Paul Lipps' Lancair 235
Story by Paul Lipps, email@example.com
Photos by Pat Panzera
Editor’s Note: We’ve already read about Paul’s revolutionary propeller in the pages of the February issue of Experimenter, so now we’d like to showcase the balance of his plane.
I purchased my Lancair kit on May 29, 1989. Its first flight occurred exactly 12 years later on May 29, 2001! The kit initially went through three different owners before I obtained it. Not too much had been done on it other than joining the fuselage halves and putting in the stub wing spars, bottom skin, and firewall. My friend Oscar Grassle had built a Lancair 235, so he helped and encouraged me. Eventually he sold me a half-share in his plane. This enabled me to gain a lot of flying experience as well as find various things that I would improve upon.
The Redesign Begins
The Lancair has poor lateral stability, so I added 16-inch tip extensions in place of the original tips, with a leading edge swept back at 50 degrees and inclined upward 30 degrees for added dihedral effect. This was done prior to similar wingtips appearing on the Diamond line of aircraft. Once the tips were completed, I made molds from them (used for clear casting-plastic) in which I cast the strobe tubes—which then became the new wingtips, after I cut off the fiberglass tips 1-1/4 inch inboard.
Paul’s redesigned wingtip (swept back at 50 degrees and inclined upward 30 degrees) incorporates a custom cast (contoured) strobe light and a flush landing light system.
It turns out that the wing’s taper, along with the new tips, is a good approximation of elliptical span loading, making for a high-efficiency wing, which should have an Oswald efficiency factor of about 0.81. At the break in the leading edge where the new tip begins, I was able to install the navigation lights and flashing recognition lights. This flashing circuit is one I designed. An article on the system appeared in the June 12, 2000, issue of Electronic Design magazine and can be found on its website.
A rather conspicuous 1-inch diameter hole on the forward edge of the canopy frame is where Paul brings in cabin air. This particular area on most planes (as well as cars) is known to be a high-pressure area. Paul takes advantage of this to help cool the cockpit.
Fresh Air Inlet
Cabin air is brought in through a 1-inch diameter hole, centered at the base of the windscreen. Incoming fresh air is routed through an expander duct to increase its pressure and slow it down, then into a 109 standard cubic feet per minute centrifugal blower, then into a duct that delivers the air into plastic eyeball vents located on the bottom of the instrument panel centered between the legs of the pilot and passenger. The electric blower has three modes of operation: on, off, and automatic. In the automatic mode a switch is operated when the canopy closes, and when the master switch is turned on, the blower turns on. When the landing gear retracts, the blower shuts off, since now there is sufficient forward speed to maintain flow. This blower certainly aids in comfort when closing that clear canopy on sunny days!
Cabin air exhausts through a 1-inch wide by 2-inch long outlet at the top of the fuselage just aft of the canopy bow. This opening is fitted with a lid hinged at its forward edge to prevent rain and bugs from entering when the plane is sitting on the ground. This duct is also fitted with a 3-inch square muffin fan (like those installed in most desktop computers) at its opening inside the cabin roof. This fan is powered by photovoltaic (solar) cells in the top of the canopy. These cells provide a trickle charge to the battery until there is sufficient sunlight to increase the cell voltage enough to run the fan. The purpose of this fan is to supply a slight flow of air through the cabin when the plane is sitting out in the sunlight with the canopy closed to keep the temperature inside the cabin from getting too hot.
Roll trim is provided by spring loading of the control sticks operated by a rotary knob on the center console. This knob is connected by reduction gears to a small cable drum. A 1/16-inch cable wraps around this drum and passes first through pulleys and then through tubes to bring it down into line with the bottom of the control stick on each side. Ten pound/inch tension springs connect from the cable ends to the bottom of the sticks. It should be understood that a lateral weight unbalance, as from wing fuel or a passenger, must be balanced by a constant force. The spring loading of the aileron provides this constant force and is independent of the plane’s forward speed. As the plane slows down, the aileron deflects more, maintaining the force. This is in opposition to an aileron trim tab, which provides a force proportional to speed-squared. A lateral unbalance with a trim tab requires the trim tab position to be increased as the airplane slows, or decreased as the speed increases.
