Some

Moil Tests By Bob Whittier

Sports. The catalogs listing them run to a total of around

INCE ITS founding over 40 years ago, the NACA (now NASA) has published a vast number of technical re-

1,000 pages! I have spent much time poring over these catalogs, looking for reports of possible interest to EAA members, and have been repeatedly impressed with the fact that most NACA publications of use to us date from the 1920s and 1930s. In those days the NACA was hard at work improving wood-wire-and-cloth airplanes in the 100 mph category. Since the mid-1930s, NACA's work has trended more and more toward improving fast military and commercial aircraft, which is as it should be. While doubtless some useful material lies hidden among the advanced tomes published in the 1940s and 1950s, it has been far from

easy to find and extract it. With those prefatory remarks out of the way, let's look at some of the NACA findings in a field which is

of very great interest to us. We are always wondering about the aerodynamic properties of our experimental, hand-made light airplane wings. What follows is presented because it is informative, thought-provoking and representative of NACA's best findings in this field. Nat-

urally, it reflects the state of the art as it was years ago, and some of our ideas have changed since then. Nevertheless, my careful scrutiny of later NACA report lists

has not revealed anything which would destroy or substantially modify the basic value of the following information. Should any reader know of specific data which I may have missed, he should by all means notify me so that pertinent follow-up material may be published. In 1927 NACA published TN 255, "Precision of Wing Sections and Consequent Aerodynamics Effects".

This report originated from the realization that some imperfection in airfoil contour is inevitable in wings made of wood and fabric. For one thing, they wanted to know how important it might be to use extreme care in making wind tunnel wing models, because if inaccuracy in real wing construction proved to be appreciable, it might also be necessary to reproduce them in model wings for the sake of obtaining reliable data in wind tunnel tests. They

wanted to find out how aging and flexing of the wood, and sag of the fabric, altered the airfoil contour. Yet another aim was to decide whether veneer-covered wings had any substantial advantage over fabric-covered ones,

as a result of greater airfoil accuracy. A wooden frame was made, Fig. 1, with which to take off contours from actual wings. The upper cross bar was removable so that the rig could be set over a wing,

and the adjustable pointers could be moved so that they just touched the wing surface. As a result it was possible to put on paper the actual airfoil contours. One measurement was made at the wing ribs and another halfway between the ribs. It was found that due to workmanship, fabric tension, etc., the outlines of the actual ribs sometimes varied noticeably from the perfect outline as plotted from airfoil coordinates. And, of course, fabric sag caused readily-measurable changes in airfoil shape

between ribs. The latter defect naturally was much greater, and had the overall effect of reducing the airfoil thickness.

Fig. 2 shows some typical results. The dashed lines showing fabric contour at the noses of some ribs show what appears to be bulging rather than sagging fabric; no explanation for this is given in the text and it must be assumed that it is the result of leading edge structural flexing between the ribs. For example, fabric tension

would cause leading edge veneer to develop bulges, and the resistance of plywood to bending would, of course, pro-

duce a bulge such as on the nose of the Gottingen 298 and Clark Y. . The worst case of sag was in the USA-35B airfoil

(not illustrated) and resulted in a reduction of airfoil thickness of .03 percent. Through computations and previous tests on airfoil thickness, it was concluded that the effect would be a drop of only .07 in the maximum

lift coefficient and no appreciable change in the drag coefficient. It was shown that the .03 percent drop occurred only at the highest angles of attack and so the sag effect for even the worst cases would have no appreciable consequences at ordinary flying attitudes. The overall conclusion was that there was no reason for incorporating simulated sag in model wings.

In 1932 thicker airfoils had become common and

many designers were using metal and plywood covered leading edges to avoid the proportionally greater sag on

thicker airfoils at the nose. The question arose of whether "Nonetheless Gregory there are those who would consider THAT merely as constructive criticism".

the extra cost and weight was warranted. An investigation was made involving the fairly thick Gottingen 387 (Continued on next page) SPORT AVIATION

IS

NACA TESTS . . .

