Pelican's Perch #8:
Go Ahead, Abuse Your Engine!
AVweb's John Deakin is asking you to abuse your engine, or so it might seem to many until he sweeps a whole slew of old wives' tales (OWTs) off the cliff. Using digital technology for data collection and simple graphs, John supports his unorthodox engine operating suggestions with data that proves the old ways may actually be worse for your engine. Wrong may well be right! If you've been taught that you must always reduce MP before reducing RPM, you're going to be forced to rethink that notion. If you think you're helping your engine live longer by reducing MP to 25 inches after takeoff, boy are you going to be amazed at how badly you've been abusing your engine. And, that's only for starters.
At a party I attended awhile back, a lady was extolling the virtues of her latest macro-biotic diet, and how she wouldn't drink coffee because it's bad, wouldn't touch salt, just on and on about her rather complex feeding routine. All the while smoking one cigarette after another! What is wrong with this picture?
So it seems with pilots, sometimes.
Many of us will hear something, and adopt it as gospel, without the slightest shred of evidence. On the other hand, present a little data, and eyes glaze over in boredom and indifference. Let's see if I can bore you with some data.
Many of these guys (and most of them are guys!) will read a half-dozen "testimonials" in a magazine ad, and seem to think that if it's in print, it must be true. Question some of these "facts" in their presence, and they get quite intense, even hostile. Defense mode, I suppose.
One of the funniest, to me, is the "Snake Oil Gambit." Also known as "Mechanic in a Can." Some pilots are eager to dump all manner of unknown ("Trade Secret, y'know") gunk in their expensive engines, on the basis of hype and ads alone, and the more costly, the better they like it.
I got involved with one of these several years back. "Microlon" was enjoying some modest success at the time, heavily hyped by its "inventor," one Bill Williams, who made the usual astonishing claims ("20% more power! Double your engine life!") in the AVSIG Forum, on CompuServe. Some of us took a few good-natured shots at his claims, which rapidly grew like Pinocchio's nose. My IO-520 engine was nearing the end of its service life, and I was eager to replace it with the IO-550 anyway, so I thought what the heck, and arranged to do a very public test of Microlon, with witnesses, at a large annual event. Bill eagerly accepted the challenge, but somehow, when the day came, he suddenly had to leave town, and later claimed the whole thing had been faked in his absence, done wrong, falsified, etc. I did a test flight, a timed, videotaped, full power climb to 10,000' with another airplane in formation early one morning, two witnesses in each airplane. Then we did "The Microlon Treatment" (at $250 for a quart of really pretty blue fluid) that same afternoon, and repeated the test flight under identical conditions the next morning. There was absolutely no difference in time to climb, of course. Hey, at least it did no harm—that I know of.
In this column, I'd like to look at a couple of things that might help your engine, and a few that might hurt it. I'll be showing you some graphs of real data, from my own engine, to illustrate my points.
Disclaimer: I will be mentioning "GAMI" often in this, and subsequent columns. GAMI is short for "General Aviation Modifications, Inc.," of Ada, Oklahoma, developer and manufacturer of "GAMIjectors," and other fine modifications for big-bore flat engines. I'd like to make it clear that the principal, George Braly, and his cohorts Tim Roehl and Mack Smith are good friends of mine. While I was of some minor assistance during development of GAMIjectors, I hold no stock or financial interest (wish I did!), nor do I work for them. I do feel they make a superior product that greatly exceeds the claims they make for it. I run the first production set of GAMIjectors in my airplane, for which I paid full-boat retail, in spite of a kind offer of a nice discount for my prior assistance.
