Optimum CG Range

Herb wrote:
Another useful approach is start at about 66% aft using manufacturer's
CG range. When making the tightest turn you normally do, if you run
out of elevator, you need to shift CG back a bit. You will probably
end up around 75%.There isn't a huge benefit in having the CG way
back, but there is a significant deterioration of handling which
requires better pilot skills to offset.The last little bit of glider
performance costs quite a bit in pilot workload until you are very
proficient. I usually take a couple pounds out of the tail in the
Spring and put it back in when my skills are back up to snuff.
FWIW
UH


Here's another gem piece of advice: With my 3D model airplanes I roll
inverted and check if I need down elevator to stay level. If so, the
cg needs to be moved further back. A well set up model will happily
fly inverted without elevator movement! Haven't tried that in my LS8,
though.
Seriously, Hanks and Eric's methods will both work well. As long as
inside the book range, find the cg that gives you good handling and
enough up elevator to stall the plane in a steep turn. It'll be at
75%-90% aft.
Why is the ability to stall in a steep turn a useful criteria? It sounds
like a safety problem to me.

--
Eric Greenwell - Washington State, USA
* Change "netto" to "net" to email me directly

Herb wrote:
Another useful approach is start at about 66% aft using manufacturer's
CG range. When making the tightest turn you normally do, if you run
out of elevator, you need to shift CG back a bit. You will probably
end up around 75%.There isn't a huge benefit in having the CG way
back, but there is a significant deterioration of handling which
requires better pilot skills to offset.The last little bit of glider
performance costs quite a bit in pilot workload until you are very
proficient. I usually take a couple pounds out of the tail in the
Spring and put it back in when my skills are back up to snuff.
FWIW
UH


Here's another gem piece of advice: With my 3D model airplanes I roll
inverted and check if I need down elevator to stay level. If so, the
cg needs to be moved further back. A well set up model will happily
fly inverted without elevator movement! Haven't tried that in my LS8,
though.
Seriously, Hanks and Eric's methods will both work well. As long as
inside the book range, find the cg that gives you good handling and
enough up elevator to stall the plane in a steep turn. It'll be at
75%-90% aft.

Why is the ability to stall in a steep turn a useful criteria? It sounds
like a safety problem to me.

--
Eric Greenwell - Washington State, USA
* Change "netto" to "net" to email me directly

Herb wrote:
...find the cg that gives you good handling and
enough up elevator to stall the plane in a steep turn. It'll be at
75%-90% aft.

Herb, J7

Reminds me of the old rule of thumb, if you could call it that:
CG too far forward, can't pull the nose up for takeoff.
CG too far aft, can't recover from a stall/spin.

This method can be expensive in airplanes though... :-)

Brian W

Thanks, guys - I've read all your responses, thought about them and
have been Googling to learn more. It looks to me like Bob W's "rant"
was correct, and that JJ's method is the best way to go.

Brian said "there is a known range of allowable CofG's in terms of
%MAC which is similar across a wide range of airframes", so I decided
to see what that might be. The best I could tell a reasonable range
was 15% to 35% of MAC. However, it's really a loosey, goosey "range" -
the HP-18 Bob K referenced has a 25% to 40% of MAC actual range, while
a DC8 has a 8% to 18% of MAC actual range! I did discover that by
using the 15% to 35% of MAC range Frank Irving's optimal CG is 75% to
100% of the allowed CG range (where 0% is the fwd limit and 100% is
the aft limit). It isn't very precise, though, and it doesn't agree
with the Akaflieg Braunschweig findings.

I then calculated the arithmetic mean chord of the LS8 wing by
dividing span by aspect ratio, and got 700mm. This looks about right,
since the root chord is 900mm, and yes, I know it's not the MAC. Then,
I found a scale drawing of the LS8 at: http://www.dg-flugzeugbau.de/Data/3s-ls8-s.pdf
If I'm reading it right, the 25% MAC will be located 225mm behind the
wing LE. Using the arithmetic mean chord of 700mm and the allowed CG
range of 280mm to 400mm behind the wing LE, I calculated that the LS8
has a CG range on the order of 33% to 50% of MAC. That seems wrong
enough that it isn't worth the bother of actually going through the
geometric excercise of calculating the MAC on the triple tapered LS8
wing. I cheerfully admit there's a possibility that I don't know what
I'm doing, but at this point I'm going to drop Irving's approach.

