【作者】 王洪金；

【导师】 龚光彩；

【作者基本信息】 湖南大学 ， 供热、供燃气、通风及空调工程， 2010， 硕士

【摘要】 空气源换热器作为一种能够一机多用，且相对安装便捷的冷热源设备在我国冬冷夏热地区广泛应用。然而在以这一区域，空气源热泵在冬季运行过程中受到当地低温高湿环境的影响，其室外换热器表面会严重结霜。尽管目前有大量的研究着眼于解决在该地区冬季工况下热泵结霜的问题，然而只有少数学者注意到在翅片的表面霜层的凝结和生长是不均匀的，这种不均匀是加剧空气阻塞的主要原因。除去部分研究阐明了霜层在平行板上的局部分布或者换热器翅片上的平均分布外，目前没有可靠的数据表明在真实换热器表面霜层的局部分布情况。因此本文着重于模拟在低温高湿空况下，管翅式换热器的局部结霜状况以及换热特性。由于在普通的管翅式换热器的翅片板厚度极薄(10-1mm)且翅片间距仅为2.5mm左右，在数值计算过程中用整场耦合的方法会过分依赖计算机的硬件配置和性能。为此本文采用了分模块迭代的方式进行耦合以及相应的离散方法。与前人的工作不同，本文中所用的模块均为三维模块，且模块各自独立，接触面上的网格也不重叠。为了验证本文所提供的计算方法的准确性本文模拟了前人试验中的案例并与实验数值对比，得到了较好的验证。随后本文利用FLUENT商用计算流体软件对换热器在不同空气入口流速的情况下对相似对照工况进行模拟。得到了在仅有显热交换下翅片表面温度分布以及对流换热系数与速度之间的关系。并且利用了本文的结果检验了通用开源软件EVSIM中对于多排管的管翅换热器不同管排对换热器总的平均换热系数的贡献权重的假设。本文发现，EVSIM中各管排等权重的假设是不妥的。以两排管的翅片换热器为例，按照EVSIM中给出的公式计算，头排的换热热流密度时第二排的1.5-2倍。而且在EVSIM中给出的管排局部对流换热系数，虽然方便与工程应用计算，但是忽略了在多排管换热气中头一排翅片管的换热温差明显大于第二排翅片管的。本文的计算同时说明了如果在严格定义翅片管的局部换热温差后，各排管的局部平局对流系数是非常接近的。在后面的工作中，本文利用FLUENT软件，采用上文中提到的耦合方式以及SIMPLE算法，将霜层的凝结生长过程简化为壁面化学反应的方式计算了霜层在翅片表面的局部分布和生长过程。计算说明，冬季工况下，事实上结霜现象最早出现在制冷剂管壁附近。翅片的温度是从制冷剂管的根部逐渐向外降低的，最终在霜层下达到平衡更多还原

【Abstract】 In this paper, the local frost mass retardition on plat-fin-round-tube exchangers has been simulated numerically with commercial CFD code FLUENT. As it is known, a great difficulty on convergence will be encounerted if the integeral coupling is applied to solve the heat transfer between fins and air, because the fin (about 0.1m thick) is much thinner than the fin spaces. To solve this problem, first, a new kind of numerical method is developed. In this method, the whole model is separated into two 3D individuals:one is air modular where fin temperature is surpposed as known and updated by the data from the second modular---fin modular, where the heat transfer between the ambient and fins are surpposed as known and updated up the data from the first modular. Comparing other existing similar methods, in this model both modular is solved with 3D and have significantly different number of mesh sizes to each othe. And an algorithm is developed to adepte this kind of method.We also compared the numerical results from this method with those experimental data and with those credible numerical ones. We find our method is successful in the cases we tested. And based on this method, we obtained data about the fin local temperature distribution at dry condition and the local heat transfer rate as well as area-wighted average ones at steady condition.The total average heat transfer coefficient of a 2-row heat exchanger is adherened to the one calculated from Webb’s j-factor equations. However, the numerical results show that in a 2-row heat exchanger, heat flux conducted through the first row is almost as twice as that through the second one. And by the same means recorded in the EVSIM, the numerical model has calculated the local average heat transfer rate of each row. And the assumption that each row’s heat transfer rate in a multiple exchanger contributes equally to the total average heat transfer rate is questioned.In the next, a unsteady model of frost growth on fins is set up by abstracting the model into a serials of wall surface reactions. The numerical results shows that frost first forms in the front of tubes, and then extands from the root of the tube to the around. The mass retardation is obvious on the sides of the tube, but much less in the rear of the first row tube. It also finds that the frost grows most fast at the frontier. On the other, it grows a little slower where is covered by frost.更多还原