Pitch trim is accomplished with a trim tab on the left elevator. The lever on the console operates the trim tab through a 1/16-inch wire using a nylon-lined bicycle brake cable housing. This housing is much lighter than a traditional bowden cable and has much less friction due to the nylon liner. The trim tab system is in contrast to the spring-loading trim system supplied with the kit. Unlike the aileron trim system described above, what is desired for pitch trim is a speed sensitive trim, which is what the tab provides. As speed increases, the tab force on the elevator will increase, moving it to raise the nose, thus trying to maintain trim speed. Since the force is proportional to speed-squared, a 1 percent change in speed will cause a 2 percent change in force on the elevator, thus showing a trim force gain to try to maintain speed equilibrium.
With the cowl removed, one can see several modifications. The most obvious is the cooling system ducting instead of the traditional baffles. Also quite noticeable are the coil packs on top of the case. Not so obvious but still visible is the harmonic balancer attached to the flywheel and the electronic ignition modules on the firewall.
I made a carbon fiber oil pan for the Lycoming O-235 engine, similar to the original aluminum casting except that where the original pan has a somewhat triangular-shaped rear portion, I have mounted my carburetor. This keeps the carb up higher than would the mount on the bottom of the pan, keeping the lower cowling line flatter and letting it blend smoothly into the fuselage. The carb feeds a fiberglass manifold that turns 90 degrees forward then branches off to each side, and then these branches in turn divide to go to the cylinders.
The ducts are kept high in the pan (out of the oil sump) to minimize heating of the induction air by the oil and so keep the volumetric efficiency high. The primer enters the turn just above the carburetor. It comes in through two jets that angle slightly outward to strike the crotch on each side where the front and rear cylinder ducts divide. This hopefully gives an even splash of fuel into all cylinder ducts.
Paul’s cooling air system. No scale.
To keep cooling drag to a minimum, I use the exhaust energy to augment the flow of the cooling air. I constructed a special, patent-pending, exhaust duct that takes the engine exhaust, spreads it out, and injects it into a box immediately below each cylinder. Here the exhaust gases and cooling air mix and flow out through a duct that exits the bottom of the cowling right near the firewall. There is one duct on each side that takes the combined flow from the front and rear cylinders. The duct, which is 5 inches wide by 2-1/2 inches high, is fitted with an outlet flap that can be opened to 3 inches high or closed down to 1 inch. These flaps are controlled by a “T” handle on the center of the instrument panel.
Paul’s panel. Note the chrome “T” handle under the transponder. This opens and closes the augmenter flaps.
During construction, Paul took this photo showing what’s inside the plenum boxes one sees when the cowl is removed.
Each cylinder is wrapped with carbon fiber reinforced with high-temperature epoxy, from the push-rod tubes around the fins and into the exhaust box. The cooling air inlets on the front of the cowling go into diverging ducts that then split and feed the two cylinders on each side. This cooling-exhaust flow is totally sealed off from the inside of the cowling. All air that enters the inlets passes through the cylinders and exits the duct. Because the exhaust is so effective in promoting the flow of the cooling air, and because all the cooling air is kept in intimate contact with the cylinder fins, it is possible to reduce the inlet ducts to only 1.5 inches by 4 inches for a total cooling area of 12 square inches (six per side). For my approximately 123-hp engine, this is 0.097 square inches per hp. This defies the long-standing rule of thumb of 0.35 square inches per hp that would call for my inlets to be 43 square inches!
The ducts at the bottom of the front two cylinders are tapped into to provide heated air for carb heat. One advantage of this is that if the engine were to stop due to carb ice, the cylinders would not cool off as rapidly as does the exhaust muff system, so heat would be available for restart.