This gets me to JJs advice. I'll follow the 2001 Akaflieg Braunschweig
method, and use the tail tank to set the CG to 65% of the allowable
range. Once I'm familiar with how she flies there, I'll move it back 5
to 10% at a time until I either reach 90% or get to JJ's criteria
point (I have to trim forward when entering a thermal), whereupon I'll
bring it forward 5%.

I'd be grateful for any further comments or suggestions.

-John

On Dec 13, 1:34 pm, jcarlyle wrote:

...I then calculated the arithmetic mean chord of the LS8 wing by
dividing span by aspect ratio, and got 700mm. This looks about right,
since the root chord is 900mm, and yes, I know it's not the MAC. Then,
I found a scale drawing of the LS8 at: http://www.dg-flugzeugbau.de/Data/3s-ls8-s.pdf
If I'm reading it right, the 25% MAC will be located 225mm behind the
wing LE...

Hmm... That doesn't seem right. Using the DJ Aerotech graphical MAC
method, the same LS8 drawing, and information on the LS8 from Thomas'
Fundamentals of Sailplane Design (thanks again, Judah!) that places
the planform break at 0.6 semispan, I got:

* MAC length of ~736mm
* MAC LE location of ~41mm aft of the wing LE at side of body _or_
* MAC LE location of ~45mm aft of the intersection of the projected
leading edge and the plane of symmetry (yeah, who uses that?)

Given that the LS8's double-trapezoid planform gives it more MAC per
unit area than the HP-18's eminently buildable single trapezoid
planform, and that the LS8's wing is unswept along the 25% chord line
as opposed to the HP-18 being unswept along the 41.25% chord, the MAC
and MAC LE numbers I got sound about right to me. But, hey, I'm a
college dropout with no engineering training, what do I know?

As regards the suggested CG location for the HP-series, Dick Schreder
typically suggested 25% to 40% MAC as the allowable range. Based on an
analysis of the margin of static stability of the HP-18 done by Steve
Smith (that's Dr. Smith to you Mythbusters fans), and based on my own
experience flying an HP-18 with CG back around 40%, I currently
recommend that HP operators limit their operation to 25% to 35% MAC.

For an extra 245 Europes, I will be glad to translate that into a
dimensional range aft of the wing leading edge at side-of-body. For
245 Australias, I will do all the above and throw in a wisecracking
reality-show cameraman. We are not accepting any other continents at
this time.

Thanks, and best regards to all

Bob K.
www.hpaircraft.com ---- now with 245% less ondulation!

On Dec 13, 9:30*pm, Bob Kuykendall wrote:

For an extra 245 Europes, I will be glad to translate that into a
dimensional range aft of the wing leading edge at side-of-body. For
245 Australias, I will do all the above and throw in a wisecracking
reality-show cameraman. We are not accepting any other continents at
this time.

Thanks, and best regards to all

Bob K.www.hpaircraft.com---- now with 245% less ondulation!

So, Bob, does this mean you are only accepting two out of continent
offers, and no in-continent offers at this time? Will this now
qualify me as the wisecracking reality show cameraman?

I have various manuals on various planes that say anything and
everything from X% to Y% of the root chord, to X% to y% of the chord .
5 meters outboard of the side of the fuselage. Best reference that
can be given for the average pilot is a set of dimensions from an easy
to identify point. Forward face of the forward drag spar in the
fuselage on an HP-18 is an excellent datum.

Steve Leonard
:-)

Bob, thanks for figuring out the MAC of the LS8. I wasn’t clear on
how to “fudge” (DJ Aerotech’s term) the winglet and the wing root to
get the required area for use in their graphical calculation of MAC.

