Advanced Problems of

Longitudinal Dynamics!

Robert Stengel, Aircraft Flight Dynamics

MAE 331, 2014"
! Angle-of-attack-rate aero effects"
! Fourth-order dynamics"
!
!
!
!

Steady-state response to control"

Transfer functions"
Frequency response"
Root locus analysis of parameter
variations"

! Nichols chart"
! Pilot-aircraft interactions"

Flight Dynamics!
204-206, 503-525!
Airplane Stability and Control!
Chapter 10!

Copyright 2014 by Robert Stengel. All rights reserved. For educational use only.!
http://www.princeton.edu/~stengel/MAE331.html!
http://www.princeton.edu/~stengel/FlightDynamics.html!

1!

Primary and Coupling Blocks of the

Fourth-Order Longitudinal Model"
# !D
V
%
% LV
VN
%
FLon = %
% MV
% L
% ! VV
N
$

!g

Mq

!D"

L"

VN

M"
! L"

VN

&
(
(
( # FPh
( = % SP
( %$ FPh
(
(
'

Ph &
FSP
(
FSP ('

! Some stability derivatives appear only in primary

blocks (DV, Mq, M!)"
! Effects are well-described by 2nd-order models"

! Some stability derivatives appear only in coupling

blocks (MV, D!)"
! Effects are ignored by 2nd-order models"

! Some stability derivatives appear in both (LV, L!)"

! May require 4th-order modeling"

2!

How do the 4th-order roots vary when

we change pitch-rate damping, Mq?"
Identify Mq terms in the characteristic polynomial"

( )
)

L"
LV
$
' 2
3
s
+
g
#
D
+
D
#
M
(
)
"
V
"
&%
)( s
VN
VN
VN
L
L
+ { D" M V # DV M " } s + g M V " V # M " V V
N
N

! Lon (s) = s 4 + DV +

( )

L"

#M q s 3 # $ DV M q + M q
%&

L"

L" '
LV
' 2
$
VN () s + M q %&( D" # g ) VN # DV VN () s
=0

3!

How do the 4th-order roots vary when

we change pitch-rate damping, Mq?"
Group Mq terms in the characteristic polynomial"

! Lon (s) = d ( s ) " M q s 3 + $ DV +

&%

L#

L# '
LV
' 2 $
VN )( s " &%( D# " g ) VN " DV VN )( s

L
L
L
= d ( s ) " M q s s 2 + $ DV + # V ' s " $( D# " g ) V V " DV # V '
&%
N)
N
N)
( &%
(
= d ( s ) + kn ( s ) = 0

n(s)
= !1
d(s)
4!

How do the 4th-order roots vary when

we change pitch-rate damping, Mq?"
! Factor terms that are multiplied by Mq to nd the 3 zeros"
! 2 zeros near origin similar to approximate phugoid roots,
effectively canceling Mq effect on them "

{ (

L" &
LV
#
VN s ! %$( D" ! g ) VN ! DV VN ('
!M q
= !1
) 4
L"
L"
LV
#
& 23
++ s + DV + VN s + %$( g ! D" ) VN + DV
VN ! M " (' s ++
*
.
L"
LV
+
+
+ { D" M V ! DV M " } s + g M V V ! M " V
+,
+/
N
N
s s 2 + DV +

!M q

L"

( )

s ( s ! z1 ) ( s ! z2 )

)(

s 2 + 2" P# nP s + # n2P s 2 + 2" SP# nSP s + # n2SP

s ( s ! z1 ) = ( s 2 ! z1s + 0 ) " s 2 + 2# P$ nP s + $ n2P

= !1

5!

Mq variation has virtually no effect on phugoid roots"

Effect on short-period roots is predicted by 2nd-order model"
!M q

(s

(s

! z1s ) ( s ! z2 )

+ 2" P# nP s + # n2P ) ( s 2 + 2" SP# nSP s + # n2SP )

(s

!M q ( s ! z2 )

+ 2" SP# nSP s + # n2SP )

6!

= !1

M! Effect on 4th-Order Roots!

Group all terms multiplied by M! to form numerator for M!"

! Lon (s) = s 4 + DV + L"

VN

# M q s3

$
'
+ &( g # D" ) LV + DV L" # M q # M q L" # M "o ) s 2
VN
VN
VN
%
(

Short
Period!

$
'
+ M q &( D" # g ) LV # DV L" ) + D" M V # DV M "o s
VN
VN (
%

+ g M V L"

VN

* d(s)+ kn(s)

# M "o LV

VN

) # !M (s + D s + g L V )
"

Phugoid!
Phugoid!

Short
Period!

7!

M! Effect on 4th-Order Roots!

! Primary effect: Same as the approximate short-period model"
! Numerator zeros"
! Same as the approximate phugoid mode characteristic
polynomial"
! Effect of M! variation on phugoid mode is small"

Short
Period!
Phugoid!

Phugoid!
Short
Period!

8!

L! /VN and LV /VN Effects on

Fourth-Order Roots!
! L! /VN: Increased damping
of the short-period"
! Small effect on the phugoid
mode"

! LV /VN: Damped natural

frequency of the phugoid"
! Negligible effect on the shortperiod"

Short
Period!

