Mechanical and erosion properties of CaC[O.sub.3]-EMAA thermal sprayed coatings.
Polymers are relatively versatile materials in terms of their processing and production. One processing technique that is increasingly being used is that of thermal spray. While this has been performed for some time with materials such as epoxy resins, the ability to thermally spray thermoplastics has become more popular, and opened up the range of potential materials that can be used and has been the subject of a recent extensive review (1).
Thermal spray has many advantages as a coating technique in that it can be used to coat complex substrate shapes without the need for environmentally unfriendly solvents (2-4). The basic process of thermal spraying is straightforward--fine particles of polymer are propelled, using a flow of carrier gas, through a flame where they are heated so they soften, towards an (often heated) substrate where they collide, flatten and assemble to form a coating from well-adhered flattened particles (5). There are a number of spray techniques used, the dominant difference being the heating source--varying from high velocity oxy-fuel to air plasma, to vacuum plasma, to a conventional flame spray where a mixture of air and (for example) propane forms the combustible mixture. In the present study, a flame spray method was used and air was used for the fluidization of particles in the powder feeder and to transport powder (or "feedstock" (6) as it is commonly termed in this area of technology) to the torch, where it participates in producing a combustion flame. This latter thermal spraying technique is the lowest cost and most physically portable of the thermoplastic thermal spray techniques--requiring polymer particle sizes of a few hundred microns. The flame itself has temperatures up to 3000[degrees]C, and thus, it is necessary for the particles to traverse the flame and become heated without sustaining catastrophic molecular damage.
Indeed, it has been shown that only a slight molecular weight change occurs during the process for the commonly used ethylene-methacrylic copolymer (or associated ionomer) (7). The main change is related to an increased polarity due to oxidation of the polymer, which is advantageous because it encourages adhesion to the polar metal substrate (8, 9). The use of copolymers such as ethylene-methacrylic acid is strategic in that the methacrylic acid/ionomer component adheres well to the metal substrate--whereas the ethylene units provide moisture and other environmental resistance to the coating--making it excellent for coating pipes, bridges and the like (10-13).
While the process is simple, there are many variables that influence the coating integrity--both in an adhesive and cohesive sense (14-16). These include flame mix (temperature), gas carrier speed, substrate temperature, torch traverse speed (as the coating is being applied), and so on (17-19). Recently we showed that such a system could also be used in a repair mode (20). If some part of the coating should be abraded, fractured, or dislodged, it was determined that there was still good adhesion between freshly sprayed molten polymer powder and the underlying old coat; indeed, it was some two times stronger than the original polymer-steel bond. Interestingly, it was found that the bond was strongest between two related but dissimilar polymers (copolymer and ionomer, both of the ethylene-methacrylic acid family), compared to bonding of the copolymer or ionomer materials to themselves.
As in many other areas of polymer science, the next-generation material usually includes the combination of polymer with other materials, be they another polymer or some ceramic or metal phase to form a composite. Composites have the advantage of maintaining the toughness and coating-forming properties of the plastic, while allowing properties such as modification of the modulus and abrasion resistance to be introduced. The added issue with the formation of thermal spray coatings is that the feedstock must be of an appropriate particle size for the thermal spray process. The ability to influence the properties mentioned above and others such as surface polarity has led to the present study of thermal spraying of an ethylene-methacrylic copolymer with calcium carbonate, a common inorganic filler, onto a steel substrate. Particle size is an important property in terms of spraying ability and final properties; therefore, CaC[O.sub.3] of three sizes is used in this work. An important area of investigation is the nature of adhesion of the composite coating to the substrate since not all of the inorganic particles pass through the flame and therefore do not melt but become physically incorporated into the coating. Likewise, the adhesion between filler particles and matrix material is a key to its integrity. Thus, a range of processing issues are investigated, as are the final properties of the coating, to ascertain the advantages and limitations of spraying such mixtures of quite different material types.
Materials and Preparation of Coatings
The polymer used in this study was an ethylene methacrylic acid (EMAA) copolymer (PF111U, 70 mesh; average diameter of 140 [micro]m) supplied by Plastic Flame Coat Systems (PFS, Texas). Three CaC[O.sub.3] powders of different particle size were used as inorganic additives. The average particle sizes were 2.8 [micro]m, 9 [micro]m, and 36 [micro]m as determined using a Micromeritics Saturn Digisizer 5200 (Micromeritics. Norcross, Georgia). The density of EMAA and CaC[O.sub.3] were determined by gas pycnometry to be 0.9392 and 2.697 g/[cm.sup.3], respectively, in agreement with established values for these materials.
