1、 Dr. Nicole Okamoto is a professor in the Department of Mechanical Engineering, San Jos State University, San Jos, CA. Dr. Andrew Sommers is an associate professor in the Department of Mechanical and Manufacturing Engineering, Miami University, Oxford, OH. Isaac Tineo and Jonathan Carlson are studen
2、ts at San Jos State University, and Christian Petty is an undergraduate student at Miami University. Dean DiBlasio, a recent SJSU graduate, is a mechanical design engineer intern at Verselus, LLC. Using Patterned Surface Wettability for Improved Frosting/Defrosting Performance Nicole Okamoto, PhD Is
3、aac Tineo Dean DiBlasio Jonathan Carlson Christian Petty Andrew Sommers, PhD Associate Member ASHRAE ABSTRACT The goal of this project was to assess the potential benefits of using a highly-controlled surface wettability to preferentially condense and locate water droplets on a heat transfer surface
4、 during the early stages of frost growth and to more completely drain melted frost during a defrost cycle. A thinner, denser frost layer may lead to improved air-side heat tranfer, and the retention of less water following defrosting might also be used to slow frost growth and lengthen the operation
5、al cycle. A baseline aluminum surface was compared with a surface with an ultra-thin hydrophobic coating, three surfaces laser etched to produce a surface microtexture in addition to the hydrophobic coating, and a fourth that was micro-milled. The results have been mixed. Contact angle measurements
6、and spray testing data have shown some benefit and the possibility for preferred drainage patterns on the enhanced surfaces. During the frosting cycles, each surface exhibited a uniform layer of frost growth. In both the second and third frosting cycles, frozen water droplets increased the retained
7、mass for all surfaces. Photographs of the frosting surfaces show water droplet retention along preferential patterns for some of the laser-etched surfaces. However, frost thickness and mass measurements do not present conclusive evidence that preferred drainage patterns improved frosting performance
8、. INTRODUCTION This study was aimed at decreasing energy costs through the development of new heat transfer fin surface designs for use in liquid-to-air heat exchangers such as those used in a range of air-cooling applications. In refrigeration systems, when the cold surface of the heat exchanger is
9、 exposed to warm humid air, condensate droplets form on the surface during early stages of frost growth. This study sought to create surfaces with an underlying surface wettability pattern or gradient that could be used to preferentially condense (and therefore locate) water droplets in desirable re
10、gions on the surface and to more completely drain the melted frost layer during a defrost cycle. Additionally, a thinner, denser frost layer might lead to improved heat conduction (and therefore improved air-side heat tranfer) and longer operational periods before defrosting became necessary. Decrea
11、sed water retention can also slow frost growth in the subsequent cycle. Water droplets also provide a site for the growth of micro-organisms such as Legionellosis bacteria. Topography-based approaches to this surface wettability modification were emphasized in this study since these techniques shoul
12、d be more robust than chemistry-based approaches which often wear off in application. Although extensive research has been reported on using topography and chemistry to alter surface wettability in other applications, there has not yet been a comprehensive study on the use of the surface wettability
13、 patterns and gradients in refrigeration systems where the energy impact (with regards to frosting and defrosting) could potentially be quite large. Although many studies have been performed to model frost growth and measure the properties of the frost layer (i.e. White and Cremers (1981), Ogawa et
14、al. (1993), Cheng and Cheng (2001), etc.), relatively few papers were found which examined the effect of the surface wettability on the growing frost layer. Somlo and Gupta (2001) prepared a weakly hydrophobic 6061 aluminum alloy surface through a dipping process involving dimethyl-n-octadecilcholor
15、osilane (DMOCS) and studied the tensile strength of the ice/surface interface. Shin et al. (2003) found that during the initial period of frost formation the shape of the micro droplets depended upon the surface energy and the process of frost growth was affected by the dynamic contact angle (DCA).