Phugoid!

9!

Longitudinal Model Transfer

Function Matrix (Hx = I, Hu = 0)"
$ nV (s) nV (s) nV (s) '
!T
!F
)
& !E
"
"
"
& n! E (s) n! T (s) n! F (s) )
)
& q
q
q
& n! E (s) n! T (s) n! F (s) )
)
& #
#
#
&% n! E (s) n! T (s) n! F (s) )(
H Lon (s) = 2
s + 2* P+ nP s + + n2P s 2 + 2* SP+ nSP s + + n2SP

$
&
&
&
&
&%

)(

!V (s) '
$ !* E(s) '
$ !* E(s) '
)
&
)
&
)
!" (s) )
=
H
=
A
s
G
H
(s)
*
T
(s)
*
T
(s)
!
!
(
)
&
)
&
)
x
Lon
!q(s) )
&
)
&
)
!* F(s) (
!* F(s) )(
%
%
!# (s) )(

10!

Elevator-to-Normal-Velocity
Transfer Function"
M " E s 2 + 2#$ n s + $ n2 Approx Ph ( s % z3 )
!w(s)
n"wE (s)
=
= 2
2
2
!" E(s) ! Lon (s) s + 2#$ n s + $ n2
s
+
2
#$
s
+
$
n
n
Ph
SP

) (

!Normal velocity transfer function is analogous

to angle of attack transfer function (!! # !w/VN)"
!z3 often neglected due to high frequency !
11!

Transfer Functions of Elevator Input

to Angle Output*"
! Elevator-to-Flight Path Angle transfer function"

(
n#" E (s)
!" (s)
L %
=
; n#" E (s) = M # E $ ' s + 1 *
T"1 )
!# E(s) ! Lon (s)
VN &
! Elevator-to-Angle of Attack transfer function"

!" (s) n#"E (s)

=
; n#"E (s) = M # E ( s 2 + 2$% n s + % n2 ) Approx Ph
!# E(s) ! Lon (s)
! Elevator-to-Pitch Angle transfer function"

$
'$
'
!" (s) n#"E (s)
=
; n#"E (s) = M # E & s + 1 )& s + 1 )
T"1 (%
T"2 (
%
!# E(s) ! Lon (s)
* Flying qualities notation for zero time constants"

12!

Feedback Control: Angles to Elevator"

!Variations in
control gain"
!Principal effect is
on short-period
roots"

13!

Frequency Response of
Angles to Elevator Input"
! Pitch angle frequency
response (!# = !$ + !!)"
! Similar to ight path
angle near phugoid
natural frequency"
! Similar to angle of
attack near shortperiod natural
frequency"

!" SS = c!# ESS

!$ SS = f !# ESS

!% SS = ( c & f ) !# ESS

14!

Elevator-to-PitchRate Frequency
Response"
! (n q) = 1"
! Negligible lowfrequency response,
except at phugoid
natural frequency"
! High-frequency
response well
predicted by 2ndorder model"

M " E s ( s $ z1 ) ( s $ z2 )
!q(s)
n q (s)
= "E
#
!" E(s) ! Lon (s) ( s 2 + 2%& n s + & n 2 ) ( s 2 + 2%& n s + & n 2 )
!

)(

Ph

M " E s s +1 T'1 s +1 T'2

( s 2 + 2%& n s + & n 2 )Ph ( s 2 + 2%& n s + & n 2 )SP

+20 dB/dec!
0 dB/dec!

+20 dB/dec!

+40 dB/dec!

20 dB/dec!

15!

Gain and Phase Margins:!

The Nichols Chart!

16!

SP

Nichols Chart: !
Gain vs. Phase Angle"
!

Bode Plot"

! Two plots"
! Open-Loop Gain (dB) vs. log10%"
! Open-Loop Phase Angle vs. log10%"

Nichols Chart"

! Single crossplot; input

frequency not shown"
! Open-Loop Gain (dB) vs. OpenLoop Phase Angle"

17!

Gain and Phase Margins"

! Gain Margin "

! Measured at the input frequency, %, for which &(j%) = 180"

! Difference between 0 dB and transfer function magnitude,
20 log10 AR(j%)"

! Phase Margin "

! Measured at the input frequency, %, for which 20 log10

AR(j%) = 0 dB "
! Difference between the phase angle &(j%), and 180"

! Axis intercepts on the Nichols Chart identify GM

and PM"
18!

Examples of Gain and Phase Margins:!

2nd-Order System with Low-Pass Filter"
Low-Bandwidth Filter"

%
" 10
%"
100 2
H blue ( j! ) = $
'$
2
2 '
j
!
+
10
) & # ( j! ) + 2 ( 0.1)(100 )( j! ) + 100 &
#(

High-Bandwidth Filter"

"
% " 100
%
10 2
H green ( j! ) = $
2
2 '$
j
!
+
100
) '&
# ( j! ) + 2 ( 0.1) (10 ) ( j! ) + 10 & # (

Nichols Chart"

Bode Plot"

19!

