You are on page 1of 6

20 Coupled Dynamic Loads Analysis

20.1 Introduction

The launch authority generally carries out a coupled dynamic load analysis
(CDLA) also called the coupled load analysis (CLA). The coupled loads analysis is
carried out in order to calculate the response behaviour and the dynamic loads
occurring in the spacecraft during launch. The design loads that are mentioned in
the user manual are obviously very general. The dynamic interaction between the
launch vehicle and the spacecraft during launch is calculated with the coupled anal-
ysis. The results are also used to avoid the satellite structure from being over-tested
during the qualification. The dynamic system of the shaker table and the spacecraft
is not the same as the launch vehicle and the spacecraft.
A complete overview for the ESA/ESTEC ARIANE 5 CDLA is given in
[Fransen 2006] and ARIANESPACE general specifications for payload dynamic
models is given in [Boland 2001]. The modal reduction shall be based on the Craig
and Bampton condensation [Craig 1968].
The launch authority (ARIANESPACE, NASA or the European space organiza-
tion ESA) is in charge of the mathematical model and the dynamic launch loads of
the launch vehicle that act on the mathematical model. The primary contractor pro-
vides the launch authority with the mathematical model of the spacecraft. The anal-
ysis stream diagram is illustrated in Fig. 20.1.
With the aid of Fig. 20.2, [Kabe 1995], it is possible to gain more insight into
the size of the complete and reduced dynamic models, as well as the relative effort
that is needed to carry out a coupled dynamic loads analysis.
Critical load-producing events during the launch phase [DiMaggio 2001]:
• Lift-off
• Atmospheric flight (gust, buffet, autopilot-induced, etc.)
• Engine ignition and shutdowns
• Staging and separation events

313
314 20 Coupled Dynamic Loads Analysis

Spacecraft
Dynamic
Model Coupled Coupled Spacecraft &
System System Launch
Model Analysis Vehicle
Launch Vehicle Responses
Dynamic
Model
Forcing
Functions

Fig. 20.1 CLA process (courtesy Quartus Engineering, USA)

Fig. 20.2 Loads analysis flow (effort breakdown), [Kabe 1998]

The mathematical models of the launch vehicle as well as the spacecraft are
converted to a so-called Craig-Bampton (CB) model. This is also done for the loads
that act on the launch vehicle. Modal synthesis methods are applied to couple the
CB reduced models and afterwards transient analysis are carried out for various
phases of the launch. Results of the calculations are:
• Max/min values in time
– interface (I/F) forces
– accelerations for the selected degrees of freedom and in the interface
– displacements for the selected degrees of freedom and in the interface
20.2 Finite Element Validation 315

• Time-history plots
– I/F forces
– I/F accelerations
– I/F displacements
• Max/min equivalent accelerations (load factors) and time histories
• Shock spectra plots of the I/F accelerations and max/min values. The I/F shock
spectrum is the sinusoidal input for the spacecraft at the base.

20.2 Finite Element Validation

The finite element model used for the coupled load analysis must be test-verified
and must satisfy the correlation requirements with the test results obtained from the
modal survey test. The recommended correlation criteria for the analytical predic-
tion and the test results include frequency and mode shape comparison [Chung
2002, NASA 1996, Ricks 1991, Fransen 2006].
• The natural frequency variation is required to be less than ± 5 %
• The cross orthogonality matrix, [XOR], of the analytical mode shapes [ Φ a ] and
the test modes [ Φ t ] with respect to the analytical mass matrix [ M a ] of the test
analysis model (TAM) is determined as follows
T T
[ XOC ] = [ Φ a ] [ M a ] [ Φ t ] with [ Φ a ] [ M a ] [ Φ a ] = 〈 I〉 and the auto-correla-
T
tion check of the test modes as defined by [ AOC ] = [ Φ t ] [ M a ] [ Φ t ] with
unity terms on the diagonal and the off-diagonal terms less than 0.1.
A XOR value close to one indicates a high degree of correlation or consistency
between two modes. A generally accepted requirement for the cross orthogonality
correlation matrix is to have all diagonal terms larger than 0.9 and all the off-diago-
nal terms less than 0.2, [Ricks 1991].

