Dick Johnson's ASH 26E Flight Test
JNewell
I read in SOARING magazine that the flight testing of the ASH-26E showed a
significant increase in L/D when the turbulators were removed form the
wings. Can anyone explain to me why this was the case? Should the
turbulators be moved to a new location or simply removed?
thanks, John Newell email: new...@njc.org
Matt S. Crane
The way I see it, turbulators cause drag, a bit like wings...but they
*can* be useful ;-).
Seriously, would it be right, though, to say that their beneficial
effect would be most evident at a high angle of attack, as this is when
detatched flow around the aerofoil is at its greatest, or have I got
the smelly end of the stick again?
Matt Crane
Astir CS
PS: May I tentatively start a thread on the pros/cons of winglets at at
circling and cruising flight....or has this been done before?
PPS: How about one on the *exact* criteria on using water ballast
(yeah, I DO have a copy of H. Reichmann's X-country Soaring)
PPPS: And, no I'm not thinking of epoxying plaster winglets to my
Astir!!
M.
Al Bowers
>Experts of the world,
Was that ex-spurts {you all know what an "ex" is and a "spurt" is a
drip under pressure...]...
>The way I see it, turbulators cause drag, a bit like wings...but they
>*can* be useful ;-).
Done right, turbulators _decrease_ drag.
>Seriously, would it be right, though, to say that their beneficial
>effect would be most evident at a high angle of attack, as this is when
>detatched flow around the aerofoil is at its greatest, or have I got
>the smelly end of the stick again?
Not quite. The problem is instability in a laminar boundary layer.
What turbulators are trying to do is to trip the flow, either just
ahead of these areas of instability so that they do not seperate, or
cause the boundary layer to trip in the seperation region (laminar
seperation bubble that reattaches). Without turbulators in these
regions you get laminar seperation which has large drag penalties.
Perhaps this might help:
laminar b.l.: low skin friction, marginal stability
turbulent b.l.: high skin friction, quite stable
You want a laminar run for as long as you can, then you _NEED_
turbulence to prevent seperation drag, the penalty which is _MUCH_
higher than the small increase in skin friction drag.
Note, use of turbulators in the design process is actually another
degree of freedom, so in theory you should be able to produce a higher
performance wing _WITH_ turbulators than without them. That doesn't
mean it is easy.
>PS: May I tentatively start a thread on the pros/cons of winglets at at
>circling and cruising flight....or has this been done before?
I say go for it.
>PPS: How about one on the *exact* criteria on using water ballast
>(yeah, I DO have a copy of H. Reichmann's X-country Soaring)
Goo look at the old Bikle articles on water ballast in Soaring magazine.
11-69 Bikle Eyeball Aerodynamics
6-70 Bikle Polars of Eight
7-70 Bikle Gear Up, Sun Up
8-70 Bikle Description, T-6 Sailplane
10-70 Bikle T-6 Performance
11-70 Bikle AS-W 12 & Libelle H301B Performance Tests
1-71 Bikle Airspeed Calibration
2-71 Bikle Flight Test Performance Summary
6-71 Bikle Polars of Eight
>PPPS: And, no I'm not thinking of epoxying plaster winglets to my
>Astir!!
I think falls under the "easier to buy a new one than mod your old
one." Although with help from people like Peter Masak, that may not
be true...
Al Bowers
Stephen Bell
: Hello,
-
My understanding is that the design of the ASH-26 airfoil (L Booermans, delft)
gives a laminar flow on both surfaces to the trailing edge (even across the flaps!!
I expect a paper on this in "Technical Soaring" as it was presented at the Omarama
OSTIV conference). If this has actually been achieved in the production ash26's then
there is no need for turbulators as their purpose is to force transition from laminar
to turbulent flow in a controlled manner , I would be curious to know where the
turbulators were fitted (factory, or during the test).
Steve.
------------------------------------------------------------------------------
-----
Stephen Bell |
Lincoln University, \--------------------(*)--------------------/
Canterbury,
New Zealand. NIMBUS II - Driver XX
E-Mail: st...@discus.lincoln.ac.nz
Bob Thrasher
>I read in SOARING magazine that the flight testing of the ASH-26E showed a
>significant increase in L/D when the turbulators were removed form the
>wings. Can anyone explain to me why this was the case? Should the
>turbulators be moved to a new location or simply removed?
On a bit different note -
I read this article last night. Wings that weigh nearly 200 lbm each. Three
people (or a special rigging device) required to assemble it. A tail that
weighs 125 lbm. Landing gear that took Mr. Johnson 10 tries to extend and
lock the first time. Early idler bearing failure. Excessive waviness in the
wings. Turbulators and blow holes that decrease performance.
On the positive side - Estimated 50/1 for an 18m ship.
And how much does this thing cost?
Tim Shea
>I read in SOARING magazine that the flight testing of the ASH-26E showed a
>significant increase in L/D when the turbulators were removed form the
>wings. Can anyone explain to me why this was the case? Should the
>turbulators be moved to a new location or simply removed?
According to the article he was unable to find any location at which the
turbulator tape was beneficial. In most cases it hurt. So they removed
it. Also the turbulator holes didn't help, so they disabled them.
Jeff Knell
>I read in SOARING magazine that the flight testing of the ASH-26E showed a
>significant increase in L/D when the turbulators were removed form the
>wings. Can anyone explain to me why this was the case? Should the
>turbulators be moved to a new location or simply removed?
Tim Shea writes:
According to the article he was unable to find any location at which the
turbulator tape was beneficial. In most cases it hurt. So they removed
it. Also the turbulator holes didn't help, so they disabled them.
===================
The turbulator holes were ineffective? Interesting.
