December 13, 1998 GAMIjectors for Lycomings

Email this article |Print this article

by	Carl Marbach

About the Author ... Carl Marbach is a co-founder of AVweb and its Publisher. Carl is a 4,000-hour pilot, and commutes between homes in Boca Raton, Florida, and Aspen, Colorado, in his 1978 Aerostar 601P. Carl was the founder and CEO of Professional Press which published five computer magazines including DEC Professional. After Professional Press was acquired by a venture capital firm, Carl founded Internetwork Publishing Corporation and now devotes himself to the emerging field of electronic publishing via the Internet. A lifelong resident of Philadelphia until 1994, Carl and his wife Helen now live in Boca Raton, Florida.

The time had finally come to replace the engines on my 1979 Piper Aerostar. N6069N's left engine had gone past its 1800-hour published TBO, and the right one was rapidly approaching that number. Both engines were running as well as they ever did, and AVweb editor Mike Busch kept advising me to keep flying until the engines started showing signs of being tired. But, I had more than just engine time in mind when I decided to take the plunge.

The Aerostar Corporation, now located in Coeur d'Alene, Idaho, had developed an STC to convert the standard Aerostar 601P engines (290 HP turbonormalized Lycoming IO-540-S1A5s) to 350 HP turbo-boosted Lycoming TIO-540-U2As. My airplane didn't have stock engines exactly — I'd installed a shorter prop that increased RPM redline to 2700, raising takeoff power to 300 HP — but the new STC still offered me a total increase of 50 HP a side. The new engines promised quite a performance boost to an already fast airplane.

While Aerostars are fast airplanes, they're not exactly known for their short-field takeoff capability. As my airplane approached max gross takeoff weight, it became quite a ground lover on takeoff. I lighten the nosewheel at 85 knots and rotate at 95 knots, then wait what seems forever to reach best single engine rate of climb speed (blue line) of 117 knots. Initial climb, particularly at max weight on a warm day, is poor enough to make Aerostar pilots wonder what it would be like with only one engine turning. I hope I never find out while I am low and slow.

I decided that adding 50 HP a side should make quite a difference, and decided not to wait any longer.

Pros and cons of the new engines

The new engines did make a big difference. With 700 HP now powering 69N, it spends less time on the runway, and reaches blue line speeds almost as soon as the gear is in the wells. Initial climb is always better than 1000 fpm and improves as the airplane climbs until above 16,000 feet where the rate of climb starts decreasing, reaching about 600-700 fpm at FL250. Pretty impressive. The takeoff and climb characteristics of the plane have been transformed from mediocre to spectacular and safety has been increased substantially. While an engine failure would not be good early after liftoff, it is now going to be more manageable than it was before the engine change.

Cruise speed with the new engines ranges from fast to blistering, dependending on what power setting I choose. At altitude, 75% produces 240-250 knots, 65% gives 220-230 knots, and 55% yields 210-220 knots (all depending on aircraft weight and flight conditions, of course).

Now for the bad news: the fuel burn is out of sight!

My 300 HP engines consumed 16.5-17.0 GPH per side at about 65% power, with Turbine Inlet Temperatures (TITs) running at a cool 1550°F (100°F below redline). Cylinder head temps also ran cool at 350°F.

With the new engines, things were very different: At 75% power, they gobbled an astonishing 24 GPH per side with the TITs running jut over 1600°F and the cylinder heads running near 400°F degrees. While these temps were below redline, I couldn't imagine they'd be good for the longevity of my expensive new engines. Reducing power to 65% brought CHTs down slightly to the high 300's and the fuel flows to 22 GPH per side. Still hot and still not exactly economical. Trying the "economy setting" of 55% reduced the head temperatures to about 375°F, but the TITs remained around 1600°F and the fuel flows were still about 20.5 GPH per side, nearly 8 GPH more than the old engines used at 65%.

