HORSES THAT EAT...

BUT By Ben Ellison, President ELLISON FLUID SYSTEMS, INC. 350 Airport Way Renton, WA 98055 For readers who are equestrain enthusiasts, I'll hasten to point out that the title refers to the ponies feeding on 100 octane rather than hay. In fairness to these gas guzzling critters, I should admit that they are all working; it's just that they don't all work at pulling the airplane. Many of them work at overcoming various types of mechanical friction as well as aerodynamic and thermal losses inside the engine. This article is aimed at reducing the work these non-propulsive ponies do, thereby allowing the fuel they've been using to remain in the tank. Some of the ideas presented herein are impractical to implement, given the 40 year old technology underlying current engines and accessories. They are mentioned, however, in hopes that the homebuilt movement will be inspired to develop some alternatives that will allow our aircraft engines to perform as well as the ones on our motorcycles and lawn mowers. First we will discuss how mixture, timing and fuel distribution affect the conversion of gasoline into heat energy, then identify several times other than the propeller that absorb the power produced by the pistons. 16 AUGUST 1985

111

NT PULL

Mixture, Timing and Fuel Distribution The effect that ignition timing, mixture setting and fuel distribution have on overall engine fuel economy is related to the reduction in flame front propagation speed that occurs when the fuel/air ratio is leaned past peak power. If we were to observe cylinder head and exhaust gas temperatures during changes in mixture (assuming an engine with perfect fuel distribution), we would see the trends illustrated in Figure 1. As the engine is leaned from full rich, the fuel/air ratio on the horizontal axis of Figure 1 will move toward lower values to the left. In going from full rich to a fuel/air ratio of around .080, we will observe a rise in EGT (Figure 1A), CHT (Figure IB),

and probably oil temperature. Additionally, we will see signs of increased power, such as increased RPM (if using a fixed pitch prop) and an increase in airspeed. Power will peak at or near this mixture setting in the wide peak shown in Figure 1C. As leaning continues and power begins dropping, EGT and cylinder head temperature will trend upward and peak at a fuel/air ratio of about .068. This fuel/air ratio is termed "stoichiometric" and is the mixture of fuel and air that provides the chemically correct amount of oxygen to burn all of the hydrogen and carbon in the fuel.

As the mixture is leaned past peak power (fuel/air ratio .080), the flame propagation speed becomes slower and slower with the combustion process terminating later and later in the power stroke of the piston. As leaning continues, the combustion rate becomes so slow that combustion is still in progress when the exhaust valve opens. When this happens, the tail end of the combustion process occurs in the exhaust pipe, bathing the EGT probe in flame. The EGT reading quite naturally turns upward again. Combustion in the exhaust pipe represents heat energy not being released in the cylinder, causing a reduction in power output as well as a drop in the cylinder head temperature. As leaning continues still further, flame remaining in the cylinder when the intake valve opens ignites the incoming charge and propagates upstream through the intake manifold in what we term a "backfire". Assuming again that the engine has perfect fuel distribution, the retarded flame propagation speed is the parameter which limits the engine's tolerance for leaning at cruise power settings. The fire can be contained in the cylinder at leaner mixtures by making more time available for the combustion process to take place prior to the opening of the exhaust valve. This can be accomplished by advancing the spark so that combustion is initiated earlier in the

BEST ECONOMY RANGE

MAX. POWER RANGE FULL RICH

"• A

.06

.07 .08 .09 FUEL/AIR RATIO

.10

.11

Figure 1

compression stroke. Because present magnetos utilize fixed timing, they must be set at a spark angle which is a compromise giving adequate power, economy and detonation margin over a wide range o.f mixture and power settings. Hopefully, the homebuilt movement will provide us with a lightweight ignition system allowing pilot adjustment of spark timing for optimizing cruise fuel economy. MAPI Engines of Eloy, AZ has developed a solid state ignition system for aircraft engines which is a step in this direction. The assumption of uniform fuel distribution which preceded the above discussion seldom exists in the real world. It is, therefore, unlikely that the chain of events outlined above occurs in all cylinders simultaneously. There are, however, things that can be done operationally to improve fuel distribution. If the engine installation utilizes a float type carburetor, fuel distribution can be much improved by exchanging it for a pressure carburetor (Bendix PS-5), fuel injection system (Bendix RSA-5), or, as we would recommend, an Ellison Throttle Body Injector. In developing the Throttle Body Injector, we have found the best indicator of good fuel distribution to be the engine's tolerance for mixture leaning below peak power. Although EGT uniformity will sometimes give an indication of fuel distribution, other variables such as spark plug gap, EGT probe location, cy-

