Poly MAX
Poly MAX
Poly MAX
Figure 3: Comparison of the measured FRFs (grey/green) with FRFs synthesized from the identified modal model (black/red). The FRFs between 2 of the 4
inputs and 1 typical output are shown.
Figure 4 Stabilization diagrams obtained by applying different parameter estimation methods to data from a partially trimmed car: (Left) FDPI; (Right) the new
LMS PolyMAX method.
LMS PolyMAX on flight strain gauges. Next to these measurable 6 shows both stabilization diagrams. Also
flutter data forces, also turbulences are exciting in this example the LMS PolyMAX method
the plane resulting in rather noisy FRFs. is clearly better than the LSCE method:
In some cases FRF data are highly Figure 5 shows some typical multiple selecting poles is intuitive, clear, and
contaminated by noise. Flight flutter coherences and the corresponding FRFs. reliable. The synthesized FRFs (Figure 7)
testing is such a case. In the example They clearly show the noisy character of validate the LMS PolyMAX estimations,
considered here, both wing tips of a plane the data. During the flight, accelerations even in the presence of high amounts of
are excited during the flight with so-called were measured at 9 locations, while data noise.
rotating vanes. These vanes generate a both wing tips were excited (2 inputs).
sine sweep through the frequency range The data were analyzed using both the
of interest. The forces are measured by LSCE and LMS PolyMAX method. Figure
Figure 5: Flight flutter test data. (Left) multiple coherences of a sensor at the wing tip close to the excitation (black/red) and a sensor at the back of the
plane (grey/green); (Right) corresponding FRFs. The frequency axis is blind for confidentiality reasons.
Figure 6: Stabilization diagrams obtained by applying different parameter estimation methods to the flight flutter test data: (Left) LSCE; (Right) the new LMS
PolyMAX method.
LMS PolyMAX: Historical
background
The LMS PolyMAX method is a further
evolution of the least-squares complex
frequency-domain (LSCF) estimation
method. That method was first
introduced to find initial values for the
iterative maximum likelihood method
[7]. The method estimates a so-called
common-denominator transfer function
model [8]. Quickly it was found that
these “initial values” yielded already
very accurate modal parameters with
a very small computational effort [7, 9,
10]. The most important advantage of
the LSCF estimator over the available
and widely applied parameter estimation
techniques [2] is the fact that very clear
stabilization diagrams are obtained. A
thorough analysis of different variants
of the common-denominator LSCF
method can be found in [10]. A complete
background on frequency-domain system
identification can be found in [11].
(2)
(3)
(4)
This procedure is similar to what happens in the time-domain 1
This method is implemented as “Frequency – domain MDOF” in
LSCE method and allows constructing a stabilization diagram Cada-X [3].
for increasing model orders and using stability criteria for 2
This method is implemented as “Time – domain MDOF” both in
eigenfrequencies, damping ratios and modal participation
Test.Lab and Cada-X [3].
factors.
Comparing LMS PolyMAX with other Modal Therefore one quickly runs into numerical conditioning
Parameter Estimators problems and severe constraints apply to both the frequency
range as the model order range of the analysis. In the past, it
LMS PolyMAX versus LSCE has been proposed to use an orthogonal polynomial basis for
As may be clear from previous section, the LMS PolyMAX the frequency-domain model to solve the numerical problems.
method proceeds along similar lines as the polyreference However this increases significantly the computational time
LSCE time-domain method: and memory requirements.
• Establishment of a set of linear equations for the The LMS PolyMAX method does not suffer from numerical
maximum required model order, from which the matrix problems as it is formulated in the -domain (i.e. a
polynomial coefficients can be computed in a least- frequency-domain model that is derived from a discrete-time
squares sense. model), whereas the existing frequency-domain methods
• Construction of a stabilization diagram by solving the use a Laplace-domain formulation (i.e. a frequency-domain
eigenvalue problem (3) for increasing model orders. The model that is derived from a continuous-time model). In
information regarding eigenfrequencies, damping ratios LMS PolyMAX, the frequency axis that extends between
and modal participation factors is contained in this and is shifted and mapped into a half unit circle in the
diagram. complex plane (2):
• Based on the user-interpretation of the stabilization
diagram, computation of the mode shapes and the lower
and upper residuals by solving (5) in a least-squares sense. , (6)
The difference between LSCE and LMS PolyMAX lies in Similar to other frequency-domain methods, the LMS
the first step. LSCE uses impulse responses to find the PolyMAX method involves the inversion of a matrix containing
polynomial coefficients, whereas LMS PolyMAX requires powers of the frequency-axis of the data. The main advantage
frequency response functions. of LMS PolyMAX is that taking powers of the -variable does
not increase the range of the values, as it boils down to a
However, this seemingly small difference has big rotation in the complex plane: . As a result,
consequences for the modal parameter estimation process. the LMS PolyMAX method can deal with a large frequency
It turns out that the new LMS PolyMAX estimator yields range and very large model orders, speeding up the modal
extremely clear stabilization diagrams making it very simple parameter estimation process considerably, as in many cases
to select the “physical” poles. In the LSCE method, the the complete frequency-band of interest can be processed at
non-physical (and sometimes even the physical) poles tend once.
to “wander” in the stabilization diagram, especially at large
model orders. The LMS PolyMAX method has the interesting There was some common belief that the numerical
property that the non-physical poles are estimated with a conditioning of frequency-domain methods is worse than time
negative damping ratio so that they can be excluded before domain methods and that broadband analyses are preferably
plotting them. Such a clear diagram does not mean that performed in time-domain [2]. When using the LMS PolyMAX
some of the poles are missing; on the contrary, more poles approach, these statements are definitely no longer true.
can be found with the LMS PolyMAX method, as evidenced
by the examples in this application note. Other validation Computational efficiency
studies also revealed that the LMS PolyMAX method has The advantages discussed here have no price in terms of
no problems in correctly estimating modes having a low computational time: LMS PolyMAX is as fast as LSCE. LSCE
damping ratio. It is sometimes stated that time-domain became the industry-standard because of its high speed
methods are preferred in case of low damping, and even for a very large number of measured outputs. A lot of
frequency-domain methods in case of high damping. The research efforts were spent to achieve this computational
LMS PolyMAX method excels in both cases. efficiency. On current PC platforms, calculation and display of
the stabilization diagram for a typical full car body model (like
LMS PolyMAX versus other frequency-domain the trimmed car body example discussed here) is in the order
methods of some seconds.
