Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

2013-01-2142
Published 09/17/2013
Copyright © 2013 SAE International
doi:10.4271/2013-01-2142
saeaero.saejournals.org

Monitor Points Method for Loads Recovery in Static/Dynamic

Aeroelasticity Analysis with Hybrid Airframe Representation
Mostafa S.A. El Sayed, Miguel Alejandro Gutierrez Contreras, and Nicholas Stathopoulos
Bombardier Aerospace

ABSTRACT
With the high design/performance requirements in modern aircrafts, the need for a flexible airframe structural modeling
strategy during the different phases of the airframe development process becomes a paramount. Hybrid structural modeling
is a technique that is used for aircraft structural representation in which several Finite Element Modeling concepts are
employed to model different parts of the airframe. Among others, the Direct Matrix Input at a Grid-Point (DMIG)
approach has shown superiority in developing high fidelity, yet, simplified Finite Element Models (FEM's). While the
deformation approach is a common choice for loads recovery in structures represented by stick models, using structural
models simulated by the DMIG representation requires the adoption of a different approach for loads recovery
applications, namely, the momentum approach.
In this paper, the Monitor Points (MP) Method is introduced as an efficient methodology for loads recovery in static
and dynamic Aeroelasticity analysis with hybrid airframe representation. MP method is a function provided in MSC
NASTRAN that hinges on the momentum approach for loads recovery as it enables the superposition of applied loads at a
user defined point and transformed into a user defined coordinate system. Here, the hybrid model is used to generate
accurate predictions of the aircraft structural kinematics in flight which by its turn generates accurate profiles for the
encountered aerodynamic and inertia loads. The MP method is then employed to generate high fidelity distributed loads
necessary to predict the different critical load cases that generate the aircraft's loads envelop.
A sensitivity analysis is conducted which showed the high convergence of the presented methodology as it is less
sensitive to modal truncation errors compared to loads recovery methods that hinge on the deformation approach.

CITATION: El Sayed, M., Gutierrez Contreras, M., and Stathopoulos, N., "Monitor Points Method for Loads Recovery in
Static/Dynamic Aeroelasticity Analysis with Hybrid Airframe Representation," SAE Int. J. Aerosp. 6(2):2013, doi:
10.4271/2013-01-2142.
____________________________________

INTRODUCTION different analyses conducted in this paper. Different levels of

The process of static/dynamic aeroelasticity analysis is structural modeling are shown in Fig. 1. A 3D FEM (Fig
multidisciplinary in nature [1]. It is an iterative process (1b)) is generated once the aircraft's layout and structural
between the structural and the aerodynamic disciplines. details are obtained. 3D FEM is commonly used in structural
Accurate predictions of the aircraft structural deformations design validation and optimization. Imposing a complex 3D
during flight have a significant effect on its aerodynamic FEM in a multidisciplinary iterative process, e.g.
performance and the fidelity of the recovered aerodynamic aeroelasticity analysis, is quite expensive. Alternatively,
loads. simplified lower fidelity FEM's of the aircraft's structure can
Aerodynamic load envelops are used for the preliminary be generated and employed [2]. The difficulty is to develop
sizing of the aircraft structure. The preliminary dimensions models that are sufficiently simple to be called thousands of
are then used to generate the aircraft structural model. times during the iterative process, but are sophisticated
To demonstrate the concept of the MP method, aircraft enough to accurately predict aircraft structural deformations
with unconventional configuration is considered for the and loads.

