Motor Head #11: One Thousand (Not So Easy) Pieces
Wonder why airplane engines cost so much? Just try to build one. AVweb's Motor Head, Marc Cook, did (well, he helped anyway) and found it to be very enlightening. And then he flew a Turbo Cirrus ...
Like many pilots and aircraft owners, I've wondered why our engines are so damn expensive. I mean, on a horsepower-per-dollar basis, they're way behind it. In a day when $30,000 does well to buy you a new 200-hp engine -- for $150/hp -- other forms of propulsion start to look pretty good. Heck, you can buy a Suzuki GSX-R1000 motorcycle for around $11,000 and get 160 hp at the rear wheel. That engine costs you about $69/hp, motorcycle included!
Yes, yes; I know that's a serious apples-to-grapefruit comparison. And I'm even coming down from my previous stance that most general-aviation engines are horrendously overpriced. After watching an engine come together in a matter of days -- an engine that I happen to own -- I'm inclined to think of the new cost as merely breathtaking rather than shamefully egregious.
There are two elements to the cost equation, of course. One is the price of the components themselves, which would seem to be very high compared to the technology in use and the amount of time a company should have amortized development costs. The second comes in the labor to assemble the engine. From what we can see of the typical piston engine, it should be a simple matter. Part A into Slot B until you're done.
Not even close.
Before jumping into the building experience, it's worth thinking a bit about the basic parts costs involved. Look at the build sheet for any engine, and the per-item costs might seem outlandish. Exhaust valve: $250. Cylinder assembly: $1200. Crankshaft: $6000. These seem bad enough, but remember that there are more than 1000 individual parts in a typical four-cylinder engine. Some of them are cheap, like $20 for a valve guide, but some of them are staggeringly expensive. How about a simple gear that drives the vacuum pump for nearly $400?
You won't find any design magic in any of these pieces. But they have to be produced to very tight tolerances, of good material, and they need to be carefully inspected at several points along the journey from raw material to a component ready to be put into an aircraft engine.
All of these aspects add to the price, but nothing like the real culprit: low volume. Neither engine manufacturer makes many of its own parts. In fact, on a recent visit to Lycoming's Williamsport, Pa., plant, I saw a lot more assembly and testing than manufacturing from raw materials. That means someone else builds the parts for them.
And while it's no doubt true that the big engine manufacturers hammer their suppliers for every nickel and dime they can get, the nature of things is that there aren't a lot of competing subcontractors who could do the same work for less money. So by the time a supplier gets up to speed and has its parts approved, the costs have gone up. Then the manufacturer gets involved, further boosting the price. You don't think that Continental or Lycoming will take a part from the supplier, perform an inspection and slap it in the appropriate colored box for nothing, do you?
Supply and demand works here, too. Have you noticed that when a part is available from an aftermarket manufacturer that the OEM -- original equipment manufacturer, which should maybe become OEB, for "boxing" -- will often respond with reduced prices to keep sales flowing? This can lead you to a couple of conclusions, both of which may be right: Either the OE has decided to take lower profits to keep volume up, or the parts were dramatically marked up when it was the only game in town.
To Build The Thing
The amazing thing is that the cost of new engines can be so strongly tied to the cost of the components, when they're not at all easy to build.
I spent the better part of a week with the crew at Barrett Precision Engines, in Tulsa, Okla. The purpose was to watch and (partly) participate in the building of an IO-390-X engine from one of Lycoming's new "kits" for Experimentals. Haven't heard much about the program outside of this column? Me neither. I've seen very little marketing on this program, which has created new avenues for kitbuilders to obtain entirely new powerplants in a number of configurations that were only available as overhauls or dramatically expensive factory-new (read: certified) forms. The silence makes me go "hmmm."
|Parts for a Barrett IO-390-X Engine (click photo for larger version)|
In any event, the economics were compelling. No doubt to get the ball rolling, Lycoming priced the new IO-390-X -- a big-bore derivative of the so-called angle-valve IO-360 -- lower than its 200-hp progenitor. I'd call an extra 10 hp for fewer greenbacks a good deal. Moreover, I'd heard from Barrett -- as in the company's founder, Monty Barrett -- that the engine was actually good for more than 210 and that, in his testing, the IO-360 really wasn't a 200-hp engine; more like 190 or so. (I haven't seen his dyno charts on that engine, but he's yet to tell me a tall one so I'll give him the benefit of the doubt on this count.)
