' I 1716
NASA Technical Paper 1716
. ' !
.Effect of Hole Geome.try and Discharge Machining (EDM) on ,Airflow. Rates; Through' Small-Diameter Holes. in Turbine Blade. Material
Steven A. 'Hippensteele a n d -Reeves P., Cochran .
Hippensteele and Reeves P. NY
NASA Technical Paper 1716
Effect of Hole Geometry and ElectricDischarge Machining (EDM) on Airflow Rates Through Small-Diameter Holes in Turbine Blade Material
Steven A. Ohio
National Aeronautics and Space Administration
Scientific and Technical Information Branch
TECH LIBRARY KAFB.II
Lewis Research Center Cleveland.
Fabrication experience has shown that the size (diameter) of the EDM electrode or the hole is not a sufficient criterion for establishing the required cooling-air flow rates through the holes because of difficulties in measuring hole diameters accurately.257 to 0. This report describes an experimental study at the NASA LewisResearch Center to determine under controlledconditionsthe effect oftwo design and two machine parameters on the air flow rates through small-diameter EDM holes. The machine parameters (electrode current and electrode currenton time) affect the nature of the EDM hole surface. of the hole. a logical specification and inspection criterionfor theseholeswould be air flow rates.462 mm. Test specimens were prepared from a typical turbine blade material (IN-100) of a typical thickness (1. The most commonly used method for putting cooling holes in the
superalloy materials used for turbine components is electric discharge machining (EDM). lnherent to the EDM process is the formation of a relatively rough surface where metal isremovedby deplating and the creation of a recast layer on this surface as the remaining metal resolidifies.Summary
An experimental investigation was conducted to determine the effect on air flow rate of two design parameters. and air-flow-rate deviations were determined functions as the of four control parameters. air flow-rates.The recastlayer is thesource of microcracks that can propagate into the base metal and can result in structural Therefore. inlet and exit shapes. Average air flow rates perhole increased linearlywith the square ofthe electrode diameter (cross-sectional area)and with electrode current-on time. electrode diameter and hole angle. and the controlled machine parameters were electrode current and electrode current-on time. Holes were electric discharge machined individually in rows of 14 holes each using solid brass electrodes. The results of the investigation showed a n exponential increase in average burn time with increasing hole length (a function of hole angle) and a linear decrease in overburn (difference between finished hole diameter and electrode diameter) with increasing electrode diameter. controlling the thickness ofthe recastlayer and the formation of microcracks within this layer are important considerations in the use of EDM for small-diameter cooling air holes. Hole characteristics such as surface roughness. failure. respectively. Surface roughness retards airflow and must be controlled or compensated for with a largerhole diameter. A cursory metallurgical examination was made of the recast layer and microcracks on the hole surfaces. A typical turbine blade material (IN-100)of a typical thickness (1.6 mm). However. Time to E D M individual holes (burn time).462 m m . and the microcracks did not extend beyond the limits of the recast layer on the examined specimens. holes angles from 40" to 75". respectively. etc.
Film-cooled gas turbine vanes and blades often contain a multiplicity of small-diameter (as small as 0.25 mm) holes for ejecting cooling air onto surfaces exposed to thehot gas path. no published information exists on the quantityand uniformity of coolant flow through EDM holes. The controlled design parameters were electrode diameter and hole angle. and hole angles (angle between surface normal and centerline of the hole) were varied from 40" to 7 5 " . The air flow rates were
. Since heat-transfer analyses of film-cooled components are based on hole cooling-air flow rates. length.6mm) was used. hole diameters. electrode current and electrode current-on time.257 to 0. The study was conducted over ranges of electrode diameters from 0. and electrode current-on times from 3 to 19 percent. The design parameters (electrode size and holeangle) obviously determine the nominal cross-sectional area and length. but changed little with changes in hole angle and electrode current.. and two machine parameters.electrode current from 20 to 40 A. increased with hole angle. The recast layer was less than 4 percent of the hole diameter. nor is there any known published information on practical tolerances for specifying deviations in flow rates through such holes.The average flow-rate deviation (from the mean flow rate for a given row of 14 holes) decreased with electrode diameter. The ranges of electrode current and electrode current-on time usedwere from 20 to 40 amperes and from 3 to 19 percent. may significantly affect the resulting hole flow rates. and changed very little with electrode current and current on-time. Electrode diameters were varied from 0.
deg Y isentropic exponent.5 cmwide. selected values ofother the parameters were used.
