Rapid Prototyping and Experimental Design

Wheel pants and gear leg fairings for the SR-1 speed plane.

In my previous “Rapid Prototyping” series of articles, I gave overviews of molding various wing parts. Well, the SR-1 wing is finally done, and since the gear attaches to the wing spar, I decided I’d finish up the gear fairings and wheel pants as well. In this first of a pair of articles, we’ll look at some of the design concepts of those parts and cover the fabrication of the gear leg fairings. In the second part of the article, I’ll show one method of making your own wheel pants.

When I originally pitched this article to editor Paul Dye, I told him I wasn’t sure if this would really grab people’s interest—I mean, how many people actually ever make their own wheel pants? In fact, my neighbor here at Santa Paula (California) Airport, Klaus Savier of Light Speed Engineering, told me to expect it to take about 200 hours, and that is almost exactly how long it took to make them. I told Paul I couldn’t imagine why anyone wouldn’t just use premade pants. For example, the ones sold by Van’s Aircraft are reasonably priced and well designed. I can’t imagine that optimized pants would gain more than a knot or two on an RV. “Well, maybe this will be good to explain to people why they wouldn’t want to make their own pants,” said Paul. So there you go—for some this may be instructional, but for most, probably a cautionary tale!

The mold being CNC’d. It’s made from a piece of 15-pound per cubic foot tooling foam.

Some Thoughts on Design

Let’s begin with the aerodynamic design of the fairings since I think they represent an interesting puzzle for which the solution is not necessarily obvious. For simplicity’s sake, we’ll assume the gear is of a circular cross section (although the following discussion can be applied to any cross sectional shape), such that the diameter of the gear leg equals the maximum thickness of the airfoil streamlining it. Our goal is to minimize drag; a secondary goal is to minimize weight. What is the optimal shape to fair the gear to minimize drag? One might grab Abbott and von Doenhoff’s classic compendium of airfoil shapes off the bookshelf and find the lowest drag coefficient airfoil in there, but this would almost certainly not work out well.

The mold is split after being CNC’d. Locator pins and holes are visible at the mold corners.

Here’s why: Coefficients are dimensionless and used for comparison purposes. Having been developed for use in wings, airfoil drag coefficients are referenced to chord since this is one of two dimensions (the other being span) that give us planform area and thus lift, which is the job of a wing. In a world where wet wing fuel capacity is your primary interest though, you could certainly have a coefficient based on the area of the foil (area times span gives volume, and thus fuel capacity).

Since the job of a fairing is not lift, but rather to provide the smallest possible drag of an enclosed object, actual chord is irrelevant, except as it affects Reynolds numbers (more on that below). In the case of foils for tubular landing gear, the appropriate drag coefficient (assuming the gear is placed at the thickest part of the foil) is based on thickness, not chord.

The mold halves assembled. You can see the half-inch-wide, .050-inch-thick joggle at the leading and trailing edges for the flange to sit in when bonding the two halves.

To make a point, let’s compare a 9% thick NACA 66-009 with a generic 30% thick fairing foil. If you wrap the 009 around a 1.25-inch diameter gear leg, the fairing will have a 13.9-inch chord, giving a drag coefficient (C_d) of approximately .004 at a Reynolds number (Re) of 2 million at 200 mph. On the other hand, Hoerner suggests that a 30% thick foil is about the optimum for a fairing. That would yield a fairing chord of 4.2 inches, with a Re of 660k at 200mph and a C_d of approximately .013. The drag calculation is as follows:

D = ρ V² S C_d

Where ρ V² is q, dynamic pressure

S is fairing area (chord x length)

C_d is the drag coefficient of the fairing foil

Dynamic pressure q and fairing length are identical for both cases, so

D ∝ chord x C_d

Therefore

D₀₀₉ = 13.9 x .004 = .056

And

D₀₃₀ = 4.2 x .013 = .055

So the drag is basically a wash. On the other hand, assuming identical construction methods, the 30% foil will only weigh one-third that of the 9% foil. Although Hoerner suggests a 30% foil is optimum for fairing bluff bodies, that assumes equal Re. There is a large drag jump for airfoil C_d as Re drops below 1 million, so depending on the actual cruise speed of the airplane, Re may favor thinner foils at lower speeds. The thinner foil forces a longer chord, which increases the Reynolds number. The wetted area increases, but the drop in drag coefficient more than offsets it.

After machining, a bond joggle is created on the leading edge with tooling wax. I generally prefer to make joggles with tooling wax rather than machining them into the mold.

So, when choosing an airfoil or performing an optimization study, be sure to use the correct Re of the foil. You also want to know if the foil has flow attachment issues like laminar bubbles, which could be corrected with zigzag tape. Finally, you’ll also want to be sure the foil can deal with small off-angles of operation, especially if they are downstream of prop wash.

For purposes of simplicity, I have ignored the effect of fairings and wheel pants on stability, but suffice it to say, depending on their size and drag characteristics, fairings and wheel pants can be significantly stabilizing or destabilizing. I know one Red Bull team that saw a large change in pitch trim after moving from stock Edge 540 wheel pants to the custom laminar-type pants teams now favor.

BID tape is first wet out on a piece of transfer plastic and then cut into strips.