Given the expense of building this feature into the
glider, you'd think the factory would have done more
thorough testing.
Revelations like these don't exactly fill you with
confidence. Makes you wonder what else they might
have overlooked.
jeff
Jonker, AS
>gives a laminar flow on both surfaces to the trailing edge (even across the flaps!!
>I expect a paper on this in "Technical Soaring" as it was presented at the Omarama
>OSTIV conference). If this has actually been achieved in the production ash26's then
>there is no need for turbulators as their purpose is to force transition from laminar
>to turbulent flow in a controlled manner , I would be curious to know where the
>turbulators were fitted (factory, or during the test).
Is the data ,coordinates, lift curves etc, avialable for these new airfoils,
or is it the property of the manufacturers???
Thanks
Attie Jonker
\ /
----------O---------- (Std Austria)
ZS-GEM
ma...@aslvx1.sugar-land.anadrill.slb.com
>>My understanding is that the design of the ASH-26 airfoil (L Booermans, delft)
>>gives a laminar flow on both surfaces to the trailing edge (even across the flaps!!
This is complete nonsense. There is no airfoil in existance to my
knowledge that has laminar flow to the trailing edge. The closest
that you might get is if you sucked air into the blowholes.
On the top surface, you normally get laminar flow back to about 40%
of chord on most sailplane airfoils; on the bottom 80-90% if
you are very lucky.
I wonder what effects the removal of the turbulator tape had on
the control effectiveness. It was probably there to avoid laminar
separation when the controls or flaps were deflected.
Regards, PM
Tim Shea
> According to the article he was unable to find any location at which the
> turbulator tape was beneficial. In most cases it hurt. So they removed
> it. Also the turbulator holes didn't help, so they disabled them.
>===================
>The turbulator holes were ineffective? Interesting.
>Given the expense of building this feature into the
>glider, you'd think the factory would have done more
>thorough testing.
>Revelations like these don't exactly fill you with
>confidence. Makes you wonder what else they might
>have overlooked.
Yes I was surprised. Also Johnson said the wing had laminar flow
along almost all of the bottom side but only to .5 - .7 (or something) of
the chord on top -- the .5 was near the spoiler junctions. Anyway
it appears that they got 50:1 LD out of it after disabling all the
turbulator features (versus something like 47:1 in the initial flight).
The tail did seem very heavy to lift, the gear took 10 tries to retract
in flight, and the self-launching propulsion system quit due to a
failed bearing. Schliecher is still using water bags. I'm probably
making things sound more negative than Johnson did in the article.
They seemed to have trouble finding enough calm days to test
performance. However I assume Johnson would only report results
he was relatively confident in.
Todd Pattist
>knowledge that has laminar flow to the trailing edge. The closest
>that you might get is if you sucked air into the blowholes.
>On the top surface, you normally get laminar flow back to about 40%
>of chord on most sailplane airfoils; on the bottom 80-90% if
>you are very lucky.
Peter, why is it that turbulator tapes, if used on the top, always
seem to be farther back than 40%? Also, I recall several Dick Johnson
reports where oil was used to check for laminar bubbles on the wing.
I *thought* the results showed laminar flow across almost the entire
top surface. Do oil tests really work to show laminar bubbles and
laminar to turbulent transition points? Thanks, in advance, for your
comments.
---------------------
Todd Pattist
pat...@ix.netcom.com
Ventus C - N210WH
Charles Crosby
: In article <41vtdv$4...@newsbf02.news.aol.com> jne...@aol.com (JNewell) writes:
: >I read in SOARING magazine that the flight testing of the ASH-26E showed a
: >significant increase in L/D when the turbulators were removed form the
: >wings. Can anyone explain to me why this was the case? Should the
: turbulator tape was beneficial. In most cases it hurt. So they removed
: it. Also the turbulator holes didn't help, so they disabled them.
This could indicate that the hoped-for laminar flow over the hinge
line is NOT achieved, and that the gap seal is probably acting as a
turbulator.
--
Charles Crosby Ph: (27)-(0)-12-4202861 (w)
Dept of Mech and Aero Eng (27)-(0)-12-3451586 (h)
University of Pretoria, South Africa Fax:(27)-(0)-12-436540
cro...@fanella.ee.up.ac.za, cro...@cpc.ee.up.ac.za, meg...@upvm2.up.ac.za
Luke Whitaker
>In article <mgias...@puknet.puk.ac.za>, mgi...@puknet.puk.ac.za
>>(Jonker, AS) writes:
>>My understanding is that the design of the ASH-26 airfoil
>>(L Booermans, delft) gives a laminar flow on both surfaces to the
>>trailing edge (even across the flaps!!
>This is complete nonsense. There is no airfoil in existance to my
>knowledge that has laminar flow to the trailing edge. The closest
>that you might get is if you sucked air into the blowholes.
>On the top surface, you normally get laminar flow back to about 40%
>of chord on most sailplane airfoils; on the bottom 80-90% if
>you are very lucky.
>I wonder what effects the removal of the turbulator tape had on
>the control effectiveness. It was probably there to avoid laminar
>separation when the controls or flaps were deflected.
>Regards, PM
A further point of confusion in this thread is that it is precisly BECAUSE
a wing has laminar flow that turbulator tape is added. It prevents laminar
stagnation/separation by inducing turbulence which brings more energetic air
to the wing surface. When did you see anyone using turbulator tape on a
wooden glider ?
In concurrence with PM, L/D isn't the only parameter that Schleicher
will have been trying to optimise (although 47:1 isn't bad). There's
climb performance, control effectiveness, high speed performance to
mention three. Its only MHO, but I suspect that the designers at
Schleicher do have some idea about how to build a glider.
