Pelican's Perch #64
Where Should I Run My Engine?
(Part 2 — The Climb)
Last month, AVweb's John Deakin started a discussion of where to run an engine during a typical flight. With so much detail needed, he ended the column just as we took off! Now he's back to talk about the climb, and as usual he has real-world data to back up his explanation.
Before continuing where we left off in last month's column, I have a little housekeeping, if I may.
AVweb has very kindly allowed me to use "firstname.lastname@example.org" as my primary e-mail address for these past five years. That has worked very well. It has been very reliable and quick, and it has also allowed some "name identification" to go with the column. That has had its advantages and disadvantages!
The new owners of AVweb feel there is a security issue with that, and I've been told that AVweb addresses will be discontinued for many of us, or perhaps all of us. Accordingly, if you do happen to have me in your address book, please note that I am reverting to my 20-year-old e-mail address at email@example.com, which can be reduced to firstname.lastname@example.org. I also use email@example.com, for e-mail related to that subject.
A reminder: Please check out Old Books and note particularly the little Wright booklet, "Basic Theory of Operation, Turbo Compound Engine," by the Wright Aeronautical Division of the Curtiss-Wright company.
This little gem has become the cornerstone of our seminars, and is one of the handouts we provide. It was something of a sleeper for all of us. It was part of the original inspiration for George Braly about seven years ago, when he first began to explore the fuel injector issue, but it got a bit lost in the shuffle ever since. We have each slowly rediscovered it, and found ourselves constantly referring to it for some of the basic principles. It is astoundingly accurate. Above all, it is a simple description of many of the issues common to ALL reciprocating, internal combustion, fixed timing, spark-fired, gasoline powered, four-stroke engines. It was mostly written by American Airlines, after they had learned how to run the huge R-3350 to TBOs of 3,600 hours and beyond, with maximum power settings producing more than 1 HP per pound of engine, and more than 1 HP per cubic inch of displacement. Our "modern" flat engines only do barely more than half that!
I cannot emphasize strongly enough that these engines all have IDENTICAL combustion events. The pressures and temperatures react in EXACTLY the same ways, and in the EXACT same relationships to each other. All have aluminum pistons, running inside steel-barreled aluminum cylinders, with fixed spark timing, running the same fuels. There are very slight, but mostly inconsequential, differences in metallurgy.
Please don't fall into the common error of thinking, "Yeah, but that's a radial engine, and I have a flat one," or, "Well, that's a turbo, and mine is normally-aspirated," or, "He's running a flat six, while I have a flat four."
Folks, one of the universal assumptions at the heart of everything we think we know is "the laws of physics are everywhere the same." (Sir Isaac Newton relied upon that concept, but I forget who he flew for.)
The combustion characteristics of ALL these engines are IDENTICAL, from your lawnmower, to the giant Wright R-3350 Turbo Compound engines in Connies and DC-7s of yesteryear.
I started reprinting and selling these old books at cost as a part of my ground schools for the C-46 for the Southern California Wing of the Commemorative Air Force, in Camarillo, Calif. Then I mentioned them in one of my early columns, and now I get a steady trickle of orders from all over the world. Color reproduction is expensive, and I've insisted on keeping the color. I'm operating at a very small loss, which is just where I want to be to avoid tax complications, and keep the fun in it. The little Wright booklet is in danger of becoming profitable, so I'm dropping the price to $20, including domestic shipping. Highly recommended, if you'd like to understand more about these incredible engines.
This Just In ...
Finally, I had submitted this column for publication, and put it behind me. I mentioned a little wistfully to George Braly on the Beech Owner's maillist that I wish I had data from an IO-520 normally aspirated engine with two different climbs, one with no leaning, the other with normal leaning as I suggest here. Within a few hours, a kindly reader, one Glenn Olsen, went out in his own airplane, flew it out from under a Class B, landed, then from that point did two identical flights to 10,000 feet, one with no leaning, one with normal leaning; downloaded it all from his JPI, and fired it off to me! Thanks, Glenn!
