NASA Technical Memorandum 103985

A Critical Assessment of UH-60Main Rotor Blade Airfoil Data

Joseph Totah

September 1993

(NASA-TM-103985) A CRITICALASSESSMENT OF UH-60 MAIN ROTORBLADE AIRFOIL DATA (NASA. AmesResearch Center) 24 p

N94-32063

Unclas

G3/02 0005471

NASANational Aeronautics andSpace Administration

https://ntrs.nasa.gov/search.jsp?R=19940027557 2018-07-03T07:40:24+00:00Z

NASA Technical Memorandum 103985

A Critical Assessment of UH-60Main Rotor Blade Airfoil DataJoseph Totah, Ames Research Center, Moffett Field, California

September 1993

NASANational Aeronautics andSpace Administration

Ames Research CenterMoffett Field, California 94035-1000

Summary

Many current comprehensive rotorcraft analyses employlifting-line methods that require main rotor blade airfoildata, typically obtained from wind tunnel tests. In order toeffectively evaluate these lifting-line methods, it is of theutmost importance to ensure that the airfoil section dataarc free of inaccuracies. A critical assessment of theSCI095 and SC1094R8 airfoil data used on the UH-60main rotor blade was performed for that reason. Ninesources of wind tunnel data were examined, all of whichcontain SCI 095 data and four of which also containSC1094R8 data. Findings indicate that the most accuratedata were generated in 1982 at the 11-Foot Wind TunnelFacility at NASA Ames Research Center and in 1985 atthe 6-inch-by-22-inch transonic wind tunnel facility atOhio State University. It has not been determined if datafrom these two sources are sufficiently accurate for theiruse in comprehensive rotorcraft analytical models of theUH-60. It is recommended that new airfoil tables becreated for both airfoils using the existing data. Addi-tional wind tunnel experimentation is also recommendedto provide high quality data for correlation with these newairfoil tables.

Symbols

b

c

Cd

Cd0

Cf

Q

C/max

C/a

"max

'"min

"recovered

correlation parameter for QJ at lowsupersonic speeds

wind tunnel model span, ft

wind tunnel model chord, ft

section drag coefficient

section drag coefficient at zero lift

mean skin friction coefficient

section lif t coefficient

maximum section lif t coefficient

dC//da = average section l if t curveslope at zero lift

maximum value of C/a at M < 1.0

min imum value of C/ at M < 1.0

recovered value of C/a at 1.0 > M > 1.1

form drag/friction drag

wind tunnel test section height, ft

K

Lp/c

(lVD)max

L

M

MI

MOD

Perror

P/

Pu

t/c

SA

va

correlation parameter for airfoil sectiondrag due to lift

airfoil perimeter/chord

maximum lift-to-drag ratio

lift, Ib

Mach number

Mach number at which the slope of C/a

changes from "+" to "-"

drag divergence Mach number

measured pressure transducer steadybias error, psia

lower surface pressure, psia

upper surface pressure, psia

thickncss-to-chord ratio

mean value of airfoil pressurecoefficient

free-stream Reynolds number

Reynolds number in drag correlation

free-stream velocity, ft/s

angle of attack, dcg or rad

zero-lift angle of attack, dcg or rad

1/P

C/Q gradient at speeds greater than M]

Cd0 gradient at speeds greater than

MOD

ratio of specific heats, 1 .4 for air

density, slugs/ft3

Y

P

Introduction

NASA, in cooperation with the U.S. Army Aviation andTroop Command (ATCOM), is engaged in a program toprovide and validate the technology and methodologyrequired to improve the performance, dynamics,acoustics, handling qualities, and cost of civil and mili taryrotorcraft. A major element of this program, the UH-60Phase II Airloads Program, consists of ground based andflight research of the UH-60 Blackhawk helicopter with a

pressure instrumented blade and a full suite of other

instrumentation.

NASA and ATCOM arc currently preparing for rigorous

analysis methodology validation using high quality data

generated from the UH-60 Phase II Airloads Program.Analysis methodology validation involves assessing andimproving state-of-the-art comprehensive analyticalmodels through exhaustive correlative studies in

performance, dynamics, and rotor structural loads andairloads. In order to assess and improve the theories and

assumptions employed in comprehensive analyticalmodels, accurate vehicle representations must be

established.

The main rotor blade airfoil section characteristics arcamong the most important parts of the vehicle repre-sentation. The airfoil sections on the UH-60 Blackhawk

helicopter are the SCI095 and SC1094R8 utilized on themain rotor blade shown in figure 1. The profiles of theseairfoils are shown in figure 2.

Sikorsky Aircraft, a Division of United TechnologiesCorporation, was tasked to provide NASA with allknown steady, 2-D wind tunnel data on the SC1095and SC1094R8 airfoils. Nine data sets (refs. 1-9) wereidentified and provided to NASA, all of which containedSCI 095 data and four of which contained SC1094R8data. This report documents an assessment of that data forboth airfoils.

An effort similar to the UH-60 Phase II Airloads Programwas performed on an H-34 helicopter by Scheiman in theearly 1960s. That experiment has long been a standard for

rotor airloads data, but it did not include high speeds.Furthermore, rotor systems have evolved dramaticallyfrom the early 1960s. The UH-60 Phase II AirloadsProgram will consider high speeds and will gather data at

much higher sample rates. The U.S. rotorcraft industry

has played a key role in defining the requirements for this

program to ensure it meets their needs. Also, a formalrecommendation resulting from a peer review of the

program in 1990 was a primary motivator for the workpresented in this report.

I would like to acknowledge and thank Mr. RobertFlemming from Sikorsky for his thorough review of the

data and for his comments, all of which have been

incorporated in this report.

Description of Data

Nine sets of UH-60 airfoil data have been considered.The sources of these data sets which contain SCI 095 and

SC1094R8 airfoil data are listed in table 1. These datasets are identified in table 1 and throughout this report,as Experiment 1, Experiment 2, and so on, throughExperiment 9. Pertinent information about the experi-

ments, the wind tunnel facilities, airfoils, and measure-ment devices are also noted in this table. Some details of

these experiments are discussed in this section.

