Home

ISSPA User`s Guide

1. aD E EB Drucker 9 Systemsteuerung LA DF Netzwerk oe iD Objektfe 2 52 MB Freier Speicher 93 2 MB EY Arbeitsplatz A Fig 6 2 3 Copy data file to data ISSPAO2 User s Guide 6 Using the ISSPA GUI 65 Convert Universal File Type 58 to UKL UR3 File button unv gt ur3 ISSPA Tools window Suchen in E data amp j ql c mjs Dateiname man_1 22 uny Dateityp uny v Abbrechen EN Explorer D project isspa garteur data Iof x Datei Bearbeiten Ansicht Wechselnzu Favoriten Extras Ei e ite Lee i Lan ee Zur ck Aufw rts Ausschneiden Kopieren Einf gen Riickgangig L schen Eigenschaften Ordner A Desktop 3 Arbeitsplatz ee 35 Diskette A amp Festplatte C Wechseldatentrager D 29 project 3 0 isspa garteur C boot data 2 587 KB Datei UNY 20 07 99 20 02 850KB Datei UR3 25 02 03 10 51 5 man_12z ur3 i F Th 8 E amp F 2 Drucker Systemsteuerung E DF Netzwerk E E cea 11 r 1 2 Objektfe 3 35 MB Freier Speicher 92 4 MB E Arbeitsplatz Fig 6 2 4 Convert Universal File Type 58 to UKL UR3 File ISSPAO2 User s Guide 6 Using the ISSPA GUI 66 Step4 Change the ISSPA working directory to b001 To change the ISSPA working directory use either the buttons of the JSSPA Plot win
2. t1 6 4979 7 all MDOF f H2 1 MDOF Fig 6 2 20 FITKOR identification Step 10 4 Check FITKOR identification results Press Run ISPRES in the ISSPA pulldown menu ISSPA Plot d project isspa garteur b001 Manage Plot ISSPA Tag Window Print Help Comparison of Measured and Recalculated Response 7 all MDOF f H2 f 6 7 8 10 all MDOF Hz 1 MDOF Fig 6 2 21 Check of FITKOR identification results 79 ISSPAO2 User s Guide 6 Using the ISSPA GUI Step 10 5 ISSPA identification Press Run ISSPALL in the ISSPA pulldown menu ISSPA Plot d project isspa garteur b001 Manage Plot ISSPA Tag Window Print Help Meas and Modified Response 7 I MDOF f Hz T all MDOF f Hz 1 MDOF Fig 6 2 22 ISSPA identification results Step 10 6 Check ISSPA identification results Press Run ISPRES in the ISSPA pulldown menu ISSPA Plot d project isspa garteur b001 Manage Plot ISSPA Tag Window Print Help Comparison of Measured and Recalculated Response II MDOF 11 6 4959 L i all MDOF fiH 1 MDOF Fig 6 2 23 Check of ISSPA identification results 80 ISSPAO2 User s Guide 6 Using the ISSPA GUI 81 Step11 identify all other modes in subdir b002 repeat all substeps of step 10 to identify all modes of the measurement data set
3. Fig 6 2 12 ISSPA identification subdirectories successfully created ISSPAO2 User s Guide 6 Using the ISSPA GUI 73 EN Explorer D project isspa garteur Biel Es Datei Bearbeiten Ansicht Wechselnzu Favoriten Extras e Fr 2 x Zur ck Wonwatts Aufw rts Ausschneiden Kopieren Einf gen Riickgangig L schen Eigenschaften Adresse D project isspa garteur gt Ordner x Dateiname GroBe Typ Ge ndertam A Desktop af C b001 Dateiordner 25 02 03 10 27 g Arbeitsplatz Ga b002 Dateiordner 25 02 03 11 15 E 3 5 Diskette 4 LA b003 Dateiordner 25 02 03 11 15 amp Festplatte C Q b004 Dateiordner 25 02 03 11 15 Wechseldatentr ger D LJ b005 Dateiordner 25 02 03 11 15 F 1 S L project CI bemb Dateiordner 25 02 03 11 16 3 isspa C data Dateiordner 25 02 03 10 27 cas 2 b001 i C data 2 E aD F 8 Drucker Systemsteuerung EA DF Netzwerk Daca ssr r 1 Objekte 0 Byte Freier Speicher 92 3 MB EJ Arbeitsplatz Fig 6 2 13 ISSPA identification subdirectories ISSPAO2 User s Guide 6 Using the ISSPA GUI 74 Step8 Create fuelleis ein in subdirectory bemb To create a fuelleis ein file use the create fuelleis ein button of the ISSPA cmb window Create FUELLEIS EIN d project isspa garteur b001 FUELLEIS EIN File fue
4. nodal excitation forces moments overall number of MDOF overall number of modes overall number of exciter forces overall number of spectral lines frequency points base acceleration file not necessarily needed ISSPAO2 User s Guide ISSPAO2 User s Guide 1 ISSPA Installation 5 1 ISSPA Installation The software package ISSPA comprises FORTRAN executables and MATHWORKS MATLAB command Files m files The FORTRAN executables are the kernel of the identification procedure and may be called independently from any other software using simple operating system commands A graphical user interface GUI has been provided to ease the application of the ISSPA modules and to produce all graphical output The GUI is implemented into MATHWORKS MATLAB using several MATLAB command files m or p Therefore ISSPA requires that MATHWORKS MATLAB Version 5 3 or higher is installed on the computer PC or UNIX workstation if the ISSPA GUT is to be used ISSPA is distributed on a media using the following directory structure I I I I I I I I I demo docu ein exe for pcode src s2_unix sgi unix contents demo ISSPA demo identification ref Appendix E docu ISSPA documentation Microsoft WORD 97 documents or pdf documents ein Templates of ISSPA input files exe ISSPA FORTRAN executables for ISSPA FORTRAN sources pcode ISSPA graphical user interface p code src ISSPA graphical user interface s2_unix ISSPA implementation t
