Author Topic: DVL data on Bf 109 roll rates  (Read 24981 times)

Offline Angus

  • Plutonium Member
  • *******
  • Posts: 10057
DVL data on Bf 109 roll rates
« Reply #180 on: October 07, 2005, 04:00:34 AM »
Spit V's were rogue anyway, - no two the same.
Do you know when they stiffened the wing? I seem to recall that it was with the VIII model, but can't find the info on it.
It was very interesting to carry out the flight trials at Rechlin with the Spitfire and the Hurricane. Both types are very simple to fly compared to our aircraft, and childishly easy to take-off and land. (Werner Mölders)

Offline gripen

  • Silver Member
  • ****
  • Posts: 1914
DVL data on Bf 109 roll rates
« Reply #181 on: October 07, 2005, 04:11:41 AM »
The wing of the Spitfire got step by step (From I to IX) thicker skin, just like the Bf 109, which increased the torsional stiffnes. I don't know the exact numbers and the thickness of the skin sheet varied from root to tip.

The final solution did not come until the Mk. 20 series which had completely redesigned wing. The VIII type airframe was something between original and Mk 20 series. The VIII had for example shorter span ailerons.

gripen

Offline Knegel

  • Nickel Member
  • ***
  • Posts: 620
DVL data on Bf 109 roll rates
« Reply #182 on: October 07, 2005, 04:22:05 AM »
Quote
Originally posted by gripen
The last sentence might be right or wrong depending the geometry of the linkage and where the loosenes is located. The pilots input might go to bend the linkage or cables.


No, the loosenes might cause changes in the leverages inside the linkage and even waste large amount of control force to linkage itself.
gripen


What you describe is again the physical law!

Quote
Originally posted by gripen

No, the RAE curves show slight bending before the force limit is reached and the report states clearly that the curves are based on flight tested data. There is not much bending below 250 mph EAS even in the case of the Spitfire I.


No, the steady rate of roll per degree aileron can be calculated from any measured rate of roll at any aileron deflection when the both are measured simultaneously as was done in the RAE tests.

 No, the calculation is rather easy; the max aileron deflection was +-17 deg and steady rate of roll per degree aileron about 9,5 deg/s at 255 mph EAS and that results about 160 deg/s (there was very little elasticity in the linkage).

gripen



I think now you messed up the reading of the graphic.

As far as i can see the graphic show the rollratio with constant alerondeflection. In the vertical i can see degree/sec (high -1),  so if there stand 10 it mean 100degree/sec, if there stand 15 its 150degree/sec.

The strait (linear) line is the rollratio without wingtwist in degree/sec , while the other line show the rollratio with wingtwist also in degree/sec. The constant round curve show that never the max stickdeflection got reached.

At 330mph the measured rollratio is roundabout 110°/sec, same like in the naca comparison, so here the stickforce is 50lb. Before this point the stickforce is smaler, behind this point the stickforce increase.

As you (maybe) can see, with increasing presure on the alerons the wingtwist increase.

So with max delection of alerons the wingtwist also should increse much with increasing speed, and this at much smaler speeds, cause the greater aleron variation cause more early a big wingtwist,  but this effect i simply cant see in the naca comparison.

The incredible roll increasement of 25°/sec due to 20lb more stickfore and the linear roll increasement of the Spitfire let me doubt that this test was realy measured! But maybe this Spit had a different wing than the other Spitfire, where the 30lb test was made with.
Maybe its the stiffned wing of the VIII model(?) like Angus mentioned. But i still doubt a that linear increasement at all.

Would be nice to see the correlating test.

Quote
Originally posted by gripen

BTW you are following Mr. Kurfürst's road now.


I would say this is the start of the polemical way. :(

I dont have a problem to offer my conclusion, although Kurfi got the similar conclusion.  If you think this is bad, you lose the path of a agrue based discussian, not me.