Luke Whitaker, Essex GC, England. K6BR and ASW15B (yes, both Schleicher).
Charles Crosby
: Somebody else wrote:
: >My understanding is that the design of the ASH-26 airfoil (L Booermans, delft)
: >gives a laminar flow on both surfaces to the trailing edge (even across the flaps!!
: >I expect a paper on this in "Technical Soaring" as it was presented at the Omarama
: >OSTIV conference).
: Is the data ,coordinates, lift curves etc, avialable for these new airfoils,
: or is it the property of the manufacturers???
To answer the first question first:
The airfoil used on the ASH-26 and ASW-27 is supposed to have laminar flow
on the LOWER surface up to 95% of the chord (ie. it stays laminar accross
the flap seal. "Blow-hole" type pneumatic turbulators are used on the
lower surface of the flap to enforce transition at 95%. The upper surface
flow is NOT laminar as far back as the flap hinge line.
To answer Attie's question, the coordinates will probably not be published,
(propriety information), but I believe that the pressure distribution has been.
It is thus possible, if somewhat tedious, to back-engineer it to get the
coordinates.
However, another poster on this thread mentioned that Dick Johnson was
unable to get any advantage from the turbulators, which might indicate that
laminar flow is NOT being maintained over the flap hinge line.
Al Bowers
>>There is no airfoil in existance to my
>>knowledge that has laminar flow to the trailing edge. The closest
>>that you might get is if you sucked air into the blowholes.
Actually, NASA Langley Research Center did test such a thing in a wind
tunnel. However, the energy required to maintain complete laminar
flow ver the entire wing is far in excess of the gain in drag
reduction (ie: you are losing more than you gain). It may have been
Dan Somers on that one, but I can't recall...
>>On the top surface, you normally get laminar flow back to about 40%
>>of chord on most sailplane airfoils; on the bottom 80-90% if
>>you are very lucky.
>Peter, why is it that turbulator tapes, if used on the top, always
>seem to be farther back than 40%? Also, I recall several Dick Johnson
>reports where oil was used to check for laminar bubbles on the wing.
>I *thought* the results showed laminar flow across almost the entire
>top surface. Do oil tests really work to show laminar bubbles and
>laminar to turbulent transition points? Thanks, in advance, for your
>comments.
Oil puddles in areas where you get seperation, and turbulent flow
`scrubs' the oil off thinner than laminar flow does. So you can see
laminar flow, laminar seperation, reattachment, turbulent flow and
finally turbulent seperation with good oil flow. the difficulty comes
in maintaining the flight condition all the way onto the ground
(though oil does take a bit of time to stabilize to a new pattern).
There are other things you can use, sublimating chemicals is one. but
wev'e had really good luck with PGME (poly glychol methyl ethelyne)
and a dye. The PGME evaporates pretty fast (like 30-40 seconds) and
so you only need to hold the flight sondition for that long. The
problem is squirting it out and stabilizing the flight condition that
quick [note: internal plumbing helps :-].
Al Bowers
Jeff Knell
Dick Johnson is not God. The validity of his
flight tests has been debated..
Who evaluates the evaluator?
<criticism of Dick's testing techniques deleted>
================================================
Apparently you believe you are qualified to evaluate
the evaluator. Please elaborate on proper technique
for performace testing of sailplanes.
It would help us evaluate your opinions if you could
include a little information about your performace
testing experience.
jeff
Lou XY
>The turbulator holes were ineffective? Interesting.
>Given the expense of building this feature into the
>glider, you'd think the factory would have done more
>thorough testing.
>Revelations like these don't exactly fill you with
>confidence. Makes you wonder what else they might
>have overlooked.
Or may be there is something wrong with the flight-evaluation..
Dick Johnson is not God. The validity of his flight tests has been
debated..
Who evaluates the evaluator? I found his way of polare measurements
always questionable since we are talking about intrinsinc error rates in
the 20-30% range. Glider comparison flights would reduce this rate.
(to minimize meteorological bias etc..). You think the drag rake is the
heavenly harp?? Get a life.
Lou.
Al Bowers
>Jeff Knell wrote:
>>The turbulator holes were ineffective? Interesting.
>>Given the expense of building this feature into the
>>glider, you'd think the factory would have done more
>>thorough testing.
>>Revelations like these don't exactly fill you with
>>confidence. Makes you wonder what else they might
>>have overlooked.
>Or may be there is something wrong with the flight-evaluation..
>Dick Johnson is not God. The validity of his flight tests has been
>debated..
>Who evaluates the evaluator? I found his way of polare measurements
>always questionable since we are talking about intrinsinc error rates in
>the 20-30% range. Glider comparison flights would reduce this rate.
Really? There are always variable beyond one's control. But for the
bulk of data which has been produced, Johnson is amongst the best.
Comparison data is only as good as the baseline to which it is
measured, and incremental errors can creep in over time. I believe
Johnson's method to be as accurate, if not more so than direct
comparisons.
>(to minimize meteorological bias etc..). You think the drag rake is the
>heavenly harp?? Get a life.
I was under the distinct impression that wake rakes were quite
accurate. The method by Betz and Jones is quite good and is accepted
world wide. That Dick Johnson would take the time to create an
instrument that `averages' and makes the data more presentable
(reducing the engineering required to understand the data) should be
appluaded, IMHO.
Is there a better technique that I am not aware of?
References:
Betz, A.; Ein Verfahren zur direkten Ermittlung des
Profilwiderstandes; ZFM 16, pg 42-44; 1925.
Jones, B.M.; The measurement of profile drag by the pitot traverse
method; ARC RM 1688; 1936.