5. US ziUy yy U 4 14 xl In the case that the base rotations can be neglected eq 4 12 yields ust us un ur U U 4 15 rel abs A Dr U U Note Either the transfer functions H re OF the relative displacements U re are used for ISSPA input 3 Definition of excitation forces To identify the modal masses in addition to the eigenfrequencies mode shapes modal damping factors the excitation forces must be known If a special force measurement device at the junction shaker structure is used they can be measured directly 15 17 If an analytical mass matrix with respect to the structural MDOFs is available an equivalent excitation force can be calculated for each MDOF using the reference excitation A In both cases all modal parameters can be calculated However if the excitation forces are unknown all modal parameters except the modal mass can be calculated from the measured data a excitation forces are unknown i e the interface reaction forces are not measured In this case a dummy unit force F 1 is applied at an arbitrary MDOF However the identified modal masses have no physical meaning and are just needed to scale the response in a recalculation synthesized response using the identified data ISSPAO2 User s Guide 4 ISSPA Input Features 36 b excitation forces are calculated from an analytical mass matrix In this case an analytical mass matrix with respect to the N structural MDOFs must be a
6. 30 40 50 all MDOF f Hz MDOF 1 MDOF Fig 6 2 25 View the comparison of measured and recalculated response The recalculated response was calculated using the combined identification results ISSPA99 User s Guide 7 ISSPA Main Results Files 83 7 ISSPA Main Results Files The identification results of each identification subdirectory are stored in UKL FR3 data files nomol1 fr3 modal masses and damping matrix nomo12 fr3 real parts of the eigenvectors imaginary parts of the eigenvectors right hand modal matrix left hand modal matrix nomo13 fr3 inverse physical stiffness matrix spectral synthesis in general incomplete nomo9 fr3 inverse physical mass matrix spectral synthesis in general incomplete Refer to app A for detailed information about the UKL FR3 format If required these data file can be read directly into MATLAB by using the readfr3 command example read eigenvalues real imaginary part and right and left hand eigenvectors from the ISSPA data file nomo12 fr3 lam_re lam_im Y X readfr3 nomo12 fr3 If module FUELLEIS has been applied to combine several identification ranges the nomol1 fr3 nomo12 fr3 nomo13 fr3 nomo9 fr3 files contain the results of the combined modal data In addition the file fuel fr3 is created fuel fr3 eigenfrequencies Hz modal masses modal damping matrix modal matrix eigenvector by eigenvector physical inverse mass matrix spectral synthesis
7. ISPMOD Purpose Selection of identification frequency ranges Input of ISSPA control parameters to specify the type of identification Input ispmod ein ISSPA control parameters Measurement data file specification supported formats UKL FR3 app A UKL UR3 app A UKL IDX app A MATLAB mat file app A SDRC Universal files Type 58 max filename length 132 characters Output ispmod aus listing file ispmod_01 mat plot file ispmod_02 mat mode indicators Limitations dynamic allocation of matrix dimensions limited to overall 5 2 Million REAL 4 values low to medium scaled problems e g N 800 MDOF m 3200 spectral lines but N m lt 2560000 Position in ISSPA sequence post none pre FITKOR ISPKON ISSPAO2 User s Guide 5 ISSPA Modules 40 5 2 FITKOR Purpose Curve fit identification of modal parameters and residual correction eigenfrequencies eigenvectors modal masses diagonal modal damping matrix Input fitkor ein FITKOR control parameters eigenfrequency iteration limit ITFMAX damping iteration limits ITDMAX Iteration start values eigenfrequencies modal damping factors Output fitkor aus listing file fitkor_O1 mat plot file residual corrected response Limitations identification using a simplified modal approach diagonal damping matrix proportional damping Position in ISSPA sequence post ISPMOD pre ISPRES ISPKON ISSPAO2 User s Guide 5 ISSPA Modules 41 3 3 ISPKON Purpose Singular value de
8. at the University of Kassel Germany UKL Prof Dr Ing M Link ISSPA comprises standalone FORTRAN executables and a graphical user interface GUI written using MATHWORKS MATLAB 10 computation and visualization software ISSPA is aimed to be used to solve low to medium scaled problems up approx 800 measurement degrees of freedom MDOF and 3200 spectral lines The developers of ISSPA put much more emphasis on the solution algorithms and the variety of applications than on graphical user interface programming Due to this the user has to take a high share on specifying all needed input data correctly ISSPA expects the user to be a so called friendly user i e he she has a basic knowledge in mechanics and structural identification to supply the needed input data and interpret the output results he she has a basic knowledge in using MATHWORKS MATLAB ISSPA is certainly not designed for users who want to proof how stupid software can be Nor should the user be disappointed if the application may fail due to erroneous input The software package ISSPA Identification of Structural System Parameters consists of nine modules ISPMOD Selection of identification frequency ranges FITKOR Curve fit identification of modal parameters and residual correction ISPKON Condensation of measurement matrices to effective degrees of freedom gt modified measurement ISSPA Identification of modal parameters from modified measurement NOMOIS
9. 433 01571 6 Kasai T Link M Identification of non proportional modal damping matrix and real normal modes Mechanical Systems and Signal Processing 2002 16 6 pp 921 934