A blend of CaC[O.sub.3] and EMAA was prepared by a mechanical mixing process in which both powders were combined and tumbled continuously in a cylindrical plastic vessel to ensure a homogeneous mixture. A Powder Pistol 124 from PFS was used both for preheating the substrate and spraying the coating. Propane (at 83 kPa) and compressed air (at 207 kPa) were used to produce the combustion flame. Compressed air was used to fluidize the powder mixture, and transport the powder blend from the fluidized bed to the torch, where it was added to the combustion gas for generation of the flame. The principle of the plastic flame spray system and process parameters has previously been described to deposit EMAA coatings (21, 22). In this work, the final coating temperature was controlled by varying the number of traverses of the spray torch during processing. The specimen temperature was measured with a hand-held infrared pyrometer for several seconds after spraying.
Substrates for peel tests were normal carbon steel bars (length 190 mm, width 25 mm, and height 10 mm). The surface of the steel bar was cleaned with acetone, but not grit blasted before spraying.
Determination of Composition of Mixture and Composite Coatings
It was observed that during spraying of the mechanical blend, the denser inorganic constituent was not entirely transferred to the coating. To determine the loss of calcium carbonate filler, the mixture was transferred into a vessel through the flame spray torch in the absence of a flame after an interval of 2 minutes. The density of the collected powder was determined by a gas pycnometer, described below. The density of sprayed material was used to calculate the mixture composition after spraying for different times. The initial composition of three CaC[O.sub.3]-EMAA mixtures with different filler sizes was 30 vol%. The filler content transferred to the final sprayed coating was also determined by gas pycnometry.
Deposition rate was determined as the volume of thermal spray coating deposited over a period of 10 minutes. The deposition rate was then expressed as a volume per minute. Deposition efficiency was determined as the ratio of the powder blend transported through the torch to the amount of material deposited in the coating. The powder transported through the torch was obtained by collecting the powder in a plastic bag in the absence of a flame. The coating was obtained by passing the powder blend through a flame and depositing onto a polytetrafluoroethylene (PTFE) pan. The coating was easy to remove after the spraying was complete.
Tensile and Peel Testing
Tensile test samples were produced by spraying a coating onto a flat PTFE-coated substrate from which it could be readily removed. A stamping tool was then used to form dogbone-like samples 25 mm in length, 3.5 mm in width, and about 1 mm in thickness. Tensile testing was conducted on an Instron 4505 (Mechanical Testing System) at a crosshead speed of 50 mm/min and room temperature, about 20[degrees]C. A minimum of five samples were tested at each condition to assess variability in coating properties. The modulus was calculated automatically using a secant modulus at 1% strain. Yield stress was determined from the stress-strain curve. Tensile strength was defined as the maximum stress obtained from the curve.
Adhesion of a single coating to mild steel (monolayer configuration) or a coating deposited on a previously sprayed layer (bilayer configuration) was determined using a 90[degrees] peel test rig (ASTM 3167-97) (23). Final coating thicknesses were about 1 mm for the monolayer samples, and about 2 mm for the bilayer samples, owing to the base and top layer as previously adopted for testing bilayered coatings (20). The peel test was conducted on the Instron 4505 Mechanical Testing System at a peel speed (crosshead speed) of 50 mm/min. A minimum of five samples were tested for each condition.
Density and X-ray Diffraction Analysis
The densities of powders and coatings were determined in an AccuPyc 1300 gas pycnometer (Micromeritics) at test conditions of five purges, 15 runs, all at room temperature of 20[degrees]C. Powder for analysis was simply poured into the sample cup but coatings sprayed onto a PTFE pan were removed and sectioned for placement into the pycnometer.
Phase composition and crystallinity of powders and coatings were determined with a Rigaku XRD Vertical Goniometer (Tokyo, Japan) set at 40 kV and 22.5 mA. The detector was set to pass through a 20 range of 1.5[degrees] to 35[degrees], at a scanning rate of 1[degrees]/min, and a step size of 0.05[degrees].
Erosion Tests and Friction Tests
Erosion testing (ASTM G76-95) (24) involved transporting erodent particles with compressed air through a metal pipe with an internal diameter of 6 mm onto a coating surface. The angle of the erodent particle jet was set by the metal pipe to either 0[degrees] (horizontal) or 90[degrees] (vertical). The erodent used in this study was garnet with a 200 [micro]m to 600 [micro]m (30~60 mesh) particle size. All erosion tests were carried out at a particle flux of 180 g/minute, particle velocity of 20 m/s, and at room temperature of about 20[degrees]C. The cumulative erosion loss was measured by weighing the sample to an accuracy of 0.0001 g at intervals of 5 minutes.
The friction and wear tests (ASTMG 133-95) (25) were conducted on a DF-PM (Japan) reciprocating sliding friction tester at load friction conditions of 1 N or 3 N, velocity of 90 mm/minute, stroke of 5 mm, room temperature of 18[degrees]C, relative humidity of 37% and 200 total cycles. The counterpart ball was a GCr15 bearing steel (SAE 52100) of 3-mm diameter. In this work, the width of wear trace measured by a digitally recording microscope was used to determine the wear behavior of composite coatings.