The kit engine arrives in a box and the components fill the top of a large workbench. This particular engine came without accessories, which would be sourced through Barrett. They arrive in due course in their own little boxes. No, the size of the box does not indicate the cost.
The first step was to measure everything: bearing surfaces, crankshaft balance, cylinder specifications including bore and choke (that's the amount the cylinder tapers right at top dead center, a necessity in an air-cooled engine that is expected to change shape at operating temperature), cam dimensions, lifter bores ... you name it. Nothing was assumed to be in new-limits specs just because the engine's parts were, in fact, new. This process, all told, takes the better part of a day.
Then it's on to the modifications, which are comparatively few in the Barrett IO-390. The cylinder's ports are cleaned up, mainly just removal of standard casting flash. But no large amount of material is removed; the standard ports flow enough for the engine in anything but racing trim. The valve seats and guides are carefully ground to be concentric and true, a critical element in making them -- particularly the hard-working, sodium-filled exhaust valve -- live for long. When necessary, the valves are hand-lapped to the seats for an ideal seal. There's another half a day for someone who really knows how to use the equipment.
Very Tight Tolerance
|Engine Inspection (click photo for larger version)|
The crankshaft is inspected and balanced to much-tighter tolerances than is permitted by Lycoming. I'm on the fence about the results, as I haven't yet flown behind this engine. On one hand, you have to think that carefully balancing the crankshaft of a four-cylinder engine whose pistons seal nearly 100 cubic inches of displacement each has got to be in the noise of all the dynamic mayhem going on between the cylinders. And yet getting the crankshaft balanced -- it's helped by the standard counterweights on the rear throw -- together with matching the component weights to half a gram on opposing cylinders and no more than a gram all together, can only help. Barrett, ever the hot-rodder, is convinced that a well-balanced engine runs better, makes more power and lasts longer because it's not trying to tear itself apart. (As much.)
I don't consider this a boutique engine. The places where the company diverges from standard practice are modest. There's no sky-high compression ratio or dramatically advanced ignition timing, two ways to achieve big power at the expense of durability.
Rather, Barrett's prides itself on keeping tight tolerances. Lycoming specifies dry tappet clearance between 0.028 and 0.080 inches. That's a relatively wide spread, one that Barrett's techs hit right in the middle, by design. Sometimes this takes a lot of time. For example, on my engine, because of the valve guide and seat work, the valve rides slightly higher in the head than it would normally. This calls for a slightly shorter pushrod to maintain proper clearances. Lycoming produces pushrods in a variety of lengths, and on most of my cylinders they were fine, hitting right in the middle of the range. But a couple didn't cooperate, requiring the grinding of the alloy tube (there's a steel ball on each end) to get the overall length between two of the standard lengths. This process took a couple of hours, but all eight clearances are in the middle of the range.
Here again, you have to ask: Does it really matter? Lycoming set the original specification assuming that the hydraulic unit in each lifter would take up all necessary slack under all normal operating conditions. For most engines, it might not matter. But, somehow, I feel better knowing mine are right down the middle.
Dressed Up for the Show
|Engine Under Construction (click photo for larger version)|
On the last day, the engine was dressed -- fitted with rebuilt, 1200-series, Bendix mags that everyone in the shop called monsters, in contrast to the petite Slicks most buyers take -- an Airflow Performance fuel servo and the sweet smell of new and placed onto Barrett's dyno. His test cell is very good, if a half a generation behind what's going on at GAMI's Carl Goulet Memorial Engine Test Stand in Ada, Okla., and affords the chance to run the engine hard before it ever sees an airframe. I was delighted to push the start button -- an unimpressive little, red button on the test console (I was hoping for a guarded switch and, maybe, a pair of keys that two highly trained crewmembers had to turn simultaneously, a la launching a nuke) -- and hear it roar to life. We watched all the parameters -- EGT, CHT, fuel flow, oil temperature and pressure, the lot -- as the 390's parts became acquainted. Then, in the second hour of running, we turned up the wick and saw, after a bit of leaning, 219 hp flash up on the screen. At a normal, full-rich mixture, it was putting down 215 hp consistently. During break-in, mind you.
