Spitfires and High Speed Roll Rate

[Anaxogoras]

I don't know why you guys keep ignoring me .  You can't even begin an apples to apples comparison between different Spits until you have figured out how to null out yaw and sideslip out of you flight tests.  There's no point to even begin comparing until you do because your numbers aren't isolated to begin to even compare the specific variables you're trying to compare.  Good luck figuring a way to null those out because when you roll, you yaw and sideslip - when you yaw, you sideslip and roll, etc.

And that's just for starters .

Tango, XO
412th FS Braunco Mustangs

We appreciate your input tango.  Generally you like to pop into these threads and throw out the most abstract analysis possible and hope that we deduce the more specific point you have in mind.  That's probably not going to happen.

Let me ask you a question:  Does your above statement imply that I cannot reach the judgment that one spitfire outrolls another, unless I "null out" yaw and slip with the scientific precision of an engineer?  Please give me some specifics.  I think I understand your point, but it would seem to be applicable to cases where the roll rates are very close, not where they are drastically different.  For example, I don't need to null out yaw and sideslip to know that the 109F outrolls the 109E at 400mph.  The former rolls without too much difficulty, and the latter almost doesn't roll at all!

[Karnak]

dtango,

Because your argument is basically that we can't know anything or even guess at anything or even use historical test results unless we know everything.  It isn't useful unless we want to just stick our heads in the sand and play know-nothings.

[715]

Why not compare to archival footage?  It always seemed to me that our Spit Is rolled way slower than archival footage of Spit Is peeling off.  Of course it is not possible to estimate airspeed from movies but for movie footage you could guess they are going between max cruise and military.

[BaldEagl]

We appreciate your input tango.  Generally you like to pop into these threads and throw out the most abstract analysis possible and hope that we deduce the more specific point you have in mind.  That's probably not going to happen.

Let me ask you a question:  Does your above statement imply that I cannot reach the judgment that one spitfire outrolls another, unless I "null out" yaw and slip with the scientific precision of an engineer?  Please give me some specifics.  I think I understand your point, but it would seem to be applicable to cases where the roll rates are very close, not where they are drastically different.  For example, I don't need to null out yaw and sideslip to know that the 109F outrolls the 109E at 400mph.  The former rolls without too much difficulty, and the latter almost doesn't roll at all!

Geez, even I have to agree with you on this one.  Who really cares about what forces are acting on what.  The question is when you crank the ailerons all the way in one direction or the other does the thing roll at a rate consistant with what it did in real life or not but as Brooke already brought up, the only way to know is to have hard data on real life roll rates.

[Angus]

Why not compare to archival footage?  It always seemed to me that our Spit Is rolled way slower than archival footage of Spit Is peeling off.  Of course it is not possible to estimate airspeed from movies but for movie footage you could guess they are going between max cruise and military.

Now that was an idea. Really!
Should of course be testable in RL, still lots of Spits around. I've seen a Mk I in the air, but think there is only one. But V and IX's are by the bundles...

[dtango]

Gavagai:

I have no issue with doing that test and saying one out-rolls the other.  But that's as much as you'll be able to deduce from the way it's tested. 

But if we want to now try to analyze WHY one out-rolls the other and HOW that's CORRECT or INCORRECT then we're going to have to do a lot more in the testing and analysis to figure that out.  From your tests results we are not able to make conclusions of the nature that people are making conclusions about.  Examine the dynamics involved then you'll understand why.  I'm sorry I have neither the time or energy to explain the physics.  I and others have explained just SOME of it before in this thread that I already referenced.  Here it is again if you want to glance at it.

http://bbs.hitechcreations.com/smf/index.php/topic,253393.0.html

The easier alternative is to do what Brooke suggests.  Look for real flight tests to compare to.  Even there you need to understand the way the've done the tests and the dynamics they are isolating.

dtango,

Because your argument is basically that we can't know anything or even guess at anything or even use historical test results unless we know everything.  It isn't useful unless we want to just stick our heads in the sand and play know-nothings.

Karnak - that's not what my argument is at all .

Tango, XO
412th FS Braunco Mustangs

[Anaxogoras]

Gavagai:

I have no issue with doing that test and saying one out-rolls the other.  But that's as much as you'll be able to deduce from the way it's tested. 

But if we want to now try to analyze WHY one out-rolls the other and HOW that's CORRECT or INCORRECT then we're going to have to do a lot more in the testing and analysis to figure that out.  From your tests results we are not able to make conclusions of the nature that people are making conclusions about. 

Fair enough.  I think that for the rest of us, because the IX and V were nearly the same airframe, we are extremely skeptical.  For my part, after testing all of the roll rates of all of our aircraft, there is nothing comparable to the V and IX.  The 109F-K are too close for my stopwatch to measure the difference (despite becoming heavier, having redesigned v-stabs, redesigned cowlings, etc.).  The 190 series shows little change, and the same is true of the P-47D series, the Yaks; even the Hurricane IID doesn't roll much worse than the Mk I or IIC.  So here I am, I observe all these examples of consistency despite small changes in weight distribution, oil cooler placement, weapon loadouts etc., and then I get to the Spitfire V and IX and they are drastically different at 400mph.  So yeah, I am skeptical, and because of the consistency across small changes in our other aircraft the burden of proof is on those who say the V and IX should be so different.

