calvi liege 2010 students
TRANSCRIPT
-
8/13/2019 Calvi Liege 2010 Students
1/144
Spacecraft Structural Dynamics & Loads - A. Calvi 1
This presentation is distributed to the students of the University of LiegeSatell ite Engineering Class November 29, 2010)
This presentation is not for further distribution
Spacecraft Structural Dynamics & Loads
An Overview
Adriano Calvi, PhDESA / ESTEC, Noordwijk, The Netherlands
-
8/13/2019 Calvi Liege 2010 Students
2/144
Spacecraft Structural Dynamics & Loads - A. Calvi 2
ForewordForeword
This half-day course on Structural Dynamics and Loads intends topresent the subject within the broad context of the development ofspacecraft structures.
Basic notions as well as some advanced
concepts are explained with minimum mathematics
The content is the result of the authors experience acquiredthrough his involvement with research and industrial activitiesmainly at the European Space Agency and Alenia Spazio
The course is specifically tailored for university students
-
8/13/2019 Calvi Liege 2010 Students
3/144
Spacecraft Structural Dynamics & Loads - A. Calvi 3
Specific Objectives of the Presentation
To provide a short overview about structural dynamics and its
importance in the development of the spacecraft structures (design,analysis & test)
To introduce the students to the logic and criteria
as regards
dynamics and loads
To point out the importance of some topics such as modaleffective mass, dynamic testing
and model validation
often notaddressed in the University Courses
To show results of some applications (satellites, launchers, etc.)
To testify the importance of structural dynamics analysis (andspecifically of some numerical methods)
-
8/13/2019 Calvi Liege 2010 Students
4/144
Spacecraft Structural Dynamics & Loads - A. Calvi 4
Spacecraft Structural Dynamics & Loads (1)
Introduction
Preliminary concepts: the launch mechanical environment
Requirements for spacecraft structures
The role of structural dynamics in a space project
Dynamic analysis types
Real eigenvalue
Frequency response
Transient response
Shock response
Random vibration
The effective mass concept
Preliminary design, design load cycles & verification loads cycle
-
8/13/2019 Calvi Liege 2010 Students
5/144
Spacecraft Structural Dynamics & Loads - A. Calvi 5
Spacecraft Structural Dynamics & Loads (2)
Payload-launcher Coupled Loads Analysis (CLA)
Mechanical tests
Modal survey test
Sinusoidal vibration test
Acoustic noise test
Shock test
Random vibration test
Overtesting, notching
and sine vibration testing
Mathematical model updating and validation
Summary and conclusive remarks
Bibliography
-
8/13/2019 Calvi Liege 2010 Students
6/144
Spacecraft Structural Dynamics & Loads - A. Calvi 6
Preliminary concepts (1)Preliminary concepts (1)
Structural dynamics is the study of structures subjected to amechanical environment which depends on time and leading to amovement
Excitation transmission types (mechanical & acoustic)
Type of time functions (sinusoidal, transient, random)
Type of frequencies involved (low frequency, broadband)
Domain of analysis (time domain, frequency domain)
Structure representation with a mathematical model (continuous ordiscrete)
-
8/13/2019 Calvi Liege 2010 Students
7/144
Spacecraft Structural Dynamics & Loads - A. Calvi 7
Preliminary concepts (2)Preliminary concepts (2)
The parameter most commonly used (in the industry) to define the motionof a mechanical system
is the acceleration
Typical ranges of acceleration of concern in aerospace structures are from0.01 g to 10,000 g.
Frequency (Hz or rad/s) and octave
Vibroacoustics, pressure (N/m 2) and Sound Pressure Level (dB)
Random vibration and (acceleration) Power Spectral Density (g 2/Hz)
Shock Response Spectrum
Root mean square (rms) = square root of the mean of the sum of all thesquares
Note 1: the decibel is a tenth of a bel, the logarithm (base 10)
of a power ratio (itis accepted that power is proportional to the square of the rms
of acceleration,velocity, pressure, etc.)
Note 2: it must be emphasized that dB in acoustics is not an unit of acousticpressure but simply a power ratio with respect to a reference pressure whichmust be stated or clearly implicit
-
8/13/2019 Calvi Liege 2010 Students
8/144
Spacecraft Structural Dynamics & Loads - A. Calvi 8
Example of satellite structural design conceptExample of satellite structural design concept
-
8/13/2019 Calvi Liege 2010 Students
9/144
Spacecraft Structural Dynamics & Loads - A. Calvi 9
Accelerations Accelerations some remarkssome remarks
The parameter most commonly used (in the industry) to define themotion of a mechanical system is the acceleration
Good reasons: accelerations are directly related to forces/stresses andeasy
to specify and measure
Some hidden
assumptions
Criteria for equivalent structural damage (e.g. shock response spectra)Note: failures usually happen in the largest stress areas, regardless if theyare the largest acceleration areas!
Rigid or static determinate junction (e.g. quasi-static loads)
Important consequences
Need for considering the actual
(e.g. test
or
flight) boundary conditions
(e.g. for the purpose of notching)
Need for a valid
F.E. model (e.g. to be used for force and stress recovery)
-
8/13/2019 Calvi Liege 2010 Students
10/144
Spacecraft Structural Dynamics & Loads - A. Calvi 10
Mechanical loads are caused by:Mechanical loads are caused by:
Transportation
Rocket Motor Ignition Overpressure
Lift-off Loads
Engine/Motor Generated Acoustic Loads
Engine/Motor Generated Structure-borne Vibration Loads
Engine/Motor Thrust Transients
Pogo Instability, Solid Motor Pressure Oscillations
Wind and Turbulence, Aerodynamic Sources
Liquid Sloshing in Tanks
Stage and Fairing Separation Loads
Pyrotechnic Induced Loads
Manoeuvring Loads
Flight Operations, Onboard Equipment Operation
-
8/13/2019 Calvi Liege 2010 Students
11/144
Spacecraft Structural Dynamics & Loads - A. Calvi 11
Launch mechanical environmentLaunch mechanical environment
Steady state accelerations
Low frequency vibrations
Broad band vibrations
Random vibrations
Acoustic loads
Shocks
Loads (vibrations) are transmitted to the payload (e.g. satellite)
through its mechanical interface
Acoustic loads also directly excite payload surfaces
-
8/13/2019 Calvi Liege 2010 Students
12/144
Spacecraft Structural Dynamics & Loads - A. Calvi 12
SteadySteady --statestate and lowand low --frequency transient accelerationsfrequency transient accelerations
-
8/13/2019 Calvi Liege 2010 Students
13/144
Spacecraft Structural Dynamics & Loads - A. Calvi 13
Acoustic Loads Acoustic Loads
During the lift off and the early phases ofthe launch an extremely high level ofacoustic noise surrounds the payload
The principal sources of noise are:
Engine functioning
Aerodynamic turbulence
Acoustic noise (as pressure waves)impinging on light weight panel-likestructures produce high response
-
8/13/2019 Calvi Liege 2010 Students
14/144
Spacecraft Structural Dynamics & Loads - A. Calvi 14
Broadband and high frequency vibrationsBroadband and high frequency vibrations
Broad band random vibrations are produce by:
Engines functioning
Structural response to broad-band acoustic loads
Aerodynamic turbulent boundary layer
-
8/13/2019 Calvi Liege 2010 Students
15/144
Spacecraft Structural Dynamics & Loads - A. Calvi 15
ShocksShocks
Mainly caused by the actuation of pyrotechnic devices:
Release mechanisms for stage and satellite separation
Deployable mechanisms for solar arrays etc.
-
8/13/2019 Calvi Liege 2010 Students
16/144
Spacecraft Structural Dynamics & Loads - A. Calvi 16
Static and dynamic environment specification (typical ranges)Static and dynamic environment specification (typical ranges)
-
8/13/2019 Calvi Liege 2010 Students
17/144
Spacecraft Structural Dynamics & Loads - A. Calvi 17
QuasiQuasi --Static Loads (accelerations)Static Loads (accelerations)
Loads independent of time or which vary slowly, so that the dynamicresponse of the structure is not significant
(ECSS-E-ST-32). Note: this is
the definition of a quasi-static event!