It was pretty clear that the fuel efficiency of the new engines was substantially worse than the old ones. With my new engines, 55% of 350 HP is 192 HP. With the old engines, 65% of 300 HP is 195 HP, about the same horsepower. Sure enough, I was getting the same airspeeds with the new engines at 55% as with the old ones at 65%. But the new engines were burning 8 GPH more! The 8 GPH was bad news for two reasons: it increased the operating cost, and more importantly it reduced my range significantly. On a four-hour trip I was using 32 more gallons than before significantly reducing my range. Look at the numbers:

So after 4:00 hrs of flying, I used to have a 2-hour reserve, now (even using a miserly 55% cruise setting) I'm down to less than one hour. My old five-hour-with-reserves airplane with had become a four-hour-with-reserves airplane. I'd lost about one hour of endurance, which at 220 knots is 220 nm less range than before. I was not happy with these numbers.

GAMIjectors to the rescue!

General Aviation Modifications, Inc. of Ada, Oklahoma has been making balanced fuel injectors for Continental engines for a few years (see Mike Busch's review article). Recently, they announced FAA approval of GAMIjectors for Lycoming engines, and their STC included the new TIO-540-U2As in my Aerostar. I knew about George Braly's team in Oklahoma and what they have done for Continental-powered airplanes, and I hoped their new Lycoming GAMIjectors might help my fuel consumption crisis.

Prior to flying to Ada, I visited GAMIjectors web site and downloaded the leaning test sheet. In order to balance the fuel flows to each of your cylinders, GAMI needs to know what the flows are to each cylinder and at what flow each cylinder reaches peak EGT. GAMI's goal is to tweak your injectors so that each cylinder reaches peak EGT at the same time (i.e., at the same mixture-control setting).

GAMI's leaning test procedure doesn't ask for specific EGT temperature readings, because they don't matter. All GAMI needs to know is the fuel flow at which each individual cylinder reach peak EGT. (You must have probe-per-cylinder EGT instrumentation such as a GEM or JPI 700 in order to obtain this information.) Specific temperatures will depend on EGT probe location, instrument calibration and internal cylinder differences; but peak EGT signifies that the "stoichiometric" fuel/air ratio has been reached. Cooler-than-peak on the rich side indicates excessive fuel, and cooler-than-peak on the lean side indicates excessive air. Conventional wisdom (and some engine manuals) advise cruising at 25°F to 75°F rich of peak, using the extra fuel in the mixture to provide additional cooling.

Lycomings are different

Due to the configuration of the runner-and-riser induction system used by most Continental engines, the individual cylinder air/fuel ratios are predictably different — the rearmost cylinders generally run lean and the frontmost ones run rich. GAMIjectors solve this problem by providing flow-balancing injectors that use slightly larger orifices in the rear cylinders and slightly smaller ones in the front cylinders. The result is an engine in which all cylinders reach peak EGT at the same mixture control setting. Because the fuel imbalance in Continentals engines is so predictable, this "standard" approach works almost all the time with Continentals. "When it doesn't work, we usually find something else wrong with the induction system like intake leaks, exhaust leaks, and so forth," George Braly told me.

But Lycoming engines do not have the same predictable imbalances. The main reason Lycoming engine cylinders do not reach peak at the same time is the sloppy tolerances in their fuel injectors. Consequently, GAMI has to custom-balance the injectors for each individual Lycoming engine.

I supplied GAMI with the engine fuel flows at which each cylinder on each of my engines peaked. It varied considerably, but "was pretty good" as Lycomings go, according to GAMI.

GAMI's goal was to provide me with a custom set of fuel injectors that would make each cylinder peak at the same time. The first step was for GAMI to remove my stock Lycoming fuel injectors and measure their actual fuel flow, using a sophisticated computerized flow bench. This step told GAMI engineers exactly what was happening in my engine to cause the cylinders to reach peak at different times.

Then, using a proprietary computer program based on my leaning results and my injector measurements, GAMI selected a custom-matched set of fuel injectors for each cylinder on each engine. These were flow-checked on GAMI's flow bench and final fine adjustments were made (by hand reaming the injectors to the precise flow specified by the program). Then the new injectors were installed on my engines, and the airplane was test-flown to verify the results.

Since each injector is custom-sized to its particular cylinder, a data plate is installed on each cylinder to indicate which injector goes where. After routine injector cleaning, it's vital that each injector be returned to its proper cylinder — they're no longer interchangeable.

