Losing time in Cloud-Streets?

[noel.wade]

All -

I've been flying with the assumption that flying slow in cloud-streets
is always a good idea. But in looking at some of my OLC flights I
wonder if my logic is faulty and I'm hurting my speed. Anyone with
lots of OLC or contest experience care to offer feedback? Details
below...

My thinking has always been that your forward speed in a circling
climb is effectively "0" (slightly more or less than 0, based on wind;
but call it 0 for the sake of argument). As a corollary, anything you
can do to maintain or gain altitude while retaining forward progress
on-course is a "win".

Therefore, I've always been slowing down in cloud-streets or under
large clouds with apparent extended areas of lift. If I'm anywhere
near cloudbase (say, less than 1000' or so), my logic has been that
I'm not gaining enough altitude to stop and circle with that zero
forward speed. Instead I try to maximize the lift by flying straight
ahead (or slightly curving my path to stay in the zone of lift) at
slow speed (say 50 kts).

Rough math to support my thoughts:
With a cruise speed of 75 knots and a time-in-climb around 25%,
average forward speed is roughly 56 knots (75% x 75 knots). This is
so close to the 50 knots (while flying in a cloud-street) that it
seems to support my behavior... But maybe I'm missing something?
Maybe the gains I'm getting are not enough to offset the slower cruise-
speed during these times? I should point out that I fly in pretty
strong thermal conditions at my local site (Ephrata, WA).

I could throw in all of my various caveats and details and thought-
processes, but rather than make this a really long post I'd love to
get some thoughts and feedback. Anyone have pointers, ideas, see
fallacies in my process, etc?

Thanks,

--Noel

[Tony[_5_]]

On May 18, 4:33*pm, "noel.wade" > wrote:
> All -
>
> I've been flying with the assumption that flying slow in cloud-streets
> is always a good idea. *But in looking at some of my OLC flights I
> wonder if my logic is faulty and I'm hurting my speed. *Anyone with
> lots of OLC or contest experience care to offer feedback? Details
> below...
>
> My thinking has always been that your forward speed in a circling
> climb is effectively "0" (slightly more or less than 0, based on wind;
> but call it 0 for the sake of argument). As a corollary, anything you
> can do to maintain or gain altitude while retaining forward progress
> on-course is a "win".
>
> Therefore, I've always been slowing down in cloud-streets or under
> large clouds with apparent extended areas of lift. *If I'm anywhere
> near cloudbase (say, less than 1000' or so), my logic has been that
> I'm not gaining enough altitude to stop and circle with that zero
> forward speed. *Instead I try to maximize the lift by flying straight
> ahead (or slightly curving my path to stay in the zone of lift) at
> slow speed (say 50 kts).
>
> Rough math to support my thoughts:
> With a cruise speed of 75 knots and a time-in-climb around 25%,
> average forward speed is roughly 56 knots (75% x 75 knots). *This is
> so close to the 50 knots (while flying in a cloud-street) that it
> seems to support my behavior... *But maybe I'm missing something?
> Maybe the gains I'm getting are not enough to offset the slower cruise-
> speed during these times? *I should point out that I fly in pretty
> strong thermal conditions at my local site (Ephrata, WA).
>
> I could throw in all of my various caveats and details and thought-
> processes, but rather than make this a really long post I'd love to
> get some thoughts and feedback. *Anyone have pointers, ideas, see
> fallacies in my process, etc?
>
> Thanks,
>
> --Noel

that sounds right to me. although i spend more like 30-40% of my time
thermalling. i always do my best to fly along at min sink or so
straight ahead in any sort of streetlike lift i find. not only do you
make forward progress but you also are gaining altitude. so now you
have gone say 5 or 10 miles in a street and you are at least at the
same altitude you started but probably 1000 or more feet higher. this
could be the equivalent climb that you would've had if you would've
stopped and circled right away except you have now made more progress
on course and still have the altitude to convert into speed.

[ryanglover1969[_2_]]

On May 18, 2:33*pm, "noel.wade" > wrote:
> All -
>
> I've been flying with the assumption that flying slow in cloud-streets
> is always a good idea. *But in looking at some of my OLC flights I
> wonder if my logic is faulty and I'm hurting my speed. *Anyone with
> lots of OLC or contest experience care to offer feedback? Details
> below...
>
> My thinking has always been that your forward speed in a circling
> climb is effectively "0" (slightly more or less than 0, based on wind;
> but call it 0 for the sake of argument). As a corollary, anything you
> can do to maintain or gain altitude while retaining forward progress
> on-course is a "win".
>
> Therefore, I've always been slowing down in cloud-streets or under
> large clouds with apparent extended areas of lift. *If I'm anywhere
> near cloudbase (say, less than 1000' or so), my logic has been that
> I'm not gaining enough altitude to stop and circle with that zero
> forward speed. *Instead I try to maximize the lift by flying straight
> ahead (or slightly curving my path to stay in the zone of lift) at
> slow speed (say 50 kts).
>
> Rough math to support my thoughts:
> With a cruise speed of 75 knots and a time-in-climb around 25%,
> average forward speed is roughly 56 knots (75% x 75 knots). *This is
> so close to the 50 knots (while flying in a cloud-street) that it
> seems to support my behavior... *But maybe I'm missing something?
> Maybe the gains I'm getting are not enough to offset the slower cruise-
> speed during these times? *I should point out that I fly in pretty
> strong thermal conditions at my local site (Ephrata, WA).
>
> I could throw in all of my various caveats and details and thought-
> processes, but rather than make this a really long post I'd love to
> get some thoughts and feedback. *Anyone have pointers, ideas, see
> fallacies in my process, etc?
>
> Thanks,
>
> --Noel

I've been thinking the same thing. I try hard to minimize the time I
spend circling. This year I'm going to slow down in lift, but not as
much as I have in the past...maybe to 60-65 knots. And I'm going to
strive harder to fly faster between thermals....say 75-80 knots on a
4-5 knot day. Ask yourself....Do I land out often? I'm guessing you
don't. I do think 50 knots is too probably too conservative under a
street. I may try to set my ring to the average expected lift under
the street, so that would be even faster than 60-65 knts....humm. Let
me know how it goes...

[Tony[_5_]]

i guess it is worth noting that my average XC speed, no wind, is
slower than minimum sink speed for my glider.

[John Cochrane]

On May 18, 4:33*pm, "noel.wade" > wrote:
> All -
>
> I've been flying with the assumption that flying slow in cloud-streets
> is always a good idea. *

I disagree. It's always MacCready theory, if correctly interpreted.

Suppose the cloudstreet offers 2 knots up everywhere with occasional 8
knot cores. (Uvalde). Option 1: fly 50 knots without losing altitude.
You average 50 knots = 57 mph. Option 2: fly 85 knots, (I.e. cruise at
Mc 4 or so) , slowly losing, but stopping to thermal in the 8 knot
cores. You average 80 mph. (Typical uvalde winning speeds) The latter
is much faster. Reichmann offers the same advice.

John Cochrane

[Dave Nadler]

On May 18, 5:33*pm, "noel.wade" > wrote:
> All -
>
> I've been flying with the assumption that flying slow in cloud-streets
> is always a good idea. *But in looking at some of my OLC flights I
> wonder if my logic is faulty and I'm hurting my speed. *Anyone with
> lots of OLC or contest experience care to offer feedback? Details
> below...
>
> My thinking has always been that your forward speed in a circling
> climb is effectively "0" (slightly more or less than 0, based on wind;
> but call it 0 for the sake of argument). As a corollary, anything you
> can do to maintain or gain altitude while retaining forward progress
> on-course is a "win".
>
> Therefore, I've always been slowing down in cloud-streets or under
> large clouds with apparent extended areas of lift. *If I'm anywhere
> near cloudbase (say, less than 1000' or so), my logic has been that
> I'm not gaining enough altitude to stop and circle with that zero
> forward speed. *Instead I try to maximize the lift by flying straight
> ahead (or slightly curving my path to stay in the zone of lift) at
> slow speed (say 50 kts).
>
> Rough math to support my thoughts:
> With a cruise speed of 75 knots and a time-in-climb around 25%,
> average forward speed is roughly 56 knots (75% x 75 knots). *This is
> so close to the 50 knots (while flying in a cloud-street) that it
> seems to support my behavior... *But maybe I'm missing something?
> Maybe the gains I'm getting are not enough to offset the slower cruise-
> speed during these times? *I should point out that I fly in pretty
> strong thermal conditions at my local site (Ephrata, WA).
>
> I could throw in all of my various caveats and details and thought-
> processes, but rather than make this a really long post I'd love to
> get some thoughts and feedback. *Anyone have pointers, ideas, see
> fallacies in my process, etc?
>
> Thanks,
>
> --Noel

A few notes:
- the glider climbs much better going straight than circling, and can
fly slower
- you can fly through the cores when you fly straight, you circle
around them
- streets have narrower cores
- all this depends on the glider (harder to avoid circling as
performance decreases)

A few weeks ago I flew 440 miles, around 15% circling, mostly during
the beginning of the flight and digging out of one hole (flew under OD
at 2nd turn):
http://www.onlinecontest.org/olc-2.0/gliding/flightinfo.html?flightId=502072323

Suggestion: Go look at the log files of the fastest pilots during last
years 18m nationals at Ephrata !

Hope this is helpful,
See ya, Dave "YO electric"

[Tim Taylor]

On May 18, 3:33*pm, "noel.wade" > wrote:
> All -
>
> I've been flying with the assumption that flying slow in cloud-streets
> is always a good idea. *But in looking at some of my OLC flights I
> wonder if my logic is faulty and I'm hurting my speed. *Anyone with
> lots of OLC or contest experience care to offer feedback? Details
> below...
>
> My thinking has always been that your forward speed in a circling
> climb is effectively "0" (slightly more or less than 0, based on wind;
> but call it 0 for the sake of argument). As a corollary, anything you
> can do to maintain or gain altitude while retaining forward progress
> on-course is a "win".
>
> Therefore, I've always been slowing down in cloud-streets or under
> large clouds with apparent extended areas of lift. *If I'm anywhere
> near cloudbase (say, less than 1000' or so), my logic has been that
> I'm not gaining enough altitude to stop and circle with that zero
> forward speed. *Instead I try to maximize the lift by flying straight
> ahead (or slightly curving my path to stay in the zone of lift) at
> slow speed (say 50 kts).
>
> Rough math to support my thoughts:
> With a cruise speed of 75 knots and a time-in-climb around 25%,
> average forward speed is roughly 56 knots (75% x 75 knots). *This is
> so close to the 50 knots (while flying in a cloud-street) that it
> seems to support my behavior... *But maybe I'm missing something?
> Maybe the gains I'm getting are not enough to offset the slower cruise-
> speed during these times? *I should point out that I fly in pretty
> strong thermal conditions at my local site (Ephrata, WA).
>
> I could throw in all of my various caveats and details and thought-
> processes, but rather than make this a really long post I'd love to
> get some thoughts and feedback. *Anyone have pointers, ideas, see
> fallacies in my process, etc?
>
> Thanks,
>
> --Noel

Noel,

The best way to think about this is to do a few thought experiments.
First you need to say what is the average climb you expect for the day
that you are willing to take, lets say 5 knots. What is the speed you
can achieve with that if you were in pure MacCready flight mode.

There are side by side cloud streets one with a 5 knot thermal every
10 miles and one one where you can fly along at 50 knots for the whole
flight. Which one will get you home faster? What speed would you fly
between the thermals on the one street.

Now combine the two together. How would you fly differently? You can
fly faster than you have in the low lift under the street, but take
advantage of the stronger cores every few miles by being below the
clouds enough to make it worth while to stop and turn in the strong
stuff. If it is a really good street you fly faster but only pull up
in the strong cores and climb back to cloud base without turning.

Reichmann covers this well in the section on Dolphin flight models and
concludes:

"For dolphin flight a new theory results which contains the "classic"
theory as a special case. The speed ring and similar devices retain
their validity. Thus, according to the theory, the ring is set
exactly the same for "classic" flight (apart from possible tactical
considerations, of course). It has been proven that "force" dolphin
flight causes cruise-speed losses."

So you speed ring or speed to fly vario is going to tell you to slow
down in lift but not all the way to climb speeds if you are cruising
with a MC 5 and your in one knot lift.