399
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

400 El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013)

A matrix reduction technique is implemented in MSC

NASTRAN [10] in what is called the DMIG approach (Fig
(1d)). In the DMIG approach, the reduced mass and stiffness
characteristics are associated to the Degrees of Freedom
(DOF) of the independent nodes (aka “the A-set”) of the
reduced FEM. The A-set nodes are selected to provide the
necessary kinematical details for each specific analysis. Two
techniques can be used to generate the DMIG model. The
first is based on the Guyan static reduction method where the
DMIG nodes carry only the stiffness characteristics of the
sructure. In this case, the mass characteristics are introduced
into the structural model using a lumped masses technique.
The second is the Craig-Bampton dynamic reduction method.
Here, the DMIG nodes carry both the mass and the stiffness
characteristics of the structure.
The increase in world-wide air transport in recent years
for both cargo and passengers increased the demand for the
development of new, large, economically and competitive
civil transport aircrafts. One way to increase competitiveness
Figure 1. Different levels of structural modeling. is to increase the size and payload capacity of aircrafts
beyond the current limits. These limits may only be overcome
by unconventional aircraft configurations. With the expected
Simplified beam FEM's, namely, Stick models (Fig (1c)) advantages of such new configurations, on the other hand,
are commonly used in aircraft industry [2, 3, 4]. The stick new challenges arise requiring adapted engineering solutions
model development process is based on simplified beam which cannot be derived from conventional engineering
theory [5]. The stiffness properties of the 3D FEM are systems. For instance, alternative hybrid structural modeling
extracted and condensed into a set of bar elements extending techniques are adopted. Figures (1e) and (1f) show two
along the principal or the elastic axes of the aircraft structure. arrangements of structural hybrid modeling as the centre
Although different shortcomings are always encountered in portion of the aircraft is presented by either 3D FEM or
handling stick models, including the loss of information in DMIG representations while the rest of the aircraft is
the wing's chord-wise and the fuselage's span-wise directions, represented by a stick-model. Due to the small number of
discontinuities in the distribution of the mass and the stiffness DOF's involved in the model, stick modeling is a preferable
parameters and errors in the model performance once loaded choice for structural representation specially, in iterative
in different load cases, however, the tolerance in the accuracy aeroelasticity analyses solutions. Hybrid modeling, on the
against the reduction in the computation cost redeemed them other hand, enables the preservation of the model simplicity
valid for the analysis of aircrafts with conventional tube and while increasing its fidelity in specific parts of the structure at
wing configurations. In these configurations the bulk which additional structural response details are required.
dimensions of the wing and the fuselage satisfy the In the current paper, the MP method is introduced for
geometrical constraints of the simplified beam theory. loads recovery in static and dynamic Aeroelasticity analyses
Another technique of structural models simplification is that include hybrid airframe representations. The MP method
the matrix reduction methods. Matrix reduction methods are hinges on the force summation concept (momentum
mathematical techniques that are used to reduce the size of approach) [11, 12] for loads recovery. Here, the hybrid model
FEM's by reducing the number of Degrees of Freedom (DOF) is used to either retrieve more details of the structural
involved in the analysis. Different matrix reduction methods response or as a higher fidelity structural modeling strategy
are available in literature including the Guyan reduction [6], that generates accurate predictions of the aircraft structural
the modal decoupling [7] and the Craig-Bampton [8] kinematics. The accurate predictions of the aircraft structural
methods. It is found that the Guyan and the Modal reduction kinematics by its turn generate accurate profiles for the
methods lack the ability of providing information of the encountered aerodynamic and inertia loads. At that point, the
physical distribution of the mass and stiffness parameters of MP method is employed to lump the loads and generate its
the actual structure. A shortcoming that is overcome by distribution along the airframe load axis.
Hashemi-Kia and Toossi [9] who developed a technique to
link the dynamically reduced mass and stiffness matrices to
the structural physical DOF's. On the other hand, the Craig-
Bampton reduction method combines the modal DOF's with
the DOF's of the boundary grids of the structure.
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013) 401