Montoya, Lawrence C.; Bikle, Paul F.; and Banner, Richard D.: "Section
Drag Coefficients from Pressure Probe Traverses of a Wing Wake at Low
Speeds," Advanced Technology Airfoil Research, Volume I. NASA
CP-2045, Part 2, 1979, pages 601-622.
Bowers, Albion H.; and Sim, Alex G.: "A Comparison of Wortmann Airfoil
Computer-Generated Lift and Drag Polars With Flight and Wind Tunnel
Results," NASA TM 86035, Dec 1984.
Al Bowers
bob gibbons
>> Or may be there is something wrong with the flight-evaluation..
>> Dick Johnson is not God. The validity of his flight tests has been
>> debated.. Who evaluates the evaluator? I found his way of polare
>> measurements always questionable since we are talking about
>> intrinsinc error rates in the 20-30% range. Glider comparison flights
>> would reduce this rate. (to minimize meteorological bias etc..). You
>> think the drag rake is the heavenly harp?? Get a life.
Lou, since Dick can not defend his procedures against your unsupported
criticism, perhaps I should put in a word in defense of the Dick Johnson
/ Dallas Gliding Association flight test procedures. Dick continues to
use the single ship absolute sink rate tests largely for two reasons,
1. It is significantly less expensive in terms of both cost and labor to
use single ship absolute measurements. Relative sink rate testing such
as the Germans use require an additional towplane and personnel to
prepare and launch both sailplanes. DGA has a limited budget and
struggles to support the flight test series as currently configured.
2. The flat land of north central Texas has extended periods of smooth
upper level winds, particularly during the winter months. Our closest
mountains are about 600 miles away. In Germany, on the other hand, wave
activity in the flight test levels is much more frequent. This is one
significant reason the Germans use comparision measurements for their
flight testing. Absolute flat air is not required in relative testing
and thus there are many more days available for testing in near mountain
terrain. This is not a factor in central Texas.
If with this additional understanding of how Dick Johnson and DGA
conduct flight tests, you still feel that comparision testing would be
advisable, I suggest you make a donation to DGA to support these
tests. DGA is chartered as a non-profit scientific and educational
corporation, so all donations to DGA are U.S. tax exempt.
Bob
--
Bob Gibbons Texas Instruments
Internet: Bob.G...@dseg.ti.com Dallas, TX
Mike Cohler
>
> >There is no airfoil in existance to my
> >knowledge that has laminar flow to the trailing edge. The closest
> >that you might get is if you sucked air into the blowholes.
> >of chord on most sailplane airfoils; on the bottom 80-90% if
> >you are very lucky.
>
> Peter, why is it that turbulator tapes, if used on the top, always
> seem to be farther back than 40%? Also, I recall several Dick Johnson
> reports where oil was used to check for laminar bubbles on the wing.
> I *thought* the results showed laminar flow across almost the entire
> top surface. Do oil tests really work to show laminar bubbles and
> laminar to turbulent transition points? Thanks, in advance, for your
> comments.
I was told that the turbulator tape 'maintains' the 'turbulent' flow
further back along the wing, so you put the tape where the turbulent
flow is about to 'separate', and not where the laminar flow is about
to become turbulent.
In other words if you look at the airflow along the wing, then the
laminar flow occurs for the first 40% along the top surface, and
then this flow becomes turbulent but stays close to the wing surface.
At this flow proceeds further back the boundary layer increases in
depth until at some point the flow separates completely from the
wing surface. If you can keep the turbulent layer attached to the
wing surface further back, then the overall lift increases.
Mike
Jens Henkner
>Mike Cohler (md...@unix.york.ac.uk) wrote:
>I was told that the turbulator tape 'maintains' the 'turbulent' flow
>further back along the wing, so you put the tape where the turbulent
>flow is about to 'separate', and not where the laminar flow is about
>to become turbulent.
>In other words if you look at the airflow along the wing, then the
>laminar flow occurs for the first 40% along the top surface, and
>then this flow becomes turbulent but stays close to the wing surface.
>At this flow proceeds further back the boundary layer increases in
>depth until at some point the flow separates completely from the
>wing surface. If you can keep the turbulent layer attached to the
>wing surface further back, then the overall lift increases
Not right. The turbulator tape should avoid laminar separation bubbles.
The flow reattaches normally after such bubbles, but is then
turbulent. On the upper surface of an airfoil you have normally a
turbulent boundary layer before a separation bubble could occur. On
the lower airfoil surface it is different, by tripping the laminar
boundary layer a separation bubble (with a large drag) can be avoided.
You could not avoid turbulent separation with a turbulator tape,
because it is already turbulent. (Suction could be useful for that).
Regards Jens
*-------------------------------------------------*
Jens Henkner Lehrstuhl fuer Fluidmechanik
TU Muenchen Arcisstr. 21 80290 Muenchen
e-mail : he...@flm.mw.tu-muenchen.de
WWW : http://www.flm.mw.tu-muenchen.de/~henky
Tel.: +49 89 2105 8366
Fax.: +49 89 2105 2505
priv: +49 89 917861
Howard Bishop
jmur...@aol.com writes:
> Unfortunately many people take the Johnson flight test as the Grail.
>It is only a tool or guide at best , but I think The testing people have
(sniff)
Waaaa.
Reminds me of how auto manufacturers used to post 1/4 mile times and then
whine and cry when "Road & Track" or someone would test it and find much
different results. Auto makers don't (generally) post performance figures
anymore, but unlike sailplane makers, they don't throw a tiff when
someone has the unmitigated gall to actually (horrors) test their
product. Sailplane makers seem to prefer accepting "calculated"
measurements or comparison glides. Real scientific stuff. "We calculated
300:1 and compared it to old man Smither's ratted out K-8 so we think
those numbers are accurate." puh-leeze.
