What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

The Back-Side of the Power Curve (or lack thereof)

Found the backside

A couple of these tests seem to stop just when it's getting interesting...

Tested again today, but the winds were a bit gusty and flying at high power and high AOA takes some practice. The best I got at near zero sink was the new data point at 40 knots shown in the plot below. I don't know how accurate the IAS is at these pitch angles though, at 20 degrees. I did get as slow as 38 knots and 92 HP, but was climbing at 400 fpm:) so I did not count that data point. Normal Vso is 42 knots.

TSVlvOzF6AaTOxSAiq23P8knDrr1oazEagyMpKkNOfrClm658aJRpbaiofZFdDED2tH7ZYP7yqORz07wFynH4kK6AQvg5Oj1tMEbNmoWMmqGYGGP2ot6xXDzPb0iOe778IaVcCIVOnbfScN0RNBxfVQo_15fzhmWdFtxnPfvE0w5-Tlf1PyZf_u1yH-L5AtOkFqP_ZTIItZtW2MexmU2LYW3EPfXd82lsuW-vOEN54PeAdBMuvR0v2MmrxXUMcn6YzBXWxxPsnZOmFY2-NUsp-eDaHblwGbeAvwGUc4fUOjPfaAVgHlyqSdvPCyG42W2uWWWAgEZ99wCa1E7QU6zsGJttl8ykLO0FFc1krRvFxkp6vlrE1HMn4WMVODOxI7svToQPXQqFMV3jrC3IoYBTlgO5QT8wyMEfV9tjecZBcAFdskYolExE9DBkRDVipGkliVPOby7IU5xuQoCJPiFRG-9cTDEEjM7DI_VMzWgIZ6mvVLgwr8_4HU57P9KL6YPGeX9NTMRoCsLG4d6wpMgW5WNtVYlMhSOuKSIda_E-1TIx4HNlOSDtRnIIPzcm4QjfCb3grNS0jRxL6_FITNpcB8aLrrmmkTfmI6ynd6D3jOyTbK79bQn9xKogzCP5HL1Sdxd_cW9rgbHxnDguBnThD-WRD2IPzwEwaXv-jMC9heBwG2ukXH3iIMSnpoE8vhISnQR2tm68B_abHOxGg=w484-h294-no
 
Lots of good flying going on out there.

A few thoughts….
Many of the curves are indeed showing that flattish bottom where the increasing drag is almost identically canceled by increasing thrust. While not linear, when you zoom in to very small regions things can look much more linear considering the full range of speed and power.

While gathering data it is best to stack conditions in your favor, i.e. picking a good day with smooth air and light winds aloft. Then holding each data point long enough to become steady state. If the resulting curve is smooth and continuous then one can be reasonably sure points are good. If you see a single point that falls out of line, it is suspect. If it is the first or last point that jumps, more intermediate point are need to see if it was a trend or anomaly.

If you hit any stall buffet then you’ve gone too far – at least in terms of following your drag polar, you’ll start getting discontinuities. And full stall, of course falls off the chart completely unless flying an Extra 300 or the V-22.

Whether or not you can get to the back side or to the left of min power required is a function of your drag polar. Required CL=sqrt(3 x CD/k). If you are able to generate a lot lift for the drag of the flaps you might get there. Simple flaps don’t make this trade-off very well, so it is a struggle.

What is quite clear is that the charts typically used to describe the phenomenon show a large area we simply can’t reach.

Annotated%20Power%20Curve.jpg
 
The important thing that Van is trying to point out is that you don?t want to unknowingly get behind the power curve. There is an airspeed that at full power you will not be able to climb and if you are low enough, you will most likely impact objects that are in your direction of flight
What folks are finding is that it is very difficult to get behind the power curve in the landing configuration. It is a very tiny sliver of airspeed.


If you have enough power available, you can actually climb at zero airspeed.
In this case you are no longer on the ?power curve?. Consider that if power is velocity x speed and speed is zero?..
Yes the engine is still making power and you are staying airborne, but you are no longer using the drag/thrust curves to derive how much power you need.
 
....I am looking for aircraft to document that can reach reversed command in the normal approach region.

No problem. My old Cessna exhibits greater sink at slower speeds, below something like 80 or 85 mph indicated. It's a good way to control the landing event. I routinely make my solo approach at 70 to 75 mph indicated.