Combination of static and low frequency loads into an equivalent
static
load specified for design purposes as C.o.G. acceleration
(e.g. NASA RP- 1403, NASA-HDBK-7004). Note: this definition is fully adequate for the
design of the spacecraft primary structure. For the design of componentsthe contribution of the high frequency loads, if relevant, is included as well!
CONCLUSION: quasi static loading means under steady-stateaccelerations
(unchanging applied force balanced by inertia loads). For
design purposes (e.g. derivation of design limit loads, selection of thefasteners, etc.), the quasi-static loads are normally calculated bycombining both static and dynamic load contributions. In this context thequasi static loads are equivalent to (or interpreted by the designer
as)
static loads , typically expressed as equivalent accelerations at the C.o.G.
-
8/13/2019 Calvi Liege 2010 Students
18/144
Spacecraft Structural Dynamics & Loads - A. Calvi 18
Typical Requirements for Spacecraft Structures
Strength
Structural life
Structural response
Stiffness
Damping
Mass Properties
Dynamic Envelope
Positional Stability
Mechanical Interface
Basic requirement : the structure shall support the payload andspacecraft subsystems with enough strength and stiffness topreclude any failure (rupture, collapse, or detrimental deformation)that may keep them from working successfully.
-
8/13/2019 Calvi Liege 2010 Students
19/144
Spacecraft Structural Dynamics & Loads - A. Calvi 19
Requirements evolutionRequirements evolution
-
8/13/2019 Calvi Liege 2010 Students
20/144
Spacecraft Structural Dynamics & Loads - A. Calvi 20
Design requirements and verificationDesign requirements and verification
-
8/13/2019 Calvi Liege 2010 Students
21/144
Spacecraft Structural Dynamics & Loads - A. Calvi 21
Examples of (Mechanical) Requirements (1)
The satellite shall be compatible with 2 launchers (potential candidates:VEGA, Soyuz in CSG, Rockot, Dnepr)...
The satellite and all its units shall withstand applied loads due to themechanical environments to which they are exposed during the service-life
Design Loads shall be derived by multiplication of the Limit Loads by a designfactor equal to 1.25 (i.e. DL= 1.25 x LL)
The structure shall withstand the worst design loads without failing orexhibiting permanent deformations.
Buckling is not allowed.
The natural frequencies of the structure shall be within adequate bandwidthsto prevent dynamic coupling with major excitation frequencies
The spacecraft structure shall provide the mounting interface to
the launchvehicle and comply with the launcher interface requirements.
-
8/13/2019 Calvi Liege 2010 Students
22/144
Spacecraft Structural Dynamics & Loads - A. Calvi 22
Examples of (Mechanical) Requirements (2)
All the Finite Element Models (FEM) prepared to support the mechanicalverification activities at subsystem and satellite level shall be delivered in
NASTRAN format
The FEM of the spacecraft in its launch configuration shall be detailed enough toensure an appropriate derivation and verification of the design loads and of themodal response of the various structural elements of the satellite up to 140 Hz
A reduced FEM of the entire spacecraft correlated with the detailed FEM shall bedelivered for the Launcher Coupled Loads Analysis (CLA)
The satellite FEMs
shall be correlated against the results of modal survey testscarried out at complete spacecraft level, and at component level
for units above
50 kg
The structural model of the satellite shall pass successfully qualification sinevibration Test.
The flight satellite shall pass successfully acceptance sine vibration test.
-
8/13/2019 Calvi Liege 2010 Students
23/144
Spacecraft Structural Dynamics & Loads - A. Calvi 23
Spacecraft stiffness requirements for different launchersSpacecraft stiffness requirements for different launchers
Launch vehicle manuals specify minimum values for the payload natural(fundamental) frequency of vibration in order to avoid dynamic coupling betweenlow frequency dynamics of the launch vehicle and payload modes
-
8/13/2019 Calvi Liege 2010 Students
24/144
Spacecraft Structural Dynamics & Loads - A. Calvi 24
The Role of Structural Dynamics in a Space ProjectThe Role of Structural Dynamics in a Space Project
Mechanical environment definition (structural response and loads
identification by analysis and test)
Launcher/Payload coupled loads analysis
Random vibration and vibroacoustic analyses
Jitter analysis
Test predictions (e.g. sine test by frequency response analysis)
Test evaluations (sine, acoustic noise)
Input to structural life analysis (e.g. generation of the loading spectrum)
Structural identification (by analysis and test)
Modal analysis
Modal survey test and experimental modal analysis
Mathematical model updating and validation
Design qualification and flight product acceptance
Qualification and Acceptance tests (sine, random, acoustic noise, shock)
-
8/13/2019 Calvi Liege 2010 Students
25/144
Spacecraft Structural Dynamics & Loads - A. Calvi 25
Spacecraft Structural Dynamics & Loads (1)
Introduction
Preliminary concepts: the launch mechanical environment
Requirements for spacecraft structures
The role of structural dynamics in a space project
Dynamic analysis types
Real eigenvalue
Frequency response
Transient response
Shock response
Random vibration
The effective mass concept
Preliminary design, design load cycles & verification loads cycle
-
8/13/2019 Calvi Liege 2010 Students
26/144
Spacecraft Structural Dynamics & Loads - A. Calvi 26
Dynamic analysis typesDynamic analysis types
Real eigenvalue analysis (undamped free vibrations)
Modal parameter identification, etc.
Linear frequency response analysis (steady-state response oflinear structures to loads that vary as a function of frequency)
Sine test prediction, transfer functions calculation, LV/SC CLA etc.
Linear transient response analysis (response of linear structures toloads that vary as a function of time).
LV/SC CLA, base drive analysis, jitter analysis, etc.
Shock response spectrum analysis
Specification of equivalent environments (e.g. equivalent sine input),
Shock test specifications, etc.
Vibro-acoustics (FEM/BEM, SEA) & Random vibration analysis
Vibro-acoustic test prediction & random vibration environment definition
Loads analysis for base-driven random vibration
-
8/13/2019 Calvi Liege 2010 Students
27/144
Spacecraft Structural Dynamics & Loads - A. Calvi 27
Reasons to compute normal modes (real eigenvalue analysis)Reasons to compute normal modes (real eigenvalue analysis)
To verify stiffness requirements
To assess the dynamic interaction between a component and itssupporting structure
To guide experiments (e.g. modal survey test)
To validate computational models (e.g. test/analysis correlation)
As pre-requisite for subsequent dynamic analyses
To evaluate design changes
Mathematical model quality check (model verification)
Numerical methods: Lanczos,
-
8/13/2019 Calvi Liege 2010 Students
28/144
Spacecraft Structural Dynamics & Loads - A. Calvi 28
Real eigenvalue analysisReal eigenvalue analysis
Note: mode shape normalizationScaling is arbitrary
Convention: Mass, Max or Point
-
8/13/2019 Calvi Liege 2010 Students
29/144
Spacecraft Structural Dynamics & Loads - A. Calvi 29
Mode shapesMode shapes
Cantilever beam
Simply supported beam
-
8/13/2019 Calvi Liege 2010 Students
30/144
Spacecraft Structural Dynamics & Loads - A. Calvi 30
Satellite Normal Modes Analysis
Mode 1: 16.2 Hz Mode 2: 18.3 Hz
INTEGRAL Satellite (FEM size 120000 DOFs)
-
8/13/2019 Calvi Liege 2010 Students
31/144
Spacecraft Structural Dynamics & Loads - A. Calvi 31
Frequency Response AnalysisFrequency Response Analysis
Used to compute structural response to steady-state harmonicexcitation
The excitation is explicitly defined in the frequency domain
Forces can be in the form of applied forces and/or enforcedmotions
Two different numerical methods: direct and modal
Damped forced vibration equation of motion with harmonicexcitation:
-
8/13/2019 Calvi Liege 2010 Students
32/144
Spacecraft Structural Dynamics & Loads - A. Calvi 32
Frequency response considerationsFrequency response considerations
If the maximum excitation frequency is much less than the lowest resonant frequency of the system, a static analysis is probably
sufficient
Undamped or very lightly damped structures exhibit large dynamic
responses for excitation frequencies near natural frequencies(resonant frequencies)
Use a fine enough frequency step size ( f) to adequately predictpeak response.