The primary objectives of three of the experiments wereto assess current technology airfoils, either stand-alone or

compared with prototypes. Experiments 3 and 8 gathered

steady, 2-D data on the SC1095 and SC1094R8 airfoilsand compared them to prototype airfoils. Experiment 7gathered SCI095 data for correlation with a computa-tional fluid dynamics code.

Evaluation of Experiment 7 data revealed grossdiscrepancies relative to the data from all the otherexperiments. The published report documenting thisexperiment noted that inaccurate tunnel wall correctionswere applied to generate the reported data (ref. 7).Regrettably, appropriate wall corrections are not availableand the tunnel configuration has since been permanently

modified.

Some experiments examined alternate methods of testing.For example, the primary objective of Experiment 2 was

the testing of a Tunnel Spanning Wing Apparatus (TSWor TSA) which fit inside a wind tunnel test section. The

TSW was evaluated in Experiments 2 and 5, and laterused in Experiment 8. Experiment 2 attributed prc-stall"bumps" in lift coefficient at high angles of attack tomodel flexibility. Experiment 5 gathered data with and

without a center span device that alleviated the modelflexibility problems noted in Experiment 2. Experiment 5

published two sets of SC1095 wake drag data, identified

as 5a and 5b. The 5a drag data accounted for the differ-ence in static pressures on each side of the wake behind

the airfoil, whereas the 5b drag data did not.

The remaining four experiments were primarily con-cerned with the study of trends. Experiment 1 consideredthe influence of various surface irregularities relative to abaseline SC1095 airfoil. Experiment 4 studied icing

conditions relative to baseline SCI095 and SC1094R8airfoil characteristics. This experiment generated rela-

tively small amounts of data under normal, non-icingconditions. Data published from two alternate lift

measurement approaches devised in Experiment 4 werealso evaluated. Experiment 6 studied the effect ofReynolds number on both the SCI 095 and the SC1094R8airfoils. This experiment documented known problems indetermining Qmax, and the airfoils used in that experi-ment were tabbed. The tabs were deflected upwardapproximately 3 degrees. The tabs also changed thethickncss-to-chord ratios to 0.091 and 0.09 for theSCI 095 and SC1094R8 airfoils, respectively. Untabbcdthickncss-to-chord ratios arc 0.095 and 0.094 for theSCI 095 and SC1094R8, respectively. Finally, Experi-ment 9 measured the effects of dynamic stall relative tobaseline SC1095 steady, 2-D characteristics. Data fromExperiment 9 were limited to speeds less than M = 0.3.

In summary, although all of the data from these nineexperiments were examined, Experiment 7 and someExperiment 4 results were not published in this report.Experiment 7 results were omitted because of theaforementioned problem with the tunnel wall corrections.Experiment 4 data gathered using the two alternate liftmeasurement approaches were also omitted because noattempt was made to address known anomalies noted atcertain test conditions. In each instance the experimenterwas consulted prior to omitting the results, andconcurrence was obtained.

Evaluation Methodology

The methodology developed by McCroskey (ref. 10) andfirst applied to NACA 0012 data was used to evaluate theSC1095 and SC1094R8 data. This methodology usesspecific criteria to separate accurate data from inaccuratedata. All the data arc then placed into one of four groupsthat further reflect varying levels of accuracy. A shortsummary of the aforementioned criteria, and thedefinitions of the four groups are given in this section.

Criteria

Generally speaking, for M < 0.6 and between106 < Rc < 107, accurate data is distinguished frominaccurate data if they exhibit the followingcharacteristics:

1. 0.10 per degree < PC/U < 2n per radian, where

b = Vl - M , 0.10 is a known boundary, and 2n is thetheoretical lift-curve slope.

2. pC/a and PC(J0 are independent of Much number.

3. PQQ and PC(J0 are slightly dependent on Reynoldsnumber.

Groups

Four groups were defined by McCroskey to distinguishvarying levels of accuracy. A graphical approach is usedto place the data into each of these groups. This approachbegins with two plots; PQa versus Re and Cd0vcrsus Re

for data less than M = 0.6 and between 106 < Rc < 107.Group 1 quality data should have values for both PQand Cd0within ±0.0005 and ±0.0002, respectively, of alog curve fit approximation of only the accurate dataidentified by the aforementioned criteria. Group 1 qualitydata are of sufficient accuracy for use in comprehensiveanalytical input models. This is further examined in theDiscussion section.

Group 2 quality data should have values for both PQand Cd0 within ±0.004 and ±0.001 , respectively, of thelog curve fit of only the accurate data identified by theaforementioned criteria. It has not been determinedwhether Group 2 data arc sufficiently accurate for use incomprehensive analytical models. This will also beexamined in the discussion section.

Group 3 quality data should have values for either PQa

or C(jowithin the Group 2 tolerances. Finally, Group 4quality data have values for both PQa and Cc)0 outsidethe Group 2 tolerances.

Once the groups have been established, C/ Qj , and(L/D)max are examined throughout the ful l r"ange of Machnumbers. The trends that these parameters exhibit as afunction of Mach number are characterized by theirinflection points, or the points at which the trendsabruptly change direction. The inflection points ofinterest arc:

C/a: C/amax' C'amin' and C/arccovcrcd

2-

3. (L/D)max: maximum value of (L/D)max

4. C/ : maximum value of Qmax max

The accuracy with which these inflection points can beestimated, in addition to the continuous and unscattcrcdbehavior of the data between the inflection points, areindications of data consistency.

It is important to realize that the groups are defined at lowspeeds for a given range of Reynolds number. This docs

not ensure that the data in any given group will retain thesame accuracy at higher speeds. It is therefore importantto plot all groups throughout the full range of Machnumbers and check the consistency of the data bothwithin the individual groups and among the groupsthemselves.