[noel.wade]

A couple of things:

1) Thanks John, that's the kind of comparison/number-checking I was
looking for. I know that cloudbase/working-band/lift-strength play a
factor in determining whether the tradeoff is worth it; but didn't
think to do a straight MacCready comparison (the idea of pushing the
glider to high speed while still in spotty lift under a cloud-street
is just still feels awkward - discounting the few times where I've
approached cloudbase and needed to "bug out")!

2) I thought flying "energy lines" was the consensus these days, and
"Dolphin flying" was out - given that variometers lag and its
incredibly hard to time your pushes and pulls when flying through
alternating narrow cores and strong sink. If your strong cores are
separated by a couple of miles, their miniscule contribution (say a 50
- 150 foot height gain) when you fly straight through them doesn't
seem worth it (especially when you start considering the risks of
hitting nasty sink before accelerating back up to cruise speed;
possibly wiping out all of your gains). I've always been much more
tempted to slow down under weak spread-out lift (for example, under a
cloud thats about to OD).


Now just to turn this around and look at it from a different
perspective: For the fast guys out there, when (if ever) *do* you
start slowing down a lot while flying straight ahead? What're the
combination of factors that encourage you to switch into this mode?

Thanks,

--Noel

[[email protected]]

On May 18, 5:33*pm, "noel.wade" > wrote:
> All -
>
> I've been flying with the assumption that flying slow in cloud-streets
> is always a good idea. *But in looking at some of my OLC flights I
> wonder if my logic is faulty and I'm hurting my speed. *Anyone with
> lots of OLC or contest experience care to offer feedback? Details
> below...
>
> My thinking has always been that your forward speed in a circling
> climb is effectively "0" (slightly more or less than 0, based on wind;
> but call it 0 for the sake of argument). As a corollary, anything you
> can do to maintain or gain altitude while retaining forward progress
> on-course is a "win".
>
> Therefore, I've always been slowing down in cloud-streets or under
> large clouds with apparent extended areas of lift. *If I'm anywhere
> near cloudbase (say, less than 1000' or so), my logic has been that
> I'm not gaining enough altitude to stop and circle with that zero
> forward speed. *Instead I try to maximize the lift by flying straight
> ahead (or slightly curving my path to stay in the zone of lift) at
> slow speed (say 50 kts).
>
> Rough math to support my thoughts:
> With a cruise speed of 75 knots and a time-in-climb around 25%,
> average forward speed is roughly 56 knots (75% x 75 knots). *This is
> so close to the 50 knots (while flying in a cloud-street) that it
> seems to support my behavior... *But maybe I'm missing something?
> Maybe the gains I'm getting are not enough to offset the slower cruise-
> speed during these times? *I should point out that I fly in pretty
> strong thermal conditions at my local site (Ephrata, WA).
>
> I could throw in all of my various caveats and details and thought-
> processes, but rather than make this a really long post I'd love to
> get some thoughts and feedback. *Anyone have pointers, ideas, see
> fallacies in my process, etc?
>
> Thanks,
>
> --Noel

Check out this site by Bob Hanson. He made a java app that directly
deals with your question with a cloud street strategy feature.
To get to that feature, look for the dropdown menu that says "These
Conditions" and choose the option at the very bottom "Cloud Street
Flight Strategy Comparison".
http://www.stolaf.edu/people/hansonr/soaring/spd2fly/

Essentially, it brute forces a lot speed to fly calculations to try to
find the optimum theoretical speed. With the given conditions it
finds the best results out of the batch it then takes the
corresponding airspeed and outputs it to the table. Some interesting
numbers it comes up with... Of course the results are subject to
interpretation based on real life results, like MacCready theory.

~Andrew

[John Smith]

Read Reichmann, he explains everything.

While real life tactics may have changed, the mathematical basics are
still the same, and it's still good advice to start with strict McCready
and adopt only later when you have understood and mastered the basics.

[Tim Taylor]

On May 18, 5:11*pm, "noel.wade" > wrote:
> A couple of things:
>
> 1) Thanks John, that's the kind of comparison/number-checking I was
> looking for. *I know that cloudbase/working-band/lift-strength play a
> factor in determining whether the tradeoff is worth it; but didn't
> think to do a straight MacCready comparison (the idea of pushing the
> glider to high speed while still in spotty lift under a cloud-street
> is just still feels awkward - discounting the few times where I've
> approached cloudbase and needed to "bug out")!
>
> 2) I thought flying "energy lines" was the consensus these days, and
> "Dolphin flying" was out - given that variometers lag and its
> incredibly hard to time your pushes and pulls when flying through
> alternating narrow cores and strong sink. *If your strong cores are
> separated by a couple of miles, their miniscule contribution (say a 50
> - 150 foot height gain) when you fly straight through them doesn't
> seem worth it (especially when you start considering the risks of
> hitting nasty sink before accelerating back up to cruise speed;
> possibly wiping out all of your gains). *I've always been much more
> tempted to slow down under weak spread-out lift (for example, under a
> cloud thats about to OD).
>
> Now just to turn this around and look at it from a different
> perspective: *For the fast guys out there, when (if ever) *do* you
> start slowing down a lot while flying straight ahead? *What're the
> combination of factors that encourage you to switch into this mode?
>
> Thanks,
>
> --Noel

Reichmann used the term "dolphin flight" to mean flying streets or
energy lines.

I will give you a simple was way to think about how fast to fly in
lift on a street. As always there are many other factors in tactics
that may make you do some things differently but those usually have to
do with some change in the conditions ahead (end of the street, big
blue hole, final glide, etc).

Lets take a simple MacCready STF ring set of numbers. These should be
about right for your DG.

0 60
-1 65
-2 70
-3 75
-4 80
-5 85
-6 90
-7 95

Invert the numbers to average base cruise speed (this is how most of
us determine our average cruise speeds so we are not just pushing and
pulling all the time.

0 60
1 65
2 70
3 75
4 80
5 85
6 90
7 95

Now on the cloud street use your target thermal strength, the one your
willing to stop for be your guide. Again let's use 5 knots. Subtract
the difference between your target speed and the STF for the lift.
Example you're in 2 knot lift, 85-(85-75)=75.

Vario Speed
0 85
1 80
2 75
3 70
4 65
5 60 (thermal if really below cloudbase)
6 55 (thermal if really below cloudbase)
7 50 (thermal if really below cloudbase)

[David[_13_]]

On May 18, 3:25*pm, John Cochrane >
wrote:
> On May 18, 4:33*pm, "noel.wade" > wrote:
>
> > All -
>
> > I've been flying with the assumption that flying slow in cloud-streets
> > is always a good idea. *
>
> I disagree. It's always MacCready theory, if correctly interpreted.
>
> Suppose the cloudstreet offers 2 knots up everywhere with occasional 8
> knot cores. (Uvalde). Option 1: fly 50 knots without losing altitude.
> You average 50 knots = 57 mph. Option 2: fly 85 knots, (I.e. cruise at
> Mc 4 or so) , slowly losing, but stopping to thermal in the 8 knot
> cores. You average 80 mph. (Typical uvalde winning speeds) The latter
> is much faster. Reichmann offers the same advice.
>
> John Cochrane

Agree!

Reichmann called Option 1 "Forced dolphin flight" if I remember
correctly.

Unfortunately this is what I most often do and I should stop.

David (GJ)

[Nine Bravo Ground]

On May 18, 5:15*pm, John Smith > wrote:
> Read Reichmann, he explains everything.
>
> While real life tactics may have changed, the mathematical basics are
> still the same, and it's still good advice to start with strict McCready
> and adopt only later when you have understood and mastered the basics.

One reason why it often pays to force yourself to fly straight ahead
rather than circling more than pure McCready theory would tell you is
that you rarely roll right into the core of a thermal. Often you spend
a number of circles getting centered and on occasion you completely
miss and make a circle in sink. The shorter the potential climb the
more this effect hurts you. Within a couple thousand feet of cloudbase
it hardly ever pays to take a circle. In theory you could roll this
into the McCready calculation, but people rarely do the extra math -
except BB.

I've found a similar effect on the last climb to make final glide -
climbing up to the altitude needed to match the McCready setting to
your climb rate almost always puts you behind the glider who leaves a
little earlier.

9B

[Ramy]

On May 20, 6:50*am, Nine Bravo Ground > wrote:
> On May 18, 5:15*pm, John Smith > wrote:
>
> > Read Reichmann, he explains everything.
>
> > While real life tactics may have changed, the mathematical basics are
> > still the same, and it's still good advice to start with strict McCready
> > and adopt only later when you have understood and mastered the basics.
>
> One reason why it often pays to force yourself to fly straight ahead
> rather than circling more than pure McCready theory would tell you is
> that you rarely roll right into the core of a thermal. Often you spend
> a number of circles getting centered and on occasion you completely
> miss and make a circle in sink. *The shorter the potential climb the
> more this effect hurts you. Within a couple thousand feet of cloudbase
> it hardly ever pays to take a circle. In theory you could roll this
> into the McCready calculation, but people rarely do the extra math -
> except BB.
>
> I've found a similar effect on the last climb to make final glide -
> climbing up to the altitude needed to match the McCready setting to
> your climb rate almost always puts you behind the glider who leaves a
> little earlier.
>
> 9B

This is theoratically correct, but I found out more often than not
that the lift right below cloudbase (keeping FAR clearance of course)
is stronger and more widespread than couple of thousands below, which
allows much faster speed without loosing altitude.

Ramy

[Andy[_10_]]

On May 20, 7:47*pm, Ramy > wrote:
>
> This is theoratically correct, but I found out more often than not
> that the lift right below cloudbase (keeping FAR clearance of course)
> is stronger and more widespread than couple of thousands below, which
> allows much faster speed without loosing altitude.
>
> Ramy

Okay, I broke out the polar and did some simple math for a "typical"
scenario.

Take a cloud street that is 16 miles long with 4 knot average thermal
strength and 10 knot peak thermal strength.

Pilot A pulls back to 70 knots and climbs in the average lift (2.4
knot net climb rate). It takes him 14 minutes to reach the end of the
street and he has gained 3,360 feet.

Pilot B climbs in the strong core, taking 30 seconds to center and
achieving a 8.4 knot average climb rate thereafter (note the higher
sink rate for circling flight). After 5 minutes he has climbed 4,250
feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
the process. Both pilots arrive at the end of the cloud street at the
same time and altitude.

Conclusion: stopping to circle in a thermal weaker than 10 knots puts
you behind the pilot who climbs straight ahead.

Taking Ramy's point about stronger lift closer to cloudbase into
account, let's assume you find an average 5 knots after a circling
climb instead of 4 knots for climbing straight ahead. In this case you
need a minimum 8.3 knot thermal before stopping to circle makes sense.
If you assume 6 versus 4 knots average lift cruising closer to
cloudbase then you only need a 6.6 knot thermal - but that starts to
feel like a pretty strong thermal strength gradient with altitude.

This analysis doesn't take into account a slight true airspeed
advantage for the pilot who climbs first due to his higher average
altitude.

Overall, I think this confirms that it pays to avoid circling under a
cloudstreet for anything but the very strongest lift. This is even
more true if you are heading into an upwind turnpoint.

9B

[Andy[_10_]]

On May 22, 9:02*am, Andy > wrote:
> On May 20, 7:47*pm, Ramy > wrote:
>
> Overall, I think this confirms that it pays to avoid circling under a
> cloudstreet for anything but the very strongest lift. *This is even
> more true if you are heading into an upwind turnpoint.
>
> 9B

BTW, I'm not arguing against McCready theory. However, I do think you
need to account for the overall conditions and things like TAS and
lift strength versus altitude, thermal centering time, winds an
proximity of turnpoints. Not all of those factors are accounted for in
the base theory and some of the outcomes are quite a bit more extreme
that you might expect.

See BB's analysis of upwind versus downwind turnpoints as an example.
Using that advice gained me something like 4 mph on a long task where
the first turn was into a 30 mph headwind.