Dave
 
Perhaps the root of the discussion in this thread is that the Lancair 360, I assume optimized for cruise, does not have enough lift to exhibit a "backside" behavior because it stalls at or near the minimum sink speed, as represented by the descent rate vs airspeed plots for that aircraft.

The Lancair is very clean in the cruise configuration and exceptionally dirty in the landing configuration. It can perform landing approaches with up to a 2,000 fpm descent rate.
Our maximum lift coefficients will all be in the same neighborhood. One difference worth noting is that the Lancair has retractable gear. The landing gear adds drag with no lift benefit. If you look at the relationship of CL to CD at minimum power you see that any additional drag pushes the airspeed at minimum power to a lower value. With just simple flaps our maximum lift coefficients are not overly impressive compared with multi-segment fowler flaps.
So the behavior is driven by the relationship of lift and drag coefficients, not just lift the coefficient.
 
No problem. My old Cessna exhibits greater sink at slower speeds, below something like 80 or 85 mph indicated. It's a good way to control the landing event. I routinely make my solo approach at 70 to 75 mph indicated.

Dave

Dave,
Excellent. Do you still have it? If so could you replicate the engine out glide test in the landing configuration and plot descent rate vs airspeed. The closest thing in the chart above is the PA-28
If you don?t still have it, what model Cessna was it?
 
The Lancair is very clean in the cruise configuration and exceptionally dirty in the landing configuration. It can perform landing approaches with up to a 2,000 fpm descent rate.
Our maximum lift coefficients will all be in the same neighborhood. One difference worth noting is that the Lancair has retractable gear. The landing gear adds drag with no lift benefit. If you look at the relationship of CL to CD at minimum power you see that any additional drag pushes the airspeed at minimum power to a lower value. With just simple flaps our maximum lift coefficients are not overly impressive compared with multi-segment fowler flaps.
So the behavior is driven by the relationship of lift and drag coefficients, not just lift the coefficient.

This comment starts to get at the root of the issue. Adding parasite drag makes the "back side" smaller and smaller, in line with the OP's personal observations with his Lancair. The min-sink point moves closer to the stall.

On the other hand, adding induced drag makes the "back side" bigger. Our RVs with low aspect ratio wing should have more separation between min sink and stall, a bigger back-side region.
 
This comment starts to get at the root of the issue. Adding parasite drag makes the "back side" smaller and smaller, in line with the OP's personal observations with his Lancair. The min-sink point moves closer to the stall.

On the other hand, adding induced drag makes the "back side" bigger. Our RVs with low aspect ratio wing should have more separation between min sink and stall, a bigger back-side region.

Steve,
There are a few drag contributors pushing in opposite directions.

A lower aspect ratio wing will indeed have higher induced drag- pushes min power to the right. On the other hand, the baseline drag of the RV is higher ? pushes min power to the left. These terms are all linear.

The bigger contributor is lift coefficient, a squared term in induced drag. Improve CL max and you can really drive up induced drag quickly.
 
Steve,
There are a few drag contributors pushing in opposite directions.

A lower aspect ratio wing will indeed have higher induced drag- pushes min power to the right. On the other hand, the baseline drag of the RV is higher ? pushes min power to the left. These terms are all linear.

The bigger contributor is lift coefficient, a squared term in induced drag. Improve CL max and you can really drive up induced drag quickly.

Well, yes of course, having a higher CL-max will allow you to fly farther into the back-side. But by the way, the wing span is also a squared term. The dimensional induced drag is Lift^2/(pi * q * b^2).

Reduce the wing span and you can really drive up induced drag quickly.
 
For constant airspeed with full flaps and 9 inHg MAP I got:
IAS Descent Rate
80 1400
75 1090
70 900
65 730
60 600
55 550
50 500
45 450
42 300* 1 knot above stall - Verified from Skyview data log