After, without benefit of a virtual cigarette or even a strong espresso, I got the privilege of walking to the front office and writing a big check. Although I don't know the breakdown -- how much Lycoming got and how much Barrett's pocketed -- my gut feeling after seeing the efforts in Tulsa is that the men who put my shiny new engine together probably didn't get what they deserved.
It's been the war cry of the super-sophisticates in aviation -- on the Experimental / Amateur-Built side, particularly -- that computer control of aircraft engines is just around the corner, and, for that matter, it's damn well about time. Against this idealistic view of GA comes the chilling fact that FADEC -- so-called full-authority digital engine control -- has been far less successful than hoped. There are a few factory-delivered Aerosance/TCM FADECs in the wild, but more than a few fitted to Experimentals. In fact, I have to admit to some surprise when Mattituck told me that more than 50 FADEC-equipped TMX engines (for Experimentals only) are in the field.
The underlying premise is that computers can make flying simpler, relieving the pilot of the arduous task of managing throttle, prop and mixture controls. (Even, beyond that, cowl flaps and boost pumps and such.) There's just too much fussing, goes the battle cry; we don't have to muck with our cars, so why our airplanes?
I've listened to this argument and slightly dismissed it. I like managing my engine the old-fashioned way. Call it my Type-A personality shining through.
But beyond that, I've had a chance to see what a properly set-up turbocharged engine can handle if you think laterally. The venue was the Cirrus SR22 that Tornado Alley Turbo is developing. I flew with company engineer George Braly during my visit to Tulsa. The short version: Turbocharging a Cirrus is as good an idea as turbocharging a Bonanza or a Cessna 206. Climb performance is maintained, true airspeeds are up -- sometimes way up -- at altitude, and more avenues for avoiding weather are open to you.
But, but, but: Turbo airplanes are hard to manage. Nonsense, at least not for this one, run in the Ada, Okla., way. On one leg of the flight, I touched the mixture control no more than three or four times from takeoff to touchdown. I nudged the combined throttle/prop control twice after setting takeoff power and before arriving in the pattern at the destination.
How could this be? As he and his crew are developing the Turbo Cirrus, Braly has kept to the power-management plan he recommends for the TAT Bonanzas: Full-rich takeoff and climb at high power; high power cruise at lean-of-peak EGT settings; descents as necessary at low power.
In more detail, then: For the takeoff we put the mixture forward and cobbed it. The automatic wastegate maintained 30 inches of manifold pressure and we used full rpm. At some point early in the climb, we reduced to 2650 rpm by nudging the throttle/prop lever (can't just call it the "power" lever, as the red knob has more effect) back a bit; manifold pressure stayed at 30. We kept going up to 17,500 feet on a standard-temperature day with the hottest CHT at 380 °F. Level off, drop prop speed to 2500 -- the throttle is still open full, showing just a bit under 30 inches now -- do the "big mixture pull" to a target fuel flow of 17 gph. Next, you sit there and wait. Nothing happened. The CHTs settled down into the 360 range, the engine ran great at what the Avidyne computer said was 90% power, and we took in the scenery.
About the only changes you might make would be on a hot day to moderate CHTs: lean to make them cooler, enrich if you'd like to go a bit faster and the CHTs will tolerate it.
At the top of the descent, I pulled it back -- at Braly's behest -- to a very low power figure, on the order of 20 inches and 2300 rpm. Leave the mixture alone all the way down and reset it on short final if you really want to. Pilots accustomed to flying this way tend to leave the mixture leaned all the way to shutdown but admonish themselves to remember to grab all the levers if a go-around is needed.
Naturally, this flight dramatically oversimplifies the process of "hands-off" mixture management and requires specific knowledge of the inter-relations of power, fuel flow, cylinder pressures and temperatures, and atmospheric effects to do safely day in and day out. But on this glorious day in Oklahoma -- clear skies, smooth winds -- it just about couldn't get any easier.
FADEC my eye.
Got motors on your mind? Check out the rest of Marc's columns.