Fly It Often
The three best things you can do for your engine, in my opinion:
1. Fly it often,
2. Install a modern digital engine monitor (I prefer the JPI EDM-700),
3. Install GAMIjectors.
First off, it seems pretty clear to me that flying an engine often is "a good thing," and it makes a lot of sense, with some pretty good data to back it up. Check out any junkyard, and you'll see that internal engine parts rust very quickly, some of them showing visible signs of the red stuff within hours, if left outside. We can look inside the cylinders of engines that don't run much, and see that familiar rusty film on the steel cylinder walls, and we'll also see a marked increase in iron in the oil samples. It's hard to imagine any beneficial effects from this! Seals dry out from disuse, rings take a set and stick, and oil runs off machined surfaces, leaving metal-to-metal contact for the first few seconds of the next run. I don't think there's any debate on this at all, except for degree. I happen to think letting an engine sit for days or weeks at a time is probably the single most harmful thing you can do to an aircraft engine. (Alas, I'm as guilty as anyone!) But check out the aircraft that fly a lot, like trainers, night cargo airplanes, check haulers, etc., and you'll generally see them going to TBO, and often beyond. Considering the mistreatment and poor maintenance many of them get, about the only thing left is "flying hard, flying often."
Monitor It Well
The next best thing you can do for any of these big-bores is to install a modern digital engine monitor. I think JPI makes the best, by far, and I prefer the EDM-700. This device portrays the EGT and CHT in each cylinder, both graphically and digitally (to one degree resolution, if desired) in a very simple, uncluttered way, and also shows battery voltage. With rather inexpensive add-ons, it will also show oil temperature, outside air temperature, turbine intake temperature, compressor discharge temperature, and above all, it can be equipped with a very inexpensive serial output that will feed any computer with all the data the instrument sees.
Finally, GAMIjectors are high-precision fuel injector nozzles that do what TCM and Lycoming should have done years ago. They are nearly magical in the way they balance the fuel to each cylinder, so that as you lean the mixture, all of them rise in lock-step, all peak at the same mixture setting, and all drop down on the lean side of peak together. This not only saves fuel and makes your engine run better at rich of peak, but it allows, for the first time, the far better mode of running lean of peak, just like the big radials used to, and just like the Malibu does.
(Yes, I know, Malibu engines had problems running lean of peak. First, there is now compelling evidence that pilots were so nervous running lean of peak, they added just a touch of fuel "to be safe," thus running CHTs far higher, which damaged their engines. Had they fully followed directions, and leaned them out as instructed, I doubt there would have been as many problems. Second, the engine was very poorly cowled by Piper. As soon as this was discovered and the nose gear door was removed and let the air get out of the cowl area, the engine cooled down, significantly. It took years and lots of toasted engines before that was discovered and fixed.)
Leaning on the Ground
I believe this is a "a very good thing" to do. Most of these engines are set up with a very rich idle mixture, to facilitate starting when cold. This mixture adjustment applies only to the very low power settings used for taxi. In most engines, somewhere at and above about 1200 RPM, the idle mixture setting is overridden by the normal functioning of the carburetor or fuel injection control, and other factors come into play.
How rich is your idle mixture? There's a very easy way to check. You want a nice warm engine for this check, so doing it at shutdown after landing is one good time to do it. Just set an RPM around the usual RPM you use for normal taxiing (I use 900 to 1,000) and start leaning, while watching the RPM very closely. Just before the engine quits, you should see a slight rise in RPM, then a quick fall. (This is very easy with the vernier type mixture controls, a bit more difficult with the push-pull knobs.)
(Note: This is not the way mechanics set the idle mixture! They use minimum idle RPM, or as called for by the book. I'm more interested in the mixture at the engine speed I use most of the time, on the ground.)
Note the amount of rise. You may see "almost nothing" on the big radials, to 50 RPM, or a bit more on some of the flat engines. If the engine was adjusted for sea level, and you do the test at Creede, Colorado (elevation 9,000 ft.), you'll see a very large increase, indeed! The mixture jet is a fixed size, so about the same volume of fuel will pass at all altitudes, but at altitude there is much less air, so the proportion changes to the rich side, and you really need to lean for all ground operations!
OK, why bother leaning on the ground? Most of these engines run ok on the ground at full rich, right? Well, not really. The unburned fuel is very dirty, and tends to foul spark plugs.
