Recently Running Out of Elevator

[Bcone1381]

Seriously consider grounding the aircraft until you find something.

[libertyman777]

(4 Place) In the past when coming in for a short landing I have been able to comfortably approach with full flap at 45-47kts. For some reason lately it feels as if I am running out of elevator closer to 55kts. I haven’t made any changes to the aircraft in quite some time. Does anyone have any suggestions?

Either I’ve been flying so many different types lately that my feel is off, something has changed or I’m going crazy.

thoughts?

Sir, we’re you able to resolve this problem?

Thanks!
Paul

[Nev]

I installed VGs under the rear stab and had to do another 5 hours of testing which I completed yesterday. Prior to this I was running out of elevator authority at 50kts when the CG was forward of 14", and low or idle power. This was observable on a steep approach when trying to get down. It could be mitigated by keeping power on and airflow over the elevators, but then I couldn't fly a steep approach at a forward CG.

The VG's made a small but perceptible difference, and I can now easily get to the "back of the drag curve" that previously I could only access with power on. I can now fly a very steep approach in full control, every time. My feeling is this would only effect those aeroplanes with a forward CG (i.e. IO540 engine). My empty CG is 8.3" - there seem to be very few aeroplanes with the CG that far forward. There were no adverse effects from the VG's. The stall speeds appeared to decrease about 1kt across all flap speeds and CG ranges (probably because the wing is now stalling ahead of the elevators), and the pitch attitude of a power on stall increased at very high power settings.

Battson then came flying with me and we spent a couple of hours flying many approaches. It was really good to see the similarities and differences between the A & B models.

[Battson]

Nev - I think it's worth pointing out that for a while now, the airspeeds posted in the landing phase have been several kts faster than they really are, due to the measurement error.

So in reality, you're running out of elevator at somewhere around 42 KTAS, which is perfectly normal for any Bearhawk at forward CG.

[Nev]

Nev - I think it's worth pointing out that for a while now, the airspeeds posted in the landing phase have been around 7 to 10kts faster than they really are, due to the measurement error.

So in reality, you're running out of elevator at somewhere around 42 KTAS, which is perfectly normal for any Bearhawk at forward CG.

Thats a good point Jon, and I definitely need to revisit this. During initial flight testing I tested for position error at cruise speed, then also at my minimum controllable IAS of 40kts. At 40kts IAS I calculated a position error of 2 kts. I did this while flying into a steady headwind, then flying downwind, then repeated it 4 times and averaged the results. But given the large discrepancy between our aircraft (and my OCD curiosity) I think I'll redo that test. I might also get the avionics guys to test my pitot specifically at those low speeds. The part that I don't understand is that if I can fly right on the edge of the stall at 40kts IAS, if there's 7-10kts of position error present then my actual stall speed would be closer to 33-30kts and I'm just not seeing that on any of the ground speed readouts. So it may be somewhere in between.

[jaredyates]

Averaging the speed isn't quite the most accurate when it comes to calculating CAS from ground speed, but that error is fairly small. TAS effects could be larger if you didn't account for those. I'd be really surprised if you were getting a power-off stall speed in the low 30s, less surprised if it involved a little propwash.

[Battson]

Thats a good point Jon, and I definitely need to revisit this. During initial flight testing I tested for position error at cruise speed, then also at my minimum controllable IAS of 40kts. At 40kts IAS I calculated a position error of 2 kts. I did this while flying into a steady headwind, then flying downwind, then repeated it 4 times and averaged the results. But given the large discrepancy between our aircraft (and my OCD curiosity) I think I'll redo that test. I might also get the avionics guys to test my pitot specifically at those low speeds. The part that I don't understand is that if I can fly right on the edge of the stall at 40kts IAS, if there's 7-10kts of position error present then my actual stall speed would be closer to 33-30kts and I'm just not seeing that on any of the ground speed readouts. So it may be somewhere in between.

To me, it looked like the error only occurred in that approach configuration, I assume the cruise ASI is accuarte - but then again, I never checked. At high AoA it appeared to come right as Jared points out. It was really just around that 50kt mark that is was clearly off, which is hard to explain.

Then again, there is a lot of digital wizardry in those black boxes now, hard to know which setting could be affecting it.

[Bissetg]

I think it’s important that we are comparing apples with apples here. I may be wrong, but I think the 4B flap area is larger than the A but the angle of deflection at full selection is less???. I was taught that to determine a true speed you fly an equilateral triangle of 120 degree legs and average the three, therefore eliminating any wind component. This is now way easier with a GPS ground speed than doing time over target calculation. (Showing my age…) This can and should be done for any configuration, cruise, or stall basic through dirty.

We shouldn’t be left guessing about KIAS vs CAS, but what we should be doing is flying an approach at Vs x 1.2 or Vs x 1.2 plus half the gust spread if condx demand it. We do that based on KIAS on the basis that we can cope with variable AUW, wind gradient etc and arrive at the intended destination in a more or less orderly fashion with some consistency, dignity and minimal damage.

The 40 hour test flying phase is in part to establish these numbers. My understanding from Nev was that he had traversed this well in his test flying and established that he had a max of 2 kt discrepancy between his KIAS and CAS. If he has been over reading by 7 to 10 kts then that should have become apparent in the test phase, alternatively if it’s a recent development it needs to be investigated.