Figure 2

linder compression, cylinder cooling, and magneto timing often have more influence on EGT profiles than mixture uniformity. Using this characteristic as an indicator, we have found that the Bendix TSA-5 injector, the Bendix PS-5 pressure carburetor, and the Ellison Throttle Body Injector all give excellent fuel distribution. On the other hand, we have found that carburetors utilizing a butterfly valve downstream of the metering jet can seldom be leaned past peak power when operating at part throttle cruise. This early roughness is caused by the butterfly acting as a turning vane, causing the dense fuel spray pattern to be deflected toward certain intake pipes and away from others. Figure 2 illustrates the part throttle biasing effect of the butterfly valve observed in the Ellison Fluid Systems, Inc. laboratory. At full throttle, this biasing influence of the butterfly is reduced, as shown in Figure 3, and the engine will tolerate a leaner mixture. Figure 4 shows that the Ellison Throttle Body Injector is free of this biasing effect even at part throttle. With a conventional float carburetor, fuel distribution may be enhanced by cruising at full throttle and using the aircraft's altitude to limit the engine to the desired power. For example, if a pilot desires to cruise at 60% power with a Lycoming O-320 engine, he may choose to fly at 3000 ft. while throttling

the engine to 23 inches of manifold pressure. Alternatively, he may cruise at 8000 ft. and operate at wide open throttle. In both cases, the power developed by the engine is the same except that at the higher-altitude, wide-open-throttle condition, the biasing effect of the butterfly plate is absent, allowing leaning to a more economical mixture setting. An additional benefit of wide open throttle operation is a significant reduction in induction pumping losses, discussed later in this article. Another operational practice that will improve the fuel distribution of float type carburetors is cruising with carburetor heat on. Heated air greatly accelerates the evaporation of fuel, thus minimizing the liquid fuel fallout occurring at the bends and corners of the induction system. Since heated air causes a slight power loss, the pilot increases power by opening the throttle (which again saves on pumping losses) or descending to a slightly lower altitude if the throttle is already wide open. It's important to understand that the use of carburetor heat during take-off or climb will reduce engine power, and perhaps induce destructive detonation if used in conjunction with excessively lean mixture settings. At power settings of 75% and below, however, proper leaning techniques along with the use of carb heat should have no adverse effect on engine operation as long as SPORT AVIATION 17

Figure 4

Figure 3

other engine limits such as cylinder head and oil temperatures are respected. These two techniques for enhancing fuel distribution, i.e., full throttle cruise at altitude and heated induction air, provide the best fuel distribution, and therefore the best fuel economy, that a conventional carburetor is capable of producing. Used in combination, these two techniques will sometimes allow leaning 25 to 50 RPM on the lean side of peak power prior to encountering engine roughness.

The Throttle Body Injector, on the other hand, will permit smooth opera-

tion 200-300 RPM on the lean side of

peak power, although such extreme leaning usually results in unacceptably low power output. All of the above suggestions are intended to allow the engine to operate smoothly and efficiently on leaner than

normal fuel-air mixtures. The combustion of lean mixtures results in exhaust products that are chem-

ically very rich in oxygen. This high temperature oxidizing environment can

erode the inside of mild steel exhaust

pipes if they are not protected with high temperature paint. For this reason, it is often cost-effective to use stainless steel for all exhaust system components. An additional chemical benefit of very lean mixtures is that their exhaust products contain less carbon monoxide (CO) than richer mixtures and are therefore less toxic. Having dealt with the combustion process in which fuel is converted into heat energy, we will now examine several losses occurring in the engine which absorb much of this energy before it reaches the propeller. 18 AUGUST 1985

Pumping Losses Pumping losses represent power that

the engine must expend internally to draw intake air past the resistance of

the throttle valve during the intake

stroke.

Any time the engine is operating at a condition in which the pressure in the

intake manifold is lower than the engine crankcase, this pressure difference between the cylinder and crankcase ends of the piston (illustrated in Figure 5) generates a force which opposes piston

movement on the intake stroke. This pressure force is greatest at low throttle settings, where manifold pressure is very low and least at wide open throttle because manifold pressure is nearly the same as the pressure in the crankcase (crankcase pressure is maintained near ambient by the crankcase breather).