Gain and Phase Margins

in Pitch-Tracking Task"
Elevator-to-Pitch-Angle
Bode Plot"

Elevator-to-Pitch-Angle
Nichols Chart"

!
!

Gain Margin: Amplitude ratio below 0 dB

when phase angle = 180"
Phase Margin: Phase angle above 180
when amplitude ratio = 0 dB"

Amplitude Ratio vs. Phase Angle"

20!

Pilot-Vehicle Interactions!

21!

Pilot Inputs to Control"

* p. 421-425, Flight Dynamics"

22!

Effect of Pilot Dynamics on

Pitch-Angle Control Task"

Open-Loop 1st-Order-Pilot/Aircraft Transfer Function"

$
!
)
,)
,
M ' E + s + 1T . + s + 1T .
&
#
! 1 / TP $
*
(1 - *
(2 H (s) = K P #
&
&# 2
2
2
2
" ( s + 1 / TP ) % # ( s + 2/0 n s + 0 n )Ph ( s + 2/0 n s + 0 n )SP &
&%
#"
23!

Effect of Pilot Dynamics on

Pitch-Angle Control Task"

! Pilot introduces neuromuscular lag while closing the control loop"

! Example"
! Model the lag by a 1st-order time constant, TP, of 0.25 s"
! Pilot


s gain, KP, is either 1 or 2"

Pilot Transfer Function =

!u ( s )
1 / TP
1 / 0.25
= KP
= KP
s +1 / 0.25
!" ( s )
s +1 / TP

24!

Effect of Pilot
Dynamics on
Pitch-Angle
Control Task"
! Gain and phase
margins become
negative for pilot
gain between 1
and 2"
! Then, pilot
destabilizes the
system (PIO)"
25!

Effect of 1st-Order Pilot Dynamics on

Elevator/Pitch-Angle Control Root Locus"

Pilot transfer function changes asymptotes of the root locus"

26!

Pilot-Induced Oscillations"

Uncommanded aircraft is stable but piloting

actions couple with aircraft dynamics to
produce instability"
F-22!

27!

Pilot-Induced Oscillations"
NASA Digital-Fly-By-Wire F-8
Simulation of Space Shuttle!

http://www.dfrc.nasa.gov/Gallery/Movie/F-8DFBW/Medium/EM-0044-01.mov!
http://www.dfrc.nasa.gov/Gallery/Movie/F-8DFBW/Medium/EM-0044-02.mov!

28!

Power Effects on 4th-Order

Longitudinal Modes!

29!

Power Effects on Stability and Control"

!
!

!
!

Gee Bee R1 Racer: an engine with

wings and almost no tail"
During W.W.II, the size of ghters
remained about the same, but
installed horsepower doubled (F4F
vs. F8F)"
Use of aps means high power at
low speed, increasing relative
signicance of thrust effects"
Short-Takeoff-and-Landing (STOL)
aircraft augment takeoff/landing lift
in many ways, e.g.,"
! Full-span aps"
! Deected thrust"

GB R1!

Grumman F4F!

Grumman F8F!

30!

Direct Thrust Effect on Speed

Stability, TV"

In steady, level ight, nominal thrust balances nominal drag"

1 2
1
"VN S ! C DN "VN2 S = 0
2
2
! Effect of velocity change"
TN ! DN = CTN

# < 0, for propeller aircraft

!T %
= $ " 0, for turbojet aircraft
!V %
> 0, for ramjet aircraft
&

! Small velocity perturbation grows if "

! Therefore"
! propeller is stabilizing for velocity change"
! turbojet has neutral effect"
! ramjet is destabilizing"

!T ! D
"
>0
!V !V
31!

Pitching Moment Due to Thrust, MV!

McDonnell Douglas MD-11!

Lake Amphibian!
Grumman F8F!

!
!
!
!

Douglas AD-1!

Negative "M/"V (Pitch-down effect) tends to increase velocity"

Positive "M/"V (Pitch-up effect) tends to decrease velocity"
With propeller thrust line above the c.m., increased velocity
decreases thrust, producing a pitch-up moment"
Tilting the thrust line can have benets"
! Up: Lake Amphibian, MD-11"
! Down: F6F, F8F, AD-1 "

32!

MV Effect on 4th-Order
Roots"
! Lon (s) = s 4 + DV + L"

VN

# M q s3

$
'
+ &( g # D" ) LV + DV L" # M q # M q L" # M " ) s 2
VN
VN
VN
%
(

$
'
+ M q &( D" # g ) LV # DV L" ) + D" M V # DV M " s
VN
VN (
%

gM " LV

VN

+ MV

D" s + g L"

VN

Short
Period!

)=0

Phugoid!

! Large positive value produces

oscillatory phugoid instability"
! Large negative value produces
real phugoid divergence"

D" = 0

33!

Steady-State Response of the 4th-Order

LTI Longitudinal Model"
!!x(t) = F!x(t) + G!u(t)
! How do we calculate the equilibrium response to control?"

