Problems of High Speed and Altitude

Robert Stengel, Aircraft Flight Dynamics MAE 331, 2010 Effects of air compressibility on ight stability Variable sweep-angle wings Aero-mechanical stability augmentation Altitude/airspeed instability

Outrunning Your Own Bullets

On Sep 21, 1956, Grumman test pilot Tom Attridge shot himself down, moments after this picture was taken Test ring 20mm cannons of F11F Tiger at M = 1 The combination of events
Decay in projectile velocity and trajectory drop 0.5-G descent of the F11F, due in part to its nose pitching down from ring low-mounted guns Flight paths of aircraft and bullets in the same vertical plane 11 sec after ring, Attridge ew through the bullet cluster, with 3 hits, 1 in engine

Copyright 2010 by Robert Stengel. All rights reserved. For educational use only. http://www.princeton.edu/~stengel/MAE331.html http://www.princeton.edu/~stengel/FlightDynamics.html

Aircraft crashed 1 mile short of runway; Attridge survived

Implications of Air Compressibility for Stability and Control

Early difculties with compressibility
Encountered in high-speed dives from high altitude, e.g., Lockheed P-38 Lightning

Effects of Air Compressibility on Flight Stability

Thick wing center section

Developed compressibility burble, reducing lift-curve slope and downwash

Reduced downwash
Increased horizontal stabilizer effectiveness Increased static stability Introduced a nose-down pitching moment

Solution
Auxiliary wing aps that increased both lift and drag

NACA WR-A-66, 1943

P-38 Compressibility Limit on Allowable Airspeed

from P-38 Pilot!s Manual

Compressibility Problems
Similar problems with P-39 Aircobra, P-47 Thunderbolt, and P-51 Mustang
Led to ight tests and greater understanding of compressibility effects

Cm(!) vs. CL(!)

NACA TR-767, 1943

Static margin increase with Mach number

increases control stick force required to maintain pitch trim produces pitch down
Perkins, 1970

Pilots warned to stay well below speed of sound in steep dive

Mach Tuck
Low angle of attack phenomenon
Shock-induced change in wing downwash effect on horizontal tail Pitch-down trim change,C mo , due to aft aerodynamic center shift with increasing Mach number

MV Effect on FourthOrder Roots

! Lon (s) = s 4 + DV +

L"

3 VN # M q s

F4D speed record ights (M = 0.98)

Low altitude, high temperature to increase the speed of sound High dynamic pressure 1.5 g per degree of angle of attack, M ! 1, dramatic trim changes with Mach number Pilot used nose-up trim control
Pull to push for pitch control in turn at end of each run Uncontrollable pitch-up to 9.1 g during deceleration at end of one run, due to pilot"s not compensating fast enough

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

Short Period Phugoid

D! = 0

Coupling derivative: Large positive value produces oscillatory phugoid instability Large negative value produces real phugoid divergence

Abzug & Larrabee

Pitch-Up
High angle of attack phenomenon Center of pressure moves forward due to tip stall F-86 trim change (right)
At t = 5 s, CN and Az are increasing (pitch-up), although elevator deection and control force are decreasing

Effects of F-86 Blunt-TrailingEdge Aileron

Mach Effect on Control of Wings-Level Flight Effect of Aileron Modication on RollControl Effectiveness and Response

Stick force

Aileron

NACA TR-1237

Mach number

NACA RM-A54C31, 1954

Mach number

Transonic Solutions
Application of outboard vortex generators to delay tip separation (Gloster Javelin example)

Supersonic Directional Instability

Reduced vertical stabilizer effectiveness with increasing Mach number Loss of X-2 on speed run F-100 solution: increased n size X-15 solution: wedge-shaped tail XB-70: fold-down wing tips
Improved supersonic lift Reduced excess longitudinal static stability

Mach number feedback to elevator on F-100 to counteract transonic trim change

Hypersonic Stability and Control

Turbojet/rocket for launch/takeoff Ramjet/scramjet powerplant for cruise High degree of coupling, not only of phugoid and short period but of structural and propulsive modes Poor lateral-directional characteristics Extreme sensitivity to angular perturbations Low-speed problems for high-speed congurations, e.g., takeoff/landing
NASA X-43

B-3 Concept DARPA HTV-2

VariableSweep/Incidence Wings

Boeing X-51A

Searching for the Right Design: The Many Shapes of the XF-91 Thunderceptor
Variable-incidence wing Tip chord > Root chord

Early Swing-Wing Designs

Translation as well as rotation of the wing (Messerschmitt P.1101, Bell X-5, and Grumman XF10F, below) Complicated, only partially successful Barnes Wallis"s Swallow (right), solution adopted by Polhamus and Toll at NACA Langley

Full nose inlet

Vee tail, large tip chord

Radome above inlet

Modied nose and tail

Boeing 2707-200 Supersonic Transport Concept

Length = 318 ft; 300 passengers; larger than the B-747 M = 2.7 (faster than Concorde) Cancelled before construction

