| OCR Text |
Show Pyrotechnic devices were used for airborne seeding and tested extensively, Two main types were utilized. o ·n e was a droppable type containing a charg of 25g of Agl and burning for 20-40 sec. while falling. The calibration for this flare is given in Figure 4. The other pyrotechnic device that was used consists of an aircraft-attached flare which contains an Agl charge of 30g and burns for a nominal period of six minutes. The calibration curve for this device is pre sented in Figure 5. Both types of unit were the subject of tracer tests, which are reported on in the diffusion section (3. 2). A non-pressurized ground g e nerator system based on the Skyfire Airborne generator was developed for remote operation. The generator system was developed to successfully ignit e or extinguish upon the operation of a single e lectrical switch. The system burns no propane and utilizes a gravity feed system. ) A fi e ld technique for generator calibration was developed. The pro cedure is depicted sch e matically in Figure 6. The technique consists of placing a vertical wind tunnel over the generator burner, so that the entire output of the gene rator is mixed with the airflow through the tunnel. An air sample is taken out of the top of the tunnel with a syringe, diluted, and inje c ted into a 100 liter m y lar bag. The sample air input into an NCAR nucleus counter is taken from this mylar bag. The technique requires some skill and practice to obtain reproducible results. First of all, a nucl e i counter operated in the vicinity of a high output generator is subject to saturation by exceedingly high nuclei concentrations unless much care is taken. Secondly, very small samples 0cc) and very high dilutions (105) are involved, so that experimental error can easily be in the orders of magnitude range. The air in the mylar bag must be nuclei free and it was found this could besl b e accomplished by several evacuations and fillings \:Vith air filtered througn a glass fibre filter. One alteration from the initial t ests was the method used to take the sample. In the initial tests, a 100 cc syringe was used and a 100cc sample was taken from the top of the wind tunnel. Ninety cc of the sample was then expelled, 90 cc of filtered air drawn into the syringe , 90 cc of this mixture expelled, and the remaining 10 cc inj ected into the mylar bag - - - giving a net sample of 1 cc of wind tunnel air. More consistently reproducible results were found to be obtained by using smaller syringes and taking ½and 1 cc samples and injecting them directly into the mylar ba~ without a dilution step. The flow through the wind tunnel w as c hecked periodically with a precision anemometer. The fan blades wer e c hanged on one occasion and a slightly lower flow factor was used subsequently (Table 1). Th e results of a series of tests are listed in Table 2. These ge n e rator outputs are plott e d in Figure 7. The values span approximately an ord e r of l 0. |