Smaller frequency spacing should be used in regions near resonantfrequencies, and larger frequency step sizes should be used inregions away from resonant frequencies
-
8/13/2019 Calvi Liege 2010 Students
33/144
Spacecraft Structural Dynamics & Loads - A. Calvi 33
Harmonic forced response with dampingHarmonic forced response with damping
-
8/13/2019 Calvi Liege 2010 Students
34/144
Spacecraft Structural Dynamics & Loads - A. Calvi 34
Transient Response AnalysisTransient Response Analysis
Purpose is to compute the behaviour of a structure subjected to time- varying excitation
The transient excitation is explicitly defined in the time domain
Forces can be in the form of applied forces and/or enforced motions
The important results obtained from a transient analysis are typically
displacements, velocities, and accelerations of grid points, and forces and stresses in elements
Two different numerical methods: direct
(e.g. Newmark) and modal
(e.g. Lanczos
+ Duhamels integral or Newmark)
Dynamic equation of motion:
-
8/13/2019 Calvi Liege 2010 Students
35/144
Spacecraft Structural Dynamics & Loads - A. Calvi 35
Modal Transient Response AnalysisModal Transient Response Analysis
-
8/13/2019 Calvi Liege 2010 Students
36/144
Spacecraft Structural Dynamics & Loads - A. Calvi 36
Transient response considerationsTransient response considerations
The integration time step must be small enough to represent accuratelythe variation in the loading
The integration time step must also be small enough to represent
themaximum frequency of interest (cut-off frequency)
The cost of integration is directly proportional to the number of time steps
Very sharp spikes in a loading function induce a high-frequency transientresponse. If the high-frequency transient response is of primaryimportance in an analysis, a very small integration time step must be
used
The loading function must accurately describe the spatial and temporaldistribution of the dynamic load
-
8/13/2019 Calvi Liege 2010 Students
37/144
Spacecraft Structural Dynamics & Loads - A. Calvi 37
ShockShock response spectrum (and analysis)response spectrum (and analysis)
Response spectrum analysis is an approximate method ofcomputing the peak response of a transient excitation
applied to a
structure or component
There are two parts to response spectrum analysis: (1) generationof the spectrum and (2) use of the spectrum for dynamic response
such as stress analysis
Note 1: the part (2)
of the response spectrum analysis has alimited use in structural dynamics of spacecraft (e.g. preliminarydesign) since the accuracy of the method may be questionable
Note 2: the term shock
can be misleading (not always a physicalshock, i.e. an environment of a short duration, is involved. Itwould be better to use response spectrum)
-
8/13/2019 Calvi Liege 2010 Students
38/144
Spacecraft Structural Dynamics & Loads - A. Calvi 38
Generation of a response spectrum (1)Generation of a response spectrum (1)
-
8/13/2019 Calvi Liege 2010 Students
39/144
Spacecraft Structural Dynamics & Loads - A. Calvi 39
Generation of a response spectrum (2)Generation of a response spectrum (2)
the peak response for one oscillator does not necessarily occur at thesame time as the peak response for another oscillator
there is no phase information since only the magnitude of peak response iscomputed
It is assumed in this process that each oscillator mass is very small relativeto the base structural mass so that the oscillator does not influence the
dynamic behaviour of the base structure
-
8/13/2019 Calvi Liege 2010 Students
40/144
Spacecraft Structural Dynamics & Loads - A. Calvi 40
Shock Response Spectrum. Some remarksShock Response Spectrum. Some remarks
The 1-DOF system is used as reference structure (since thesimplest) for the characterization of environments (i.e.
quantification of the severity equivalent environments can bespecified)
In practice, the criterion used for the severity is the maximumresponse which occurs on the structure
(note: another criterion
relates to the concept of fatigue damage)
A risk in comparing two excitations of different nature is in theinfluence of damping on the results (e.g. maxima are proportionalto Q for sine excitation and variable for transient excitation!)
The absolute acceleration spectrum is used, which providesinformation about the maximum internal forces and stresses
The shock spectrum is a transformation of the time history which
is
not reversible (contrary to Fourier transform)
-
8/13/2019 Calvi Liege 2010 Students
41/144
Spacecraft Structural Dynamics & Loads - A. Calvi 41
Shock Response SpectrumShock Response Spectrum
(A) is the shock spectrumof a terminal peaksawtooth (B) of 500 Gpeak amplitude and 0.4millisecond duration
-
8/13/2019 Calvi Liege 2010 Students
42/144
Spacecraft Structural Dynamics & Loads - A. Calvi 42
Random vibration (analysis)Random vibration (analysis)
Random vibration is vibration that can be described only in astatistical sense
The instantaneous magnitude is not known at any given time;rather, the magnitude is expressed in terms of its statisticalproperties (such as mean value, standard deviation, and probabilityof exceeding a certain value)
Examples of random vibration include earthquake ground motion,wind pressure fluctuations on aircraft, and acoustic excitation dueto rocket and jet engine noise
These random excitations are usually described in terms of a
power spectral density (PSD) function
Note: in structural dynamics of spacecraft, the random vibrationanalysis is often performed with simplified techniques (e.g. basedon Miles
equation
+ effective modal mass models)
-
8/13/2019 Calvi Liege 2010 Students
43/144
Spacecraft Structural Dynamics & Loads - A. Calvi 43
Random noise with normal amplitude distributionRandom noise with normal amplitude distribution
-
8/13/2019 Calvi Liege 2010 Students
44/144
Spacecraft Structural Dynamics & Loads - A. Calvi 44
Power Spectral Density (conceptual model)Power Spectral Density (conceptual model)
-
8/13/2019 Calvi Liege 2010 Students
45/144
Spacecraft Structural Dynamics & Loads - A. Calvi 45
Sound Pressure Level (conceptual model)Sound Pressure Level (conceptual model)
-
8/13/2019 Calvi Liege 2010 Students
46/144
-
8/13/2019 Calvi Liege 2010 Students
47/144
Spacecraft Structural Dynamics & Loads - A. Calvi 47
Modal effective mass (1)Modal effective mass (1)
It may be defined as the mass terms in a modal expansion of thedrive point apparent mass of a kinematically
supported system
Note: driving-point FRF: the DOF response is the same as the excitation
This concept applies to structure with base excitation
Important particular case: rigid or statically determinate junction
It provides an estimate of the participation of a vibration mode, interms of the load it will cause in the structure, when excited
Note: avoid using: it is the mass which participates to the mode!
