This incident actually has been mentioned in another thread but I think it warrants its own.
Transport Canada has released an accident report of a Mooney M20F which lost control in IMC following the failure of both EFIS Garmin GI275 with which the airplane was equipped. During the upset, the airplane stalled several times at it’s cruising altitude of 15000 ft and in a nose dive exceeded Vne by approximately 70 kts. The pilot was able to regain control at 8100 ft and was able to extract himself from a mountainous area and return to his homebase. The pilot holds an ATPL and beyond 6000 hours.
The aircraft continued to fly normally, even though later inspection found damage to fuel tanks and landing gear deficiencies which were repaired. The airplane was returned to flight after the repairs.
The report goes on to make several comments and recommendations regarding partial panel skills. It also describes the failure occurrence:
the aircraft entered instrument meteorological conditions (IMC). Shortly thereafter, the aircraft’s attitude direction indicator (ADI)Footnote 3 displayed the “AHRS ALIGN” (attitude and heading reference system alignment) message, and indications of attitude (pitch and bank) were lost while indications of airspeed, altitude, and vertical speed were retained (Figure 1).
…..
At the same time, the aircraft’s horizontal situation indicator (HSI)Footnote 4 also indicated a failure, displaying a red X over the HDG (heading) annunciation. The pilot attempted to switch the HSI to the ADI page using the instrument’s touch screen function and selector knob, but was unsuccessful.
So we have a vintage Mooney M20F which has two top state of the art EFIS fail and provoke an incident which could well have cost the guy his life. Maybe it is not such a bad idea after all to keep a vaccuum horizon. At the same time, it is quite remarkable how the airplane came out of this ordeal: 70 kt over Vne which means 242kt IAS is quite a deviation. The pilot never knew this until the recordings of the EFIS were read out (which still recorded multiple parameters despite the failure.)
The question in my mind is what causes the GI275 to fail in this way and why both, while continuing to record flight parameters. I guess most people who own Garmin ADI’s will want to know.
https://www.bst-tsb.gc.ca/eng/rapports-reports/aviation/2021/A21P0001/A21P0001.html
@Mooney_Driver there is a GI275 double failure thread on Beechtalk, I will have a careful read, and it references this report.
I believe the wing spar, and general construction of the Mooney is amongst the strongest in the industry, although perhaps tailplanes may have failed in spiral dive mishandled recoveries.
I would probably keep a turn coordinator, or install a properly independent ADI. This is the typical logic in a glass cockpit where the standby EFIS runs off a separate AHRS and ADC, with an independent battery.
Alternatively a Garmin 345 would provide an independent AHRS to an iPad running ForeFlight.
Crikey.
I don’t think the question of how/why the Garmin units failed is an important as acknowledging that they can and allowing for this in your panel strategy.
People’s excitement that they can “rip out the vacuum system” when installing these units baffles me. Don’t they want redundancy? Any piece of electronics can just die on you, and in mitigating against that having two of the same unit isn’t much of a strategy, especially if they share common inputs.
With a G5 and a vacuum horizon (plus an electric turn coordinator) in our TB10 we have excellent redundancy. Three very different systems, of which the likelihood of them all failing at once is so vanishingly small as to be irrelevant.
@Graham someone I know bought a nice Commander 115 from a 747 pilot, who had helpfully installed four artificial horizons and a turn coordinator.
I kept my vaccum horizon when I upgraded, not least because I read some comments about failure modes of the Aspen here. But also because I needed to keep the vaccum system anyway as my passenger step retracts using vaccum.
I agree that panel strategy has a lot to do with it. Also the acknowledgment that in quite a few cases, partial panel skills have proven to be insufficient in a real life situation and that not only with newbies but also with seasoned veterans. Hence airliners have at least 3 more or less independent horizons. Which as not prevented some accidents to happen in recent years. What I read in many reports is that it is not only full failures which prompt loss of control, but particularly failures which are not apparent as such, at least in the initial stage. This may lead into disorientation combined with loss of confidence in the displays which are there.
Personally, I have the Aspen and a vaccum horizon plus a D1 which I’d have running separately in real life IFR or night flights. 3 is better than 2 exactly for the reason that if you don’t know which horizon is correct, with 3 you have a chance that 2 show the same while the broken one won’t.
As for the Mooney wingspar… so far to my knowledge only one Mooney has ever broken apart in the air, which happened in a thunderstorm many years ago. IMHO that is quite a track record for the M20. I admit it gives a lot of confidence.
which has two top state of the art EFIS fail
More accurately:
“which has two identical firmware version 1.0.0.0 instruments, with the same bugs, same failure modes, same types of AHRS components, and he got rid of his redundancy by removing his vacuum AI, and having spent some 20k on these two boxes, saw them both fail”
I have a collection of photos of Aspen PFDs doing the same thing, in IMC, from the late Stefan in Germany. Never got posted on forums of course, because everybody with kit covered by warranty has to maintain the “dealer relationship”.
Well in all fairness we do not know if it’s a dual ADI failure – as when a second GI275 is installed as primary HSI it will not auto-revert to ADI mode if it detects a fault in its counterpart, nor can the user manually select ADI mode as that function is not available for primary HSI units.
Peter wrote:
I have a collection of photos of Aspen PFDs doing the same thing, in IMC, from the late Stefan in Germany.
That is why one needs a 2nd ADI with a single Aspen. And obviously it would have helped to have one here too.
T28 wrote:
Well in all fairness we do not know if it’s a dual ADI failure – as when a second GI275 is installed as primary HSI it will not auto-revert to ADI mode if it detects a fault in its counterpart, nor can the user manually select ADI mode as that function is not available for primary HSI units.
So then the question arises why this installation did not mandate the installation of a 2nd ADI which is totally independent of the GI275 installation, particularly if they are interconnected.
A dual installation will normally have an auto reversion to ADI in the HSI, but on this occasion they installed the HSI and then the ADI separately?
In theory this installation would be correct from a reg, but not as Garmin intended, if the turn coordinator was retained, but then query why didn’t the ATP pilot keep right side up on a traditional limited panel- there wouldn’t have been the insidious gyro failure of a vacuum system, but a clear CAS message to advise to switch to the turn coordinator.
On the Aspen , the MAX version with SW version 2.11 has different logic which makes it IAS-independent hence supposedly more reliable.
I do have an Aspen but I elected to retain the old Vacuum AI plus the GTX345 IRU displaying on the iPads, plus the turn coordinator.
If you have any intention of flying IMC or at night, attitude display redundancy is way more important for keeping alive than engine redundancy or most any other system redundancy. In my case, even more than electrical system redundancy, since I have the Aspen battery as well as the vacuum AI.