Surface images and fracture cross sections of polymer coatings were observed on a JEOL JSM-840A scanning microscope (Tokyo, Japan). To prevent plastic deformation of the polymer coatings at normal temperature. SEM samples were fractured at liquid nitrogen temperature, and gold sputter-coated with 12 short cycles, each cycle consisting of a 10-second coating time and a 10-second rest time. Accelerating voltages of 15 kV were used in the SEM during observation.
RESULTS AND DISCUSSION
Composition of CaC[O.sub.3]-EMAA Coatings
It has been previously established that the composition of thermal spray composite coatings differs from the powder mixture used for spraying in those circumstances where a mechanical powder blend is employed (26). Usually, the amount of the heavier filler is less in the final coating than in the powder mixture. The source of the filler loss has been considered to be mainly attributed to the morphology, density, and size difference of the filler and polymer powder, with no clear knowledge on the dominant factor influencing filler loss. To better understand the filler loss, it is necessary to follow the change in mixture composition during the spray process itself. Both the amount and kinetics of filler loss were investigated in this work for three different sizes of filler particles. These were transported into a collection vessel (plastic bag) without a flame to allow the change in composition over time to be monitored.
Each powder blend exhibited a differing loss of CaC[O.sub.3] powder, the degree of loss dependent on the filler size as shown in Fig. 1. The initial concentration was 30 vol% in all cases. The powder blend with the largest inorganic filler exhibited the least loss and produced a relatively constant 25 vol% of inorganic filler during powder transport for a period of 40 minutes. Samples containing the smaller filler sizes led to a lower filler content after 2 minutes, and a continued greater loss with time. The smallest filler size blended with the polymer powder produced nearly a two-fold loss in the filler, amounting to 18 vol% initially, decreasing to 10 vol% after 40 minutes. Some error in the density measurements may have arisen from the attraction of CaC[O.sub.3] filler to the plastic bag. However, the results reflect a trend with respect to mixture composition. Delivery of the blended polymer and inorganic powder showed that some fraction of the fine inorganic particles are lost, the lighter particles to a greater degree, while the larger, heavier, 36-[micro]m particles are more effectively transferred with the polymer powder.
[FIGURE 1 OMITTED]
The same trend in the lower filler transfer was found by examining the fluidized bed powder feeder. It occasionally appeared that some filler powder adhered to the internal wall of the fluidized bed powder feeder bucket wall. The density of the accumulated powder on the inside wall of the fluidized bed above the top powder level was higher than the initial powder blend placed into the powder feeder. This suggests a high filler content on the inside wall of the feeder vessel. Smaller inorganic filler particles with their lower weight appeared to relocate preferably onto the bucket walls. For example, the initial composition of the CaC[O.sub.3]-EMAA powder blend with a 9-[micro]m size powder was 30 vol%, but after powder transport for 30 minutes, the filler content of the transported powder was 16 vol%, while the mixture remaining in the fluidized bed was 20 vol%. However, the powder collected on the bucket wall reached a filler content as high as 46 vol%. The individual filler particles are, thus, too light to be fully co-fluidized with plastic particles. Thus, when fillers are blended with plastic particles, the closer the filler particle weight to the plastic particle, the better fluidization of both powder types in the mixture. For the EMAA powder used in this study, with an average particle size of 140 [micro]m, the average weight of the mean-sized particle is 1.35 X [10.sup.-6] g. The average weight of individual CaC[O.sub.3] filler particle was 3.10 X [10.sup.-11] g, 1.03 X [10.sup.-9] g, and 0.66 X [10.sup.-7] g for the fillers with a mean particle size of 3, 9, and 36 [micro]m, respectively, all of which are significantly lighter than the EMAA particles. To match this mass, the average inorganic filler size would have to be much larger, ~99 [micro]m. Note that the more rounded shape of the filler would also provide less buoyancy compared to the polymer particles and reduce slightly the optimal size accordingly.
Additional filler is lost upon passing the composite feedstock through the flame during the deposition process. Fig. 2. An increase of filler in the powder blend, within the fluidized bed powder feeder, leads to a linear increase in filler incorporated into the coating. A larger filler size produces a steeper gradient in the line, indicating a lower filler loss during the spray process. The higher filler loss of small particles during deposition is attributed to inflight filler loss and rebound from the substrate. The polymer has a higher momentum and deposits effectively because of the high particle softness imparted from the heating process. Incorporation of hard filler particles depends mainly on the softness of the underlying surface. The small sizes are more sensitive to gas flow, approaching the substrate at a right angle, but undergo a change at the substrate to follow the substrate surface. It is expected that very small particles will be entrained in the gas flow path and may not reach the soft polymer surface, and get carried away as part of the effluent.