Greetings, Knegel

Offline gripen

  • Silver Member
  • ****
  • Posts: 1914
DVL data on Bf 109 roll rates
« Reply #183 on: October 07, 2005, 05:42:12 AM »
Quote
Originally posted by Knegel
What you describe is again the physical law!


I'm merely saying here that we don't know for sure what would be the effect of the loosenes in the linkage. The graph shows similar values at low deflections as other tests but at higher deflection the graph show different and unlogical values. The rest is speculation.

Quote
Originally posted by Knegel

As far as i can see the graphic show the rollratio with constant alerondeflection. In the vertical i can see degree/sec (high -1),  so if there stand 10 it mean 100degree/sec, if there stand 15 its 150degree/sec.


No, the graph (Fig. 3) [edited, I forgot the Fig number] gives the steady rate of roll per degree aileron and for example at about 255 mph EAS the plane did about 9,5 deg/s/degree aileron and with full 17 deg deflection that results about 160 deg/s.





Quote
Originally posted by Knegel
The strait (linear) line is the rollratio without wingtwist in degree/sec , while the other line show the rollratio with wingtwist also in degree/sec. The constant round curve show that never the max stickdeflection got reached.


No, the straight line is also rate of roll per degree aileron as stated in the right side of the graph ie deg/s/degree aileron.  The round curve shows (as stated in the right side of the graph) the measured rate of roll per degree aileron which can be calculated from any known   steady roll rate and aileron deflection combination. Note that the straight line is calculated from the round  line.

Quote
Originally posted by Knegel
I would say this is the start of the polemical way. :(

I dont have a problem to offer my conclusion, although Kurfi got the similar conclusion.  If you think this is bad, you lose the path of a agrue based discussian, not me.


Well, I have said several times in this thread that it's up to you to choose which data you want to believe. But all I see here is that you have problems to understand the Fig. 3 in the RAE 1231.

The report as well as the graph text state directly that the values are measured in the flight test and therefore the wing twist is included and report also states that the Spitfire, Mustang and Typhoon were measured similar way. Here you can choose Mr. Kurfürst's road or believe what the report and graph actually say.

gripen
« Last Edit: October 07, 2005, 06:36:30 AM by gripen »

Offline Angus

  • Plutonium Member
  • *******
  • Posts: 10057
DVL data on Bf 109 roll rates
« Reply #184 on: October 07, 2005, 06:54:32 AM »
Well, Gripen, you beat me to it.
From Knegel:
"As far as i can see the graphic show the rollratio with constant alerondeflection. In the vertical i can see degree/sec (high -1), so if there stand 10 it mean 100degree/sec, if there stand 15 its 150degree/sec."

It got me pondering, - you read the graph wrong, for there is hardly no rollrate of a WW2 fighter that insane!!!

And as I did point at before, - no matter how deep you get into stickforce math, deflections, stretchings and so on, it is the actual roll rate at the receiving end that counts.
It was very interesting to carry out the flight trials at Rechlin with the Spitfire and the Hurricane. Both types are very simple to fly compared to our aircraft, and childishly easy to take-off and land. (Werner Mölders)

Offline Knegel

  • Nickel Member
  • ***
  • Posts: 620
DVL data on Bf 109 roll rates
« Reply #185 on: October 07, 2005, 01:13:14 PM »
Hi,

Now its getting funny. :)

Quote
Originally posted by gripen
I'm merely saying here that we don't know for sure what would be the effect of the loosenes in the linkage. The graph shows similar values at low deflections as other tests but at higher deflection the graph show different and unlogical values. The rest is speculation.
gripen


You dont know for sure what is the reason for the different result, but you 1st was sure that the leverage is not the reason, but the linkage, then the geometry(leverage) did change cause a loose linkage.
There is no loose linkage, which have 15°/sec roll performence decreasement as result. A mechianic would call this damaged and i doubt that they made excessive test with a damaged plane.

If cou would know a bit about the physical law in general and special linkages, you would know that this different results only can happen due to a wanted different leverage.

