Glider performance testing

[[email protected]]

I recently reread Richard Johnson's article, "Sailplane Performance
Flight Test Methods" in the May, 1989 issue of Soaring magazine. In
the article Johnson mentions the Deceleration Measurement method of
determining a performance polar. It had occured to me a couple of
years ago that such a method could be used in conjunction with GPS data
as a fairly simple way of measuring glider performance. I envision
importing the GPS time, altitude, and position data into a spread sheet
application to do the number crunching. Anyone out there ever tried
anything like that? Anyone have suggestions or caveats about doing it?

Myles Bradley

[Bill Daniels]

> wrote in message
ups.com...
>I recently reread Richard Johnson's article, "Sailplane Performance
> Flight Test Methods" in the May, 1989 issue of Soaring magazine. In
> the article Johnson mentions the Deceleration Measurement method of
> determining a performance polar. It had occured to me a couple of
> years ago that such a method could be used in conjunction with GPS data
> as a fairly simple way of measuring glider performance. I envision
> importing the GPS time, altitude, and position data into a spread sheet
> application to do the number crunching. Anyone out there ever tried
> anything like that? Anyone have suggestions or caveats about doing it?
>
> Myles Bradley
>


I don't know if it has been actually tried but I have heard it discussed
several times. If, and it's a big if, you know that the glider can be flown
perfectly level as it slows down, then you measure the rate at which
airspeed is lost, the whole polar can be computed from a single pass.

Back in the 1960's it was proposed to build an "alley" of helium balloons
staked out in two parallel rows exactly1000 feet above a dry lake to avoid
ground effect. The glider pilot would use the balloons as an eyeball guide
in flying a level path. An on-board strip chart recorder would record
airspeed against time as it decayed from Vne to stall.

I don't think it was ever tried. Obviously, there are considerable problems
with this approach not the least is that it will require a dead calm day and
very cool hand on the controls. If there is an advantage it is that only
one tow is needed and higher performance glider take longer to slow down so
resolution improves. I don't see how GPS would help unless, somehow, GPS
altitude could be made to provide centimeter level accuracy. Vertical motion
in the atmosphere would still skew the results.

There is a third potential method with its own difficulties. L/D is the
angle between exact level and the free stream flow. If you had an inertial
reference unit that could establish exact level and a pitch vane on a long
nose probe, the angle of the vane measured against level is the L/D. The
problem is that the angles are very small. (40:1 L/D = 1.43 degrees , 60:1
= .0.95 degrees) and small angular errors generate large L/D errors.

All considered, the timed descent series method that Dick Johnson uses is
the least difficult and if done carefully, produces very good results.


Bill Daniels

[[email protected]]

I'm looking forward to seeing what else comes from this post.

Can't the pressure altitude recorded in the log file be used? The
altitude used doesn't need to be accurate in the absolute sense (does
it?); rather just show the accurate altitude delta over the measurement
window.

On the topic of flight tests, I usually hear crickets when I ask this
question, but does anyone know where the Ventus 2 polar came from in
the SeeYou database (Tools | Polar)?

~ted/2NO

[Gary Evans]

At 00:00 28 February 2006, Bill Daniels wrote:
>
> wrote in message
ups.com...
>>I recently reread Richard Johnson's article, 'Sailplane
>>Performance
>> Flight Test Methods' in the May, 1989 issue of Soaring
>>magazine. In
>> the article Johnson mentions the Deceleration Measurement
>>method of
>> determining a performance polar. It had occured to
>>me a couple of
>> years ago that such a method could be used in conjunction
>>with GPS data
>> as a fairly simple way of measuring glider performance.
>> I envision
>> importing the GPS time, altitude, and position data
>>into a spread sheet
>> application to do the number crunching. Anyone out
>>there ever tried
>> anything like that? Anyone have suggestions or caveats
>>about doing it?
>>
>> Myles Bradley
>>
>
>
>I don't know if it has been actually tried but I have
>heard it discussed
>several times. If, and it's a big if, you know that
>the glider can be flown
>perfectly level as it slows down, then you measure
>the rate at which
>airspeed is lost, the whole polar can be computed from
>a single pass.
>
>Back in the 1960's it was proposed to build an 'alley'
>of helium balloons
>staked out in two parallel rows exactly1000 feet above
>a dry lake to avoid
>ground effect. The glider pilot would use the balloons
>as an eyeball guide
>in flying a level path. An on-board strip chart recorder
>would record
>airspeed against time as it decayed from Vne to stall.
>
>I don't think it was ever tried. Obviously, there
>are considerable problems
>with this approach not the least is that it will require
>a dead calm day and
>very cool hand on the controls. If there is an advantage
>it is that only
>one tow is needed and higher performance glider take
>longer to slow down so
>resolution improves. I don't see how GPS would help
>unless, somehow, GPS
>altitude could be made to provide centimeter level
>accuracy. Vertical motion
>in the atmosphere would still skew the results.
>
>There is a third potential method with its own difficulties.
> L/D is the
>angle between exact level and the free stream flow.
> If you had an inertial
>reference unit that could establish exact level and
>a pitch vane on a long
>nose probe, the angle of the vane measured against
>level is the L/D. The
>problem is that the angles are very small. (40:1 L/D
>= 1.43 degrees , 60:1
>= .0.95 degrees) and small angular errors generate
>large L/D errors.
>
>All considered, the timed descent series method that
>Dick Johnson uses is
>the least difficult and if done carefully, produces
>very good results.
>
>
>Bill Daniels