Typical for many frequency-domain parameter estimation
methods is that they involve the inversion of a matrix
containing powers of the frequency-axis of the data.
LMS PolyMAX: Executing Modal References [9] VAN DER AUWERAER H., P.
Testing Jobs at Unrivaled Speed GUILLAUME, P. VERBOVEN AND S.
[1] VAN DER AUWERAER H., C. VANLANDUIT. Application of a fast-
and Accuray
LIEFOOGHE, K. WYCKAERT AND J. stabilizing frequency domain parameter
DEBILLE. Comparative study of excitation estimation method. ASME Journal of
With the new LMS PolyMAX method,
and parameter estimation techniques on Dynamic Systems, Measurement, and
a breakthrough in Experimental Modal
a fully equipped car. In Proceedings of Control, 123(4), 651–658, 2001.
Analysis has been achieved. Whereas
the method equals or even outperforms IMAC 11, the International Modal Analysis
Conference, 627–633, Kissimmee (FL), [10] VERBOVEN, P. Frequency domain
the current standard LSCE technique
USA, 1–4 February 1993 system identification for modal analysis.
on common test structures, it brings a
PhD Thesis, Vrije Universiteit Brussel,
solution for problems -like trimmed body
[2] HEYLEN W., S. LAMMENS AND P. Belgium, 2002.
and flutter data- where current EMA
technology has shown its limits. SAS. Modal Analysis Theory and Testing.
Department of Mechanical Engineering, [11] PINTELON R. AND J. SCHOUKENS.
Katholieke Universiteit Leuven, Leuven, System Identification: a Frequency
By substantially simplifying the analysis
Belgium, 1995. Domain Approach. IEEE Press, New York,
process, LMS PolyMAX will be enjoyed
2001.
by many new users in the field. For
advanced applications, its powerful [3] LMS INTERNATIONAL. The LMS
Theory and Background Book, Leuven, [12] GUILLAUME P., P. VERBOVEN, S.
clear stabilization, and the quality of the
Belgium, 2000. VANLANDUIT, H. VAN DER AUWERAER
modal parameter estimation are real
AND B. PEETERS. A poly-reference
breakthroughs, widely expanding the
[4] LEMBREGTS F., J. LEURIDAN, L. implementation of the least-squares
application range and drastically reducing
ZHANG AND H. KANDA. Multiple input complex frequency-domain estimator. In
the number of iterations needed.
modal analysis of frequency response Proceedings of IMAC 21, the International
LMS PolyMAX is not yet-another-
functions based direct parameter Modal Analysis Conference, Kissimmee
parameter-estimation-technique, but a
identification. In Proceedings of IMAC (FL), USA, February 2003.
global solution for Experimental Modal
Analysis. 4, the International Modal Analysis
Conference, 589–598, Los Angeles (CA), [13] PEETERS B., P. GUILLAUME, H.
USA, 1986. VAN DER AUWERAER, B. CAUBERGHE,
The new LMS PolyMAX is part of the
P. VERBOVEN AND J. LEURIDAN.
LMS Test.Lab Structures solution for
[5] LEMBREGTS F., R. SNOEYS AND J. Automotive and aerospace applications
Modal testing and analysis. LMS Test.
LEURIDAN. Application and evaluation of the LMS PolyMAX modal parameter
Lab Structures is an integrated suite
of multiple input modal parameter estimation method. In Proceedings of
of applications covering the range
estimation. International Journal of IMAC 22, Dearborn (MI), USA, January
of structural dynamic engineering
Analytical and Experimental Modal 2004.
completely. Dedicated applications
serve impact hammer measurements, Analysis, 2(1), 19–31, 1987.
single shaker testing and advanced
multiple-input, multiple-output (MIMO) [6] BROWN D.L., R.J. ALLEMANG,
analysis. These starting points measure R. ZIMMERMAN AND M. MERGEAY.
the motion/force transfer or Frequency Parameter estimation techniques for
Response Functions (FRF) required modal analysis. Society of Automotive
for modal analysis. A dedicated modal Engineers, Paper No. 790221, 1979.
analysis module automatically accesses
these measurements to compute the [7] GUILLAUME P., P. VERBOVEN AND
modal parameters: mode shape vector, S. VANLANDUIT. Frequency-domain
resonant frequency, damping factor and maximum likelihood identification of
modal mass. All applications are tightly modal parameters with confidence
integrated so that data streams smoothly intervals. In Proceedings of ISMA 23, the
from acquisition, through analysis, to International Conference on Noise and
display and reporting. Vibration Engineering, Leuven, Belgium,
16–18 September 1998.
LMS Virtual.Lab
LMS Virtual.Lab is the world’s first
integrated software environment for the
functional performance engineering of
critical design attributes, such as noise
and vibration, ride, handling, comfort,
safety and durability.
LMS Test.Lab