DomgEoXST_jKCnNRqQrwWpH3QkxK9X8fnMnCqV09hpoS_M5HWEIJ_0H9xIJSFocVo23iHxOlI5Ka-v5J8WZ2GHm747sPAv1VA_M_W8Zw4wlSGONl7Snj4R_Uwgx6iwvudUr_gnxZ8b1NAiJUBFqTpfjZktHwZcYB8ifQDXSFiMxrLTxCzOPSu3Z7tw2RSYksKBYRuUnlVcgUxpv-aabDaaYFao2BEcdlc3ekLsWK4i5v-EUC0NQHKNzhxU_aLHgHHAovr4oYbOOxWXfzIz9q1bsf8R9GgY0aEVvgVNz40vXln8p88Y_HODU_1jAEYZvtO5OzLvZRzBXTZAF5VigNppzH9zWKY1XdjqsZWChSwkqP-v1Yf9uKGzRY47NfzgaGR4unVk-hKIOTEsfOmBYHfyUt00V4qQPP8nR0vdISfel2Tt58hEe6TQwU-RQWi5sFmuzlozJuXGr6sUG5NSoypkGiaOVOFYSgsF7PqA8QaFIED4b4F9nezpflEuAriBD-2fwlhF8WUdDzVRsm0wekwkednEj2RjX3iduQZxlFOk-15aOfB_oqfMhX4bUHsPz5dvOlVbFixbLwfL19gAO1fn4rolAd9qdo-A80k2vm2IBmWECQ7-_cYSHatopFapdgEonha4zNf-CR516Sms3B-WXiAFNeBBxQOaEBDZb1xAYqawBHtLZMvdQuodIfPDz6iV0PdEXVmPRzBhDOWw=w484-h289-no

Brice:
Can you explain what you are doing for this test? You note Constant Airspeed in the text (Constant Velocity in the figure) but the table and graph both show a range of IAS. I must be missing something.

Thanks,
Scott
 
Maybe Chris can adjust his gray Lancair 235 line accordingly to show it a little flatter on the bottom end.


Everyone should do this stuff to learn the numbers for their own plane.

John,
The best thing to do is to generate a completely new plot from max flap extended speed all the way down to just above stall. These are qualitative tests to simplify things, but that means we need to gather all the data points in the same flight to eliminate variables.
 
Brice: Can you explain what you are doing for this test?
Not Brice, but I can explain (I posted a short video, above). At idle power, you descend at a series of constant airspeeds and document what your rate of descent is at each different airspeed. You can use your VSI but, for best accuracy, it's better to time your descent through 1,000' of altitude.
 
Sure

Brice:
Can you explain what you are doing for this test? You note Constant Airspeed in the text (Constant Velocity in the figure) but the table and graph both show a range of IAS.

Scott, each data point is a constant IAS glide test where the sink rate is measured. Several data points were collected during a single glide by varying the airspeed and holding. As Snopercod mentioned, it would be best to be established at the airspeed and measure at the same altitude each time. The last data point is very near stall speed and something interesting is happening there. The blue line is data collected by hand while flying. the orange line is extracted from the data log.
 
RV-9A

This comment starts to get at the root of the issue. Adding parasite drag makes the "back side" smaller and smaller, in line with the OP's personal observations with his Lancair. The min-sink point moves closer to the stall.

On the other hand, adding induced drag makes the "back side" bigger. Our RVs with low aspect ratio wing should have more separation between min sink and stall, a bigger back-side region.

I found that with the RV-9A, there is a very small range of airspeed where the power increase is substantial to fly on the backside of the curve. I was expecting a much larger airspeed region where this occurs, but it occurs in the last 3 knots prior to stall. I tested 45 knots and 40 knots, and normal stall speed is 42 knots. I was definitely able to fly slower than 42 knots with added power. That day was a little gusty so it was hard to tell where any stall buffet might have been.

This is not a flight regime I would ever expect to fly the plane. The pitch angle is high, and the plane is on the verge of stalling.

The -9 flaps do not generate much drag. With full flaps (30 degrees) and a fixed pitch prop, I need to be at idle from midfield down wind to touch down at 65 knots or slower to keep from floating too much.
 
Scott, each data point is a constant IAS glide test where the sink rate is measured. Several data points were collected during a single glide by varying the airspeed and holding. As Snopercod mentioned, it would be best to be established at the airspeed and measure at the same altitude each time. The last data point is very near stall speed and something interesting is happening there. The blue line is data collected by hand while flying. the orange line is extracted from the data log.

Thanks Snopercod and Brice. Now I understand.
Scott
 
I found that with the RV-9A, there is a very small range of airspeed where the power increase is substantial to fly on the backside of the curve. I was expecting a much larger airspeed region where this occurs, but it occurs in the last 3 knots prior to stall. .