The addition of filler was found to influence the coating deposition rate. To maintain the filler content at about 6 vol% within the coating for all composite coatings, the powder blends were prepared using the data of Fig. 2. To achieve this, the powder blend was prepared with 20, 25, and 30 vol% filler, with filler particle sizes of 36, 9, and 3 [micro]m, respectively. The deposition rate was measured for the various powder mixtures, as the amount of coating deposited per unit time, Table 1. The deposition rate decreased with the addition of inorganic filler (not surprising, because of inorganic particle loss). It appears that the addition of 6 vol% filler into the coating does not follow a rule of mixtures, since a 94 vol% deposition rate of the polymer would lead to decrease in polymer deposition of 26% to 24.4%, and the actual decrease in total solid content is to 21% for the 36-[micro]m filler and 10% for the 2.8-[micro]m filler. The deposition rate decreases with a smaller filler particle size, since the smaller particles are more susceptible to being lost away from the substrate--as explained above. The deposition efficiency (Table 2) is a combination of rate of deposition and rate that material leaves the fluidized bed. Deposition efficiency exhibited the same trend as the deposition rate when the filler content within the coating was maintained at about 6 vol%. Table 2. Pure PF111 exhibits a high deposition efficiency of 90%, while blended powders show a lower deposition linked to both the filler particle size and content, as discussed above.
[FIGURE 2 OMITTED]
X-ray diffraction was used to monitor the influence on matrix crystallinity of filler addition. Blends of filler and powder showed diffraction peaks for both constituents, but a new peak arose, most likely related to the polymer phase. The diffraction patterns of the filler, polymer powder, and coating are shown in Fig. 3a. In the composite coating, the new peaks at 23.0[degrees] and 29.4[degrees] are due to the filler. Fig. 3b. The small peak at 23.0[degrees] becomes more defined with large filler particle size owing to a larger crystallite size. The ratio of the crystalline peak at 21.2[degrees] from the polymer to the amorphous peak height remains the same, despite the presence of the filler. The new peak occurs at 18.6[degrees], and is likely due to the effect of the calcium carbonate particles' encouraging nucleation of a different polymer crystalline structure.
The surface of the CaC[O.sub.3]-EMAA composite coating, as observed in the scanning electron micrograph, remains quite smooth, typical of thermally sprayed polymer coatings, despite the inclusion of filler particles that are loosely adhered to the surface, Fig. 4a. Fillers are more clearly observed in the fractured cross section as rougher areas within the coating, Fig. 4b. Failure occurs in close proximity to the filler particles. It is not clear if the polymer between separate filler particles has provided the path of least resistance through the material. The rough appearance of the failure surface suggests that failure can lead to detachment of the polymer from the filler particles.
[FIGURE 3 OMITTED]
Tensile and Peel Strength of Composite Coatings
The change in mechanical behavior with filler added to the neat PF111 polymer clearly leads to a large decrease in ductility and tensile strength of the coating. Table 2. The 1% secant modulus of composites increased marginally with a higher filler content. It is well known that the addition of inorganic filler into a polymer matrix will increase its modulus and hardness and decrease the ductility of polymer. The strain at fracture and tensile strength of pure PF111 are 512% and 12.6 MPa, while for the composite with a filler content of 7.25%, these properties are only 48% and 10.4 MPa. The filler-enriched boundary areas of the deposited polymer particles in the thermal spray coating result in a more significant decrease in mechanical properties than for composites made by compression molding. Filler particles in the coating are not as evenly distributed as in compression-molded samples, but are expected to be confined to the outer perimeter of each flattened polymer particle. Filler particles are supposedly deposited onto hot polymer flattened particles and become diffused into the periphery of the polymer regions from the heat dissipation through the coating during molten droplet deposition. This diffusion is supported by the lower molecular weight of polymer chains on the outer layer of the polymer particle, resulting from exposure to the intense heating conditions of the flame (7). Furthermore, the high heat capacity of the inorganic filler particles promotes diffusion as heat is released during cooling of the filler particles. The extent of diffusion will be lowered as the filler interacts with the cooler flattened polymer core containing the higher molecular weight EMAA. Therefore, a contribution as small as 2.5 vol% can lead to large microstructural changes resulting in a markedly lower elongation to fracture.
[FIGURE 4 OMITTED]
The precise modulus, hardness, and ductility of the coating would be expected to be heterogenous throughout the sample on a micro-level. Inside a deposited polymer particle, the modulus and hardness would be lower and ductility would be higher. The total deformation within these regions can influence the surrounding particle-rich areas with higher modulus and hardness, and less-ductile regions that may fracture if the total deformation imposed is too great.
The effect of filler size on the tensile behavior of the 5 vol% filler composite coatings for the various filler particle sizes was determined. Table 3. All of the tensile curves exhibit the same shape, except for slightly lower ductility of the composite with the largest particles, possibly due to a greater possibility for points of weakness to occur in that blend.