I have enough education and experience in this field, its not a speculation!

Quote
Originally posted by gripen

No, the graph (Fig. 3) [edited, I forgot the Fig number] gives the steady rate of roll per degree aileron and for example at about 255 mph EAS the plane did about 9,5 deg/s/degree aileron and with full 17 deg deflection that results about 160 deg/s.

No, the straight line is also rate of roll per degree aileron as stated in the right side of the graph ie deg/s/degree aileron.  The round curve shows (as stated in the right side of the graph) the measured rate of roll per degree aileron which can be calculated from any known   steady roll rate and aileron deflection combination. Note that the straight line is calculated from the round  line.
gripen

 
Next to the strait line and the other line you clearly can read rate of roll, rate of roll get measured in degree/sec, not in degree/sec/degree.
Steady rate of roll per degree aleron, this mean a constant aleron deflection. They didnt wrote  "Steady rate of roll per x degree aleron", and of course this wouldnt make sence in relation to different speeds. We would have two variables, its pretty not usefull to display such things.
With a constant aleron deflection and a rigid wing the roll ratio increase linear with the speed. Thats what the strait line show.

They was up to show the influence of the wingtwist, therfor they did calculated strait line, with rigid wings. This they did before the flight or after the flight, but they didnt need the measured curve for this.

Quote
Originally posted by gripen

Well, I have said several times in this thread that it's up to you to choose which data you want to believe. But all I see here is that you have problems to understand the Fig. 3 in the RAE 1231.
gripen


I dont need to believe, i take the tests like they are and try to explain why they are different, thats all. That we shouldnt believe we can see in the mistake that the NACA testers made regarding the wingtwist.

Quote
Originally posted by gripen

The report as well as the graph text state directly that the values are measured in the flight test and therefore the wing twist is included and report also states that the Spitfire, Mustang and Typhoon were measured similar way. Here you can choose Mr. Kurfürst's road or believe what the report and graph actually say.

gripen


The other NACA test also state clearly that the wing dont twist!

I dont know if the tests are realy measured, but i would like to know why the different Spitcurves dont fit to each other.

I dont say the Spit didnt roll that fast, i only wanna know what they did change or if the testers made mistakes again.

Angus,
i only did try to explain that the leverage in the different Spits did change(similar to the P36 to P40). Its important to know that there was changuings to be able to say which spit did roll with what speed.


Greetings, Knegel

Offline gripen

  • Silver Member
  • ****
  • Posts: 1914
DVL data on Bf 109 roll rates
« Reply #186 on: October 07, 2005, 03:05:16 PM »
Quote
Originally posted by Knegel

You dont know for sure what is the reason for the different result, but you 1st was sure that the leverage is not the reason, but the linkage, then the geometry(leverage) did change cause a loose linkage.
There is no loose linkage, which have 15°/sec roll performence decreasement as result. A mechianic would call this damaged and i doubt that they made excessive test with a damaged plane.


We do know that the NACA tested Spitfire V had unlogical roll rate curve if compared to RAE and RAAF tested planes. We also know that the plane saw extensive flying in the US. It's my speculation that this is caused by loose linkage because at low speed and/or low aileron deflections reached values are normal. At high aileron deflection values differ considerably from the other tests and show characters which are not seen in other tests.

Quote
Originally posted by Knegel
I have enough education and experience in this field, its not a speculation!


I have no idea about your education or experience, but I can draw conclusions from your postings.

 
Quote
Originally posted by Knegel
Next to the strait line and the other line you clearly can read rate of roll, rate of roll get measured in degree/sec, not in degree/sec/degree.