16、Although the technical literature is replete with articles describing frost properties, frost growth models, and the impact of frost on air-side heat transfer, few papers were found which specifically address the effect that surface wettability has on these items. Most examined only a very narrow ra
17、nge of contact angles or were limited to hydrophilic surfaces with DCA less than or equal to 90. Conflicting information was found in the literature regarding the effect that hydrophilic/hydrophobic coatings have on frost properties. Most of these surfaces were modified using surface coatings which
18、poses longevity concerns and were not tested following multiple cycles. The other novel aspect of this work was the use of surface tension gradients for frost management. One of the earliest works involving surface tension gradients was by Chaudhury and Whitesides (1992) who were able to induce the
19、upward motion of small water drops on tilted silicon surfaces. Daniel et al. (2001) later observed the movement of water droplets on a surface with a radial surface tension gradient. The authors reported that an increase in velocity (over earlier studies) was due to the coalescence of droplets. Shas
20、try et al. (2006) described a rough superhydrophobic surface (produced in silicon) with a contact angle gradient. Droplets were propelled down these gradients by mechanical vibration. These findings suggest that wettability gradients (based solely in topographical variation) might be used to facilit
21、ate drainage prior to the start of frost growth. Although the use of surface tension gradients has received considerable interest in recent years, the authors are not aware of any study that has specifically examined the use of surface tension gradients as a potential means of affecting the properti
22、es of a frost layer and/or improving defrosting effectiveness. In summary, the objectives of this project were to investigate new ways of both (i) influencing frost growth and frost properties during the condensation period by controlling the initial distribution of water on the surface, and (ii) im
23、proving water drainage during the defrost cycle through the creation of preferential drainage paths on the surface using regions of high hydrophilicity and hydrophobicity. Accompanying this objective was the goal of using surface tension gradients and/or wettability patterns to create a water drople
24、t driving force on the surface to steer droplets to “preferred” locations on the surface. In application, this could be especially desirable during the early stages of frost growth to help reduce the rate of frost accumulation and blockage in the most critical areas of the heat exchanger. SURFACE PR
25、EPARATION Laser etching and micro-milling were chosen as the methods of manufacturing the surface microtexture, and plasma-enhanced chemical vapor deposition (PECVD) was used for depositing an ultra-thin hydrophobic coating on the surface. The surfaces that were chosen for experimentation were the l
26、inear surface tension gradient and radial surface tension gradient along with a parallel microchannel design. The gradient was intended to increase the coalescence of tiny water droplets through the creation of a net surface tension force that would allow for the possibility of spontaneous water dro
27、plet movement on the surface (Sommers et al. 2013). The parallel micro-groove structure was chosen because it was expected that this surface structure would facilitate improved water drainage while simultaneously preventing condensate blow-off. In addition, parallel channels were expected to be easi
28、er to manufacture than other geometries using current industrial practices. Although uniform posts were also considered, posts might be more susceptible to cracking in refrigeration systems due repeated “freeze-thaw” cycles. The test surfaces were produced from plates of aluminum alloy 5052 having a
29、 brushed finish. Plate dimensions were 3.125” 3.875” 0.125” (7.9 9.8 0.32 cm). As shown in Table 1, Sample 1 was an unmodified Al 5052 plate and served as a baseline. Sample 2 contained a hydrophobic coating but was otherwise identical to Sample 1. Sample 3 contained variably spaced, micron-sized ch
30、annels which formed a radial surface tension gradient pattern. Each circle in the pattern was constructed with a surface tension gradient to create a net force that might propel water droplets towards the center of the circle and in doing so promote droplet coalescence (and thus drainage). Sample 4
31、also contained variably spaced channels, but the channels formed a linear gradient along the surface. Channels were variably spaced from 25 m to 800 m to create the net surface tension force. One complete wettability zone contained two gradient patterns, each a mirror image of the other. In this way
32、, droplets were propelled from both directions along the gradient towards a central “drainage duct” on the surface. It was also thought that two droplets might meet in these regions and coalesce thereby facilitating the drainage process. Sample 5 was designed to have the same raised contact area as