As for tensile properties of flame-sprayed composites, the influence of filler on the peel strength to steel (or any substrate) has not yet been reported. Peel strength curves of composite coatings of a different filler size. with a filler content of about 5 vol%, are shown in Fig. 5. The 2.8-[micro]m filled composite attains its maximum peel strength much earlier in the extension portion of the test, whereas the larger-particle ceramic-filled materials exhibit much earlier local yielding. A low initial peel strength could be attributed to the slightly higher filler content. No significant change in the peel strength was evident with an increase in filler size, providing values of 1.46 [+ or -] 0.36, 1.37 [+ or -] 0.41 and 1.29 [+ or -] 0.27 N/m.
[FIGURE 5 OMITTED]
The effect of filler content (in the case for a 9-[micro]m powder) was to decrease the adhesion strength and strain to failure, Fig. 6. The peel strength for PF111, 2.54%, 5.46% and 7.25% filler was determined as 2.2 [+ or -] 0.76, 1.54 [+ or -] 0.56, 1.23 [+ or -] 0.41, and 0.94 [+ or -] 0.33 N/m, respectively. The lower adhesion strength is attributed to less interaction between the polymer and the substrate. The filler does not interact as adhesively with the substrate as does the polymer. preventing interaction between the polymer and the underlying metal substrate. Observation of the peeled surface shows some uncovered filler particles exposed on the sample surface. There is also an absence of the stick-slip phenomenon seen in the pure plastic coatings as shown in Fig. 6. A steady force for peeling composite coatings produces a smooth fracture surface as seen in Fig. 7.
A comparison between Fig. 5 and Fig. 6 also shows that the preheat temperature is an important parameter that influences the adhesion between the coating and the substrate. The peel strength decreases from about 1 N/mm for a preheat temperature of 100[degrees]C to 0.7 N/mm for a preheat temperature of 60[degrees]C. A higher preheat temperature is desirable for a stronger coating/ substrate adhesion, as previously reported for pure PF111 coatings (20). Bilayer coatings of PF111 on CaC[O.sub.3]-PF111 were then prepared by thermal spray to investigate the adhesion between composite coatings with the pure polymer coating. where the PF111 coating was used as a top layer. Peel strength for neat layers sprayed on composite materials decreased with a higher filler content, Fig. 8. This decrease is analogous to the decrease in peel strength of the monolayer composite coatings onto steel itself, as shown in Fig. 6.
Peel strength between PF111 and CaC[O.sub.3]-PF111 within the double-layer coating is higher than that between the CaC[O.sub.3]-PF111 monolayer and the substrate and even higher than the peel strength between PF111 alone and the substrate. Addition of an intermediate polymer coating, thus, improves the bonding between the polymer composite coating and the substrate. Previous work between two neat PF111 layers investigated after a peel test revealed many "weld points" (20). These points contributed to a higher peel strength for the bilayer coatings. Bilayer coatings of PF111 on PF111 composite also exhibited weld points on the peeled surface, but it is likely that the density of weld points would be lower than for PF111 on PF111 bilayer coatings, reduced by the presence of inorganic particles, Fig. 9. Even though the majority of the filler particles become completely covered by the polymer, the thin polymer coating on the inorganic particles can be seen in Fig. 9 to be readily peeled off to expose the filler particle.
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
The Erosion and Tribological Behavior of Composite Coatings
Erosion and abrasion characteristics of flame-sprayed polymeric coatings have not been widely reported in the literature. The wide use of these coatings for corrosion protection subjects them to mechanical abuse. most commonly under erosion environments (27, 28). Recent work on flame-sprayed alumina-reinforced nylon coatings has revealed the improvement of erosion corrosion resistance at low particle velocities (29). The material loss in composite coatings with different filler size but a similar filler volume fraction was investigated at an impingement angle of 45[degrees]. Figure 10 presents the cumulative volume loss due to erosion from coatings of a different filler size. It seems that the smallest filler produces the best erosion resistance, despite the lowest filler content of all the composite coatings. It is noteworthy that the coating with the smallest filler size was the only coating that was more abrasion-resistant than a coating of the polymer itself. The improved erosion resistance could be related to a more homogeneous coating at the micrometer level that increases the effective hardness of the coating.
[FIGURE 10 OMITTED]
The erosion volume loss of composite coatings with different filler content was also determined for this system; the results are shown in Fig. 11. An increase in filler content within the coating produced a higher material volume loss through erosion.
Erosion is considered a complex process related to fatigue, tribological behavior, and mechanical properties of the matrix. Any decrease in ductility of the matrix tends to lead to a lower fatigue resistance. which is detrimental to the erosion resistance of the matrix. An increase in hardness should, for materials exhibiting a well-bonded composite microstructure, benefit the erosion resistance of the matrix (30). These contrary factors coexist in composite materials. When the filler content is low, the influence of hardness will dominate the erosion resistance of the matrix. When the filler content is high, the converse influence of fatigue will become more accented (31).
