Transfer efficiency for airless painting systems.

Spray transfer efficiency (TE) is defined as the mass fraction of sprayed paint which is deposited on the intended target, the remainder of the sprayed paint becomes undesirable overspray. The relationship between TE and gun supply pressure (or paint mass flow rate), gun-to-target distance, gun traverse speed, the angle of the spray gun relative to the target (gun-to-target angle), plus spray cone angle is reported herein for a typical fan spray system. Experimental results indicate that spray momentum rate (SMR) and droplet size dictate the TE for the various combinations of parameters considered here. The key finding is that TE correlates with SMR and spray mean drop size (Sauter mean diameter, or [D.sub.32]) via an expression of the form TE = a + b SMR - c (SMR)[.sup.2] + d [D.sub.32], where a, b, c, and d are coefficients determined by fitting the experimental data, and SMR is estimated via SMR = [m.sup.2]/[rho]A, where the paint mass flow rate m, the paint density is [rho], and the gun exit orifice effective tip cross sectional area is A. This expression accounts for physical phenomena that govern sprayed droplet deposition characteristics, such as entrainment, bounce-back, and drop size.

Experimental results also show that, for the range of parameters studied, gun traverse speed has no effect on TE, but increasing the angle of the spray gun relative to the target (gun-to-target angle), increasing the spray cone angle, or increasing the gun-to-target distance will decrease TE.

Keywords: Application methods, latex, pollution, spray application, transfer efficiency

**********

Transfer efficiency (TE) is a key concern for commercial spray painting processes. Typical targets include building interior and exterior surfaces, furniture and wood products, and automotive refinishing. A number of applicator types are available and in use, but airless systems largely dominate the market. Consequently, airless systems were considered here.

Regardless of target surface and applicator type, TE influences painting costs (higher TE results in less waste and therefore reduced materials costs), environmental pollutants (higher TE minimizes solid and vapor pollutants), and operator safety (higher TE means less solvents and potentially respirable paint particles for operators to inhale). Improving TE is therefore of considerable interest.

Various parameters have been shown to influence TE. They can be divided into the physical (paint percent solids, density, viscosity, and surface tension), the operational (operator skill level, spray equipment type, gun supply pressure, paint mass flow rate, spray cone angle, angle of spray relative to target, gun-to-target distance, gun traverse speed), and the geometrical (target size and shape). Several previous studies shed light on how some of these quantities influence TE. These are discussed below.

Walberg (1) and Snowden-Swan and Worner (2) reported that operator skill level and efficiency has a major influence on TE--maximum TE approached 70% for an experienced operator and only 60% for a novice. Their results are consistent with those of Ehrenhofer, (3) who suggested that switching to automated spray systems for finishing plants (which relied primarily on manual spraying), apart from increasing TE, can also improve finish quality, reduce labor overhead, lower VOC emissions, reduce paint waste, and lessen the health hazard for the employees. In a later study, Lamancusa (4) stated that converting to robotic spray applicators might improve TE by 10%.

Snowden-Swan and Worner (2) also conducted experiments with various spray systems using paints of different percent solids to check its influence on TE. Perhaps not surprisingly, they failed to observe any clear relationship between paint percent solids and TE.

Workpiece geometry and size are uncontrollable factors influencing TE, but the distance of the spray gun to the target (painting distance), and the angle and speed of the spray gun are adjusted by the operator. The effect of increasing painting distance is generally to reduce transfer efficiency, as was observed by Hicks et al. (5) They found that increasing gun-to-target distance from 17.8 to 35.6 cm (at a differential air pressure of 262 kPa for a paint viscosity of 57 cSt) decreased TE from 78 to 54%. The decrease in TE with increased painting distance was attributed to the accompanying increase in spray cross section. The working fluid for this study was an aqueous solution of Staley high fructose corn syrup (HFCS) with a specific gravity of 1.3.

[FIGURE 1 OMITTED]

Hicks et al. (5) also reported that increasing paint viscosity increases paint transfer efficiency by shifting the atomized drop size distribution to larger drops, which exhibit greater drop transfer efficiency, i.e., the fraction of paint droplets of a particular size class that are deposited on the target. They found that increasing paint viscosity from 57 to 106 cSt (at a painting distance of 25.4 cm and a differential air pressure of 262 kPa) increased TE from 68 to 75%. Muir, (6) however, argued that viscosity cannot be the only parameter for estimating TE, noting that the energy of atomization competes with viscosity and has a greater effect on TE. He further stated that paint viscosity is a physical factor that determines the amount of energy needed to obtain a desired atomization level. Lowering the viscosity will reduce the energy required to atomize the paint, leading to higher TE. Muir (6) does agree with Hicks et al. (5) when he notes that for a given amount of atomization energy, a higher viscosity yields a higher TE. Dependence of atomization and TE on paint physical parameters has been studied extensively and is discussed in greater detail below.