No, let's have look:



First the text in the bottom: STEADY RATES OF ROLL PER UNIT AILERON ANGLE
Which means the steady rate of roll per degree aileron angle because the aileron angle is given in the degrees in the report. So it can be written:

(rate of roll)/(aileron angle)

Or with units:

(deg/s)/deg

Same is said with a bit other words in the left side of graph: STEADY RATE OF ROLL PER DEGREE AILERON (DEG.SEC^-1)

And same is said directly in the text: "The variation of steady rate of roll per unit aileron angle with speed is shown in Fig.3"

Lets take an example: What steady rate of roll at 350 mph EAS can be reached with max available aileron deflection with 50 lbs stick force assuming flexible wing?

First we must take a look to stick force graph:



We can read from the stick force graph that at 359 mph EAS max aileron deflection is about 8 degrees with 50 lbs and at 308 mph EAS about 12 degrees so we can estimate that at 350 mph EAS we can reach roughly 8,8 deg aileron deflection. So now we can simply read from the Fig. 3 what kind rate of steady rate of roll per unit aileron angle can be reached at 350 mph EAS and the graph gives about 11 (deg/s)/deg for flexible (ie measured) wing so the rate of roll would be:

8,8deg x 11(deg/s)/deg = 96,8 deg/s

Note that unneeded degs "eat each other" and disapear :)

If we assume rigid wing, the Fig. 3 gives about 15 (deg/s)/deg and that results:

8,8deg x 15(deg/s)/deg = 132 deg/s

Now we know that if the roll rate curve gives about 97 deg/s at 350 mph EAS it contains wing twist and if it gives about 132 deg/s, the wing is assumed to be rigid. So let's have look to Fig. 6:



The chart gives about 97 deg/s so it contains wing twist. Lets look the NACA chart:



It gives the same value ie 97 deg/s so it contains wing twist too. Lets read what RAE 1231 says about the test data:

"The best method of comparison of the rolling performance of different aircraft is based on the steady rate of roll a pilot can generate using a definite stick force, say 50 lb., or full aileron if this requires less than 50 lb. on the stick. This course has been adopted in Fig.6 which shows the results obtained for the F.W.190, Mustang, Typhoon, and Spitfire V (metal covered ailerons) with both standard and clipped wings. On all these aircraft instrumental records of rolling performance have been obtained at the R.A.E, similar to those under discussion for the Fw 190. In this connection it is worth noting that "instrumentation" is essential when obtaining the curves of Fig.6. Stop-watch measurements of time to bank on fighters are rarely of sufficient accuracy, since the times to be measured are so small."

So now we have 100% certainty that the RAE test data on the Fw 190, Spitfire, Mustang and Typhoon include wing twist. And the meaning of the grahs should be clear for all.

Quote
Originally posted by Knegel
The other NACA test also state clearly that the wing dont twist!


So it's says but RAE data proves that claim errorneous:



Note that again the steady rate of roll is given per degree aileron deflection just like in the wing twist graph of the Fw 190.

gripen
« Last Edit: October 07, 2005, 03:09:34 PM by gripen »

Offline niklas

  • Nickel Member
  • ***
  • Posts: 418
DVL data on Bf 109 roll rates
« Reply #187 on: October 07, 2005, 04:43:50 PM »
Well maybe it depends on the aileron angel for which the rollrate was measured. Assuming a linear dependence of aileron angle and rollrate is maybe not realistic.
Or the measured rollrate was too high for some guys in the RAF...

niklas

Offline Angus

  • Plutonium Member
  • *******
  • Posts: 10057
DVL data on Bf 109 roll rates
« Reply #188 on: October 07, 2005, 07:42:27 PM »
Ah yeasssh.
The 109 rolled so well it couldn't be measured.
The 190 rolled nicely also. Almost as well. :D
It was very interesting to carry out the flight trials at Rechlin with the Spitfire and the Hurricane. Both types are very simple to fly compared to our aircraft, and childishly easy to take-off and land. (Werner Mölders)

Offline Knegel

  • Nickel Member
  • ***
  • Posts: 620
DVL data on Bf 109 roll rates
« Reply #189 on: October 08, 2005, 01:18:53 AM »
Hi,

with increasing aleron variation and increasing speed, at same time, the steady rate of roll(rigid wing) dont would increase linear!!!