!x SS = "F "1 G !u SS
! For the longitudinal model"
$
&
&
&
&
&
%

!VSS
!" SS
!qSS
!# SS

$ *D
V
' &
L
&
)
V
VN
) &
) = *& M
V
) &
&
)
( & * LV
VN
%

*g

Mq

*D#

L#

VN

M#
* L#

VN

*1

'
)
)
)
)
)
)
)
(

$ 0
&
& 0
&
& M +E
& 0
%

T+T

L+ F / VN

*L+ F / VN

'
)$ !+ ESS
)&
)& !+TSS
)& !+ F
SS
)%
(

'
)
)
)
(
34!

Algebraic Equation for

Equilibrium Response"
$
$
'
*gM + E L# )
0
&
[ gM # L+F / VN ]
VN (
%&
&
& $
' $
'
L
L
L
L
& &% DV # V * D# V V M + E )( &% M V # V * M # V V T+T )( $%( D# M V * DV M # ) L+ F / VN '(
N
N
N
N
&
&
0
0
0
' &
$
'
L
V
) &
0
[ L+F / VN ]
&%*gM + E VN )(
) %
)=
g M V L# * M # LV
)
VN
VN
)
(

$
&
&
&
&
&
%

!VSS
!" SS
!qSS
!# SS

! Roles of stability and control

derivatives identied"
! Result is a simple equation
relating input and output"

$
&
&
&
&
&
%

!VSS ' $ a
) &
!" SS ) & c
)=
!qSS ) & 0
&
!# SS )( % f

'
)
)
)
)
)
)
)
) $ !+ ESS
(&
& !+TSS
&
% !+ FSS

0 b '$ !* E
)
SS
d e )&
& !*TSS
0 0 )&
0 g )(% !* FSS

'
)
)
)
(

35!

4th-Order Steady-State Response May

Be Counterintuitive"
!VSS = a!" ESS + ( 0 ) !" TSS + b!" FSS
!# SS = c!" ESS + d!" TSS + e!" FSS

!qSS = ( 0 ) !ESS + ( 0 ) !" TSS + ( 0 ) !" FSS

!$ SS = f !" ESS + ( 0 ) !" TSS + g!" FSS
! Observations"
!
!
!
!

Thrust command"
Elevator and ap commands"
Steady-state pitch rate is zero!
4th-order model neglects air density gradient effects!

Steady-state pitch angle"

!" SS = !# SS + !$ SS = ( c + f ) !% ESS + d!% TSS + ( e + g ) !% FSS

36!

'
)
)
)
(

Ercoupe Approach to Safe, Affordable Flying"

! Limited control authority; 2-control cockpits (rudder
interconnected to aileron)"
! Limited center-of-mass travel"
! Limited speed range"
! Wing leveling and lateral stability (stable spiral mode)"
! Fixed, tricycle landing gear"

37!

Less-Safe Airplanes"

However -- many designers of personal aircraft had other ideas"

! Mignet Flying Flea (Homebuilt, pivoting main wing, no ailerons, unrecoverable

dive)"
! V-tail Beechcraft Bonanza Model 35 (10,000 built, 250 in-ight structural failures) "
! American Yankee AA-1 (BD-1, hot
, stalls and spins) "
! Bede BD-5 (Home-built, unforgiving ying qualities)"

38!

!
!
!

Do Safe Airplanes Have

Fewer Accidents?"

The Ercoupe
s safety record is about

average"
Many J-3 Cubs involved in fatal stall/spin
accidents"
Cirrus SR-20/22 has had 26 fatal
crashes, killing 52 to date (1999-2006)"

Probable cause of Ercoupe fatal

accidents: 1994-2004"
!
!
!
!
!
!

VFR ight in IFR conditions"

Pilot


s lack of night-ying experience"
Pilot had no ight training"
Inadvertent stall"
Loss of engine power"
In-ight breakup (corrosion)"

Top causes of general aviation accidents "

!
!
!
!
!
!

Maneuvering"
Weather"
Takeoff/Climb"
Descent/Approach"
Engine/Prop"
Fuel Management"

In comparison, there were 69

fatalities during 36 years of
ying the F-14 ghter"

39!

2004 General Aviation Accidents !

(AOPA)"

Nall Report, 2012-13, http://www.aopa.org/-/media/Files/AOPA/

Home/Pilot%20Resources/Safety%20&%20Prociency/Accident
%20Analysis/Nall%20Report/Scorecard_1213.pdf"
40!

How Safe is Safe?!