Test Specimens Test specimens were made vacuun-cast from IN-100 (nickel alloy)strips 1. and 40 A and current-ontimes of 3. K t burn time.3 kPa. and 19 percent were used. Identicalround holeswere electric dischargemachinedindividually into these strips to form specimensrows of 14 holes each. Airflow Measuring System The systemused to measure the cooling-air flow rates through the small-diameter EDM film-cooling holes is shown in figure 1 . 30. Only acursorymetallurgicalexamination ofthe recast layer and its indigenous microcracks was performed.462 mm and hole angles of 40°. The machine two parameters investigated were electrode current and electrode current-on time. mm hole humber in row. and 0. Eighteen strips. EDM Equipment The EDM machine used in the preparation of test specimens for this experimental investigation had a powersupplythat was adapted for the drillingof small-diameter holes. Only the 3 percent current on time was used with all variations of each of other the control parameters. were prepared. this system is similar to and is based on experience from the one described in reference 1 .00127 mm. 0. These values covered ranges of interest for typical the holes in film-cooled turbine blades and vanes.361. 0. kPa diameter. All machining was done with solid brass electrodes.5 percent. 1.and 75" were used. sec W flow rate per hole.40 7 percent time that the electrode current is turned on A flow-rate deviation of the holes in a row. Therangeof current levels covered slow to fast burn rates. 1 1 . Electrode diameter of 0. The then air passed throughrotameter a a and throttling valve to the test strip holder. The diameters of the electrodes used in this study were measured within 0. mm hole length.6 mm thick. In each specimen row the 14 holes making up the row were machined with all four control parameters held constant.8 kPa. Air exited the test strip through one the EDM holes in one of the of test specimen rows. The accuracy of the rotameter f1 was percent of the full-scale reading.
barometric pressure downstream of holes. 1 to 14 total pressure upstream of holes.0127-mm tolerance on the diameter. Pressurized air at about 860 kPa entered the measuring system throughacontrol valve and passed through and a filter a pair of pressure regulators which reduced the pressure to 206.3 K. For the other current-on times. all other holesin the test strip were sealed with vinyl tape a during given measurement.7 cm long.2 kPa above barometric pressure T absolute temperature. The two design parameters investigated were electrode diameter and hole angle (the anglebetween the normal surface and the hole centerline). 55. Currents of 20.7 kPa)
B D L n P
per strip and a total of 62 rows. Industry standards for these size electrodes allow *0. with a maximum of 5 rows
. The range of current-on permitted times an evaluation of the oil flushing effectiveness on burn rate surface and condition. Accuracy of the airflow measuring system was f1. 60". 4. 1 to 14 s standarizedcondition(temperatureof 289.measured for individual using holes roomtemperature air discharging to atmospheric pressure at aconstantpressure drop acrossthe holes. burn rates affect surface roughness and the development of the recast layer with its attendant rnicrocracks. percent Subscripts av mean average in a row of 14 alike holes e electrode h finished hole rn measured condition n hole number in row. 6. the accuracy of the pressure gages was f0.8 K and barometric pressure of 101.257.272. and 3.. and the accuracy of the temperature gages was *0. The results are presented as plots of average-hole flow rate and average-hole flow-rate deviation as functions of eachof the controlled parameters. mg/sec e hole angle measured between the surface normal and the hole centerline.
hole diameter was measured using a set of round gage pins graduated in 0.6 mm--
. internal pressure. burn time (in seconds) was measured for individual holeswith a stop Subsequently. Accurate hole diameters arevery difficult to measure because thesurfacesoftheEDM holes are very irregular and gage pins are calibrated in large increments compared with the diameters. Temperature. All holes in the test strip except one were sealed with vinyl tape beforetesting. Measurement of Air Flow Rates Individual measurements of air flow rates through each of the EDM holes were made. watch.0254-mm increments.the hole diameter will always be larger than the diameter of the electrode used to make the hole.2 kPa abo
"Pressure regulators Figure 1.