But does it help?

So now I have a set of engines that have almost perfectly balanced fuel/air ratios for every cylinder. How does that help?

Actually, the GAMIjectors don't help much so long as the same old rich-of-peak leaning procedures are used. The real benefit of the new injectors is that they allow me to lean the engines far more aggressively than before, and run them in the lean-of-peak regime. Before installing the GAMIjectors, I didn't have that option because the engines would start running rough if I tried leaning them that far.

Here's why. When an engine is operating on the rich side of peak, small variations in mixture don't have much effect on power output. If some cylinders are running 75°F rich of peak and others are running 25°F rich of peak, they're still producing about the same horsepower and so the engine runs smoothly. Even with uneven fuel flows, most engines won't start running rough so long as you stay on the rich side of peak.

But what happens if we move to the lean side of peak EGT? As you can see from the accompanying graph, both horsepower and CHT starts dropping off fairly rapidly as the mixture gets leaner. In fact, on the lean side of peak, horsepower varies almost proportionally with fuel flow. (As a rule of thumb, HP is equal to about 14 times GPH. Stash this tidbit away for later...we'll come back to it.)

If you lean a "stock" Lycoming engine lean of peak, it starts running rough. Why? Because some cylinders are running richer (say 25°F lean of peak) and others are running leaner (say 75°F lean of peak), so their power outputs are noticeably different. These horsepower imbalances between cylinders are perceived as a "rough-running engine." But with properly flow-balanced GAMIjectors, the cylinders run at near-identical mixtures, and put out near-identical horsepower. The net result is that you can lean the engine well into the lean-of-peak regime without any onset of roughness.

We're now ready for the unconventional wisdom: lean-of-peak operation. If we lean to about 50°F lean of peak, CHTs run some 25°F cooler than they do at peak. The EGTs at 50°F lean of peak are the same as they are at 50°F rich of peak, of course. The horsepower produced at 50°F lean of peak is somewhat lower, so we compensate for that by bringing up the manifold pressure by 2 or 3 inches restore the lost power. The result is that the engine puts out the same horsepower at lower fuel flow and cooler CHTs. Which sounds like a good thing, doesn't it?

Is this legal? Is it wise? Will it hurt anything? GAMI has found this excerpt from a Lycoming 540 engine manual:

Lean the mixture until EGT peaks and continue to lean until [the EGT] drops 25 to 50 degrees on the gauge. Flying on the lean side is permissible if extended range and cooler engines are desired. Operation at peak EGT is only recommended for mixture control adjustments or when induction icing occurs. [The] amount of temperature drop can be determined by resultant fuel consumption and engine smoothness.

When operating on the lean side of the power curve, the pilot may observe that airspeed and power are less. If you desire to regain lost airspeed and continue to fly on the lean side of the curve, two steps are important. If sufficient throttle is available at the lower altitudes; first add two inches of manifold pressure to the standard cruise setting and then lean 25 to 50 degrees (lean of peak). Occasionally, some pilots prefer to fly on the rich side of the power curve; this is permissible. Adjust the mixture until EGT peaks and then enrich mixture until you get 25 to 50 degree drop on the EGT gauge. Acceptable continuous (cylinder) head temperature is an important reference here.

To sum up: according to the chart and Lycoming's own explanation quoted above, running 25°F to 50°F lean of peak EGT will result in lower CHTs, lower fuel consumption and slightly reduced power (which can be restored by increasing manifold pressure 2-3 inches). Because the power curve drops sharply on the lean side of peak, the engine will run more roughly lean of peak unless all the cylinders are on the same part of the curve (and thus producing the same horsepower). This will only happen if the fuel/air ratios are closely matched for all the cylinders.

That's why you need the matched GAMIjectors.

Actual results

In my Aerostar with the 350 HP engines equipped with GAMIjectors, I have found that I can fly at about 75°F lean of peak without losing too much power. Peak TIT depends on altitude — the higher I go, the higher the peak TIT at any given power (as expected). The limiting factor is the 1650°F TIT limit, so at higher altitudes I will have to run at reduced power settings in order to keep the TITs below redline.