(NTSB)"
Accidents, Fatalities, and Rates, 1986 through 2005,
for U.S. Air Carriers Operating Under 14 CFR 121

Accidents, Fatalities, and Rates, 1986 through 2005,

U.S. General Aviation
Accidents

Fatalities

Year
All
Fatal Total Aboard Flight Hours
1986 2,581
474
967
879
27,073,000
1987 2,495
446
837
822
26,972,000
1988 2,388
460
797
792
27,446,000
1989 2,242
432
769
766
27,920,000
1990 2,242
444
770
765
28,510,000
1991 2,197
439
800
786
27,678,000
1992 2,111
451
867
865
24,780,000
1993 2,064
401
744
740
22,796,000
1994 2,021
404
730
723
22,235,000
1995 2,056
413
735
728
24,906,000
1996 1,908
361
636
619
24,881,000
1997 1,844
350
631
625
25,591,000
1998 1,905
365
625
619
25,518,000
1999 1,905
340
619
615
29,246,000
2000 1,837
345
596
585
27,838,000
2001 1,727
325
562
558
25,431,000
2002 1,715
345
581
575
25,545,000
2003 1,739
352
632
629
25,706,000
2004 1,617
314
558
558
24,888,000
2005 1,669
321
562
557
24,401,000

Accidents
per 100,000
Flight Hours
All
Fatal
9.49
1.73
9.18
1.63
8.65
1.66
7.97
1.52
7.85
1.55
7.91
1.57
8.51
1.82
9.03
1.74
9.08
1.81
8.21
1.63
7.65
1.45
7.19
1.36
7.44
1.41
6.5
1.16
6.57
1.21
6.78
1.27
6.69
1.33
6.75
1.36
6.49
1.26
6.83
1.31

Accidents

Fatalities

Year All
Fatal Total
Aboard Flight Hours
1986
21
2
5
4
9,495,158
1987
32
4
231
229
10,115,407
1988
29
3
285
274
10,521,052
1989
23
8
131
130
10,597,922
1990
21
6
39
12
11,524,726
1991
22
4
62
49
11,139,166
1992
16
4
33
31
11,732,026
1993
22
1
1
0
11,981,347
1994
18
4
239
237
12,292,356
1995
30
1
160
160
12,776,679
1996
31
3
342
342
12,971,676
1997
43
3
3
2
15,061,662
1998
41
1
1
0
15,921,447
1999
40
2
12
11
16,693,365
2000
49
2
89
89
17,478,519
2001
41
6
531
525
17,157,858
2002
35
0
0
0
16,718,781
2003
51
2
22
21
16,887,756
2004
23
1
13
13
18,184,016
2005
32
3
22
20
18,728,000

Accidents
per 100,000
Flight Hours
All
Fatal
0.211 0.011
0.306
0.03
0.266 0.019
0.217 0.075
0.182 0.052
0.198 0.036
0.136 0.034
0.184 0.008
0.138 0.033
0.235 0.008
0.239 0.023
0.285
0.02
0.258 0.006
0.24 0.012
0.28 0.011
0.216 0.012
0.209
0.302 0.012
0.126 0.005
0.171 0.016

41!

Next Time:!
Flying Qualities Criteria!
Flight Dynamics!
419-428, 525-533!
Airplane Stability and Control!
Chapter 21!

42!

Supplemental
Material

43!

Apparent Mass and

Unsteady Aerodynamics!

44!

Distinction Between Angle-ofAttack Rate and Pitch Rate"

!! With no vertical motion of the
c.m., pitch rate and angle-ofattack rate are the same "

!! = q

!! With no pitching, vertical heaving

(or plunging) motion of the c.m.,
produces angle-of-attack rate but
no pitch rate"

Vertical velocity
distribution induced
by pitch rate!

!! " 0 " q; q = 0

45!

Angle-of-Attack Rate
Has Two Effects "
!! Pressure variations at wing
convect downstream,
arriving at tail !t sec later"
!! Lag of the downwash"
!! Delayed tail-lift/pitchmoment effect"

!! Vertical force opposed by a

mass of air (apparent
mass) as well as airplane
mass"
!! Vertical acceleration
produces added lift and
moment"

Flight Dynamics, pp. 204-206, 284-285"

46!

Angle-of-Attack-Rate Effects Principally

Affect the Short-Period Mode"
Lift and pitching moment proportional to angle-of-attack rate"

!q! = M q !q + M " !" + M # E !# E + M "! !"!

% L
(
!"! = '1$ q * !q $ L"
!" $ L# E
!# E $ L"!
!"!
V
V
V
VN
&
N)
N
N

( ) (

( )

Bring effects to left side"

!q! " M #! !#! = M q !q + M # !# + M $ E !$ E

( V ) !#! = %'&1" L V (*)!q " ( L V ) !# " ( L V ) !$E

!#! + L#!

$E

Vector-matrix form"
#
&
# 1
M"
Mq
!M "!
(# )q! & %
%
( = % * Lq #
& (%
%
L"!
L"
% 0 %$1+
VN (' ($% )"! (' % ,+1! VN /. !
VN
'
$
$

( )

( )

&
#
M 0E
(# )q & %
(+ %
(%
L0 E
(%$ )" (' % !
VN
$
'

&
(
( )0 E
(' 47!

Angle-of-Attack-Rate Effects"
Pre-multiply both sides by inverse"
)1
& 1#
# 1
Mq
M"
)M
!
"
# !q! & %
( 33%
%
(=%
#
& ( 2% * L
L
L
%$ !"! (' % 0 %1+ "! V ( ( 3% ,1) q V / ) " V
$
'
N
+
.
N
N
' 43$
$

( )

( )

&
#
M 0E
(# !q & %
(+ %
(%
L0 E
(%$ !" (' % )
VN
$
'

5
&
33
(
( !0 E 6
3
('
37

Inverse of the apparent mass matrix"

#
%
!1
%
&
# 1
!M "!
%
(
%
$
L"!
%
#
& ( =
% 0 %$1 +
VN (' (
'
$

( )

( )

L"!
#
&
1
+
%$
VN (' M "!
0
1

( )

L"!
#
&
1
+
%$
VN ('

&
(
(
(
'
48!