[FIGURE 2 OMITTED]

Muir (6) and Ewert et al. (7) state that atomizing air pressure greatly influences TE in air guns. The higher air pressure is believed to cause turbulent atomization at the gun tip, which generally decreases TE. Turbulence, however, is desired near the workpiece to increase TE by enhanced normal velocity fluctuations. (8) Small drops which would be swept past the target by the velocity component parallel to the target get caught up in turbulent eddies and the fluctuating normal component of velocity transports them to the target surface.

Ehrenhofer, (3) in a general summary article which did not specify the fluids sprayed, suggested that there are three key factors in nonelectrostatic spray system design and operation that directly influence TE. These are forward velocity, bounce-back, and turbulence. Forward velocity is the speed at which the paint travels towards the target and is directly related to gun supply pressure--the higher the gun supply pressure the higher the speed. If the coating leaves the gun too quickly, the spray pattern may overshoot the target, increasing overspray and requiring booth cleanup. Bounce-back refers to the paint droplets impacting the target and then bouncing off without adhering. Other investigators support this view, noting that an excessively high velocity also causes bounce-back and lowers TE. (9,10) The importance of turbulence is discussed below.

Contrary to these findings, Hicks et al. (5) were surprised to observe that while an increase in atomization pressure caused a shift in drop size distribution towards smaller drops, there was no appreciable difference in TE when varying air supply pressures. They attributed this insensitivity of drop transfer efficiency to air supply pressure variations to two competing effects--entrainment and axial momentum. Sprays created using higher air pressures have increased entrainment of ambient air, resulting in greater airflows and higher air velocities near the workpiece (target). Under these conditions, it is more difficult for drops to penetrate the boundary layer near the workpiece and deposit on the target. However, the larger air velocities near the gun impart greater axial momentum (directed towards the workpiece) to the drops. The higher droplet axial momentum is apparently able to compensate for the increase in air velocities and boundary layer turbulence intensity above the workpiece. The net effect results in little change in TE.

[FIGURE 3 OMITTED]

A possible conclusion from the results obtained in references 3 and 5 is that combining the effect of paint viscosity and gun supply pressure implies that drop size distribution becomes critical in influencing TE. This will be verified in the current study.

There is also some sentiment in the spray painting industry that TE cannot be predicted. Shaffer (11) claims that because TE depends on numerous combinations of different parameters, there can be no standard choice of paint or workpiece. He also claims it is improper to assign a value of TE to each particular type of spray system, as each one used for a different purpose would have its own set of advantages. These claims are based largely on his experience of working in the painting industry. Shaffer (11) also conjectured that traverse velocity would influence TE, but no previously published work has established this fact.

In addition to the dependence on such operational parameters, rheological properties and physical parameters of the paint will influence atomization and drop size, and hence TE. Xing et al. (12) studied non-Newtonian waterborne latex coatings with various sizes of latex particles and different thickener compounds. They found a relationship between Sauter mean diameter and orifice diameter for airless sprays originating from fan nozzles. The correlation depended upon surface tension, extensional viscosity, injection densities of the air and coating, and injection pressure differential. Other parameters such as the storage moduli, at least for low deformation rates, did not correlate with drop sizes.

In any paint spray many factors influence the drop size and the velocity at which drops move toward the workpiece. Larger drops are usually able to penetrate the boundary layer due to their higher momentum and are, therefore, more likely to get deposited on the surface, increasing TE. Smaller drops have lower momenta and are unable to penetrate the boundary layer, resulting in overspray. Larger workpieces tend to have higher TEs, and TE decreases as the gun-to-target distance increases. Gun supply pressure has a mixed influence on TE. Tip geometry is another important parameter, in particular orifice diameter. As orifice size increases, TE also increases because a larger orifice nozzle most often results in the formation of larger drops.