This adverse pressure force can be more easily understood using the numbers in the following example: The surface area of a Lycoming O320 piston is 20.6 square inches. When operating at 1000 ft. and 75% power, the manifold pressure is 25 in. Hg. The pressure in the crankcase is equal to the ambient pressure at 1000 ft. or 29.00 in. Hg. The pressure difference

between the two sides of the piston is

29 - 25 = 4 in. Hg or 1.97 psi. The actual pressure gradient would vary between 1.97 psi and zero over the entire intake stroke, so for this calculation we assume an average value of 1.0 psi. Multiplying this pressure by the 20.6 sq. in. piston area yields a force of 20.6 pounds. This force exists on each piston during its intake stroke which is 3.875 inches long. If the engine is cruising at 2500 RPM, the adverse pressure force absorbs 12.1 hp.* These 12.1

non-propulsive horses consume 5.4

Ibs. of fuel per hour. * This analysis ignores a small power benefit occurring during the first portion of the compression stroke. This 12.1 HP loss can be nearly eliminated by cruising the engine at full throttle at an altitude that limits the engine

to the desired power as described ear-

lier. Operation at full throttle causes the pressure in the intake manifold to equal

the crankcase pressure. The adverse pressure gradient opposing the piston on the intake stroke is gone and the pumping loss is essentially zero.

This loss is illustrated graphically by the two indicator diagrams below which plot cylinder pressure versus piston

travel as the piston goes through its four

stroke cycle. Figure 6 represents part throttle operation while Figure 7 represents wide open throttle. Examining Figure 6, compression of the fuel/air mixture begins at point A and continues until the piston reaches top dead center (TDC) at point B. During the compression stroke, combustion is initiated by the spark which occurs at point S. The pressure increase following ignition is naturally quite steep owing to the pressure rise of combustion. The combustion pressure rise continues through the first part of the power stroke, then cylinder pressure falls due to increasing cylinder volume until reaching point E. At E, with the cylinder still substantially above atmospheric pressure, the exhaust valve opens, causing a sudden blow-down to point A at bottom dead center (BDC). The cylinder is exhausted by the advancing piston from point A to C, at which time the exhaust valve closes and the intake valve opens at or near TDC. As the piston moves from TDC to BDC along the intake stroke (C to A), the pressure in

valve opens before the piston reaches

bottom dead center (BDC) at E on Figure 6 and 7 and the pressure in the cylinder is still considerably above atmos-

pheric. The loud acoustic explosion that

MANIFOLD PRESSURE 25 IN. Hg. (12.3 PSI) CRANKCASE PRESSURE 29 IN. Hg. (14.2 PSI)

Figure 5

the cylinder is roughly equal to manifold pressure, which is assumed to be 25 in. Hg. in the above example. The area enclosed in the compression power loop represents work done on the piston by the expanding combustion gases, while the area enclosed in the exhaust-intake loop (cross-hatched area) represents the mechanical work that the engine must do to breathe. The power output to the crankshaft is equal to the open area minus the cross-hatched area. This considerable energy expended to induct mixture through a partly closed throttle valve (represented by the crosshatched area) is something that anyone suffering from a bad cold or chronic

asthma can understand from personal experience. Figure 7 illustrates full throttle operation of an engine with essentially no pumping loss. Keep in mind that eliminating this loss through wide open throttle operation will not make more power available, but rather will save the fuel those pumping ponies were consuming.

Slowdown Losses Slowdown losses occurring in a typical piston engine are illustrated by the area enclosed by the dotted line on the indicator diagrams of Figure 6 and 7. They originate because the exhaust

occurs when the valve opens is the sonic boom from the supercritical flow past the exhaust valve. This high residual pressure in the cylinder has the potential for doing a lot more mechanical work before its heat energy is dissipated. This work can be extracted in a turbo supercharger, a power recovery turbine, or some other form of secondary expansion device. Brown, Boveri & Co. of Baden, Switzerland (Ref. 1) makes an ingenious device that uses high pressure slugs of exhaust gas as pistons acting in small tubular chambers to compress fresh induction air to a high level of boost pressure. This device, called a Comprex(R) PressureWave Supercharger, is just waiting to be discovered by innovative homebuilders. Without incurring the added complexity and expense of a secondary expansion device, wasted exhaust energy can be used most simply by turning it rearward to get the benefit of a small amount of jet thrust. Using a little more sophistication, the exhaust can be expanded in an ejector pump built into the cooling system to pump cooling air. When designed properly, such a system can provide negative cooling drag which means that the exhaust/cooling air combination provides a small amount of propulsive thrust. Ref. 2 describes how exhaust pulses can interact with cooling air in a divergent nozzle behaving like a piston pump. Since the large amount of energy remaining in the exhaust stream has al-

Pambient -• Pmanifold

TDC

PISTON STROKE

Figure 6

BDC

TDC

PISTON STROKE

BDC

Figure 7 SPORT AVIATION 19

FRICTION 1HORSEF»OWER FORVy\RIOUS OISPUICEMEIY T A V C 0LYC OMING ENGIN ES

jjjlijji

-"t:rt-;:5 irt:?!^: rd.r;ji:t

60——

Friction Losses '../a: 4; TT^ITT:

SOrttr1

0-360

:::::tt4i-

—t-*-, . »_(_)-4.