Also, over time, these unburned products work their way into the valve guides, causing them to stick, especially when cold (aka "Morning Sickness," from the first start of the day). Eventually this may lead to a valve sticking open enough that the piston will start beating on it, and that's not a good thing!
I've faithfully leaned on the ground for the past 800 hours or so, and have never once had a fouled plug, or a problem with an exhaust guide.
The only downside I know of to leaning on the ground is that it creates the possibility of forgetting to go full rich, and taking off that way. But there is an excellent method for preventing this—simply lean so brutally that taking off is impossible. As soon as the engine has stabilized after starting, lean it out until you get that RPM rise, then an RPM fall, or until the engine begins to run rough, then enrich it just slightly, just barely enough to make it run smooth. No, you cannot hurt the engine by doing this since you're pulling virtually no power.
Look at this simple chart, which is actual data from my IO-550 for an engine start, runup, and beginning of the takeoff roll. There are several interesting things here. There will be additional charts, all in this same format, and all recorded from a JPI EDM-700 Engine Monitor, connected to a personal computer, and later graphed in Excel 97 (one of Microsoft's finest efforts). To reduce clutter, I have graphed only the hottest EGT and CHT (#2 on my engine, by a very small margin) and fuel flow. In order to get everything nicely placed on the graph for clarity, I have simply multiplied the very low fuel flow by 100, so both fuel flow (times 100) and CHT can be read on the right scale, with EGT on the left scale. All data points are captured at six-second intervals, which doesn't miss much.
Note the long, slow, even rise of CHT. This is very typical of all changes in CHT, for the engine is a massive heat sink, and it takes a long time to change its temperature in either direction. This CHT probe is buried deep in the cylinder casting, and is an excellent measure of internal engine temperature. Watching these CHT traces also gives the lie to some of the "shock cooling" theories, but that's for another time.
By contrast, note how quickly the EGT changes. It's instantaneous, and this is very useful in flight. It is possible see a CHT that is too hot, or increasing, and just tweak the mixture enough to get a 20F or 50F change in EGT, then wait for a minute or three, and the CHT will gradually follow.
Finally, look at the fuel flow trace. Before turning the engine, I prime it with the boost pump for a few seconds, throttle cracked, mixture full rich, and that runs about 8 gph into the engine. Boost off, and immediately hit the starter.
You see the EGT rise rapidly as the engine starts, and the fuel flow stabilizes at about 4 gph, in full rich. As soon as the engine is running smoothly I lean it right out until the engine falters, then enrich just barely enough to get it to smooth out. This is so very lean, I cannot get much more than about 1200 RPM, and it's about 2.7 GPH.
I’m Not Being Cheap
Now, airline captains are cheap, but even I am not cheap enough to worry about wasting that 1.3 GPH for a few minutes on the ground. But if you've ever wondered why spark plugs foul during ground operations, I think you're looking at the reason.
For runup, just a touch of extra mixture will allow 1500 to 1700 RPM for the prop check, and a quick mag check, then back to about 1,000 RPM, and a quick twist back on the mixture again. This is important, remember, to prevent taking off with that mixture not in full rich. In my opinion, if you will not lean it right out, almost to the point of roughness, then never mind leaning at all on the ground, just leave it in full rich at all times on the ground. I'd rather see you with fouled plugs, than a ruined engine, or worse!
You will note that EGT always rises during the time you operate on one mag, during the mag check. This is because with only one plug firing, more of the mixture is still burning as it exits through the opening exhaust valve, and since the EGT probe is just outside that valve, it "sees" the still-burning mixture, and shows a higher value. This is just one of many things that will cause a misleading EGT indication. Additionally, if you check the mags while leaned out, the mixture is burning much more slowly, so even more burning mixture escapes. Since this burning is not taking place within the combustion chamber, even more power is lost, so you'll see a greater mag drop during a check. If this makes you uncomfortable, go full rich for the check, and then just leave it there for the takeoff.