Results

The methodology described in the previous section wasapplied to the data from the experiments for both airfoils.The results of the evaluation of the data are presented inthis section. Table 2 lists pertinent information about thewind tunnel facilities used in all nine experiments, tunnelwall corrections, and known accuracies of experimentsthat generated the NACA 0012 data previously evaluatedby McCroskey.

SC1095 Airfoil

Evaluation of PQa- Figure 3 shows derived PQa

values from the experiments plotted versus log(Rc).Figure 4 shows a log curve fit of the data only between0.10 < pC/a < 2n, along with Group 1 and Group 2tolerances. Balance data from Experiment 2 and pressuredata from Experiments 3 and 6 values are wi thin theGroup 2 tolerance; however, none of the experiments arcconsistently within the Group 1 tolerance. The implica-tion is that Experiments 2 (balance), 3, and 6 producedGroup 2 quality lift coefficient data because the derivedPQa values are within the Group 2 tolerance.

Evaluation of Cd0— Figure 5 shows Q]0 values from theexperiments plotted versus log(Re), along with a logcurve fit of that data, and Group 1 and Group 2tolerances. Wake drag data from Experiments 1, 4, 5a,6, and 8 appear to be within or very near the Group 1tolerances. Wake drag data from Experiments 2, 3, 5b,and 9 are all within the Group 2 tolerances. Theimplication is that all of the experiments produced Group2 quality drag coefficient data because the C{j0 values arcwithin the Group 2 tolerance.

Groupings- Based entirely on the above evaluations ofPQQ and Cd0 as presented in figures 3 through 5 thegroupings for the SC1095 data are:

Group 1 None

Group 2 Experiments 2 (balance and wake drag), 3,and 6

Group 3 Experiment 1, 2 (pressure), 4, 5, 8, and 9

Group 4 None

Results for C/Q- The variation of C/Q throughout the fullrange of Mach numbers is shown in figure 6, with eachgroup duly noted. An examination of the Group 2 datareveals that there is a smooth and consistent trend in thevariation of C/Q with Mach number up to M = 0.84. Thistrend is noticeably different than that exhibited by theGroup 3 data, and less scattered than the Group 3 data aswell. A maximum value of C/a occurs at M = 0.84 and aminimum value occurs at M = 0.90, with a small recoveryat speeds greater than M = 0.95. Maximum, minimum,and recovered values of C/a can be roughly estimatedfrom the data shown in figure 6. The McCroskcy-Smithexpression superimposed on figure 6 will be discussed inthe next section.

In summary, no lift coefficient data exist beyondM = 1.10, the best lift coefficient data available are foundto be Group 2 quality, and that the data are only consis-tent at speeds up to M = 0.84.

Results for Cd0- The variation of C<j0 throughout the fullrange of Mach numbers is shown in figure 7, with eachgroup duly noted. There is a consistent trend in thevariation of the Group 2Cd0 with Mach number up toM = 0.80. Experiment 8 balance data appear to be higherthan the established trend beyond M = 0.70. A maximumvalue of C(j0 can be roughly estimated at M = 0.98. TheMcCroskcy-Smith expression superimposed on figure 7will be discussed in the next section.

In summary, no drag coefficient data exist beyondM = 1.10, the best drag coefficient data available arefound to be Group 2 quality, and the data arc onlyconsistent at speeds up to M = 0.80.

Results for (L/D)max and C/max- Figures 8 and 9 showthat (L/D)max and Q data from Experiments 2(balance and wake drag), 3, and 6 are consistent at speedsbetween 0.50 < M < 0.84. However, scatter belowM = 0.5 is evident. McCroskey showed that good datatend to exhibit high (L/D)max

ar|d Qmax va'ucs at 'ovv

speeds. (L/D)max and C/max data from Experiment 3 datawere noticeably higher at low speeds than the otherexperiments that produced Group 2 quality data. Based onthese figures, it appears that (L/D)max and C/max occur atroughly M = 0.3.

The inflection points of interest for C/a, C(j0, (L/D)max,and C/max for the SCI095 airfoil data arc given in table 3.

SC1094R8 Airfoil

Evaluation of pc/a— Figure 10 shows derived PQa

values from the experiments plotted versus log(Re).Figure 11 shows a log curve fit of the data within0.10 < PQa < 2n, along with Group 1 and Group 2tolerances. Some of the data from each experiment arcoutside the Group 2 tolerances. At least half of thebalance data from Experiment 8 and half of the pressuredata from Experiments 3, 4, and 6 are scattered within theGroup 2 tolerances. None of the experiments are consis-tently within the Group 1 tolerances. It can be concludedthat all of these experiments produced a certain amount ofGroup 2 quality lift coefficient data because the derivedPQa values are within the Group 2 tolerance.

Evaluation of C,j0- Figure 12 shows C^Q values fromall the experiments plotted versus log(Re), along with alog curve fit of that data, and Group 1 and Group 2 toler-ances. Most or all of the data from Experiments 3, 6,and 8 arc within the Group 2 tolerances. Data fromExperiment 8 are scattered, with a few points outside theGroup 2 tolerance boundary on the high side. Data fromExperiment 4 were not used in deriving the log curve fitand arc outside of the Group 2 tolerance. The implicationis that Experiments 3, 6, and 8 produced a certain amountof Group 2 quality drag coefficient data because the Cd0

values are within the Group 2 tolerance.

Groupings- Based entirely on the evaluation of PC/Q andC(jo as presented in Figures 10 through 12 the groupingsfor the SC1094R8 data arc

Group 1 None

Group 2 Experiments 3, 6, and 8

Group 3 Experiment 4

Group 4 None

Results for C/_— The variation of C/n throughout the ful lrange of Mach numbers is shown in figure 13. An exami-nation of the Group 2 data reveals that there arc slightlyconflicting trends in the variation of C/a with Machnumber. Experiment 6 values tend to be higher thanthe trend established by the other experiments belowM = 0.60. A maximum value occurs at M = 0.83, but nomin imum or recovered values can be established. Thedata tend to be more scattered beyond M = 0.70 than at

lower speeds, regardless of the groupings. The Smithexpression superimposed on figure 13 will be discussed inthe next section.