Measurement of Burn Time and Finished Hole Diameter During the machining of the holes.Hole angle--
Standardization of Measured Flow Rates The measured airflowratesthroughtheEDM holes were standardized to correct for the effect of air density variations. isentropic flow equations to adjust for differences in pressure ratio in reference 2:
. 1.pressure.and flow readings at the rotameter and temperature."-\
Test strip thickness. hole Because oftheinherent"overburn"inEDM. .Flow measuring apparatus with test strip holder.2 kPa between the pressure inside the holder and atmospheric (barometric) pressure. This value of constant pressure drop was chosen because experience has shown that the flow coefficients of the holes are less sensitive to pressure drop variations at this value (approximately half an atmosphere) than at other values. and barometricpressure readings at the test specimen holder were recorded for flow through each individual hole. '\
Test pressure (55..Then the strip was mounted in the specimen holder in the flow measuring system and
subjected to a constant pressure difference of 55. The following equation comes from the standard. one dimensional. The hole diameter was defined as the average of the diameter of the largest pin that would pass through a given hole and the smallestpin that would not pass through the hole.
hole length was directly related to hole angle. and themachine two parameters.07 to 6.
n= 1 4
Average Burn Time and Average Finished Hole Diameter The average burn time (timein seconds to EDM an individual hole) for the experimental specimen rows are tabulated on table I.
. average flow rate deviation.0127 mm had been allowed. EDM holes in 1. Because the test strip thickness in this investigation was constant. The analysis of variance for these data showed thatthesignificantinfluencing parameter was holelength. standard a barometricpressureof 101. Where variations due to the control parameters distinguishable. This averaging was done using the following relationship. analytical This technique shows whether variations in testresultsthat aredueto planned changes in the control parameters can be distinguished from variations that are due to system errors (ref. fit This comparison is based on electrode diameters that were measured within 0.8 K.n=l -
Deviations of Standardized Flow Rates The variation offlow rate in the individual holes in each row of 14 holes was evaluated by determining an average deviation of the standardized flow rate for each row using the following relationship: The average burn time increased exponentially from 37 to 154 secas the holelength increased from 2. 4). The maximum variation from the curve fit was ~ t 2 4percent the over full range of control parameters. the hole diameters recorded for the two smallest electrodes are identical. Because the “go.9 kPa. air flow rates.13 mm. 3) were made of the hole flow rate.the variation would have been much greater.no-go” pin measurementshole for diameter were very approximate. A least square curve fit of these data is
C wuu . If industry standards of &0.4 percent.) Using a standard temperature of 289.7 kPa. Each plotted represents point several measurements of measuredholediameter.
The maximum variation of the measured finished hole diameter from the curve was only 4. and a standard upstream pressure of 156.00127 mm. are the data will be presented as functions of these parameters.
Averaging of Standardized Flow Rates The standardized flow rates for the individual holes in each row of 14 holes were averaged to obtain a single value of flow rate to be used in correlations with the controlled parameters.andflow-ratedeviations for small-diameter. The relationship between electrodediameterand finishedholediameter is shown in figure 3. ref. A least squares curve fit of the average finished hole diameter for each of the four electrode diameters is
Statistical Analyses Statistical analyses (analysis of variance of factorial experiments.(All symbols are defined in Symbols. Figure is a 2 plot of average the burn times for all 3-percent electrode current-on time data against hole length. electrodecurrent and electrodecurrent-ontime. finished holes diameters. electrode diameter and hole angle. this correction factor becomes:
Results and Discussion
The burn times. A brief discussion the of recast layer and the microcracks that resulted from the EDM process is also given. and burn time data.
All measured flow rates were multiplied by this correctionfactortoarriveatacommon basis for comparison.6-mm-thick IN-100 strips using solid brass electrodes are presented herein as functions of the two design parameters.
31'1 60.839 42.15 60.74 5.57 15.88 2.93 35.283 25.02 2.92 2.910 25.14 41.92 135.00121 mm b * ~ .871 59.819 41.Dl 3.19 5.29 61.06 1.01 39.53 1.29 3.122 51.81 5.340 61.54 3.366 60.81 2.36 42.663 0.36 64.62 123.48 4.00 53.31 6.020 62.20 1.329 42.26 1.43 155.032 41.R3 1.36 46.52 1.356 28.316 28.17 43.