A side note on the big radials, which have "Auto Rich," "Auto Lean" and "Idle Cutoff" positions. There is a small, but noisy group of people who insist on operating them on the ground in Auto Lean, thinking they are accomplishing some of the above "good things." Not true, on those engines, at and below about 1200 or 1400 RPM, on the ground, there is no difference at all between Auto Rich and Auto Lean. If there was a difference, you'd see a slight RPM change. In order to enjoy the above benefits, those engines must be manually leaned, almost back to the Idle Cutoff position, to do any good. They are also generally set up to idle much leaner than flat engines, so unless they are maladjusted, it's better to just operate in Auto Rich on the ground. The danger with the Auto Lean position on a big radial is that it will allow takeoff power to be set, with potentially catastrophic results from detonation.
At higher elevations, all these engines should probably be manually leaned, as the idle mixture gets richer with altitude.
Partial Power Takeoffs
In recips, they're dumb, dumb, dumb, no matter how you look at them (jets are a different subject, not covered here). But maybe if you do them, you haven't really looked at them, and have just listened to some ABM ("Airport Big-Mouth") spout off. Without thinking about them, you may very well fall for the Old Wives Tale "Gee, why beat up the engine, when I don't need to?"
First, any partial-power takeoff will leave you lower and slower, for a longer time. This does not enhance safety. If you are flying a twin, a partial power takeoff enormously complicates the engine-out case, because not only do you have the engine loss to handle, you must get the remaining engine(s) up to full power to get any climb performance at all. But there is more, when it comes to "high-performance" engines (big-bore flat sixes, and big radials).
To the best of my knowledge, all these engines have some sort of method or device to greatly enrich the mixture at full rated power, for cooling. On some, it's called a "Power Enrichment Valve" and it's a straight mechanical device. Others may call it by different names, and it's done a little differently on the big radials, but the purpose is the same, to cool the engine with excess fuel. Look at the exhaust from any of these engines at full power, and you'll see a dirty, sooty stream of unburned "stuff" coming out.
If you don't pull enough power to activate that device, you are pulling very high power, without the enriched mixture for cooling. Some engines may tolerate that better than others, but it's hard on all of them.
A Cool but Complex Subject
Engine cooling is a very complex subject. In theory, the aircraft, the engine, the cowling, and the mixture are all rather carefully designed to work together to produce a safe takeoff. Remember how slowly that CHT rises? The same thing happens on takeoff. In the early stages of the takeoff roll, the CHT hasn't come up enough to do any damage, and in the later stages, there is enough cooling airflow to stop further rise. If you make a partial power takeoff (using climb power, for example), you will be too lean, which will make the CHTs rise much faster. You will also delay the aircraft's acceleration, which means you'll take a longer time to get to the airspeed that is sufficient to cool the engine at climb power. Bad thing, all the way around. Full rated power, on all takeoffs, please. It is always easier on the engine, it is always recommended by the manufacturer, and it is always safer.
Here's a fairly conventional takeoff, full throttle all the way, 2700 RPM until liftoff, 2500 RPM right after gear up. Again, note the slow, steady rise of the CHT, which will eventually stop around 350F. I have changed fuel flow multiplier to "times ten" (instead of "times 100") for this chart, to keep it in a reasonable position on the chart. (A fuel flow of "250" is really 25.0 GPH.) Note the fuel flow drops automatically, when the RPM is reduced to 2500, from 2700 (primarily for noise, my prop really howls at 2700).
Now here's the way a lot of folks do it, as taught by one major training outfit. Full throttle, full rich mixture, 2700 RPM for the takeoff. That's fine, so far. But then we pull it up to climb at 95 to 1,000' agl, dramatically reducing the cooling airflow. During that climb, the CHT has soared through 350F, headed for 400F, plus! At 1,000' agl, we reduce manifold pressure to 25," and RPM to 2500, but look at the CHT! It never even slows down, keeps right on headed for the moon, in spite of an increase in airspeed to 120 concurrent with the power reduction! What has happened here is that by pulling the throttle off the full power stop, we have cut out the "Power Enrichment," and actually leaned the engine quite a bit.
