Conclusions - 2012

Summary

Droplet Density

Due to the high volume of spray material and the dynamics of the spray plume in an airblast sprayer, the results for droplet density should not be utilized for either the NESareScan or DepositScan programs. Overlapping droplets and droplet shapes modified by the high velocity air make it difficult to determine the droplet density.

Droplet Size

Droplet size was difficult to measure because of droplet overlap and various droplet shapes created by the blast of high velocity air. In determining droplet size, the NESareScan program calculates the volume median diameter which is written as VMD or Dv0.5. Once the percent coverage reaches 20%, the droplet size becomes very difficult to measure. The measurement is the most accurate at a very low percent coverage. In calculating droplet size only cards with less than 5% coverage should be used. The droplet size increases as the percentage coverage increases with both the NESareScan and DepositScan programs.

The blue spray pattern indicator gave statistically smaller droplet sizes on the water sensitive paper in both the NESareScan and DepositScan programs. No statistical differences were found with the red food coloring in either computer program. The NESareScan program gave smaller droplet sizes than DepositScan but only once was the result statistically significant.

Percent Coverage

Percent coverage proved to be the most accurate of the three parameters tested and it can be used to adjust the spray pattern. Differences were found based on the paper, dye or program utilized. With the blue spray pattern indicator, the percent coverage means were higher and statistically significant with the three photographic papers as compared to the the water sensitive paper at all heights between 2.5 and 6.5 feet when using the NESareScan program. Almost no statistical significance was found with the same results with the blue spray pattern indicator and the DepositScan program.

When using the red food coloring very few significant differences were found between the various papers with the NESareScan program. Only four significant differences were found out of 210 possible combinations. With the DepositScan program and the red food coloring, a total of 12 statistically significant differences were found, but no clear trends emerged.

Results from the first and second tests were used to compare the blue spray pattern indicator with the red food coloring. The blue dye had a significantly greater percent coverage than the red food coloring with the gloss, semi-gloss and matte papers at nearly all the heights with both the NESareScan and Deposit Scan programs. No significant differences were found for the water sensitive paper. The use of red food coloring will give a more accurate determination of the percent coverage.

The results from each test can be compared for any significant difference in the percent coverage between the NESareScan and DepositScan programs using a two-tailed t-test. The only trend that appeared was the NESareScan program tended to give higher results for the gloss and semi-gloss papers at a high percent coverage.

Out of 168 statistical comparisons run between the NESareScan and DepositScan programs for percent coverage, 26 showed a statistical differences or 15.5% of the total. Of the 26 statistically significant differences, 19 times the NESareScan had a higher result and 7 times the DepositScan had a higher result.

Conclusions - 2010 versus 2012

In the 2010 project it could not be determined which design actually gave a more accurate representation of the true spray pattern generated by an airblast sprayer. The current project attempted to answer that question and in the process develop an even lower cost patternator utilizing photographic printer paper and dye. Image analysis software was developed so the results could be expressed quantitatively.

The SARE and modified Cornell patternator from 2010 were retested in 2012. The modified Cornell patternator captured 24% of the spray material while the SARE patternator captured 53%. The SARE patternator captured significantly (1% level of significance) more spray material on 5 of the 7 panels. These results mirrored those obtained in 2010.

Every design of patternator has its limitations. It can be difficult to get accurate measurements and the measurements can vary for different reasons depending on the patternator design. In 2010, a varying loss of spray material depended on the design of the patternator. The patternators in the 2010 project collected spray for one minute, which helped eliminate any variability due to sprayer output, but spray material was lost. This loss reduces the accuracy.

While the paper patternators used in the 2012 project record all of the spray that hits the card, the spray is collected for a very short period of time. The amount of spray recorded depends on the speed of the tractor and the width of spray plume. The width of the spray plume depends on the nozzle, nozzle setting and pressure. Our tractor speed in 2012 and also in 2010 was 2.68 miles per hour and the width of the spray plume was approximately 30 inches wide at 4.5 feet from the center line of the sprayer. With these parameters a card at 3.5 feet above the ground is exposed to spray for only 0.64 seconds. At 3 miles per hour the time would be 0.57 seconds and at 4 mph the time would be 0.43 seconds. The extremely short sample time can increase the variability. Multiple tests should be run to reduce the inherent variability from a short collection time. A high degree of variance can be found between some of the replicates.

With the card patternator there is also a loss of accuracy due to the overlap of droplets. As the amount of spray increases more and more spray hits an area that was already covered. After about 20% coverage the cards become less accurate because of spray overlap. This inaccuracy increases as coverage increases. The size of a droplet can easily double if only a small amount of the droplet overlaps another droplet. Cards are very accurate in determining the actual percent coverage, which correlates to the coverage on plants, although a mature grape vine or any plant is a much more complex system than a flat card. Some areas of the plant receive an amount of spray analogous to that found on a card while other areas of the plant receive almost no spray because of overlapping leaves and fruit.

One outstanding advantage of a paper card patternator is its ability to be used with any type of sprayer. It will work equally well with backpack sprayers, horizontal boom sprayers or airblast sprayers. The total cost of a vertical paper patternator that can be used with an airblast sprayer is just $45. For a boom sprayer, the cost of the patternator would be only $34.

One of the original intentions of this project was to determine which patternator developed in 2010 gave the best representation of the spray pattern from an airblast sprayer. It was thought the paper card patternator would reveal whether the SARE or modified Cornell patternator gave results closest to the actual spray pattern. Once the paper card patternator was tested it was readily apparent the problem of spray overlap and extremely short collection time would make the determination of the best patternator design difficult.

Based on the results from this project the SARE patternator gives a more accurate spray pattern determination, but this cannot be determined positively because of spray overlap on the card patternator. The SARE patternator more closely mimics the actual results in a vineyard or orchard versus the modified Cornell patternator. The use of any patternator reduces costs and increases productivity from reduced pesticide use and better application.