A study using an experimental design of dispersion coating with PTFE ambient cured process is presented. The effects of five major process parameters of the coating, i.e. polymer concentration, dip coating time, presence of immersion rinsing, spray rinsing, and air blow drying on surface residue defect, coating thickness, and wear resistance were investigated. The study was conducted starting with two level factorial experiments to identify the most significant parameters in the coating process. and concluding with response surface methodologies to establish an optimum operating condition for the coating in anodizing aluminum process. It was observed that dipping time and water immersion rinsing time were identified to be the key parameters in the PTFE air-dry coating process. With the key parameters, an operating process condition for the dispersion coating with PTFE that provided elimination of residue defects and satisfactory wear resistence was determined to be a PTFE dipping time of 2 minutes with water rinsing time of 1 minute.
(Keywords: PTFE, dispersion coating, response surface methodology)
Sanford Hardlube™ is a proprietary anodizing process developed by Sanford Process. Inc. It is an electrochemical process of a combination of “Teflon” and Sanford Hardcoat™ for building a lubricious aluminum oxide coating on aluminum. The process converts the aluminum surface to aluminum oxide (Al2O3H2O) and replaces the H2O with Sanford Hardlube™. In this process, PTFE is suspended in water with surfactant, applied by dipping the unsealed oxide surface and cured at ambient for 2 hours. Sanford Hardlube® has superior features  such as:
- Low Coefficient of Friction: A continuous film which penetrates the surface of the hard anodized aluminum and has lower friction by as much as fivefold.
- Non-Stick: Low wettability prevents most solid substances from permanently adhering to the coating surface. Almost all substances release easily (see Fig.1).
- High Abrasion Resistance: Hardcoat is aluminum oxide. It is “file hard”, has been measured as high as Rockwell 70C, and has wearing qualities superior to case hardened steel.
- High Corrosion Resistance: Hardcoat considerably increases corrosion resistance of aluminum while the fluorocarbon resin is normally unaffected by chemical environments.
PTFE and/or Teflon can be coated by spraying or plated on by impregnating it into other plating materials such as nickel. Unlike common techniques of drying and baking the substrate to remove volatile components at high temperature, ambient temperature curing coating process has revealed a significant issue of white stains on the surface. This is mostly due to an uneven distribution of PTFE particles. The white stains indicate that areas where less PTFE is incorporated do not have the same physical and mechanical properties as in areas where the PTFE content is uniform. 
Figure 1. Water drops illustrate low wettability of Sanford Hardlube
In this paper the PTFE dispersion coating process was investigated using parametric design techniques to rectify white stain defect on the surface during drying step. Of the most common statistical design methods, the Taguchi-type factorial design as well as the central composite design (CCD) was used in the present work to analyze the performance of PTFE coating process cured at ambient temperature. The methodology provides several important operating guidelines to prevent forming white stains from the surface.
2 EXPERIMENTAL SETUP
PTFE aqueous dispersions used in our studies were representative samples of Sanford Hardlube™. The samples were characterized for percent solids, particle size, pH, and surfactant content. Table 1 summarized the characteristics of the dispersion.
TABLE 1. Characteristics of Sanford Hardlube®
2.2 Coating Procedure
6061 Aluminum coupons of 4” by 4” in size were coated with the Sanford Quantum™ hard anodize process. Sanford Quantum™, which is a low voltage (average 12 VDC) hardcoating process in room temperature (70ºF) provided a coating thickness of 1 mil (25 μm) with little darkening of the oxide. To make a clear contrast with white PTFE particles the hardcoated coupons were dyed black. Before and after test coupons processed by Sanford Quantum™ with black dying shown in Fig.1.
Figure 2. Pictures of coupon before and after processing Sanford Quantum™
Preweighed coupons having a 1.0 mil of hardcoat processed by Sanford Quantum™ were immersed into Sanford Hardlube® dispersion solution with a gentle air agitation for a given period. Immersion rinsing was performed using stir-bar agitated bath and followed by either water spray or air blow off with dry compressed air at room temperature. The PTFE coated coupons were allowed to dry for at least 12 hours. Once dry, the coupons were weighed to calculate average coating thickness. Efficiencies were calculated for each coupon to determine the amount of PTFE applied. Visual inspection to identify surface residue defect was carried out by applying a residue index from 0 (no defect) and 1 (worse defect) to each test coupon in the design matrix. Taber abrasion test per MIL-A-8625F section 4.5.5 was also conducted to investigate PTFE coating wear resistance.