It's too late to integrate the resulting charts from those flights into this column properly, for I'd have to start over. However, they are so pertinent to this column, I'm going to add them on at the end, with more comments. As you read this column, you might want to flip to the end (the section titled "Time For the Final Chart,") take a peek, and then come back.
I apologize for the awkwardness.
Back to Last Month
I rudely left you hanging last month, right at liftoff, full power. It's worth noting that you could have left the engine running at that power setting for the entire month, and two months more. Assuming a total of about 750 hours in a month, almost all these "big-bore" flat six (and flat four) engines should happily run continuously at that power setting for three full months, and probably well beyond, as continuous running is known to be "better" for them. I would not be even a little surprised at 10,000 hour TBOs on any of these engines, if run continuously, paying attention to temperatures and pressures.
It's not the power you pull that determines maintenance, repairs, and TBO, it's the temperatures and internal pressures you run, and how you abuse your engine with cold starts to instant high RPM. Sadly, during the last ten years or so, it has been more an issue of quality control at the factories than engine management. Lycoming and TCM crankshafts have broken, and factory-installed exhaust valves on TCM engines appear to fail prematurely, no matter how you run the engine.
Please get away from that feeling that you are somehow abusing your engine, and that you need to reduce from takeoff power right away to "save the engine."
The sole exception is when you have an engine with a LIMITATION regarding takeoff power, sometimes five minutes, sometimes less. Please do observe that limitation, there MAY be something the factory knows. Again, I refer ONLY to a genuine LIMITATION, in the "Limitations" section of the POH, and NOT the text written in the "How to fly" section.
I will also repeat from the previous column, YOU MUST BE GETTING FULL REDLINE FUEL FLOW at full takeoff power AT SEA LEVEL! In truth, the factory redlines are often a bit on the low side, and flows should be tweaked up just a bit. Even a half-gallon per hour makes a BIG difference in CHTs during climb.
In general, and speaking very roughly, if you see EGTs anywhere over about 1,300°F (lower will not hurt a thing and is probably “better”) during a sea-level takeoff, or CHTs above about 360°F right after takeoff, YOUR FUEL FLOW IS TOO LOW. Having a good understanding of the proper relationship between the EGTs and the fuel flow at very rich mixture settings will always give you a good cross-check on whether or not you are getting adequate fuel flow -- even if your fuel flow needle breaks off and falls to the bottom of the instrument!
If your fuel flow is too low on takeoff, don't let your mechanic talk you out of setting it up a bit, and if necessary, find a mechanic who will do it. VERY few mechanics understand the importance of this. Far too many mechanics consider the fuel flow redline a maximum, and they consider a little bit less as "good," or "better for the engine," thinking they are being "conservative." After all, we all like to stay a little bit (or a lot) away from all sorts of redlines, right? That fuel flow redline is a MINIMUM, not a maximum. Treat it as such.
OK, On to Our Climb, At Last
Climb power? Use the same power you used for takeoff (unless there is an explicit limitation). If you pulled the RPM back for less noise, you can leave it there, or run it back up to full redline RPM once noise is no longer a concern. The choice is yours. Or pick a smooth RPM in between, if there is one. Simplify your life, and just leave the throttle wide open.
As the altitude increases, the manifold pressure drops, and the engine receives less and less air (or less oxygen, to be more correct). But the fuel orifices remain the same, the fuel pump is turning at the same RPM, so the fuel flow remains roughly the same in GPH or PPH (give me a little room, here -- as the TCM altitude compensating fuel pump is "better" in this regard, and will automatically lean the engine during the climb when set up properly).
(There are also some carburetors and fuel servos that do attempt to measure the MASS of air entering, and regulate the fuel flow accordingly, but let's skip those, for now.)
Less air, same fuel. That means "richer," and if you leave it alone in the climb, at some point you'll be losing a lot of power from a much too-rich mixture, and leaving a dirty trail of soot behind you. Without leaning, the engine will eventually be running so rich that it will lose enough power to stop the climb many thousands of feet below the true maximum altitude.
Obviously, we need to manually lean most normally aspirated engines during the climb. There are exceptions like the GSO-480s on the Twin Bonanza, but they are not common today. Turbo’d engines also remain fully rich for the most part, but this column is about normally aspirated engines.