10. 6 2 16 Setting of the fuelleis ein file The user may also edit the fuelleis ein file This ASCII file is of type UKL FR3 To close the Create FUELLEIS EIN figure press close ISSPAO2 User s Guide 6 Using the ISSPA GUI 76 Step9 View measurement data and mode indicator functions of the complete data set Step 9 1 Change the ISSPA working directory to b005 Step 9 2 Start ISPMOD from the JSSPA Linear Identification window Step 9 3 Press plot to view the measurement data En _lAlx ISSPA Plot d project isspa garteur b005 Manage Plot ISSPA Tag Window Print Help Measured Response 40 MDOF f H2 modify residuals 30 40 50 all MDOF fIHzl Moor MDOF ISPHOD execution complete Ready NUM Astar A e A O ultsE K Micro PAMATL Biss TEE gAissP Wiss TEE TEE via 1203 Fig 6 2 17 View measurement data ISSPAO2 User s Guide 6 Using the ISSPA GUI 77 Step 9 4 Press MIF to view the mode indicator functions ISSPA Plot d project isspa garteur b005 ork g Manage Plot ISSPA Tag Window Print Help Envelope amp Single Mode Indicator create fuelleis ein FUELLEIS add Residuals 0 Envelope f Hz f Sea ee 50 Single Mode Indicator f Hz 1 MDOF ISPHOD execution complete Ready Mstar A e AO ultek W Mico PAMATL iss Fisse Fissp Ed ssP Risse Bunt Fig 6 2 18 View mo
11. Assembly of the inverse physical system matrices DAMPF Damping matrix improvement using a modal approach ISPRES Recalculation using identified data FUELLEIS Assembly of the results of different frequency ranges ISPGEN Calculation of dynamic responses from given modal data The theory on which the software package is based is described in 1 2 3 13 18 A comprehensive introduction to the theory is given in chap 3 of this manual ISSPAO2 User s Guide 2 General 8 The ISSPA software package comprises two different identification procedures FITKOR an identification approach based on modal parameters ISSPA an identification approach based on physical parameters Both procedures share a common data basis This data basis consists of several data files and an internal program control file These files manage the data transfer between the modules and are generally not used or modified by the user To perform a FITKOR identification the following modules are used successively ISPMOD data preprocessing FITKOR identification using a simplified modal approach residual correction ISPRES recalculation using identified data To perform an ISSPA identification the following modules are used successively ISPMOD data preprocessing FITKOR residual correction ISPKON condensation of measurement matrices ISSPA identification using a physical approach NOMOIS assembly of inverse system matrices DAMPF damping matrix improvement if required ISPRE
12. Dateiordner 25 02 03 10 27 2 Objektfe O Byte Freier Speicher 93 2 MB Arbeitsplatz Fig 6 2 1 Create subdirectories data and b001 ISSPAO2 User s Guide 6 Using the ISSPA GUI 63 Step2 Start ISSPA within MATLAB Step2 1 Press isspa within MATLAB E BIEN oranan ARE Manage Help ISSPA ISSPAD2_D1_b001 UKL Root Edition actual working directory select working directory ISSPA02_01_b001 UKL Root Edition isspa gt gt Ready NUM Astar A e a O UltraEdit 32 F Microsoft wor fA MATLAB Com SY Explorer D p s JISSPA Mai E LView Pro 1 0 4 EQ 10 35 Fig 6 2 2 Start ISSPA within MATLAB The JSSPA Plot window and the ISSPA Main window become accessible ISSPAO2 User s Guide 6 Using the ISSPA GUI 64 Step3 Copy the measurement data to directory data If required convert the data file format It is assumed that a file of measurement data is available from test This file must be of type UKL FR3 app A UKL UR3 app A UKL IDX app A MATLAB mat file app A SDRC Universal files Type 58 Refer to app for detailed information about the data formats The user can convert the original data set from unv gt ur3 Convert Universal File Type 58 to UKL UR3 File unv gt mat Convert Universal File Type 58to MATLAB mat File using the appropriate buttons of the JSSPA Tools window If the original data set is none of these types the user m
13. be organized and named as follows data measurement data file b001 first identification subdirectory All other subdirectories will be created automatically in step 7 All necessary steps will be illustrated by the data of the garteur example The complete garteur example is given on the demo garteur directory The user can autorun this example by pressing garteur within MATLAB from the demo garteur autorun directory It is recommended that prior to program execution the user should copy the complete demo directory to a temporary working directory to keep the original demo directory save Additional demonstration examples are given in the demo directory which can be started accordingly These examples demonstrate different identification types and the application of the different measurement data formats Step1 Create subdirectories data and b001 example ISSPA identification parent directory D project isspa garteur Datei Bearbeiten Ansicht Wechselnzu Favoriten Extras Ki eo gt fw BB amp x Zur ck Vorvarts Aufw rts een Kopieren Einf gen R ckg ngig L schen Eigenschaften Adresse D project isspa garteur gt Ordner x A Desktop a 2 3 Arbeitsplatz H E 3 5 Diskette A ES Festplatte C o Wechseldatentrager D Bg project S E isspa HAN garteur a E a F i Drucker Systemsteuerung E DF Netzwerk 3 Scheduled Tasks ji 2 ketone x 25 02 03 10 27