The cumulative erosion volume loss of coatings decreased significantly when the incidence angle was changed from 45[degrees] to 90[degrees], Fig. 12. The same behavior occurs with carbon steel (32) and is explained by a number of aspects of the tribological behavior of materials. The wear resistance of PF111 is much lower than that of metals under normal friction conditions. Since the fatigue and wear resistance properties are two important factors for the erosion resistance of materials, the influence of these two factors will be modified as the incidence angle is changed. The fatigue resistance is the decisive factor for vertical impingement. With a decrease of impingement angle, the influence of wear resistance becomes the decisive factor (33, 34). PF111 polymer possesses poor wear resistance at the lower impingement angle and, thus, the erosion volume loss is very significant.
[FIGURE 11 OMITTED]
[FIGURE 12 OMITTED]
Erosion resistance can also be related to other tribological characteristics of the composite coating. The coefficient of friction and wear track width of composite coatings decrease with increasing filler content, Fig. 13. This smaller track width indicates that the addition of the filler improves the wear resistance of the polymer coating.
Examination of the wear trace by scanning electron microscopy revealed severe plastic deformation of the PF111 coating. Fig. 14a. The low ease of deformation of neat PF111 does not support the loads imposed by the ball-on-plate configuration, resulting in a high wear and friction coefficient. Inclusion of 2.54% CaC[O.sub.3] in the coating indicated less plastic deformation, suggesting an improvement in the hardness of the coating, Fig. 14b. The appearance of a ribbon at the edge of the wear trace implies that the ductility is still present. Increasing the filler content to 7.25% shows that no plastic defomation is visible, Fig. 14c. Slight scraping of the surface revealed regions of separation particles in patches that could represent a collection of filler particles and lead to a reduction in plastic deformation.
[FIGURE 13 OMITTED]
Micrographs of the erosion scar display a different surface depending upon the filler loading. Fig. 15. The scar on pure PF111 exhibits ripples that represent the erosion process of a plastic material, Fig. 15a. The ripples arise from the ability of the polymer to undergo plastic deformation in response to forces imposed by the erodent particles. In addition, erodent particles become more easily captured in the coating. Many cracks become visible in coatings containing filler (Fig. 15b), and this is attributed to the decrease in ductility with the addition of filler. Fatigue resistance of polymer coatings with filler is lowered concurrently with the strain. Coatings become damaged and large erosion debris leaves the surface in the form of chips as the filler produces more strain-sensitive regions within the coating. Fig. 15c. The strain-sensitive region will form around the inorganic filler since the inorganic filler cannot be deformed easily compared to the polymer matrix. The erosion process is, thus, considered a periodic compression deformation process induced by the impaction of erodent flux (35).
The results suggest that mechanical properties such as elastic modulus, hardness. coefficient of friction, and erosion resistance can be improved; however, composite coatings manufactured from mechanically blended filler and polymer powder appear sensitive to the filler size and content. This behavior is attributed to the localization of the filler particles within the boundaries and between the deposited polymer particles, where a high filler concentration is quickly established. Further work is necessary to investigate the mechanical properties of composite coatings using larger filler particles than those used here, and which can be fluidized to the same degree as polymer particles and may well establish a different microstructure within the composite coating. Larger filler particles can alter the extrusion of particles from the composite and hence alter the erosion resistance (36). The microstructure would be expected to be a more homogeneous distribution. as filler particles have a higher momentum and can embed into the soft polymer surface more effectively. Of further interest would be the fracture toughness of these coatings. This property can also be used to calculate the brittleness index, the ratio of hardness to fracture toughness, which has been found to provide a good indicator of erosion resistance for polymeric materials (37).
Although mechanical powder blends reported in this work offer a means of combining different materials into the spray material, it should be noted that another potential means of producing composite coatings (not studied here) involves mechanically incorporating both materials into each powder particle (38). The powder then is more expensive, because of the additional expense in powder processing, but would provide a higher deposition efficiency. and ensure a uniform content of the inorganic constituent. One means of incorporating an inorganic material with the polymer feedstock would be mechanofusion. Previous work has shown the layering of an inorganic material onto polymer particles (39). It should be noted, however, that the resulting microstructure can be different with mechanically alloyed and mechanofused powders owing to the different distribution of inorganic particles in the coating.
[FIGURE 14 OMITTED]
[FIGURE 15 OMITTED]
Inorganic filler content in mechanically blended powders delivered by a fluidized bed powder feeder through a flame spray torch decreases over a period of time. This decrease is greater for finer particles, owing to the different fluidization within the feeder and transport properties within the fast gas stream of the flame. Deposition rate and efficiency decrease with increasing filler content within the powder blend. The best filler transport characteristics, deposition efficiency, and deposition rate were obtained with the largest (36-[micro]m) CaC[O.sub.3] filler used.
Filler size does not have a significant effect on the overall tensile properties and peel strength of coatings. Filler content has a larger influence and decreases the tensile strain at fracture, but increases modulus. Peel strength decreases with filler loading for both the monolayer and bilayer coatings. However, the adhesion between a composite coating and a pure polymer is significantly higher than the composite coating on mild steel. The use of composite coatings is, thus, enhanced with the presence of a pure PF111 coating as the bonding layer.