9B

[Eric Greenwell]

On 5/18/2010 3:28 PM, Dave Nadler wrote:
> On May 18, 5:33 pm, > wrote:
>
>> All -
>>
>> I've been flying with the assumption that flying slow in cloud-streets
>> is always a good idea. But in looking at some of my OLC flights I
>> wonder if my logic is faulty and I'm hurting my speed. Anyone with
>> lots of OLC or contest experience care to offer feedback? Details
>> below...
>>
snip
>
> A few weeks ago I flew 440 miles, around 15% circling, mostly during
> the beginning of the flight and digging out of one hole (flew under OD
> at 2nd turn):
> http://www.onlinecontest.org/olc-2.0/gliding/flightinfo.html?flightId=502072323
>
> Suggestion: Go look at the log files of the fastest pilots during last
> years 18m nationals at Ephrata !
But keep in mind, a modern ballasted 18 meter glider will not have to
circle nearly as often as your unballasted DG 300 will!

--
Eric Greenwell - Washington State, USA (netto to net to email me)

[[email protected]]

On May 18, 6:25*pm, John Cochrane >
wrote:
> On May 18, 4:33*pm, "noel.wade" > wrote:
>
> > All -
>
> > I've been flying with the assumption that flying slow in cloud-streets
> > is always a good idea. *
>
> I disagree. It's always MacCready theory, if correctly interpreted.
>
> Suppose the cloudstreet offers 2 knots up everywhere with occasional 8
> knot cores. (Uvalde). Option 1: fly 50 knots without losing altitude.
> You average 50 knots = 57 mph. Option 2: fly 85 knots, (I.e. cruise at
> Mc 4 or so) , slowly losing, but stopping to thermal in the 8 knot
> cores. You average 80 mph. (Typical uvalde winning speeds) The latter
> is much faster. Reichmann offers the same advice.
>
> John Cochrane

I would only add that the other important factor is the view ahead.
Sometimes slowing in moderate lift for a while- maybe to max l/d or
something as opposed to min sink, can keep you in contact with the
lift without having to circle. Modern gliders don't give up much
straight ahead climbing ability when flown a bit above l/d max.
Qualifier- if you slow down so much that you climb to where you can't
read the clouds effectively, you have made an important mistake.
You ask good questions.
UH

[Tim Taylor]

On May 22, 10:02*am, Andy > wrote:
> On May 20, 7:47*pm, Ramy > wrote:
>
>
>
> > This is theoratically correct, but I found out more often than not
> > that the lift right below cloudbase (keeping FAR clearance of course)
> > is stronger and more widespread than couple of thousands below, which
> > allows much faster speed without loosing altitude.
>
> > Ramy
>
> Okay, I broke out the polar and did some simple math for a "typical"
> scenario.
>
> Take a cloud street that is 16 miles long with 4 knot average thermal
> strength and 10 knot peak thermal strength.
>
> Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> knot net climb rate). It takes him 14 minutes to reach the end of the
> street and he has gained 3,360 feet.
>
> Pilot B climbs in the strong core, taking 30 seconds to center and
> achieving a 8.4 knot average climb rate thereafter (note the higher
> sink rate for circling flight). After 5 minutes he has climbed 4,250
> feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> the process. Both pilots arrive at the end of the cloud street at the
> same time and altitude.
>
> Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> you behind the pilot who climbs straight ahead.
>
> Taking Ramy's point about stronger lift closer to cloudbase into
> account, let's assume you find an average 5 knots after a circling
> climb instead of 4 knots for climbing straight ahead. In this case you
> need a minimum 8.3 knot thermal before stopping to circle makes sense.
> If you assume 6 versus 4 knots average lift cruising closer to
> cloudbase then you only need a 6.6 knot thermal - but that starts to
> feel like a pretty strong thermal strength gradient with altitude.
>
> This analysis doesn't take into account a slight true airspeed
> advantage for the pilot who climbs first due to his higher average
> altitude.
>
> Overall, I think this confirms that it pays to avoid circling under a
> cloudstreet for anything but the very strongest lift. *This is even
> more true if you are heading into an upwind turnpoint.
>
> 9B

Andy,

Ok, I tried to take your scenario and put some numbers to it. I used
the polar for an LS-4 (no DG 300 numbers handy). I Assumed 4 knots of
lift along the 16 mile street and a 10 knot thermal at the end. The
glider starts at 3000 feet below the clouds and ends at cloud base at
the end of the street (similar to Reichmann). I added 20 seconds of
centering penalty and a 20% higher sink rate while thermalling.

Don't know if the table will get scrambled on most viewers but here is
the data. It shows that flying near MC speeds is optimum as long as
you can find the strong thermal ahead. If you expect to find a strong
thermal ahead don't slow down to minimum sink speeds. I think I will
use a MC speed of about (expected climb from thermal - average street
strength) as a good compromise MC setting to fly. In this case 10 - 4
= MC setting of 6 to fly.



MC miles/hr time minutes Average speed
45 14.99 64.05
50 14.53 66.06
55 14.14 67.89
0 60 13.80 69.56
1 68 13.34 71.96
2 76 13.18 72.84
3 84 12.62 76.05
4 90 12.33 77.86
5 94 12.18 78.80
6 98 12.07 79.55
7 104 11.95 80.35
8 112 11.87 80.85
9 118 11.87 80.88
10 124 11.90 80.66

TT

[Tim Taylor]

On May 26, 1:09*pm, Tim Taylor > wrote:
> On May 22, 10:02*am, Andy > wrote:
>
>
>
>
>
> > On May 20, 7:47*pm, Ramy > wrote:
>
> > > This is theoratically correct, but I found out more often than not
> > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > is stronger and more widespread than couple of thousands below, which
> > > allows much faster speed without loosing altitude.
>
> > > Ramy
>
> > Okay, I broke out the polar and did some simple math for a "typical"
> > scenario.
>
> > Take a cloud street that is 16 miles long with 4 knot average thermal
> > strength and 10 knot peak thermal strength.
>
> > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > knot net climb rate). It takes him 14 minutes to reach the end of the
> > street and he has gained 3,360 feet.
>
> > Pilot B climbs in the strong core, taking 30 seconds to center and
> > achieving a 8.4 knot average climb rate thereafter (note the higher
> > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > the process. Both pilots arrive at the end of the cloud street at the
> > same time and altitude.
>
> > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > you behind the pilot who climbs straight ahead.
>
> > Taking Ramy's point about stronger lift closer to cloudbase into
> > account, let's assume you find an average 5 knots after a circling
> > climb instead of 4 knots for climbing straight ahead. In this case you
> > need a minimum 8.3 knot thermal before stopping to circle makes sense.
> > If you assume 6 versus 4 knots average lift cruising closer to
> > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > feel like a pretty strong thermal strength gradient with altitude.
>
> > This analysis doesn't take into account a slight true airspeed
> > advantage for the pilot who climbs first due to his higher average
> > altitude.
>
> > Overall, I think this confirms that it pays to avoid circling under a
> > cloudstreet for anything but the very strongest lift. *This is even
> > more true if you are heading into an upwind turnpoint.
>
> > 9B
>
> Andy,
>
> Ok, I tried to take your scenario and put some numbers to it. *I used
> the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> lift along the 16 mile street and a 10 knot thermal at the end. *The
> glider starts at 3000 feet below the clouds and ends at cloud base at
> the end of the street (similar to Reichmann). *I added 20 seconds of
> centering penalty and a 20% higher sink rate while thermalling.
>
> Don't know if the table will get scrambled on most viewers but here is
> the data. *It shows that flying near MC speeds is optimum as long as
> you can find the strong thermal ahead. *If you expect to find a strong
> thermal ahead don't slow down to minimum sink speeds. *I think I will
> use a MC speed of about (expected climb from thermal - average street
> strength) as a good compromise MC setting to fly. *In this case 10 - 4
> = MC setting of 6 to fly.
>
> MC * * *miles/hr * * * *time minutes * *Average speed
> * * * * 45 * * *14.99 * 64.05
> * * * * 50 * * *14.53 * 66.06
> * * * * 55 * * *14.14 * 67.89
> 0 * * * 60 * * *13.80 * 69.56
> 1 * * * 68 * * *13.34 * 71.96
> 2 * * * 76 * * *13.18 * 72.84
> 3 * * * 84 * * *12.62 * 76.05
> 4 * * * 90 * * *12.33 * 77.86
> 5 * * * 94 * * *12.18 * 78.80
> 6 * * * 98 * * *12.07 * 79.55
> 7 * * * 104 * * 11.95 * 80.35
> 8 * * * 112 * * 11.87 * 80.85
> 9 * * * 118 * * 11.87 * 80.88
> 10 * * *124 * * 11.90 * 80.66
>
> TT

Sorry found a sign error in my calculations. Here are the corrected
values:

MC miles/hr time minutes Average speed
45 14.99 64.05
50 14.53 66.06
55 14.14 67.89
0 60 13.80 69.56
1 68 13.34 71.96
2 76 13.25 72.46
3 84 12.90 74.43
4 90 12.75 75.27
5 94 12.70 75.57
6 98 12.69 75.68
7 104 12.71 75.54
8 112 12.82 74.88
9 118 12.95 74.11
10 124 13.12 73.16

[Nine Bravo Ground]

On May 26, 2:16*pm, Tim Taylor > wrote:
> On May 26, 1:09*pm, Tim Taylor > wrote:
>
>
>
>
>
> > On May 22, 10:02*am, Andy > wrote:
>
> > > On May 20, 7:47*pm, Ramy > wrote:
>
> > > > This is theoratically correct, but I found out more often than not
> > > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > > is stronger and more widespread than couple of thousands below, which
> > > > allows much faster speed without loosing altitude.
>
> > > > Ramy
>
> > > Okay, I broke out the polar and did some simple math for a "typical"
> > > scenario.
>
> > > Take a cloud street that is 16 miles long with 4 knot average thermal
> > > strength and 10 knot peak thermal strength.
>
> > > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > > knot net climb rate). It takes him 14 minutes to reach the end of the
> > > street and he has gained 3,360 feet.
>
> > > Pilot B climbs in the strong core, taking 30 seconds to center and
> > > achieving a 8.4 knot average climb rate thereafter (note the higher
> > > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > > the process. Both pilots arrive at the end of the cloud street at the
> > > same time and altitude.
>
> > > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > > you behind the pilot who climbs straight ahead.
>
> > > Taking Ramy's point about stronger lift closer to cloudbase into
> > > account, let's assume you find an average 5 knots after a circling
> > > climb instead of 4 knots for climbing straight ahead. In this case you
> > > need a minimum 8.3 knot thermal before stopping to circle makes sense..
> > > If you assume 6 versus 4 knots average lift cruising closer to
> > > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > > feel like a pretty strong thermal strength gradient with altitude.
>
> > > This analysis doesn't take into account a slight true airspeed
> > > advantage for the pilot who climbs first due to his higher average
> > > altitude.
>
> > > Overall, I think this confirms that it pays to avoid circling under a
> > > cloudstreet for anything but the very strongest lift. *This is even
> > > more true if you are heading into an upwind turnpoint.
>
> > > 9B
>
> > Andy,
>
> > Ok, I tried to take your scenario and put some numbers to it. *I used
> > the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> > lift along the 16 mile street and a 10 knot thermal at the end. *The
> > glider starts at 3000 feet below the clouds and ends at cloud base at
> > the end of the street (similar to Reichmann). *I added 20 seconds of
> > centering penalty and a 20% higher sink rate while thermalling.
>
> > Don't know if the table will get scrambled on most viewers but here is
> > the data. *It shows that flying near MC speeds is optimum as long as
> > you can find the strong thermal ahead. *If you expect to find a strong
> > thermal ahead don't slow down to minimum sink speeds. *I think I will
> > use a MC speed of about (expected climb from thermal - average street
> > strength) as a good compromise MC setting to fly. *In this case 10 - 4
> > = MC setting of 6 to fly.
>
> > MC * * *miles/hr * * * *time minutes * *Average speed
> > * * * * 45 * * *14.99 * 64.05
> > * * * * 50 * * *14.53 * 66.06
> > * * * * 55 * * *14.14 * 67.89
> > 0 * * * 60 * * *13.80 * 69.56
> > 1 * * * 68 * * *13.34 * 71.96
> > 2 * * * 76 * * *13.18 * 72.84
> > 3 * * * 84 * * *12.62 * 76.05
> > 4 * * * 90 * * *12.33 * 77.86
> > 5 * * * 94 * * *12.18 * 78.80
> > 6 * * * 98 * * *12.07 * 79.55
> > 7 * * * 104 * * 11.95 * 80.35
> > 8 * * * 112 * * 11.87 * 80.85
> > 9 * * * 118 * * 11.87 * 80.88
> > 10 * * *124 * * 11.90 * 80.66
>
> > TT
>
> Sorry found a sign error in my calculations. Here are the corrected
> values:
>
> MC * * *miles/hr * * * *time minutes * *Average speed
> * * * * 45 * * *14.99 * 64.05
> * * * * 50 * * *14.53 * 66.06
> * * * * 55 * * *14.14 * 67.89
> 0 * * * 60 * * *13.80 * 69.56
> 1 * * * 68 * * *13.34 * 71.96
> 2 * * * 76 * * *13.25 * 72.46
> 3 * * * 84 * * *12.90 * 74.43
> 4 * * * 90 * * *12.75 * 75.27
> 5 * * * 94 * * *12.70 * 75.57
> 6 * * * 98 * * *12.69 * 75.68
> 7 * * * 104 * * 12.71 * 75.54
> 8 * * * 112 * * 12.82 * 74.88
> 9 * * * 118 * * 12.95 * 74.11
> 10 * * *124 * * 13.12 * 73.16

Nice job Tim. At what speed do you get to top of lift before you
reach the 10-knotter?