Of course the old Soaring Symposia articles indicate that a lot of work
is needed to get the most out of a glider. I suspect this is still the
case. Remember the FTE from about ten years ago of the DG300 that came
from the factory with about half of the blow holes clogged? I'd be a
little reluctant to blame anyone, even a Texan, if my delivered product
preformed worse than a factory's demonstration ship.
At any rate, I'm pretty glad that SOMEONE is doing some testing.
Manufacturers have historically put out some pretty suspect numbers
(Speed Astir advertised at 42-1 in 1979, 1-26 pretends to be 23-1, the
Libelle was supposed to be a true 40-1, etc, etc). Somebody needs to
hold their feet to the fire.
hcb
WILLIAM BERLE
Sure, someone needs to hold their feet to the fire, I agree. But
perhaps someone who is not a personal friend of one manufacturer's
chief designer. Or perhaps someone who has the budget to perform tests
with a high degree of accuracy, not a few flights in not so still air.
Johnson is currently the best flight tester we have who is willing to
do the tests and publish them. We will have to live with a few
inconsistencies of his work until someone with a wind tunnel and a
research budget will take his place. I respect Johnson's ability and
willingness to do the tests. I also remember that he agreed with a
friend of his' party line about the location of the static ports under
the wingroot to "keep rain out". It also just happened to make the
airspeed system read higher at the high end and lower at the low end,
so the pilots would think their ship was doing better than it was.
I also think that Johnson should lead off with the testing of a
competition prepared ship, after all this will be the real world we fly
in, and then put in the original factory rough finish ship's data. All
of my sailplanes were noticeably improved after the hours of tinkering
by a good shop. Most everyone else's ships were tuned up too, and this
is the conditions of the race. If a Schempp-hirth glider was 35 to 1
out of the molds, but 40 to 1 after tuning, this is a much more valid
test condition of interest to readers and buyers. No factory production
line ship is in peak tune due to cost. Not a Schleicher, Schempp, or
Schneider.
Furthermore, still air testing is the easiest to do, but actual
racing brings out other factors indeed. Ask any old AS-W20 pilot about
the flexible wings, they made useable lift out of plain turbulence.
That would not have interested Dick Johnson in a flight test, but it
got many of us home on marginal days!
Bill Berle VB
In <434kf3$d...@condor.ic.net> Howard Bishop <bis...@asc-tech.com>
writes:
Larry D. Goddard
: In other words if you look at the airflow along the wing, then the
: laminar flow occurs for the first 40% along the top surface, and
: then this flow becomes turbulent but stays close to the wing surface.
: At this flow proceeds further back the boundary layer increases in
: depth until at some point the flow separates completely from the
: wing surface. If you can keep the turbulent layer attached to the
^^^^ ^^^ ^^^^^^^ ^^^^ ^^^^^^^^^
Mike:
Shouldn't this be "then the overall drag decreases"? May be two sides of
the same coin, but...
Larry Goddard
"01" LS-3a
JMurrayLX
>> Lou, since Dick can not defend his procedures against your unsupported
>>criticism, perhaps I should put in a word in defense fo Dick Johnson /
Dallas >>Gliding Association flight test procedures.
The following is support for criticism of Dick Johnson' s flight tests;
I am obviously not a"non-biased observer"as the US importer for
Schleicher. There are many people world wide who find a lot of fault with
your test procedures. As a representative of one of the manufacturers I
think that often your test results are not contained in what is considered
common sense or common knowledge ,yet you publish it anyway. This seems
unfair. Then if criticzed, your main defense is "lack of funds, we are
doing the best we can.with the resources at hand" The most flagrant
examples which come to mind quickly are as follows.
Feb 1984 - LS4 tests at 36-1! It had just won the last three years
nationals and was beyond reproach as at least 40-1. Later you admitted
that you would retest another ship to see is this example LS-4 was bad.
If you had results this bad which you know not to be accurate, why publish
it? In the same isue you claim that if 16.5 m wing tips are installed on a
Ventus the glide is 50;1. Again, history has proven this quite
optimistic.
Nimbus 3 test for Johnson's personal ship-60-1! Quite impressive! he
certainly enjoys a distinct advantage over the field, why did Schempp then
produce the Nimbus 4 if they already had 60-1? The answer is they did not
have it .
ASH-25 test 52-1. Interesting. Most competitors in the open class
admit that there is a slight difference between the single and two place
ships ,whether Nimbus 3d or ASH-25. The difference is however so small
people will still argure if it is in fact there. I personally think that
the difference is small but real. If you are a very serious world champs
contender get an ASW-22B or a Nimbus 4 . If you also enjoy sharing the
experience with people get a '25 or a '4d. The point is that your tests
show a full 10% difference between the ships. That is very inaccurate. Why
print it?
ASH-26E- You are showing 48 or 50-1. We at schleicher do not agree
with your data. We think it is 50 or 52-1. Time will tell who is correct.
Turbulators- It is interesting that lately you always seem to find a
major fault with the designer's turbulator design. This makes me think
that either all the German designers are in error or perhaps you are. I
think the difference in performance in different locations of turbulators
is less than the scatter in your data .
I strongly offer that perhaps the scatter in your data is and has been
too great to draw the kinds of conclusions you are constantly arriving at.