33、Sample 4 but without the gradient. Sample 5 channels were spaced 100 m apart. Sample 6 contained nearly the same topographical gradient as Sample 4 but was manufactured using micro-milling techniques instead of laser etching. The width of the channels for Samples 3-6 was fixed at 25 m. All the surfa
34、ces (minus Sample 1) contained a 100-125 nm (3.910-6 - 4.910-6 in) thick PTFE-like hydrophobic coating that was deposited via plasma-enhanced chemical vapor deposition (PECVD). Three different criteria were then used to evaluate the effectiveness of the new fin surface designs: (a) measurement of wa
35、ter droplet contact angles on all fabricated samples, (b) measurement of the water drainage characteristics during spray tests, and (c) measurement of frost thickness and frost mass accumulation during cyclic frosting experiments. Table 1. Surface Fabrication Matrix Sample Use of Gradient? Type of G
36、radient Fabrication Method Surface Details 1 N - - Baseline aluminum 2 N - - Baseline aluminum with hydrophobic coating 3 Y radial laser etching Radial design with a topographical surface tension gradient 4 Y linear laser etching Parallel micro-channels with a topographical surface tension gradient
37、5 N - laser etching Parallel micro-channels (evenly spaced) 6 Y linear micro-milling Parallel micro-channels with a topographical surface tension gradient CONTACT ANGLE MEASUREMENTS AND SPRAY TESTING Both static and dynamic contact angles were measured on the prepared surfaces using a Ram-Hart preci
38、sion contact angle goniometer. The stated accuracy of this equipment was 0.1. The overall total uncertainty associated with these measurements was higher due to other sources of possible error (i.e. repeatability, evaporative losses, etc.). For the advancing contact angle, the average standard devia
39、tion for all samples was 4.0 while for the receding contact angle, the average standard deviation for all the samples was 4.9 For these tests, water droplets were injected on the surface using a high precision micro-syringe. The advancing and receding angle were then measured by either increasing or
40、 decreasing the volume of the water droplet on the surface using standard goniometer testing procedures. The contact angle that a liquid droplet forms on a horizontal surface is described by the classical equation by Young (1855). These measured data are shown below in Table 2. These data show that
41、Sample 1 (baseline) and Sample 3 had the highest contact angle hysteresis values (i.e. 108.7and 111.3 respectively), while Sample 5 had the lowest contact angle hysteresis (i.e. 12.3) of the tested samples. Sample 4 also possessed very low contact angle hysteresis at the hydrophobic end of the gradi
42、ent pattern (i.e. 21.8). Contact angle (CA) hysteresis is the absolute difference between the advancing and receding contact angles and is often used as a gauge of hydrophobicity. Large contact angle hysteresis implies a large affinity for water retention. Additionally, the PTFE-like coating was ben
43、eficial in reducing the CA hysteresis. The CA hysteresis for Sample 2 was more than 20less than Sample 1. Nonetheless, both of these values are large suggesting a strong affinity for water Table 2. Surface Contact Angle Data Sample CENTER LOCATION EDGE LOCATION Static CA Advancing CA Receding CA Sch
44、ematic Static CA Advancing CA Receding CA Schematic 1 109.6 (n = 28) 134.4(n = 29) 25.7(n = 28) - N/A - - N/A - 2 132.1(n = 21) 157.5(n = 22) 69.5(n = 27) - N/A - - N/A - 3 137.7 (n = 98) 155.9 (n = 27) 44.6 (n = 25) 164.1 (n = 48) - - 4 150.0 (n = 30) 166.9 (n = 26) 71.5 (n = 26) 170.8 (n = 27) 172
45、.1 (n = 26) 150.3 (n = 26) 5 172.4(n = 30) 173.1(n = 31) 160.8(n = 28) - N/A - 6 136.2 (n = 38) 165.8 (n = 25) 85.1 (n = 25) 152.8 (n = 40) 164.0 (n = 29) 132.1 (n = 28) retention. The micro-scale roughness of Samples 3-5 served to reduce contact angle hysteresis. Also, as expected, the CA hysteresi
46、s on the gradient surfaces varied from the “hydrophobic” end to the “hydrophilic” end of the gradient design. On Sample 4, the CA hysteresis was 21.8 on the “hydrophobic” end while it was 95.4 on the “hydrophilic” end of the gradient. This suggests the existence of a strong surface tension driving f
47、orce for droplet motion. Finally, these data suggest that Sample 5 should be the best performer in terms of overall water drainage. It had the highest overall contact angle of all the surfaces (i.e. most hydrophobic), but it also possessed the lowest CA hysteresis. Spray tests were performed to asse
48、ss the water drainage characteristics of the surfaces and to determine whether preferential drainage was achieved. A spray bottle was used to create a fine mist of water droplets to wet the surface. Figure 1 shows representative images. Water collected as designed in specific locations on the gradie
49、nt surfaces (3, 4, and 6) but showed no such preference on the baseline surfaces (1 and 2) as well as Sample 5 which did not contain a gradient. On Sample 3, the water collected in the center of the circles (hydrophilic end of the radial gradient) as well as the large regions between the circles. On Sample 4, the water collected at the end of the linear gradient in designated “drainage ducts” as shown in Fig. 1. Droplet movement and coalescence were responsible for the large droplets tha