Filler size also has little influence on the erosion of composites when the filler is added at 5 vol%. A filler content greater than 5 vol%, however, produces a coating with less erosion resistance than a pure PF111 coating. At an incidence angle of 90[degrees], the fatigue resistance dominates, while at lower incidence angles, the wear resistance plays a more important role. Despite the decrease in the coefficient of friction with filler loading, the mechanical properties of the composite are undesirable for erosion when the filler loading exceeds 5 vol%. Overall, low filler loadings can give simultaneous improvements in hardness, modulus, and erosion resistance.
Table 1. The Deposition Rate and Efficiency for Powder Blends With Different Filler Sizes and Amounts. + 36 [micro]m + 9 [micro]m + 3 [micro]m Particles Particles Particles PF111 (5.8 vol%) (5.2 vol%) (4.7%) Deposition rate 26 21 16 9 ([cm.sup.3]/min) Deposition 90 70 50 30 efficiency (%) Table 2. Tensile Properties of Coatings Containing a 9-[micro]m CaC[O.sub.3] Filler at Different Volume Contents. 2.5% PF111 CaC[O.sub.3] Tensile strength (MPa) 12.6 [+ or -] 0.9 11.8 [+ or -] 0.4 Strain at fracture (%) 512 [+ or -] 53 101 [+ or -] 14 1% Secant modulus (MPa) 196 [+ or -] 10 208 [+ or -] 12 5.5% 7.3% CaC[O.sub.3] CaC[O.sub.3] Tensile strength (MPa) 10.7 [+ or -] 0.5 10.4 [+ or -] 0.5 Strain at fracture (%) 72 [+ or -] 8 48 [+ or -] 6 1% Secant modulus (MPa) 198 [+ or -] 15 229 [+ or -] 19 Table 3. Tensile Properties of Coatings With Different Filler Sizes But Constant Volume Content. 2.8 [micro]m PF111 CaC[O.sub.3] Filler content (%) 0 4.66 Tensile strength (MPa) 12.6 [+ or -] 0.9 10.34 [+ or -] 0.7 Strain at fracture (%) 512 [+ or -] 53 79 [+ or -] 17 1% Secant modulus (MPa) 196 [+ or -] 10 203 [+ or -] 11 9 [micro]m 36 [micro]m CaC[O.sub.3] CaC[O.sub.3] Filler content (%) 5.24 5.76 Tensile strength (MPa) 10.45 [+ or -] 0.5 10.39 [+ or -] 0.6 Strain at fracture (%) 67 [+ or -] 13 63 [+ or -] 15 1% Secant modulus (MPa) 210 [+ or -] 8 212 [+ or -] 16
The authors thank NSF INT 9513462, which enabled this work to be performed. Dr. F. Y. Yan was supported by NSFC (National Natural Science Foundation of China) with grant # 59925513 and Dr. K. A. Gross by the Australian Research Grant with grant # F10017027.
1. E. Petrovicova and L. S. Schadler, Int. Mater, Rev., 47, 169 (2002).
2. L. Pawlowski. The Science and Engineering of Thermal Spray Coatings, Wiley, Chichester, England (1995).
3. D. A. Gerdeman and N. L. Hecht, Arc Plasma Technology in Materials Science, Springer-Verlag, New York (1972).
4. R. P. Krepski. "Thermal Spray Coating Applications in the Chemical Process Industries," MTI Publication No. 42 for NACE Intl. (1994).
5. C. C. Berndt and R. H. Unger, The Finishing Line, 18 (2002).
6. C. C. Berndt, "Preparation of thermal spray powders." Education module on thermal spray. Pub. ASM International. Ohio (1992).
7. J. A. Brogan. C. C. Berndt, G. P. Simon, and D. Hewitt, Polym, Eng. Sci., 38, 1873 (1998).
8. J. A. Brogan, J. Margolies, S. Sampath, H. Herman, C. C. Berndt, and S. Drozdz, in Advances in Thermal Spray Science and Technology, pp. 521-526, C. C. Berndt and S. Sampath, eds., ASM International, Materials Park, Ohio (1995).
9. C. C. Berndt. J. A. Brogan. G. Montavon. A. Claudon, and C. Coddet, in Proc. Int. Conf. on Composites Engineering, pp. 185-86. SSPC, Pittsburgh (1997).
10. M. L. Allan, C. C. Berndt, and D. Otterson. Geothermal Resources Council Transactions, 22, 20 (1998).
11. J. A. Brogan and C. C. Berndt, J. Mater. Sci., 32. 2099 (1997).
12. J. A. Brogan, R. Lampo, and C. C. Berndt. in Proc. 4th World Congress on Coating Systems for Bridges and Steel Structures, pp. 200-212. Missouri (1995).