The curve would go upward, cause the force due to increasing speed and increasing alerondeflection together would let increase the rollratio
progressively !!

Yes its the steady rate of roll per 1 unit aleron delfection, not per x unit alerondelfection!!

How do you think they will measure a linear increasement of alerondelfection and speed at same time?? Thats simply not possible!

You will realy tell me that they had exact 10° alerondeflection at 230mph and 13° at 300mph??

Thats rubbish!  And dont make sence!! We would have the speed, the alerondeflection and the rollratio as variable, thats simply not usefull!

To show the influence of the wingtwist they need a constant alerondeflection with increasing speed. Otherwise we cant read anything!

With a constant aleron deflection (one unit, the exact value isnt important)  the physical law say that the rollratio increase linear with the speed(up to roundabout mach 0,75).
With different aleron deflections we dont know the exact curve, cause the effectiveness of the alerons with different alreon variations depends much to the aerodynamic of the wing and alerons.
Therfor it dont make sence to make wingtwist tests with a variable alerondeflection, specialy not, if the speed change at same time.

If you dont understand that, we can stop here!

Greetings, Knegel
« Last Edit: October 08, 2005, 01:22:07 AM by Knegel »

Offline Knegel

  • Nickel Member
  • ***
  • Posts: 620
DVL data on Bf 109 roll rates
« Reply #190 on: October 08, 2005, 01:39:47 AM »
btw, the DLV made the same tests with the 109F2, this is what they write:

"Um den Einfluß der elastischen Flügelverdrehung auf die Drehgeschwindigkeit zu untersuchen, ist es jedoch erforderlich, die Abhängingkeit der Rollgeschwindigkeit von der Fluggeschwindigkeit für konstante Querruderausschläge zu kennen."

Translated:
"To investigate the influence of the elastic wingtwist to the roll speed, its needed to know the dependence of the roll speed to the flight speed for constant aleron deflections."

And if you look here, to the 3° or 6° aleron deflection line, it looks almost the same like the FW190 and Spit line with constant alerondeflection.




Greetings, Knegel

Offline niklas

  • Nickel Member
  • ***
  • Posts: 418
DVL data on Bf 109 roll rates
« Reply #191 on: October 08, 2005, 03:28:30 AM »
Well, the 109F chart says the peak rollrate is at higher speed for lower deflection. This shouldn´t be surprising. The difference looks minor on the other hand, less than 100km/h.
So a single chart for all aileron deflection is at least for higher speeds (non-linear characteristics) a simplification.

Furthermore aileron angle is a bad base for such a chart because the aileron angle get lower due to elastic deformation at higher speeds, see that F2 chart. I think it would have been better to base it on stick angle. Or does the mean aileron angle is the theoretical angle without elastic deformation? Is it a corrected one for this influence?

niklas

Offline gripen

  • Silver Member
  • ****
  • Posts: 1914
DVL data on Bf 109 roll rates
« Reply #192 on: October 08, 2005, 04:28:15 AM »
Quote
Originally posted by Niklas
Well maybe it depends on the aileron angel for which the rollrate was measured. Assuming a linear dependence of aileron angle and rollrate is maybe not realistic.
Or the measured rollrate was too high for some guys in the RAF...


The data set was measured with quite wide variety of aileron angles (roughly from 2,5 to 14 deg) and generally the RAE report praises the ailerons of the Fw 190 as well as pretty much all other British sources. In addition report notes that the tested plane had rather heavier ailerons than other two they had.

Quote
Originally posted by Knegel

with increasing aleron variation and increasing speed, at same time, the steady rate of roll(rigid wing) dont would increase linear!!!


I must honestly say that I have no anykind of interest to know if the curve for the rigid wing is right or wrong, I'm merely interested about the measured roll rates at measured aileron deflections. If you are not happy with curve for the rigid wing, you have got the data and the field is open for your analysis. But maybe you should open a new thread for that because this very little to do with the subject of this thread.

