FWIW, Lancair and Gipsland both lost airframes to their spin test campaigns. The pilots survived in both cases, but the pilot of the GA-8, had trouble getting out and got hit by the prop on his way out... I'd call that a close scrape.

I will try to dig out the paper Imhabe that discusses this.

Spin testing is anything but easy (or cheap).

Some believe, as I do, they wanted to make sure the chute was required equipment. What better way than to make that official by using it as an ELOS.

It is FAR more expensive to certify a chute than to do a few spins, so the "chute as spin ELOS" argument, especially given the plane recovers from spins fine, seems motivated by something other than the cost of spin testing.

Mike C.

One, the FAA doesn't do the flight testing, Cirrus does, so they were not dependent on FAA resources.

Two, if spin testing was going to cause months of delay, that speaks to the incompetence of Cirrus, not the nature of spin testing.


It takes months do to do an ELOS, especially this one which involved researching the history of spin accidents:

http://rgl.faa.gov/Regulatory_and_Guida ... E-96-5.pdf

The ELOS was granted approximately a YEAR before first deliveries.

I find the theory the ELOS was done to speed up deliveries decidedly implausible.

Mike C.

As far as certification goes, the only positive benefit the chute got was to relieve Cirrus of spin testing via the ELOS. Had the ELOS not been granted, Cirrus would have had to do spin testing, and the chute would have been looked at only to make sure it wasn't reducing safety.

Part 23 of the FAA rules don't address having a chute.

Mike C.

I am fairly certain the repack revenue is not the primary motivation.

The motivation is to reduce liability and the theory was, at least prior to production, that having a chute eliminates most fatal accidents. It was easy to look at every fatal accident and find a point where, if they had a chute, the accident would not happen.

Cirrus wanted to be sure the chute was always there to reduce their liability exposure.

As history unfolded, the chute didn't do such a great job of eliminating fatal accidents, one could argue it made it worse, so the theory of the chute and the practical impact didn't match.

Mike C.

Do you know for a fact that Cirrus lost those airframes? I think they did 6ish different deployments- and I believe for the tests they had a system to cut the chute away. I can't find any evidence of the facts though, my information is simply from hanging out with some of the guys who worked there during that time and talking about it.

I recall being told a story by someone who worked on the program at the time of a tense moment where the cutaway system didn't work as intended on the first try. I think in that case, he was successful in getting it cut away.

Don Coburn- were you around Cirrus at that time?

Here are a couple interesting papers on the topic:

In the first one, Mr. Roberts says that the tests were done in two days of flying, but I think that misses the point a little. I don't think that fully accounts for the build up and gradual expansion, I'm guessing he means to go out and fly the tests for certification credit in an airplane that you know will pass.

In the second one, Ms. Englert says that it took 1.5 years to get the LC-400 through the spin certification process and she goes on to detail quite a few changes.

In any case, I don't think Cirrus did the chute because they couldn't pass the spin tests, it was simply that they (Alan and Dale) had made the decision that there would be a chute, period. Why spend the time and money to certify the chute AND spins- when you can get the ELOS and avoid what might turn out to be a 1.5 year ordeal to get the spins sorted?


errr--- and now I'll actually attach the documents

That isn't the time for the spin test, that is failing the test and having to modify and fix the airframe. The test did its job.

We know the SR20/22 did not fail the test, it passed EASA spin testing.


And yet, they did do the spin tests for EASA.

And yet, they are doing the spin tests for SF-50.

Mike C.

Lots of mis-information in this thread.

Regarding the chute being more revenue stream vs technical requirements, that is entirely false. Cirrus went through a spin testing program for European certification, and they were not able to get across the line. The SR line is a relatively hard airplane to get into a spin, but it's also a relatively hard plane to get out of a spin. It was always recoverable but required full down elevator, which was deemed unlikely for an average pilot to apply. So the chute was their path to meeting the spin requirements. The proposed Part 23 reform removes the spin requirement entirely, so perhaps there is an avenue there to remove the chute through STC, but as it stands today to remove the chute you're going to have to come up with some other ELOS avenue to meet spin requirements.

As far as the SF50 meeting spin requirements, I have no inside info, but it's reasonable to assume that the SF50 meets spin requirements just as well as say a PC12 meets requirements. It doesn't. However there are stall protections such as stick shakers, flight envelope protections, etc and the FAA signs off as an ELOS. Have to think that's the path Cirrus is going down. That being said, I've been told the parachute will be required equipment per the POH on the SF50 even though the FAA doesn't care one way or another. So there you could argue its partly motived by revenue, I'm sure there are other good arguments such as limiting liability.

Also I'm not aware of any airframes being lost during parachute testing. For the SR line there was a mechanism to release the chute after deployment. I've read the same approach is being planned for SF50 parachute deployment tests, if they indeed continue forward with that in light of the latest FAA requirements. That doesn't mean its zero risk, just that there is plan to do testing without an airframe loss.

