Desalination 280 (2011) 363–369
Contents lists available at ScienceDirect
Desalination j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / d e s a l
Experimental investigation of dropwise condensation on hydrophobic heat exchangers. Part II: Effect of coatings and surface geometry Jorge R. Lara ⁎, Mark T. Holtzapple Department of Chemical Engineering, Texas A&M University, 3122 TAMU, College Station, TX 77843-3122, USA
a r t i c l e
i n f o
Article history: Received 9 March 2011 Received in revised form 11 July 2011 Accepted 12 July 2011 Available online 9 August 2011 Keywords: Dropwise condensation Electroless Ni–P–PTFE Hydrophobic heat exchangers Vertical-grooved heat exchanger sheets Vapor-compression desalination
a b s t r a c t This is Part II of an experimental investigation of hydrophobic heat exchangers. Two plates were studied: (a) 0.127-mm-thick titanium grade 2 and (b) 0.203-mm-thick copper. Titanium plate had round-dimpled spacers. Copper had either round-dimpled spacers or round-shaped vertical-grooved spacers. Titanium was bare but copper had electroless Ni–P–PTFE hydrophobic coating. Two chemical compositions of the hydrophobic coating were employed: lead-containing and lead-free. For some studies, the coating thickness was varied from 0.635 to 127 μm. To measure the overall heat transfer coefficient, the plates were mounted in a sealed two-chamber apparatus with condensing saturated steam on one side and forced-convective boiling liquid water on the other. The best overall heat transfer coefficient was U = 240 kW/(m2·°C) (0.203-mm-thick copper plate, round-shaped vertical grooves, 2.54-μm-thick lead-free Ni–P–PTFE, P = 722 kPa, T = 160 °C, ΔT = 0.20 °C, saturated liquid velocity vliq = 1.57 m/s, shearing steam vsteam = 0.23 m/s, and flow ratio R ≈ 0.6 kg shearing steam/kg condensate). Published by Elsevier B.V.
1. Introduction Part I of this study [1], used a variety of heat transfer enhancement techniques such as dropwise condensation, forced convective boiling, roughening on boiling surface, boiling stones as nucleation agents, and condensation with shearing steam. Condensation on a hydrophobic surface using shearing steam and forced-convective boiling with nucleation agents produced very efficient heat transfer. The low surface energy of titanium promotes dropwise condensation, which increases the heat flux [2]. Furthermore, titanium resists abrasion, corrosion, and fouling. Over time, the resulting overall heat transfer rate of titanium surfaces is often comparable to metals that have higher thermal conductivity. The literature [3] suggests that compared to flat surfaces, vertical grooves deliver about 25% higher overall heat transfer coefficients. For dropwise condensation, liquids form microscopic droplets on the condensation surface, followed by droplet growth, coalescence/ growth, and downflow. The thermal resistance of liquids attached to the metal surface dominates dropwise condensation [4]; rapidly shedding liquid droplets is an important factor that increases heat flux [5,6]. Round-shaped vertical grooves on the condensing surface help channel the condensing steam so it sheds quickly, which increases the heat flux [3,7,8]. This study measures the effects of shearing steam on the overall heat transfer coefficient of hydrophobic surfaces of thin
⁎ Corresponding author. E-mail address:
[email protected] (J.R. Lara). 0011-9164/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.desal.2011.07.017
copper plates. Electroless Ni–P–PTFE is employed as an economical and robust hydrophobic coating. Table 1 exhibits a literature review of previous work with electroless Ni–P–PTFE hydrophobic coatings on vertical plates. Past studies focused on characterizing the coating resistance to corrosion and fouling [9–17]. The overall heat transfer coefficients reported were very low compared to the results of the present study. None of the experiments reviewed was conducted with high-pressure steam. In this review, the highest measured overall heat transfer coefficient was U = 17 kW/(m2·°C) at comparatively large ΔT (2 to 7 °C) and low P (200 kPa) for a corrugated plate-and-frame heat exchanger coated with electroless Ni–P–PTFE. Desalination technologies for municipal drinking water must meet NSF STD 61 certification. Lead is a common contaminant in most Ni–P–PTFE hydrophobic coatings. To overcome this problem, lead-free chemistry should be employed in systems that produce drinking water [18]. In this paper, the first study quantifies heat transfer in titanium plates with round dimples. The second study uses copper with round dimples; the thickness of the lead-containing Ni–P–PTFE coating was varied. The third study uses copper with vertical grooves with leadcontaining Ni–P–PTFE coatings. The fourth study is similar to the third, but it employs lead-free Ni–P–PTFE. 2. Materials and methods Experiments were conducted using bare 0.127-mm-thick titanium grade 2 and 0.203-mm-thick copper coated with either lead-containing or lead-free electroless Ni–P–PTFE coating.