In summary, no lift coefficient data exist beyondM = 0.90, the best lift coefficient data available areGroup 2 quality, and these data are only consistent atspeeds up to M = 0.70.

Results for Cd0- The variation of Cd0 throughout the fullrange of Mach numbers is shown in figure 14. There arcconflicting trends in the variation of Cd0 with Machnumber beyond M = 0.70. Wake data from Experiment 6,and to a lesser extent from Experiments 3 and 8, exhibitlower drag values than the balance data from Experi-ment 8. Figure 14 also shows that some Experiment 8wake drag data arc high at low speeds, and those datapoints correspond to the high drag values noted infigure 12. A maximum value of Cd0 cannot be deter-mined. The Smith expression superimposed on figure 14will be discussed in the next section.

In summary, no drag coefficient data exist beyondM = 0.90, the best drag coefficient data available arcGroup 2 quality, and the data arc only consistent at speedsup to M = 0.70.

Results for (L/D)max and Qmax- Figures 15 and 16show (L/D)max and C/max data, respectively, from all theexperiments. These data arc only consistent at speedsbetween 0.60 < M < 0.84. Scatter below M = 0.6 isevident. Experiment 6 again appears to exhibit a different(L/D)max and C/ trend than the other experiments,which all tend to be in better agreement. Experiments 3and 8 exhibit slightly higher values of (L/D)max andC/max. Based on these figures it appears that (L/D)max

and C/max occur at roughly M = 0.3.

The inflection points of interest for C/a, Cd0, (L/D)max,and C/max for the SC1094R8 airfoil data arc given intable 3.

The results of this evaluation show that none of theexperiments produced Group 1 quality data. Some of theexperiments produced Group 2 quality data. Experiment 3produced Group 2 quality data for both the SCI095 andthe SC1094R8 airfoils. Experiment 8 produced Group 2quality data for the SC1094R8 airfoil. The SC1095data was found to be consistent up to M = 0.84 for l i f tcoefficient and M = 0.80 for drag coefficient. TheSC1094R8 data was found to be consistent up toM = 0.70 for both lift and drag coefficient, exceptfor some scattered drag data at low speeds. Other

experiments that produced Group 2 quality data werefound to exhibit slightly different trends, inconsistencies,or lower values of (L/D)max and C/max relative to theaforementioned experiments.

During the initial phases of this evaluation, the experi-menters responsible for the publication of the SCI 095 andSC1094R8 data were contacted. They were sent somepreliminary results and were asked to comment on thoseresults. The following responses were obtained and werefactored into the results presented in figures 3 through 16.

1. In general, at low drag levels, a balance sized to havehigh drag level capability does not give adequate reso-lution or precision.

2. In transonic or rotational flow balance drag data canbe more accurate than the total probes of a wake rakebecause a rake can not capture all of the losses.

3. When experimental angle-of-attack increments aretoo coarse to accurately derive reasonable values for C(j0,assume that QJO = Cda=Q- Note that the zero-lift angle ofattack for the SCI 095 and SC1094R8 arc -0.3° and -1.4°,respectively (rcf. 11).

4. Integration of surface pressures on a model with alimited number of pressure taps is generally inaccurate.

5. Tunnel wall porosity affects C/ and C/m, and isr J *Gt 'maxdiscussed in the published report for Experiment 5(rcf. 5).

Discussion

Discussion is warranted on the Group 1 and Group 2tolerances relative to accuracies in experimental measure-ments. To illustrate this, consider the derivation of l i f tfrom measured data as

1 2Lmcasurcd = C/measured ~PV (1)

where

C/'measured

=— f (Cp. -CP )dg£ I \ '/ ru /

Assuming there arc inaccuracies in the pressuretransducer measurements in the form of a simple bias

error, Pcrror. tncn

P / - P + P ,crror

P - P -Pru ' 'error-

substituting

'measured 2P.error

measured -2P,error

(2)

(3)error

Note that on the UH-60 Phase II Airloads Programpressure blade, the transducer measurements arc accurateto within 0.1% of their maximum range of 20 psia. Thus,

measured

error

equals 0.04 psia.

Now, consider the maximum error in the calculation oflift for a given group tolerance, as follows:

Croup 1

error

+ 0.0005JAa-pV2

-O.OOOSJAa-pV2

PAL( Group 1

error

PAL•Group 2

• O.OOlAa-pV2

• 0.008Aa-pV2

(4)

(5)

(6)error

Figure 17 is a plot of equations (3), (5), and (6) versusairspeed for a nominal anglc-of-attack range of 1.0°. Itcan be seen that a large region exists beyond M = 0.40,which indicates that the assumed bias error in liftmeasured by the UH-60 pressure blade is smaller thanthe maximum possible error that can be obtained whencalculating lift using Group 2 wind tunnel data. This isnot the case for Group 1 quality data.

The purpose of figure 17 is to show that Group 1 data aresufficiently accurate to use in predicting UH-60 airloads.It is not meant to imply that Group 2 quali ty data are notsufficiently accurate. Such a determination is dependenton many factors, such as:

1. The desired accuracy of the predictions.

2. Aspects of the physical representation of the UH-60that arc inaccurate or cannot be modeled, that overshadowany inaccuracies realized by using Group 2 quality data.

3. Limitations of the comprehensive analytical modelthat may overshadow any inaccuracies realized by usingGroup 2 quality data.

Until these issues have been resolved, it cannot be deter-mined whether Group 2 data arc sufficiently accurate fortheir intended use in the UH-60 model.

This last issue is the primary concern when evaluatinglifting-line methods employed in comprehensive rotor-craft analyses. It is not known if current methodologiesarc sensitive to errors introduced by using Group 2quality data. This question forms the basis for therecommendations discussed in the next section.

Also, it should be noted that consistent Group 2 qualitydata do not exist beyond speeds of roughly M = 0.70 orM = 0.80, depending on the airfoil. The advancing bladeMach numbers for the SC1095 and the SC1094R8 airfoilsarc 1.012 and 0.90, respectively, at the "do not exceed"velocity of 192 knots and 20,000 feet.