Based on the general knowledge of spray painting processes outlined above, the following parameters and their relationship to TE were considered during this study:

[FIGURE 4 OMITTED]

* Gun supply pressure (equivalent to paint mass flowrate for a given nozzle orifice)

* Gun-to-target distance

* Gun traverse speed

* Angle of the spray gun relative to the target (gun-to-target angle)

* Tip geometry (both effective orifice diameter and spray cone angle)

The current study considers how each of these five factors influences transfer efficiency.

In contrast to previous studies (almost all of which report data from spray experiments that were performed manually, resulting in variations in transfer efficiency values without taking into account the operator efficiency factor), operating conditions in this study are well-defined: paint mass flow rate is controlled to within [+ or -] 2.5%; gun traverse speed is controlled to within [+ or -] 5.0%; gun-to-target angle is controlled to within [+ or -] 1[degrees]; and gun-to-target distance is known to within [+ or -] 1%. These well-defined operating conditions were achieved through use of an automated traverse.

Finally, although previous studies provide information on how TE is affected by various operational and geometrical factors, the fundamental mechanisms causing these changes have not been identified. The current study establishes two key physical factors, spray momentum rate and droplet size, as being primarily responsible for TE. A correlation for TE, in terms of spray momentum rate and droplet size, is presented. It can be used to guide future spray tip designs.

EXPERIMENTAL APPARATUS

The experimental investigation utilized a commercially available airless gun (atomizer assembly) to spray paint onto a target, a Malvern particle size analyzer (to measure drop size distributions), a test facility with a computer-controlled three-axis traverse that positioned the gun relative to the target, a convection oven for drying paint on the targets, a mass flow rate meter to monitor paint supplied to the gun, and a centigram balance to determine the amount of paint deposited on each target plate. Recall that a mechanical traverse system was chosen to eliminate the influence of operator variability on transfer efficiency. Sherwin-Williams A100 latex flat white was used as the working fluid. This is a water-based paint with characteristics typical of automotive or industrial paint behavior. The manufacturer reports the following properties: VOC (less exempt solvents): 147 g/L (1.22 lb/gal) volume solids: 32 [+ or -] 2%, weight solids: 48 [+ or -] 2%, and weight per gallon: 10.9 lb. Settles (13) reported that A100 paint exhibits rapid initial shear thinning for strain rates of 0-40 se[c.sup.-1] and a final, nearly constant asymptotic value at high strain rates (measured up to 550 se[c.sup.-1]). All heavily loaded paints are expected to exhibit non-Newtonian or viscoelastic behavior. Settles shows a plot of shear viscosity as a function of strain rate and states that the extensional viscosities are expected to be as much as 1000 times greater. No rheological analysis of the paint was done in the current study.

[FIGURE 5 OMITTED]

The spray gun was mounted on a computer-controlled three-axis Unislide[TM] traverse, which allowed it to be positioned relative to the target with an accuracy of [+ or -] 1 mm. The traverse initially accelerated the gun to the desired traverse speed, maintained that speed while the spray was over the target, and then decelerated the gun back to rest. The entire motion was programmed and repeatable.

The traverse assembly was located above a ventilation box with a honeycomb top cover, which provide a uniform downward velocity field without effectively disturbing the spatial mass distribution of the spray. (12) It also supported the target. Sprayed paint was vented through an exterior wall by an exhaust fan.

[FIGURE 6 OMITTED]

Paint flow rate was monitored by a Micro Motion[TM] mass flow rate meter having an accuracy of [+ or -] 0.03 kg/min, as determined by calibration. Test samples were dried using a Blue MTM single wall transite convection oven.

The procedure to determine TE closely followed ASTM standards [designation D 5327-97]. The following protocol was used: (a) agitate the test paint in a closed container for at least 30 min before paint samples are sprayed; (b) select a target panel such that it has a minimum length of 15 cm above and below the edge of the spray pattern major axis, then measure the width of the target panel; (c) cut a strip of aluminum foil longer than the length of the target panel; (d) weigh the foil strip using a centigram balance and record its weight; (e) mount the foil over the panel and position it below the portion of the paint gun traverse path having uniform velocity; (f) choose a particular combination of parameters for the test, i.e., paint pressure at the spray gun, spray gun-to-target distance, spray gun traverse speed, and gun-to-target angle., then establish a mass flow rate for that parameter set using the mass flow rate meter; (h) use the three-axis traverse to move the paint gun across the target in a single pass; (i) remove the aluminum foil from the target and heat it to 120[degrees]C for four hours in the oven; (j) remove the foil from the oven and let it cool for 15 min; (k) weigh the paintcoated foil; (l) the mass of the paint solids deposited is the difference between the weight of the foil after painting and baking, and the weight of the foil before painting; and (m) calculate TE as described below.