Chris C. comments prompted me to do a little searching and I found the following interesting article:
http://www.kineticlearning.com/pilots_w ... 06_03.html

Tim

Do you know for a fact that Cirrus lost those airframes? I think they did 6ish different deployments- and I believe for the tests they had a system to cut the chute away. I can't find any evidence of the facts though, my information is simply from hanging out with some of the guys who worked there during that time and talking about it.

I recall being told a story by someone who worked on the program at the time of a tense moment where the cutaway system didn't work as intended on the first try. I think in that case, he was successful in getting it cut away.

Don Coburn- were you around Cirrus at that time?

Article on Avweb on SF50:

http://www.avweb.com/avwebflash/news/Ae ... 110-1.html

The part I am most curious about:

"The SF50’s unique automatic flight control system will be an integral part in the CAPS deployment sequence and will intercede to pitch the airplane into slower flight if the pilot commands a deployment outside the system’s airspeed envelope."

This raises some questions:

If the system is smart enough to put the airplane inside the flight envelope at a known attitude and airspeed, why would you want the chute to deploy then? Can the human handle the plane now that the computer "fixed" it?

Why was this needed? The obvious supposition is that at the weight and speed of the SF50, the chute has to be deployed in a very restrictive envelope or it fails.

What if the reason you are deploying the chute is that the flight control system or the instruments it depends on are faulty? How will the system handle that case? Will it chase the faulty data before the chute deploys? What happens if the faulty data means the deployment envelope is never reached?

It sounds like there is definitely software between the chute handle the pilot pulls and the chute deployment itself. Namely, there is some code that handles putting the aircraft into the right attitude before the chute is fired. What happens if that software fires the chute erroneously without pilot command? What happens if that software isn't functioning? Will the chute be disabled or is the deployment envelope procedure bypassed?

What if the aircraft never can attain the proper attitude for chute deployment? Does the chute never deploy? Or is there a timeout?

If the system delays chute deployment to try and achieve proper attitude, could that delay mean the difference between success and failure in chute deployment?

What if the problem is a trim runaway and the autopilot servos are not strong enough to overcome the out of trim condition? The servos MUST be set so the human can overpower them.

What happens if the human overpowers the controls after pulling the chute handle and the deployment envelope is never reached?

What if the failure is structural, a wing or control surface comes off. How can the system put the plane in the right attitude if the plane is uncontrollable? If the software tries to solve that problem for some number of seconds and fails, will the plane may get so far out of the chute deployment envelope that it no longer works?

Does the system depend on electrical, bleed, hydraulic, or other power from the airframe? Must be some electrical needs given the autopilot computer and servos are being used. What happens if that power is missing?

I begin to understand why Cirrus wants the chute to not be required for certification. This allows them to avoid all sorts of system and software certification hurdles that would otherwise be required. The fact there is software in the chute deployment control path is mind boggling to me from the start. It is "chute by wire". I bet there will be cases where pilots call for the chute and DON'T GET IT.

It is going to be REAL interesting what the chute deployment history of the SF50 will turn out to be. For one thing, the customers will be the FIRST to deploy it as Cirrus is not going to do a full up test. Making customers test pilots is downright scary. For another, the complexity of the system is extreme particularly when one considers the failure modes that are possible in all the chute deployment scenarios. And lastly, the chute on the SF50 likely covers far less of the operational envelope of the airplane than the chute on the SR series, hence the control system requirements being implemented.

Mike C.

Mike,

Your points are all very well thought out and presented from an engineering perspective, however for the current Cirri owners who want to upgrade to a jet, can you imagine having to explain to SWMBO that the new plane doesn't have a chute?

I've met quite a few new plane owners and one thing that seems to resonate is they bought what they wanted that would fit their mission. It is pretty apparent that most of them (including a couple of women owners) could realistically spend a lot more without feeling any pain, but elected to stay with a plane they felt capable in as PIC.

Cirrus will sell a lot of jets for those looking for automation and simplicity.

Peace,
Don

What if.....



What if.....

.....there are now a dozen more factors in the SF50 discussion to worry about?

Sounds no harder than having to explain it has only one engine.

Further, you have to explain why, compared to the twin jets, it is so slow, so low altitude, and just as expensive to fly.

You can also explain the safety record for twin jets which is outstanding.

You also have to explain the chute doesn't work near the ground.

If your spouse is incapable of understanding those arguments, you could probably put a button on the panel labeled "CHUTE" and they would feel comforted by that, despite the fact the button does nothing.


Single plus chute is not simpler or easier than twin jet for the pilot. The single plus chute pilot has to deal with the complexity of failures in those areas where the chute does not work (near the ground, for example). Even at higher altitude, engine failure is far more complex for the single than the twin.

Having one engine does not make the plane simpler, nor meaningfully change the pilot training effort. Buying a single jet thinking it will be simpler is a false premise.

Mike C.