] i§ 4 0 —— i

:rr^ip?

§P

^U-320 : i::|j: iiiljjj

••3**

i' ';;»u;

* • § 30H5? — •;;—

•H^

0-235

•

§

HH:.

= 20^=

1£

10 —— ff^T"

~-s^

i*^

:-^: —————, ——

__; _: _LI ___ . _

. . . . -

20

22

24 2€ 32 28 30 C:RANKSSHAFTS>PEED-I3PM x1 00

34

Figure 8

LONG EZ COWL

COOLING AIR IN

OIL COOLER HEAT LOAD LYCOMING O-320

100 90

7

iiiiliiil

1 !1 1 1 1 1 1 i

!1 11111111f 1rrr\t i

::;::::::

• i ;;::::::

70

11

cc

o

I

cc

60

a.

i 50 40

30

fe

iiptr

*. .......

ill

^0

22

24 26 28 30 32 CRANKSHAFT SPEED - RPM H- 100

Figure 10 20 AUGUST 1985

/•::::::

/:::::::

HI

Figure 8 (Ref. 3) shows the horsepower that is being consumed by friction for several Lycoming engines as a function of engine RPM. These losses include the obvious friction losses of pistons sliding in cylinders, shafts turning in bearings and gear teeth gnashing together, as well as more subtle losses such as the power required to operate non-discretionary accessories like magnetos, prop governors and oil and fuel pumps. Accessories such as alternators, hydraulic and vacuum pumps are treated separately because their use is at the discretion of the pilot. Generally, friction losses in an engine are a function of engine RPM and may be minimized by cruising the engine at the lowest possible RPM and highest manifold pressure consistent with the engine's ability to produce the desired cruise power. In addition to minimizing engine friction, operation at high manifold pressure and low RPM will also minimize induction pumping losses and allow the propeller to operate nearer its peak efficiency. With a fixed pitch propeller, of course, a heavy propeller pitch will adversely affect take-off and climb performance. Although it is a slight simplification, it is fairly safe to assume that half of the friction losses are absorbed as heat by the engine oil system and must be dissipated through the oil cooler. The other half of the friction losses are dissipated as heat through the cylinder walls and cooling fins. Cooling Power Losses

Figure 9

80

ready been paid for in the fuel which was burned in the cylinders, it certainly makes sense to use it before dumping it overboard.

34

Part of the power delivered by the crankshaft is devoted to cooling the engine just as though a water pump and radiator fan were attached. This power is consumed by bringing outside air into the aircraft's cowling, where it is first decelerated to raise its static pressure, then routed through ducts, baffles, cylinder fins and oil coolers before it is

dumped overboard again. The typical cooling system, designed to cool the engine at full power on a hot day at low airspeeds, delivers excessive cooling air when operating at cruise. The ability to turn this excess cooling air off using cowl flaps will significantly reduce cooling power losses. To avoid wasting cylinder cooling air, tight fitting cylinder baffling should be configured as necessary to allow all cylinders to operate at equal cylinder head temperatures. A cowl flap allowing pilot control of cooling air will yield minimum cooling power losses when (Continued on Page 62)

1,

Glider pilot, balloon pilot and ultralight pilot. The FAA has determined that the safety record of these aircraft Operations does not warrant from the medical standpoint a required airman medical certificate. Were it not so, the FAA would long ago have instituted such a requirement in keeping with its statutory air safety mission. I have personally reviewed accident statistics relating to these three pilot categories and can state that no medical justification exists for instituting a requirement for periodic medical examinations. From time to time some well meaning but misguided group, with no supporting data in hand, proposes that the FAA institute periodic medical certificate requirements for one or another of the above three categories. My recommendation to all pilots is to fight any unnecessary extension of regulations and

potential loss of flight freedom for any segment of pilots. It is often more difficult to get a regulation off the books than to put it on. If medical evidence is developed that clearly warrants, in the interest of safety, extending the medical certificate requirement to any of these three categories, I will be at the head of the line promoting such an extension. However, such is not the case. I also recommend that all pilots relate to a regulatory agency in the same fashion that a prudent person relates to a business, an insurance company or a bank, when matters that restrict freedom of flight are being covered. Oral promises and handshakes in the above respect are very risky as a basis for long term understanding. Top level personnel change and adminstration changes occur and the previous promises of departed officials become meaningless. In

addition, the human memory is very fallible. Those matters that don't restrict the freedom of flight, as with those matters that don't involve one's financial security, can be dealt with by a handshake. But I know from long direct involvement in regulatory activities that, for the reasons given above, an oral promise, a handshake and a smile by a Government representative concerning critical regulatory matters, often constitute the prelude to future animosity, adversarial relationships and unnecessary trouble on all sides, all to the detriment of progress in aviation. Freedom . . . promote it or lose it!