In most of these engines, CHT is usually somewhere around 250°F on the ground, before takeoff. The CHT will gradually and smoothly rise through the takeoff roll, and continue rising well into the initial climb until the cooling airflow is enough to stop the rise. If the engine compartment baffling is well-engineered and fuel flow is properly set, it should stop rising and stabilize at something around 320°F to 360°F in the climb, once cooling airflow is present. One of the strong reasons for using full power on all takeoffs is to get to that cooling speed as soon as possible, to stop that CHT rise. That's also why I don't like the Vx and Vy climbs unless they are absolutely necessary; they run the CHTs higher under any given conditions, needing even more fuel flow.
From that point, if left to its own devices, the engine will gradually cool off as power drops off with altitude, the mixture goes richer and richer, and the EGTs and CHTs drop (see added charts, below).
Note the EGT and CHT
Do you ever note your EGT shortly after liftoff, during the first 1,000 feet of climb (at sea level)? You should, and here's why.
If your engine is set up for the correct fuel flow at sea level and full power, and you note that EGT shortly after takeoff, you have a good EGT setting for the rest of the climb. Simply tweak the mixture knob to roughly that EGT from time to time during the climb, and you will get excellent results, both in terms of economy and safety for the engine.
Note the simplicity? Full throttle, full prop, full rich for takeoff. Tweak the mixture knob now and then in the climb for a loose "target EGT." That's it! No tweaking the MP every thousand feet to bring the MP back up to some oddball MP setting that never made sense in the first place.
Oh, you say you never take off at sea level? (There's always a heckler back there in the back.) Simple: Lean to about 1,250ºF, and make sure no CHT goes over about 360ºF, with 340º being "better."
Oh, you don't have an engine monitor? Sorry pal, I can't help you; you're on your own. I'm really hard-line on engine monitors; I think it's stupid to operate one of these big, expensive, life-supporting engines without one.
Another technique in normally aspirated flat sixes that are not set up properly with enough fuel flow on takeoff is to watch the CHT on the hottest cylinder. If it is above about 350°F or so, I know there's not enough fuel flow. Try the boost pump! That may not help much, it may help a lot, it may even be "too much." If the latter, simply lean manually to get the fuel flow where you want it. If that's not enough, or it doesn't work, continue the climb, and the natural enriching process will soon take over and cool the CHTs. Once that happens, note the EGT at that point, and lean to that EGT for the remainder of the climb.
Then get that fuel flow set higher!
Alert readers may note that I don't use consistent temperatures on some of these settings. That's deliberate, because I don't want you to be picky on these, either. Just ballpark it. FAR too many pilots see a number, and become a slave to it. RELAX, for crying out loud! This is supposed to fun, albeit with a serious side. These techniques REDUCE your workload.
If you see a really high and rising CHT, flip the boost pump on to give the engine some extra fuel flow. Low if you have it, or just ON if you don't have a two-speed pump. On some engines, this may produce too much flow, and you may have to manually lean the mixture to bring it down a little, but still more than it was before you turned the boost on.
(At low power settings on the ground, most boost pumps will flood the engine, sometimes killing it. At full power, few boost pumps are powerful enough to do that, but still, they may produce too much of a good thing, and there may be some power loss from "much too rich." The mixture control will handle this nicely, as again, the engine doesn't care how it gets the fuel, just how much it gets.)
You might think in terms of three scenarios. One is a properly set up engine, where you simply note the EGT passing 500 to 1,000' and lean to it for the climb. A second scenario is where you must "do something a little different" (boost pump, wait for CHTs to drop). The third is when it is really lean, where you either need to abort the flight, or just pull the mixture right back to an LOP setting, and do the climb LOP.
So basically, I'll start leaning to keep that EGT roughly the same as it was right after takeoff, when it stabilized. That, in turn, will maintain the CHT in about that same 330°F range, which is my goal. Why not use CHT in the first place? You certainly could, but CHT indications change very slowly, and it takes too much attention and fiddling. We're trying to make it easy and simple. By using the quick response of the EGT, I can lean it about 20°F quickly, and then come back a minute later and check what that did to the CHT. We continue to climb, note the EGT drop, and lean to bring the EGT back up, then later double-check the CHT. Repeat as necessary.