2.3 Experimental Plan
Taguchi-type fractional factorial experiment was conducted to screen a large number of variables and to identify the most significant parameters with a very small number of experiments. The factorial design was followed by a central composite design (CCD) to determine the approximate parameter space for the optimization of the coating process. The CCD is the most popular response surface methodology (RSM) design . It contains an imbedded fractional factorial design with center points that is augmented with a group of `star points’ that allow estimation of curvature, i.e., the data to be fitted to a second-order Taguchi polynomial (see Fig. 3).
The statistical analysis was accomplished with Design-Ease and Design-Expert software . The design evaluated the effects of the coating processing variable on the visually measured response, i.e., surface residue defect. The net effect of each variable is calculated by determining the difference between the positive and negative levels divided by numer of positive level tests as shown in the formula 
Where E is the effect of variable, R is the response and the number of + levels equals the number of – levels. Design is summarized in Table 2.
Figure 3: The example points of a Central Composite design with three input parameters (CCD factor = b/a).
TABLE 2. Design Summary for Central Composite Design (CCD)
3 RESULTS AND DISCUSSION
The Design-Expert program was used to analyze the response of surface residue defect using the following methodology: (a) ANOVA analysis was conducted to determine the adequacy of linear, quadratic, and cubic models; (b) one model was then chosen for in-depth regression analysis; (c) diagnostic evaluation of the robustness of the model was determined; and (d) response surface analysis was conducted to optimize the coating process parameters.
TABLE 3. Model Fit Summary
Table 3 showed that F value of 79.3 was obtained for linear model, which includes only main effect (A, B and so on), with corresponding probability value of 0.01%. A large F value and a small probability value indicate the model validity. Therefore, the linear model was chosen for further model verification study.
Table 4 showed analysis of variance for the selected linear model with 2 factor interaction. Again, the model F-value of 60.5 implies the model is significant. There is only 0.07% chance could be occurred due to noise. Values of “Prob > F” less than 0.0500 indicate model terms are significant. In this case C (Air Blow), E (Water Spray) are significant model terms. The “Lack of Fit F-value” of 0.62 implies the Lack of Fit is not significant relative to the pure error. There is a 70.51% chance that a “Lack of Fit F-value” could be occurred due to noise.
Table 4. Analysis of Variance (ANOVA)
Diagnostic evaluation of the robustness of the model was determined using the normal probability plot for the studentized residuals. The results are given in Fig. 4. Since there are no departures from a straight line, the diagnosis of residuals does not reveal any statistical problems in the regression analysis. Analysis was then conducted to determine the optimum parameters for CCD. Figures 5 to 7 show contours and response surface for the residue index.
Figure 4. Normal probability plot of residues
Figures 5 and 6 plotted the two parameters affecting residue index as a function of air blow, water rinse, and water spray modes. For all rinsing modes off, lower polymer concentration and higher immersion provided lower residue index. Figure 7 showed three dimensional surfacel plot of defect index as function of rinsing time and PFTE concentration. It showed 60 seconds of water rinsing time provided residue defect free surface without respect to ploymer concentration.
Table 5 showed results on Taber abrasion testing with and without PTFE coating. The results show that the consistency number of weight loss indicates that PTFE particles are effectively coated on the surface regardless of water rinsing time and coating concentration. Compared with a control value, i.e., without PTFE coating, the wear resistance was increased by two fold.
Figure 5. Contour plot with air blow, water rinse, and water spray all off mode.
Figure 6. 3 Dimensional surfacel plot with air blow, water rinse, and water spray all OFF mode.
TABLE 5. Taber Abrasion Testing w/wt PTFE Coating
Figure 7. Three dimensional surfacel plot of defect index as function of rinsing time and PFTE concentration
From the results presented above, it can be concluded that:
1. Of the five control variables studied, coating dipping time, water spraying rinse, and air blowing drying are the most significant.
2. Central composite design was used to establish an optimum operating condition with the key parameters to eliminate residue defects and satisfactory wear resistance. It was determined to be a coating dipping time of 2 minutes with flowing water rinsing time of 1 minute.
3. Taber abrasion test was conducted to compare with test coupons processed with and without PTFE coating. The wear resistance was increased by two fold with the coupon coated with PTFE.
 Sanford Hardlube product description, www.sanfordprocess.com, Sanford Process, Inc., Natick, MA.
 Pietsch, K.H.: Dispersion Coatings with PTFE, PF Online, Electroless Nickel ’99 Conference, Indianapolis, Indiana, November (1999).
 Lorenzen, T.J. and Anderson, V.L. (1993), Chapter 2 of Design of Experiments: A No-Name Approach, Marcel Dekker, New York.
 Design-Expert Ver 6., State-Ease, Inc. Minneapolis, MN.
 Box, G.E.P., Hunter, W.G., and Hunter. J.S. (1978) Chapter 6 of Statistics for Experimenters, John
Wiley & Sons, New York.