There are currently many sources of SCI095 andSC1094R8 data circulating throughout the aerospacecommunity in a variety of different formats. Not all thesources of data arc traceable. In fact, many of thosesources may be inaccurate. An easy way to check theaccuracy of those data sets would be to plot the variationof C/Q and Cd0 versus Mach number against the resultspresented in this report for either airfoil. Alternately, theresults presented in this report can be approximated usingsemi-empirical expressions developed by McCroskcy(rcf. 10) for the NACA 0012 airfoil in supersonic flowand by Smith (rcf. 12) for a variety of different airfoils insubsonic flow. Correlation of the semi-empirical expres-sions with the results presented in this report are shown infigures 6 and 7 for the SC1095 airfoil, and in figures 13and 14 for the SC1094R8 airfoil. The combined, orcomposite, McCroskcy-Smith expressions used togenerate the approximations shown on those figures arcpresented in the appendix.

Conclusions

The primary motivation for this evaluation was to preparefor rigorous analysis methodology validation as part ofthe UH-60 Phase II Airloads Program. Analysis method-

ology validation consists of assessing and improvingstate-of-the-art comprehensive analytical models throughexhaustive correlative studies in performance, dynamics,and rotor structural loads and airloads. In order to pro-ductively assess and improve the theories and assump-tions employed in comprehensive analytical models,accurate vehicle representations must be established. Themain rotor blade airfoil section characteristics arc amongthe most important parts of the vehicle representation.Ultimately, the improvements in analysis methodologyand in vehicle representations will be judged relative toimprovements in the correlation of predictions withexperimental measurements.

This report shows that the most accurate data arc Group 2quality. The experiments that generated this quality ofdata were performed in 1982 at the 11-Foot Wind TunnelFacility at NASA Ames Research Center and in 1985 atthe 6-Inch by 22-Inch Transonic Wind Tunnel Facility atOhio State University. It has not been determined whetherGroup 2 quality data arc sufficiently accurate to use incomprehensive analytical models to predict experi-mentally measured airloads data.

Furthermore, the McCroskey methodology used to •evaluate the airfoil data tend to work best when there arelarge amounts of data, a significant portion of which areGroup 1 and 2 quality throughout the desired ranges ofMach and Reynolds numbers. Although it can be arguedthat a significant percentage of the data presented hereinare Group 2 quality, they are not nearly as much asdesired, nor are they as consistent as desired. Therefore, itis important that further synthesis and experimentation beperformed in order to generate Group 1 quality data. Thisconclusion warrants specific recommendations, discussedin detail in the next section.

Recommendations

The results presented in this report show that the mostaccurate data arc Group 2 quality. It has not been deter-mined whether this is of sufficient accuracy to use inpredicting UH-60 airloads. If it is determined that theaccuracy is not sufficient, then the following recom-mendations should be interpreted as requirements.

It is recommended that further wind tunnel experimenta-tion be performed to obtain Group 1 quality data, and thatthis effort be preceded by a synthesis similar to thatperformed by Tanner (ref. 13). Candidate facilities forthe experimentation include those shown by McCroskey

to produce Group 1 quality data. However, speed, angle -

of-attack range, and Reynolds number range should be

considered before choosing a wind tunnel. It is under-stood that the aforementioned wind tunnels may not be

able to satisfy the high speed requirement, and this shouldweigh heavily in the selection of a wind tunnel facility.Further, large positive and negative angle-of-attack rangesshould be considered in increments small enough to

identify the exact values for C/max and Cdo. It isimportant that both the synthesis and experimentation be

performed for full scale Reynolds numbers. Actual

SCI 095 and SC1094R8 contours as measured on thePhase II Airloads Program pressure instrumented bladeshould be used if new wind tunnel models of those airfoil

sections arc to be fabricated. Determination of severalcritical parameters should be the priority of both thesynthesis and the experimentation. These parameters

include, but are not limited to, C/^ , Cin . .'«max '«min

Cin . Mnn, CH^ , the maximum values of'"recovered" UL" a°max'

(L/D)max and C/ , and the Mach numbers at which

they occur.

where

m ,

Finally, it is recommended that a methodology bedeveloped to evaluate pitching moment coefficient, andthat the synthesis and additional experimentation treatpitching moment coefficient with the same level of detailas lift and drag coefficient.

Appendix

For the SC1095 airfoil data, the combined, or composite,

McCroskey-Smith expressions are superimposed onfigures 6 and 7. In the M s 0.6 region the empirical log

curve fit of the C/Q data is given as

= 0.0531 + 0.0081 log(Re) (7)

The expressions developed by Smith for C/a are

Os Ms(M1 =0.84)

C'a(SM,) =C'a(M-0,Rc-6x,06) ^ + T7

(8)

andM1 < M s 0.93

, , ., ,'a(sM|)

-(M-M,) (9)

dC,S--5.4

dM

and the expression developed by McCroskey for C/Q is0.93<Msl.l

l)M2(t/c)]~ (10)

In the M s 0.6 region the log empirical curve fit of the

Cd0data is given as

Cdo -0.01 43 -0.001 Olog(Rc) (11)

The expressions developed by Smith for Cdo arc

0^ Ma (MDD = 0.8)

cf.L/ Cs\ K[(0.01745)|aZL|]

2-7

c\ CF •6x10°(12)

where

c f -•

aZL--°-3dcg

SA-1.14

2.022c

Zs- - 0.02CF

RNo -Rc =6x106

K-2.12

and MDD < M < 1-0

0.4552.58

do(>MDD) " do(MDD) 3M

where

dCn• 0.4

-MDD) 03)

dM

and the expression developed by McCroskey for Cdo is

1.0s Ms 1.1

1-1/3 where

where

a = 5.0

For the SC1094R8 airfoil data, all of which were subsonicdata, the Smith expressions arc superimposed on fig-ures 13 and 14. In the M s 0.55 region the empirical logcurve fit of the C/a data is given as

pC/a =0.0977+ 0.0009 log(Re) (15)