For example, at FL180, 65% power (2200 RPM and 32" MP) burns 21 GPH per side and yields a TIT of about 1600°F with CHTs about 360°F if I lean the old way (rich of peak). With the GAMIjectors, I can continue leaning past peak EGT to a miserly 16 GPH without any engine roughness, which yields TITs of about 1640°F and CHTs about 335°F. Adding a couple of inches to the manifold pressure (to 34") to restore the lost power doesn't seem to affect these temps. Note that the CHTs are extremely cool and the fuel flows far lower than before, but the TITs are running close to the limit and are definitely the limiting factor.

Where'd that 16 GPH figure come from? Well, 65% of 350 HP is about 227 HP. Using the magic HP=14*GPH formula I mentioned earlier, fuel flow at 227 HP should be 227 divided by 14 or 16.2 GPH.

At lower altitudes, the TITs are not as close to 1650°F, and the CHTs run cooler yet! At higher altitudes, I have to reduce cruise power to 55% or 60% to keep the TITs below redline. Reducing RPM to 2100 at the lower power settings helps lower the TITs.

When I mentioned to George Braly of GAMI that I was uncomfortable with such high TITs, George pointed out that since my airplane has an all-Inconel exhaust system (which can handle 2100°F temps), the only real disadvantage of running near-redline would be that it might shorten the life of my turbochargers somewhat (due to accelerated turbine wheel "blade stretch" at those high temperatures). But he pointed out that with the 10 GPH fuel saving per hour, I would save $20/hr or $20,000 per 1,000 hours of operation. With that kind of savings, maybe I could put up with slightly reduced turbocharger life. At the same time, the dramatically lower CHTs that result from lean-of-peak operation should produce longer cylinder life. Since a turbocharger overhaul costs only a small fraction of what a top overhaul does, this seems like a pretty good tradeoff.

So, I have gotten my range back, reduced cruise fuel flows, and lowered cruise CHTs all at the expense of slightly higher TITs and some one-time expense for the GAMIjectors. All good tradeoffs I think. The engines run smoothly at 75°F lean of peak and my speeds are about 10 knots below book speeds for the power settings I have chosen. I now have choices:

• run at higher power and higher speeds at the cost of high fuel flows and high CHTs,
• run at somewhat reduced power and lower all temps and fuel flows a little, or
• run at reduced power lean of peak and reduce fuel flows significantly, increase range, lower CHTs all at the expense of higher TITs.

The lean-of-peak controversy

Does everyone agree that running the engines this way is a good thing? No. There are many (including certain engine overhaul shops) who are dead set against running on the lean side of peak and will blame any engine problems on that practice.

One school of thought is that on the lean side of peak there is less fuel and more oxygen, and in theory that could lead to more oxidation, possibly of the cylinder walls or valve assemblies, and contribute to increased wear. GAMI says it has seen no evidence whatsoever to validate this theory, and that the effect of any increase in oxygen on cylinders and valves would be more than offset by reduced operating temperatures.

Others insist that since the practice is not specifically spelled out in the Lycoming operating manual for their particular engine model, they must conclude that running lean of peak has not been endorsed by the manufacturer. But if temperature limits are respected and power settings are such that detonation is not a problem, it's hard to see how Lycoming would object to leaning aggressively.

GAMIjectors has run their own engines this way for several years and many hundreds of hours, and their customers have done the same for tens of thousands of hours (cumulatively) with the same excellent results. "The engines look better running lean of peak than we would normally expect them to look using more traditional methods of leaning," GAMI's service department told us.

When I mentioned that most pilots would prefer not to be guinea pigs for a new method of engine management, GAMI pointed out that back in the days of piston-powered airliners, the Wright Aeronautical Division (who produced the turbo-compound piston engines that powered the DC-6, DC-7 and the famous Lockheed Constellations) recommended that these engines be operated on the lean side of peak to lower cylinder head temperatures, extend the range of the aircraft and to avoid the possibility of detonation at high power settings. Thousands of these planes flew for millions of hours using lean-of-peak operation. So, says GAMI, when we fly in this fashion, we aren't pioneers — we're historians.