What's Too Hot?
There is now very real data to support the idea that anything over 400F can be very harmful to these engines, notwithstanding the factory limit of 460F. Modern test instrumentation (courtesy of GAMI) has demonstrated that there can be more than 150 degrees difference around the circumference of a cylinder. Single probe CHT instruments may not have the probe installed on the hottest cylinder. Even multi-probe CHT instruments may not be measuring the hottest spot on a given cylinder.
Accordingly, this business of reducing to 25" right after takeoff may be one of the most harmful things you can do to your engine. Far better to just leave it at full power (limitations permitting), or, if you have a noise problem, just pull the prop back a couple hundred RPM to keep the neighbors happy.
Now, some will yell about this, based on the Old Wives' Tale that goes "Always reduce MP before reducing RPM." But look at the logic. On my 550, that 200 RPM drop amounts to a 15hp loss. If you watch the JPI (see chart), you will see the EGT drop, the CHT will remain about the same or a bit less, and the actual pressures inside the combustion chamber remain essentially the same. Please tell me how this can be harmful to the engine? Anyone?
In fact, TCM did exactly the above, by simply limiting the RPM to 2500 on the same engine, to satisfy the German noise requirements for Beech A-36s delivered there! They just call it a 285 hp engine, instead of 300, modify the performance data to suit, with no other changes.
On the big radials, there ARE times when the order of increasing/reducing MP and RPM are very important. For simplicity, a lot of that got reduced to "rules of thumb" that get used all the time (sometimes unnecessarily), and those carried over into the flat sixes as they began pulling more and more performance out of them. It may even be true of some of the flat engines with gear-driven superchargers, or turbos, but on most GA engines, it's simply a non-issue.
OWTs Passed On
Years ago, I was getting a quickie checkout from a young CFI in a 182 one day, and we had climbed out to about 7,000' msl (3,000' agl) for the airwork, when he requested cruise power, and specified 23 inches; and 2300 RPM. Since full throttle was already producing about 23 inches, I started pulling the prop back from about 2500, and he came unglued, with a huge lecture about how important it was to pull the throttle back first! He went on and on, regurgitating the OWTs he'd been taught about detonation, over-pressure, bearing failure, blah, blah, blah. Sad. So I pulled the throttle back to 20 inches, the RPM back to 2300, then reapplied full throttle, and he was just as happy as he could be. I can only shake my head at the ignorance, and weep at how many students he has instilled with that same misinformation.
By contrast, on the big radials (and a very few GA airplanes) a reduction to METO power after some speed is attained is not harmful, as the manufacturers provided for this. But it is not a good idea to use METO for takeoff, for the cooling reasons mentioned above, and the engine failure case.
(Note: METO ("Maximum Except Take Off") power is not often mentioned in GA publications, but if your POH says something like "Takeoff Power is limited to X minutes," then the recommended power after that is the functional equivalent of METO in the bigger engines.)
Alert readers will see some oddities in these EGT traces. EGT is often not an accurate reflection of the real combustion temperatures. EGT is extremely useful, but not as absolute numbers. Rather, watch the trends, and develop a feel for what they should show at certain key points of flight. It is not at all necessary that EGTs be equal to each other, because even minute differences in installation position can have a large effect. What is important is that they move together when conditions change, and this is what GAMIjectors do so well.
Now, would you like to see something really interesting? Picture this, if you please. A normal takeoff with my IO-550, full throttle, 2700 RPM, full rich. We lift off, get the gear up, and reduce RPM to 2500 for noise, accelerate to 120 knots for the climb. I really like the higher speed, for cooling, and for visibility. Fuel flow is about 28 GPH at full power, drops to 25 GPH after the RPM reduction.