Brice,
Did you get the sense some flow was already separating? Any buffet?
The typical drag and power curves, especially the theoretical ones, are nice and smooth and predictable because they assume no discontinuities along the way. If you can get part of the inboard wing to stall, as in the buffet region, you are effectively making the flying portion smaller while still dragging the stalled portion along for the ride. Given enough power one could transition into a vertical hover. This would explain the very narrow speed band. Once part wing is separating, the rest has to make up the difference and it doesn't take long for the separated region to grow rapidly.
 
One last datapoint

Did you get the sense some flow was already separating?

Weather cleared and I had another attempt at collecting data at high power and slow speed. I updated the plot below with one other data point at 41 knots. The air was very smooth and I was able to maintain a stable condition for about 10 seconds. There were no signs of buffeting that I could tell. In an attempt to get this number, I did stall it multiple times and it did buffet just before the nose dropped.

From earlier plots, the "backside" flight envelope is marked as beginning at the minimum of the curve, so that puts the region starting around 60 knots for my plane. But nothing dramatic occurs until less than 45 knots or 3 knots above the no-power level stall speed. I did get the plane down to 39 knots IAS again, but with 126 HP it was climbing at 425 ipm still.

furQYuadelP6w6TkLi0aVl5CrzaLdj6WkxanLrlwPQDvAMk3julwoiT_jh5J6wMrd0l2KidtO5yptWWSDXFVBbfFs8ntQdlIMjeIzTSf0qOJSFPQMWwCfLesextzxr2-sZisH7K-BfGE98TIrWMFUpAY-zqfZLmderqywci0fOHU2nTD_2N441NW7HWb_wKQdwtI-xY7Arg_rmAV3uLSatfXjoJy8DAgGVtOu-ELWWmphBfBjKZlKYYJd_liFn4Ouy6Yg4eU_gfbNyFeN-Hi3tSodyVnn5yg-xBlHDsVNEqyd3kuSG23lq-LVENTfQTyWhSB93dlN7AlyjNdkBBnUHTaR9Ht3toxG7EnfmMonJyjmJgm5U9QA3mGdw2c2A98H4EsGdTp9ZwMyivN4nPmpzXlM2yXAFhAifkkIo2ZR2qmbSWvbNeCka3a9TVUC1L0dC8bswUi-2fwrsI3C3fCpHZSIa5pIvykVV-JId70Izv8ExQ0cD7m_WzwrukAMK2l-A41T_-RZ2lGn2sG6LJ8eN5h0VtQvGHDpk6VQK6yzxnxt-b0J3sermvDrimYdHC2kNBUjUucDizgNvEellfqpnS1G15PVbXeCkuAAhsIfGh9kNbKkzNgFFabTx-R91LaQGRCtBIa_N9ffRk-jYpXxD5GNEx-KcV4MZ6Bz7GgrbJ7HES-EG5xYxKN_UcdzBV0lEgjOv0MgDwJKDd5qg=w484-h294-no
 
Dave,
Excellent. Do you still have it? If so could you replicate the engine out glide test in the landing configuration and plot descent rate vs airspeed. The closest thing in the chart above is the PA-28
If you don?t still have it, what model Cessna was it?

It's a 1955 Cessna 180, I still have it, and finding both time to do it and still air is iffy. Here on the lee of the Rockies, virtually any wind includes up and down flowing air.

It would be interesting to do it with different flap configurations. The info I included was based on 40 degrees flap, and flaps are very powerful on that airplane.

Dave
 
One day at a bbq I witnessed a perfect demonstration of the back side of the curve on an RV6. The pilot was a fireman who had worked late the previous night and later confided that he probably should not have flown. It was a tight grass strip with trees all around and he got behind. Of course everyone was watching and I was filming and he just fell out of the sky. The nose came way up and as it did the AOA and sink rate went way up. We heard the power come up just as he hit, but since he was probably 500ft away it took a bit of time for the sound to get to us, but either way he did not quite catch it in time. He hit hard, tailwheel first, then main gear which splayed out. I think there was some green on the prop tips but he got away with it. Scared the **** out of himself and did it in front of 50 people which is always good for the ego.

Of course the Vans wing design will be susceptible to this. It is a short stubbing wing. High induced drag is desirable on approach if you are at the right speed, short wings make for low roll damping so sporty handling, and it makes for lower wing bending moments so a lighter wing than a high aspect ratio wing and the straight planform makes it easy to build. It is a great design, but every design has its particular characteristics. This one is well known and if you are familiar with it it is a non issue. But fall asleep on final and react with stick instead of power and things will get exciting very quickly. This event really stuck in my head. I think most airplanes will do this to a point but I think an RV will do it in a bit more pronounced way. It's a "feature" :D
 
Bryce,
It looks like you were able to make it to the back side. It is great to see folks out there exploring the slow end of the envelope. Too many (especially Lancair) pilots are scared to death by it.