The workload is a bit high the first few times you do this, because it's strange, and you have to think about what you're doing. But once you do it a few times, you will know what to expect, and the adjustments will be quick and natural, taking very little or no attention, and far less attention than "the old way."
Parking the Engine
It is worth noting at this point another crucial concept we teach in the seminars. We call it "parking the engine." Any time we adjust the engine controls, we want to leave the engine in a situation where it will not, if left alone or neglected, change by itself into a "bad" setting. That way, if we get busy, or have to concentrate on ATC, another airplane system, a radio setup, or a baby barfing, the engine will be safe. In the climb I have just described, the worst that can happen is that you forget to lean, the engine slowly goes richer and gets cooler. The engine and its controls are "parked."
A reminder here, "On the rich side of peak, leaner is hotter, but on the lean side, leaner is cooler." That's a crucial concept! Repeat that to yourself, until you "get it."
Some like climbing LOP. This works, as the engine doesn't know what the airplane is doing; all it cares about is fuel, air, spark, and cooling airflow. But if you set up a LOP climb in a normally aspirated engine and forget it, the mixture will gradually go richer and richer, perhaps getting too hot. LOP climbs violate our general principle of "parking" the engine safely. It doesn't mean you can't do it, or shouldn't do it, it just takes a little more care. The alarms on the JPI make this a much more feasible operation, and it will save a little fuel, perhaps enough to give you just a bit more range to make that long non-stop.
Don’t be anal about adjusting the climb mixture, please? Many people fiddle it to death, and that’s not necessary. Once you make an adjustment, it will be just fine for a thousand feet of climb, or more. Tweak it again, and it’s good for another thousand feet or two. There's a subtle (and graceful) distinction between tortuously trying to set it within a degree or two, and just setting it so that it drifts into what you want. It doesn't matter if the climb CHT runs 320, 340, 360, or even 380. If it's cooler than that, and you're still ROP, lean it a bit. If the CHT is sneaking up slowly, wait a bit, leave it alone, and altitude will take care of it.
[Note: There are some notoriously bad engine installations in which the cooling air flow baffling is so poorly designed, installed, and maintained that holding CHTs under 380 to 400°F on a normally aspirated engine in cruise is a problem. If yours is one of these, you should spend some serious time and effort to get that corrected. There are people that understand these issues and who know how to get it done properly.]
Over the Hump
So, after takeoff, "climb a little, lean a little, EGT rises," right?
At some point, somewhere above 7,000 feet or so (this altitude depends on how good your baffling is, and may be as high as 9,000), you’ll notice that when you lean the mixture, the EGT will drop, instead of rise. This indicates you’ve gone "over the hump," ("When LOP, leaner is cooler") and you're now climbing lean of peak EGT. Never fear, by this time the engine power is low enough that it won’t matter what mixture you run, it won’t damage the engine.
"'Doesn't matter,' he says?" Well, it doesn't matter for stress or internal pressures, but where you set the mixture will have a MAJOR effect on mileage and speed! If you’re continuing to a higher altitude, and want the climb performance, note the peak EGT you just saw, and set the mixture slightly richer, to around 50°F ROP. That will give you the most power you can extract from the engine. (Actually, 50°F to about 100°F ROP is "best power," but 50°F ROP will give you the power for a bit less fuel and a bit higher CHT than 100 ROP.)
At this point, you’ll see the hottest CHT start dropping, perhaps after a small increase as you transition to 50°F ROP. When that starts, it’s time to start closing the cowl flaps to reduce drag, and I generally just close them all the way. The CHT might tick up around 10°F or so, but not much more. I figure cowl flaps on most Bonanzas are good for about 10°F CHT, about three to five knots in TAS, and 1,000 to 2,000 feet of absolute ceiling.