The expressions developed by Smith for C/a areO s M s ( M =0.83)

(16)

and M, < M < 0.9

where

(17)

dCifa_dM

• 5A

In the M < 0.55 region, the empirical log curve fit of the

Cd0data is given as

Cdo =0.0146 -0.0009 log(Rc) (18)

The expressions developed by Smith for C(j0 arc0 s M =s (MDD = 0.7)

SA-M+^U-CF/I

ZL|]2-7

f = 6x10°(19)

0.4552.58

"ZL --l-4deg

SA-1.14

- - 2.022c

•s- . 0.02CF

R N o - R e - 6 x 1 0 °

K-2.12

and MOD < M < 0.9

Cd0(>MDD) '

where

acd^--0.15

dM

The above expressions compare well with the notedtrends of C/a and C(j0 versus Mach number for both theairfoils, with a small discrepancy noted in C/Q for theSC1094R8 data between M = 0.5 and M = 0.8. Theseexpressions are meant only to quantify observed trends.An analysis of the correlation is not within the scope ofthis paper.

References

1. Griffin, T.: Wind Tunnel Tests of Sikorsky SC1095Two Dimensional Airfoil with Various SurfaceIrregularities. UARL Report M432170-1,October 17, 1973.

2. Vogt, P.: An Evaluation of NACA 0012 and SC 1095Airfoil Tests in the UTRC Large Subsonic WindTunnel. UTRC75-121, December 31,1975.

3. Lcdniccr, D.; and Owen, S.: An Evaluation of ElevenHelicopter Airfoil Sections in the Ohio StateUniversity 6 x 22 Transonic Facility.SER-760603, November 15, 1985.

4. Flcmming, R. J.; and Lcdniccr, D. A.: High SpeedIce Accretion on Rotorcraft Airfoils. NASACR-3910,1985.

5. Jepson, W. D.: Two Dimensional Test of Four AirfoilConfigurations with an Aspect Ratio of 7.5 and a16 inch Chord up to a Mach Number of 1.1.SER-50977, April 5, 1977.

6. Noonan, K. W.; and Bingham, G. J.: AerodynamicCharacteristics of Three Helicopter AirfoilSections at Reynolds Numbers From ModelScale to Full Scale at Mach Numbers from 0.35to 0.9. NASA TP-1701, September 1980.

7. Hicks, R. M.; and Collins, L. J.: An Evaluation ofThree Helicopter Rotor Sections. NASATM-86719, August 1985.

8. Flemming, R. J.: An Experimental Evaluation ofAdvanced Rotorcraft Airfoils in the NASAAmes Eleven-Foot Wind Tunnel. NASACR-166587, April 1982.

9. McCroskey, W. J.; McAlistcr, K. W.; Carr, L. W.;and Pucci, S. L.: An Experimental Study ofDynamic Stall on Advanced Airfoil Sections.NASA TM-84245, July 1982.

10. McCroskey, W. J.: A Critical Assessment of WindTunnel Results for the NACA 0012 Airfoil.NASA TM-100019, October 1987.

11. Shanlcy, J. P.: Validation of UH-60A CAMRAD/JAInput Model. SER-701716, November 1991.

12. Smith, R. L.: Closed-Form Equations for the Lift,Drag, and Pitching-Moment Coefficients ofAirfoil Sections in Subsonic Flow. NASATM-78492, 1978.

13. Tanner, W. H.: Charts for Estimating Rotary WingPerformance in Hover and at High ForwardSpeeds. NASA CR-114, 1964.

10

Table 1. SC1095 and SC1094R8 wind tunnel data reference list

Experiment facility Mach rangereport no.

Experiment 1 0.3-0.75LSWTSftUARLM432170-1

Experiment 2 0.3-0.75LSWTSftUTRC75-121

Experiment 3 0.3-0.85Ohio State 6 in. x 22 in.SER-760603

Experiment 4 0.3-0.87NRC high speed icing

facilityNASACR-3910

Experiments 0.3—1.1NSRDC 7 ft x 1 0 ftSER-50977

Experiment 6 0.35-0.9LaRC 6 in. x 28 in.NASATP-1701

Experiment 7 0.2-0.88Ames 2 ft x 2 ftNASATM-86719

Experiments 0.3—1.1Ames 1 1 ftN ASA CR-1 66587

Experiment 9 0.11-0.3Ames 7 ft x 10ftNASA TM-84245

Rc range Model and tunnel(range x 106) dimensions

2.52-4.97 b/c = 2.06h/c = 5.8

2.54-4.94 b/c = 7.75h/c = 5.8

2.88-7.57 b/c = 3.67h/c =1.0

1 .80-4.60 b/c = 2.0h/c = 2.0

1.29-3.96 b/c = 7.5h/c = 5.25

1.60-6.70 b/c =1.94h/c = 9.03

1 .87^1.00 b/c = 4.0h/c = 4.0

3.60-6.05 b/c = 8.21h/c = 8.21

1.44-3.88 b/c = 3.49h/c = 5.0

Remarks

SCI 095Limited rights data,balance, wake

SC1095Limited rights data,pressure, wake

SC1095, SC1094R8Partial limitation,pressure, includingtripped data

SC1095, SC1094R8Domestic limitation,

pressure, wake

SCI 095Balance, pressure,

wake, includingstrut data

SC1095, SC1094R8Pressure, +3° (up)TE tabs

SC1095Pressure

SC1095, SC1094RSBalance, pressure,

wake

SC1095Pressure, wake

LSWT = large-scale wind tunnelNSRDC = Naval Ship Research and Development Center

11

Table 2. Wind tunnel facility reputation and wall corrections used in the experiments

Experiment

Experiment 1Experiment 2Experiment 3

Experiment 4

Experiment 5

Experiment 6

Experiment 7Experiment 8Experiment 9

Wind tunnel

LSWTSftLSWTSftOhio St. 6 in. x 22 in.