Angle-of-Attack-Rate Effects"
Substitute"

#
%
%$

#
%
%
!q! & %$
(=
!"! ('

&
#
&
L"!
M "! (
%$1+
(
VN '
( 0#
Mq
M"
0
1 (' 22%
1% * Lq L
#
&
L"!
2% ,1) V / ) " V
+
.
%$1+
N
N
VN ('
23$

( )

4
&
22
(# !q &
( +"5
(%
2
(%$ !" ('
26
'

( )

Multiply matrices"
0#
7%
7%
7%
7%
# !q! &
7%$
1
%
(=
1
&7
%$ !"! (' #1+ L"!
%$
VN (' 7
7
7
7
2

( )

0#
3 &
* L
-3 0#
&
&
L
L
L
1%1+ "! V ( M q + M "! ,1) q V /4 1%1+ "! V ( M " ) M "! " V 4 (#
+
N '
N .5
N '
N 5 (
2$
2$
!q
(%
%
* Lq !
($ "
) L"
,1) V /
V
('
+
N.
N

( )

#
%
+%
%
%
$

( )

#
&
L"!
M ) M "! L6 E
%$1+
VN (' 6 E
VN

( )

) L6 E

( )

&

) ((!6E

VN

(
(
'

49!

Simplication of Angle-of-AttackRate Effects"

Lq and L!! have small effects for large aircraft*

!! Typically"

M q and M !! are same order of magnitude

and have more significant effects

* but not for small aircraft, e.g.,
R/C models and micro-UAVs"

!! Neglecting"
#
# !q! & %
("%
%
!
"
!
(' %
$%
%
%$

{M

Lq and L!!

+ M "!
1

{ ( )}
( )
M " ) M "!
)

L"

L"

VN

VN

&
#
L
M * E ) M "! * E V
(#
& %
!q
N
(%
(+%
( % !" ( %
L
' %
) *E V
($
N
('
$

&
(
( !* E
(
(
'
50!

3
7
&7
(7
('7
7
4
7
7
7
7
7
5

2nd-Degree Characteristic Polynomial with"

Lq and L!! " 0

Short-period characteristic polynomial"

! ( s) =

&
$s " ( M q + M #! )& "$ M # " M #! L#
%
' (%
VN )'

( )
$
&
L
(%s + ( V ))'

"1

$
& $
&
= $%s " ( M q + M #! )&'(s + L#
" M " M #! L#
VN ') (% #
VN ')
%

( )

( )

#
& )#
= s 2 + % L!
" M + M !! )( s + *% M ! " ( M q + M !! ) L!
VN ( q
VN
$
' +$

( )&('+ M ( L V ),-.

( )

!!

Damping is increased"
Natural frequency is unaffected"

$
' *$
'# ( M q + M "! )) s + +& M " # M q L"
! ( s ) = s 2 + & L"
VN
VN )(./
%
( ,%

( )

( )

= s 2 + 201 n s + 1 n2 = 0

51!

Comparison of Bizjet Fourth- and

Second-Order Models and Eigenvalues"
Fourth-Order Model
F=
-0.0185
0.0019
0
-0.0019

-9.8067
0
0
0
0 1.2709
0 -1.2794 -7.9856
0
1 -1.2709

Phugoid Approximation
F=
-0.0185
0.0019

-9.8067
0

Short-Period Approximation
F=
-1.2794
1

-7.9856
-1.2709

G=
0
0
-9.069
0
G=
4.6645
0
G=
-9.069
0

Eigenvalue
4.6645
0
0
0

-8.43e-03 + 1.24e-01j
-8.43e-03 - 1.24e-01j
-1.28e+00 + 2.83e+00j
-1.28e+00 - 2.83e+00j
Eigenvalue
-9.25e-03 + 1.36e-01j
-9.25e-03 - 1.36e-01j
Eigenvalue
-1.28e+00 + 2.83e+00j
-1.28e+00 - 2.83e+00j

Damping
6.78E-02
6.78E-02
4.11E-01
4.11E-01
Damping
6.78E-02
6.78E-02
Damping
4.11E-01
4.11E-01

Freq. (rad/s)
1.24E-01
1.24E-01
3.10E+00
3.10E+00
Freq. (rad/s)
1.37E-01
1.37E-01
Freq. (rad/s)
3.10E+00
3.10E+00

Approximations are very close to 4th-order values

because natural frequencies are widely separated"
52!

A Little More About Output Matrices"

With Hx = I and Hu = 0!
!y = !x = H x !x; then H x = I 4
and
"
$
$
$
$
$
#

!y1 % "
'
!y2 ' $
=$
!y3 ' $
' $
!y4 ' #
&

1
0
0
0

0
1
0
0

0
0
1
0

Only output is !V!