Now, I don't believe you're saying they should be different.  You're right that we need test data to be sure.  Let's come full circle and agree that there is reason for HTC to check their numbers.  Need I remind everyone of how much some of their flight models have changed over the years?  We had the Spitfire V and IX in Warbirds and they almost rolled the same.  Check out http://www.rdrop.com/users/hoofj/wbroll.htm  Hoof's numbers are also for one 360 degree roll.

Roll rates:  150    200   250   300   350    400
SpitV     4.6s      3.6s  4.1s   5.1s  7.1s     10.4s
SpitIX    5.0s      3.7s  4.2s  5.4s    7.9s    11.4s

So the same person, HT, modeled those Spits and they rolled about the same.  The difference is that the Spit V rolls much better at 400mph than it did then!

[jocko-]

Finding real period test results would help. Knowing that the Mk VIII's roll rate is based on extended wingtips it makes more sense that it lags behind the others but the IX's roll rate gives me pause!

[dtango]

Gavagai:

I understand your logic.  However you and others have made too many aerodynamic assumptions.

For starters are you sure you removed the effect of yaw or sideslip from your roll tests between the Spit V and Spit IX?  Because yaw and sideslip will change your roll performance.

For grins let's say that you have and assume we can simplify the analysis just to one degree of freedom.  The following is the general equation for estimating roll rate in the one degree of freedom case:



It covers both the roll accleration and steady state roll.  It's already been mentioned that the Spit IX has two radiators vs. the Spit V's one.  Why does a small thing like that matter?  Because the roll transient time constant 1/T, which determines roll accleration is impacted by the rolling moment of inertia (Ixx).  Here's the equation for 1/T:



where:
Q = dynamic pressure (.5*air_density*velocity^2)
S = wing surface area
b = wing span
Clp = coefficient of rolling moment due to rolling
Ixx = rolling moment of inertia
V = velocity

The 2 radiators under each wing increases the rolling moment of inertia.  Increase the moment of inertia reduces 1/T which reduces roll acceleration.

And that's just one variable.  There are others.  Roll performance is complicated and there are more details to be considered that can affect it than the factors folks have brought up in this thread.

Why do the WB roll rates differ from AH for the Spit V and IX?  Beats the heck outta me .

Should HTC look the AH Spit V and Spit IX numbers?  Sure thing.

Is there something wrong?  Maybe / Maybe Not, but I sure the heck can't tell from speculating from your data points and the simple factors that people have thrown out so far for why it is wrong because we're totally ignoring other key aerodynamic factors as well.

Tango, XO
412th FS Braunco Mustangs

[Anaxogoras]

I can see why the Spit IX's roll acceleration would be slower based on your clear explanation of the second formula.  I am still studying the link you gave in this thread, so I'm not going to ask you for help with the first formula yet.

On the other hand, roll acceleration rates are minimally responsible for the different rates I measured.  Suppose I had nulled out yaw and side slip, and suppose that we weren't interested in roll acceleration, but only roll rate.  What sort of changes in the Spitfire IX impact roll rate alone?

[SgtPappy]

Ah yes. Thanks Tango for another great page of explanations The math behind turning page is also favourite'd.

[Angus]

Anaxo, - the load would. Or rather/also the weight distribution.

[Anaxogoras]

Anaxo, - the load would. Or rather/also the weight distribution.

Like the 109/p47/190/hurricane/yak series?

[fudgums]

Gavagai:

I understand your logic.  However you and others have made too many aerodynamic assumptions.

For starters are you sure you removed the effect of yaw or sideslip from your roll tests between the Spit V and Spit IX?  Because yaw and sideslip will change your roll performance.

For grins let's say that you have and assume we can simplify the analysis just to one degree of freedom.  The following is the general equation for estimating roll rate in the one degree of freedom case:

(Image removed from quote.)

It covers both the roll accleration and steady state roll.  It's already been mentioned that the Spit IX has two radiators vs. the Spit V's one.  Why does a small thing like that matter?  Because the roll transient time constant 1/T, which determines roll accleration is impacted by the rolling moment of inertia (Ixx).  Here's the equation for 1/T:

(Image removed from quote.)

where:
Q = dynamic pressure (.5*air_density*velocity^2)
S = wing surface area
b = wing span
Clp = coefficient of rolling moment due to rolling
Ixx = rolling moment of inertia
V = velocity

The 2 radiators under each wing increases the rolling moment of inertia.  Increase the moment of inertia reduces 1/T which reduces roll acceleration.

And that's just one variable.  There are others.  Roll performance is complicated and there are more details to be considered that can affect it than the factors folks have brought up in this thread.

Why do the WB roll rates differ from AH for the Spit V and IX?  Beats the heck outta me .

Should HTC look the AH Spit V and Spit IX numbers?  Sure thing.

Is there something wrong?  Maybe / Maybe Not, but I sure the heck can't tell from speculating from your data points and the simple factors that people have thrown out so far for why it is wrong because we're totally ignoring other key aerodynamic factors as well.

Tango, XO
412th FS Braunco Mustangs

I am so retarded 

[Angus]

Like the 109/p47/190/hurricane/yak series?

Not sure what you're getting at, but look at the difference when wing tanks are full/empty for instance.