Dynamic amplification factor
Modal reaction forces
Base (junction) excitation
-
8/13/2019 Calvi Liege 2010 Students
48/144
Spacecraft Structural Dynamics & Loads - A. Calvi 48
Modal effective mass (2)Modal effective mass (2)
The effective mass matrix can be calculated either by the modalparticipation factors
or by using the modal interface forces
Normally only the values on the leading diagonal of the modaleffective mass matrix are considered and expressed in percentage
of the structure rigid body properties (total mass and secondmoments of inertia)
The effective mass characterises the mode and it is independentfrom the eigenvector normalisation
Gen. mass
Resultant of modal interface forces
i-th mode Rigid body modes
Modal participation factorsEffective mass
Eigenvector max value
-
8/13/2019 Calvi Liege 2010 Students
49/144
Spacecraft Structural Dynamics & Loads - A. Calvi 49
Modal effective mass (3)Modal effective mass (3)
For the complete set of modes the summation of the modaleffective mass is equal to the rigid body mass
Contributions of each individual mode to the total effective masscan be used as a criterion to classify the modes
(global or local)
and an indicator of the importance of that mode, i.e. an indication ofthe magnitude of participation in the loads analysis
It can be used to construct a list of important modes for thetest/analysis correlation and it is a significant correlation parameter
It can be used to create simplified mathematical models (equivalentmodels with respect to the junction)
-
8/13/2019 Calvi Liege 2010 Students
50/144
Spacecraft Structural Dynamics & Loads - A. Calvi 50
Example of Effective Mass table(MPLM test and FE model)
-
8/13/2019 Calvi Liege 2010 Students
51/144
Spacecraft Structural Dynamics & Loads - A. Calvi 51
Spacecraft Structural Dynamics & Loads (1)
Introduction
Preliminary concepts: the launch mechanical environment
Requirements for spacecraft structures
The role of structural dynamics in a space project
Dynamic analysis types
Real eigenvalue
Frequency response
Transient response
Shock response
Random vibration
The effective mass concept
Preliminary design, design load cycles & verification loads cycle
-
8/13/2019 Calvi Liege 2010 Students
52/144
Spacecraft Structural Dynamics & Loads - A. Calvi 52
A5 Typical Sequence of events A5 Typical Sequence of events
-
8/13/2019 Calvi Liege 2010 Students
53/144
Spacecraft Structural Dynamics & Loads - A. Calvi 53
A5 Typical Longitudinal Static Acceleration A5 Typical Longitudinal Static Acceleration
-
8/13/2019 Calvi Liege 2010 Students
54/144
Spacecraft Structural Dynamics & Loads - A. Calvi 54
Sources of Structural Loadings (Launch)Sources of Structural Loadings (Launch)
Axial-Acceleration Profile for the Rockot Launch Vehicle
g8
7
6
5
43
2
1
0t,150 200 250 300100500
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1-20
-10
0
10
20
30
40
50
60
70
80
a c c e
l e r a t i o n
[ m / s
2 ]
t [s]
-
8/13/2019 Calvi Liege 2010 Students
55/144
Spacecraft Structural Dynamics & Loads - A. Calvi 55
Axial Acceleration at Launcher/Satellite Interface (Engines Cut-off)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1-20
-10
0
10
20
30
40
50
60
70
80
a c c e l e r a t
i o n
[ m / s
2 ]
t [s]
-
8/13/2019 Calvi Liege 2010 Students
56/144
Spacecraft Structural Dynamics & Loads - A. Calvi 56
Load Factors for Preliminary Design (Ariane 5)Load Factors for Preliminary Design (Ariane 5)
-
8/13/2019 Calvi Liege 2010 Students
57/144
Spacecraft Structural Dynamics & Loads - A. Calvi 57
QuasiQuasi --static loads for different launchersstatic loads for different launchers
-
8/13/2019 Calvi Liege 2010 Students
58/144
Spacecraft Structural Dynamics & Loads - A. Calvi 58
QuasiQuasi --Static Flight Limit loads for Dnepr and SoyuzStatic Flight Limit loads for Dnepr and Soyuz
-
8/13/2019 Calvi Liege 2010 Students
59/144
Spacecraft Structural Dynamics & Loads - A. Calvi 59
Example of Physical (and Modal) Mass Acceleration Curvefor preliminary design of payload hardware or equipment items
100
10
11 10 100 1000
Physical
Modal
Effective Mass, kg
D y n a m
i c R e s p o n s e
A c c e
l e r a
t i o n ,
g
-
8/13/2019 Calvi Liege 2010 Students
60/144
Spacecraft Structural Dynamics & Loads - A. Calvi 60
Load Combination Criteria for Components(International Space Station Program)
-
8/13/2019 Calvi Liege 2010 Students
61/144
Spacecraft Structural Dynamics & Loads - A. Calvi 61
Loads and Factors
Expendable launch vehicles,pressurized hardware andmanned system Test Logic
Common Design LogicSatellitesTest Logic
Limit Loads - LL
Design Limit LoadsDLL
x Coef. A
DYL
x Coef. B
DUL
x Coef. C
x KQ x KA
QL AL
x KQ x KA
QL
AL
I n c r e a s
i n g
L o a
d
L e v e
l
ECSS E-ST-32-10
-
8/13/2019 Calvi Liege 2010 Students
62/144
Spacecraft Structural Dynamics & Loads - A. Calvi 62
Some definitionsSome definitions
Design:
The process used to generate the set information describing the
essential characteristics of a product (ECSS-P-001A)
Design means developing requirements, identifying options, doing analyses and trade studies, and defining a product in enough detail so
it can be built (T. P. Sarafin)
Verification:
Confirmation by examination and provision of objective evidence thatspecified requirements have been fulfilled (ISO 8402:1994)
Verification means providing confidence through disciplined steps thata product will do what it is supposed to do (T. P. Sarafin)
Note: we can prove
that the spacecraft satisfies the measurable criteriawe have defined, but we cannot prove
a space mission will be successful
-
8/13/2019 Calvi Liege 2010 Students
63/144
Spacecraft Structural Dynamics & Loads - A. Calvi 63
Design Loads CyclesDesign Loads Cycles A load cycle is the process of:
Generating and combining math models for a proposed design
Assembling and developing forcing functions, load factors, etc. tosimulate the critical loading environment
Calculating design loads and displacements for all significantground, launch and mission events
Assessing the results to identify design modifications or risks
Then, if necessary, modifying the design accordingly or choosing
to
accept the risk
-
8/13/2019 Calvi Liege 2010 Students
64/144
Spacecraft Structural Dynamics & Loads - A. Calvi 64
Design loads cycle processDesign loads cycle process
-
8/13/2019 Calvi Liege 2010 Students
65/144
Spacecraft Structural Dynamics & Loads - A. Calvi 65
Final VerificationFinal Verification
Consist of:
Making sure all requirements are satisfied ( compliance )
Validating the methods and assumptions used to satisfyrequirements
Assessing risks
-
8/13/2019 Calvi Liege 2010 Students
66/144
Spacecraft Structural Dynamics & Loads - A. Calvi 66
Final Verification (crucial points)Final Verification (crucial points)
To perform a Verification Loads Cycle for structures designedand tested to predicted loads
Finite element models correlation with the results of modal andstatic testing
Loads prediction with the current forcing functions
Compliance with analysis criteria (e.g. MOS>0)
To make sure the random-vibration environments used toqualify components were high enough (based on datacollected during the spacecraft acoustic test)
Note: in the verification loads cycle instead of identifying requireddesign changes (design loads cycle) the adequacy of the structurethat has already been built and tested is assessed
-
8/13/2019 Calvi Liege 2010 Students
67/144
Spacecraft Structural Dynamics & Loads - A. Calvi 67
Criteria for Assessing Verification Loads (strength)Criteria for Assessing Verification Loads (strength)
Analysis: margins of safety must me greater that or equal to zero
Test: Structures qualified by static or sinusoidal testing
Test loads or stresses as predicted
(test-verified math model and test
conditions) are compared with the total predicted loads during themission
(including flying transients, acoustics, random vibration,
pressure, thermal effects and preloads)
Test: Structures qualified by acoustic or random vibration testing
Test environments are compared with random-vibration environmentsderived from system-level acoustic testing
-
8/13/2019 Calvi Liege 2010 Students
68/144
Spacecraft Structural Dynamics & Loads - A. Calvi 68
Spacecraft Structural Dynamics & Loads (2)
Payload-launcher Coupled Loads Analysis (CLA)
Mechanical tests
Modal survey test
Sinusoidal vibration test
Acoustic noise test
Shock test
Random vibration test
Overtesting, notching
and sine vibration testing
Mathematical model updating and validation
Summary and conclusive remarks
Bibliography
-
8/13/2019 Calvi Liege 2010 Students
69/144
UPPERCOMPOSITE
-
8/13/2019 Calvi Liege 2010 Students
70/144
Spacecraft Structural Dynamics & Loads - A. Calvi 70
Ariane-5 Dynamic Mathematical ModelPAYLOAD
EAP+EAP-
EPC
Dynamic effects up to about 100 Hz
3D FE models of EPC, EAP, UC
Dynamic Reduction using Craig-Bampton
formulation
Incompressible or compressible fluids models for liquidpropellants
Structure/fluid interaction
Nearly incompressible SRB solid propellant modeling
Pressure and stress effects on launcher stiffness
SRB propellant and DIAS structural damping
Non-linear launch table effects
-
8/13/2019 Calvi Liege 2010 Students
71/144
Spacecraft Structural Dynamics & Loads - A. Calvi 71