The strips are then positioned on pieces of peel ply about 1 inch wider than the BID tape. Once positioned on the peel ply, the transfer plastic is removed.

After the BID tape cures, the leading edge flange is trimmed off.

The outer leading-edge bond joggle is sanded smooth, then wiped with acetone to prep for bonding the outer leading-edge BID tape. Use the industry standard double-wipe method described in Step 9.

Gear Fairing Construction

Construction of gear fairings is relatively straightforward. Assuming the intersection fairings are molded separately, the gear fairing will be symmetric left and right, and a single mold can be used for both. For these kinds of small, symmetrical molds, I prefer to machine the mold in a single piece, then split it down the middle. Here are the steps involved with making the fairings.

1. The mold blank is constructed in the usual manner. This mold was made from a piece of scrap 15-pound per cubic foot tooling foam (Coastal Enterprises’ PB Board). To ensure alignment, locator pins and holes were placed at the four corners of the mold. When bonding the two halves, the flanges sit in a half-inch wide joggle at the leading and trailing edges. The joggle thickness is exactly equal to the thickness of the combined flanges.
2. Surface prep: Tooling foam molds can be released using either the double bagging method (see related articles, “Black Beauty,” April 2015 and “Rapid Prototyping,” November 2017), or by squeegeeing on a coat of epoxy, or epoxy plus Cab-O-Sil (approximately cup Cab per 100 grams epoxy). When epoxy sealing with Cab, be sure to squeegee the coat as thin as possible to avoid adding thickness to the outer mold line (OML). The sealed surface should then be sanded with 220 to ensure a smooth surface. (We’ll assume you are not looking for a Class A surface with these molds, since you are not going to get that with lightweight tooling foam or MDF anyway). This should be followed by three coats of mold wax (i.e., Meguiar’s #8), followed by a coat of PVA or a release like Chemlease. MDF molds should be sealed with a minimum of three coats of Minwax bare wood sanding sealer, then waxed/released.
3. After machining, a bond joggle is created on the leading edge with tooling wax. I generally prefer to make joggles with tooling wax rather than machining them into the mold. The reason for this is that once the part is assembled, the tooling wax can be removed from the mold, and the finished part will nest correctly in the mold. If the joggle is a permanent (machined) part of the mold, the finished part will no longer nest.
4. The left and right halves are laid up in a single operation. Three layers of 6-ounce cloth should be sufficient for most aircraft.
5. The part is demolded, cleaned up, and split into left and right halves. Sand about 1 inch of the inner leading edge to prep for bonding the leading edge.
6. The left and right halves are secured in their respective molds. Gel-type super glue is applied to the leading edge flange area, one dab every one to two inches, and the mold halves are assembled with the locator pins. (Be sure to include these as part of the CAD model!) Be careful not to glue the fairing to the mold.
7. After the glue cures, the fairing is removed and a liberal bead of penetrating super glue is applied to the entire inner leading edge join seam. Do not glue the trailing edge at this time—you need it open for the next step.
8. After the penetrating super glue has cured, a strip of BID tape (bidirectional cloth oriented 45 degrees to the long axis of the tape) is applied to the inner joint line, with access from the trailing edge. Be careful not to pry the trailing edge too far open, or you may cause the leading edge flanges to split apart.
9. After the BID tape cures, the leading edge flange is trimmed off. Next, the outer leading edge bond joggle is sanded smooth, then wiped with acetone to prep for bonding the outer leading edge BID tape. Use the industry standard double wipe method shown in the photo: A clean piece of gauze is swiped after the gauze soaked with acetone. This picks up any contaminants dissolved by the acetone before the solvent has a chance to evaporate and redeposit the contaminant.
10. The exterior leading edge is now joined with BID tape. BID is first wet out on a piece of transfer plastic and cut into strips. The strips are then positioned on pieces of peel ply about 1 inch wider than the BID. Once positioned on the peel ply, the transfer plastic is removed. This makes for a neat, clean taping operation.
11. The BID tape is then placed on the leading edge. Strips of peel ply are wrapped over the leading edge and taped in place with foil tape to hold the BID tape securely in place.
12. There are various ways of finishing the trailing edge of the fairing. The trailing edge can be held closed by intersection fairings, bonded together, or joined with a piano hinge. (The latter method is used on RVs and is well documented on www.VansAirForce.net.) Likewise, there are various methods of attaching the fairing to the gear, including permanent bonding, piano hinges, clamps, and screws/nut plates.

BID tape placed on the leading edge.

Strips of peel ply are wrapped over the leading edge and taped in place with foil tape to hold the BID tape securely in place.

An epoxy/micro shim pad ensures a tight fit of the fairing to the attach bracket. It also provides extra material thickness to allow for countersinking the attach screws.

The gear fairing attach bracket is simply two nut plates bonded to the gear leg.

For those interested in reading further about 2-D fairing design, the following is a good paper: “Design of a 2-D Fairing for a Wind Turbine Tower” by O’Connor and Loth. The books referenced in the drag calculations are Theory of Wing Sections by Abbott and von Doenhoff, and Chapter 6 of Hoerner’s Fluid-Dynamic Drag.

Next month we’ll continue the discussion to look at design and construction of wheel pants.