NRC Icing

N S R D C V f t x 10ft

LaRC6in. x 28 in.

Ames 2 ft x 2 ftAmes 1 1 ftAmes 7 ft x H)ft

Documentedreputation"

Group 3Group 3Group 4

Group 4

Group 4(capable ofGroup 3)

Group 2

Tunnel wall corrections

Linear wall corrections, solid wallsLinear wall corrections, solid wallsIndependent plenums for top and bottom walls, porouswalls

Linear plenum, surface corrections, and high turbulencelevels

Large lift interference, with and without wake dragcorrections, high porosity, and slotted walls

AOA corrected, side wall boundary layer effects on shockposition, and C/max, and slotted walls

AOA corrections and slotted wallsWall corrections but none on AOA, and slotted wallsLinear wall corrections, and solid walls

"Established by McCroskey in his evaluation of NACA 0012 data.

Table 3. Summary of results

SCI 095

SC1094R8

ValueMach

number

Value

Mach

number

Cd°max Cd°MDD

0.09 0.00750.98 0.80

* 0.00770.70

(L/D)max

1050.30

(115)0.30

C'max

(1.35)0.30

(1.75)0.30

/amax /amin

(0.24) (0.05)0.84 0.90

0.20 *

0.83

'a recovered

(0.085)0.93

*

() denotes scattered data* denotes no data

12

O.OORI

I

I

I

I

I

0.11111111

1:

7R

0.1 9R11 SCI 095

airfoil

|

1

0.47R1

1

'o.SOR1 11 |

H l .

i |

1 1

0.82R l.OOR1 1

1 1

!0.85R '1 1 1

SC1094R8 | | SC1095 |airfoil airfoil

H i ,_, . ir^ ^i

1 1 1

" 1Transition Transition Transition

Figure 1. UH-60A Black Hawk main rotor blade airfoil section locations.

SC1095airfoil

SC1094R8airfoil

Figure 2. SC1095 and SC1094R8 airfoil sections.

13

o

.16 -

.14 -

.12 -

.10 -i

.08 -

.06 -

.04

Tripped data

Experiment

H 1, Balanceo 2, BalanceD 2, PressureD 3, Pressure* 4, PressureB 5, Balance* 5a, Pressure• 5b, PressureA 6, Pressureo 8, Balance• 8, Pressurex 9, Pressure

106 107

Log (Re)Figure 3. SC1095 Mach corrected lift curve slope versus Reynolds number, M < 0.6.

o>o

o

.16 n

.14 -

.12 -

.10 -

.08 -

.06 -

.04

Experimento 2, BalanceD 3, PressureA 6, Pressure

0.053131 + 0.0081188 Log (R e)

Tripped data / 1

rad\ /0.10

::::: } Group 1 range

;;;;;;; } Group 2 range

10 6 107

Log (Re) .Figure 4. SC1095 Mach corrected lift-curve slope versus Reynolds number with Group 1 and Group 2 tolerances,M < 0.6.

14

.0150 -,

.0125 -

.0100 -

.0050 -

.0025 -

Tripped data

Experiment

D 1,Wake

o 2, Wake

a 3, Wake

• 4, Wake

• sa,Wake

• 5b, Wake

• 6, Wake

• 8, Wake

x 9, Wake

0.014282 - 0.0010138 Log (R e)

I!I!I } Group 1 range

"I" } Group 2 range

10 10Log (Re)

Figure 5. SC1095 drag coefficient at zero-lift versus Reynolds number with Group 1 and Group 2 tolerances. M < 0.6.

.3 n

.2 -

o>o>

O

.1 -

Composite McCroskey-Smith expression Group 2, Experiment

o 2, Balanceo 3, PressureA 6, Pressure

Group 3, Experiment

• 1, Balancea 2, Pressure» 4, Pressure

B 5, BalanceA 5a, Pressure• 5b, Pressure• 8, BalanceM 8, Pressurex 9, Pressure

.4 .6M

1.0 1.2

Figure 6. SC1095 lift-curve slope versus Mach number.

15

Composite McCroskey-Smith expression

X(0

150 -

125 -

100 -

75-

50 -

25 -

Group 2, Experiment

o 2, Wake

a 3, Wake

A 6, Wake

Group 3, Experiment

• 1,Wake* 4, WakeB 5, Balance

• 5a, WakeA 5b, Wake

* 8, Balance

8, Wake

9, Wake

.2

Figure 7. SC1095 drag coefficient at zero lift versus Mach number.

J i 5H I

Group 2, Experiment

o 2, Balance/wakeo 3, Pressure/wakea 6, Pressure/wake

Group 3, Experiment

• 1, Balance/wakea 2, Pressure/wake• 4, Pressure/wake

B 5, Balance/wakeA 5a, Pressure/wake

• 5b, Pressure/wake

o 8, Balance/wake• 8, Pressure/wake

x 9, Pressure/wake

'«B

.2 .4 .6 .8 1.0M

1.2

Figure 8. SC1095 maximum lift-to-drag ratio versus Mach number.

16

O

Xaj

2.00 -•

1.75 -•

1.50 -

1.25 -

1.00 -

.75 -

.50 -

.25 -

0

Grouo 2,

o 3,A 6,

x x x8 * rX Grou

X I • 1

I 8 • 2,'I o * 4,

A A iA • H 5'

A A ^5*1

• A "5b,* 8,• 8,

• H

.2 .4 .6 .8 1.0 1.1M

.16 -]

.14 -

.12 -

O)0)

o.10 -

.08 -

.06 -

.04

Group 2, Experiment

2, Balance

3, Pressure6, Pressure

Group 3, Experiment

, Balance, Pressure

4, Pressure, Balance

, Pressurei, Pressure8, Balancei, Pressure

i, Pressure

Figure 9. SC1095 maximum lift coefficient versus Mach number.

oX

Experiment

o 3, Pressure

x 4, Pressure

o 6, Pressure

o 8, Balance

A 8, Pressure

2

106 107

Log (Re)

Figure 10. SC1094R8 Mach corrected lift curve slope versus Reynolds number, M < 0.55.