Looking at your data and weight shown (didn?t change between flights), the lift coefficient at 40 KIAS is a way up there ~2.3. There appears to be a little bit of a contradiction between the glide test and the level test at the extreme low end. In the glide, the left most point had power required going down, whereas in level flight it went up. The high power to maintain level flight may be enabling these lower flight speeds ? airflow and direct lift (deck angle is probably ~20 deg).

It also struck me that the low end points could only be maintained for a few seconds? It should be a stable condition. In general it is really difficult to determine if one has a steady state level flight condition in just a few seconds. You are juggling power and trying to get altitude and airspeed to remain constant.

I wanted to rewind real quick back to the initial post about drag vs power curves and their differences and relate back to what you are seeing. I also found some additional theoretical curves on-line that show a good side-by-side comparisons of drag vs power.

combined.jpg


The drag curve minimum is somewhere in the middle of approach speed region. It is not immediately apparent where this point is in that you would need to plot out sink rate vs forward speed or descent angle to find it. Sink rate doesn?t start climbing when passing this point. It is the ratio of D/L that starts climbing.

The power curve is obtained by multiplying the drag curve by velocity. This pulls the left end of the drag curve down dramatically. The minimum of this curve occurs at a much lower speed than minimum drag. The onset of any significant rise in power required in your plot showed how close to stall you had to get.

I think what confuses many is this. Having read about the back-side concern when on approach, they pull the nose up and see the speed decay. The initial thought is that it must be the back-side of the power curve when in fact this is a normal response even on the front side of the power curve. A flatter descent angle requires more power to maintain speed. If power is not added during the angle change, speed decays. The only difference is that it will decay faster on the back-side because you now have an additive effect. What your exercise showed is how close to stall your aircraft has to be before the induced drag portion kicks in as a significant contributor. Normal approach speeds should keep you well clear.
 
I would love to see a video of a 172 in level flight indicating 15-20 KIAS. Unfortunately, once you hit buffet or stall you are not operating on the power curve being discussed.

I've been at 25-30 knots in my Cessna at level flight. I'll get video of it next time I do it.
 
Power curve video

I thought this would be an appropriate 4 minute video for this thread. It demonstrates an RV flying formation with a float plane. It appears that both pilots are fairly experienced and thus could happen to anyone. The float plane pilot got the airplane to slow on the approach and to late on the throttle.

https://www.youtube.com/watch?v=fC5yscm9dsI
 
I thought this would be an appropriate 4 minute video for this thread. It demonstrates an RV flying formation with a float plane. It appears that both pilots are fairly experienced and thus could happen to anyone. The float plane pilot got the airplane to slow on the approach and to late on the throttle.

https://www.youtube.com/watch?v=fC5yscm9dsI

I remember seeing this a while back. Despite the title of the video, there is no way to tell if it really had anything to do with the 'back side' or if it was simply an approach that was just too slow. Normal approach speeds (1.3 Vso) have far more energy than needed to round out and flare from any descent.
 
I thought this would be an appropriate 4 minute video for this thread. It demonstrates an RV flying formation with a float plane. It appears that both pilots are fairly experienced and thus could happen to anyone. The float plane pilot got the airplane to slow on the approach and to late on the throttle.

https://www.youtube.com/watch?v=fC5yscm9dsI

The low wing airplane is not an RV. Despite the grainy video, the landing gear is in the wrong location on the wing, and the shape of the rudder and vertical stabilizer is wrong. Perhaps it is a Zenith?

Not sure if both pilots are fairly experienced. Something about not flying within 500' of a vessel or person. I doubt they had a waiver.

And definitely not "back side of the power curve". The floatplane was too slow with no power on. If it was "back side..." power would have been fully on.
 
It's all about energy...

I just posted some AOA and energy management training resources in the transition training thread at the top of this page. If you're interested in the back side of the power curve and energy management, you might find some of the subject matter helpful. There are embedded videos. Since the AOA system we are using is tone-based, even if you don't have an AOA system in your airplane, it's easy to follow along in the videos.

Cheers,

Vac
 
Back
Top