364
J.R. Lara, M.T. Holtzapple / Desalination 280 (2011) 363–369
NR
NR NR
NR
No
Yes
No
Yes
NR NR High wear resistance NR
Yes 15 × 10 Vertical coupon No No No Ni–Cu–PPTFE 0.35
23
Ni–P–PTFE Ni–P–PTFE
ð2Þ
U q m hfg A ΔT
overall heat transfer coefficient (kW/m 2·°C) heat flux (kW/m 2) condensate collected from the apparatus (kg/s) latent heat of condensation (kJ/kg) effective heat transfer area = 0.0645 m 2 temperature differential across the plate (°C)
The first plate was round-dimpled 0.127-mm-thick titanium grade 2 (k = 22 W/(m·°C)) with chemical composition: carbon 0.80% max., nitrogen 0.03% max., oxygen 0.25% max., iron 0.30% max., hydrogen 0.015% max., titanium balance [19]. The condensing metal surface was bare. The plates were 305 mm × 305 mm. One hundred equally distributed round dimples (19.1-mm diameter and 3.18-mm deep separated by 25.4-mm centers) were formed on each plate. Because the mounting mechanism blocked some of the plate, the effective heat transfer area was 254 mm × 254 mm or 0.0645 m 2. The second plate was 0.203-mm-thick copper (k = 400 W/(m·°C)) with round dimples. Both plate surfaces were modified with leadcontaining Ni–P–PTFE hydrophobic coatings of different thicknesses. The third plate was 0.203-mm-thick copper (k=400 W/(m·°C)) with round-shaped vertical grooves. The plates were 305 mm×305 mm. Twenty-seven equally distributed round grooves (8-mm diameter and 3.18-mm deep) were formed on each plate. The effective heat transfer area was 254 mm ×254 mm or 0.0645 m2. Both plate surfaces were modified with a 0.635-μm-thick lead-containing Ni–P–PTFE hydrophobic coating by Micro Plating, Inc. (Erie, PA). In this study, two hydrophobic coatings were used; one uses lead acetate as stabilizer in the electroless bath and the other is lead-free. Table 2 shows the chemical composition of lead-containing Ni–P–PTFE [17] compared to lead-free Ni–P–PTFE [18]. The fourth plate was 0.203-mm-thick copper with round-shaped vertical grooves that was coated with lead-free 2.54-μm-thick Ni–P–PTFE.
Cu 100 NR [16]
100 100 NR NR [14] [15]
200 NR [13]
100 [12]
NR
SS 304 SS 316
NR NR
NR 1.28−23
No No
No No
No No
Corrugated plate-and frame HX Rotating cylinder Common heat exchangers No No No NI–P–PTFE 0.35 Cu SS 304
NR
No No No Ni–Cu–PPTFE 1.0 SS 304
23
17
NR Ni-P-PTFE improved processing skim milk and tomato juice NR
Yes
Yes
Minimized microbial adhesion by over 96−98% Yes NR
Yes 17 No Yes (Oil) No Ni–P–PTFE 200 2−7 [11]
Pool boiling 101.3 [10]
NR
SS 316
NR
NR
No No No Ni–P–PTFE 0.35 Cu SS 304
Various
No No No 0.254−2.54 0.35 Cu SS 304 LC steel Pool boiling 101.3 [9]
NR
Ni–P–PTFE
0.35 Cu SS 304 Pool boiling 101.3
Ni–Cu–PPTFE
=A
Table 2 Bath composition for electroless Ni–P–PTFE hydrophobic coating.
NR [8]
and
2.2. Test surfaces
Base metal thickness (mm)
23
No
No
No
15 × 10 vertical coupon 15 × 10 vertical coupon 15 × 10 vertical coupon Corrugated plate-and frame HX 15 × 10 vertical coupon
5
Inhibited formation of CaSO4 scale Reduced adhesion of CaSO4 scale Yes NR
NR
NR
Superior to Ni-P-PTFE Coating Yes Yes 7
No
Dropwise condensation (DWC) Corrosion resistance Antifouling Heat transfer coefficient (kW/(m2 °C)) Surface geometry (mm) Nucleation sites Liquid side agitation Shear steam
ð1Þ
where:
Substrate metal
Coating thickness (μm)
!q" ΔT
! " q = mhfg
Psat (kPa)
Coating
The plates were placed in a two-chamber apparatus described in Part I [1]. Measured overall heat transfer coefficients U are obtained from U=
ΔT (°C) Ref no.