A sensitivity analysis is also conducted which showed the From eqn (2), the slave DOF's can be expressed in terms
fast convergence of the MP method compared to loads of the master DOF's as:
recovery methods that hinge on the deformation approach
[12, 13, 14, 15, 16, 17]. This result is in agreement with other
(3)
studies available in literature [12, 16, 18].
The Guyan reduction transformation equation is then
The paper is organized in six sections. After this given as:
introduction, the fidelity of different levels of structural
modeling is tested in section 2. Section 3 describes the MP
method. The accuracy of loads recovered from different
structural models using the MP method is presented in
section 4. Section 5 on the other hand, presents a sensitivity (4)
analysis to determine the convergence of the MP method by
testing its sensitivity to modal truncation errors. The paper is where T and Ts are the Guyan reduction transformation
concluded in section 6. matrices; I is the identity matrix.
DMIG models generated by the Guyan reduction method
FIDELITY OF DIFFERENT LEVELS are only efficient in static analysis. Involving such models in
OF STRUCTURAL MODELING a dynamic analysis requires other strategies for mass
representation.
In this section, the fidelity of different levels of structural The modal decoupling transformation equation is
modeling is tested by comparing the modal characteristics of expressed as:
four different models for the main airframe of a generic
aircraft. The four models include a 3D FEM, a stick model, a
full DMIG model and a hybrid model, as shown, respectively, (5)
in Figs (1b), (1c,) (1d) and (1f). In the hybrid model, the
where Φ is a modal matrix that is constructed by
center portion of the aircraft is represented by the DMIG
concatenating the Mode shapes of the structure into its
approach and the rest of the airframe is modeled by the
columns; q is the modal DOF's. In this method, the physical
traditional stick modeling method.
DOF's of the structure are reduced by expressing them in
Before proceeding with the comparison, we review in the
terms of the modal DOF's.
next sub-section the main matrix reduction methods available
The Craig-Bampton method combines DOF's of the
in literature.
boundary points of the physical structure with the structural
Matrix Reduction Methods normal modes. The transformation equation of the Craig-
Bampton method can be expressed as:
Three methods of matrix reduction are reviewed, namely,
Guyan, Modal Decoupling and Craig-Bampton methods.
To illustrate the concept of matrix reduction, consider the (6)
equation of motion of a FEM of order N as:
where ΦCB and rCB are, respectively, the Craig-Bampton
transformation matrix and DOF's. Craig-Bampton DOF's are
(1)
a combination of the modal DOF's and the DOF's of the
where M ∈ RN×N and K ∈ RN×N are, respectively, the mass physical boundary points of the structure.
and the stiffness matrices of the FEM, r ∈ RN×1 and R ∈ DMIG models generated by the Craig-Bampton reduction
RN×1 are, respectively, the nodal deformations and applied method are efficient for both static and dynamic types of
forces vectors. Damping terms are ignored. analysis.
In the Guyan reduction method, the displacement vector is
split into master DOF's rm ∈ RNm×1 and slave DOF's rs ∈
Comparison of Modal Characteristics of
RNs×1. The master DOF's are those DOF's associated with the Different Structural Models
A-set nodes of the FEM which are retained after applying the The modal analysis solution in MSC NASTRAN [10],
matrix reduction method. As the Guyan method hinges on a solution 103, is used to recover the modal characteristics of
static reduction technique, then, the static part of equation (1) the four models of the main airframe of the generic aircraft.
can be rewritten as: The natural frequencies of their corresponding flexible modes

(2)
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

402 El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013)

are compared. The 3D FEM is used as a reference model as loads developed in element II of the structure, where element
the percentage error in a natural frequency is computed as: II is located just in-board of section A-A. This implies that
the applied loads to portion I of the wing are equivalent to set
L2 (V2M2T2F2) of internal loads developed in element II.
This is the core idea of the MP method as the distributed
(7) internal loads in an element can be predicted by the
superposition of the applied loads to the proper portion of the
where S, H and FD denote, respectively, stick, hybrid and full
aircraft structure. The superposition of the applied loads is
DMIG models. ωi is the natural frequency of the ith model.
performed by the MONPNT1 function in MSC NASTRAN
Figure (2) shows the percentage error of the natural [10].
frequencies of the stick, the hybrid and the full DMIG models
compared to the corresponding natural frequencies of the 3D
FEM.

Figure 2. Error percentage of the natural frequencies of Figure 3. Schematic Drawing of Loads balance in an
the stick, hybrid and full DMIG models compared to the aircraft wing structure
3D FEM