Note how close all three inlets are to the spinner. This is contrary to most designs, but Paul’s propeller allows for this highly efficient placement. The opening under the spinner is the second-generation induction inlet. The current and third generation is described in the “Induction” chapter below.
The cooling inlets at the front of the cowling are kept close to the spinner (again, against conventional thinking) so that the inner surface of the inlet is in plane with the spinner. In this way, the higher velocity air displaced by the spinner is injected into the inlet aperture. And with installation of the ELIPPSE propeller—one that actually makes thrust all the way to the root—the system is complete.
The tiny rectangular inlet just aft of the spinner supplies enough fresh air to adequately cool the alternator. The snorkel below the spinner has been replaced twice since this photo was shot, most recently with a modified NACA submerged inlet to supply the induction system.
The oil cooler inlet shown on top of the cowl in the photo below feeds the composite plenum shown in the above photo.
The oil cooler is mounted to the firewall on the passenger side. It is fed from a streamlined external duct on top of the cowling (see above photo). This inlet measures 1 inch tall by 2 inches wide. With an immediate 90-degree downward turn, the cooling air enters an expander duct before hitting the oil cooler fins. The heated outlet air continues on into a contracting duct that then turns 90 degrees toward the rear and exits the cowling just above the cooling-exhaust duct and does not disrupt the smooth flow of air over the fuselage.
The oil cooler outlet duct is fitted with a cockpit-controlled flapper valve that can divert the heated air into a passage that safely penetrates the firewall to provide cabin heat. This helps to eliminate the carbon monoxide (CO) threat. Since the air in the cowling is not pressurized and the air in the oil cooler outlet is at a higher pressure, there is little possibility for cowling air to enter the cabin through a leaky oil cooler duct, bringing CO with it.
Figure 1. The first of three different induction air inlets used on Paul’s 235.
Figure 2. Version 2.0 had the opening high and tight.
The first iteration used a snorkel like those we are used to seeing on high-performance composite aircraft (see Figure 1 above). The second generation (see Figure 2 above) used a 1-inch by 2-inch carburetor inlet duct, and like the cooling inlets, it was up close to the spinner so that the top of the duct was in plane with the spinner, providing higher velocity flow for the carb. The duct expanded before entering a chamber that housed a 4-inch by 10-inch K&N filter. The carb heat control butterfly is ahead of this filter so that all induction air, heated or otherwise, passes through the filter. A 1-inch by 4-inch filter bypass is provided to allow air to divert under the center of the filter. This valve is controlled by a spring-loaded rod that is pushed shut by the nose gear in the down position. Retracting the nose gear permits the bypass to open, allowing air to pass through the filter and also under it to the outlet. This minimizes induction pressure loss. The reasoning is that most filtering is necessary only near the ground.
The new design includes a redesigned inlet. A curved-divergent-submerged (NACA) inlet in the bottom of the cowling immediately ahead of the carburetor now feeds the system. This was done because the engine showed a tendency to hunt in rpm during a preflight run-up at 1700 rpm. I suspected that this occurred from a resonance in the long induction system, and some simple tests confirmed this. Testing after this installation showed that the rpm is now steady during run-up, but the manifold absolute pressure during flight was about 0.6 inches lower than what I felt should be obtained with the stagnation pressure from forward speed. So I made a forward-facing stagnation inlet with a 2-inch wide by 1-inch high opening and temporarily fastened it over the entrance of the submerged inlet near the opening using No. 2 self-tapping screws—fitting a sheet metal cover over the forward portion of the submerged inlet—essentially mimicking the version 2.0 inlet only farther back. I then went flying and gathered data. These flights were followed within a half-hour by a flight with these mods removed, and guess what? I got exactly the same results! Back to the drawing board.
I made my own capacitance pick-ups for the header tank and the two wing tanks. I designed these to give me a linear change of capacitance versus fuel level in the tanks. These pick-ups are used to vary pulse-width modulators that then drive 10-segment LED indicators. I pick off two of the LED segment driver outputs for the header tank to give me a 40 percent fuel level (about 4.7 gallons) and 100 percent level. These go to an alarm indicator that flashes a very bright LED right-smack ahead of me on the panel. The 100 percent output is tied in with the wing-to-header fuel pump switches. When the header reaches 100 percent and either pump is on, the LED flashes. This helps keep me from over-filling the header.