Table 1 Literature review of Ni–P–PTFE hydrophobic coatings.
2.1. Calculation of heat transfer coefficients
Lead-containing Ni–P–PTFE [17]
Lead-free Ni–P–PTFE [18]
NiSO4 · 6H2O 25 g/L NaH2PO2 · H2O 30 g/L a Na3C6H5O7 · 2H2O 18 g/L Sodium acetate 18 g/L (CH2) CS 1 ppm Lead acetate (stabilizer) 3 ppm PTFE (60 wt.%) 10 mL/L C20H20F23N2O4I (FC-4) 0.4 g/L pH 4.8 T (°C) 88
NiSO4 · 6H2O 30 g/L NaH2PO2 · H2O 30 g/L Lactic acid 25 mL/L Sodium acetate 10 g/L Accelerator 4 g/L KIO3 (stabilizer) 5 ppm PTFE (60 wt.%) 4–50 mL/L
a
Sodium citrate.
pH 4.6–5 T (°C) 90 ± 2
365
J.R. Lara, M.T. Holtzapple / Desalination 280 (2011) 363–369 Condenser Pressure (kPa)
80
25
70
827 (Projected) 60
20
Heat Flux q (kW/m2)
Overall Heat Transfer Coefficient (kW/(m2 . °C))
90
50 40 30 Condenser Pressure (kPa)
20
722 550 432
10
722 15
550 10
432 5
0 0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
Differential Temperature across Plate ∆T (°C)
0 0
Fig. 1. Overall heat transfer coefficients for 0.127-mm-thick titanium grade 2. Forced convection boiling vliq = 1.57 m/s, shearing steam vsteam = 0.15 m/s. Optimal shearing steam on the condensing surface (see Fig. 3). Smooth curves are given by Eq. 3.
In the lead-free coating bath, it was observed that PTFE precipitated on the surface. The reason is unknown; however, this effect allowed more PTFE particles to deposit at the surface, which enhanced the hydrophobic effect of the coating.
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Differential Temperature across Plate ∆T (°C) Fig. 2. Heat flux for 0.127-mm-thick titanium grade 2. Forced-convection pool boiling. Optimal shearing steam from Fig. 3. Curves were created using Eqs. 4 and 5. Dashed line is a projection to desired operating pressure.
As described in Part I [1], for a given condenser pressure P, the empirical equation β
U = α ðΔT Þ 3. Results for titanium Fig. 1 shows the overall heat transfer coefficients for titanium as a function of temperature differences across the plate at different saturated steam pressures. The overall heat transfer coefficient U increases at high pressures and small ΔT. Forced convection is imposed in the saturated liquid side with vliq = 1.57 m/s. The best overall heat transfer coefficient was U = 78 kW/(m2·°C) (P = 722 kPa, T = 166 °C, ΔT = 0.20 °C, saturated liquid velocity vliq = 1.57 m/s, shearing steam velocity vsteam = 0.12 m/s, and flow ratio R = 1.5 kg shearing steam/kg condensate). (Note: The shearing steam velocity vsteam is reported at the exit of the heat exchanger.) After the experiments, the surface exposed to boiling water was in good condition.
ð3Þ
describes each of the curves in Fig. 1. Table 3 reports the parameters α and β. As observed, α and β are both P-dependent and ΔT-dependent. To calculate the heat flux, the empirical equation is 1+β
q = UΔT = α ðΔT Þ
ð4Þ
Fig. 2 shows heat flux q across the plate for different ΔT and P. Fig. 3 shows variations of the overall heat transfer coefficient with shearing steam for a given ΔT. For each condition, there is an optimal shearing steam velocity (or flow ratio R). The maximum overall heat transfer coefficient measured was U = 78 kW/(m 2·°C) (P = 722 kPa, T = 166 °C, vliq = 1.57 m/s, vsteam = 0.12 m/s, and R = 1.5 kg shearing steam/kg condensate).