Once you make that adjustment to 50°F ROP, you won't have a fixed point to lean to, and your engine will probably continue self-enriching at higher altitudes. But experiment with it yourself, in YOUR engine. During the continued climb, every couple thousand feet just lean it a bit to peak, then back to 50 ROP. With a few flights like this, you'll know what it takes to keep your engine 50 ROP in the higher altitudes. If this sounds like a lot of work, remember that at the higher altitudes, it's going to take you ten minutes or so to climb another thousand feet! This procedure is to expand YOUR knowledge, of YOUR engine, and once you have the picture, it's a quick adjustment, with little thought.
If you’re really pushing altitude, and want to take your normally aspirated engine into the flight levels (I've been to 21,000 feet with my old IO-520), you might very, very slowly lean to peak again, using the tiniest movements on the mixture you can. Note the peak and then the drop as you lean, then reverse and enrich past peak again, very slowly, noting the higher EGT of the two. (For reasons we do not yet understand, the maximum EGT value will almost always be a little higher when leaning in one direction.) From that, enrich to about 50°F to 80ºF ROP, and that should do it.
If you're patient, you can actually find the "absolute altitude" for your current weight and conditions, leaned for best power. From that point, enrich the mixture to about 120 ROP, or lean it to peak, and you'll lose altitude! It's a vivid demonstration of the power control that is available from mixture alone.
During that climb, if the engine begins running rougher and rougher as you lean, your fuel distribution is not correct, and you will be stuck with enriching until the engine runs smoothly, and leaving it there. This is the problem GAMIjectors are designed to correct.
Another trick. If you really want a "minimum fuss" climb, try this for fun. After takeoff, established in the climb, lean until you drive the hottest CHT up to 400ºF, then just let it be. The "natural enrichment" from the climb will slowly enrich the mixture, and that CHT will slowly drop. When it gets down to about 300ºF, lean it out again. You might even get all the way to your cruise altitude that way, without touching the engine controls again!
I know. That's a bit lackadaisical, even for a lazy old airline pilot like me, and it's not the best way to manage your engine. But, it might be a fun experiment, and a way to learn a bit about your engine.
Next month we'll talk about the various methods for cruise flight. I can promise you, they won't look much like your POH! You can take a black marker, and block out all references to "percentage" of power. You don't want to deface your POH? Ok, ok, block 'em out mentally, then!
Give Me the Data
(The following was inserted at the last minute before publication. This is Glenn Olsen's data.)
Ok, let's talk about some of the charts we use to back up our suggestions. If you have an engine monitor that will record and download your data, you can chart what's going on using Microsoft's magnificent spreadsheet program, "Excel." I have rarely seen a software package with more sheer dazzling power than this one.
For those into such thing, a brief tutorial. For the rest, skim through this until you get to "Time For the Final Chart."
A downloaded data file might look like this, after you download it to your computer:
"EDM- 700 V 281 J.P.Instruments (C) 1998"
"Aircraft Number N9766Y_"
"Flight #47A 12/27/2 22:0:50"
"Eng Deg F OAT Deg F F/F GPH"
"Duration 1.08Hours Interval 6 seconds "
This will normally be named a "TXT" or "PRN" file. It'll have the flight information, then a "header" row that identifies the data to follow, and finally, the data, separated by commas.
Excel will open this file, recognize it's not quite ready for prime time, and will ask a few questions. Delimited? Yes, Next. What with? A Comma, Next, Next, and Finish.
That should produce a straight-forward spreadsheet like this (click for full-size version - 71 Kb):
The "Time" column is a bit of a bother, but since I don't really care what time of day all this happened, I just format the whole column for "Time," "00:00:00", then put a zero in the first data line, add "00:00:06" to it, and continue that down the whole column. This gives "Elapsed time" from zero, which is more useful, most of the time.
Charting the Data
If you then highlight the whole mess from "Time" on down to the bottom, and over to the right, you can hit "Insert," "Chart," "Line," (choose the line chart in the upper left corner), "Next," "Next," "Next," then "As a New Chart," "Finish," you should see something like this:
The only things I've done here are to make the background white so the skinny colored lines show a little better here, and I've circled two interesting things. When I first saw those, right out of the box I knew I had some really good data.