Sizing flight events (CLA with VEGA Launcher)Sizing flight events (CLA with VEGA Launcher)
1. Lift-off (P80 Ignition and Blastwave)
2. Mach1/QMAX Gust
3. P80 Pressure Oscill ations
4. Z23 Igni tion
5. Z23 Pressure Oscillations
6. Z9 Ignition
-
8/13/2019 Calvi Liege 2010 Students
72/144
Spacecraft Structural Dynamics & Loads - A. Calvi 72
CLA OutputCLA Output
LV-SC interface accelerations
Equivalent sine spectrum
LV-SC interface forces
Equivalent accelerations at CoG
Internal responses
Accelerations,
Displacements
Forces
Stresses
-
8/13/2019 Calvi Liege 2010 Students
73/144
Spacecraft Structural Dynamics & Loads - A. Calvi 73
Payload / STS CLAPayload / STS CLA
Lift-off Force Resultant in X [lbf]
Lift-off Main Fitting
I/F ForceX Dir. [N]
Lift-off Main Fitting
I/F Force
Z Dir. [N]
Lift-off Keel Fitting
I/F ForceY Dir. [N]
-
8/13/2019 Calvi Liege 2010 Students
74/144
Spacecraft Structural Dynamics & Loads - A. Calvi 74
Spacecraft Structural Dynamics & Loads (2)
Payload-launcher Coupled Loads Analysis (CLA)
Mechanical tests
Modal survey test
Sinusoidal vibration test
Acoustic noise test
Shock test
Random vibration test
Overtesting, notching
and sine vibration testing
Mathematical model updating and validation
Summary and conclusive remarks
Bibliography
-
8/13/2019 Calvi Liege 2010 Students
75/144
Spacecraft Structural Dynamics & Loads - A. Calvi 75
Testing techniquesTesting techniques Introduction (1)Introduction (1)
Without testing, an analysis can give completely incorrect results
Without the analysis, the tests can represent only a very limited reality
Two types of tests according to the objectives to be reached:
Simulation tests for structure qualification or acceptance
Identification tests (a.k.a. analysis-validation tests) for structureidentification (the objective is to determine the dynamic characteristics ofthe tested structure in order to update
the mathematical model)
Note: identification and simulation tests are generally completelydissociated. In certain cases (e.g. spacecraft sine test) it is technicallypossible to perform them using the same test facility
-
8/13/2019 Calvi Liege 2010 Students
76/144
Spacecraft Structural Dynamics & Loads - A. Calvi 76
Testing techniquesTesting techniques Introduction (2)Introduction (2)
Generation of mechanical environment
Small shakers (with flexible rod; electrodynamic)
Large shakers (generally used to impose motion at the base)
Electrodynamic shaker
Hydraulic jack shaker
Shock machines (pyrotechnic generators and impact machines)
Noise generators + reverberant acoustic chamber (homogeneous and
diffuse field)
Measurements
Force sensors, calibrated strain gauges
Accelerometers, velocity or displacement sensors
-
8/13/2019 Calvi Liege 2010 Students
77/144
Spacecraft Structural Dynamics & Loads - A. Calvi 77
Classes of tests used to verify requirements (purposes)Classes of tests used to verify requirements (purposes)
Development test
Demonstrate design concepts and acquire necessary information fordesign
Qualification test
Show a design
is adequate by testing a single article
Acceptance test
Show a product
is adequate (test each flight article)
Analysis validation test
Provide data which enable to confirm critical analyses or to change(update/validate) mathematical models and redo analyses
-
8/13/2019 Calvi Liege 2010 Students
78/144
Spacecraft Structural Dynamics & Loads - A. Calvi 78
Tests for verifying mechanical requirements (purposes)Tests for verifying mechanical requirements (purposes)
Acoustic test
Verify strength and structural life by introducing random vibration
through acoustic pressure (vibrating air molecules)
Note: acoustic tests at spacecraft level are used to verify adequacy ofelectrical connections and validate the random vibration environmentsused to qualify components
(Pyrotechnic) shock test
Verify resistance to high-frequency shock waves caused by separationexplosives (introduction of high-energy vibration up to 10,000 Hz)
System-level tests are used to verify levels used for component testing
Random vibration test
Verify strength and structural life by introducing random vibrationthrough the mechanical interface (typically up to 2000 Hz )
-
8/13/2019 Calvi Liege 2010 Students
79/144
-
8/13/2019 Calvi Liege 2010 Students
80/144
Spacecraft Structural Dynamics & Loads - A. Calvi 80
RockotRockot Dynamic SpecificationDynamic Specification
Marketed by: Eurockot
Actually flight qualified
Manufactured by: Khrunichev
Capability: 950 kg @ 500 km
Launch site: Plesetsk
Environment Level
Sine vibrationLongitudinal = 1 g on [5-10] Hz 1.5 g at 20 Hz
1 g on [40-100] HzLateral = 0.625 g on [5-100] Hz
Acoustic
31.5 Hz = 130.5 dB63 Hz = 133.5 dB125 Hz = 135.5 dB250 Hz = 135.7 dB
500 Hz = 130.8 dB1000 Hz = 126.4 dB
2000 Hz = 120.3 dB
Shock100 Hz = 50 g700 Hz = 800 g1000 Hz 1500 Hz = 2000 g
4000 Hz 5000 Hz = 4000 g10000 Hz = 2000 g
-
8/13/2019 Calvi Liege 2010 Students
81/144
Spacecraft Structural Dynamics & Loads - A. Calvi 81
Modal survey test (identification test)Modal survey test (identification test)Purpose : provide data for dynamic mathematical model validationNote: the normal modes are the most appropriate dynamiccharacteristics for the identification of the structure
Usually performed on structural models (SM or STM) in flightrepresentative configurations
Modal parameters (natural frequencies, mode shapes, damping,effective masses) can be determined in two ways:
by a method with appropriation of modes, sometimes called phaseresonance, which consists of successively isolating each mode by anappropriate excitation and measuring its parameters directly
by a method without appropriation of modes, sometimes called phaseseparation, which consists of exciting a group of modes whoseparameters are then determined by processing the measurements
-
8/13/2019 Calvi Liege 2010 Students
82/144
Spacecraft Structural Dynamics & Loads - A. Calvi 82
Different ways to get modal data from testsDifferent ways to get modal data from tests
Hammer test
Vibration test data analysis
Dedicated FRF measurement & modal analysis
Full scale modal survey with mode tuning I n c r e a s i n g e
f f o r t
D a
t a c o n s i s t e n c y
-
8/13/2019 Calvi Liege 2010 Students
83/144
Spacecraft Structural Dynamics & Loads - A. Calvi 83
Modal Survey Test vs. Modal Data extracted from the Sine Vibration Test
Modal Survey :
requires more effort (financial and time)
provides results with higher quality
Modal Data from Sine Vibration :
easy access / no additional test necessary
less quality due to negative effects from vibration
fixtures / facility tables not indefinitely stiff
higher sweep rate (brings along effects like beating or control instabilities)
-
8/13/2019 Calvi Liege 2010 Students
84/144
Spacecraft Structural Dynamics & Loads - A. Calvi 84
Ariane 5 Ariane 5 -- Sine excitation at spacecraft base (Sine excitation at spacecraft base (sine-equivalent dynamics)
-
8/13/2019 Calvi Liege 2010 Students
85/144
Spacecraft Structural Dynamics & Loads - A. Calvi 85
Sine vibration for different launchers (longitudinal)Sine vibration for different launchers (longitudinal)
-
8/13/2019 Calvi Liege 2010 Students
86/144
Spacecraft Structural Dynamics & Loads - A. Calvi 86
Sine vibration for different launchers (lateral)Sine vibration for different launchers (lateral)
-
8/13/2019 Calvi Liege 2010 Students
87/144
-
8/13/2019 Calvi Liege 2010 Students
88/144
Spacecraft Structural Dynamics & Loads - A. Calvi 88
Herschel on Hydra
( b )( b )
-
8/13/2019 Calvi Liege 2010 Students
89/144
Spacecraft Structural Dynamics & Loads - A. Calvi 89
Acoustic test (objectives) Acoustic test (objectives)
Demonstrate the ability of a specimen towithstand the acoustic environment duringlaunch
Validation of analytical models
System level tests verify equipmentqualification loads
Acceptance test for S/C flight models
-
8/13/2019 Calvi Liege 2010 Students
90/144
Spacecraft Structural Dynamics & Loads - A. Calvi 90
Ariane 5 Ariane 5 Acoustic noise spectrum under the fairing Acoustic noise spectrum under the fairing
A i f diff l hA i f diff l h
-
8/13/2019 Calvi Liege 2010 Students
91/144
Spacecraft Structural Dynamics & Loads - A. Calvi 91
Acoustic spectra for different launchers Acoustic spectra for different launchers
Sh k t t Obj ti d kSh k t t Obj ti d k
-
8/13/2019 Calvi Liege 2010 Students
92/144
Spacecraft Structural Dynamics & Loads - A. Calvi 92
Shock test. Objectives and remarksShock test. Objectives and remarks
Demonstrate the ability of a specimen towithstand the shock loads during launchand operation
Verify equipment qualification loadsduring system level tests
System level shock tests are generallyperformed with the actual shockgenerating equipment (e.g. clamp bandrelease)
or by using of a sophisticated pyro- shock generating system (SHOGUN for
ARIANE 5 payloads)
Shock response spectra for different launchers (spacecraft separShock response spectra for different launchers (spacecraft separ ation)ation)
-
8/13/2019 Calvi Liege 2010 Students
93/144
Spacecraft Structural Dynamics & Loads - A. Calvi 93
Shock response spectra for different launchers (spacecraft separ Shock response spectra for different launchers (spacecraft separ ation)ation)
Note: for a consistentcomparison, data
should refer tothe same adapter.