Table 3 Empirical constants. U = kW/(m2 °C) ΔT = °C P = kPa q = kW/m2
U = α(ΔT)β q = α(ΔT)1 + β
Plate description
P
U = a Pb 6
a = ∑ ai ðΔT Þi i=0 6
b = ∑ bi ðΔT Þi i=0
α
β
0.127-mm-thick titanium grade 2 Bare Forced convection (vliq = 1.57 m/s) Shearing steam (vsteam = 0.15 m/s)
722 550 432
17.1 11.2 7.3
–0.906 –0.592 –0.703
0.203-mm-thick copper with vertical grooves Coated Lead-containing Ni–P–PTFE (0.635 μm) Forced convection (vliq = 1.57 m/s) Shearing steam (vsteam = 0.16 m/s)
722 646 515
40.3 32.2 24.3
–0.955 –0.932 –0.562
0.203-mm-thick copper with vertical grooves Coated Lead-free Ni–P–PTFE (2.54 μm) Forced convection (vliq = 1.57 m/s) Shearing steam (vsteam = 0.23 m/s)
722 653 584
61.1 39.8 25.9
–0.915 –0.821 –0.771
i 0 1 2 3 4 5 0 1 2 3 4 5 6 0 1 2 3 4
ai
bi −6
5.76 × 10 − 6.60 × 10− 5 4.08 × 10− 4 − 1.10 × 10− 4 1.49 × 10− 4 − 4.45 × 10− 5 − 1.98 × 10− 5 2.58 × 10− 4 –1.27 × 10− 3 2.98 × 10− 3 − 1.64 × 10− 3 3.10 × 10− 3 − 7.41 × 10− 4 2.20 × 10− 4 1.04 × 10− 3 1.05 × 10− 3 − 1.21 × 10− 3 4.21 × 10− 4
3.0065 –3.8753 5.6446 –5.1313 2.5015 –0.5014 4.0360 –10.484 27.079 –46.190 46.581 –25.128 5.5556 2.2024 –1.3611 1.3216 –0.7324 0.1648
366
J.R. Lara, M.T. Holtzapple / Desalination 280 (2011) 363–369 180
Overall Heat Transfer Coefficient (kW/(m2 . °C))
Overall Heat Transfer Coefficient (kW/(m2 . °C))
100
80
60
(a) 40
(b) 20
(c)
160 140 120
Ni-P-PTFE hydrophobic coating thickness (µm)
100 80
0.635
60
12.7
40
1.27
20
127
0
0 0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0
0.4
0.1
Shearing Steam Velocity v (m/s) Fig. 3. Heat transfer coefficient as a function of shearing steam for bare titanium. Forced-convection pool boiling. (a) ΔT = 0.19 °C, P = 722 kPa; (b) ΔT = 0.2 °C, P = 550 kPa; (c) ΔT = 0.38 °C, P = 432 kPa.
Fig. 1 showed heat transfer coefficient U as a function of ΔT for three pressures P. Fig. 4 shows the same data where U is a function of P for a given ΔT. Fig. 4 was created using Eq. 5, U = aP
b
ð5Þ
where condenser pressure (kPa)
P 6
i
a = ∑ ai ðΔT Þ i=0
6
0.2
0.3
0.4
0.5
0.6
i
b = ∑ bi ðΔT Þ : i=0
Table 3 gives values for the empirical parameters. Fig. 4 shows the projected heat transfer coefficient for condensing pressure P = 827 kPa, which the literature indicates is the maximum pressure where dropwise condensation ocurrs [20]. 4. Results for dimpled copper with varying thicknesses of leadcontaining hydrophobic Ni–P–PTFE The thermal performance of lead-containing Ni–P–PTFE hydrophobic coatings with different thicknesses was measured with round-
1 = U
ðkx Þcopper + 2ðkxÞcoating 0
$4
1
1 B2:03 # 10 mC + =@ A kW kW 47 2 ∘ 0:4 ∘ m· C m ·C copper
40
Heat Flux q (kW/m2)
Overall Heat Transfer Coefficient (kW/(m2 . °C))
45
0.2
60 0.4 0.6 0.8 1.0 1.2
20
0.9
dimpled 0.203-mm-thick copper plates. The following thicknesses were tested: 0.635, 1.27, 12.7, and 127 μm. Fig. 5 shows the overall heat transfer coefficients measured at a constant saturated steam pressure (P = 722 kPa) with different temperature differences across the plate. Fig. 6 shows measured heat fluxes. Fig. 7 shows the effect of the Ni–P–PTFE hydrophobic coating thickness on overall heat transfer coefficients. For high-phosphorous lead-containing electroless Ni–P–PTFE hydrophobic coating, the literature [21] reports a thermal conductivity k = 5.44 W/(m·°C). This value can be independently verified by using the heat transfer data reported herein. The lowest measured overall heat transfer coefficient (U = 47 kW/(m 2·°C)) corresponds to 127-μm-thick coating. Assuming the resistance of the boiling liquid and condensing steam are negligible,
∆T (°C)
80
40
0.8
Fig. 5. Overall heat transfer coefficient of 0.203-mm-thick copper plates with leadcontaining Ni–P–PTFE hydrophobic coatings of different thicknesses. Saturated steam at constant pressure P = 722 kPa. Forced convective saturated liquid (vliq = 1.57 m/s). Optimal shearing steam.