Performing at least three tests for each condition ensures statistical confidence in the transfer efficiency measurements. The fractional uncertainty in transfer efficiency, i.e., [delta](TE)/TE, is kept within [+ or -] 5% through the use of instruments described above.

A Malvern particle size analyzer was employed to obtain drop size distributions as follows. A spray was established for a given combination of parameters required for a test. The particle size analyzer was positioned to collect drop size distribution data along a line of sight parallel to the minor axis of the elliptical cross section spray. The spray was traversed along the major axis while the drop sizing instrumentation was held stationary. This configuration is ideally suited for ensuring low beam obscuration (due to the short illuminated sample length) and, therefore, avoids problems associated with multiple scattering of light caused by a high density spray or a long path length.

In order to directly correlate the effect of varying input parameters on the resulting drop size distribution for the entire spray, some type of global drop size distribution must be calculated. Snyder (14) defines two different techniques for determining global drop size distributions. One involves the weighted combination of a series of distinct line-of-sight measurements. The other involves continuous sampling of the spray as the gun is swept across the Malvern particle size analyzer laser beam. We have used the latter technique because it allows faster measurement of spray drop sizes. The decreased data collection time not only reduces testing time, but also reduces the amount of paint required. This is beneficial for both cost and safety concerns.

The resulting drop size distribution is processed by a custom software routine operating on a PC. A Rosin-Rammler distribution of weight fractions is assumed and [D.sub.32] (Sauter mean diameter) calculated. [D.sub.32] was chosen because transfer efficiency is dictated by droplet ballistics, which in turn is controlled by the droplet volume-to-surface area due to the competition between inertia (proportional to diameter cubed, [d.sup.3]) and aerodynamic drag (proportional to [d.sup.2]).

Uncertainty in transfer efficiency measurements was estimated using a method similar to that suggested by Fox and McDonald. (15) We assume that the equipment was constructed correctly and calibrated properly to eliminate fixed errors.

Transfer efficiency is defined as follows:

TE = [([m.sub.f] - [m.sub.i]) X 100]/[m.sub.out] (1)

where [m.sub.f] = mass of the sample after painting and drying, [m.sub.i] is the initial mass of the sample, and [m.sub.out] = mass of paint exiting the nozzle during spraying. The uncertainty in TE is given by:

[u.sub.TE] = [+ or -] [square root of (([m.sub.f] X [[u.sub.mf]/[[m.sub.f] - [m.sub.i]]])[.sup.2] + ([m.sub.i] X [[u.sub.mi]/[[m.sub.f] - [m.sub.i]]])[.sup.2] + [u.sub.mout.sup.2])] (2a)

where

[u.sub.mf] = [+ or -] [0.01/[m.sub.f]] and [u.sub.mi] = [+ or -] [0.01/[m.sub.i]] are the least count of the centigram balance. The total mass of paint sprayed is computed from

[m.sub.out] = [dot.m] X t X %sol (2b)

where [dot.m] = paint mass flow rate, t = time the spray is activated (i.e., the target length, s, divided by the gun traverse speed, v), and %sol = percent solids present in the paint.

Given the relative uncertainties in gun traverse speed, target length, mass flow rate, and percent solids, as follows:

[u.sub.v] = [+ or -] 0.05, [u.sub.s] = [+ or -][1.5/s], [u.sub.mdot] = [+ or -]0.025, and [u.sub.%sol] = [+ or -]0.02,

the corresponding uncertainty in TE ranges from 2.5 to 3.4%. Note that [u.sub.v] and [u.sub.%sol] are manufacturer specifications and [u.sub.mdot] is determined from fluctuations observed when reading the mass flow rate meter.

RESULTS AND DISCUSSION

A typical fan spray system was used and TE was measured for gun supply pressures from 10.4 to 17.3 MPa, gun-to-target angles of 0[degrees] to 45[degrees], gun traverse speeds between 75 and 300 mm/sec, gun-to-target distances from 30 to 45 cm, tip geometries (effective orifice diameters) of 0.38 to 0.48 mm, and nominal spray cone angles between 30[degrees] and 60[degrees]. See Figure 1 for a definition of all quantities. Data illustrating the effect of each quantity on TE are presented below. (Note that each tip geometry is identified by a four-digit code, with the first two digits representing the effective orifice diameter in thousandths of an inch and the second two specifying the nominal spray cone angle. For example, 15.50 describes a spray tip having an effective diameter of 0.015 in. (0.38 mm) and a nominal cone angle of 50[degrees].)