Shock machine (metalShock machine (metal metal pendulum impact machine)metal pendulum impact machine)
-
8/13/2019 Calvi Liege 2010 Students
94/144
Spacecraft Structural Dynamics & Loads - A. Calvi 94
Shock machine (metalShock machine (metal --metal pendulum impact machine)metal pendulum impact machine)
-
8/13/2019 Calvi Liege 2010 Students
95/144
Random vibration test with slide tableRandom vibration test with slide table
-
8/13/2019 Calvi Liege 2010 Students
96/144
Spacecraft Structural Dynamics & Loads - A. Calvi 96
Random vibration test with slide tableRandom vibration test with slide table
Random vibration test: dataRandom vibration test: data processing bandwidthprocessing bandwidth
-
8/13/2019 Calvi Liege 2010 Students
97/144
Spacecraft Structural Dynamics & Loads - A. Calvi 97
Random vibration test: dataRandom vibration test: data --processing bandwidthprocessing bandwidth
The figures show how the data-processing bandwidth can affect acalculated power spectral density. Whether a PSD satisfies criteriafor level and tolerance depends on the frequency bandwidth usedto process the measured acceleration time history.
Spacecraft Structural Dynamics & Loads (2)
-
8/13/2019 Calvi Liege 2010 Students
98/144
Spacecraft Structural Dynamics & Loads - A. Calvi 98
Spacecraft Structural Dynamics & Loads (2)
Payload-launcher Coupled Loads Analysis (CLA)
Mechanical tests
Modal survey test
Sinusoidal vibration test
Acoustic noise test
Shock test
Random vibration test
Overtesting, notching
and sine vibration testing
Mathematical model updating and validation
Summary and conclusive remarks
Bibliography
Overtesting:Overtesting:di d i
-
8/13/2019 Calvi Liege 2010 Students
99/144
Spacecraft Structural Dynamics & Loads - A. Calvi 99
Overtesting:Overtesting:an introductionan introduction(vibration absorber effect)(vibration absorber effect)
Introduction to overtesting and notching
-
8/13/2019 Calvi Liege 2010 Students
100/144
Spacecraft Structural Dynamics & Loads - A. Calvi 100
Introduction to overtesting and notching
The qualification of the satellite to lowfrequency transient is normallyachieved by a base-shake test
The input spectrum specifies theacceleration input that should excitethe satellite, for each axis
This input is definitively different fromthe mission loads, which are transient
Notching: Reduction of accelerationinput spectrum in narrow frequencybands, usually where test item hasresonances
(NASA-HDBK-7004)
-
8/13/2019 Calvi Liege 2010 Students
101/144
Spacecraft Structural Dynamics & Loads - A. Calvi 101
GOCE on ESTEC Large Slip Table Herschel on ESTEC Large Slip Table
-
8/13/2019 Calvi Liege 2010 Students
102/144
Shock Response Spectrum and Equivalent Sine Input
-
8/13/2019 Calvi Liege 2010 Students
103/144
Spacecraft Structural Dynamics & Loads - A. Calvi 103
S oc espo se Spect u a d qu va e t S e put
A shock response spectrum is aplot of maximum response
(e.g.
displacement, stress, acceleration)of single degree-of-freedom(SDOF) systems to a given input
versus some system parameter,generally the undamped naturalfrequency.
-
8/13/2019 Calvi Liege 2010 Students
104/144
ESI for Spacecraft
-
8/13/2019 Calvi Liege 2010 Students
105/144
Spacecraft Structural Dynamics & Loads - A. Calvi 105
CLA (Coupled Load Analysis)
0 10 20 30 40 50 60 70 800
50
100
150
200
250
S R S [ m
/ s 2 ]
frequency [Hz]
SRSQ
SRS ESI
ESI
12
Q
SRS ESI
SRS
Difference is negligible for small damping ratios
/[ sm 2.412DOF2DOF
]/[ sm01.0
Hz23frequencynatural
SDOFSDOF
-
8/13/2019 Calvi Liege 2010 Students
106/144
Spacecraft Structural Dynamics & Loads - A. Calvi 106
][ s
[ Hz
]/[ sm 2.46
][s
[ Hz
Hz23frequencynatural 1.97 01.0
]/[ sm
Transient responseTransient response
Frequency response at ESI levelFrequency response at ESI level
The effects of the sine sweep rate on the structural response
-
8/13/2019 Calvi Liege 2010 Students
107/144
Spacecraft Structural Dynamics & Loads - A. Calvi 107
The acceleration enforced by the shakeris a swept frequency function
The sweep is amplitude modulated
Acceleration transient response can besignificantly lower that the steady-statefrequency response
2 oct/min
0 10 20 30 40 50 60-5
0
5
time [s]
a c c e
l a r a
t i o n
[ m / s
2 ] 4 oct/min
Effect of sweep rateEffect of sweep rate on isolated peak for increasing and decreasingon isolated peak for increasing and decreasing
frequency sweepsfrequency sweeps
-
8/13/2019 Calvi Liege 2010 Students
108/144
Spacecraft Structural Dynamics & Loads - A. Calvi 108
The sweep rate V
has 3effects:
a variation (sign of V ) ofthe frequency of the peak: f
A decrease of the peakamplitude: A
An increase of the peakwidth (with loss ofsymmetry):
-
8/13/2019 Calvi Liege 2010 Students
109/144
Spacecraft Structural Dynamics & Loads - A. Calvi 109
Notching
ESI(equivalent sine)
-
8/13/2019 Calvi Liege 2010 Students
110/144
Spacecraft Structural Dynamics & Loads - A. Calvi 110
-
8/13/2019 Calvi Liege 2010 Students
111/144
Spacecraft Structural Dynamics & Loads - A. Calvi 111
SineSine --burst load testburst load test The sine burst test is used to apply a quasi static load to a test item in
-
8/13/2019 Calvi Liege 2010 Students
112/144
Spacecraft Structural Dynamics & Loads - A. Calvi 112
The sine-burst test is used to apply a quasi-static load to a test item inorder to strength qualify the item and its design for flight
A secondary objective is to minimize potential fatigue damage to
thetest item
For components and subsystems, the fixture used for vibration testingoften can also be used for sine-burst strength testing. For this reason,
strength qualification and random vibration qualification can often beperformed during the same test session which saves time and money
Since the test is intended to impart a quasi-static load to the test item,the test frequency must be
(in principle) below the fundamental
resonant frequency of the test item
The sine-burst test is a cost effective alternative to either static loadsor to centrifuge testing
(Sine) quasi static load test(Sine) quasi static load test (sine burst)(sine burst)
-
8/13/2019 Calvi Liege 2010 Students
113/144
Spacecraft Structural Dynamics & Loads - A. Calvi 113
Random vibration test: notching of test specificationRandom vibration test: notching of test specification
-
8/13/2019 Calvi Liege 2010 Students
114/144
Spacecraft Structural Dynamics & Loads - A. Calvi 114
Illustration of notching of random vibration test specification,at the frequencies of strong test item resonances