Your values of 736 mm for MAC, with a MAC LE of 41 mm aft of the LE of
the wing root, jibe quite well with my arithmetic chord of 700 mm and
my arithmetic chord LE of 50 mm aft of the LE of the wing root. I’m
not clear why you said it didn’t seem right.

But thanks to you I can now calculate the LS8 CG limits in terms of
MAC. The fwd CG limit is 280 mm aft of the LE of the wing root, and
the aft CG limit is 400 mm aft of the LE of the wing root (both values
from TCDS G14CE). So:

Fwd CG limit = 100 * (280 - 41) / 736 = 32% MAC
Aft CG limit = 100 * (400 - 41) / 736 = 49% MAC

This result (a) makes the LS8 odd from the perspective of SE light
aircraft (typical CG range from 15% to 35% of MAC) and the HP-18 (was
25% to 40%, now 25% to 35%), and (b) means that the 1981 Frank Irving
optimum CG guideline of 30% to 35% of MAC isn’t useful. So that
answers my original questions.

-John

On Dec 13, 10:30 pm, Bob Kuykendall wrote:
Hmm... That doesn't seem right. Using the DJ Aerotech graphical MAC
method, the same LS8 drawing, and information on the LS8 from Thomas'
Fundamentals of Sailplane Design (thanks again, Judah!) that places
the planform break at 0.6 semispan, I got:

* MAC length of ~736mm
* MAC LE location of ~41mm aft of the wing LE at side of body _or_
* MAC LE location of ~45mm aft of the intersection of the projected
leading edge and the plane of symmetry (yeah, who uses that?)

Given that the LS8's double-trapezoid planform gives it more MAC per
unit area than the HP-18's eminently buildable single trapezoid
planform, and that the LS8's wing is unswept along the 25% chord line
as opposed to the HP-18 being unswept along the 41.25% chord, the MAC
and MAC LE numbers I got sound about right to me. But, hey, I'm a
college dropout with no engineering training, what do I know?

As regards the suggested CG location for the HP-series, Dick Schreder
typically suggested 25% to 40% MAC as the allowable range. Based on an
analysis of the margin of static stability of the HP-18 done by Steve
Smith (that's Dr. Smith to you Mythbusters fans), and based on my own
experience flying an HP-18 with CG back around 40%, I currently
recommend that HP operators limit their operation to 25% to 35% MAC.

jcarlyle wrote:
Fwd CG limit = 100 * (280 - 41) / 736 = 32% MAC
Aft CG limit = 100 * (400 - 41) / 736 = 49% MAC

John,
My experience and judgement tell me your figures are way too far aft.
I suspect the distance aft of the leading edge to the zero MAC is in
error. One sure way to check this is to mark the MAC on both wings
then assemble the wing without the fuselage on saw-horses. Then snap a
string from both zero MAC's and measure the distance from the string
to the leading edge at root rib.
Cheers,
JJ

JJ, I place great weight on your experience and judgement! I agree,
what I calculated is way far back. But I've looked over my figures
very carefully a number of times, and if there's an error I sure can't
find it.

As for the LE to zero MAC dimension causing the problem, we can get a
figure of merit by calculating the CG limits for the root chord:

Fwd CG limit = 100 * 280 / 900 = 31% root chord
Aft CG limit = 100 * 400 / 900 = 44% root chord

Of course it's different from the 32% to 49% MAC, or the 33% to 50%
for the arithmetic chord, but all three result are in the same ball
park. Measuring the actual wings as per your write-up is a good
suggestion, but due to available work area and the weather I won't be
able to try that for about 5 months at the earliest.

-John

On Dec 15, 10:49 am, JJ Sinclair wrote:
jcarlyle wrote:

Fwd CG limit = 100 * (280 - 41) / 736 = 32% MAC

Aft CG limit = 100 * (400 - 41) / 736 = 49% MAC

John,
My experience and judgement tell me your figures are way too far aft.
I suspect the distance aft of the leading edge to the zero MAC is in
error. One sure way to check this is to mark the MAC on both wings
then assemble the wing without the fuselage on saw-horses. Then snap a
string from both zero MAC's and measure the distance from the string
to the leading edge at root rib.
Cheers,
JJ