Below the three fuel level indicators is the throttle quadrant. The large blue knob that would normally be used for prop pitch is used for carb heat in Paul’s plane.
The wing tanks hold 16-17 gallons each, and the header tank holds 11-12 gallons. I usually flight plan for a 6 to 7 gph average fuel consumption on trips. This gives me an endurance of five to six hours with 7 gallons remaining, giving a no-wind range of 1,000 to 1,200 miles. To take advantage of this endurance, it is necessary to have a liquid overboard dispensing system, which I installed and have made much use of on many occasions, as inhabitants of California, Oregon, Washington, Nevada, Utah, and Wyoming can attest!
The engine is equipped with dual Light Speed Engineering (LSE) plasma ignition systems. One system is triggered from an LSE magneto plug-in Hall-effect trigger, the other from a Hall-effect trigger board I made that is mounted on the front of the engine. This gives me two entirely separate and electrically isolated triggers. I have a temporary digital display that can show me timing and absolute manifold pressure from each ignition by means of a rotary selector switch. The unit also has a potentiometer for varying timing of both units +/- 5 degrees.
The LSE Hall-effect crank position sensor located in the stock accessory case, where the magneto once resided.
I decided to install a horn in the right wing behind where the gear retracts. The two horns are pointed forward and down and can be readily heard on the ground when flying over at 1,000 feet AGL with the gear down. Of course, horns for an AIRplane must be AIR horns!
Systems keep evolving on my airplane—I guess that’s why we call it experimental aviation. After building, installing, and flying the ELIPPSE propeller described in the February 2009 issue of Experimenter, I have since modified the tips narrowing them to a point; this brought about an increase of static rpm from about 2190 rpm to about 2210 rpm, but I can’t say that it had a noticeable effect on cruise speed.
Paul Lipps presenting a propeller forum at the 2004 COPPERSTATE Fly-In.
A second and oh so wonderful modification was the addition of a TruTrak heading autopilot and altitude hold. The Lancair has a fairly narrow-chord wing, so the available center of gravity percentage change becomes only a very small change in distance. As a result, just reaching forward and placing my hand on the instrument panel would cause about a 200 fpm rate-of-descent; putting my hands behind my head would cause about the same increase in rate of climb. Reaching across the cabin to grab a chart or bottle from the right seat would also yield undesirable results. I would raise my head and find the plane in a dive to the right. Now I just select my Garmin handheld as my navigation source, set my altitude, and I’m on my way! This is really a benefit when I’m doing speed runs to calibrate my indicated airspeed (IAS) and static ports, or just to see the effect of a change to reduce airframe drag. It becomes possible, under controlled conditions, to see a 1 mph change.
I have a copy of the article from the August 1983 Sport Aviation that gave the details of Dan Somer’s NLF(1)-0215F “Eagle 1” airfoil used on my Lancair. This airfoil was designed to make use of flap reflex, raising the trailing edge up to reduce the airfoil’s camber and the position of its drag bucket relative to the center of lift (CL). From it I was able to approximate the desired reflex based on aircraft weight, true airspeed, and density altitude all of which determines the aircraft’s required CL. I did some experiments with reflex on a trip from Casper, Wyoming, to Santa Maria, California, and was able to affect a 3 mph increase in going from 8 degrees to 6 degrees, then a 3 mph drop from 6 degrees to 4 degrees, and a 4 mph drop from 4 degrees to 2 degrees. The other advantage was that as I reduced reflex, the body angle decreased, giving me more visibility over the nose. I doubt I could have played with the flaps and maintained consistent readings without the help of the TruTrak.