Example
Two 2-DOFs systems need to be correlated with each other. The first system
represents the analytical system and the second one the tested system. The system
is illustrated in Fig. 20.3. We will apply the correlation specification as stated in
[Ricks 1991].
316 20 Coupled Dynamic Loads Analysis

m1

k1
m2

k2

Fig. 20.3 2 DOFs dynamic system.

The characteristics of the dynamic systems are given in Table 20.1.

Table 20.1 Characteristics of dynamic system


m1 m2 k1 k2
Model (kg) (kg) (N/m) (N/m)
Analytical 1.0 1.0 10000 10000
Test 1.2 0.95 9500 11000

The intermediate results are given in Table 20.2.

Table 20.2 Intermediate results


Natural frequencies
Model { f } (Hz) Mode Shapes [ Φ ]
Analytical
⎧ ⎫
{ f a } = ⎨ 9.8363 ⎬ [ Φ a ] = – 0.8507 – 0.5257
⎩ 25.7518 ⎭ – 0.5257 0.8507

Test
⎧ ⎫
{ f t } = ⎨ 9.5099 ⎬ [ Φ t ] = – 0.8185 – 0.4891
⎩ 26.6358 ⎭ – 0.5263 0.8807

The auto orthogonality of the test modes with respect to the analytical mass
matrix becomes
T
[ AOC ] = Φ a [ M a ] [ Φ a ] = 1 0 ,
0 1

T 0.9469 – 0.0632 .
[ AOC ] = Φ t [ M a ] [ Φ t ] =
– 0.0632 1.0149
T
The test modes are scaled such that the diagonal terms of Φ t [ M a ] [ Φ t ] become
unity. Hence the scaled test mode shape becomes
20.2 Finite Element Validation 317

[ Φ t ] = – 0.8411 – 0.4855 ,
– 0.5408 0.8743

and the [ AOC ] becomes

T 1.0000 – 0.0645 .
[ AOC ] = Φ t [ M a ] [ Φ t ] =
– 0.0645 1.0000

The cross orthogonality matrix is

T 0.9998 – 0.0467 .
[ XOC ] = [ Φ a ] [ M a ] [ Φ t ] =
– 0.0178 0.9989

The differences of the natural frequencies between the analytical and test model
is less than 5%, however, for the off-diagonal terms in the [XOC], they are less
than 0.2, [Ricks 1991].
End of example
318 20 Coupled Dynamic Loads Analysis

20.3 Literature

Boland, Ph., Bourgain, M., 2001, General Specification for payload Dynamic Models, AE/DI/S/
ES No 102/01, A5-SG-0-01, Issue 4, Arianespace, France.
Chung, Y.T., Foist, B.L. and Sernaker, M.L., 2002, Validation of Payload Acoustic Model based
on Acoustic Test Results, AIAA-2002-1621.
Craig, R.R., Jr., Bampton, M.C.C., July 1968, Coupling of Substructures for Dynamic Analysis,
AIAA Journal, Vol. 6, No. 7, pages 1313–1319.
DiMaggio, S., Structural Design and Verification of Space and Launch Vehicles, The Aerospace
Corporation, Presentation at Columbia University New York, NY, April 29, 2002,
www.civi.columbia.edu/CEE1201/02lect26SD.pdf
Kabe, A.M., 1998, Design and Verification of Launch and Space vehicle Structures, AIAA-98-
1718, pages 175–189.
Fransen, S. H., J., A., 2006, Methodologies for Launcher-Payload Coupled Dynamic Analysis,
ISBN 90-9020293-5, PhD work Delft University of Technology, European Space Agency.
NASA, 1996, Load Analysis of Spacecraft and Payloads, NASA-STD-5002, June 21.
Ricks, E.G., 1991, Guidelines for Loads Analysis and Dynamic Model Verification of Shuttle
Cargo Elements, MSFC-HDBK-1974, October 15.

20.4 Exercises

20.4.1 Internet search

Find on the internet publications about coupled dynamic load analysis?

You might also like