14. terms Purge Delete ISSPA intermediate results main Start tools ctr Switch ISSPA window lin cmb plot id log list Fig 6 1 7 ISSPA Combine Identification window ISSPAO2 User s Guide 6 Using the ISSPA GUI 60 4 ISSPA Identification Process c user weiland m lab isspa demo fhg5 b003 Window Print Help ISSPA02_01_b001 UKL Root Edition Fig 6 1 8 ISSPA Identification Process Log window for development use only ISSPAO2 User s Guide 6 Using the ISSPA GUI 61 ISSPA Identification Results c user weiland m lab isspa demo fhg5 b003 Manage Window Print Help actual path c user weiland matlab isspa demo fhgS bO03 eigenfrequency Hz modal damping modal mass 4 7005099e 001 9 6309643e 001 3 7011466e 000 6 3173374e 001 9 9685636e 001 3 0555536e 000 5 6520631e 001 9 1441730e 001 3 6172567 e 000 List modal parameters of actual working directory eigenfrequencies Hz modal damping factors Percent of critical damping diagonal terms of the damping matrix modal masses w r eigenvectors normalized to max value equal 1 Fig 6 1 9 ISSPA Identification Results window ISSPAO2 User s Guide 6 Using the ISSPA GUI 62 6 2 Getting Started It is recommended to use the following steps for an ISSPA identification of a measurement data set The data set will in general not be identified in one single identification run Therefore a set of subdirectories must be allocated which should
15. yields the transfer functions with respect to a non fixed base In this case all other base excitation DOFs are treated as structural DOFs ISSPAO2 User s Guide 4 ISSPA Input Features 34 2 Calculation of the relative response at the structural MDOFs with respect to the base A In the case of the transfer function matrix eq 4 9 10 the rigid body part must be removed from the transfer functions l H He H T 4 1 1a b imag _ ryimag H e H where Are 3 He real imaginary part of H with respect to the base A T rigid body transformation matrix relating the unit base displacements in excitation direction and the structural MDOFs responses This modification effects only the real part and depends on the orientation of the structural MDOFs with respect to the XYZ coordinate system defined by the base DOFs So in the case e g that the direction of a structural MDOF coincides with the base excitation direction the value 1 must be subtracted from the real part of the transfer function In the more general case the relative response of a structural MDOFs is given by Ure T Uis u TU 4 12 where Ua relative response Ups measured response T rigid body transformation matrix 100 0 z y i hMDOF inx T 0 1 0 z 0 x i thMDOF iny 4 13 001 yf xi O i thMDOFinZ ISSPAO2 User s Guide 4 ISSPA Input Features 35 Example Calculation of the relative response of the first structural MDOF in X direction U USS
16. 5 16 17 18 19 Link M Ulrich H Weiland M ISSPA Guidelines Modal Identification using Base Excitation Test Data Laboratory of Lightweight Structures and Structural Mechanics UKL Kassel April 1999 Link M Qian G Identification of Dynamic Models for Substructure Synthesis Using Base Excitation and Measured Reaction Forces Revue Francaise de Mecanique No 1 1994 Weiland M Link M Direct Parameter Estimation of Weak Nonlinear Systems Using Vibration Test Data Proc ASME Conf on Noise amp Vibration Boston ISBN 0 7918 1718 0 1995 Schedlinski C Link M Identification of Frequency Response Functions and Modal Data From Base Excitation Tests Using Measured Interface Forces Proc ASME Conf on Noise amp Vibration Boston ISBN 0 7918 1718 0 1995 Weiland M Link M A Direct Parameter Estimation Method for Weak Nonlinear Systems Proc of Int Modal Analysis Conf IMAC 14 ISBN 0 912053 49 Dearborn USA Feb 1996 Schedlinski C Link M Identification of Rigid Body Properties Using Base Excitation and Measured Interface Forces Proceedings of the 1996 ESA Conference on Spacecraft Structures Materials and Mechanical Testing European Space Agency ESA Noordwik The Netherlands 1996 Kr tzig W B Meskouris K and Link M Baudynamik und Systemidentifikation In Der Ingenieurbau Grundwissen Hrsg G Mehlhorn Verlag Ernst amp Sohn Berlin 1995 ISBN 3
17. 7140E 03 0 1616170E 05 0 4712373E 04 0 8415350E 05 0 1437794E 03 0 1406293E 03 0 4294511E 05 0 1363841E 03 0 3075632E 05 0 4994640E 04 0 3231501E 05 0 1315818E 05 0 6000249E 07 0 3330285E 04 0 2405812E 04 0 2041526E 05 0 1146013E 04 0 6943534E 05 0 1996569E 04 0 7146348E 05 0 1468075E 05 0 1470365E 03 0 1436025E 03 0 1817207E 05 0 1416077E 03 0 2184509E 05 0 4620782E 04 0 9366609E 05 0 1371530E 03 0 1364532E 03 0 3364921E 05 0 1346868E 03 0 2668122E 05 0 5092345E 04 0 1219600E 05 0 1342383E 05 0 1625624E 04 0 1380897E 05 0 1437719E 06 0 6818987E 05 0 8851293E 07 0 1870318E 06 0 2267865E 05 0 1513406E 05 0 3353644E 05 0 1816545E 04 0 7081912E 06 0 1385029E 04 0 4607240E 06 0 1596397E 04 ISSPA99 User s Guide 8 Reference and Related Literature 85 8 Reference and Related Literature 1 2 3 4 5 6 7 8 9 10 11 Link M Theory of a Method for Identifying Incomplete Matrices from Vibration Test Data Z d Flugwissenschaft und Weltraumforschung ZFW 9 1985 Heft 2 3 Link M Structural System Identification Using Single and Multi Axial Vibration Test Data Proc Spacecraft Structures CNES Toulouse 1985 ESA SP 238 1986 Link M Vollan A Identification of Structural System Parameters from Dynamic Response Data Z d Flugwissenschaft und Weltraumforschung ZFW 2 1978 Heft 3 Link M Weiland M Moreno Barragan J Direct Physical Matrix Identification as Co