(Continued from page 15)

airfoil on a Fairchild FC-2 airplane. Exact contours were taken off at the rib, and between ribs. A solid metal model wing was made and tested, first with smooth surface, then was surface machined to represent a fabriccovered wing with leading edge sheathing, and then a fabric-covered wing with no leading edge sheathing. Fig. 3

shows the sag which resulted. Tests were made and published in TN 428, "Characteristics of an Airfoil as Affected by Fabric Sag". Fig. 4 is taken from this report and you may see for yourself how little variation there is in aerodynamic characteristics. The report itself concludes that "fabric sag between ribs has a very small effect on the aerodynamic characteristics of a wing". It was noted that on the FC-2 where there was a break in the contour, at the rear end of the leading edge sheathing, the angle between tangents of the reinforced and sagging surfaces was approximately seven degrees. This small structural defect is considered to have more deleterious effect than sag alone. The prime purpose of leading edge reinforcement is to hold the fabric up against high-speed air pressures. By 1933 airplane speeds had increased appreciably, and designers began to wonder what could be gained by polishing wings. Particular attention was given to nose smoothness, and the results published in TN 457, "The Aerodynamic Characteristics of Airfoils as Affected by Surface Roughness". It was found that as the Reynolds Number increases, there is marked increase in sensitivity to roughness. Irregularities as small as approximately .0002 times the chord can cause adverse results. Fig. 5 graphically shows the difference between a highly polished and a rough surface. The polished surface in this test was equivalent to wet sanding with fine abrasive paper, followed by hand rubbing and waxing, while the rough surfaces were simulated by coating the surface with No. 180 carborundum grains. The implication given by part of the text is that the surface laps in an all-metal wing's skin can cause more drag than the relatively unbroken surface of a fabric-covered wing. Common laps facing aft were shown to cause 12 pounds of drag at 200 mph and to consume 6.5 hp. Scratches of as little as .0002 x chord, if located on the leading edge material, can cause increased drag. If a wing walk near the root has its rough surface carried all the way to the leading edge, the result can be considerable resistance as compared to a wing walk located on the after part of the wing root only, in fast airplanes. The tendency of a rough wing walk to cause drag is increased by the factor of wing-fuselage interference.

When the year 1939 arrived, all-metal airplanes were

in common use and NACA undertook an investigation of the drag caused by rivets, spot welding, plain laps and jogged laps. The rows of rivets and the several lap joints in an all-metal wing can produce a surface that is very far indeed from the original smooth airfoil outline, as in Fig. 6. As a result, TN 695, "The Effects of Some Common Surface Irregularities on Wing Drag" was published.

It concludes that rivets spaced % in. apart in 13

spanwise rows on top and bottom surfaces of a five-foot

chord airfoil increased the drag by from 6 percent for

countersunk rivets, to 27 percent for 3/32 in. brazier head

rivets. About 70 percent of this drag was due to the rivets on the forward 30 percent of the airfoil. Lapped joints, six on each surface, increased drag by from 4 percent for joggled laps to 9 percent for plain laps. The roughness due to bad spray painting increased drag by 14 percent, and coating the wing with grains .0013 in. in diameter increased it by 42 percent. Manufacturing irregularities such as bulges and wrinkles in a metal wing increased its drag by 8 percent of the smooth-wing drag, over and above drag caused by rivets and laps. On a transport plane doing 250 mph, it would take 500 hp to overcome the drag of rivets and laps alone on a wing such as was used in these tests, but if the forward 30 percent of the wing were to be smoothed up, the power required to overcome rivet and lap drag would drop to 160 hp! A few months after the appearance of the above report, TN 724 made its appearance, and was entitled "The Effects of Surface Waviness and of Rib Stitching on Wing Drag". A smooth test wing was made and tested to get basic figures, then modified by attaching protuberances to its surface. First, several spanwise waves were created, each 3.0 in. wide and .048 in. high, to duplicate the several spanwise strips of aluminum making up a metal skin. When they covered only the rear 67 percent of both surfaces they increased drag by only 1 percent, but when they covered the rear 92 percent of the surfaces they increased drag by 10 percent. A single such wave located at the 10.5 percent chord position on the top surface, caused an increase of drag of 6 percent by causing premature flow separation. This substantiates the earlier findings that a ridge or break in the upper surface where the fabric runs off the nose reinforcing can be more deleterious than is commonly supposed. The same series of tests were run with imitation ribstitching attached to the wing at six-inch intervals. This resulted in a 7 percent increase in wing drag, about one-third of this increase being due to premature separation of the airflow at the forward ends of the ribstitching.

The practical significance of this is that it pays to do a neat stitching and taping job. Before putting on the pinked tape, go over all the ribstitching knots, pressing them down into the fabric at one side of the rib with your thumb to avoid the lumps caused by knots. When performing stitching, pull up each knot really tight, to bed the ribstitching cord down into the reinforcing tape and so minimize the amount it will project. When putting on the pinked tape, do not leave bubbles under it where it humps up along the ribs, just hoping that further doping will pull it down smooth. Deliberately and carefully stick it down onto the wing cloth, working out all humps and bubbles over the ribs.

In short, good workmanship can pay off in increased

performance. Put your wings together and cover them with constant thought to providing a smooth, unbroken surface. "Why don't you watch your step?!!!"

16

1962