Applicability of the MP Method for Loads

The Root Mean Square (RMS) value for the errors in the
first 21 flexible modes is computed for each model. It is Recovery Applications
found that the RMS values are 11.6%, 5.7% and 0.3% for the To check the accuracy and the applicability of the MP
stick, the hybrid and the full DMIG models, respectively. Method, distributed loads are recovered from the wings and
From this result, it can be concluded that the DMIG the fuselages of the hybrid and the stick models using the MP
approach involving the Craig-Bampton reduction method is Method (MONPNT1 function in MSC NASTRAN [10]) and
an efficient methodology that generates high fidelity the deformation method (ELFORCE function in MSC
structural models. NASTRAN [10]). The aircraft structural models are
employed in a static and a dynamic aeroelasticity analyses.
MONITOR POINTS METHOD No control surfaces are attached to the wings. The Doublet
In linear elasticity, applied loads to kinematically Lattice Method [19] is adopted for aerodynamic modeling.
determinate structures generate infinitesimal structural Loads recovered by the MP and the ELFORCE methods are
deformations. Those deformations develop internal loads that compared. Three load cases are considered for this analysis:
change with each external load increment until a state of 1. Static aeroelastic load case whereas a rigid aircraft is
static or dynamic equilibrium is achieved. At that point, the considered in a 1-G heave acceleration.
internal loads are in balance with the applied external loads.
Figure (3) shows a sectioned aircraft wing that illustrates 2. Dynamic aeroelastic frequency response load case
such state of equilibrium. Here, the aerodynamic and inertia whereas a flexible aircraft is considered in a vertical gust
loads applied to portion I of the wing are in balance with set excitation. Here, the exciting signal is given by a Von-
L1 (V1M1T1F1) of internal loads at section A-A. Set L1, by Karman Spectrum, represented by a Power Spectral Density
its turn, is in balance with set L2 (V2M2T2F2) of the internal (PSD) function.

3. Dynamic aeroelastic transient response load case

whereas flexible aircraft is considered in a Tuned Discrete
Gust (TDG) excitation.
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013) 403

Static load results in a 1-G trim load case

A static aeroelastic load case of 1-G heave acceleration is
considered for this analysis. The trim conditions included a
Mach number of M = 0.82 and a dynamic pressure of q =
0.99. Concentrated forces are applied forward on the center
of gravity of the two engines of the aircraft to simulate engine
thrust effects. Main load results are obtained from the Stick
Model (SM) and the Hybrid Model (HM) and are plotted in
Figs (4) and (5) for the wings and fuselage, respectively.
From Figs (4) and (5) it can be realized that distributed
loads recovered from the stick model using the methods of
ELFORCE (ELF) and MP conform very well which indicates
the accuracy of the MP method (MPM). On the other hand,
distributed loads recovered from the hybrid model using the
MP method are slightly deviated from those recovered from
the stick model also using the MP method. This deviation is
interpreted due to the deviation in the modal characteristics of
the two models, as previously discussed in section 2.
Incremental load results in PSD vertical gust case
A PSD vertical gust case is chosen for this analysis. The
Von-Karman Spectrum input signal is considered with RMS
gust intensity of 79 ft/sec True Air Speed (TAS); Mach
Figure 4. Comparison of main distributed loads in the number: M=0.7686 and dynamic pressure: q=2.355 psi. The
wings of the aircraft recovered by MP and ELFORCE Loads are recovered from element 1001 (end B) which is the
methods last inboard element in the right wing of the stick model.
Since the center portion of the hybrid model of the aircraft
is modeled as a DMIG, then, the MP method is the only mean
available to recover loads from that portion of the aircraft.
Fig (6) shows the DMIG zone in the hybrid model that
corresponds to element 1001 in the stick model. The dynamic
loads recovered from this element in the hybrid and the stick
models are presented in Figs (7a) and (7b). Loads are
recovered from the hybrid model using only the MP method
and from the stick model using both the MP and the
ELFORCE methods. Accordingly, three set of data are
present in each of figs (7a) and (7b).

Figure 6. Hybrid Model of the aircraft with indication of

Figure 5. Comparison of main distributed loads in the the DMIG zone that corresponds to element 1001 on the
fuselage of the aircraft recovered by MP and ELFORCE stick model
methods
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

404 El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013)

Figure 7. Comparison of main distributed loads in the Figure 8. Comparison of main distributed loads in the
wing of the aircraft recovered by MP and ELFORCE wing of the aircraft recovered by MP and ELFORCE
methods methods

Incremental load results in TDG case From Figs (7) and (8), it can be realized that the two data
Similar to the PSD gust analysis, the response of the stick sets representing loads recovered from the stick model using
and the hybrid models to a TDG input signal is measured. the methods of MP and ELFORCE conform very well. On
Again, loads are recovered from element 1001 (end B). Three the other hand, loads recovered from the hybrid model are
data sets are shown in Figs (8a) and (8b) representing significantly deviated from those recovered from the stick
dynamic loads recovered using the MP and the ELFORCE model. These observations prove the accuracy and the
method applied to the stick model and the MP method applicability of the MP method. The deviations in the loads
applied to the hybrid model as element 1001 is located in the recovered from the stick and the hybrid models, although
DMIG zone of the hybrid model. recovered by the same recovery method, are interpreted due
to differences in the modal characteristics of the two models.
It should be noted that loads recovered in the PSD gust
analysis using the MP and the ELFORCE methods represent
the Frequency Response Function of the structural response
which is given as:
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013) 405

