学的研究人员估计一台装有单核处理器的典型台式计算机要用 460 年才能模拟实验可度量的流速上限。为使模拟扩展至速度约 1 厘米/秒或更低，那么典型的实用设备需要 400 倍或更多的计算时间，也就是 184000 年。并且，要模拟碳纳米管孔径的典型范围将需要 10 到 100 倍的时间，这导致总计算时间超过 100 万年。

按照科学方法，必须研究这一低速区域，以便将模拟与实验进行比较，而不是简单地尝试从高速模拟进行推论。此类推论具有明显问题，因为在低速时可能会发生非线性现象。例如，随着速度的降低，在固体摩擦上出现沾滑现象，而这可能在与碳纳米管直接接触的水层中扮演重要角色，因为水在接近碳纳米管表面时会变为类似于冰的方式。

由于进行此项研究需要非常庞大的计算资源，这远远超过了清华大学团队自己的计算机集群的能力，因此 World Community Grid 以及和您一样的志愿者们可以发挥重要作用，为研究人员贡献原本需要从其他方面获得的更为强大的计算能力。

该项目的成果使我们不仅能够测试 Navier 的推测，因此对纳米级别的流体力学贡献基础知识，而且还能够深入了解如何进一步优化液体流经碳纳米管隔膜和其他形态的纳米级别隔膜的方式。

清华大学的团队非常期望更好地真正了解最适合流速的最佳孔径，这将指导以后高效碳纳米管过滤器隔膜的合成和制作，并且提供新方法来生成廉价的水过滤系统。

Computing for Clean Water
Project Status and Findings:  
Information about this project is provided on the web pages below and by the project scientists on the Computing for Clean Water website. If you have comments or questions about this project, please visit the Computing for Clean Water forum.

Mission
The mission of Computing for Clean Water is to provide deeper insight on the molecular scale into the origins of the efficient flow of water through a novel class of filter materials. This insight will in turn guide future development of low-cost and more efficient water filters.

Significance
Lack of access to clean water is one of the major humanitarian challenges for many regions in the developing world. It is estimated that 1.2 billion people lack access to safe drinking water, and 2.6 billion have little or no sanitation. Millions of people die annually - estimates are 3,900 children a day - from the results of diseases transmitted through unsafe water, in particular diarrhea.

Technologies for filtering dirty water exist, but are generally quite expensive. Desalination of sea water, a potentially abundant source of drinking water, is similarly limited by filtering costs. Therefore, new approaches to efficient water filtering are a subject of intense research. Carbon nanotubes, stacked in arrays so that water must pass through the length of the tubes, represent a new approach to filtering water.

Approach
Normally, the extremely small pore size of nanotubes, typically only a few water molecules in diameter, would require very large pressures and hence expensive equipment in order to filter useful amounts of water. However, in 2005 experiments showed that such arrays of nanotubes allow water to flow at much higher rates than expected. This surprising result has spurred many scientists to invest considerable effort in studying the underlying processes that facilitate water flow in nanotubes.

This project uses large-scale molecular dynamics calculations - where the motions of individual water molecules through the nanotubes are simulated - in order to get a deeper understanding of the mechanism of water flow in the nanotubes. For example, there has been speculation about whether the water molecules in direct contact with the nanotube might behave more like ice. This in turn might reduce the friction felt by the rest of the water, hence increasing the rate of flow. Realistic computer simulations are one way to test such hypotheses.

Ultimately, the scientists hope to use the insights they glean from the simulations in order to optimize the underlying process that is enabling water to flow much more rapidly through nanotubes and other nanoporous materials. This optimization process will allow water to flow even more easily, while retaining sources of contamination. The simulations may also reveal under what conditions such filters can best assist in a desalination process.