The curves are pretty flat on the slow end down to just above best L/D
speed (say Mc=1). Also notice that once you get to 3 knots on the
McCready there is only 13 seconds of difference between the best and
worst times up to Mc=9. That means you MUST center a thermal on the
first circle to make it worthwhile to stop. Your 20% higher sink for
circling flight is about right for 25-30 degrees of bank - at 45
degrees the sink rate is around 70% higher, so the thermals would have
to be fat AND easy to core. More typically I find that around 1/3 of
the time when I turn in a thermal I wish I hadn't and another 1/3 of
the time it takes me a few turns to center it - that's on a good day.
I don't know if that nets out to a 20 second expected loss or not.

I'm normally pickier about stopping to circle at the beginning of the
street versus when I'm getting towards the end - though I am now
reconsidering this view. Higher TAS with altitude and likely stronger
lift closer to cloudbase would argue for climbing up early rather than
late.

My -27 is a little better on the run than my old LS-4 which I think
biases the optimal a bit more to straight ahead climbing over
circling. My flying style has certainly migrated in that direction.

9B

[Nine Bravo Ground]

On May 27, 4:28*am, Nine Bravo Ground > wrote:
> On May 26, 2:16*pm, Tim Taylor > wrote:
>
>
>
>
>
> > On May 26, 1:09*pm, Tim Taylor > wrote:
>
> > > On May 22, 10:02*am, Andy > wrote:
>
> > > > On May 20, 7:47*pm, Ramy > wrote:
>
> > > > > This is theoratically correct, but I found out more often than not
> > > > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > > > is stronger and more widespread than couple of thousands below, which
> > > > > allows much faster speed without loosing altitude.
>
> > > > > Ramy
>
> > > > Okay, I broke out the polar and did some simple math for a "typical"
> > > > scenario.
>
> > > > Take a cloud street that is 16 miles long with 4 knot average thermal
> > > > strength and 10 knot peak thermal strength.
>
> > > > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > > > knot net climb rate). It takes him 14 minutes to reach the end of the
> > > > street and he has gained 3,360 feet.
>
> > > > Pilot B climbs in the strong core, taking 30 seconds to center and
> > > > achieving a 8.4 knot average climb rate thereafter (note the higher
> > > > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > > > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > > > the process. Both pilots arrive at the end of the cloud street at the
> > > > same time and altitude.
>
> > > > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > > > you behind the pilot who climbs straight ahead.
>
> > > > Taking Ramy's point about stronger lift closer to cloudbase into
> > > > account, let's assume you find an average 5 knots after a circling
> > > > climb instead of 4 knots for climbing straight ahead. In this case you
> > > > need a minimum 8.3 knot thermal before stopping to circle makes sense.
> > > > If you assume 6 versus 4 knots average lift cruising closer to
> > > > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > > > feel like a pretty strong thermal strength gradient with altitude.
>
> > > > This analysis doesn't take into account a slight true airspeed
> > > > advantage for the pilot who climbs first due to his higher average
> > > > altitude.
>
> > > > Overall, I think this confirms that it pays to avoid circling under a
> > > > cloudstreet for anything but the very strongest lift. *This is even
> > > > more true if you are heading into an upwind turnpoint.
>
> > > > 9B
>
> > > Andy,
>
> > > Ok, I tried to take your scenario and put some numbers to it. *I used
> > > the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> > > lift along the 16 mile street and a 10 knot thermal at the end. *The
> > > glider starts at 3000 feet below the clouds and ends at cloud base at
> > > the end of the street (similar to Reichmann). *I added 20 seconds of
> > > centering penalty and a 20% higher sink rate while thermalling.
>
> > > Don't know if the table will get scrambled on most viewers but here is
> > > the data. *It shows that flying near MC speeds is optimum as long as
> > > you can find the strong thermal ahead. *If you expect to find a strong
> > > thermal ahead don't slow down to minimum sink speeds. *I think I will
> > > use a MC speed of about (expected climb from thermal - average street
> > > strength) as a good compromise MC setting to fly. *In this case 10 - 4
> > > = MC setting of 6 to fly.
>
> > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > * * * * 45 * * *14.99 * 64.05
> > > * * * * 50 * * *14.53 * 66.06
> > > * * * * 55 * * *14.14 * 67.89
> > > 0 * * * 60 * * *13.80 * 69.56
> > > 1 * * * 68 * * *13.34 * 71.96
> > > 2 * * * 76 * * *13.18 * 72.84
> > > 3 * * * 84 * * *12.62 * 76.05
> > > 4 * * * 90 * * *12.33 * 77.86
> > > 5 * * * 94 * * *12.18 * 78.80
> > > 6 * * * 98 * * *12.07 * 79.55
> > > 7 * * * 104 * * 11.95 * 80.35
> > > 8 * * * 112 * * 11.87 * 80.85
> > > 9 * * * 118 * * 11.87 * 80.88
> > > 10 * * *124 * * 11.90 * 80.66
>
> > > TT
>
> > Sorry found a sign error in my calculations. Here are the corrected
> > values:
>
> > MC * * *miles/hr * * * *time minutes * *Average speed
> > * * * * 45 * * *14.99 * 64.05
> > * * * * 50 * * *14.53 * 66.06
> > * * * * 55 * * *14.14 * 67.89
> > 0 * * * 60 * * *13.80 * 69.56
> > 1 * * * 68 * * *13.34 * 71.96
> > 2 * * * 76 * * *13.25 * 72.46
> > 3 * * * 84 * * *12.90 * 74.43
> > 4 * * * 90 * * *12.75 * 75.27
> > 5 * * * 94 * * *12.70 * 75.57
> > 6 * * * 98 * * *12.69 * 75.68
> > 7 * * * 104 * * 12.71 * 75.54
> > 8 * * * 112 * * 12.82 * 74.88
> > 9 * * * 118 * * 12.95 * 74.11
> > 10 * * *124 * * 13.12 * 73.16
>
> Nice job Tim. *At what speed do you get to top of lift before you
> reach the 10-knotter?
>
> The curves are pretty flat on the slow end down to just above best L/D
> speed (say Mc=1). Also notice that once you get to 3 knots on the
> McCready there is only 13 seconds of difference between the best and
> worst times up to Mc=9. That means you MUST center a thermal on the
> first circle to make it worthwhile to stop. Your 20% higher sink for
> circling flight is about right for 25-30 degrees of bank - at 45
> degrees the sink rate is around 70% higher, so the thermals would have
> to be fat AND easy to core. More typically I find that around 1/3 of
> the time when I turn in a thermal I wish I hadn't and another 1/3 of
> the time it takes me a few turns to center it - that's on a good day.
> I don't know if that nets out to a 20 second expected loss or not.
>
> I'm normally pickier about stopping to circle at the beginning of the
> street versus when I'm getting towards the end - though I am now
> reconsidering this view. Higher TAS with altitude and likely stronger
> lift closer to cloudbase would argue for climbing up early rather than
> late.
>
> My -27 is a little better on the run than my old LS-4 which I think
> biases the optimal a bit more to straight ahead climbing over
> circling. My flying style has certainly migrated in that direction.
>
> 9B

Just as a hypothetical, I figured out how much "centering loss" you
might experience rolling into a 10-knot thermal. If I assume 1/3 of
the time you get 10 knots from the first turn, 1/3 of the time you
spend a turn at 2 knots and a turn at 5 knots before centering and 1/3
of the time you burn a circle on a thermal that turns out not to be
there, then the expected centering loss is the equivalent of 40
seconds circling at zero sink.

9B

[Tim Taylor]

On May 27, 8:25*am, Nine Bravo Ground > wrote:
> On May 27, 4:28*am, Nine Bravo Ground > wrote:
>
>
>
>
>
> > On May 26, 2:16*pm, Tim Taylor > wrote:
>
> > > On May 26, 1:09*pm, Tim Taylor > wrote:
>
> > > > On May 22, 10:02*am, Andy > wrote:
>
> > > > > On May 20, 7:47*pm, Ramy > wrote:
>
> > > > > > This is theoratically correct, but I found out more often than not
> > > > > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > > > > is stronger and more widespread than couple of thousands below, which
> > > > > > allows much faster speed without loosing altitude.
>
> > > > > > Ramy
>
> > > > > Okay, I broke out the polar and did some simple math for a "typical"
> > > > > scenario.
>
> > > > > Take a cloud street that is 16 miles long with 4 knot average thermal
> > > > > strength and 10 knot peak thermal strength.
>
> > > > > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > > > > knot net climb rate). It takes him 14 minutes to reach the end of the
> > > > > street and he has gained 3,360 feet.
>
> > > > > Pilot B climbs in the strong core, taking 30 seconds to center and
> > > > > achieving a 8.4 knot average climb rate thereafter (note the higher
> > > > > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > > > > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > > > > the process. Both pilots arrive at the end of the cloud street at the
> > > > > same time and altitude.
>
> > > > > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > > > > you behind the pilot who climbs straight ahead.
>
> > > > > Taking Ramy's point about stronger lift closer to cloudbase into
> > > > > account, let's assume you find an average 5 knots after a circling
> > > > > climb instead of 4 knots for climbing straight ahead. In this case you
> > > > > need a minimum 8.3 knot thermal before stopping to circle makes sense.
> > > > > If you assume 6 versus 4 knots average lift cruising closer to
> > > > > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > > > > feel like a pretty strong thermal strength gradient with altitude..
>
> > > > > This analysis doesn't take into account a slight true airspeed
> > > > > advantage for the pilot who climbs first due to his higher average
> > > > > altitude.
>
> > > > > Overall, I think this confirms that it pays to avoid circling under a
> > > > > cloudstreet for anything but the very strongest lift. *This is even
> > > > > more true if you are heading into an upwind turnpoint.
>
> > > > > 9B
>
> > > > Andy,
>
> > > > Ok, I tried to take your scenario and put some numbers to it. *I used
> > > > the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> > > > lift along the 16 mile street and a 10 knot thermal at the end. *The
> > > > glider starts at 3000 feet below the clouds and ends at cloud base at
> > > > the end of the street (similar to Reichmann). *I added 20 seconds of
> > > > centering penalty and a 20% higher sink rate while thermalling.
>
> > > > Don't know if the table will get scrambled on most viewers but here is
> > > > the data. *It shows that flying near MC speeds is optimum as long as
> > > > you can find the strong thermal ahead. *If you expect to find a strong
> > > > thermal ahead don't slow down to minimum sink speeds. *I think I will
> > > > use a MC speed of about (expected climb from thermal - average street
> > > > strength) as a good compromise MC setting to fly. *In this case 10 - 4
> > > > = MC setting of 6 to fly.
>
> > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > * * * * 45 * * *14.99 * 64.05
> > > > * * * * 50 * * *14.53 * 66.06
> > > > * * * * 55 * * *14.14 * 67.89
> > > > 0 * * * 60 * * *13.80 * 69.56
> > > > 1 * * * 68 * * *13.34 * 71.96
> > > > 2 * * * 76 * * *13.18 * 72.84
> > > > 3 * * * 84 * * *12.62 * 76.05
> > > > 4 * * * 90 * * *12.33 * 77.86
> > > > 5 * * * 94 * * *12.18 * 78.80
> > > > 6 * * * 98 * * *12.07 * 79.55
> > > > 7 * * * 104 * * 11.95 * 80.35
> > > > 8 * * * 112 * * 11.87 * 80.85
> > > > 9 * * * 118 * * 11.87 * 80.88
> > > > 10 * * *124 * * 11.90 * 80.66
>
> > > > TT
>
> > > Sorry found a sign error in my calculations. Here are the corrected
> > > values:
>
> > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > * * * * 45 * * *14.99 * 64.05
> > > * * * * 50 * * *14.53 * 66.06
> > > * * * * 55 * * *14.14 * 67.89
> > > 0 * * * 60 * * *13.80 * 69.56
> > > 1 * * * 68 * * *13.34 * 71.96
> > > 2 * * * 76 * * *13.25 * 72.46
> > > 3 * * * 84 * * *12.90 * 74.43
> > > 4 * * * 90 * * *12.75 * 75.27
> > > 5 * * * 94 * * *12.70 * 75.57
> > > 6 * * * 98 * * *12.69 * 75.68
> > > 7 * * * 104 * * 12.71 * 75.54
> > > 8 * * * 112 * * 12.82 * 74.88
> > > 9 * * * 118 * * 12.95 * 74.11
> > > 10 * * *124 * * 13.12 * 73.16
>
> > Nice job Tim. *At what speed do you get to top of lift before you
> > reach the 10-knotter?
>
> > The curves are pretty flat on the slow end down to just above best L/D
> > speed (say Mc=1). Also notice that once you get to 3 knots on the
> > McCready there is only 13 seconds of difference between the best and
> > worst times up to Mc=9. That means you MUST center a thermal on the
> > first circle to make it worthwhile to stop. Your 20% higher sink for
> > circling flight is about right for 25-30 degrees of bank - at 45
> > degrees the sink rate is around 70% higher, so the thermals would have
> > to be fat AND easy to core. More typically I find that around 1/3 of
> > the time when I turn in a thermal I wish I hadn't and another 1/3 of
> > the time it takes me a few turns to center it - that's on a good day.
> > I don't know if that nets out to a 20 second expected loss or not.
>
> > I'm normally pickier about stopping to circle at the beginning of the
> > street versus when I'm getting towards the end - though I am now
> > reconsidering this view. Higher TAS with altitude and likely stronger
> > lift closer to cloudbase would argue for climbing up early rather than
> > late.
>
> > My -27 is a little better on the run than my old LS-4 which I think
> > biases the optimal a bit more to straight ahead climbing over
> > circling. My flying style has certainly migrated in that direction.
>
> > 9B
>
> Just as a hypothetical, I figured out how much "centering loss" you
> might experience rolling into a 10-knot thermal. *If I assume 1/3 of
> the time you get 10 knots from the first turn, 1/3 of the time you
> spend a turn at 2 knots and a turn at 5 knots before centering and 1/3
> of the time you burn a circle on a thermal that turns out not to be
> there, then the expected centering loss is the equivalent of 40
> seconds circling at zero sink.
>
> 9B