What you have here is a seriously flawed methodology which lead you to
very inconsistant conclusions.. When asked to defend the conclusions, you
cite the costs involved in improving the methodology. I think a good first
step would be to at least admit that many of your tests are often not what
even you must know to be accurate . Then if that is the case, write a very
strong defense of the product stating that you feel that experience shows
it better than the tests but the test data was as presented. Then the real
step is to change your methodology so that you are not faced with the
glaring inconsistencies that you presently print in "SOARING". If the
Nimbus 3 is 60-1 and you know from racing experience that the 'ASH-25 is
so close to
equal that it is extremely difficult to to beat, then your data showing
10% difference in L/D must be flawed. Why print data which is so flawed
when it has such deep effects onthe manufacturers of ships?
The comment on the net was something like "Turbulator holes
ineffective? one would think that Schleicher would have done more testing
and been more careful. Makes you wonder what else they overlooked."
Unfortunately many people take the Johnson flight test as the Grail.
It is only a tool or guide at best , but I think The testing people have
allowed their results to become "larger than life". As long as this
attitude prevails, the public will continue to take your tests too
seriously and the manufacturers will tend to view your testing as half
crap shoot and half science on a budget.
John Murray
Eastern Sailplane Inc.
David H Noyes
Jens Henkner <he...@flm.mw.tu-muenchen.de> wrote:
>
>The flow reattaches normally after such bubbles, but is then
>turbulent. On the upper surface of an airfoil you have normally a
>turbulent boundary layer before a separation bubble could occur.
Does this mean that separation does not occur on the upper surface?
That is, there is laminar flow, then turbulent flow to the trailing edge?
>On the lower airfoil surface it is different, by tripping the laminar
>boundary layer a separation bubble (with a large drag) can be avoided.
Does this mean that the flow goes from laminar directly to separated on
the lower surface? And that by tripping it to turbulent it remains
turbulent to the trailing edge?
Jens: I have always been puzzled as to why the turbulator tape was on
the *lower* surface but was afraid to ask....... :-)
Leider, bin ich kein Fluidmechaniker :-(
Eric Greenwell
article on the ASH26E, because there have been some important
changes since Don Pollard's glider (the one tested) was built in May '94.
-Later ones, like mine (built Dec. '94), have wing panels that
are much lighter at 160 pounds. My wife and I can assemble mine
without assistance or an assembly dolly. The earlier gliders
were built for a 10 lb/sq. ft wing loading; the later ones for
9.2 lb/sq.ft.
-Even with my 175 lb weight, including parachute, I need only 9
pounds nose weight for the optimum CG. Ballasted, I can legally
carry 19 gallons. Empty weight is 830 lbs vs. Don's 934 lbs.
-My glider has the optional wing tip wheels and steerable tail
wheel, which lets me taxi from the ramp to the runway for
takeoff. The steerable wheel can be removed and replaced with
the standard fixed pneumatic wheel in a few minutes, should I
desire the ultimate performance for a contest.
-The tail is heavy for a good reason. It prevents the glider
from nosing over when full power is applied, as other
designs are prone to do. Good steering is maintained over a wider
range of conditions.
-My landing gear was difficult to retract, also, until I
discovered two over-tightened bolts and a weak assist spring.
Since changing these, I can retract it without difficulty.
-I've done no formal flight testing, but the -26E seemed at least
the equal of the DG-800s when I flew in the Motor Glider
Nationals this year at Minden. Don and I each won a day.
I agree completely with Don's comments. The engine starts quickly and
easily, has less noise and vibration than the Rotax powered models, less
fuel consumption, and no oil mixing with the gasoline. The glide angle
does not degrade as much with the propellor out as it does on gliders
where the engine is also in the slipstream.
I am especially impressed that such an innovative propulsion system has
had so few problems. Schleicher and Heide (the H in ASH26E and the
principal designer) are to be congratulated for this.
Remember that a self-launching sailplane is far more complex than a plain
glider. Talk to a bunch of motorglider pilots and you will discover that
everyone has problems regardless of brand.
The ASH26E has features in additon to the propulsion system that are
important to me:
1) The massive, shock absorbing landing gear that extends the wheel fully
from the fuselage is designed to progressively fail in crash landing to
reduce the stress on the pilot. It also has a Cleveland hydralic disc
brake to provide the best braking possible.
2) I believe the fuselage, a further development of the ASW24 fuselage, is
the most crashworthy design on the market. It is also very comfortable.
3) The 38 degree landing flaps (like the ASW20B and C models) makes
slower, steeper approaches possible, and shorter stops once I touch down.
Handling is excellent because the ailerons go up (washout) as the flaps
go down.
My ASH26E glides and climbs noticeably better than my ASW20C did, so I'n
not concerned about the turbulators or how far back the laminar flow
persists. If there is a problem with the airfoil, I'm sure Schleicher
will fix it. In the meantime, I am enjoying the heck out of it. For my
purposes, there isn't a better motorglider available.
--- WinQwk 2.0b#1462
Al Bowers
I wasn't going to add gasoline to the flames here, but I cannot resist
any longer.
>Sure, someone needs to hold their feet to the fire, I agree. But
>perhaps someone who is not a personal friend of one manufacturer's
>chief designer.
I see Dick Johnson doing a pretty good testing job. He doesn't
_delete_ data unless he has a valid reason to do so, and he only
ignores the data, not deletes it. the scatter is as he publishes, and
you are free to draw your own conclusions based on his data.
Manufacturers have a problem in that they have to sell their product
to survive. And there are usually big steps between a prototype's
performance and what eventually pops out of the mold for the customer
(this gets worse as production number go up). Johnson does have some
in-built biases, but so do all of us.
>Or perhaps someone who has the budget to perform tests
>with a high degree of accuracy, not a few flights in not so still air.