Figures 2 and 3 show how TE scales with gun supply pressure (equivalent to TE versus paint mass flow rate). In these two figures, we compare the performance of tips 15.50, 17.50, and 19.50 at supply pressures between about 10 and 20 MPa for gun-to-target distances of 30 cm (Figure 2) and 45 cm (Figure 3). Both figures show that TE first increases with gun supply pressure, passes through a maximum value, then decreases with a further increase in gun supply pressure, regardless of tip type and gun-to-target distance. Note in Figure 2 that the TE variation from tip to tip is within experimental uncertainty for gun supply pressures from 13.8 to 17.2 MPa. In contrast, this figure shows tips 17.50 and 19.50 have higher TE values than tip 15.50 at the lower gun supply pressure of 10.3 MPa. In Figure 3, note that a gun supply pressure of 13.8 MPa provides the optimum TE for all three tips. Also note from Figure 3 that all three tips exhibit very similar variations in TE with gun supply pressure, although the TE for tip 17.50 is higher than that for tips 15.50 and 19.50 at this gun-to-target distance.

[FIGURE 7 OMITTED]

In preparation for a discussion of physical mechanisms underlying TE, we note at this point that an increase in gun supply pressure imparts higher momentum to the spray, which permits drops to penetrate the gas-phase boundary layer formed at the target surface. An increase in gun supply pressure also results in a reduction in drop size. These effects compete to set the level of TE. Because smaller drops penetrate the boundary layer formed on the target surface less easily than larger ones due to their relatively lower momentum, we would expect a decrease in TE. Furthermore, since larger sized tips are expected to generate larger drops based on general atomization physics, we anticipate larger TE values for tips 19.50 and 17.50 than for tip 15.50.

While either Figure 2 or 3 illustrates the effect of gun supply pressure on TE, one must compare data from both figures to determine how TE varies when gun-to-target distance is changed. Comparing the figures clearly shows that increasing gun-to-target distance decreases TE. This behavior is consistent with that of previous studies. (5)

Figure 4 shows typical TE versus gun traverse speed data, in this case taken for tip 17.50 at gun-to-target distances of 30, 45, and 59 cm. Note that TE does not vary significantly over the range of gun traverse speeds tested. Since the data of Figure 4 are qualitatively similar to data from all tips and gun-to-target distances, we conclude that there is no variation in TE with gun traverse speed. This result addresses Shaffer's (11) conjecture that gun traverse speed would affect TE--evidently it does not over a range of speeds typical of manual spraying operation (75 to 300 mm/sec).

[FIGURE 8 OMITTED]

It is not possible to further explain why our results differ from the opinions of Shaffer (11) for at least three reasons. First, the article by Shaffer is an editorial based on the author's opinions (backed by years of practical experience) but not by experimental data. Second, Shaffer does not specify the types of paints sprayed, instead remarking only that they are a number of types (lacquers, which he claims are nearly Newtonian, as well as other types which are non-Newtonian). Third, Shaffer does not indicate what types of equipment were used to spray the paint, nor how transfer efficiency was evaluated.

Figure 5 demonstrates how TE changes with gun-to-target angle. Again, sprays were formed using three tips: 15.50, 17.50, and 19.50. Data presented here are typical of those collected; we have arbitrarily chosen results for a gun-to-target distance of 45 cm and a gun supply pressure of 13.8 MPa. As Figure 5 shows, TE does not depend on gun-to-target angle at the larger gun-to-target distance for tips 17.50 and 19.50, but does for tip 15.50. Thus, tips 17.50 and 19.50 not only have higher TE values than tip 15.50, but also show a more desirable independence of TE with respect to gun-to-target angle. This minor variation in TE with gun-to-target angle may be partly responsible for operator-related variations in TE reported by earlier groups. (1-4) An experienced operator may learn that keeping the gun axis perpendicular to the target improves TE.

A second set of tips (15.30 and 15.60) was used to investigate the influence of spray cone angle on TE. These two tips produce sprays having two different cone angles (nominally 30[degrees] and 60[degrees], respectively) but the same orifice size as tip 15.50 (the baseline case). Tip speed in all cases was 250 mm/sec.