Kachine parametel Ilectrode current.29 47.38
" " "
46 41 48 49 50 51 52 53
30 20 30
" " "
" " "
" " "
" " "
" " "
" " "
" " "
2 3 4
6 I 8 20 25 9 10 11 12 13 21 22 23 26 21 28 29 14 15 16 24 30 17 18 19
29.86 67.024 45.90 .786 28.602 41.546 42.51 28.43 41.41 3.57 4.83 .21 146.57 66.200 45.67 56.095 45.86 29.21 €0.529 42.86 51. deg
40 40 40
deviatlon.819 41.01 53.90 1.83 4.13 1.01 33.212
20 30 40 20 30 30 30 40 20
48.01 180.025 60.51 31.17 4.047 29. percent
31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 _____
" " " " " "
10 19 3 3 3 6 11 3 3 3
53.823 44.92 0.56 2.72 1.53 1.23 28. mm
54 55 56 51 58 59 60 61 62
0.43 7.72 2.136 40.233 14.83 44.29 54.464 26.19 36.31 1.14 51.105 40.936 57.928 64.80 158.21 52.36 58.356 27.64 56.159 44.93 2.712 24.64 47.85 .86 1.09 0.EXPERIMENTAL DATA
burn t h e .92 2.16 6.131 27.438 46.915 41.12 2.680 42.76 2.41 2.71 48.41 2.
Hole mgle.11 63.114 61.542 42.055 29.TABLE I.93 125.50 2.53 3.141 62.462
15 15 15
30 40 20 30 30 40 20
"+o.44 3.131 62.61 1.54 .29 96.069 41.99 .93 3.36 63.533 60.29 3.51 3.57 44.93 49.41 1.313 26.16 .109 61.w .516 45.581 40.411 60.84 .35 2.21 54. m ml ~ ~ ~ .21 2.830 41.48 6.
.283 26.198 26.58 4.R60 41.901 29.234 41.701 42.531 24. sec
Design parameters Electrode dlnmetera.123 38.29 31.035 40.
mm Figure 3
? Md I
.Finished hole diameter
as function of hole diameter.
.Average burn time as function of hole length.Current.30
Electrode diameter. A
if thecorresponding finished hole diameters from figure 3 are substituted for the electrode diameters. a 2. it also showed that there was a verysmall effect of the interaction of electrode diameter and electrode current.462 mm.20
-Least-squares-cu ve fit: 1 Wav 232 74 De + 11. The electrode current had little effecton the average flow rate per hole. a e 3. the average increased from 26. 8 mg/sec or +12. The average flow rate per hole as a function of electrodecurrent is shown in figure6forthe full ranges of hole angles and electrode diametersand for an electrode current-on of time 3 percent.I5 Electrode diameter squared.05
.74 D2 e
e.theflow rate remained within &7.25
A 4 0
'0 30 O H ) 0 3 0
. As expected.
. D .However.It also can be seen from the figure that the overburn(defined as Dh -De).06
.3 percent for the full range of hole angles when thesameelectrodediameter and electrodecurrent were used.4 to 60. as would be expected.:
. The lines in the figure connect the mean values of points having the same electrode diameter and electrode current at each hole angle.5 percent for the full range of electrode currents when thesame hole angle and electrodediameter were used. which remained within +6. The average flow rate perhole as a function of hole angle is shown in figure 5 for the full ranges of electrode diametersand electrode currents and for an electrode current-on time of 3 percent.09 to 0. The maximum variation from the curve fit of figure4 was ~ 5 .25.
Average Hole Flow Rate The average flow rate per hole as a function of electrode diameter squared is shown in figure 4 for the full ranges of hole angle and electrode current and for an electrode current-on time of 3 percent.7 mg/sec. then the area ratio (or ratio of finished hole diameters squared) is 2.The analysis ofvarianceconfirmedthatthe electrode current had little effect.3-fold increase.3 percent of the flow rate. = 232.10
flow rate per hole Based on this curve fit.06mm (or from 35 percent to 13 percent) as the elettrode diameterincreasedfrom 0. The average flow rate perhole as a function of electrode current-on timeis shown in figure 7 for the full rangeofelectrodecurrents andfor electrode
Angle.23-fold increase. decreasedlinearly from 0. as D2 increased from 0.Flow rate per hole as function
of electrode diameter squared. It can be seen from the figure that hole angle (or hole length) has very little effect on average flow rate per hole. A least-squares-curve fit of the averageof the flow rates for each value of electrode diameter squared is
W.066 to 0.. the analysis of
variance confirmed that the flow rate per hole was a very strong function of electrode diameter. This ratio agrees closely with the ratio of the flow rates. This trend can be seen in figure 6 where the curves are not all parallel.257 to 0.