>Johnson is currently the best flight tester we have who is willing to
>do the tests and publish them. We will have to live with a few
>inconsistencies of his work until someone with a wind tunnel and a
>research budget will take his place. I respect Johnson's ability and
>willingness to do the tests. I also remember that he agreed with a
>friend of his' party line about the location of the static ports under
>the wingroot to "keep rain out". It also just happened to make the
>airspeed system read higher at the high end and lower at the low end,
>so the pilots would think their ship was doing better than it was.
Scatter in flight data is usually much worse than what Johnson gets!
He does an exemplary job at what he does. Look at the old data Bikle
published in Soaring back in the early 70s to see how much scatter can
result on a day which feels perfectly smooth. He also does an
airspeed cal to remove the biases introduced by that as well (see his
old report on the std Cirrus B).
> I also think that Johnson should lead off with the testing of a
>competition prepared ship, after all this will be the real world we fly
>in, and then put in the original factory rough finish ship's data. All
>of my sailplanes were noticeably improved after the hours of tinkering
>by a good shop. Most everyone else's ships were tuned up too, and this
>is the conditions of the race. If a Schempp-hirth glider was 35 to 1
>out of the molds, but 40 to 1 after tuning, this is a much more valid
>test condition of interest to readers and buyers. No factory production
>line ship is in peak tune due to cost. Not a Schleicher, Schempp, or
>Schneider.
I thought the test of the original Discus A did exactly that. And the
first results he got were somewhat poorer than after a little prep
work (quick check, from ~39 to 42.5 L/D_max).
> Furthermore, still air testing is the easiest to do, but actual
>racing brings out other factors indeed. Ask any old AS-W20 pilot about
>the flexible wings, they made useable lift out of plain turbulence.
>That would not have interested Dick Johnson in a flight test, but it
>got many of us home on marginal days!
When you figure out how to quantify that on a repeatable basis, let us
all know. But that particular phenomena has been reported many times.
And Schlicher ships are much better at it than the stiffer `Brand SH'.
however, I don't see the `Brand SH" pilots/owners complaining with
quite the same vitriol as I percieve here (or am I taking things the
wrong way?).
Taking flight data is a very thankless task. Reporting is even worse.
the data is the data, but if you disagree, I'm sure that the DGA will
allow you to bring your own prepped ship in and test it. And nothing
is stopping you from repeating the tests to see if you are right or
not. Then you can fair the curve where you want to...
Al Bowers
bob gibbons
JM> The following is support for criticism of Dick Johnson' s flight
JM> tests;
... text deleted ...
JM> The most flagrant examples which come to mind quickly are as follows.
JM> ASH-26E- You are showing 48 or 50-1. We at schleicher do not
JM> agree with your data. We think it is 50 or 52-1. Time will tell who
JM> is correct. Turbulators- It is interesting that lately you always
JM> seem to find a major fault with the designer's turbulator
JM> design. This makes me think that either all the German designers are
JM> in error or perhaps you are. I think the difference in performance
JM> in different locations of turbulators is less than the scatter in
JM> your data .
JM> I strongly offer that perhaps the scatter in your data is and
JM> has been too great to draw the kinds of conclusions you are
JM> constantly arriving at. What you have here is a seriously flawed
JM> methodology which lead you to very inconsistant conclusions.. When
JM> asked to defend the conclusions, you cite the costs involved in
JM> improving the methodology. I think a good first step would be to at
JM> least admit that many of your tests are often not what even you must
JM> know to be accurate . Then if that is the case, write a very strong
JM> defense of the product stating that you feel that experience shows
JM> it better than the tests but the test data was as presented. Then
JM> the real step is to change your methodology so that you are not
JM> faced with the glaring inconsistencies that you presently print in
JM> "SOARING". If the Nimbus 3 is 60-1 and you know from racing
JM> experience that the 'ASH-25 is so close to equal that it is
JM> extremely difficult to to beat, then your data showing 10%
JM> difference in L/D must be flawed. Why print data which is so flawed
JM> when it has such deep effects onthe manufacturers of ships?
I asked Dick Johnson if he cared to comment on your post. This is his
reply.
Prior to the Dallas testing of the ASH-26, Don Pollard performed
numerous still air comparision tests with Stan Nelson's 17.6m Ventus CM,
and they saw essentially equal performance. Various flap settings were
included.
Dick Johnson
bob gibbons
WB> In article <434927$h...@newsbf02.news.aol.com> JMurrayLX,
WB> jmur...@aol.com writes:
>> Unfortunately many people take the Johnson flight test as the Grail.
>> It is only a tool or guide at best , but I think The testing people
>> have
WB> (sniff) Waaaa. Reminds me of how auto manufacturers used to post
WB> 1/4 mile times and then whine and cry when "Road & Track" or someone
WB> would test it and find much different results. Auto makers don't
WB> (generally) post performance figures anymore, but unlike sailplane
WB> makers, they don't throw a tiff when someone has the unmitigated
WB> gall to actually (horrors) test their product. Sailplane makers
WB> seem to prefer accepting "calculated" measurements or comparison
WB> glides.
... text deleted ...
Amen, I am reminded of the late 1960's and early 1970's, you could not
release a new sailplane without claiming at least 5% better performance
than your existing competitors. I seem to recall ASW-15's claiming 42:1
L/D. It was not until Paul Bikle started his flight test series in the
early 1970's that the manufacturers were pulled back to reality, kicking
and screaming, as I recall.
Dick's measurements may not be perfect (a statistical sample of one),
but they are the best we have today.
Bob
Jens Henkner
>Does this mean that the flow goes from laminar directly to separated on
>the lower surface?
After a separation bubble there is normally (not always) a turbulent
reattachment.