Spacecraft Structural Dynamics & Loads (2) Payload launcher Coupled Loads Analysis (CLA)
-
8/13/2019 Calvi Liege 2010 Students
115/144
Spacecraft Structural Dynamics & Loads - A. Calvi 115
Payload-launcher Coupled Loads Analysis (CLA)
Mechanical tests
Modal survey test
Sinusoidal vibration test
Acoustic noise test
Shock test
Random vibration test
Overtesting, notching
and sine vibration testing
Mathematical model updating and validation
Summary and conclusive remarks
Bibliography
Validation of Finite Element Models
(with emphasis on Structural Dynamics)
-
8/13/2019 Calvi Liege 2010 Students
116/144
Spacecraft Structural Dynamics & Loads - A. Calvi 116
Everyone believes the test data except for theexperimentalist, and no one believes the finiteelement model except for the analyst
All models are wrong, but some are still useful
Verification and Validation Definitions(ASME Standards Committee: V & V in Computational Solid Mechanics )
V ifi i ( f d l l i ) P f d i i h
-
8/13/2019 Calvi Liege 2010 Students
117/144
Spacecraft Structural Dynamics & Loads - A. Calvi 117
Verification (of codes, calculations) : Process of determining that amodel implementation accurately represents the developers
conceptual description of the model and the solution to the model
Math issue: Solving the equations right
Validation: Process of determining the degree to which a model isan accurate representation of the real world from the perspective ofthe intended uses of the model
Physics issue: Solving the right equations
Note: objective of the validation is to maximise confidence inthe predictive capability of the model
Terminology: Correlation, Updating and Validation Correlation:
-
8/13/2019 Calvi Liege 2010 Students
118/144
Spacecraft Structural Dynamics & Loads - A. Calvi 118
the process of quantifying the degree of similarity and dissimilarity betweentwo models (e.g. FE analysis vs. test)
Error Localization:
the process of determining which areas of the model need to be modified
Updating:
mathematical model improvement using data obtained from an associatedexperimental model (it can be consistent or inconsistent )
Valid model :
model which predicts the required dynamic behaviour of the subjectstructure with an acceptable degree of accuracy, or correctness
Some remarks on the validation of critical analyses
Loads analysis is probably the single most influential task in
-
8/13/2019 Calvi Liege 2010 Students
119/144
Spacecraft Structural Dynamics & Loads - A. Calvi 119
y p y g
designing a space structure
Loads analysis is doubly important because it is the basis for statictest loads as well as the basis for identifying the target responsesand notching criteria
in sine tests
A single mistake in the loads analysis can mean that we design andtest the structure to the wrong loads
We must be very confident in our loads analysis, which means wemust check the sensitivity of our assumptions and validate theloads analysis that will be the basis of strength analysis and statictesting
Note: Vibro-acoustic, random and shock analyses are usually notcritical
in the sense that we normally use environmental tests to
verify mechanical requirements
ASME V&V Guide ASME V&V Guide vs.vs. Validation of FEM for CLAValidation of FEM for CLA
Reality of interest : satellite / low frequency transient environment
-
8/13/2019 Calvi Liege 2010 Students
120/144
Spacecraft Structural Dynamics & Loads - A. Calvi 120
y q y
Intended use of the model : launcher/satelli te CLA (to predict systembehaviour for cases that will not be tested)
Response features of interest : CLA loads (forces, accelerations, etc.)
Validation testing : modal survey test or base-drive sine test
Experimental data : accelerations (and forces) (time histories)
Experimental features of interest : natural frequencies, mode shapes
Metrics : relative errors (e.g. natural frequencies), MAC, etc.
Accuracy requirements : e.g. ECSS-E-ST-32-11
Computational model : NASTRAN F.E. model (eigenmodes analysis)
Validation documentation : ECSS-E-ST-32 (DRD Test/analysiscorrelation)
Targets of the correlation(features of interest for quantitative comparison)
-
8/13/2019 Calvi Liege 2010 Students
121/144
Spacecraft Structural Dynamics & Loads - A. Calvi 121
Characteristics that most affect the structure response to applied forces
Natural frequencies
Mode shapes
Modal effective masses
Modal damping
Total mass, mass distribution
Centre of Gravity, inertia
Static stiffness
Interface forces
Correlation of mode shapesCorrelation of mode shapes
-
8/13/2019 Calvi Liege 2010 Students
122/144
Spacecraft Structural Dynamics & Loads - A. Calvi 122
Spacehab FEM coupled to the test rig model & Silhouette
GOCE modal analysis and survey testGOCE modal analysis and survey test
-
8/13/2019 Calvi Liege 2010 Students
123/144
Spacecraft Structural Dynamics & Loads - A. Calvi 123
CrossCross --Orthogonality Check (COC) and Modal Assurance Criterion (MAC)Orthogonality Check (COC) and Modal Assurance Criterion (MAC)
The cross-orthogonality between the analysis and test mode
-
8/13/2019 Calvi Liege 2010 Students
124/144
Spacecraft Structural Dynamics & Loads - A. Calvi 124
shapes with respect to the mass matrix is given by:
The MAC between a measured mode and an analytical mode isdefined as:
aT m MC
as
T
asmr
T
mr
asT mr
rs MAC
2
Note: COC and MAC do not give a useful measure of the error!
Columbus: Cross-Orthogonality Check up to 35 Hz (target modes) TEST
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
FEM Err.% [Hz] 13.78 15.80 17.20 23.81 24.23 24.65 25.36 25.59 26.59 27.19 27.53 28.87 30.19 30.55 32.73 33.15 33.86 34.57 35.21 36.16
-
8/13/2019 Calvi Liege 2010 Students
125/144
Spacecraft Structural Dynamics & Loads - A. Calvi 125
1 -2.94 13.37 1.00
2 -0.95 15.65 1.00
3 -1.73 16.90 0.99
4 -3.26 23.03 0.93 0.35
5 -1.16 23.95 0.34 0.93
6 -2.00 24.16 0.95
7 -1.98 24.86 0.95 0.27
8 -0.12 25.56 0.86
9 -0.95 26.34 0.22 0.90
10 -2.65 26.47 0.95
11 -0.40 27.42 0.26 0.96
12 -3.65 27.82 0.82 0.27
13 -6.00 28.38 0.46 0.89
15 1.19 30.91 0.26 0.95
17 1.63 33.26 0.94 0.21 0.32
18 - 33.71 0.64 0.34 0.62
19 -4.72 34.45 0.95
20 1.21 34.99 0.57 0.81
-
8/13/2019 Calvi Liege 2010 Students
126/144
Spacecraft Structural Dynamics & Loads - A. Calvi 126
MPLM Modal CorrelationMPLM Modal Correlation
MPLMMPLMModalModalEffectiveEffective
-
8/13/2019 Calvi Liege 2010 Students
127/144
Spacecraft Structural Dynamics & Loads - A. Calvi 127
EffectiveEffectiveMassesMasses(Final Correlation)(Final Correlation)
Soho SVM Cross-Orthogonality CheckF.E.M.