Variable Sweep and Incidence

General Dynamics F-111

Variable sweep
High aspect ratio for lowspeed ight
Landing and takeoff Loiter

Low aspect ratio for highspeed ight

Reduction of transonic and supersonic drag

Boeing 2707-300

Variable incidence
Improve pilot"s line of sight for carrier landing

LTV F-8

Advanced Variable-Sweep Designs

Fairing of wing trailing edge to stabilizer leading edge at high sweep
reduces downwash at the tail and corresponding pitch stability effectively forms a delta wing

Swing-Wing Solutions
Fuel shift to move center of mass aft as wing sweeps aft Forward wing surface that extends as wing sweeps aft Advanced stability augmentation systems

Wing glove/leading-edge extension and outboard rotation point

provides greater percentage of lift at high Mach number and angle of attack
Grumman F-14 Tomcat BAE Tornado Rockwell B-1 Grumman F-14 Tomcat

Dassault Mirage G8

Oblique Wing Concepts

High-speed benets of wing sweep without the heavy structure and complex mechanism required for symmetric sweep Blohm und Voss, R. T. Jones , Handley-Page concepts Improved supersonic L/D by reduction of shock-wave interference and elimination of the fuselage in ying-wing version

NASA Oblique Wing Test Vehicles

Stability and control issues abound: The fact that birds and insects are symmetric should give us a clue (though they use huge asymmetry for control)
Strong aerodynamic and inertial longitudinal-lateral-directional coupling High side force at zero sideslip angle Torsional divergence of the leading wing

Test vehicles: Various model airplanes, NASA AD-1, and NASA DFBW F-8 (below, not built)

B-52 Control Compromises to Minimize Required Control Power

Limited-authority rudder, allowed by
Low maneuvering requirement Reduced engine-out requirement (1 of 8 engines) Crosswind landing gear

Aero-Mechanical Stability Augmentation

Limited-authority elevator, allowed by

Low maneuvering requirement Movable stabilator for trim Fuel pumping to shift center of gravity

Small manually controlled "feeler" ailerons with spring tabs

Primary roll control from powered spoilers, minimizing wing twist

Internally Balanced Control Surface

Boeing B-52

B-52 Mechanical Yaw Damper

Combined stable rudder tab, low-friction bearings, small bobweight, and eddy-current damper for B-52 Advantages
Requires no power, sensors, actuators, or computers May involve simple mechanical components

B-52 application
Control-surface n with exible seal moves within an internal cavity in the main surface Differential pressures reduce control hinge moment

Problems
Misalignment, need for high precision Friction and wear over time Jamming, galling, and fouling High sensitivity to operating conditions, design difculty

C H ! C H" " + C H# # + C H pilot input

Boeing B-47 Yaw Damper

Yaw rate gyro drives rudder to increase Dutch roll damping Comment: The plane wouldn"t need this contraption if it had been designed right in the rst place. However, mode characteristics -especially damping -- vary greatly with altitude, and most jet aircraft have yaw dampers Yaw rate washout to reduce opposition to steady turns

Northrop N-9M

Northrop YB-49 Yaw Damper

Minimal directional stability due to small vertical surfaces and short moment arm Clamshell rudders, like drag aps on the B-2 Spirit The rst stealth aircraft, though that was not intended Edwards AFB named after test pilot, Glen Edwards, Princeton MSE, killed testing the aircraft B-49s were chopped up after decision not to go into production Northrop had the last word: it built the B-2

Northrop/Grumman B-2

Northrop YB-49

High-Altitude Stall-Mach Buffet

Increased angle of attack and lift coefcient leads to Stall buffet Intermittent ow separation at transonic speed and Mach buffet The place where they meet = Cofn Corner Can induce an upset (loss of control) U-2 operates in Cofn Corner Citation X (M = 0.92) has wide buffet margin

Altitude/Airspeed Instability

U-2

Citation X

Supersonic Altitude/Airspeed Instability

Inability of XB-70, Concorde, and YF-12A/SR-71 to hold both altitude and airspeed at high speed cruise
Phugoid mode is lightly damped Height mode brought about by altitude-gradient effects Exacerbated by temperature/density gradients of the atmosphere Oblique engine-inlet shock is "spit out," decreasing thrust and increasing drag Can trigger large longitudinal or lateral-directional oscillations

Effect of Supersonic Mach Number on Altitude/Airspeed Stability

Characteristic polynomial for 2nd-order approximation
2 !(s) = s 2 + DV s + gLV / VN = s 2 + 2"# n s + # n

Engine unstart

Ph

Need for closed-loop, integrated control of altitude and airspeed

In supersonic ight (M > 1)

% M2 ( DV = 2!" nPh # ' 2 & M $ 1* )