Andy,

I agree with BB, Ramy and your comments. Everything depends on where
you are flying and the strength and variability of the conditions.
Out west at Logan or Parowan you can risk going faster and lower
beneath the cloud streets, at Uvalde I learned the hard way not to
drop more than 2000 feet below the clouds or risk not getting back up
or if you do very slowly (I had one save from 400 feet agl). The best
lift was within 1000 feet of cloudbase on most streets.

I will work on refining the model when I get time so I can run more
scenarios easily. The general things to take aways from the exercise
is that slowing to minimum sink is not usually a good idea unless you
really need to climb, best l/d or slightly faster is much better.
That being slightly conservative costs very little on potential speed
and reduces the risk factor of not finding that next big thermal. The
difference of flying at MC 4 to MC 7 is very little and you lose much
less altitude at the lower MC to stay in the stonger lift and
connected to the clouds.

[Chris Reed[_2_]]

Tim Taylor wrote:
> On May 27, 8:25 am, Nine Bravo Ground > wrote:
>> On May 27, 4:28 am, Nine Bravo Ground > wrote:
>>
>>
>>
>>
>>
>>> On May 26, 2:16 pm, Tim Taylor > wrote:
>>>> On May 26, 1:09 pm, Tim Taylor > wrote:
>>>>> On May 22, 10:02 am, Andy > wrote:
>>>>>> On May 20, 7:47 pm, Ramy > wrote:
>>>>>>> This is theoratically correct, but I found out more often than not
>>>>>>> that the lift right below cloudbase (keeping FAR clearance of course)
>>>>>>> is stronger and more widespread than couple of thousands below, which
>>>>>>> allows much faster speed without loosing altitude.
>>>>>>> Ramy
>>>>>> Okay, I broke out the polar and did some simple math for a "typical"
>>>>>> scenario.
>>>>>> Take a cloud street that is 16 miles long with 4 knot average thermal
>>>>>> strength and 10 knot peak thermal strength.
>>>>>> Pilot A pulls back to 70 knots and climbs in the average lift (2.4
>>>>>> knot net climb rate). It takes him 14 minutes to reach the end of the
>>>>>> street and he has gained 3,360 feet.
>>>>>> Pilot B climbs in the strong core, taking 30 seconds to center and
>>>>>> achieving a 8.4 knot average climb rate thereafter (note the higher
>>>>>> sink rate for circling flight). After 5 minutes he has climbed 4,250
>>>>>> feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
>>>>>> the process. Both pilots arrive at the end of the cloud street at the
>>>>>> same time and altitude.
>>>>>> Conclusion: stopping to circle in a thermal weaker than 10 knots puts
>>>>>> you behind the pilot who climbs straight ahead.
>>>>>> Taking Ramy's point about stronger lift closer to cloudbase into
>>>>>> account, let's assume you find an average 5 knots after a circling
>>>>>> climb instead of 4 knots for climbing straight ahead. In this case you
>>>>>> need a minimum 8.3 knot thermal before stopping to circle makes sense.
>>>>>> If you assume 6 versus 4 knots average lift cruising closer to
>>>>>> cloudbase then you only need a 6.6 knot thermal - but that starts to
>>>>>> feel like a pretty strong thermal strength gradient with altitude..
>>>>>> This analysis doesn't take into account a slight true airspeed
>>>>>> advantage for the pilot who climbs first due to his higher average
>>>>>> altitude.
>>>>>> Overall, I think this confirms that it pays to avoid circling under a
>>>>>> cloudstreet for anything but the very strongest lift. This is even
>>>>>> more true if you are heading into an upwind turnpoint.
>>>>>> 9B
>>>>> Andy,
>>>>> Ok, I tried to take your scenario and put some numbers to it. I used
>>>>> the polar for an LS-4 (no DG 300 numbers handy). I Assumed 4 knots of
>>>>> lift along the 16 mile street and a 10 knot thermal at the end. The
>>>>> glider starts at 3000 feet below the clouds and ends at cloud base at
>>>>> the end of the street (similar to Reichmann). I added 20 seconds of
>>>>> centering penalty and a 20% higher sink rate while thermalling.
>>>>> Don't know if the table will get scrambled on most viewers but here is
>>>>> the data. It shows that flying near MC speeds is optimum as long as
>>>>> you can find the strong thermal ahead. If you expect to find a strong
>>>>> thermal ahead don't slow down to minimum sink speeds. I think I will
>>>>> use a MC speed of about (expected climb from thermal - average street
>>>>> strength) as a good compromise MC setting to fly. In this case 10 - 4
>>>>> = MC setting of 6 to fly.
>>>>> MC miles/hr time minutes Average speed
>>>>> 45 14.99 64.05
>>>>> 50 14.53 66.06
>>>>> 55 14.14 67.89
>>>>> 0 60 13.80 69.56
>>>>> 1 68 13.34 71.96
>>>>> 2 76 13.18 72.84
>>>>> 3 84 12.62 76.05
>>>>> 4 90 12.33 77.86
>>>>> 5 94 12.18 78.80
>>>>> 6 98 12.07 79.55
>>>>> 7 104 11.95 80.35
>>>>> 8 112 11.87 80.85
>>>>> 9 118 11.87 80.88
>>>>> 10 124 11.90 80.66
>>>>> TT
>>>> Sorry found a sign error in my calculations. Here are the corrected
>>>> values:
>>>> MC miles/hr time minutes Average speed
>>>> 45 14.99 64.05
>>>> 50 14.53 66.06
>>>> 55 14.14 67.89
>>>> 0 60 13.80 69.56
>>>> 1 68 13.34 71.96
>>>> 2 76 13.25 72.46
>>>> 3 84 12.90 74.43
>>>> 4 90 12.75 75.27
>>>> 5 94 12.70 75.57
>>>> 6 98 12.69 75.68
>>>> 7 104 12.71 75.54
>>>> 8 112 12.82 74.88
>>>> 9 118 12.95 74.11
>>>> 10 124 13.12 73.16
>>> Nice job Tim. At what speed do you get to top of lift before you
>>> reach the 10-knotter?
>>> The curves are pretty flat on the slow end down to just above best L/D
>>> speed (say Mc=1). Also notice that once you get to 3 knots on the
>>> McCready there is only 13 seconds of difference between the best and
>>> worst times up to Mc=9. That means you MUST center a thermal on the
>>> first circle to make it worthwhile to stop. Your 20% higher sink for
>>> circling flight is about right for 25-30 degrees of bank - at 45
>>> degrees the sink rate is around 70% higher, so the thermals would have
>>> to be fat AND easy to core. More typically I find that around 1/3 of
>>> the time when I turn in a thermal I wish I hadn't and another 1/3 of
>>> the time it takes me a few turns to center it - that's on a good day.
>>> I don't know if that nets out to a 20 second expected loss or not.
>>> I'm normally pickier about stopping to circle at the beginning of the
>>> street versus when I'm getting towards the end - though I am now
>>> reconsidering this view. Higher TAS with altitude and likely stronger
>>> lift closer to cloudbase would argue for climbing up early rather than
>>> late.
>>> My -27 is a little better on the run than my old LS-4 which I think
>>> biases the optimal a bit more to straight ahead climbing over
>>> circling. My flying style has certainly migrated in that direction.
>>> 9B
>> Just as a hypothetical, I figured out how much "centering loss" you
>> might experience rolling into a 10-knot thermal. If I assume 1/3 of
>> the time you get 10 knots from the first turn, 1/3 of the time you
>> spend a turn at 2 knots and a turn at 5 knots before centering and 1/3
>> of the time you burn a circle on a thermal that turns out not to be
>> there, then the expected centering loss is the equivalent of 40
>> seconds circling at zero sink.
>>
>> 9B
>
> Andy,
>
> I agree with BB, Ramy and your comments. Everything depends on where
> you are flying and the strength and variability of the conditions.
> Out west at Logan or Parowan you can risk going faster and lower
> beneath the cloud streets, at Uvalde I learned the hard way not to
> drop more than 2000 feet below the clouds or risk not getting back up
> or if you do very slowly (I had one save from 400 feet agl). The best
> lift was within 1000 feet of cloudbase on most streets.
>
> I will work on refining the model when I get time so I can run more
> scenarios easily. The general things to take aways from the exercise
> is that slowing to minimum sink is not usually a good idea unless you
> really need to climb, best l/d or slightly faster is much better.
> That being slightly conservative costs very little on potential speed
> and reduces the risk factor of not finding that next big thermal. The
> difference of flying at MC 4 to MC 7 is very little and you lose much
> less altitude at the lower MC to stay in the stonger lift and
> connected to the clouds.
>
>
>
I made a summary of Reichmann for the simple-minded (i.e. me) a few
years ago.

MC 0 = Bad, unless you're struggling to stay airborne.

The penalty for flying MC 1, as opposed to what your flight computer
says, is pretty small, and the gains in range are big in comparison.

As a rule of thumb, flying 1/2 the MC recommended by the theory/computer
is a good strategy.