$4
2 # 1:27 # 10 m kcoating
!
Ni-P-PTFE hydrophobic coating thickness (µm)
120
100
0.7
Differential Temperature across Plate ∆T (°C)
0.635
35
1.27
30
12.7
25 20 15
127
10 5
0 300
400
500
600
700
800
900
1000
Operating Pressure (kPa)
0 0
0.2
0.4
0.6
0.8
1
1.2
1.4
Differential Temperature across Plate ∆T (°C) Fig. 4. Overall heat transfer coefficient for bare 0.127-mm-thick titanium grade 2 plate. Shearing steam on the condensing surface and forced convective saturated pool boiling (v liq = 1.57 m/s). Smooth curves were determined using Eq. 5. Solid line is interpolation. Dashed line is extrapolation.
Fig. 6. Heat flux across 0.203-mm-thick copper plates with lead-containing Ni–P–PTFE hydrophobic coatings of different thickness. Saturated steam at P = 722 kPa. Forced convection saturated liquid (vliq = 1.57 m/s).
367
J.R. Lara, M.T. Holtzapple / Desalination 280 (2011) 363–369
Condenser Pressure (kPa)
160 60 140 120
Heat Flux q (kW/m2)
Overall Heat Transfer Coefficient (kW/(m2 . °C))
180
100 80 60 40 20
50
827 (Projected)
40
722 646
30
515 20
10
0 0
20
40
60
80
100
120
140
Ni-P-PTFE Hydrophobic Coating Thickness (µm)
0 0
Fig. 7. Overall heat transfer coefficients on 0.203-mm-thick copper substrate with varying thicknesses of lead-containing Ni–P–PTFE hydrophobic coating. Saturated steam at P = 722 kPa with ΔT = 0.20 °C across the plate. Forced convective pool boiling with vliq = 1.57 m/s.
kcoating = 12:2
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
Differential Temperature across Plate ∆T (°C) Fig. 9. Heat flux for 0.203-mm-thick copper plate with round-shaped vertical grooves. Fully coated with 0.635-μm-thick lead-containing Ni–P–PTFE hydrophobic coating. Forced convection in saturated pool boiling. vsteam is the optimal corresponding value (see Fig. 10). Smooth curves were calculated using Eq. 4. Dashed line is a projection to desired operating pressure using Eq. 5.
W : m⋅ ∘ C
This calculated value of kcoating represents a lower limit. With resistance from boiling liquid and condensing steam, the value of kcoating would be even higher. The difference between the experimental and the literature values could result from differences in coating formulation or differences in experimental approach. The literature source does not provide enough information to resolve the discrepancy.
5. Results for vertical-grooved copper with lead-containing Ni–P–PTFE hydrophobic coating Fig. 8 shows the overall heat transfer coefficients for 0.203-mmthick copper with round-shaped vertical grooves coated with 0.635-μm-thick lead-containing Ni–P–PTFE. The temperature differences across the plate were varied for different constant saturated steam pressures. Forced convection is imposed in the saturated liquid side with vliq = 1.57 m/s. Eq. 3 describes each of the curves shown in Fig. 8. Fig. 9 shows heat flux q across the plate for different ΔT and P. Table 3 reports the empirical parameters used in Eq. 3 and 4.
Fig. 10 shows variations of the overall heat transfer coefficient with shearing steam velocity for a given ΔT. Curve (a) shows U is more sensitive to small variations of ΔT at high P. Previously, Fig. 8 showed heat transfer coefficient U as a function of ΔT for a constant P. Fig. 11 shows the same data where U is a function of P for a given ΔT. Fig. 11 was created using Eq. 5. Table 3 shows the empirical parameters. The maximum overall heat transfer coefficient measured was U=191 kW/(m2·°C) (P=722 kPa, T=166 °C, vliq =1.57 m/s, vsteam = 0.16 m/s, and R=1.5 kg shearing steam/kg condensate). Fig. 11 shows the projected heat transfer coefficient for condensing pressure P=827 kPa.