Whatever data is in the first column will become the "X-Axis," which is elapsed time, across the bottom. The "Y-Axis" up the left side will scale automatically to the highest number in the dataset, which in this case is RPM. Each column B through the last column will plot on the chart, with a line connecting the data points. Great stuff!
The single yellow line above all the rest is the RPM trace. By having it up there, the left side scale has to run from zero to at least 2,700, which squishes all the rest of the data down low. The multicolored traces between 1,000 and 1,500 are all six of the EGT traces. The problem with this is that it leaves the huge scale on the left side, and all the small numbers hide down in the clutter at the very bottom.
Back to the raw data. Make the RPM column RPM/10 (2700 RPM becomes "270"), the Manifold pressure becomes MAP*10 (24" becomes "240," and finally, the Fuel Flow becomes FF*10 (19 GPH becomes "190." These are much more workable numbers.
We run through the charting process again, adjust the background, change some colors, rescale the primary (left) axis, label a couple of things, and we have:
Now we can see a bit more clearly what the engine monitor has been trying to tell us!
We often want to take a very careful look at all six of those EGT traces, and all six CHT traces for troubleshooting, but in this case, we're interested in the overall picture. I went back and moved all the data over 26 columns, then made the new column A represent time, Column B became an AVERAGE of the EGT (=Average(AB2:AG2)), CHT became an average of all six CHTs, then four more columns for Manifold Pressure (MAP*10), RPM (RPM/10), Fuel Flow (FF*10), and altitude (ALT/10).
Back to the familiar charting process, and we now get pretty much the same view, but with one line for EGT (average), one line for CHT (average), and a line each for the other parameters that are important at this time. A few more labels, a few more color changes, and we see:
We rescale the left axis from 0 to 1500. We click on each of the data traces down at the bottom in turn, "Format," Selected Data," "Axis," "Secondary." This moves all those items to the right side of the chart (the secondary axis), which digs them up from the basement of the chart, expands them greatly, and gives a much clearer picture, especially if we rescale that axis to show zero to 400. We also chop off the first 16 minutes, the last 20 minutes, and 14 minutes out of the middle, and we get two side-by-side graphs that vividly illustrate the difference between a climb with no leaning, and a climb with normal leaning.
Time For The Final Chart
Ok, all you folks who skipped the dirty details, here's where we pick you up again!
The top red trace is the average EGT in both graphs. The rapid rise is the rise during the takeoff roll. Be careful not to think of that as the climb to altitude!
The climb to altitude is shown by the diagonal brown line that starts near the lower left corner of each graph. Note that RPM (blue line) remains at 2685 throughout both charts, the manifold pressure (black line) peaks at takeoff, and drops linearly with altitude. Those are the only two parameters that are the same on both graphs.
Climbs were from about 700 feet msl to 10,000 feet msl, on a cool day (50°F on the ground).
Without leaning, both EGT and CHT drops. The fuel flow remains precisely the same, because the fuel pump is turning at the same RPM throughout.
By leaning to maintain the approximate EGT during the climb (right graph, GREAT JOB, Glenn!), we see that the CHT also says pretty constant (between 300 and 312°F), but now the fuel flow drops as we climb and lean manually. Glenn obviously paid attention to keep that EGT so even, but we don't need to do that in normal operations. Lean it to the desired EGT every couple thousand feet, and otherwise leave it alone, once you know the drill.
As in all normally-aspirated engines, even with leaning, the power drops with altitude, because less air is available. Manual leaning is one way to get more power in the climb, but overall, it still falls off.
However, if you don't lean, the power drops even more, because the engine cannot burn the extra fuel. If you carry this too high, the fuel will literally put the fire out, and the engine won't run at all. This power difference is illustrated by the fact that the unleaned climb to 10,000 feet took 12 minutes, and the leaned climb took only nine!
Now that you've seen the charted data, may I suggest you reread this column, and see if it doesn't make a bit more sense? Again, I apologize for the poor layout, but my choice was simple, either leave the charts out entirely, or do it this way. "The Book" will do it much better.
Be careful up there!