This was for individual thermals, not streets, but the calculations
above suggest it works for streets as well.

[Andy[_10_]]

On May 27, 3:25*pm, Tim Taylor > wrote:
> On May 27, 8:25*am, Nine Bravo Ground > wrote:
>
>
>
>
>
> > On May 27, 4:28*am, Nine Bravo Ground > wrote:
>
> > > On May 26, 2:16*pm, Tim Taylor > wrote:
>
> > > > On May 26, 1:09*pm, Tim Taylor > wrote:
>
> > > > > On May 22, 10:02*am, Andy > wrote:
>
> > > > > > On May 20, 7:47*pm, Ramy > wrote:
>
> > > > > > > This is theoratically correct, but I found out more often than not
> > > > > > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > > > > > is stronger and more widespread than couple of thousands below, which
> > > > > > > allows much faster speed without loosing altitude.
>
> > > > > > > Ramy
>
> > > > > > Okay, I broke out the polar and did some simple math for a "typical"
> > > > > > scenario.
>
> > > > > > Take a cloud street that is 16 miles long with 4 knot average thermal
> > > > > > strength and 10 knot peak thermal strength.
>
> > > > > > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > > > > > knot net climb rate). It takes him 14 minutes to reach the end of the
> > > > > > street and he has gained 3,360 feet.
>
> > > > > > Pilot B climbs in the strong core, taking 30 seconds to center and
> > > > > > achieving a 8.4 knot average climb rate thereafter (note the higher
> > > > > > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > > > > > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > > > > > the process. Both pilots arrive at the end of the cloud street at the
> > > > > > same time and altitude.
>
> > > > > > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > > > > > you behind the pilot who climbs straight ahead.
>
> > > > > > Taking Ramy's point about stronger lift closer to cloudbase into
> > > > > > account, let's assume you find an average 5 knots after a circling
> > > > > > climb instead of 4 knots for climbing straight ahead. In this case you
> > > > > > need a minimum 8.3 knot thermal before stopping to circle makes sense.
> > > > > > If you assume 6 versus 4 knots average lift cruising closer to
> > > > > > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > > > > > feel like a pretty strong thermal strength gradient with altitude.
>
> > > > > > This analysis doesn't take into account a slight true airspeed
> > > > > > advantage for the pilot who climbs first due to his higher average
> > > > > > altitude.
>
> > > > > > Overall, I think this confirms that it pays to avoid circling under a
> > > > > > cloudstreet for anything but the very strongest lift. *This is even
> > > > > > more true if you are heading into an upwind turnpoint.
>
> > > > > > 9B
>
> > > > > Andy,
>
> > > > > Ok, I tried to take your scenario and put some numbers to it. *I used
> > > > > the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> > > > > lift along the 16 mile street and a 10 knot thermal at the end. *The
> > > > > glider starts at 3000 feet below the clouds and ends at cloud base at
> > > > > the end of the street (similar to Reichmann). *I added 20 seconds of
> > > > > centering penalty and a 20% higher sink rate while thermalling.
>
> > > > > Don't know if the table will get scrambled on most viewers but here is
> > > > > the data. *It shows that flying near MC speeds is optimum as long as
> > > > > you can find the strong thermal ahead. *If you expect to find a strong
> > > > > thermal ahead don't slow down to minimum sink speeds. *I think I will
> > > > > use a MC speed of about (expected climb from thermal - average street
> > > > > strength) as a good compromise MC setting to fly. *In this case 10 - 4
> > > > > = MC setting of 6 to fly.
>
> > > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > > * * * * 45 * * *14.99 * 64.05
> > > > > * * * * 50 * * *14.53 * 66.06
> > > > > * * * * 55 * * *14.14 * 67.89
> > > > > 0 * * * 60 * * *13.80 * 69.56
> > > > > 1 * * * 68 * * *13.34 * 71.96
> > > > > 2 * * * 76 * * *13.18 * 72.84
> > > > > 3 * * * 84 * * *12.62 * 76.05
> > > > > 4 * * * 90 * * *12.33 * 77.86
> > > > > 5 * * * 94 * * *12.18 * 78.80
> > > > > 6 * * * 98 * * *12.07 * 79.55
> > > > > 7 * * * 104 * * 11.95 * 80.35
> > > > > 8 * * * 112 * * 11.87 * 80.85
> > > > > 9 * * * 118 * * 11.87 * 80.88
> > > > > 10 * * *124 * * 11.90 * 80.66
>
> > > > > TT
>
> > > > Sorry found a sign error in my calculations. Here are the corrected
> > > > values:
>
> > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > * * * * 45 * * *14.99 * 64.05
> > > > * * * * 50 * * *14.53 * 66.06
> > > > * * * * 55 * * *14.14 * 67.89
> > > > 0 * * * 60 * * *13.80 * 69.56
> > > > 1 * * * 68 * * *13.34 * 71.96
> > > > 2 * * * 76 * * *13.25 * 72.46
> > > > 3 * * * 84 * * *12.90 * 74.43
> > > > 4 * * * 90 * * *12.75 * 75.27
> > > > 5 * * * 94 * * *12.70 * 75.57
> > > > 6 * * * 98 * * *12.69 * 75.68
> > > > 7 * * * 104 * * 12.71 * 75.54
> > > > 8 * * * 112 * * 12.82 * 74.88
> > > > 9 * * * 118 * * 12.95 * 74.11
> > > > 10 * * *124 * * 13.12 * 73.16
>
> > > Nice job Tim. *At what speed do you get to top of lift before you
> > > reach the 10-knotter?
>
> > > The curves are pretty flat on the slow end down to just above best L/D
> > > speed (say Mc=1). Also notice that once you get to 3 knots on the
> > > McCready there is only 13 seconds of difference between the best and
> > > worst times up to Mc=9. That means you MUST center a thermal on the
> > > first circle to make it worthwhile to stop. Your 20% higher sink for
> > > circling flight is about right for 25-30 degrees of bank - at 45
> > > degrees the sink rate is around 70% higher, so the thermals would have
> > > to be fat AND easy to core. More typically I find that around 1/3 of
> > > the time when I turn in a thermal I wish I hadn't and another 1/3 of
> > > the time it takes me a few turns to center it - that's on a good day.
> > > I don't know if that nets out to a 20 second expected loss or not.
>
> > > I'm normally pickier about stopping to circle at the beginning of the
> > > street versus when I'm getting towards the end - though I am now
> > > reconsidering this view. Higher TAS with altitude and likely stronger
> > > lift closer to cloudbase would argue for climbing up early rather than
> > > late.
>
> > > My -27 is a little better on the run than my old LS-4 which I think
> > > biases the optimal a bit more to straight ahead climbing over
> > > circling. My flying style has certainly migrated in that direction.
>
> > > 9B
>
> > Just as a hypothetical, I figured out how much "centering loss" you
> > might experience rolling into a 10-knot thermal. *If I assume 1/3 of
> > the time you get 10 knots from the first turn, 1/3 of the time you
> > spend a turn at 2 knots and a turn at 5 knots before centering and 1/3
> > of the time you burn a circle on a thermal that turns out not to be
> > there, then the expected centering loss is the equivalent of 40
> > seconds circling at zero sink.
>
> > 9B
>
> Andy,
>
> I agree with BB, Ramy and your comments. *Everything depends on where
> you are flying and the strength and variability of the conditions.
> Out west at Logan or Parowan you can risk going faster and lower
> beneath the cloud streets, at Uvalde I learned the hard way not to
> drop more than 2000 feet below the clouds or risk not getting back up
> or if you do very slowly (I had one save from 400 feet agl). *The best
> lift was within 1000 feet of cloudbase on most streets.
>
> I will work on refining the model when I get time so I can run more
> scenarios easily. *The general things to take aways from the exercise
> is that slowing to minimum sink is not usually a good idea unless you
> really need to climb, best l/d or slightly faster is much better.
> That being slightly conservative costs very little on potential speed
> and reduces the risk factor of not finding that next big thermal. *The
> difference of flying at MC 4 to MC 7 is very little and you lose much
> less altitude at the lower MC to stay in the stonger lift and
> connected to the clouds.

Yup,

It was a little while back that I realized that speed in soaring
contests mostly has to do with the pilot's ability to estimate
probabilities. How fast or slow you fly in cruise matters little
within a broad range. How well you thermal matters only a bit. The
thing that takes you to the next level is being able to answer a
simple question over and over again "if I fly straight ahead on course
right now instead of circling, will I find a better thermal before I
get desperate?" The great pilots know when to press on and when to
stop and climb.

9B

[[email protected]]