HORSES THAT EAT . . . BUT DON'T PULL

ing system sized to provide a maximum of 1.5 Ibs./sec. of cooling airflow. Cowl Flap Open: Aircraft velocity Vo = 292ft./sec. (200 MPH) Cooling air exit velocity Vexit = 90 ft./sec. Cooling airflow Wa = 1.5 Ib./sec.

If the master switch is left off, the electric load is gone, but the friction remains. It is not unusual for some aircraft to provide cabin heat through an electric resistance heater of 300 watts or more. A 300 watt heater representing .4 HP plus a 10 amp electrical load at .162 HP plus .10 HP belt and pulley friction consume .662 HP and .3 Ibs. of fuel each hour. This again represents horses being fed that are not pulling the aircraft. In an all-out efficiency competition, it would pay to operate avionics on a storage battery with the alternator belt removed and to remove the vacuum pump.

(Continued from Page 20)

adjusted to yield the maximum acceptable cylinder head temperature on all cylinders. Airflow through the engine oil cooler also imposes a power loss on the engine. Most oil cooler installations are configured so that operation of the engine cowl flap throttles the oil cooler airflow as well as the cylinder cooling airflow. This arrangement is deficient because when airflow through the cooler provides more cooling capacity than the oil system needs, the oil temperature control valve reacts by bypassing oil around the cooler. The airflow through the cooler, and therefore the drag of the cooler, is the same regardless of the oil cooler's heat load. A much more efficient system provides a separate valve controlling airflow through the cooler. The oil cooler airflow and therefore the cooling drag, can be reduced when maximum cooling capacity is not required. Figure 9 shows the cooling configuration used on Long-EZ N17BE which allows pilot control of both cylinder and oil cooling air. As with cylinder cooling, the power loss from oil cooling is minimum when the oil cooler airflow is adjusted to yield the maximum acceptable oil temperature. In addition, engines that make use of that transfer from oil to induction air, such as the Lycoming engine series, will experience better fuel distribution and vaporization when oil temperature is kept at this maximum acceptable level. The following example illustrates the reduction in cooling power losses in the Ellison Fluid Systems' Long-EZ. Lycoming O-320 engine with a cool62 AUGUST 1985

Cooling Power

= Wa x Vo (Vo Vexit) = (1.5) (292) (202) / (32.2) (550) = 4.94 HP

The equivalent engine shaft horsepower is 4.94 / Prop. Eff. or 4.94 / .75 = 6.58

If you wish to contact the author for additional information, please write to Stanley R. Mohler, M.D., Wright State University, P.O. Box 927, Dayton, OH 45401.

Cooling HP = 6.58 HP

If, under these flight conditions, the engine requires only half as much air to maintain the desired cylinder head temperature, then closing the cowl flap to .75 Ibs./sec. will yield the following cooling power loss: Cowl Flap Closed: Vo =292 ft./sec. Vexit = 90 ft./sec. Wa = .75 Ib./sec. Cooling Power = (.75) (292) (202) / (32.2) (550) = 2.47 HP 2.47/.75 = 3.29 shaft HP This represents a savings of 3.29 HP or 1.48 pounds of fuel per hour in fuel consumption. Discretionary Losses Engine-driven accessories not critical for flight will aid fuel economy if not used. A vacuum pump driving the basic IFR gyro instruments absorbs .85 HP, which consumes .38 Ibs. of fuel per hour. An alternator, at 10 amps output, consumes .162 HP for electrical power and about .10 HP in belt and pulley friction.

Conclusion Although the external aerodynamics of the airplane receive most of our attention when we design and build our dream machines, the thermodynamic and aerodynamic processes occurring inside the engine compartment are of equal importance. Careful control of these processes through proper engine compartment design as well as skillful engine operation will pay handsome dividends in extended aircraft range and reduced fuel bills.

References 1. George Gyarmathy: "How Does the Comprex(R) Pressure-Wave Supercharger Work?", SAE Paper 830234, 1983. 2. W. S. Johnson and T. Yang: "A Mathematical Model for the Prediction of the Induced Flow in a Pulsejet Ejector with Experimental Verification", ASME Paper 68-WA/FE-33, 1968. 3. AVCO Lycoming: "Determination of Installed Horsepower for Lycoming Reciprocating Engines."