1 2 3 4 8 9 10 15 21 26 29 30 31TEST Err. % Freq. Hz 34.83 37.24 44.07 45.19 51.51 52.68 55.46 62.18 70.92 77.99 81.53 82.25 84.42
-
8/13/2019 Calvi Liege 2010 Students
128/144
Spacecraft Structural Dynamics & Loads - A. Calvi 128
1 2.87 35.86 0.87 0.462 0.00 37.24 0.47 0.873 4.17 45.99 0.874 4.78 47.46 0.77 0.24
5 -3.39 49.82 -0.33 0.766 0.96 53.19 0.75 0.227 2.10 56.65 0.79 0.218 58.67 -0.28 -0.35 -0.229 60.24 -0.30 -0.22 0.46 0.46
10 3.30 64.30 0.6111 66.40 0.21 0.3212 67.50 0.45 -0.43
13 68.73 -0.38
14 69.6815 71.6916 72.71 0.37 -0.3317 3.30 73.34 0.21 0.8518 74.7819 75.6320 78.77 -0.2421 82.1222 7.72 84.51 0.7623 5.54 86.31 0.87 -0.2324 7.21 88.64 0.6425 5.45 89.29 -0.33 0.6326 94.4427 97.1528 99.56
GOCEGOCE -- MAC and Effective MassMAC and Effective Mass
-
8/13/2019 Calvi Liege 2010 Students
129/144
Spacecraft Structural Dynamics & Loads - A. Calvi 129
Aeolus STM: comparison of transfer functions Aeolus STM: comparison of transfer functions
-
8/13/2019 Calvi Liege 2010 Students
130/144
Spacecraft Structural Dynamics & Loads - A. Calvi 130
Sine test response , FEM predicted response and post-test (updated FEM) response
Lack of Matching between F.E. Model and Test
Modelling uncertainties and errors (model is not completely physicallyrepresentative)
-
8/13/2019 Calvi Liege 2010 Students
131/144
Spacecraft Structural Dynamics & Loads - A. Calvi 131
representative)
Approximation of boundary conditions
Inadequate modelling of joints and couplings
Lack or inappropriate damping
representation
The linear assumption of the model versus test non-linearities
Mistakes
(input errors, oversights, etc.)
Scatter in manufacturing
Uncertainties in physical properties
(geometry, tolerances, material properties)
Uncertainties and errors in testing
Measured data
or parameters contain levels of errors
Uncertainties in the test set-up, input loads, boundary conditions
etc.
Mistakes
(oversights, cabling errors, etc.)
Test-Analysis Correlation CriteriaThe degree of similarity or dissimilarity establishing that the correlationbetween measured and predicted values is acceptable
-
8/13/2019 Calvi Liege 2010 Students
132/144
Spacecraft Structural Dynamics & Loads - A. Calvi 132
ECSS-E-ST-32-11 Proposed Test / Analysis Correlation Criteria
Model Updating Using Design Sensitivity and Optimisation(traditional basic assumptions)
-
8/13/2019 Calvi Liege 2010 Students
133/144
Spacecraft Structural Dynamics & Loads - A. Calvi 133
Some appropriate objective functions, within design optimisationcodes, can be used to drive a F.E.M. to behave in the samemanner as the real structure portrayed by a set of numerical test
results
The test results accurately depict the true behaviour of the structure
-
8/13/2019 Calvi Liege 2010 Students
134/144
Model updating: example of objective functionsNatural Frequency:
e
eaf f
f f d
-
8/13/2019 Calvi Liege 2010 Students
135/144
Spacecraft Structural Dynamics & Loads - A. Calvi 135
Mode Shape:e
ead ( : modal scale factor)
Effective Transmissibilities:e
eaT~
T~
T~
T~
d
Effective Masses: eea
M~
M~
M~
M~
d
Objective Function
bW bF T with
M~T~
f
d
d
d d
b and
M~
T~
f
w
w
ww
W
Location of Modelling Errors and Selection of the Design
Variables Based on Sensitivity AnalysisCriterion (selection of design variables):
-
8/13/2019 Calvi Liege 2010 Students
136/144
Spacecraft Structural Dynamics & Loads - A. Calvi 136
The design variables should be selected for those elements or element groupswhich have an influence on the eigenfrequencies
and mode shapes which are
targeted during the correlation/updating process (in addition to analysts knowledgeof uncertain modelled regions of the structure and/or results of
other error
localisation analyses)
Two basic approaches are poss ible:
Initial model sensitivities (e.g. initial derivative approach)
A Posteriori
Approach (at the end of a preliminary optimisation process)
Criterion (error localisation):
To determine how effective certain physical properties changes might be in reducingthe difference between measured and calculated data (however high sensitivity isnot generally a sufficient reason for the selection of a candidate parameter!)
Limitations of the sensitivity and optimisation approach
Largest changes can be in the most sensitive parameters ratherthan those in error ( inconsistent updating and misleading error
-
8/13/2019 Calvi Liege 2010 Students
137/144
Spacecraft Structural Dynamics & Loads - A. Calvi 137
localization)
Errors of insensitive regions cannot be detected
The success of the updating procedure can strongly depend on the selection of the design parameters to be updated (it could be
necessary to consider several sets of design parameters to detecterroneous regions of the structure)
The approach could be short-sighted
(possible convergence tolocal minima)
ExerciseExercise
Calculate natural frequencies and modeshapes of the 2-DOF satellite represented
-
8/13/2019 Calvi Liege 2010 Students
138/144
Spacecraft Structural Dynamics & Loads - A. Calvi 138
p pin the figure
Consider a perturbed model, representingthe real (tested) structure, having
k1
= 60 E5 N/m
k
2
= 130 E5 N/m
Calculate natural frequencies and modeshapes for the perturbed (test) model
Correlate the 2 models, i.e. calculate:
Natural frequency deviations
Mode shapes cross-orthogonality
check
Solutions of the exerciseSolutions of the exercise
-
8/13/2019 Calvi Liege 2010 Students
139/144
Spacecraft Structural Dynamics & Loads - A. Calvi 139
-
8/13/2019 Calvi Liege 2010 Students
140/144
Summary and Conclusive Remarks
The role of structural dynamics in a space project
-
8/13/2019 Calvi Liege 2010 Students
141/144
Spacecraft Structural Dynamics & Loads - A. Calvi 141
The role of structural dynamics in a space project
Dynamic analysis types
The effective mass concept
Design load cycles & verification loads cycle
Payload-launcher Coupled Loads Analysis
Mechanical tests
Overtesting
& notching
Mathematical model updating and validation
BibliographyBibliography
Sarafin T.P. Spacecraft Structures and Mechanisms, Kluwer, 1995 Craig R.R., Structural Dynamics An introduction to computer methods, J. Wiley
-
8/13/2019 Calvi Liege 2010 Students
142/144
Spacecraft Structural Dynamics & Loads - A. Calvi 142
and Sons, 1981
Clough R.W., Penzien
J., Dynamics of Structures, McGraw-Hill, 1993
Ewins
D.J., Modal Testing
Theory, practice and applications, Research Studies
Press, Second Edition, 2000
Wijker
J., Mechanical Vibrations in Spacecraft Design, Springer, 2004
Girard A., Roy N., Structural Dynamics in Industry, J. Wiley and
Sons, 2008
Steinberg D.S., Vibration Analysis for Electronic Equipment, J. Wiley and Sons,2000
Friswell
M.I., Mottershead
J.E., Finite Element Model Updating in StructuralDynamics, Kluwer
1995
Ariane
5 Users Manual, Arianespace, http://www.arianespace.com/
BibliographyBibliography -- ECSS DocumentsECSS Documents
ECSS-E-ST-32 Space Project Engineering - Structural
-
8/13/2019 Calvi Liege 2010 Students
143/144
Spacecraft Structural Dynamics & Loads - A. Calvi 143
ECSS-E-ST-32-03 Structural finite element models
ECSS-E-ST-32-10 Structural factors of safety for spaceflight
hardware
ECSS-E-ST-32-02 Structural design and verification of
pressurized hardware
ECSS-E-ST-32-11 Modal survey assessment
ECSS-E-ST-32-01 Fracture control
ECSS-E-10-02 Space Engineering - Verification
ECSS-E-10-03 Space Engineering - Test ing
THE END!
-
8/13/2019 Calvi Liege 2010 Students
144/144
Spacecraft Structural Dynamics & Loads - A. Calvi 144
Acknowledgements:
ALENIA SPAZIO, Italy, for the data concerning the projects GOCE, COLUMBUS,MPLM and SOHO
EADS ASTRIUM, UK, for the data concerning the project AEOLUS andEarthCARE
ESA/ESTEC, Structures Section, NL, for the data concerning ARIANE 5 FEmodel and LV/SC CLA