[FIGURE 9 OMITTED]

Figure 6 shows typical TE versus gun supply pressure data for tips 15.30 and 15.60. In this case we have arbitrarily chosen a gun-to-target distance of 30 cm. The figure shows that TE decreases as gun supply pressure increases in both cases. This follows the trend for tips 15.50, 17.50, and 19.50 reported earlier. More importantly, this figure shows that TE decreases as spray cone angle increases (compare the 15.30 data to the 15.60 case).

Figure 7 shows TE versus gun-to-target distance for tips 15.30 and 15.60 at a gun supply pressure of 13.8 MPa and a gun-to-target angle of 0[degrees]. Note that TE generally decreases as gun-to-target distance increases from 30 to 45 cm, although the experimental uncertainty bars overlap somewhat. As demonstrated in the previous figure, TE decreases as spray cone angle increases, but experimental uncertainty is high.

Figure 8 shows how TE varies with gun-to-target angle. Three tip types (15.30, 15.50, and 15.60) were used and three gun-to-target angles (0[degrees], 22.5[degrees], and 45[degrees]) considered, all at a gun-to-target distance of 30 cm and a gun supply pressure of 13.8 MPa. Again, the data show that TE generally decreases as spray cone angle increases, with the decrease being most significant for the 22.5[degrees] and 45[degrees] cases. The small decrease and leveling off is within experimental uncertainty for the 0[degrees] case. Also note that TE decreases with an increase in gun-to-target angle, except for tip 15.50 where TE is nearly the same at 22.5[degrees] and 45[degrees].

Once the behavior of TE as a function of atomizer type (i.e., tip diameter and spray cone angle), target geometry (i.e., gun-to-target distance and gun-to-target angle), and operating conditions (i.e., gun supply pressure and gun traverse speed) was established, further investigations were performed to study the mechanisms responsible for these changes, since previous studies (5, 16) suggest that spray momentum rate (SMR) plays an important role. SMR is the force required to hold an infinitely large target plate stationary and perpendicular to the spray. In this case, paint leaving the gun will form a spray whose axial direction will be turned 90[degrees] by the presence of the plate (a classical stagnation point flow). Since the liquid exiting the gun possesses axial momentum, and the plate transforms the axial spray motion into radial spray motion, a force must be exerted. This force equals the spray momentum rate. SMR was estimated for different tips operating under various conditions using the following relationship:

SMR = [dot.m.sup.2] / [rho]A (3)

where [dot.m] is the paint mass flow rate, [rho] is the paint mass density, and A is the effective cross-sectional area of the tip orifice. Because SMR is proportional to the square of the mass flow rate, and since the mass flow rate is proportional to the tip area when gun supply pressure is kept constant, SMR is greater for tip 19.50 than for tip 15.50 at equal gun supply pressures. Also, increasing gun supply pressure for a given tip size increases the mass flow rate resulting in a higher value of SMR.

In summary, gun supply pressure and orifice size are the key factors that influence SMR. Furthermore, for the paint and atomizer used, the gun traverse speed can be ignored since we have shown that changes in this parameter have little impact on TE. Finally, the influence of gun-to-target angle has been ignored, because we assume that an experienced operator will keep the gun axis perpendicular to the target surface and that a robotic paint system will be programmed to do so. Under these assumptions, only the influence of SMR and spray mean drop size need be considered.

[FIGURE 10 OMITTED]

Typical spray mean drop size data are reported in Figures 9 and 10 as Sauter mean diameter ([D.sub.32]). Each value was obtained by averaging at least three ensembles of drop data acquired in a plane perpendicular to the spray. Uncertainty bars are too small to plot because statistical scatter in the mean drop size data was on the order of the symbol size. Both figures demonstrate the expected decrease in [D.sub.32] with an increase in gun supply pressure. Figure 9 also shows that [D.sub.32] increases as tip orifice diameter increases for a fixed gun-to-target distance of 30 cm. This behavior is also expected. Figure 10 shows that [D.sub.32] values for the 17.50 case are highest, followed by those for the 19.50 and 15.50 cases for a fixed gun-to-target distance of 45 cm. The position of the 17.50 case data is somewhat unexpected.

[FIGURE 11 OMITTED]

[FIGURE 12 OMITTED]

All TE data for the 1X.50 series tips were plotted versus SMR in Figure 11. Based on these results, it was first assumed that TE is related to SMR via a quadratic expression, an assumption consistent with the explanation of Hicks and Senser (8) that an increase in SMR should result in an increase in TE, because drops having higher axial momenta are usually more able to penetrate the boundary layer along the target and deposit on it. Quadratic behavior of TE versus SMR is also consistent with the work of McCarthy (16) who showed that droplet bounce-back phenomena became more important as spray momentum rate was increased. Thus, higher SMR values should eventually lead to lower TE because droplet bounce-back will eventually dominate.