13. M. L. Allan, C. C. Berndt. J. A. Brogan, and D. Otterson, in Thermal Spray Meeting the Challenges of the 21st Century. pp. 13-18, C. Coddet, ed., ASM International, Materials Park, Ohio (1998).
14. D. J. Varacalle, Jr., D. P. Zeek, K. W. Couch, D. M. Benson, and S. M. Kirk, in Thermal Spray: A United Forum for Scientific and Technological Advances, pp. 231-238. C. C. Berndt, ed., ASM International, Materials Park. Ohio (1997).
15. T. J. Steeper, W. L. Rigg II, A. J. Rotolico. J. E. Nerz. D. J. Varcalle, Jr., and G. C. Wilson, in Thermal Spray Coatings: Research, Design and Applications, pp. 31-36, C. C. Berndt and T. F. Bernecki, eds., ASM International. Ohio (1993).
16. P. Ostojic and C. C. Berndt, J. Surf. Coat. Tech., 34. 43 (1988).
17. I. A. Fisher. Int. Metall. Rev., 17, 117 (1972).
18. T. D. Fender, Mater. Technol., 11, 16 (1996).
19. D. S. Parker. Plast. Surf. Eng., 82, 20 (1998).
20. F. Y. Yan. K. A. Gross, G. P. Simon, and C. C. Berndt. J. Appl. Polymers, 88, 214 (2003).
21. P. J. Loustaunau and D. Horton, Materials Performance, 33, 32 (1994).
22. R. Knight, X. Fang, and T. E. Twardowski, in Thermal Spray 2001: New Surfaces for a New Millennium, pp. 361-368. C. C. Berndt, K. A. Khor, and E. F. Lugscheider. eds., ASM International, Ohio (2001).
23. ASTM D3167-97 Standard Test Method for Floating Roller Peel Resistance of Adhesives (1997).
24. ASTM G76-02 Standard Test Method for Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets (2002).
25. ASTM G133-02 Standard Test Method for Linearly Reciprocating Ball-on-Flat Sliding Wear (2002).
26. J. A. Brogan, C. C. Berndt, A. Claudon, and C. Coddet, in Thermal Spray: Meeting the Challenges of the 21st Century, pp. 1173-1178, C. Coddet, ed., ASM International, Materials Park, Ohio (1998).
27. T. Race, V. Hock, and A. Beitelman. J. Protective Coatings & Linings, 6, 37 (1989).
28. A. Chmiel, V. Mottola, and J. Kauffman, J. Protective Coatings & Linings, 6, 23 (1989).
29. H. Chen, H. Zhao. J. Ou, H. Shao. and S. Zhao S. Wear, 233-35. 431 (1999).
30. J. Li and I. M. Hutchings, Wear, 135, 293 (1990).
31. A. Brandstadter. K. C. Goretta, J. L. Routbort, D. P. Groppi, and K. R. Karasek, Wear, 155 (1991).
32. N. M. Barkoula and J. Karger-Kocsis, J. Mater, Sci., 37. 3807 (2002).
33. M. G. Gee, R. H. Gee, and I. McNaught, in Proc. 14th Int. Conf. on Wear of Materials, Washington, D. C. (2003).
34. Q. Chen and D. Y. Li. in Proc. 14th Int. Conf. on Wear of Materials, Washington, D. C. (2003).
35. T. S. Chow and R. C. Penwell, in Metal-filled Polymers, pp. 227-256, S. K. Bhattacharya, ed., Marcel Dekker Inc., New York (1986).
36. G. P. Tilly, Wear, 16, 447 (1970).
37. K. J. Friedrich, J. Mater. Sci., 21. 3317 (1986).
38. L. Sun, C. C. Berndt, and K. A. Gross, J. Biomater. Sci. Polym. Ed., 13, 977 (2002).
39. J. A. Brogan. K. A. Gross, Z. Chen, H. Herman, and C. C. Berndt, in National Thermal Spray Conference, pp. 159-164. C. C. Berndt and S. Sampath. eds., ASM International, Cleveland (1994).
F. Y. YAN (1,2), K. A. GROSS (1*), G. P. SIMON (1), and C. C. BERNDT (3)
(1) School of Physics and Materials Engineering
VIC 3800, Australia
(2) State Key Laboratory of Solid Lubrication
Lanzhou Institute of Chemical Physics
Chinese Academy of Sciences
Lanzhou 730000, People's Republic of China
(3) Department of Materials Science and Engineering
Stony Brook University
Stony Brook, NY 11794-2275
*To whom correspondence should be addressed.
|Printer friendly Cite/link Email Feedback|
|Author:||Yan, F.Y.; Gross, K.A.; Simon, G.P.; Berndt, C.C.|
|Publication:||Polymer Engineering and Science|
|Date:||Aug 1, 2004|
|Previous Article:||Indentation of poly(formaldehyde), reinforced nylons and poly(ethyl terephthalate).|
|Next Article:||Modeling and simulation of stretch blow molding of polyethylene terephthalate.|