"
(
!y = !V = "# 1 0 0 0 $% (
(
(
(#

0
0
0
1

% "$
'$
'$
'$
'$
&
#

!x1 % "
' $
!x2 ' $
!
!x3 ' $
' $
!x4 ' $#
&

!V %
'
!( '
!q '
'
!) '&

!V and !! are measured"

!V $
)
!& )
!q )
)
!' )%

"
" !y1 % " !V % " 1 0 0 0 % $
$
'=$
!y = $
'=$
'
$# !y2 '& # !( & # 0 0 0 1 & $
$
$#

!V %
'
!) '
!q '
'
!( '&

53!

A Little More About Output Matrices"

! Output (measurement) of body-axis velocity and pitch rate and
angle"
! Transformation from [!V, !$, !q, !#] to [!u, !w, !q, !!]"

#
%
%
%
%
$

& # cos ) N
( %
( = % sin ) N
( %
0
%
(
!" ' %$
0

!u
!w
!q

0 0 *VN sin ) N & #

(%
0 0 VN cos ) N ( %
(%
0 1
0
(%
1 0
1
(' %$

! Separate measurement
of state and control
" !y1 %
' " 1 0 0 0 %
$
perturbations!
" !x %
!y = $
' = H x !x + H u !u
# !u &

$
$
$
$
$
$
$
#

!y2 ' $
$
!y3 ' $
'=
!y4 ' $
' $
!y5 ' $
!y6 ' $#
&

0
0
0
0
0

1
0
0
0
0

0
1
0
0
0

0
0
1
0
0

'"
'$
'$
'$
'$
' $#
'&

!V &
(
!+ (
!q (
(
!) ('

" 0
!V % $ 0
' $
!( ' $ 0
+
!q ' $ 0
' $
!) &' $ 1
$# 0

0
0
0
0
0
1

%
'
'
' " !* E %
' $ !* T '
&
'#
'
'&
54!

Elevator-to-Normal-Velocity Numerator"
(L#E = 0)"

Transform though !N back to body axes"

H x Adj ( sI ! FLon ) G = # sin " N

$

0 0 VN cos " N

#
(
(
%(
&(
(
(
$

nVV (s) n'V (s) nqV (s) n"V (s) %

)# 0
nV' (s) n'' (s) nq' (s) n"' (s) ) ( 0
)(
nVq (s) n'q (s) nqq (s) n"q (s) ) ( M * E
)( 0
nV" (s) n'" (s) nq" (s) n"" (s) ) $
&

%
)
) = n*wE (s)
)
)
&

Scalar transfer function numerator"

%
nqV (s) )
nq' (s) )
w
#
)M
n! E (s) = $ sin " N 0 0 VN cos " N
!E
q
nq (s) )
)
nq" (s) )
&
= M ! E #$(sin " N ) nqV (s)+ (VN cos " N ) nq" (s)%&
#
(
(
%(
&(
(
(
$

Elevator-to-NormalVelocity Frequency
Response"
!(n q) = 1"
!Complex zero
almost (but not
quite) cancels
phugoid
response"

55!

M " E s 2 + 2$% n s + % n2 Approx Ph ( s & z3 )

!w(s)
n"wE (s)
=
#
!" E(s) ! Lon (s) s 2 + 2$% n s + % n2
s 2 + 2$% n s + % n2 SP
Ph

) (

0 dB/dec!
+40 dB/dec!
0 dB/dec!
40 dB/dec!

20 dB/dec!

56!

Airspeed Frequency Response to

Elevator and Thrust Inputs"
Response is primarily through the lightly damped
phugoid mode"

57!

Altitude Frequency Response

to Elevator and Thrust Inputs"

Altitude perturbation: Integral of the ight path angle perturbation"

t

!z(t) = "VN % !# ($ )d$

0

!z(s)
$ V ' !* (s)
= #& N )
% s ( !" E(s)
!" E(s)

!z(s)
$ V ' !* (s)
= #& N )
% s ( !" T (s)
!" T (s)

58!

High- and Low-Frequency Limits of

Frequency Response Function"
H ij ( j" ) = AR(" ) e j# (" )
H ij (j! " #) "

q
q$1
kij %&( j! ) + bq$1 ( j! ) + ...+ b1 ( j! ) + b0 '(

%( j! )n + an$1 ( j! )n$1 + ...+ a1 ( j! ) + a0 '

&
(

"

kij

( j! )n$q

&
k ij b0
, b0 % 0
(
k ij ( j" ) + bq$1 ( j" ) + ...+ b1 ( j" ) + b0
(
a0
H ij ( j" # 0) #
#'
n
n$1
( k ij j(0)b1 , b = 0, b % 0, etc.
( j" ) + an$1 ( j" ) + ...+ a1 ( j" ) + a0
0
1
() a0

q$1

59!