18. 834872E 01 0 8335010E 00 0 1886097E 00 0 1060285E 01 0 1453784E 01 0 2052904E 00 0 1514559E 01 0 3442339E 01 0 3644113E 01 0 1000000E 01 0 5018267E 01 0 2950925E 01 0 9275818E 00 0 2108194E 01 0 9619984E 02 0 1490270E 01 0 1224379E 02 0 1097076E 01 0 6956457E 02 0 1316239E 02 0 6194564E 02 0 6055269E 01 0 2822213E 02 0 8184624E 01 0 1373446E 01 0 1025655E 00 0 1000000E 01 0 4829212E 01 0 3005042E 01 0 9103230E 00 0 1975775E 01 0 1041523E 01 0 1635938E 01 0 3307970E 00 0 5121922E 01 0 4001315E 01 0 2341951E 00 0 1025470E 01 0 5793894E 01 0 4974201E 01 0 1918605E 02 0 8239701E 02 0 1147643E 01 0 1912182E 02 0 2152015E 02 0 5058206E 01 0 1045360E 01 0 9565794E 01 0 5013299E 02 0 6882320E 01 24 24R 4 inverse mass matrix row by row 0 1784582E 01 0 3506804E 00 0 2511585E 01 0 1089068E 01 0 3571489E 01 0 5694401E 01 0 1353878E 01 0 4111475E 00 0 2512510E 00 0 6817310E 01 0 8913288E 01 0 3146711E 01 0 1964661E 01 0 4726983E 01 0 1815625E 03 0 1298116E 01 0 4074537E 01 0 3792671E 01 0 2377156E 01 0 1896085E 00 0 3374030E 02 0 2537450E 00 0 1244929E 01 0 1950248E 00 0 3506804E 00 0 2872380E 00 0 1362605E 01 0 2262529E 00 0 5061503E 02 0 6673619E 02 0 1873330E 01 0 2377561E 01 0 6364847E 01 0 1635967E 03 0 2948286E 03 0 4762087E 02 0 2170958E 02 0 3634376E 01 0 2022791E 00 0 3744663E 02 0 1676566E 00 0 9851424E 02 0 1796962E 00 24 24R 4 inverse stiffness matrix row by row 0 1806364E 03 0 1470365E 03 0 8851558E 06 0 114
19. ISP File Options UR2 Plot Data Truncation Limit TRUMLIM lo 3 0 excitation forces INFNR RKRAFT frequency range selection IANF IEND IPEAK ge WRITE CLOSE Fig 6 2 7 Setting of the ispmod ein file The user may also edit the ispmod ein file This ASCII file is of type UKL FR3 To close the Create ISPMOD EIN figure press close ISSPAO2 User s Guide 6 Using the ISSPA GUI 69 Step6 Create fitkor ein in subdirectory b001 To create a fitkor ein file use the create fitkor ein button of the JSSPA Start window Create FITKOR EIN d project isspa garteur b001 FITKOR EIN File No of estimated modes MFG Iteration step limits TDMA ITF MAX Data Reduction IREDM IREDN 20 Residual Terms IRES estimated eigenfrequencies viscous damping Percent of critical damping 0 000000e 000 0 000000e 000 CLOSE Fig 6 2 8 The Create fitkor ein window to create modify the fitkor ein file The user must carefully fill out this template FITKOR EIN File press the New button to select the location of the fitkor ein file if the fitkor ein file already exists use the Browse button No of estimated modes MFG Iteration step limits ITDMAX ITDMAX 0 no damping iteration ITDMAX 20 max 20 damping iteration steps recommended ITFMAX ITFMAX 0 no eigenfrequency iteration ITDMAX 20 max 20 eigenfrequency iteration steps recomm Data Reduction IREDM IREDM 1 5 spe
20. Internet l K A S S E L UKA Fachgebiet Leichtbau und woe Laboratorium fiir Strukturmechanik RS ITAT Prof Dr Ing Michael Link 49 0 561 804 2654 2632 49 0 561 804 3631 link uni kassel de www uni kassel de fb14 leichtbau ISSPA 02 Identification of Structural System Parameters Version ISSPA02 03 User s Guide extract Revision 19 January 2004 ISSPAO2 User s Guide Table of Contents 1 ISSPA Installation 2 General 3 Theoretical Background 4 ISSPA Input Features 5 ISSPA Modules 5 1 ISPMOD 5 2 FITKOR 5 3 ISPKON 5 4 ISSPA 5 5 NOMOIS 5 6 DAMPF 5 7 ISPRES 5 8 FUELLEIS 5 9 ISPGEN 6 Using the ISSPA GUI 6 1 General 6 2 Getting started 7 ISSPA Main Results Files 8 Reference and Related Literature 27 39 39 40 41 42 43 44 45 46 47 49 49 62 83 85 ISSPAO2 User s Guide Abbreviation MDOF UKL ISSPA MATLAB GUI FRF X Y Z XX VY ZZ Ux Uy Uz Uxx Uvy Uzz Fx Fy Fz Fxx Fyy Fzz n N r MFG Dr m filename measurement degree of freedom University of Kassel Department of Civil Engineering Laboratory of Lightweight Structures and Structural Mechanics Identification of Structural System Parameters high performance numeric computation and visualization software MATHWORKS graphical user interface frequency response function MDOF direction translational MDOF rotational MDOF nodal excitation forces
21. PA GUI ISSPAMan BEIE Manage Help Buttons main Start tools ctr ISSPA lin cmb ISSPAO2_01_b001 plot id log list close UKL Root Edition Working Directory actual working directon select working directory main lin cmb id log list Fig 6 1 2 ISSPA Main window Switch ISSPA window close ISSPA Main window Select working directory 53 ISSPAO2 User s Guide 6 Using the ISSPA GUI lISSPA Stat RIE Manage Help Buttons create ispmod ein create fitkor ein create ISSPA subdirs create ispmod ein create fitkor ein Create ISSPA Subdirs main Start tools ctr main lin cmb lin cmb plot id log list plot id log Fig 6 1 3 ISSPA Start window 54 Create Modify ispmod ein Create Modify fitkor ein Create ISSPA identification subdirectories b001 b050 and the directory bemb to combine identification results from identification subdirectories Switch ISSPA window ISSPAO2 User s Guide 6 Using the ISSPA GUI ISSPA Tools PETES Manage Help Buttons unv gt ur3 Convert Universal File Type 58 to UKL UR3 File Data File Conversion unv gt mat Convert Universal File Type 58 to MATLAB mat File unv gt ur3 formum Convert UKL data files unformatted lt formatted unv gt mat formum main start tools ctr Switch ISSPA window main lin cmb lin cmb plot id log list plot id log list close Fig 6 1 4 ISSPA Tools