Drop size is also believed to play a role in influencing TE. Recall that Hicks and Senser (8) suggested that TE decreases as [D.sub.32] decreases. Thus, SMR and [D.sub.32] should compete to influence TE. A correlation of TE against a quadratic function of SMR and a linear function of [D.sub.32] was developed for the 1X.50 series data as follows:

TE = 55.2 + 20.1 SMR - 3.12 SM[R.sup.2] + 0.118 [D.sub.32] (4)

In the expression above, TE is expressed in percent, SMR in Newtons (N), and [D.sub.32] in [micro]m.

Equation (4) achieves agreement between experimental and calculated TE to within [+ or -] 1.5%, as shown in Figure 12. The correlation has an [r.sup.2] value of 0.898 (goodness of fit parameter from the least-squares quadratic fit). Results for the 15.30 and 15.60 tips are also plotted on this figure to assess the validity of the fit that was determined from the 1X.50 series tips. The agreement is quite reasonable.

We now further discuss points raised regarding the behavior of TE in Figures 2 and 3. An explanation is provided based on how TE depends on SMR and [D.sub.32]. In Figure 2 the value of TE is approximately constant (within experimental uncertainty) for all three tips in the range of gun supply pressures from 13.8 to 17.2 MPa, i.e., the highest pressures. Below this pressure range the value of TE becomes significantly lower for the 15.50 tip. SMR for tip 15.50 was estimated using equation (3) and found to be 1.25 at 10.3 MPa, whereas it is 1.95 and 2.18, respectively, for tips 17.50 and 19.50. Recall also from Figure 9 that [D.sub.32] is lower at 10.3 MPa for tip 15.50 ([D.sub.32] = 58 [micro]m) than for tips 17.50 and 19.50 ([D.sub.32] = 60 and 65 [micro]m, respectively). Thus, the primary reason that TE for tip 15.50 is substantially lower at 10.3 MPa than for the other two tips is that SMR is lowest for that case. As noted, [D.sub.32] for that case is also the lowest which serves to decrease TE even further.

It was observed in Figure 3 that TE for tip 17.50 is higher than that for tips 15.50 and 19.50 at the 45 cm gun-to-target distance for the pressure range of 10.3 to 17.2 MPa. This is because the relative SMR values for tips 15.50, 17.50 and 19.50 are the same as those for the 30 cm target distance (Figure 2), but [D.sub.32] for tip 17.50 is unexpectedly greater than that for tips 19.50 and 15.50 for all values of pressure in this range. Refer again to Figure 10. The larger drop size for tip 17.50 counters the effect of SMR and increases TE.

This discussion underscores the importance of both SMR and [D.sub.32] on TE. These parameters can combine constructively or destructively, and thus influence the degree of overspray. Both effects are accounted for in equation (4).

SUMMARY AND CONCLUSIONS

Transfer efficiency was measured for a typical fan spray using five gun tips of different orifice sizes (effectively 0.38 to 0.48 mm, or 0.015 to 0.019 in.) and spray cone angles (nominally 30[degrees], 50[degrees], and 60[degrees]). Four operational parameters (gun-to-target angle, gun-to-target distance, gun traverse speed, and gun supply pressure) were varied to test their effect on TE and spray drop size. Experimental results are summarized as follows:

* TE increased as gun supply pressure increased from 10.3 to 17.2 MPa, at which point it started to diminish.

* The angle of the spray gun relative to the target (gun-to-target angle) did not have much influence on TE, except for smaller orifice tips for which TE decreased as the gun-to-target angle was increased to 45[degrees].

* TE showed no significant dependence on gun traverse speed, over a range of speeds typical of those used by spray equipment operators.

* TE decreased as the cone angle of the spray increased.

* Drop size decreased as gun supply pressure increased.

* Drop size increased as orifice size increased.

Based on the experimental data, TE was related to SMR through the polynomial relationship TE = 55.2 + 20.1 SMR - 3.12 SM[R.sup.2] + 0.118 [D.sub.32]. The corresponding correlation coefficient is [r.sup.2] = 0.898. Here, TE is expressed in percent, SMR in N, and [D.sub.32] in microns.