Elevator-to-Pitch-Rate"
Numerator and Transfer Function"
"
(
(
H x Adj ( sI ! FLon ) G = "# 0 0 1 0 $% (
(
(
(
#

nVV (s) n&V (s) nqV (s) n'V (s) $

)" 0
&
&
&
&
nV (s) n& (s) nq (s) n' (s) ) ( 0
)(
q
q
q
q
nV (s) n& (s) nq (s) n' (s) ) ( M * E
)( 0
nV' (s) nV' (s) nV' (s) nV' (s) ) #
%

M " E s ( s $ z1 ) ( s $ z2 )
!q(s)
n"qE (s)
=
# 2
!" E(s) ! Lon (s) s + 2%& n s + & n2
s 2 + 2%& n s + & n2
Ph

) (

$
)
) = n*qE (s)
)
)
%

SP

Free s in numerator differentiates

pitch angle transfer function"
60!

Transfer Functions of Thrust

Input to Angle Output"
Thrust-to-Flight Path Angle transfer function"

!" (s)
L
n#"T (s)
=
; n#"T (s) = T# T V s 2 + 2$% n s + % n2
!# T (s) ! Lon (s)
VN

Approx SP

Thrust-to-Angle of Attack transfer function"

!" (s)
n#"T (s)
$
'
=
; n#"T (s) = T# T s & s + 1 T )
%
"T (
!# T (s) ! Lon (s)
Thrust-to-Pitch Angle transfer function"

!" (s)
n" (s)
$
'
= #T
; n#"T (s) = T# T & s + 1 T )
%
"T (
!# T (s) ! Lon (s)

61!

Root Locus Analysis of Angular

Feedback to Thrust (4th-Order Model)"
Flight Path Angle!

Pitch Angle!

Pitch Rate!

Angle of Attack!

62!

Frequency Response of Angles

to Thrust Input"

! Primarily effects ight path angle and low-frequency pitch angle"

63!

Control Surface Dynamic

Coupling!

64!

Dynamic Model of a Control

Surface Mechanism"
Approximate control dynamics by a 2nd-order LTI system"

!!! " H !!!! " H ! ! = H # # + H command + ...

mechanism dynamics = external forcing

Bring all torques and inertias to right side"

H
!!!E = elevator =
I elevator

C H elevator

1 2
"V Sc
2

I elevator

= $%C H!!E !! E + C H! E ! E + C H# # + C H command

1 2
"V Sc
2
&
+ ...'
I elevator

( H !! E!! E + H ! E! E + H # # + H command + ...

65!

Coupling of System Model and

Control Mechanism Dynamics"
! 2nd-order model of control-deection dynamics"
! Command input from cockpit"
! Forcing by aerodynamic effects"

! Control surface deection"

! Aircraft angle of attack and angular rates"

"E
!!x" E = F" E !x" E + G" E !u" E + FSP
!x SP

# ! & # 0
% !" E ( ) %
%$ !"!!E (' $% H " E

# 0
1 &# !" E & # 0 &
(%
( !" Ecommand + %
(+ %
H "!E ('$ !"!E ' %$ *H " E ('
%$ H q

0 &# !q &
(%
(
H + (%$ !+ ('
'

! Short period approximation"

! Coupling with mechanism dynamics"
!!x SP = FSP !x SP + G SP !u SP
$
M#
$ !q! ' & M q
&
)*&
L
&% !#! )( & 1 + # V
N
%

= FSP !x SP + F"SPE !x" E

'
$ M
"E
)$ !q ' &
&
)
+
)
& L" E
)&% !# )( &% +
VN
(

0 '$
) !" E '
&
)
0 )% !"!E (
)(

66!

Short Period Model Augmented by

Control Mechanism Dynamics"

State Vector!

Augmented dynamic equation"

!!x SP /" E = FSP /" E !x SP /" E + G SP /" E !" Ecommand

!x SP '

$
&
=&
&
&
&%

!q '
)
!" )
!# E )
)
!#! E )(

Augmented stability and control matrices"

FSP/! E

"
FSP
= $ !E
$ FSP
#

F!SPE
F! E

"
$
% $
'=$
' $
& $
$
#

Mq
1
0
Hq

M(
) L(

M !E

VN

) L! E

0
H(

VN

0
H !E

0
1
H !!E

%
'
'
'
'
'
'
&

G SP /! E

" 0
$
0
=$
$ 0
$ H
# !E
67!

Roots of the Augmented Short

Period Model"
Characteristic equation for short-period/elevator dynamics"

(s # M )
q

! SP/" E ( s ) = sI n # FSP/" E =

#1
0
#H q

#M $

(s + L V )
$

0
#H $

Short Period"

! SP /" E ( s ) = s 2 + 2# SP$ nSP s + $ n2SP

#M " E

L" E

VN

s
#H " E

=0

#1
( s # H "!E )

Control Mechanism"

)( s

+ 2#" E$ n" E s + $ n2" E

)
68!

%
'
'
'
'
&

Roots of the Augmented Short

Period Model"

! Coupling of the modes

depends on design
parameters"
M!E ,

L! E

VN , H q , and H "

! Desirable for mechanical natural

frequency > short-period natural
frequency"
! Coupling dynamics can be
evaluated by root locus analysis"
69!