22. S recalculation using identified data Note if no residual correction is required FITKOR can be skipped All control input to a module is given in a ASCII input file with a fixed filename This filename consists of the module name and the extension ein e g ISPMOD EIN All input files except ISPGEN EIN are specified using the UKL FR3 file specifications which is given in appendix A Templates of the input files are given in appendix B and in the ein directory Each module creates a ASCII file which lists input data program flow and results The filename is fixed and consists of the module name and the extension aus e g ISPMOD AUS In order to manage the internal data flow between the modules unformatted and formatted sequential files are created A list of these files is given in appendix C The files follow the UKL FR3 UR3 file specifications They have no further relevance to the user and should be deleted after the results of an identification range are excepted The ISSPA GUI provides the Purge button to properly perform this task In practical applications the identification of a complete measurement data set is not performed in a single step Usually the overall data set is divided into several identification ranges which are identified one by one It is recommended to use a separate directory for each identification range e g bOOI b002 When the identification of each separate identification range is complete the moda
23. SSPA subdirs to create Actual value 5 ml HL Fig 6 2 10 Select Number of ISSPA subdirectories to create Use the slider to change the overall number of ISSPA subdirectories to create Dateiname E Dateityp M files m v Abbrechen ZL Fig 6 2 11 Select the parent directory of the ISSPA identification subdirectories b001 bD0xx bemb ISSPAO2 User s Guide 6 Using the ISSPA GUI 72 Select either ispmod ein or fitkor ein as source files to be copied in K Ea Suchen in apn amp f wl cl Dateiname fitkor ein Dateityp ein v Abbrechen Z Fig 6 2 11 Select either ispmod ein or fitkor ein file to be copied into the ISSPA identification subdirectories b001 b0xx bemb The ispmod ein fitkor ein and if exist is2tdas ein files of the selected directory are copied into the ISSPA identification subdirectories b001 bOxx The ispmod ein and if exist is2tdas ein are copied to the bcmb directory as well If any of the subdirectories and the respective input files exist the user will be notified He will get detailed information on the MATLAB command window and he will be asked to keep or overwrite the respective subdirectories and files If the user presses the yes to all button ISSPA will automatically overwrite all existing subdirectories files After copying all data the user will be notified Information X Ci ISSPA subdirectories created and input files copied
24. Step12 combine ISSPA identification subdirectories b001 bOxx in subdirectory bcmb It is assumed that the complete data set was identified using the three ISSPA identification subdirectories b001 b003 Step 12 1 Change the ISSPA working directory to bemb Step 12 2 Edit fuelleis ein to select the subdirectories b001 b003 Press Create fuelleis ein of the ISSPA Main window Press radio buttons of b001 b003 to select the subdirectories Press save to store actual settings Create FUELLEIS EIN d project isspa garteur bemb FUELLEIS EIN File d project isspa garteur bemb fuelleis ein fuel fr3 file format options IDPFD INVMA INVSTEN List Subdirectories ADDON E oo E oo E oo E oos E oos E oo E oos E oo active WRITE CLOSE Fig 6 2 24 Select subdirectories b001 b003 They are used to assemble the overall identification results file fuel fr3 Step 12 3 Press fuelleis to run FUELLEIS ISSPAO2 User s Guide 6 Using the ISSPA GUI 82 Step13 Calculate response data using the combined modal data set Step 13 1 Press ispmod from the ISSPA Linear Identification window to run ISPMOD Step 13 2 Press ispres from the ISSPA Linear Identification window to run ISPRES Step 13 3 Press cmp of ispres to view the comparison of measured and recalculated response Manage Plot ISSPA Tag Window Print Help Comparison of Measured and Recalculated Response 50 all MDOF f Hz
25. ally set by ISSPA Data Dimensions M overall number of spectral lines M 4096 N overall number of MDOF N 24 Data Type IHARMO measurement data is FRF data single point excitation Data Options IRESP if data file is of type UNV IRESP 0 response fixed IRESP 1 response moving Print Options INFUNV if data file is of type UNV INFUNV 1 list FRF header in ispmod aus INFUNV 0 no listing of FRF header in ispmod aus ISSPAO2 User s Guide 6 Using the ISSPA GUI 68 File Options IUR2 Create UKL UR plot files IUR2 1 createRuntime Options IDISP Amount of runtime display IDISP 0 none IDISP 1 errors IDISP 2 errors warnings IDISP 3 all Plot Data Truncation Limit ITRUNLIM excitation forces INFNR MDOF number INFNR 4 RKRAFT force amplitude FRF gt RKRAFT 1 0 frequency range selection frequeny range limits IANF spectral line number left ANF 1 IEND spectral line number right IEND 4096 IPEAK spectral line number of peak frequency if automatic frequency line reduction is required IPEAK 0_ no spectral line reduction After supplying all needed input the user must press write to store the actual settings Create ISPMOD EIN d project isspa garteur b001 ISPMOD EIN File D project isspa garteurbO01 ispmod ein ISSPA Data File D project isspa garteur data man_122z unyv Data Dimensions M N Data Type IHARMO DATA Options RESP Zs Print Options INFUN Y Runtime Options ID