17

.14 -

.12 -

O

.08

- 0.097663 + 0.0008767 Log (R)

--- 0.10

•;•;; } Group 1 range

;:;;; } Group 2 range

Experiment

o 3, Pressurex 4, Pressure

D 6, Pressure

° 8, Balancea 8, Pressure

10'Log (Re)

Figure 11. SC1094R8 Mach corrected lift-curve slope versus Reynolds number with Group 1 and Group 2 tolerances,M < 0.55.

.0150 -

.0125 -

.0100 -

0° .0075 -

.0050 -

.0025 -

.0000

Experiment

o 3, Wakex 4, Wake

a 6, WakeA 8, Wake

/•> A A "'" •-•-•

x x

0.014616 -0.0009379 Log (R@)

~'.~" } Group 1 range

"II } Group 2 range

10' 10'Log (Re)

Figure 12. SC1094R8 drag coefficient at zero-lift versus Reynolds number with Group 1 and Group 2 tolerances,M < 0.55.

18

i- Q)

Group 2, Experimento 3, Pressurea 6, Pressure« 8, Balance* 8, Pressure

Group 3, Experimentx 4, Pressure

Smith expression

.2

.06-1

Figure 13. SC1094R8 lift-curve slope versus Mach number.

O

.05-

.04-

.03-

.02-

.01 -a9.

ft

Group 2, Experimento 3, PressureD 6, Pressure« 8, BalanceA 8, Pressure

Group 3, Experimentx 4, Pressure

Smith expression

.2 .4 .6M

.8 1.0

Figure 14. SC1094R8 drag coefficient at zero lift versus Mach number.

19

150 -

125 •

100 -

E

10

50 -

25 -

oax

oAA

Group 2, Experimento 3, Pressure/wakea 6, Pressure/wakeo 8, Balance/wakeA 8, Pressure/wake

Group 3, Experimentx 4, Pressure/wake

.2 .4 .6M

.81

1.0

Figure 15. SC1094R8 maximum lift-to-drag ratio versus Mach number.

XCO

h»

2.00-

1.75-

1.50-

1.25-

1.00-

.75-

.50-

.25-

0

o

ioA Group

D 1 ° 3'

B D 6,

x 9 * 8,

* ftn a Group

x 4-

• 1 • 1 • 1 • 1 • 1

.2 .4 .6 .8 1.0M

2, ExperimentPressurePressureBalancePressure

3, Experiment

Pressure

Figure 16. SC1094R8 maximum lift coefficient versus Mach number.

20

.25n

.20-

a -15 H

.10-

.05-

PALGroup 1

error

Group 2

error

Measure

error

Region where the error inmeasured lift is less thanthe error associated with

maximum group 2 tolerances

0.2 0.4 0.6 0.8M

1.0

Figure 17. Comparison of the error in lift derived from measured pressure transducer steady bias error, and Group 1 andGroup 2 tolerances versus Mach number for a nominal 1 degree change in angle of attack.

21

REPORT DOCUMENTATION PAGEForm Approved

OMB No. 0704-0188

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time tor reviewing instructions, searching existing data sources,gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of thiscollection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate tor information Operations and Reports, 1215 JeffersonDavis Highway. Suite 1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project (0704-0188), Washington. DC 20S03.

1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE

September 19933. REPORT TYPE AND DATES COVERED

Technical Memorandum4. TITLE AND SUBTITLE

A Critical Assessment of UH-60 Main Rotor Blade Airfoil Data

6. AUTHOR(S)

Joseph Totah

5. FUNDING NUMBERS

505-59-36

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)

Ames Research CenterMoffett Field, CA 94035-1000

8. PERFORMING ORGANIZATIONREPORT NUMBER

A-93004

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)

National Aeronautics and Space AdministrationWashington, DC 20546-0001

10. SPONSORING/MONITORINGAGENCY REPORT NUMBER

NASA TM-103985

11. SUPPLEMENTARY NOTESPoint of Contact: Joseph Totah, Ames Research Center, MS 237-5, Moffett Field, CA 94035-1000

(415) 604-5057Presented at llth Annual Applied Aerodynamics Meeting, August 9—13,1993, Monterey, California

12a. DISTRIBUTION/AVAILABILITY STATEMENT 112b. DISTRIBUTION CODE

Unclassified — UnlimitedSubject Category 02

13. ABSTRACT (Maximum 200 words)

Many current comprehensive rotorcraft analyses employ lifting-line methods that require main rotorblade airfoil data, typically obtained from wind tunnel tests. In order to effectively evaluate these lifting-linemethods, it is of the utmost importance to ensure that the airfoil section data are free of inaccuracies. Acritical assessment of the SC1095 and SC1094R8 airfoil data used on the UH-60 main rotor blade wasperformed for that reason. Nine sources of wind tunnel data were examined, all of which contain SC1095data and four of which also contain SC1094R8 data. Findings indicate that the most accurate data weregenerated in 1982 at the 11-Foot Wind Tunnel Facility at NASA Ames Research Center and in 1985 at the6-inch-by-22-inch transonic wind tunnel facility at Ohio State University. It has not been determined if datafrom these two sources are sufficiently accurate for their use in comprehensive rotorcraft analytical modelsof the UH-60. It is recommended that new airfoil tables be created for both airfoils using the existing data.Additional wind tunnel experimentation is also recommended to provide high quality data for correlationwith these new airfoil tables.

14. SUBJECT TERMS

UH-60, Airfoils, Assessment

15. NUMBER OF PAGES

2216. PRICE CODE

A03

17. SECURITY CLASSIFICATIONOF REPORT

Unclassified

18. SECURITY CLASSIFICATIONOF THIS PAGE

Unclassified

19. SECURITY CLASSIFICATIONOF ABSTRACT

20. LIMITATION OF ABSTRACT

NSN 7540-01-280-550O Standard Form 298 (Rev. 2-89)Prescribed by ANSI Std. Z39-18298-102