analysis. The MP method is employed to recover loads from

an outboard wing station, as indicated in Fig (9). For
comparison, loads recovered from the four models are plotted
(5) in Figs (9a) and (9b). From Figs (9a) and (9b), it can be
realized that loads recovered from the full DMIG model and
where Φ0(ω) and Φi(ω) are the output Power Spectral the 3D FEM conform very well. On the other hand, loads
Density Function (PSDF) and the input Von-Karman PSD recovered from the hybrid and the stick models are
signal. ω is the frequency. significantly deviated from those recovered from 3D FEM.
These observations prove high fidelity of the DMIG approach
ACCURACY OF LOADS for structural modeling.
RECOVERED FROM DIFFERENT
SENSITIVITY ANALYSIS
STRUCTURAL MODELS USING THE
The sensitivity of the MP and the ELFORCE methods for
MP METHOD modal truncation is tested and compared. For this purpose,
Now, as the accuracy and the applicability of the MP the stick model of the generic aircraft is employed in a
method is verified, it is used to check the accuracy of loads dynamic aeroelasticity PSD vertical gust analysis and the lift
recovered from the different levels of structural models used forces are recovered at a specific wing station using the MP
to represent the generic aircraft, namely, the 3D FEM, the and the ELFORCE methods. The Lanczos method [10] is
stick, the hybrid and the full DMIG models, each model is employed for modal extraction. The analysis is conducted at
employed in a PSD vertical gust dynamic aeroelasticity different frequency ranges that are prescribed for the modal
extraction method. The modal extraction frequency ranges
considered are (0:5) Hz with increment of 5 Hz up to (0:100)
Hz.
Figures (10), (11) and (12) show the lift forces at a
specified wing station recovered using the MP and the
ELFORCE methods at modal extraction frequency ranges of,
respectively, (0:5) Hz, (0:40) Hz and (0:65) Hz.

Figure 10. Comparison of lift load in the wing of the

aircraft recovered by the MP and the ELFORCE
methods with frequency range of (0-5 Hz) for modal
extraction

Figure 9. Comparison of main distributed loads in the

wing of the aircraft recovered by MP method
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

406 El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013)

Figure 13. Comparison of lift load in the wing of the

aircraft recovered by the MP and the ELFORCE
methods with different frequency ranges for modal
extraction

Figure 11. Comparison of lift load in the wing of the

aircraft recovered by the MP and the ELFORCE
methods with frequency range of (0-40 Hz) for modal
extraction

Figure 14. Comparison of error in RMS value of PSDF

lift load in the wing of the aircraft recovered by the MP
and the ELFORCE methods with different frequency
ranges for modal extraction

From Fig (14) it can be realized that loads recovered by

the MP method converges at modal extraction frequency
range of (0:40) Hz while the loads recovered by the
ELFORCE method converges at modal extraction frequency
range of (0:65) Hz. This result indicates that the convergence
rate of the MP method is 162.5% higher than the ELFORCE
method. This result can also be noted from Fig (15) as the
error in the RMS value of the load converges to 0% at
frequency ranges of (0:40) Hz and (0:65) Hz for loads
Figure 12. Comparison of lift load in the wing of the recovered by, respectively, the MP and the ELFORCE
aircraft recovered by the MP and the ELFORCE methods.
methods with frequency range of (0-65 Hz) for modal
extraction CONCLUSION
In this paper, a new methodology for static and dynamic
The RMS of the PSDF output loads are determined for the loads recovery is presented, namely the MP method. The MP
different ranges of frequency, as shown in Fig (13). Using Fig method hinges on the momentum approach as applied loads
(13), the Convergence of the load values is determined and to the aircraft structure are lumped as distributed loads along
the errors in the recovered loads are compared, as shown in the aircraft load axis. The MP method provided a structural
Fig (14).
 Downloaded from SAE International by University of British Columbia, Sunday, July 29, 2018

El Sayed et al / SAE Int. J. Aerosp. / Volume 6, Issue 2(December 2013) 407

modeling flexibility as the DMIG approach can be employed

for structural representation at different parts of the airframe.
Advantages of the DMIG modeling approach include the
generation of high fidelity structural models at a user defined
cloud of grid points, reducing significantly the time and
efforts required to generate the reduced structural models as
the method is automated within MSC NASTRAN and
providing flexibility in selecting the Grid points at which
loads and deformations are required.
On the other hand, advantages of the MP method include
generating high fidelity loads, less sensitivity to modal
truncation errors in dynamic analysis and allowing flexibility
in selecting the level of condensation of structural model
without affecting the loads recovery algorithm.