6. Results for vertical-grooved copper with lead-free Ni–P–PTFE hydrophobic coating Lead-free 2.54-μm-thick Ni–P–PTFE hydrophobic coating was applied to 0.203-mm-thick copper plates with round-shaped vertical grooves. Fig. 12 shows the overall heat transfer coefficients corresponding to different temperature differences across the plate at constant saturated steam pressures. Forced convection was imposed on the saturated liquid side with vliq = 1.57 m/s. Fig. 13 shows the heat
250
200
Overall Heat Transfer Coefficient (kW/(m2 . °C))
Overall Heat Transfer Coefficient (kW/(m2 . °C))
250
150
100 Condenser Pressure (kPa)
50 722 646 515
0 0
0.2
0.4
0.6
0.8
1
1.2
1.4
Differential Temperature across Plate ∆T (°C)
200
150
(b)
(a)
100
(c)
50
0 0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Shearing Steam Velocity v (m/s) Fig. 8. Overall heat transfer coefficients for 0.203-mm-thick copper plate with roundshaped vertical grooves. Fully coated with 0.635-μm-thick lead-containing Ni–P–PTFE hydrophobic coating. Forced-convection saturated pool boiling (vliq = 1.57 m/s). Optimal shearing steam on the condensing surface (see Fig. 10). Condensing steam at different pressures. Smooth curves are given by Eq. 3.
Fig. 10. Heat transfer coefficient as a function of shearing steam velocity vsteam (a) P = 722 kPa, (b) P = 646 kPa, (c) P = 515 kPa. 0.203-mm-thick copper with vertical grooves and 0.635-μm-thick hydrophobic lead-containing Ni–P–PTFE coating. Forced convection in saturated liquid side vliq = 1.57 m/s.
368
J.R. Lara, M.T. Holtzapple / Desalination 280 (2011) 363–369 Condenser Pressure (kPa)
∆T (°C) 300
80 70
250
827 (Projected) 722
Heat Flux q (kW/m2)
Overall Heat Transfer Coefficient (kW/(m2 . °C))
0.20
200
150
0.40 100
0.60 0.80
50
1.00
0 400
500
600
700
800
60 50 653
40 30
584
20 10
900
0
1000
0
0.2
Operating Pressure (kPa)
0.4
0.6
0.8
1
1.2
1.4
1.6
Differential Temperature across Plate ∆T (°C)
Fig. 11. Overall heat transfer coefficient for 0.203-mm-thick copper plate with roundshape vertical grooves coated with 0.635-μm-thick lead-containing Ni–P–PTFE hydrophobic coating. Force-convection shearing steam on the condensing surface and forced convective saturated pool boiling (vliq = 1.57 m/s). Smooth curves were determined using Eq. 5. Solid line is interpolation. Dashed line is extrapolation.
Fig. 13. Heat flux for copper coated with 2.54-μm-thick lead-free Ni–P–PTFE hydrophobic coating. Forced-convection pool boiling. Optimal shearing steam from Fig. 14. Curves were created using Eqs. 4 and 5. Dashed line is a projection to desired operating pressure.
flux. Fig. 14 shows variations of the overall heat transfer coefficient with shearing steam for given ΔT. For this case, the maximum overall heat transfer coefficient measured was U = 240 kW/(m 2 ·°C) (P = 722 kPa, T = 166 °C, ΔT = 0.20 °C, v liq = 1.57 m/s, v steam = 0.23 m/s and R = 0.6 kg shearing steam/kg condensate). Fig. 15 shows U as a function of P for a given ΔT, which was calculated using Eq. 5 with parameters from Table 3. At P = 827 kPa and ΔT = 0.2 °C, the projected overall heat transfer coeffcient is U = 270 kW/(m 2·°C).
Overall Heat Transfer Coefficient (kW/(m2 . °C))
300
7. Conclusions
250
200
(a) 150
100
(b)
50
(c) 0 0
In Part II, for 0.127-mm-thick bare titanium, the best overall heat transfer coefficient was U = 78 kW/(m 2·°C) (P = 722 kPa, T = 166 °C, ΔT = 0.20 °C, vliq = 1.57 m/s, and vsteam = 0.12 m/s). For copper with vertical round-shaped grooves coated with 0.635-μm-thick leadcontaining Ni–P–PTFE hydrophobic coating, the best overall heat transfer coefficient was U = 191 kW/(m 2·°C) (P = 722 kPa, T = 166 °C, ΔT = 0.20 °C, vliq = 1.57 m/s, and vsteam = 0.16 m/s). For copper with vertical round-shaped grooves coated with 2.54-μm-thick leadfree Ni–P–PTFE hydrophobic coating, the best overall heat transfer
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Shearing Steam Velocity v (m/s) Fig. 14. Heat transfer coefficient for copper coated with 2.54-μm-thick lead-free Ni–P– PTFE hydrophobic coating. Forced-convection pool boiling. (a) ΔT = 0.20 °C, P = 722 kPa; (b) ΔT = 0.2 °C, P = 653 kPa; (c) ΔT = 0.23 °C, P = 584 kPa.