From the results of this study, it is concluded that spray momentum rate (SMR) has the strongest influence on TE, while mean drop size ([D.sub.32]) has a lesser effect on TE. Both SMR and [D.sub.32] are influenced by gun supply pressure, but in opposite ways. Increasing gun supply pressure causes [D.sub.32] to decrease, but at the same time SMR increases due to the corresponding increase in paint mass flow rate. The data show that SMR dominates, and is thus the controlling factor. In addition, the data also show that the change in [D.sub.32] over the range of pressures considered in this study (10.3 to 17.2 MPa) is small, which is partly why it has only a minor effect on TE.

In a practical sense, the results of this study demonstrate that various combinations of operational parameters should be chosen to increase SMR, and therefore maximize TE. In addition, the results show that gun traverse speed has little influence on TE for values between 75 and 300 mm/sec (4 and 12 in./sec), but that gun-to-target angle can have a significant effect on TE. Therefore, operators should find that keeping the spray perpendicular to the target surface is more important than the speed at which the gun moves.

References

(1) Walberg, A.C., "Boost Overall Transfer Efficiency," Industrial Finishing, 66, No. 5, 20-30 (1990).

(2) Snowden-Swan, L. and Worner, P., "Determining Transfer Efficiency and VOC Emissions," Metal Finishing, 91, No. 6, 73-78 (1993).

(3) Ehrenhofer, K.R., "Selecting Automatic Spray Guns," Industrial Finishing, 63, No. 11, 31-35 (1987).

(4) Lamancusa, J.P., "Pollution Prevention Steps for Plastic Coaters--A Simplified Approach. Part II," Plating & Surface Finishing, 81, No. 3, 25-32 (1994).

(5) Hicks, P.G., Senser, D.W., Kwok, K.C., and Liu, B.Y.H., "Drop Transfer Efficiency in Air Paint Sprays," presented at the Engineering Society of Detroit Advanced Coatings Technology Conference, November 9-11, Dearborn, MI (1993).

(6) Muir, G.L., "Improving Liquid Spray Transfer Efficiency," Product Finishing, 59, No. 6, 62-66 (1995).

(7) Ewert, S.A., Felstein, S.R., and Martinez, T., "Low-Cost-Transfer-Efficient Paint Spray Equipment," Metal Finishing, 91, No. 8, 59-64 (1993).

(8) Hicks, P.G. and Senser, D.W., "Simulation of Paint Transfer in an Air Spray Process," J. Fluids Eng.--Transactions of the ASME, 117, No. 4, 713-719 (1995).

(9) Bunnell, M., "The Case for Turbine Sprays," Product Finishing, Vol. 53, No. 2, pp. 72-74 (1988).

(10) Zhou, Y., Lee, S.W., McDonell, V.G., Samuelsen, G.S., Kozarek, R.L., and Lavernia, E.J., "Characterization of Linear Spray Atomization and Deposition for Continuous Production of Aluminum Alloys," J. Mat. Synthesis and Processing, 5, No. 1, 111-116 (1997).

(11) Shaffer, P.D., "Realities of Paint Transfer Efficiency," Industrial Finishing, 63, No. 1, 32 (1987).

(12) Xing, L., Glass, J.E., and Fernando, R.H., "Parameters Influencing the Spray Behavior of Waterborne Coatings," JOURNAL OF COATINGS TECHNOLOGY, 71, No. 890, 37 (1999).

(13) Settles, G.S., "A Flow Visualization Study of Airless Spray Painting," Proc. of the 10th Annual Conference on Liquid Atomization and Spray Systems ILASS-Americas 97, Ottawa, Canada, 145-149 (May 1997).

(14) Snyder, H.E., "Drop Size Investigation of an Electrostatically-Assisted Fan-Spray Atomizer," M.S.M.E. Thesis, Purdue University (August 1988).

(15) Fox, R.W. and McDonald, A.T., Introduction to Fluid Mechanics, 5th ed. Wiley, 1998.

(16) McCarthy, J.E., "Basic Studies on Spray Coating Drop Rebound from a Small Workpiece with a Conventional Air Applicator," M.S.M.E. Thesis, Purdue University (December 1991).

Michael W. Plesniak, ([dagger]) Paul E. Sojka, and Anshul K. Singh--Purdue University*

*Maurice J. Zucrow Laboratories, School of Mechanical Engineering, West Lafayette, IN 47907-1003.

([dagger]) Author to whom correspondence should be sent. Voice: 765.494.1537; fax: [phone]; email: [e-mail].