26. cal user interface ISSPA GUI has been provided to ease the application of the ISSPA modules and to produce all graphical output The GUI is implemented into MATHWORKS MATLAB using several MATLAB command files m or p Therefore ISSPA requires that MATHWORKS MATLAB Version 5 3 or higher is installed on the computer PC or UNIX workstation if the ISSPA GUI is to be used To run the ISSPA GUI within MATLAB the MATLABPATH has to be extended to include the UKL ISSPA SRC or UKL ISSPA PCODE MATLAB and the UKL ISSPA TBOX directories ref chap 1 After MATLAB start the actual setting of the MATLABPATH can be checked at any time by pressing path The user must also set the operating system program search path to include UKL ISSPA EXE and UKL TBOX Refer to the User s Manual of your computer to properly set the program search path The ISSPA GUL is started by pressing isspa on the MATLAB command window After pressing isspa the ISSPA Plot window and the ISSPA Main window become visible first From these windows all other ISSPA windows and all ISSPA commands are accessible using the respective pulldown menus or buttons To exit ISSPA use the Mange pulldown menu of an JSSPA window The ISSPA GUI consists of several subroutines which are called from the ISSPA windows Almost all variables used in the ISSPA GUI are local variables which will be cleared after an ISSPA task has been performed so that the user has no access to any of these variables Th
27. composition of the measurement matrices to determine the number r of effective modes in the identification range rank of measurement matrices Condensation of measurement matrices to r effective degrees of freedom Input ispkon ein ISPKON control parameters or interactively no of effective modes Output ispkon aus listing file ispkon_01 mat plot file modified measurement Limitations Position in ISSPA sequence post FITKOR ISPMOD if no residual correction is required pre ISSPA ISSPAO2 User s Guide 5 ISSPA Modules 42 3 4 ISSPA Purpose Identification of modal parameters from modified measurement matrices eigenfrequencies eigenvectors modal masses non diagonal modal damping matrix Input Output isspa aus listing file Limitations Position in ISSPA sequence post ISPKON pre NOMOIS ISSPAO2 User s Guide 5 ISSPA Modules 3 5 NOMOIS Purpose Assembly of the inverse stiffness and mass matrices Deletion of noise modes Input nomois ein NOMOIS control parameters or interactively no of ISSPA mode shapes to be deleted Output nomois aus listing file Limitations Position in ISSPA sequence post ISSPA pre ISPRES DAMPF 43 ISSPAO2 User s Guide 5 ISSPA Modules 44 3 6 DAMPF Purpose Damping matrix improvement using a modal approach Input dampf ein DAMPF control parameters or interactively no of spectral lines left right to eigenfrequency to use for improving t
28. ctral lines to the left and right of the estimated eigenfrequencies are used at the first iteration steps Finally all spectral lines are used if the iteration converges IREDN reserved Residual Terms ISSPAO2 User s Guide 6 Using the ISSPA GUI 70 IRES IRES 1 residual displacement only IRES 2 residual displacement residual acceleration IRES 2 residual acceleration only estimated eigenfrequencies viscous damping Percent of critical damping input of MFG estimated eigenfrequencies and viscous damping factors After supplying all needed input the user must press write to store the actual settings Create FITKOR EIN d project isspa garteur b001 FITKOR EIN File D project isspa garteur b001 fitkor ein No of estimated modes MFG 1 Iteration step limits ITDMAX ITFMAX Data Reduction IREDM IREDN Residual Terms IRES estimated eigenfrequencies viscous damping Percent of critical damping 0 000000e 000 WRITE eee Fig 6 2 9 Setting of the fitkor ein file The user may also edit the fitkor ein file This ASCII file is of type UKL FR3 To close the Create FITKOR EIN window press close ISSPAO2 User s Guide 6 Using the ISSPA GUI 71 Step7 Create ISSPA identification subdirectories b001 b0xx bemb To create ISSPA identification subdirectories b001 bOxx bcmb use the create ISSPA subdirs button of the JSSPA Start window 4 ISSPA Question X How many I
29. de indicator functions Step 9 5 Save MIF plot to select identification ranges ISSPA Plot pulldown menu Keep Figure Parameter lower subplot only Note Window name JSSPA PLOT is changed to Figure No 1 ISSPA Plot lower plot of fig 6 2 19 ISSPAO2 User s Guide 6 Using the ISSPA GUI 78 Step10 identify first mode in subdir b001 Step 10 1 Change the ISSPA Working Directory to b001 Step 10 2 Select identification range around first eigenfrequency from Figure No 1 ISSPA Plot window ISSPA Select identification ranges En ISSPA Plot d project isspa garteur b001 BEE Manage Plot ISSPA Tag Window Print Help Measured Response FITKOR all MDOF f Hz DAMPF modify BZ ar residuals RSC Mee all MDOF fIHz MDOF 1 MDOF Envelope amp Single Mode Indicator R 60 NUM ba a 50 MStart SA MATLAB C QJ Figure No FJISSPA Line BJISSPA Pi BS LView Pro via 12 13 Fig 6 2 19 Select identification range near first eigenfrequency ISSPAO2 User s Guide 6 Using the ISSPA GUI Step 10 3 FITKOR identification Pick Eigenfrequency in ISSPA Plot ISSPA Pick Eigenfrequencies After Picking the first eigenfrequency FITKOR is automatically started 4 ISSPA Plot d project isspa garteur b001 Manage Plot ISSPA Tag Window Print Help Compare of Original amp FITKOR Residual Corr Response MDOF