I have calculated speed to fly/flap postion for my
DG808B by running a pre-planned gps course of 5 mile
per setting in a straight line and calculating L/D
for each speed from the data log. While the actual
L/D may be off slightly due to wind the relative values
should be accurate. This provides the best speed to
fly each flap position. I have also run pure L/D tests
also using a pre-planned gps course over a 15 mile
course in both directions to cancel out wind effect
which is what is loaded into my flight computer polar.

[Bert Willing]

"Bill Daniels" <bildan@comcast-dot-net> wrote in message
. ..
> I don't know if it has been actually tried but I have heard it discussed
> several times. If, and it's a big if, you know that the glider can be
> flown perfectly level as it slows down, then you measure the rate at which
> airspeed is lost, the whole polar can be computed from a single pass.
>
> Back in the 1960's it was proposed to build an "alley" of helium balloons
> staked out in two parallel rows exactly1000 feet above a dry lake to avoid
> ground effect. The glider pilot would use the balloons as an eyeball
> guide in flying a level path. An on-board strip chart recorder would
> record airspeed against time as it decayed from Vne to stall.
>
> I don't think it was ever tried. Obviously, there are considerable
> problems with this approach not the least is that it will require a dead
> calm day and very cool hand on the controls. If there is an advantage it
> is that only one tow is needed and higher performance glider take longer
> to slow down so resolution improves. I don't see how GPS would help
> unless, somehow, GPS altitude could be made to provide centimeter level
> accuracy. Vertical motion in the atmosphere would still skew the results.
>
> There is a third potential method with its own difficulties. L/D is the
> angle between exact level and the free stream flow. If you had an
> inertial reference unit that could establish exact level and a pitch vane
> on a long nose probe, the angle of the vane measured against level is the
> L/D. The problem is that the angles are very small. (40:1 L/D = 1.43
> degrees , 60:1 = .0.95 degrees) and small angular errors generate large
> L/D errors.
>
> All considered, the timed descent series method that Dick Johnson uses is
> the least difficult and if done carefully, produces very good results.
>

All this stuff has been used by the Idaflieg in the 80's, and they now work
on a differential GPS measurement system.

The point with the vane obviously works only if its base is in the free
flow, and if it is the base which is levelled horizontally. If the base is
connected to the fuselage, it will give you the fuselage pitch and nothing
else.
The point with all measurements done without a reference glider is that you
can't cancel vertical airmass movements. That's ok if you are measuring a
2-33, but on anything like a 40+ ship an airmass movement of a few cm/s will
result in substantial differences. An error of 1 cm/s in the sink of a 50:1
ship gives ore takes almost 1 point. A couple of cm/s sink is something you
often have associated with high pressure regions over a wide area (that's
actually how they function).
That is the main reason the Idaflieg measurements are usually trusted far
more than Dick's measurement. Dick is putting an enormous effort in getting
good results, but he just can go so far.

[MickiMinner]

wrote:
> Anyone have suggestions or caveats about doing it?
>
> Myles Bradley

I would contact Dale Kramer from Canada. He has done the same thing
(using computer spreadsheet number crunching) for "Pilot Performance".
I bet if you contact him, you wouldn't have to waste a lot of time
re-inventing the wheel (oops, I meant the measurement criteria, and
number crunching)
Micki

[[email protected]]

wrote:
> I recently reread Richard Johnson's article, "Sailplane Performance
> Flight Test Methods" in the May, 1989 issue of Soaring magazine. In
> the article Johnson mentions the Deceleration Measurement method of
> determining a performance polar. It had occured to me a couple of
> years ago that such a method could be used in conjunction with GPS data
> as a fairly simple way of measuring glider performance. I envision
> importing the GPS time, altitude, and position data into a spread sheet
> application to do the number crunching. Anyone out there ever tried
> anything like that? Anyone have suggestions or caveats about doing it?
>
> Myles Bradley

Reply:
The other columns on the spread sheet will be for airmass movement.
Figure out how to deal with this and it will be as easy as it looks.
This is why Idafleig uses a "master" glider in formation and compares
performance.
UH