REFERENCES
1. Love, M., and Bohlman, J., “Aeroelastic Tailoring and Integrated Wing
Design,” Recent Advances in Multidisciplinary Analysis and
Optimization, NASA CP-3031, Pt. 1, 1989, pp. 431-444.
2. Elsayed, M.S.A., Sedaghati, R. and Abdo, M., “Accurate Stick Model
Development for Static Analysis of Complex Aircraft Wing-Box
Structures,” AIAA JOURNAL, Vol. 47, No. 9, September 2009.
3. Abdo, M., and Pepin, F., “Transonic Aerodynamics of Flexible Wings”,
CASI 48th Annual Conferences Proceedings, Canadian Aeronautics and
Space Institute, Toronto, Ontario, 2001, pp. 47-53.
4. Abdo, M., Piperni, P., Isikveren, A., and Kafyeke, F., “Optimization of a
Business Jet”, CASI Annual General Meeting, Aircraft Design &
Development Symposium, Toronto, Ontario, April 2005.
5. Timoshenko, S., “History of strength of materials”, McGraw-Hill New
York, 1953.
6. Guyan, R. J., “Reduction of Stiffness and Mass Matrices”, AIAA
Journal, Vol. 13, No. 2, Feb. 1965, pp. 380.
7. Wright, J.R., Cooper, J.E., “Introduction to Aircraft Aeroelasticity and
Loads”, John Wiley & Sons Ltd, 2007.
8. Craig, R.R. Jr., and Bampton, M.C.C., “Coupling of Substructures for
Dynamic Analyses”, AIAA Journal, vol. 6, no. 7, July 1968, pp.
1313-1319.
9. Hashemi-Kia, M., and Toossi, M., “Development and Application of a
Technique for Reducing Airframe Finite Element Models for Dynamics
Analysis”, NASA, CR 187448, Oct. 1990, McDonnell Douglas
Helicopter Company, Mesa, AZ.
10. Raymond, M., and Miller, M., MSC/NASTRAN, “Quick Reference
Guide”, Ver. 68, 1994.
11. Reschke, C., “Integrated Flight Loads Modelling and Analysis for
Flexible Transport Aircraft,” PhD thesis, Stuttgart University, Germany,
2006.
12. Bisplinghoff, R. L., Ashley, H. and Halfman, R. L., “Aeroelasticity,”
Dover Publications, Inc., 1955.
13. Engelsen, F. and Livne, E., “Mode Acceleration Based Random Gust
Stresses in Aeroservoelastic Optimization,” Journal of Aircraft, 41(2):
335-347, 2004.
14. Fransen, S.H.J.A., “An Overview and Comparison of OTM
Formulations on the basis of the Mode Displacement Method and the
Mode Acceleration Method,” Worldwide Aerospace Conference &
Technology Showcase, 2001.
15. Kalman, T. P., Rodden, W. P. and Giesing J. P., “Application of the
Doublet-Lattice Method to Nonplanar Configurations in Subsonic
Flow,” Journal of Aircraft, 8(6):406-413, 1971.
16. Pototzky, A. S., “New and Existing Techniques for Dynamic Loads
Analysis of Flexible Airplanes,” Journal of Aircraft, 23(4):340-347,
1985.
17. Craig, R. R., “Structural Dynamics,” Wiley, 1981.
18. Karpel, M. and Presente, E., “Structural Dynamic Loads in Response to
Impulsive Excitation,” International Forum on Aeroelasticity and
Structural Dynamics, pages 1059-1075, May 1993.
19. Albano, E.; Rodden, W. P., “A doublet-lattice method for calculating lift
distributions on oscillating surfaces in subsonic flows,” AIAA Journal,
vol. 7, issue 2, pp. 279-285