coefficient was U = 240 kW/(m 2·°C) (P = 722 kPa, T = 166 °C, ΔT = 0.20 °C, vliq = 1.57 m/s, vsteam = 0.23 m/s). Including both Parts I [1] and II, Table 4 summarizes the results of the experimental investigation of hydrophobic heat exchangers. The
∆T (°C)
300
0.20
250
Overall Heat Transfer Coefficient (kW/(m2 . °C))
Overall Heat Transfer Coefficient (kW/(m2 . °C))
300
200
150
Condenser Pressure (kPa)
100
50
722 653 584
250
200 0.40
150
100
0.70 0.85
50
1.40
0 0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
Differential Temperature across Plate ∆T (°C)
0 300
400
500
600
700
800
900
1000
Operating Pressure (kPa) Fig. 12. Overall heat transfer coefficients for copper coated with 2.54-μm-thick leadfree Ni–P–PTFE hydrophobic coating. Forced-convection pool boiling. Optimal shearing steam on the condensing surface (see Fig. 14). Condensing steam at different pressures. Smooth curves are given by Eq. 3.
Fig. 15. Overall heat transfer coefficient for copper coated with 2.54-μm-thick lead-free Ni–P–PTFE hydrophobic coating. Forced-convection pool boiling. Curves were created using Eq. 5. Dashed line is a projection to desired operating pressure.
369
J.R. Lara, M.T. Holtzapple / Desalination 280 (2011) 363–369 Table 4 Summary of results of the experimental investigation on hydrophobic heat exchangers. Part
I
Figure
2 4
II
6 9 10 13 14 17 1 5
8 11 12 15 a
Overall heat transfer coefficient (kW/(m2 °C))
Plate substrate
Coating thickness (μm)
Coating chemistrya
Surface geometry
Shearing steam (m/s)
Liquid convection (m/s)
Liquid-side nucleation
Boiling stones
Min. ΔT across plates (°C)
Max. steam pressure (kPa)
Naval brass 464 0.762-mm-thick Naval brass 464 0.762-mm-thick Naval brass 464 0.762-mm-thick Copper 0.203-mm-thick Copper 0.203-mm-thick Titanium 0.127-mm-thick Copper 0.203-mm-thick
No
Ni–P–PTFE Leaded Ni–P–PTFE Leaded Ni–P–PTFE Leaded Ni-P-PTFE Leaded Ni–P–PTFE Leaded Ni–P–PTFE Leaded Ni–P–PTFE Leaded
Round dimples
0.08
No
Sand blasted
No
0.4
722
17
Round dimples
0.08
No
Sand blasted
No
0.20
722
70
Round dimples
0.08
1.57
Sand blasted
No
0.20
Round dimples
0.4
1.57
No
No
0.20
Round dimples
0.5
1.57
No
Yes
0.20
Round dimples
0.15
1.57
No
No
0.20
Round dimples
1.57
No
No
0.20
99 119 159 202 184 248 78 94.5 141 131
Ni–P–PTFE Leaded Ni–P–PTFE Lead-free
Vertical grooves
NR NR NR 0.16
722 827 722 827 722 827 722 827 722
1.57
No
No
0.20
Vertical grooves
0.23
1.57
No
No
0.20
722 827 722 827
191 272 240 281
Copper 0.203-mm-thick Copper 0.203-mm-thick
0.635 0.635 0.635 0.635 No 1.27 12.7 127 0.635 2.54
Microplating, Inc.; extrapolated values are shown in bold italics; NR = not reported.
