Fred,

 

That is a fascinating description of that flight.

 

Based on having run a similar engine with 8.5:1  (TSIO-550) on a test stand with internal cylinder pressure transducers,  I would predict that it would be difficult to avoid detonation under your flight conditions and at 320 Hp at any fuel flow less than about 0.66 to 0.68 BSFC.  That is about  35 to 37 gph.  

 

I don’t know what you planned for enroute fuel flows - -  but that would have been a  detonation challenge at planned fuel flows much less than that, given your OAT and altitude conditions.

 

Regards,  George

 

 

I am delighted to rejoin the LML after an absence of perhaps 5 years during which time I retired, moved to the southwest corner of Australia, built a house and a hangar, started another tech company (very early days yet), got embroiled in the EngineAir V8 deal (which consumed my engine budget) and finally restarted my ancient Lancair IV project which has lain quiescent in the hangar until this weekend.  On Saturday, the fuselage moved to the shop at home and I now hope to finish it within a year.  I have re-subscribed to LML, and watched the chatter with much interest.

 

With regard to the many discussions about compression ratios, lean of peak operation, turbocharging, and such I offer a bit of additional data and commentary.  I rode with Brent Regan when we won the Denver to Oshkosh race in the mid-90’s, and the following year when we got beat.  (Such is racing.) The race was a non-event after our attempt to set a new San Francisco to Denver speed record (overhead to overhead) on the way to Denver.  That was an interesting 2.5 hours, and we cut the attempt short when it was clear that we were 4 minutes off the pace and could not a make it up.  One lesson: don’t try to break a speed record set in winter by flying in summer.  Here’s why…

 

Brent recently wrote of the work with his engine, a modified Lycoming 540 fitted out a lot like a Malibu Mirage, with a lot of Regan specialty items.  Brent chose to use 8.5 compression ratio compared to the 7.3 on the Continental TSIO 550 usually used in the Lancair IV.  The result was the desired power at lower manifold pressure, 35 instead of 38 inches if memory serves correctly. 

 

Brent and I did a lot of flight testing, and Brent did a lot of work on improving cooling while reducing cooling drag since management of engine heat is the major challenge to making horsepower at altitude.  Our design condition was to run SFO-DEN at 27,000 feet, 90% power (about 320 HP) at about 320 knots TAS (Mach 0.53, hard Mach limit of 0.58 for descent).  Aerodynamic heating was about 25F as I recall and had to be incorporated into calculations to obtain TAS and Mach number.  We modeled the entire flight via computer.

 

Here was the rub.  We departed San Jose with full fuel, two of us (not lightweights) and toothbrushes as we sent our baggage via UPS.  The summer temperature aloft was forecast at ISA+20C, that is, equivalent to a 104F day at sea level.  We knew the long climb would burn fuel, and we wanted to start out high, and dive across the SFO VOR to start the clock.  The temperature was high enough that cooling during climb was a problem requiring extra fuel.  When Brent asked if he could circle over the Pacific once more before crossing SFO, I did a quick calculation that showed we were already fuel critical for the mission before crossing the starting line.  So we launched without the altitude benefit.

 

The challenge I faced while Brent flew the airplane was managing the mixture to have enough fuel to land while avoiding detonation.  We watched EGTs and CHTs like a hawk, and as I leaned, I could detect incipient detonation as it started in one cylinder.  Stopping it required immediate enrichment, and jockeying with RPM and manifold pressure while we tried again.  CHTs were right around 400F plus or minus a bit.

 

The airplane was NEVER STABLE the entire trip.  We fiddled and fussed and calculated the entire time, banged along in the high clouds, got a good dose of static charge build up in the cirrus that caused everything to blink and reset to zero (even the clock, it gets your attention) until I could not make the numbers work over western Colorado and we called it quits.  My head stayed down on the engine gages and the calculator for two hours straight as the mixture and power fiddling continued.  We simply did not have enough fuel on board to make the power required for the length of time it would take.  We were trying to operate around 30 gallons per hour (320 HP).  It was not enough for those conditions.

 

After that run, the trip DEN-OSH was a piece of cake.  It was cooler, and the mission was much shorter (starting line being brake release) so we could burn fuel with abandon, and we did.

 

So what was happening?  In light of recent discussions about higher compression ratios to make more power and yield more efficiency (and maybe create problems?) I decided to go back and do some calculations.  They were enlightening.

 

Here is the scoop.  At 27,000 feet, ambient pressure is 10.2 inches of mercury.  The manifold pressure was 32 inches.  However, to this you must add the intercooler pressure loss, normally about 1.5 inches for cruise, bur probably more like 3.5 inches given the high flow rates and very high compressor discharge temperature (higher volumetric flow rates).  The pressure ratio across the compressor (neglecting inlet losses) was then about 3.48.  I don’t have the compressor map for Brent’s turbos, but for a TO3 Garrett which probably has a similar characteristic, one can extrapolate the “islands of efficiency” on the pressure map and guess that the compressor efficiency was probably 60% because the compressor was working way above its optimum pressure ratio.

 

The OAT we measured was ISO+25C, or 8F, including the aerodynamic heating, a bit cooler than forecast but still warm.  Put this air through the compressor and you get a compressor temperature rise of about 332F or 340F discharge temperature for the ambient plus aero heating we were experiencing.  The intercoolers were also way off design point particularly with the warm, thin air at 27,000 feet. Normally intercooler effectiveness (ability to cool) is about 60%, but I would guess we were operating about 50% meaning that the intercooler discharge temperature( which equals the intake manifold temperature) was about 170F, toasty.  

 

Now if you take 170F as dry air and compress it 8.5 to 1 at 90% efficiency (guess for compression inside the cylinder, probably too high) the peak air temperature (neglecting heat transfer) goes over 1400F, WAY over the auto ignition temperature for even hard to light fuels like methane.  In practice, however, we add gasoline which evaporates and lowers the temperature nicely.  Then we pour in MORE gasoline for more cooling, and then we pour in even MORE gasoline so that droplets are evaporating during compression and soaking up heat, reducing temperatures and increasing the effective octane number.  My calculator says we were probably pouring about 9 gallons per hour more gasoline in the engine than it could burn.  It wasn’t enough.  We needed 10 or 11 gallons per hour of excess fuel.

 

So, what does this mean for higher compression and turbo charged engines?  To me it means that if you do not flog the poor beast (90% power, 27,000 feet, ISO+20C) and keep the CHT down, you can avoid detonation.  Just add fuel.   How much depends on how hard you flog.

 

Want to lean it out?  Then put on liquid cooled heads (180F instead of 400F) or back off on the manifold pressure.  If you keep it cool, you will be just fine.  If you let your baffles decay (even good ones are junk, allowing too much leakage), watch temperatures, and recognize that speed costs lots of Avgas, then you will be fine.  If you want fuel economy at high altitude on a warm day while going as fast as you can, fuggetaboutit. 

 

Fast, High, Cheap.  Pick any two. 

 

Glad to be back.

 

Fred Moreno

Out of Hibernation