Lines are mean per hole (same electrode diameter and angle).257 20
40 50 Holeangle.70
\3 30 30 0 0 30 2 1 4 0 A 40
P L .Flow rate per hole as function of electrode current.361
deg 40 60 75 40 60 75 40 60 75
. deg 70
75 40 60 75
'. Lines are mean of average f l o w rate per hole (same electrode diameter and current). . . e. A
of average flow rate
20 10 20
Electrode current. .0.272-.462
20 20 20
\3 30 0 3 0
0 3 0
A 4 0
. mrn .
Electrode diameter.Flow rate per hole as function of hole angle (measured between the surface normal and the hole centerline).
8 12 Electrode current-on time.Average flow rate per hole as function
of percent electrode current-on time.
e. Lines are least-squares-curves fits of mean of average flow rate deviation at same hole angles. .0.35 Electrodediameter. A
." 1 16 20
./ Electrode mm
0 3 0 60 A 4 0 60 Mean of average flow rate per hole (same electrode diameter. mm
0 .40 De. .
75 40 60
75 40 60 15
. 4 7 5 ~+ 58.Average flow rate deviation (from mean among all holes in a row) as function of electrode diameter.50 . Least-squares-curve fit for the mean of average flow rate per hole (same electrode diameter). T percent . A
Figure 9.(a) Magnification.Microphotographs of cross section of typical EDM hole in IN-100 t-naterial. . X250.
aholeangleof 40". For the larger electrodediameter. slightly by hole angle (or hole length). .rows of holes with like diameters and angles would possibly be machinedsimultaneously.
Aav = 10.-EDM holes in IN-100 specimens were subjected to metallographic inspection to determine the general nature factors. Therefore. and little by electrode current and current-on time. that the average flow-rate deviation affected was significantly by electrode diameter.
Aav =6. based on these fits.14De
Summary of Results
Theresults ofanexperimental investigation to It can be seen fromthe figure thattheaverageflow-determinethe effects oftwo design parameters rate deviationdecreasedlinearly with electrode(electrodediameter and holeangle) andtwomachine diameter and increased with hole angle. data the were plotted as a function of the electrode diameter.of. and hole angles used for these samples.3 percent for an increase in currenton time from 3 to 1 1 percent..The solid lines in the figure connect the mean values ofpointshaving thesame electrode diameter.electrode current. the recast layer did not exceed 0. for 75" hole angle. and the average burn times presented are on an individuai hole basis. for 60" holeangle. but the apparent exaggerated taper of the hole in figure 9(a) is duetothe fact thatthe specimen was notcut exactly on the centerline of the hole. of small-diameter that machined holeswere
.12.13-20.therecorded average burn times per hole may not be meaningful for production estimates.individualhole electric discharge machining a feasible way to is produce test models. Therefore. andthemicrocracksdidnot extend beyond the limits of the recast layer. This relationship is shown in figure 8 where all points are at an electrode current-on time of 3 percent.0127 mm had been allowed. In this investigation the electrode diameters were measured within 0. some samples.35 . and electrode current at each electrode current-on time.462 mm and holeangles of 40" and 60". Average Hole Flow Rate Deviations The variation in flow rates through the 14 holes in each row (average flow-rate deviations the from mean flow rate for the row) were determined by using equation ( ) The analysis of variance showed 4.diameters of 0. it can be concluded that for the range of parameters tested flow rates are reproducible within acceptabletolerances andthat theeffectsof the controlparameters used (electrode diameter. However. For the full parameters (electrode current and electrode currentrangeofparameterstested. If industry standards of &0. the average flow rate per hole.361 mm and
for an electrodediameterof 0. However. andan electrode current of 20 A. from amanpower/machinetimestandpoint.and electrodecurrent-on time) can be evaluated accurately orare negligible. the average flow rate per hole increased 6.Fortheelectrode currents.For the smaller electrodeand a hole angle of 60°.2 percent for an increase in electrode current-on time from 3 to 19 percent. Least square curve fits of these mean values are
Recast Layer and Microcracks The scope of this experimental investigation did not include an in-depth study of the formation of a recast layer and microcracks that areinherent in the EDM process. curve increased 14. the variation in flow rate would have been much greater.016 mm (4percent of theholediameter).462 mm. these times are applicable to model making and do show that. electrode current-on times. In a production effort.33De
On the basis of the results from this experimental investigation of air flow rates through small diameter EDM holes.themaximumaverageontime)ontheairflowand physical characteristics flow rate deviation was less than 7 6 percent. Least square curve fits of the data are for 40"hole angle. electrode diameters. A l l of the holes in this experimental investigation were electricdischargemachinedindividually. hole angle.)
for an electrode diameter of 0.00127 mm. hole angle. shows of these Figure 9 microphotographsof a sample.361 and 0. (Note: Some slight taper in an EDM hole is normal.
