Cirrus Accidents
I have an aerospace manufacturing background and when I went to the Cirrus factory last spring, I was impressed with the manufacturing discipline and the redundant fail-safe features of the design.
They have done a lot to assure that the systems are easy to learn and very pilot friendly. This will sell a lot of airplanes. The Cirrus is a delight to fly and most everything about it is attractive.
But the fact that you are told not to spin it, and if you do, to use the ballistic chute, is a serious problem, in my opinion. At best, IFR flying for a low-time pilot is a challenge. Recovery from unusual attitudes is part of the IFR agenda.
Also, many pilots will seek to get their private certificates in this aircraft. Spins in IFR conditions and inadvertent spins in VFR conditions are, at the least, challenging for any pilot. We all know that the FAA/NTSB will label accidents resulting from spins as pilot error.
But what about the FAA error in not insisting on the airplane be recoverable under normal pilot technique? The Cirrus parachute is a great back-up when all else fails, but as a primary recovery from a spin?
In my view, this is dangerous and we will lose a lot of pilots to this design . How many VFR/IFR pilots do you know who will recognize a spin and be able to recover by pulling the parachute before they hit 900 feet AGL? So far, per the current accidents, it doesnt look like this is a good recovery technique.
This aircraft is a tremendous step forward in quality, performance and handling. Why cant we bottom out the spin issue so we dont lose a bunch of folks to a correctable problem? Socata had to spin test the TBM700 per French government demand. Thats a lot more airplane than the Cirrus. Do we want to teach people stalls in an aircraft that cant be spun?
Use your influence as a major aviation voice. Table this problem so we can save lives and make a truly great design and product really safe for the low-time pilots.
-John Cordes
via e-mail
———-
I read with interest your article addressing safety with Cirrus and the ambiguity of safety-related numbers in general. Two aspects of the uncertainty you eluded to are based on small total hours flown and a small number of accumulated accidents.
Probability-based safety information could supplement measured accident rates. For example, if Cirrus had eight accidents in 77,000 hours, the measured rate is about 0.000104 accidents per hour flown. But a different perspective would produce a 90 percent confidence that the accident probability is between about .0000515 and .000201 for an accident in a given operational hour. The same rate with higher total hours (such as 16 accidents in 154,000 hours) would produce a tighter probability distribution, which is a way of showing the numbers are more reliable.
Similarly, if the Mooney Ovation had four accidents in 41,000 fleet hours, one could be about 90 percent confident that the probability of an accident in a given hour is between about .0000366 and .000241, with a 5 percent chance of being below this range and another 5 percent chance of being above.
This broader range is the result of statistics based on smaller numbers. With four accidents in 41,000 hours, the chance of eight or more accidents in the next 77,000 hours is about 59.1 percent, providing nothing has affected the fleet accident likelihood.
This supports your contention that no reliable conclusion regarding relative (Mooney/Cirrus) accident rates can be drawn.
Likewise, if Lancair had zero post-production accidents out of, say, 6000 early fleet hours, then we can be 90 percent confident that the probability per hour of an accident is below about 0.000384.
Because we have no minimum here, we have no way to draw a conclusion that any of these three aircraft has fewer accidents per hour flown than another, based on limited described accident data.
On the other hand, your provided Katana information of 24 accidents in 800,000 hours would suggest a 90 percent confidence range of .000020 to 0.000044 chance of an accident in a given hour. This suggests a lower hourly accident probability than that based on your Cirrus numbers.
Confidence ranges can supplement traditional accident rates that are based on simple division because readers can see how fuzzy the numbers are.
-S. Scott Murray
Fremont, California
———-
Cirrus Speed
Several times in the last few months you have published comparative numbers for the Cirrus Design SR20. For example, in the August issue, you compare the SR20 to the Diamond Star. The numbers you use for the SR20 are 160 knots cruise, 1150 pounds useful load, 2900 pounds gross weight.
Well, I dearly love my early model SR20, but those numbers are not accurate. Since in this article and in earlier articles, you emphasize the fact that you are using real-world numbers and not marketing numbers, here are some real-world numbers:
Cruise: I am consistently about 5 to 8 knots slower than the Cirrus POH. Every SR20 owner I have talked to says theyre anywhere from 3 knots to 10 knots slower than the POH.Cirrus claims that bugs or water on the wing or high weight will cut the cruise speed. Perhaps.