Prepare to Scream
But, now the wild one, prepare to scream. At about 1,000' agl, still wide open throttle, 2500 RPM, I reach down, grab the red mixture knob, and firmly and quickly pull it back to about 15 GPH! I have done this with a number of very experienced pilots, and most of them jump right out of their skin, in horror.
Before you do the same, look at the data, and follow me through, here.
This chart is a continuation of "Takeoff 1." At time 4 minutes and 25 seconds (00:04:25) or so, the CHT is gradually drifting up to 400F, when I rudely deprive the engine of about 44% of its fuel, in about one or two seconds. Note the EGT takes a huge jump, and then falls, indicating we have passed through peak EGT (have we ever!). This instantly halts the CHT rise, and it stabilizes at about 370F. After a minute or so, I note the CHT starts dropping a little, and since I have arbitrarily decided to keep the fire burning at about 380F, I add just a little fuel to the fire. The fuel flow kicks up to a little over 15 GPH, the EGT rises about 70F to 1470, and the CHT slowly eases up to 380 and continues, so I ease the fuel back a bit, until the EGT settles around 1450F, and the CHT stabilizes nicely at 380.
Note: A word of caution. While, from considerable personal experience, and from the experience of a number of others, this works very well on the big bore normally aspirated TCM engines (IO-520 and IO-550) , there are some engines that this should not be tried on. In particular, there is some possibility that this kind of mixture reduction could, under some adverse circumstances, cause a detonation problem on some of the Lycoming normally aspirated engines, and on some of the TCM and Lycoming turbo-charged engines, which have different detonation margins. Of course, an engine that has the timing set to the wrong value or that has contaminated fuel can always cause detonation, but leaning in this manner might aggravate one of those maintenance or operating problems.
Furthermore, I want to make it perfectly clear I'm not recommending you run your engine this way! It is immoral and fattening, will deprive the big oil companies of their rightful revenue, make aviating much cheaper, and may well grow hair on the palms of your hands! I want no lawsuits, here!
With those notes of caution firmly in mind, let us return to the graph. Why aren't we burning this engine up? Full power, with a very lean mixture? Really goes against the grain, doesn't it. But you have to realize, the engine is not really pulling full power! We have leaned it out so much that even with full MP and 2500 RPM, our power is back significantly. Additionally, we're so lean, that excess air is doing the cooling, instead of excess fuel.
The engine is running cooler (380F instead of 400+F), cleaner (no excess fuel), and cleaner and cooler is "better." This is reason enough to do this, but there's more.
Leaner Is Better — If You Can
Of course, this mode of operation is impossible without GAMIjectors, because the mixture distribution without them is so bad that the engine begins to run rough long before we can lean this much.
(Well, that's not entirely true. The injector nozzles that TCM puts in their engines are out of tolerance almost as often as they are within tolerance, so if you're lucky enough to have those nozzles in the right cylinders, you could have the same effect you'd have from GAMIjectors. If you're lucky enough to enjoy this situation mark those injectors, and make very sure they get back into the same cylinders after any work!)
These charts were derived from data taken during two flights. The first was my standard operation, leaning the engine out drastically, and climbing that way to 5,000', straight out. Second was done at full rich. Here's the data:
The first one hit 5,000' at 13.9 GPS miles from the airport. The second hit 5,000' at 10.9 GPS miles, and I just flew level at 5,000' to 13.9, then noted the data. So climbing lean will take longer, but the time to get to any given point and altitude "downrange" will be very close.
Okay, that's 1.2 gallons saved in the climb to 5,000', big deal. But wait, that's only seven minutes! Project that out to a long climb to 13,000' or so, and you will end up with about seven gallons more fuel at a given point in space. That's more than half an hour's fuel saved, and is about the minimum legal reserve in my airplane! Yes, it takes a bit longer to get to altitude, so what?
I've been running my IO-550 this way for about 500 hours now, and there is every indication the engine loves this sort of thing. A number of others have more than 1,000 hours of this same operation, so it can't be too harmful. Only time will tell if this will allow longer TBOs, of course. But, I cannot see any reason why it won't.
Be careful up there!