Experience has shown that change in characteristics is often the harbinger of trouble. Whether it’s the change of relationship between IAS and CAS or the control effectiveness the change needs to be quantified and the cause determined. If it’s simply fine tuning and improving control effectiveness for a given aircraft configuration then that’s fine too, so long as we understand that that’s what it is. We just need to know, or at least Nev does.

Just my two cents worth but in my experience with my B model the elevator is effective through the speed range and it is a very predictable aeroplane albeit with a much lighter motor than the IO 540 up front and therefore a different C of G.

[svyolo]

Modern EFIS's are stupid easy to check their accuracy. Straight and level, fly straight upwind or down wind at one speed. Do a 180, same speed, the wind (calculated by EFIS) should be the same. If the headwind and tailwind are not calculated as the same, your pitot static is off. Check it at various speeds, including approach speeds. Correcting ground speed/ wind/TAS to CAS are fairly simple calculations that we have all done.

The EFIS and GPS make this all a lot easier than steam gauges. You should be able to assume the GPS ground speeds are correct in steady state, stabilized flight. The rest are calculated with various other inputs.

[Battson]

I thought you had to fly a triangle to account for all the variable, and apply the standard equation?

[svyolo]

With steam gauges, absolutely fly a triange, although if you are using DME for distance they also have errors for slant angle vs ground track distance. The general aviation EFIS's (which I have never flow behind) are pretty simple and I understand what they use because I looked into it extensively. In non-accelerated flight (including not turning), your GS should be accurate, and you can check it against your phone's GPS or a second standalone GPS.
Pitot static and OAT are inputs and assumed correct by the EFIS (even if it is not) and other factors are calculated off of those.

The inertial measurement stuff built in is in the same class as smartphones, and maybe not as good as current year smartphones. But since you can fly IFR off of it for years not, it is good enough.

If you plotted the inertial measured position over time it comes out looking like a flower or drawing of an atom. During those changes in position, the GS (filtered with GPS data) will vary a little, I am not sure how much with a general aviation EFIS's. With commercial stuff it was generally within 0-4 knots variation, but that was flying 4-5 times faster than we do.

I learned on steam for the first 600 hours, transitions to first generation military glass for the next almost 2k hours, which was very accurate and IFR certified. It was awesome, but had little backup other than a couple of small steam gauges. The rest commercial glass which was generally triple redundant.

If I remember right I think someone on this forum diagnosed a Pitot static inaccuracy with his EFIS. I can't remember for sure but I think it might have been Jared.

[Battson]

I think it’s important that we are comparing apples with apples here. I may be wrong, but I think the 4B flap area is larger than the A but the angle of deflection at full selection is less???

Flap defections are more or less the same for the 4-place and Bravo as I understand it, around 50 degrees is specified. The Patrol has less, around 40.

I found it was important to use full flaps or no flaps for slips. Slips at half flap can be risky, as you can lose elevator control in my experience. This has never happened at full flaps.

The plane is also easier to fly on approach and touchdown with full flaps, as there is more drag. This also requires more elevator input as there is more reaction and the plane tends to be slower, so this does impact the elevator effectiveness discussion.

[Bissetg]

Interesting re the slipping. I haven’t found the need for it given the effectiveness of the flaps but still feeling my way. I do like the ball in the middle though.


Knowing Nev I anticipate a thorough testing and analysis of his Pitot Static system and stall speeds in various configurations imminently.

It will be interesting to compare our flap deflections when we all get together next week.

[Battson]

It is amazing how badly one can botch an approach, and still hit the threshold at Vso x 1.1 using a slip with full flaps. She comes down like a stone.
I probably use a sustained full slip once a fortnight, it takes a little practice for me, so I keep hold of the technique.

[Nev]

Airpeed Indicator position error test

I was very curious to see if my elevator authority issue was a result of my IAS over-reading significantly on approach (and therefore flying slower than I realised), or was more the result of the forward CG, so today I re-tested for position error at a range of speeds.

I used a spreadsheet from the National Test Pilots School. The inputs are Groundspeed and Track from 4 headings at a constant IAS. I repeated this for each different IAS. Outputs are calculated TAS, and wind direction and speed, as well as a standard deviation for the data. I flew this at 2000 pressure altitude, with the CG at 12.5" and 1950 lbs AUW.

The results show that for my B model 4-place, (with an IO540 and empty CG of 8.3") an approach speed of 50 KIAS is over-reading by 1.9 kts (TAS 48.1kts). Therefore most of my approaches have been flown at an actual TAS of 48 kts. The VG's have made a small but noticeable improvement to elevator authority at that speed. The position error often present on many aircraft at very high AOA wasn't observed, and the IAS was actually slightly higher than TAS at high AOA.

I've summarized the data in the small table below, and included a screenshot of the full spreadsheet below that.

The model B Bearhawk flies very well for me at this approach speed and leaves me a risk margin. I'm able to see the aim point over the cowl all the way down, and if I get very high on profile I can correct it by either lowering the nose without a large increase in speed, or by raising the nose further. The landing distances seem very respectable and with one or two POB are usually in the range of 100-150m. Overall I'm very happy with the aircraft.

1A741F5E-E88F-4036-9576-34EDBD9FD439.jpg


8850E194-3929-4BF5-AB7C-C9C4C712D667.jpg

[Bissetg]

Ahem, your testing and analysis didn’t disappoint ….