maximum overall heat transfer coefficient found in the literature from previous works is U = 17 kW/(m2·°C), which is 14.2 times lower than the best overall heat transfer coefficient U = 240 kW/(m 2·°C) measured in this study for thin copper sheets with round-shaped vertical grooves and lead-free Ni–P–PTFE hydrophobic coating. For all cases, the hydrophobic surface exposed to boiling water was in good condition after the experiments were concluded. 8. Legal notice This desalination technology has been licensed to Terrabon, Inc. The information, estimates, projections, calculations, and assertions expressed in this paper have not been endorsed, approved, or reviewed by any unaffiliated third party, including Terrabon, Inc., and are based on the authors' own independent research, evaluation, and analysis. The views and opinions of the authors expressed herein do not state or reflect those of such third parties, and shall not be construed as the views and opinions of such third parties. References [1] J.R. Lara, M.T. Holtzapple, Experimental Investigation of Dropwise Condensation on Hydrophobic Heat Exchangers Part I: Dimpled Sheets, Desalination 278 (2011) 165–172. [2] Titanium Metals Corporation, Titanium, Products and Applications, http://www. timet.com/fab-p09.htm November 2004. [3] M. Izumi, S. Kumagai, R. Shimada, N. Yamakawa, Heat transfer enhancement of dropwise condensation on a vertical surface with round shaped grooves, Experimental Thermal and Fluid Science 28 (2004) 243–248. [4] A. Leipertz, A.P. Froba, Improvement of condensation heat transfer by surface modifications, Heat Transfer Engineering 29 (2008) 343–356. [5] Y.-X. Wang, J.L. Plawsky, P.C. Wayner Jr., Optical measurement of micro scale transport processes in dropwise condensation, Microscale Thermo physical Engineering 5 (2001) 55–69.
[6] C. Yamali, H. Merte Jr., Influence of sweeping on dropwise condensation with varying body force and surface sub cooling, Int. J. Heat Mass Transfer 42 (1999) 2943–2953. [7] A.S. Lexmond, C.W.M. van der Geld, The effect of fluid flow on the detachment of drops in the wake of a flat plate, Experimental Thermal and Fluid Science 29 (2005) 363–370. [8] A.S. Lexmond, C.W.M. van der Geld, The effect of plate thickness, surface tension and fluid flow on detachment of drops from plate, Experimental Thermal and Fluid Science 29 (2005) 813–819. [9] Q. Zhao, Y. Liu, Investigation of graded Ni–Cu–P–PTFE composite coatings with antiscaling properties, Applied Surface Science 229 (2004) 56–62. [10] Q. Zhao, Y. Liu, S. Wang, Surface modification of water treatment equipment for reducing CaSO4 scale formation, Desalination 180 (2005) 133–138. [11] Q. Zhao, H. Muller, Graded Ni–P–PTFE coatings and their potential applications, Surface and Coatings Technology 155 (2002) 279–284. [12] V. Berndt, S. Zunft, Steinhagen Muller-, Theoretical and experimental study on dropwise condensation in plate heat exchangers, 5th European Thermal-Sciences Conference, The Netherlands, 2008. [13] Q. Zhao, Y. Liu, Development and evaluation of Ni–Cu–P–PTFE composite coatings to minimize microbial adhesion, ECI Symposium Series, Volume RP2, Proceedings of 6th International Conference on Heat Exchanger Fouling and Cleaning, Germany, June, 2005, pp. 5–10. [14] H. Muller, Smart Surfaces for Improved Heat Exchangers, https://www.htri.net/ Public/prodsvcs/HMS_Victoria2.pdf. [15] A. Ramalhoa, J.C. Miranda, Friction and wear of electroless Ni–P– PTFE coatings, Wear 259 (2005) 828–834. [16] S. Balasubramanian, V.M. Puri, Thermal energy savings in pilot-scale plate heat exchanger system during product processing using modified surfaces, Journal of Food Engineering 91 (2009) 608–611. [17] Q. Zhao, Y. Liu, E.W. Abel, Effect of Cu content in electroless Ni–Cu–P–PTFE composite coatings on their anti-corrosion properties, Materials Chemistry and Physics 87 (2004) 332–335. [18] Y.Z. Zhang, Y.Y. Wu, K.N. Sun, M. Yao, Characterization of electroless Ni–P–PTFE composite deposits, Journal of Materials Science Letters 17 (1998) 119–122. [19] http://mdmetric.com/tech/tiG2cp.htm, Maryland Metrics, March 2011. [20] J.B. Dolloff, N.H. Metzger, L.H.S. Roblee Jr., Dropwise condensation of steam at elevated pressures, Chem. Eng. Sci. 24 (1969) 571–583. [21] http://www.techmetals.com/EngineeredFinishes/TM117p.pdf. TechMetals Inc, March 2011.