On May 28, 2:42*am, Andy > wrote:
> On May 27, 3:25*pm, Tim Taylor > wrote:
>
>
>
>
>
> > On May 27, 8:25*am, Nine Bravo Ground > wrote:
>
> > > On May 27, 4:28*am, Nine Bravo Ground > wrote:
>
> > > > On May 26, 2:16*pm, Tim Taylor > wrote:
>
> > > > > On May 26, 1:09*pm, Tim Taylor > wrote:
>
> > > > > > On May 22, 10:02*am, Andy > wrote:
>
> > > > > > > On May 20, 7:47*pm, Ramy > wrote:
>
> > > > > > > > This is theoratically correct, but I found out more often than not
> > > > > > > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > > > > > > is stronger and more widespread than couple of thousands below, which
> > > > > > > > allows much faster speed without loosing altitude.
>
> > > > > > > > Ramy
>
> > > > > > > Okay, I broke out the polar and did some simple math for a "typical"
> > > > > > > scenario.
>
> > > > > > > Take a cloud street that is 16 miles long with 4 knot average thermal
> > > > > > > strength and 10 knot peak thermal strength.
>
> > > > > > > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > > > > > > knot net climb rate). It takes him 14 minutes to reach the end of the
> > > > > > > street and he has gained 3,360 feet.
>
> > > > > > > Pilot B climbs in the strong core, taking 30 seconds to center and
> > > > > > > achieving a 8.4 knot average climb rate thereafter (note the higher
> > > > > > > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > > > > > > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > > > > > > the process. Both pilots arrive at the end of the cloud street at the
> > > > > > > same time and altitude.
>
> > > > > > > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > > > > > > you behind the pilot who climbs straight ahead.
>
> > > > > > > Taking Ramy's point about stronger lift closer to cloudbase into
> > > > > > > account, let's assume you find an average 5 knots after a circling
> > > > > > > climb instead of 4 knots for climbing straight ahead. In this case you
> > > > > > > need a minimum 8.3 knot thermal before stopping to circle makes sense.
> > > > > > > If you assume 6 versus 4 knots average lift cruising closer to
> > > > > > > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > > > > > > feel like a pretty strong thermal strength gradient with altitude.
>
> > > > > > > This analysis doesn't take into account a slight true airspeed
> > > > > > > advantage for the pilot who climbs first due to his higher average
> > > > > > > altitude.
>
> > > > > > > Overall, I think this confirms that it pays to avoid circling under a
> > > > > > > cloudstreet for anything but the very strongest lift. *This is even
> > > > > > > more true if you are heading into an upwind turnpoint.
>
> > > > > > > 9B
>
> > > > > > Andy,
>
> > > > > > Ok, I tried to take your scenario and put some numbers to it. *I used
> > > > > > the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> > > > > > lift along the 16 mile street and a 10 knot thermal at the end. *The
> > > > > > glider starts at 3000 feet below the clouds and ends at cloud base at
> > > > > > the end of the street (similar to Reichmann). *I added 20 seconds of
> > > > > > centering penalty and a 20% higher sink rate while thermalling.
>
> > > > > > Don't know if the table will get scrambled on most viewers but here is
> > > > > > the data. *It shows that flying near MC speeds is optimum as long as
> > > > > > you can find the strong thermal ahead. *If you expect to find a strong
> > > > > > thermal ahead don't slow down to minimum sink speeds. *I think I will
> > > > > > use a MC speed of about (expected climb from thermal - average street
> > > > > > strength) as a good compromise MC setting to fly. *In this case 10 - 4
> > > > > > = MC setting of 6 to fly.
>
> > > > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > > > * * * * 45 * * *14.99 * 64.05
> > > > > > * * * * 50 * * *14.53 * 66.06
> > > > > > * * * * 55 * * *14.14 * 67.89
> > > > > > 0 * * * 60 * * *13.80 * 69.56
> > > > > > 1 * * * 68 * * *13.34 * 71.96
> > > > > > 2 * * * 76 * * *13.18 * 72.84
> > > > > > 3 * * * 84 * * *12.62 * 76.05
> > > > > > 4 * * * 90 * * *12.33 * 77.86
> > > > > > 5 * * * 94 * * *12.18 * 78.80
> > > > > > 6 * * * 98 * * *12.07 * 79.55
> > > > > > 7 * * * 104 * * 11.95 * 80.35
> > > > > > 8 * * * 112 * * 11.87 * 80.85
> > > > > > 9 * * * 118 * * 11.87 * 80.88
> > > > > > 10 * * *124 * * 11.90 * 80.66
>
> > > > > > TT
>
> > > > > Sorry found a sign error in my calculations. Here are the corrected
> > > > > values:
>
> > > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > > * * * * 45 * * *14.99 * 64.05
> > > > > * * * * 50 * * *14.53 * 66.06
> > > > > * * * * 55 * * *14.14 * 67.89
> > > > > 0 * * * 60 * * *13.80 * 69.56
> > > > > 1 * * * 68 * * *13.34 * 71.96
> > > > > 2 * * * 76 * * *13.25 * 72.46
> > > > > 3 * * * 84 * * *12.90 * 74.43
> > > > > 4 * * * 90 * * *12.75 * 75.27
> > > > > 5 * * * 94 * * *12.70 * 75.57
> > > > > 6 * * * 98 * * *12.69 * 75.68
> > > > > 7 * * * 104 * * 12.71 * 75.54
> > > > > 8 * * * 112 * * 12.82 * 74.88
> > > > > 9 * * * 118 * * 12.95 * 74.11
> > > > > 10 * * *124 * * 13.12 * 73.16
>
> > > > Nice job Tim. *At what speed do you get to top of lift before you
> > > > reach the 10-knotter?
>
> > > > The curves are pretty flat on the slow end down to just above best L/D
> > > > speed (say Mc=1). Also notice that once you get to 3 knots on the
> > > > McCready there is only 13 seconds of difference between the best and
> > > > worst times up to Mc=9. That means you MUST center a thermal on the
> > > > first circle to make it worthwhile to stop. Your 20% higher sink for
> > > > circling flight is about right for 25-30 degrees of bank - at 45
> > > > degrees the sink rate is around 70% higher, so the thermals would have
> > > > to be fat AND easy to core. More typically I find that around 1/3 of
> > > > the time when I turn in a thermal I wish I hadn't and another 1/3 of
> > > > the time it takes me a few turns to center it - that's on a good day.
> > > > I don't know if that nets out to a 20 second expected loss or not.
>
> > > > I'm normally pickier about stopping to circle at the beginning of the
> > > > street versus when I'm getting towards the end - though I am now
> > > > reconsidering this view. Higher TAS with altitude and likely stronger
> > > > lift closer to cloudbase would argue for climbing up early rather than
> > > > late.
>
> > > > My -27 is a little better on the run than my old LS-4 which I think
> > > > biases the optimal a bit more to straight ahead climbing over
> > > > circling. My flying style has certainly migrated in that direction.
>
> > > > 9B
>
> > > Just as a hypothetical, I figured out how much "centering loss" you
> > > might experience rolling into a 10-knot thermal. *If I assume 1/3 of
> > > the time you get 10 knots from the first turn, 1/3 of the time you
> > > spend a turn at 2 knots and a turn at 5 knots before centering and 1/3
> > > of the time you burn a circle on a thermal that turns out not to be
> > > there, then the expected centering loss is the equivalent of 40
> > > seconds circling at zero sink.
>
> > > 9B
>
> > Andy,
>
> > I agree with BB, Ramy and your comments. *Everything depends on where
> > you are flying and the strength and variability of the conditions.
> > Out west at Logan or Parowan you can risk going faster and lower
> > beneath the cloud streets, at Uvalde I learned the hard way not to
> > drop more than 2000 feet below the clouds or risk not getting back up
> > or if you do very slowly (I had one save from 400 feet agl). *The best
> > lift was within 1000 feet of cloudbase on most streets.
>
> > I will work on refining the model when I get time so I can run more
> > scenarios easily. *The general things to take aways from the exercise
> > is that slowing to minimum sink is not usually a good idea unless you
> > really need to climb, best l/d or slightly faster is much better.
> > That being slightly conservative costs very little on potential speed
> > and reduces the risk factor of not finding that next big thermal. *The
> > difference of flying at MC 4 to MC 7 is very little and you lose much
> > less altitude at the lower MC to stay in the stonger lift and
> > connected to the clouds.
>
> Yup,
>
> It was a little while back that I realized that speed in soaring
> contests mostly has to do with the pilot's ability to estimate
> probabilities. How fast or slow you fly in cruise matters little
> within a broad range. How well you thermal matters only a bit. The
> thing that takes you to the next level is being able to answer a
> simple question over and over again "if I fly straight ahead on course
> right now instead of circling, will I find a better thermal before I
> get desperate?" The great pilots know when to press on and when to
> stop and climb.
>
> 9B- Hide quoted text -
>
> - Show quoted text -

Expanding on your thought.
Right and left decisions are much more important than speed to fly
because of energy lines as well as improving the probability of
reduced sink and hopefully, that great circling climb.
I don't hear very many pilots I consider to be really good worrying
about how fast to fly, but they ALL pay attention to where to fly.
UH

[Papa3]

If anyone has a lot of time on their hands, the data from the recent
Mifflin Contest day 4 would make for some fascinating empirical
review. There was a single cloudstreet (super highway) running from
about the SW end of Nittany Mountain up to about Nisbet. Some guys
took a slower climb at the beginning of the street to climb all the
way up to cloudbase. Others "bumped up". Still others ran then
circled. I strongly believe that the biggest two determinants of the
speed in that section were:

a. Which part of the street you followed (i.e. the lateral decisions)
b. Which climb you took to get connected.

Link to the raw files is here:
http://www.ssa.org/members/contestreports/contestflighttraces/2010-05-21-ALL.zip

If someone wants to crunch the numbers, I can help by giving you some
sample flights to show the area/time in question.

P3




On May 28, 7:56*am, wrote:
> On May 28, 2:42*am, Andy > wrote:
>
>
>
>
>
> > On May 27, 3:25*pm, Tim Taylor > wrote:
>
> > > On May 27, 8:25*am, Nine Bravo Ground > wrote:
>
> > > > On May 27, 4:28*am, Nine Bravo Ground > wrote:
>
> > > > > On May 26, 2:16*pm, Tim Taylor > wrote:
>
> > > > > > On May 26, 1:09*pm, Tim Taylor > wrote:
>
> > > > > > > On May 22, 10:02*am, Andy > wrote:
>
> > > > > > > > On May 20, 7:47*pm, Ramy > wrote:
>
> > > > > > > > > This is theoratically correct, but I found out more often than not
> > > > > > > > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > > > > > > > is stronger and more widespread than couple of thousands below, which
> > > > > > > > > allows much faster speed without loosing altitude.
>
> > > > > > > > > Ramy
>
> > > > > > > > Okay, I broke out the polar and did some simple math for a "typical"
> > > > > > > > scenario.
>
> > > > > > > > Take a cloud street that is 16 miles long with 4 knot average thermal
> > > > > > > > strength and 10 knot peak thermal strength.
>
> > > > > > > > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > > > > > > > knot net climb rate). It takes him 14 minutes to reach the end of the
> > > > > > > > street and he has gained 3,360 feet.
>
> > > > > > > > Pilot B climbs in the strong core, taking 30 seconds to center and
> > > > > > > > achieving a 8.4 knot average climb rate thereafter (note the higher
> > > > > > > > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > > > > > > > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > > > > > > > the process. Both pilots arrive at the end of the cloud street at the
> > > > > > > > same time and altitude.
>
> > > > > > > > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > > > > > > > you behind the pilot who climbs straight ahead.
>
> > > > > > > > Taking Ramy's point about stronger lift closer to cloudbase into
> > > > > > > > account, let's assume you find an average 5 knots after a circling
> > > > > > > > climb instead of 4 knots for climbing straight ahead. In this case you
> > > > > > > > need a minimum 8.3 knot thermal before stopping to circle makes sense.
> > > > > > > > If you assume 6 versus 4 knots average lift cruising closer to
> > > > > > > > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > > > > > > > feel like a pretty strong thermal strength gradient with altitude.
>
> > > > > > > > This analysis doesn't take into account a slight true airspeed
> > > > > > > > advantage for the pilot who climbs first due to his higher average
> > > > > > > > altitude.
>
> > > > > > > > Overall, I think this confirms that it pays to avoid circling under a
> > > > > > > > cloudstreet for anything but the very strongest lift. *This is even
> > > > > > > > more true if you are heading into an upwind turnpoint.
>
> > > > > > > > 9B
>
> > > > > > > Andy,
>
> > > > > > > Ok, I tried to take your scenario and put some numbers to it. *I used
> > > > > > > the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> > > > > > > lift along the 16 mile street and a 10 knot thermal at the end. *The
> > > > > > > glider starts at 3000 feet below the clouds and ends at cloud base at
> > > > > > > the end of the street (similar to Reichmann). *I added 20 seconds of
> > > > > > > centering penalty and a 20% higher sink rate while thermalling.
>
> > > > > > > Don't know if the table will get scrambled on most viewers but here is
> > > > > > > the data. *It shows that flying near MC speeds is optimum as long as
> > > > > > > you can find the strong thermal ahead. *If you expect to find a strong
> > > > > > > thermal ahead don't slow down to minimum sink speeds. *I think I will
> > > > > > > use a MC speed of about (expected climb from thermal - average street
> > > > > > > strength) as a good compromise MC setting to fly. *In this case 10 - 4
> > > > > > > = MC setting of 6 to fly.
>
> > > > > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > > > > * * * * 45 * * *14.99 * 64.05
> > > > > > > * * * * 50 * * *14.53 * 66.06
> > > > > > > * * * * 55 * * *14.14 * 67.89
> > > > > > > 0 * * * 60 * * *13.80 * 69.56
> > > > > > > 1 * * * 68 * * *13.34 * 71.96
> > > > > > > 2 * * * 76 * * *13.18 * 72.84
> > > > > > > 3 * * * 84 * * *12.62 * 76.05
> > > > > > > 4 * * * 90 * * *12.33 * 77.86
> > > > > > > 5 * * * 94 * * *12.18 * 78.80
> > > > > > > 6 * * * 98 * * *12.07 * 79.55
> > > > > > > 7 * * * 104 * * 11.95 * 80.35
> > > > > > > 8 * * * 112 * * 11.87 * 80.85
> > > > > > > 9 * * * 118 * * 11.87 * 80.88
> > > > > > > 10 * * *124 * * 11.90 * 80.66
>
> > > > > > > TT
>
> > > > > > Sorry found a sign error in my calculations. Here are the corrected
> > > > > > values:
>
> > > > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > > > * * * * 45 * * *14.99 * 64.05
> > > > > > * * * * 50 * * *14.53 * 66.06
> > > > > > * * * * 55 * * *14.14 * 67.89
> > > > > > 0 * * * 60 * * *13.80 * 69.56
> > > > > > 1 * * * 68 * * *13.34 * 71.96
> > > > > > 2 * * * 76 * * *13.25 * 72.46
> > > > > > 3 * * * 84 * * *12.90 * 74.43
> > > > > > 4 * * * 90 * * *12.75 * 75.27
> > > > > > 5 * * * 94 * * *12.70 * 75.57
> > > > > > 6 * * * 98 * * *12.69 * 75.68
> > > > > > 7 * * * 104 * * 12.71 * 75.54
> > > > > > 8 * * * 112 * * 12.82 * 74.88
> > > > > > 9 * * * 118 * * 12.95 * 74.11
> > > > > > 10 * * *124 * * 13.12 * 73.16
>
> > > > > Nice job Tim. *At what speed do you get to top of lift before you
> > > > > reach the 10-knotter?
>
> > > > > The curves are pretty flat on the slow end down to just above best L/D
> > > > > speed (say Mc=1). Also notice that once you get to 3 knots on the
> > > > > McCready there is only 13 seconds of difference between the best and
> > > > > worst times up to Mc=9. That means you MUST center a thermal on the
> > > > > first circle to make it worthwhile to stop. Your 20% higher sink for
> > > > > circling flight is about right for 25-30 degrees of bank - at 45
> > > > > degrees the sink rate is around 70% higher, so the thermals would have
> > > > > to be fat AND easy to core. More typically I find that around 1/3 of
> > > > > the time when I turn in a thermal I wish I hadn't and another 1/3 of
> > > > > the time it takes me a few turns to center it - that's on a good day.
> > > > > I don't know if that nets out to a 20 second expected loss or not..
>
> > > > > I'm normally pickier about stopping to circle at the beginning of the
> > > > > street versus when I'm getting towards the end - though I am now
> > > > > reconsidering this view. Higher TAS with altitude and likely stronger
> > > > > lift closer to cloudbase would argue for climbing up early rather than
> > > > > late.
>
> > > > > My -27 is a little better on the run than my old LS-4 which I think
> > > > > biases the optimal a bit more to straight ahead climbing over
> > > > > circling. My flying style has certainly migrated in that direction.