DV decreases as M increases Phugoid stability is reduced in supersonic ight

Effect of Atmosphere Variation on Aerodynamics

Air density and sound speed vary with altitude, z

Third-Order Model Including Altitude Effects

!! height (t) = Fheight !x height (t) + G height !" T (t) x
Neglecting Mz and short-period dynamics
# )DV # !V & % ! ( % LV % % !"! ( = % VN % !! ( % $ z ' 0 % $ )g 0 )VN Lz )Dz & ( # !V & # T* T & ( ( % !" ( + % ( % 0 ( !* T VN ( % ( % !z ( % 0 ( ' ' $ 0 ($ '

! ( z ) = !SL e" z

#! ( z ) = "!SL e" z #z

a ( z ) = a zref !a ( z ) !a = !z !z

( )

!a + z " zref !z

These introduce altitude effects on lift, drag, and pitching moment

. # ( 1 ( 1 +% 2 +1 'V 2 S - 5 ! / #CT ( M ) " C D ( M ) % * ! 4CL ( M ) * $ & ) 2m 'V S - 2 ,3 ) 2m ,& ; Lz = $ ; Dz ! 0 !z !z # ( 1 +% ! 4Cm ( M ) * 'V 2 Sc - 5 ) 2I yy ,5 4 & Mz = $ !z
M= V a

3rd-degree characteristic polynomial

sI ! Fheight = " ( s ) = s 3 + DV s 2 + g
2 = ( s ! 'h ) s 2 + 2( P) nP s + ) nP = 0

LV

Lz LV % # VN + Lz s + VN $ DV VN ! Dz VN & = 0

Oscillatory phugoid mode Real height mode

(!

, " nP

#h

Approximate Roots of the 3rd-Order Equation

Assume phugoid response is fast compared to height mode response
Phugoid Mode Height Mode

3rd-Order Steady-State Response

"1 !x SS = "Fheight G height !# TSS

sI ! Fheight = " ( s ) = s 3 + DV s 2 + g

LV

Lz LV % # VN + Lz s + VN $ DV VN ! Dz VN & = 0

% !VSS ' ' !$ SS ' !z SS &

% "DV ( ' * ' L * = "' V V N * ' ) ' 0 &

"g 0 "VN

"Dz ( * Lz * VN * * 0 * )

"1

% T# T ' ' 0 ' 0 &

( * * !# TSS * )

From Flight Dynamics, pp. 476-480, with negligible Dz

# T) T % DV % & 0 ( % (=% LV % ( + T . VN ' % * - )T 0 % , DV / Lz % VN $ & ( ( ( ( !) TSS ( ( ( ( '
2 nd - order Approximation !VSS = 0 !" SS = T# T !# TSS g

Phugoid Mode

! nP " g

LV

VN + Lz ; # P "

DV 2 g LV VN + Lz

# !VSS % % !" SS % !z SS $

Height Mode

!h " #

VN $ DV %

Lz

VN # Dz VN + Lz

LV

LV

& VN '

Steady-state response to constant thrust increase

Bounded airspeed increase Horizontal ight path Bounded altitude increase

Phugoid and Height Modes of 5th-Order Longitudinal Model*

M=3 Altitude = 70,000 ft
Phugoid Period, s Phugoid Wavelength, ft x 10-6

Altitude Response of 5th-Order Longitudinal Model

Control Effects Disturbance Effects
Vertical Wind Step Thrust Control

Lift Control

Horizontal Wind Step

Phugoid Damping Ratio

Height Mode Time Constant, s

Random Vertical Wind Moment Control

* Short-period, phugoid, and height modes

!x T = [ !V !" !q !# !z ]

Stengel, 1970

Stengel, 1970

Next Time: Maneuvering and Aeroelasticity

!"##$%&%'()$ *)(%+,)$

F-86 Flight Test: Attempt to Hold Load Factor at 3 in Transonic Windup Turn

Transonic Pitchup Problem

Sign reversal of Cm! with increasing angle of attack
Combined effect of Mach number and changing downwash effects on horizontal tail
Mach number Stick force, FE Elevator deection, "E

Just below Pitch-up

Just above Pitch-up

F-86 Sabre wind-up turn

Turn at high bank angle, constant load factor, decreasing velocity, and increasing angle of attack
NACA TR-1237
Normal force coefcient, CN Normal load factor, Az

Pitch rate

Pitch moment coefcient, Cm NACA RM-A54C31, 1954

Time, s

Time, s

Flying Tail of the XF10F

Variable-sweep successor to the F9F-6 Cougar and precursor to the F-14 Tomcat T-tail assembly with controllable canard and no powered control
Like a small airplane afxed to the n Pitching moment was inadequate during landing

Handley-Page Oblique Wing Concepts

Advantages
10-20% higher L/D @ supersonic speed (compared to delta planform) Flying wing: no fuselage Which way do the passengers face? Where is the cockpit? How are the engines and vertical surfaces swiveled? What does asymmetry do to stability and control?

Issues

Boeing 2707-300 Supersonic Transport

Variable-sweep wing dropped in favor of more conventional design Final conguration had weight and aeroelastic problems Project cancelled in 1971 due to sonic boom, takeoff sideline noise and cost problems

B-52 Rudder Control Linkages