The electrode overburn (difference between finished diameter hole and electrode diameter) decreasedlinearlyfrom 0. 505-04. Convection. RonaldPress Company. 1980. 1974. Hippensteele. 3 . 1964.
. to be affected little ( k 6 .06 mm with an increase in electrode diameter from 0. April 14.3 for an 80 percent increase in diameter) b. not to be affected significantly by electrode current and electrode current-on time d . 1 . Ohio. Vol. Owen L.3 times increase in electrode current-on time. Hicks. The average flow rate deviation (from the mean flow rate for a given row of 14 holes) was found a.13 mm. 1956. Rinehartand Winston.257 to 0. 3. to be affected little (&7. im
2.462 mm. ed. The ranges of the parameters tested were electrode diameter from 0.: Pressure-Loss and Flow Coefficients
Inside a Chordwise-Finned.. 2. Impingement.: The Design and Analysis of Industrial Experiments.8 b.: Fundamental Concepts in Design the of Experiments Holt. 1953. hole angle from 40" to 75". electrode current from 20 to 40 A. 4. to decrease linearly by a factor of about 3 as the electrode diameter increased by a factor of 1. NASA TM X-3028. and electrode current-on time from 3 to 19 percent. Cleveland. Over these ranges of the control parameters 1 . Charles R.
Lewis Research Center. to increase linearly by 14.09 to 0. to increase linearly with the square of the electrode diameter (by a factor of 2.462 mm .l
individually into superalloy material by electrical discharge machining @DM) are summarized below.5 percent) by doubling the electrode current d.2 percent for a 6. A cursory metallographic examination showed that the thickness of the recast layer was less than 4 percent of hole diameter and that the microcracks present did not extend beyond the limits the recast of layer. 5 . Hafner Publishing Company. Shapiro. 3 percent) by hole angle (related to hole length) c.257 to 0. to increase as the hole angle increased c. to have maximum deviations of less than 7. Archer H.
1 . The average burn time per increased hole exponentially (from 37 to 154 sec) with an increase in hole length from 2.: The Dynamics and Thermodynamics of Compressible Fluid Flow.07 to 6.6 percent for all tested values of the control parameters. Steven A.
4. Davies. The average flow rate per hole was found a. and
F l Air-Cooled Turbine Vane. National Aeronautics and Space Administration.
Work Unit No. The holes were machined individually in rows of 14 each through 1. Sponsoring AgencyNameandAddress
11. and two machine parameters. Supplementary Notes
* F o r s a l e by the National Technical Information Service. of Pages
The effects of two design parameters. Title and Subtitle
NASA TP-1716 EFFECT OF HOLE GEOMETRY AND ELECTRIC-DISCHARGE MACHWING (EDM) ON AIRFLOW RATES THROUGH SMALL-DIAMETERHOLES IN TURBINE BLADE MATERIAL
13.6-mm-thick IN.462 mm) electric-discharge-machined (EDM) holes were measured. Price'
NASA-Langley. Performing Organization Code
8. The data showed linear increase in air flow rate with increases in electrode cross-sectional area and current-on time and little change with changes in hole angle and electrode current. Springfield.'
Steven A. Cochran
9. Contract or Grant No. Turbine cooling.
2 Government Accession No. 20546 C. .
3.257 to 0. Security Classif. Hippensteele and Reeves P. No. Fabrication. Film cooling. Performing Organization Name and Address
10. Recipient's Catalog No. Type of Report and Period Covered
National Aeronautics and Space Administration Washington. Virginia
12. The average flow-rate deviation (from the mean flow rate for a given row) decreased linearly with electrode diameter and increased with hole angle. Report No. (of this report)
Unclassified unlimited STAR Category 07
4. on air flow rates through small-diameter (0. Security Classif. Key Words (Suggested
18. D. electrode diameter and hole angle. (of this page)
21. Report Date November 1980 6. Sponsoring AgencyCode
16. Electric discharge machining (EDM) 19. Burn time and finished hole diameter were also measured.100 strips. electrode current and current-on time. Performing Organization Report No.
National Aeronautics and Space Administration Lewis Research Center Cleveland. Distribution Statement
Gas turbine engines.