Quote
Originally posted by Knegel
Yes its the steady rate of roll per 1 unit aleron delfection, not per x unit alerondelfection!!


No one has claimed otherwise here except that you claimed above that:

"As far as i can see the graphic show the rollratio with constant alerondeflection. In the vertical i can see degree/sec (high -1), so if there stand 10 it mean 100degree/sec, if there stand 15 its 150degree/sec."

Is there a remote possibility that you could admit that you did not understand the Fig. 3 until I explained it to you?

Quote
Originally posted by Knegel
How do you think they will measure a linear increasement of alerondelfection and speed at same time??


What they did was simply that they flew at given speed (205, 257, 308, 359 and 409 mph EAS), rolled the plane and recorded the aileron deflection, steady rate of roll and stick force (and probably many other things too) and repeated that several times at all tested speeds. The steady rate of roll per unit (degree) aileron can be easily calculated from any known aileron deflection and steady roll rate combination simply:

steady roll rate per unit = (measured steady roll rate)/(measured aileron deflection)

And to keep units with:

(deg/s)/deg = (deg/s)/deg

As an example, if measured roll rate at given speed is 160 deg/s and measured aileron deflection was 17 deg, the steady rate of roll per unit (degree) aileron is simply:

(160deg/s)/(17deg) = 9,41 (deg/s)/deg

Quote
Originally posted by Knegel
To show the influence of the wingtwist they need a constant alerondeflection with increasing speed. Otherwise we cant read anything!


No, they just needed to collect enough test points and fit the curve. There is some thirty measured data points in the set.



Quote
Originally posted by Knegel
If you dont understand that, we can stop here!


Well, the understanding problem is in your side.

gripen

Offline Knegel

  • Nickel Member
  • ***
  • Posts: 620
DVL data on Bf 109 roll rates
« Reply #193 on: October 08, 2005, 09:49:49 AM »
Hi,

you be right, it did need a bit, but now i got it! :)

This graph show the "roll effectiveness"(rate of roll per aleron delfection) of the wing.

The entrys 'measured rate of roll', 'rate of roll with rigid wings' and 'loss of rate of roll due to wing torsion' drove me to the wrong conclusion!  
This describtions are simply wrong. Thes curves display the 'wing torsion factor' related to the 'measured rate of roll' and 'wing torsion factor' of a rigid wing, the different of both display the 'loss of roll effectivity' due to wing torsion(decreased drill rigidity).

All three wrong describtions whould have been measured in degree/sec, thats why i did stumble.

Out of this factors, as you wrote, we can calculate the rate of roll, with different aleron deflections.

So the FW190A curve is the measured one.


Greetings, Knegel

Offline Knegel

  • Nickel Member
  • ***
  • Posts: 620
DVL data on Bf 109 roll rates
« Reply #194 on: October 10, 2005, 01:24:41 AM »
Btw.

what i find to be very strange is that they assume a absolute rigid linkage.
Up tp 60lb stickforce the alreron deflection increase absolute linear with the stickforce(double stickforce = double alerondeflection).

The 109F did show a smooh decreasing of the alerondeflection with the speed. This had a roll ratio lost of 15°/sec as result on the huighest peak (roundabout max 80 °/sec to 95°/sec).

If the measured of course include the softness in the linkage, so the Different to the rigid wing curve include the softness of the linkage too. But its wrong to calculate from this curve to a curve at higher speed or higher alerondeflection, cause the the influence of the softness of the linkage differ with the speed and/or alerondeflection.

Since it looks like they did calculate the 50lb curve out of the tested curves, its possible that they dont include the softness of the linkage at given speeds with max aleron deflection, cause they assume a absolute rigid linkage! (if there was a softness in the linkage at all).

Dont we have a german FW190 roll test, similar to the DLV one? They did split all this influences into different tests.

Greetings, Knegel