But I know of no real-world SR20 that consistently matches POH numbers. I typically flight plan 150 knots. 160 knots is just not possible without a tailwind.
Gross Weight: Most SR20s have upgraded to 3000-pound gross weight. All new SR20s are at 3000 pounds, I believe.
Useful Load: My airplane, an older C model, has an empty weight of 2080 pounds for a useful load of 920 pounds. Before the gross weight upgrade, our useful load was 820 pounds, which was a significant problem.
The empty weight of SR20s varies from airplane to airplane, but from what I can tell, mine is a relatively light C model. Very early marketing literature listed a useful load of 1100 pounds, but that was back when they thought they could bring the empty weight in at 1800 pounds.
-Brand Fortner
Clarksville, Maryland
———-
And Tiger, Too
This is relative to the Tiger comparison chart on page 7 of the July 2002 issue. Under the column,Cruise @ 65 percent, it appears that the numbers given on six of the eight airplanes are at 75 percent, rather than 65 percent power.
The numbers for the Tiger and C182 are for 65 percent. The numbers for the other aircraft are incorrect. For the Cirrus SR20, the 65 percent power numbers for a pressure altitude of 10,000 feet are 153 KTAS and 10.5 GPH.
The Aviation Consumer does a great job for us. The value is in credibility through accurate reporting. It appears to me that somebody did a poor job of research here.
Did you ever notice the wide variation in fuel flows reported by different manufacturers on like engines at like power settings? My theory is that some manufacturers 75 percent power is more robust than others. At least they are honest about the fuel flow!
-Jim Scott
Wilmington, Delaware
If weve learned anything in 25 years of comparing airplanes, its that nobody seems to agree on speed claims. Our data comes from a combination of real-world experience, owner reports and POH. We agree with you that the Cirrus speed is optimistic. In a future report, well try to pin these numbers down.
———-
Diesel Economics
You guys are so fascinated with the fact that the Thielert engine runs on kerosene that you missed the really important breakthrough that it represents: This is an automotive block in a certified engine.
Even at this early stage, built in small quantity, it is cheaper to purchase outright than brand C orL. The boating industry has used automotive blocks exclusively for their small four-cycle enginessince day one. The cost of customblocks, diesel or gas,haskept them out of all but the most specialized uses.
-Tom Munson
La Mesa, California
On the other hand, for a variety of reasons related to cost, certification, liability exposure and weight, no one has successfully brought an automotive-based engine into the general aviation market. Toyota certified its Lexus V-8 but chose not to build it for aircraft. Yes, economy of scale on the automotive side is attractive. But its hardly a cure all.
———-
Fan Feedback
Your article on avionics cooling fans in the August issue is excellent. Several years ago, I had a Cherokee 180 which, after two hours of flight, would require directing the floor vent upwards to keep the transponder working.
The avionics shop had said no cooling fan was needed. I do not doubt the 30-degree differential between OAT and stack temps, but I wonder if all of that 30 degrees is due to heat generated by the avionics. After all, the top of the instrument panel is usually blackaluminum and is always in the sun.
How much of that heat is solar heat? Would we and the aircraft manufacturers be smart to install some sort of insulation under the instrument panel deck to reduce the solar heat?
If Larry Anglisano still has his heat probegear, he could measure the temperature difference between the cabin andthe instrument stack, with the radios off, and answer this question quickly. Regardless of this next finding, I agree with his conclusion: Avionics fans are cheap insurance for longer life of your avionics. I suggest that insulation will be cooling on the cake.
-Charles E. Truthan
Via e-mail
———-
Datalink Demands
From the land of Billy Bobs Bait and Aircraft (including beer, wine, guns and ammo), a comment on your excellent article, A Flood of Datalink, in the September issue.
You make the appropriate observations that we wont be surprised to see another datalink service emerging before the end of the year and there is almost certain to be a painful shakeout before we know who the reliable, long-term providers of this service will be.
Very true, in an industry coupled to computer technology that jumps a generation every four to six months or so with only more acceleration in sight.
A suggestion on how to identify the winners or, at least, those that have the potential to be: Do a survey of the dozen or so you identified and ask if they used your article, especially the comments above, to motivate their developmental and (particularly) sales and marketing people.
You can bet those that did are good coaches and have a shot at the title.
-Henry Oldham
via e-mail