To hedge against a gear-up landing, I designed and installed a circuit that operates from pitot and static pressure that will flash an LED immediately below my gear switch and in front of me if my speed drops below 85 mph IAS while the gear is retracted—to remind me to put the gear down—or should my speed exceed 120 mph IAS with the gear down—to remind me to retract it. Details and instructions for building the system will be published in a future issue of CONTACT! Magazine.
Belleville Washers for Prop Retention
The best that could be said about my use of Belleville washers is for Pat to publish, with the permission of Vance Jacqua’s next of kin, the analysis that he performed. I have been using these washers with my propeller for at least three years, and the thing that I note about them is that in removing them, it always takes several turns before the bolts will spin free. That is opposite the condition that is normally seen with prop bolts in that the torque is usually gone within one-quarter to one-half. The other valuable observation Vance discovered in his analysis is the amount of compression that should be achieved through torque on the bolts. (Also see Vance’s article, “Theory of Rod Bolts and Other Prestressed Bolts” published in the May 2009 issue of EAA’s Experimenter e-Newsletter.)
Paul is not the only one to take advantage of the Belleville washer for prop retention. Cozy Mark IV builder/flier Marc Zeitlin has a successful record with them. His installation is shown above. A propeller departure over the California desert (including structural damage to the wing) had Marc rethinking his prop installation and prodded him to consider using the washers. Photo courtesy Marc Zeitlin.
In the yet-to-be-published article, Vance analyzes this both in terms of the required compression to achieve the necessary friction between the propeller hub and the crankshaft flange, as well as in terms of the crushing strength of various woods. It can be seen from some of these torque numbers that some propeller manufacturers are specifying too great of a torque value; this could be crushing the wood and leaving the prop in a state where it could collapse further under loading, leading to the possibility of losing the torque on the bolts with disastrous results! The Belleville washers, by having repeatable compression loading versus deflection, lend themselves perfectly to the task of providing the best means to ascertain this compression through seeing and measuring that the washers have not been taken to their flat value.
Tom Aberle’s Phantom biplane with Paul’s four-blade prop.
2004 was the banner year for Tom Aberle’s Phantom biplane (see the July 2009 issue of EAA’s Experimenter e-Newsletter); he obtained a qualifying speed increase of 20 mph over 2003’s record-setting speed. This was done with an ELIPPSE three-blade propeller design while averaging 250 rpm less than in 2003, a 7.1 percent decrease in power.
For 2005, a four-blade prop was designed and built that restored the 250 rpm deficit. Because of some problems that were later determined to be in the fuel system, it was decided to go with the previous year’s three-blade.
This same year saw Jeffrey Lo, in his new Miss Gianna biplane, flying an identical three-blade ELIPPSE propeller, with which he took first in qualifying at 237.403 mph. Phantom was being flown by Tom’s partner, Andrew Buehler, and qualified second at 232.71. Miss Gianna took the first heat at 224.043, and Phantom was third at 205.303. The positions changed on the second heat, with Phantom first at 216.018, followed by Miss Gianna at a close second at 214.392. The biplane Gold went to Phantom at 230.827, and Miss Gianna placed second at 220.443. Charlie Greer was flying his Formula One Miss B Haven (not in the biplane class) behind another ELIPPSE prop and qualified third at 249.099, just behind David Hoover’s new Endeavor at 249.815. This prop was not well-matched to Charlie’s drag and power, and he was never able to get the design rpm.
2006 saw Phantom, piloted again by Tom Aberle and with the four-blade prop, set a new qualifying record of 249.106 and winning the Gold race at an astounding, for biplanes, 251.958! Had they been using the course distance adjusted for the faster speeds of the Formula One racers, Phantom’s speed would have been 252.2 mph! Miss Gianna was no slouch either, qualifying at 241.136, and taking second in the Gold at 231.685. Tom was not able to compete in 2007 due to engine problems. But in 2008, Tom set a new record of just over 252 mph, and now his biplane was actually faster than both the T-6s and the F1s!With the 2009 Reno event happening just days after this e-newsletter is published, it will certainly be interesting to follow the affect of the ELIPPSE propeller.