[Nine Bravo Ground]

On May 28, 4:56*am, wrote:
> On May 28, 2:42*am, Andy > wrote:
>
>
>
>
>
> > On May 27, 3:25*pm, Tim Taylor > wrote:
>
> > > On May 27, 8:25*am, Nine Bravo Ground > wrote:
>
> > > > On May 27, 4:28*am, Nine Bravo Ground > wrote:
>
> > > > > On May 26, 2:16*pm, Tim Taylor > wrote:
>
> > > > > > On May 26, 1:09*pm, Tim Taylor > wrote:
>
> > > > > > > On May 22, 10:02*am, Andy > wrote:
>
> > > > > > > > On May 20, 7:47*pm, Ramy > wrote:
>
> > > > > > > > > This is theoratically correct, but I found out more often than not
> > > > > > > > > that the lift right below cloudbase (keeping FAR clearance of course)
> > > > > > > > > is stronger and more widespread than couple of thousands below, which
> > > > > > > > > allows much faster speed without loosing altitude.
>
> > > > > > > > > Ramy
>
> > > > > > > > Okay, I broke out the polar and did some simple math for a "typical"
> > > > > > > > scenario.
>
> > > > > > > > Take a cloud street that is 16 miles long with 4 knot average thermal
> > > > > > > > strength and 10 knot peak thermal strength.
>
> > > > > > > > Pilot A pulls back to 70 knots and climbs in the average lift (2.4
> > > > > > > > knot net climb rate). It takes him 14 minutes to reach the end of the
> > > > > > > > street and he has gained 3,360 feet.
>
> > > > > > > > Pilot B climbs in the strong core, taking 30 seconds to center and
> > > > > > > > achieving a 8.4 knot average climb rate thereafter (note the higher
> > > > > > > > sink rate for circling flight). After 5 minutes he has climbed 4,250
> > > > > > > > feet. He then cruises at 110 knots for 16 miles, giving up 890 feet in
> > > > > > > > the process. Both pilots arrive at the end of the cloud street at the
> > > > > > > > same time and altitude.
>
> > > > > > > > Conclusion: stopping to circle in a thermal weaker than 10 knots puts
> > > > > > > > you behind the pilot who climbs straight ahead.
>
> > > > > > > > Taking Ramy's point about stronger lift closer to cloudbase into
> > > > > > > > account, let's assume you find an average 5 knots after a circling
> > > > > > > > climb instead of 4 knots for climbing straight ahead. In this case you
> > > > > > > > need a minimum 8.3 knot thermal before stopping to circle makes sense.
> > > > > > > > If you assume 6 versus 4 knots average lift cruising closer to
> > > > > > > > cloudbase then you only need a 6.6 knot thermal - but that starts to
> > > > > > > > feel like a pretty strong thermal strength gradient with altitude.
>
> > > > > > > > This analysis doesn't take into account a slight true airspeed
> > > > > > > > advantage for the pilot who climbs first due to his higher average
> > > > > > > > altitude.
>
> > > > > > > > Overall, I think this confirms that it pays to avoid circling under a
> > > > > > > > cloudstreet for anything but the very strongest lift. *This is even
> > > > > > > > more true if you are heading into an upwind turnpoint.
>
> > > > > > > > 9B
>
> > > > > > > Andy,
>
> > > > > > > Ok, I tried to take your scenario and put some numbers to it. *I used
> > > > > > > the polar for an LS-4 (no DG 300 numbers handy). *I Assumed 4 knots of
> > > > > > > lift along the 16 mile street and a 10 knot thermal at the end. *The
> > > > > > > glider starts at 3000 feet below the clouds and ends at cloud base at
> > > > > > > the end of the street (similar to Reichmann). *I added 20 seconds of
> > > > > > > centering penalty and a 20% higher sink rate while thermalling.
>
> > > > > > > Don't know if the table will get scrambled on most viewers but here is
> > > > > > > the data. *It shows that flying near MC speeds is optimum as long as
> > > > > > > you can find the strong thermal ahead. *If you expect to find a strong
> > > > > > > thermal ahead don't slow down to minimum sink speeds. *I think I will
> > > > > > > use a MC speed of about (expected climb from thermal - average street
> > > > > > > strength) as a good compromise MC setting to fly. *In this case 10 - 4
> > > > > > > = MC setting of 6 to fly.
>
> > > > > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > > > > * * * * 45 * * *14.99 * 64.05
> > > > > > > * * * * 50 * * *14.53 * 66.06
> > > > > > > * * * * 55 * * *14.14 * 67.89
> > > > > > > 0 * * * 60 * * *13.80 * 69.56
> > > > > > > 1 * * * 68 * * *13.34 * 71.96
> > > > > > > 2 * * * 76 * * *13.18 * 72.84
> > > > > > > 3 * * * 84 * * *12.62 * 76.05
> > > > > > > 4 * * * 90 * * *12.33 * 77.86
> > > > > > > 5 * * * 94 * * *12.18 * 78.80
> > > > > > > 6 * * * 98 * * *12.07 * 79.55
> > > > > > > 7 * * * 104 * * 11.95 * 80.35
> > > > > > > 8 * * * 112 * * 11.87 * 80.85
> > > > > > > 9 * * * 118 * * 11.87 * 80.88
> > > > > > > 10 * * *124 * * 11.90 * 80.66
>
> > > > > > > TT
>
> > > > > > Sorry found a sign error in my calculations. Here are the corrected
> > > > > > values:
>
> > > > > > MC * * *miles/hr * * * *time minutes * *Average speed
> > > > > > * * * * 45 * * *14.99 * 64.05
> > > > > > * * * * 50 * * *14.53 * 66.06
> > > > > > * * * * 55 * * *14.14 * 67.89
> > > > > > 0 * * * 60 * * *13.80 * 69.56
> > > > > > 1 * * * 68 * * *13.34 * 71.96
> > > > > > 2 * * * 76 * * *13.25 * 72.46
> > > > > > 3 * * * 84 * * *12.90 * 74.43
> > > > > > 4 * * * 90 * * *12.75 * 75.27
> > > > > > 5 * * * 94 * * *12.70 * 75.57
> > > > > > 6 * * * 98 * * *12.69 * 75.68
> > > > > > 7 * * * 104 * * 12.71 * 75.54
> > > > > > 8 * * * 112 * * 12.82 * 74.88
> > > > > > 9 * * * 118 * * 12.95 * 74.11
> > > > > > 10 * * *124 * * 13.12 * 73.16
>
> > > > > Nice job Tim. *At what speed do you get to top of lift before you
> > > > > reach the 10-knotter?
>
> > > > > The curves are pretty flat on the slow end down to just above best L/D
> > > > > speed (say Mc=1). Also notice that once you get to 3 knots on the
> > > > > McCready there is only 13 seconds of difference between the best and
> > > > > worst times up to Mc=9. That means you MUST center a thermal on the
> > > > > first circle to make it worthwhile to stop. Your 20% higher sink for
> > > > > circling flight is about right for 25-30 degrees of bank - at 45
> > > > > degrees the sink rate is around 70% higher, so the thermals would have
> > > > > to be fat AND easy to core. More typically I find that around 1/3 of
> > > > > the time when I turn in a thermal I wish I hadn't and another 1/3 of
> > > > > the time it takes me a few turns to center it - that's on a good day.
> > > > > I don't know if that nets out to a 20 second expected loss or not..
>
> > > > > I'm normally pickier about stopping to circle at the beginning of the
> > > > > street versus when I'm getting towards the end - though I am now
> > > > > reconsidering this view. Higher TAS with altitude and likely stronger
> > > > > lift closer to cloudbase would argue for climbing up early rather than
> > > > > late.
>
> > > > > My -27 is a little better on the run than my old LS-4 which I think
> > > > > biases the optimal a bit more to straight ahead climbing over
> > > > > circling. My flying style has certainly migrated in that direction.

[John Cochrane]

As a rule of thumb, flying 1/2 the MC recommended by the theory/
computer
is a good strategy.

If flying 1/2 the Mc recommended by "theory" is right, that means you
have the wrong theory! Understanding a correct theory is even
better.

Lowering the Mc setting to account for centering time, bottom to top
average rather than best surge of lift, need to stay connected to the
clouds, character of sky ahead, altitude remaining, number of options
left, etc. etc. is all part of good theory.

This does not mean it ever makes sense to cruise at a setting lower
than the weakest lift you would currently take, which is a common
mistake. If you're cruising at Mc 1, but rejecting smooth easy to core
3 knot (bottom to top) thermals while waiting for the 5 knotter, you
would do better by cruising at Mc 2 and taking the occasional 3
knotter, but not going too high in it.

As UH emphasizes, all this is third order. Flying in the right place
and altitude band is the key decision!

John Cochrane
BB

[Chris Reed[_2_]]

John Cochrane wrote:
> As a rule of thumb, flying 1/2 the MC recommended by the theory/
> computer
> is a good strategy.
>
> If flying 1/2 the Mc recommended by "theory" is right, that means you
> have the wrong theory! Understanding a correct theory is even
> better.
>
> Lowering the Mc setting to account for centering time, bottom to top
> average rather than best surge of lift, need to stay connected to the
> clouds, character of sky ahead, altitude remaining, number of options
> left, etc. etc. is all part of good theory.
>
> This does not mean it ever makes sense to cruise at a setting lower
> than the weakest lift you would currently take, which is a common
> mistake. If you're cruising at Mc 1, but rejecting smooth easy to core
> 3 knot (bottom to top) thermals while waiting for the 5 knotter, you
> would do better by cruising at Mc 2 and taking the occasional 3
> knotter, but not going too high in it.
>
> As UH emphasizes, all this is third order. Flying in the right place
> and altitude band is the key decision!
>
> John Cochrane
> BB

I don't think I made it clear that I was translating into a rule of
thumb for the average pilot (say, me). I've read your paper, and am
convinced, but translating that into my own flying requires a precision
that I just don't have.

So for my purposes I have two options:

Blindly follow the PDA. This would work if I did it, I guess, but
distractions (ooh, shiny!), inattention, picking the wrong route, etc
mean I don't.

Work out a (an?) heuristic which approximates the theory. For me, that's
to take the general reading on my averager during the last thermal
(well. more accurately the next thermal I would accept, but if that part
of the day is consistent the last thermal will do), and then estimate
the relevant MC to fly to the next thermal.

My achieved climb is probably 2/3 of what I see on the averager (UK,
lower cloudbases than many places), so I could choose that as my MC
figure. However, I also know from reading the theory that the difference
in achieved speed over the ground from flying a little faster or slower
is a few percent (2-5).

Thus if I fly half what I saw on the averager, I'm close to (though
below) the optimum speed, but have extended my range. Or, to put it
another way, I'm flying less than 5% slower than someone who can fly to
the theory, but giving myself a larger margin to compensate for the
errors I know I will make.

This is not a strategy for winning contests, but it works quite well for
completing the task. It's also pretty simple in the UK because a day
when I see 6kt consistently on the averager is execptional. Thus I only
need to memorise four speeds to fly (MC 0-3) and remember how to divide
by 2.