> Jens: I have always been puzzled as to why the turbulator tape was on
>the *lower* surface but was afraid to ask....... :-)
>
>Leider, bin ich kein Fluidmechaniker :-(
>
Not everybody can. Imagine that! I' wouldn't live in a world that
consist only of fluidmecanics! (so much for the philosophy)
Tom Ruscitti
watch, especially in this internet vacuum that excludes his active
participation. Soaring is not a poor sport, either in cash or
expertise. If those who care so much about this had instead been
billing (at their chose profession) all the time spent writing about
this, we could buy the DGA 50 more tows plus some calibrated flight
test equipment from Cambridge. Turn off your computers. Send DGA a
check. Go flying.
formerly TR.
--
Tom Ruscitti Phone: 814-863-1276
Statistical Consulting Center FAX: 814-863-7114
Penn State University email: trus...@stat.psu.edu
John Cochrane
soaring article while at region 10 this summer. His answer: Althaus wants to measure
absolute drag; Johnson is only interested in measuring changes in drag, e.g. as flap
settings or turbulators are changed. For that purpose, the errors documented by
Althaus are probably not all that bad, so the previous Johnson results are likely
to be ok.
BTW, I ran into Dick one day at the hotel swimming pool and noticed him intently
studying a piece of paper. It turned out to be the entrant list; Dick was trying
learn the names of all the new contestants. Just one more display of a gracious and
modest attitude all too often missing in this sport, especially among well-known
personalities.
John Cochrane
ThermlRydr
Ruscitti) writes:
> Turn off your computers. Send DGA a
>check. Go flying.
Well said !
Al Bowers
>Dick's measurements may not be perfect (a statistical sample of one),
>but they are the best we have today.
Bob is right. But there are ways to improve the statical sample of
the data you get from a single ship. It isn't easy. I have devoted
some thought to the process by which one gathers data and presents it
to reduce `engineering judgement'. And while the technique I (and my
colleagues) propose may not be perfect, we will have a single
reproducable technique from which to build.
First, using s stopwatch and a dial guage instrument needs to be
replaced. For less than $600 you can pick up a data logger device
which reads analog inputs and records it to digital RAM (12 bit).
Then you need some pressure transducers to attach these to, again
these are simple and make raw output voltage from pressure.
Temprature compensation (Wheatstone bridges) and calibration are
needed to get good results from these things.
For all practical purposes, you can use the aircraft's static sources,
provided you can get a good airspeed calibration. If not, a Kiel
probe and a trailing static line can do this for you (you can think of
this as total energy for your airspeed system).
Once stabilized on an airspeed, data recording can begin. 20 samples
a second is probably sufficient, but the minimum time to get a good
delta (assuming partial glide method here) is pretty long, on the
order of 60 seconds or longer (probably a delta in altitude is better,
like 200 feet or more). Because of `bit flopping' you cannot get a
clean signal of sink rate from sample to sample, and you need to
average out any variation in air mass motion. Several repeats of the
same flight condition (at different times) is a good idea, perhaps 3
to 5 repeats if possible.
Now it starts to get complicated. using the time histories, you need
to determine how good the data is from any particular run. There are
several statistical methods to do this, the one I am most familiar
with is Cramer-Rao bounds; though Dr Gene Morelli at NASA Langley
research center has a new technique I have not yet tried. Cramer-Rao
(CR) isn't a direct measure of the `goodness' of the data, but it is
about the best one can expect from time history data. A good CR
(small uncertainty) doesn't necessarily mean a good data point, but a
bad CR (large uncertainty) does mean a less reliable data point. but
with the mean and the uncertainty of each run, it gives a good feel
for where the data should go. This technique has been used to analyze
flight data for about 20 years by Dr Ken Iliff at NASA Dryden Flight
Research Center, with good results.
Now comes the part that is hard. Until the present, engineering
judgement was used to fair a curve through the data points produced,
with some defference given to the uncertainty of each point. However,
it could be argued that a spline of some sort (cubic spline seems to
hold the most favor) could be faired through the data, and use the
inverse of the uncertainty as the weighting for each data point. this
would remove the engineering judgement required to fair a line through
the data. The data itself would determine it's own goodness or
badness. This is much harder to execute than it is to say though
(hence the long term use of engineering judgement to fair a curve).
Also, I did go back and find the Althaus article in Tech Soaring.
Very well done, though his conclusions about Johnsons' rake are
somewhat in error. Althaus states that the Johnson rake is (direct
quote) "not suitable for optimization of the flap setting." Upon
examining figure 7 (the figure in question) the comparison of the
Althaus rake and the Johnson rake is shown. It is evident looking at
the figure that the minimization point for both rakes is identical.
The relative values and the absolute values are off (though the
Althaus rake is much, much closer), but that isn't the point.
Optimization _IS_ the point. And the errors introduced by the Johnson
rake are quite small for the example given. Perhaps Althaus has other
data which shows differences in the minimization of Cd between his
instrument and that of Johnson, if so it is not presented. That said,
the Althaus rake is a much more elegant and superior design...
Al Bowers ...a nobody in the flight test world...
ma...@aslvx1.sugar-land.anadrill.slb.com
>In article <434927$h...@newsbf02.news.aol.com> JMurrayLX,
>jmur...@aol.com writes:
>>It is only a tool or guide at best , but I think The testing people have
>
the conclusions that Dick draws. Many more flights are needed, and
a statistical based, least squares regression technique should be
used to define the polar. From what I can tell, Dick just pulls
out his french curve and draws his best fit curve by eye.
..
About ten years ago, Rudy Alleman and Eric Greenwell did some
flight testing of sailplanes using the absolute height drop
technique. They got much better results statistically than
Johnson did on the same sailplanes.
It's a shame that they didn't continue their testing.
Peter Masak