30. dify residuals 6 Using the ISSPA GUI 58 Start ISPMOD Plot measurement data Plot Mode Indicator Function Start FITKOR Plot measurement data residual corrected data Autorun ISPKON ISSPA NOMOIS Start ISPKON Plot substitute measurement data Plot substitute measurement data measurement data Start ISSPA Start NOMOIS Start DAMPF Start ISPRES Plot recalculated data Plot recalculated data measurement data Modify residual terms create Create residual terms after identification from the deviation of measured and delete ISSPA XL FITKOR NP Purge edit ein aus ISPMOD main start tools ctr lin cmb plot id log list recalculated response Delete residual terms for development use only for development use only Delete ISSPA intermediate results edit or type ISSPA module input output files Edit ispmod ein ispmod aus type ispmod aus on MATLAB command window Switch ISSPA window Fig 6 1 6 ISSPA Linear Identification window ISSPAO2 User s Guide 6 Using the ISSPA GUI 59 F ISSPA Combine MIFA Buttons Manage Help create fuelleis ein Create Modify fuelleis ein ein aus disp_all Edit fuelleis ein fuelleis aus type fuelleis aus on MATLAB command window add residuals add residual terms from existing identification subdirectories create Create residual terms after identification from the deviation of measured and recalculated response delete Delete residual
31. dow and the ISSPA Main window or the pulldown menu Manage fig 5 1 2 Ex EEF SEs ISSPA Main PiE E Manage Help ISSPA ISSPAN2_01_b001 UKL Root Edition actual working directory d pr garteur bO01 ISSPA02_01_b001 UKL Root Edition Ready NUM MStart A SS OQ ultraEdit 32 WF Microsoft AMATLAB C Y Explorer D ER ssPa m EissPa Tool E Lview Pro m OM 10 54 Fig 6 2 5 The ISSPA Plot window and ISSPA Main window ISSPAO2 User s Guide 6 Using the ISSPA GUI 67 Step5 Create ispmod ein in subdirectory b001 To create an ispmod ein file use the create ispmod ein button of the ISSPA Start window Create ISPMOD EIN d project isspa garteur b001 ISPMOD EIN File Data Dimensions M N Data Type IHARMO DATA Options IRESP 0 Print Options INFUNY Runtime Options IDISP File Options IUR2 Plot Data Truncation Limit TRUMLIM jo fo excitation forces INFNR RKRAFT frequency range selection IANF IEND IPEAK Saag SSS 0 0 M CLOSE Fig 6 2 6 The Create ispmod ein window to create modify the ispmod ein file The user must carefully fill out this template ISPMOD EIN File press the New button to select the location of the ispmod ein file if the ispmod ein file already exists use the Browse button ISSPA Data File press the Browse button to select the data file D project isspa garteur data man_12z unv Filetype the filetype is set automatic
32. e ISSPA GUI also uses a set of global variables which are set when the ISSPA start command isspa 1s pressed and cleared after the ISSPA exit command is executed These global variables are of no further relevance to the user You can list these variables by pressing whos global on the Matlab Command window after ISSPA was started The Matlab Command window is available at any time so that the user may perform other calculations within MATLAB e g MATFEM while the ISSPA windows are active All data and identification results are stored in data files which are written in the UKL FR3 format or the MATLAB mat file format level 1 compatible to MATLAB Version 4 and 5 Refer to app A C for a detailed listing of the file format and index The user can read from these data files by using the readfr3 command e g read eigenvalues real imaginary part and right and left hand eigenvectors from the ISSPA data file nomo12 fr3 lam_re lam_im Y X readfr3 nomo12 fr3 Press help readfr3 for detailed information about the input output arguments of this MATLAB call ISSPAO2 User s Guide 6 Using the ISSPA GUI 50 Data from MATLAB mat files can be loaded using the load command e g load measurement data of the actual identification range load ispmod01 Refer to the MATLAB User s Guide or press help load for detailed information about the load command There are several ISSPA pulldown menus available in the ISSPA windows which main conten
33. g e g a single shaker or a modal impulse hammer Single point excitation at MDOF k with force Fx 1 In this case the response is equivalent to the k th column of the frequency response functions FRF matrix U jo Hijo Note If more than one exciter configuration was used for testing like in classical modal survey testing with force appropriation the test data must be analyzed separately for each configuration ISSPAO2 User s Guide 4 ISSPA Input Features 29 Case Il Harmonic constant single axis base acceleration Data preparation outside ISSPA necessary case II is thereby reduced to case I ISSPA control parameters IHARMO 1 IFUSS 0 It is assumed that the following basic data is available ref fig 4 1 e the measured acceleration frequency response of N degrees of freedom MDOFs well distributed on the structure to gain a sufficient spatial resolution The responses are measured at m frequency points spectral lines e the measured acceleration frequency response of the pilot accelerometers mounted on the rigid interface shaker structure e if available measured reaction forces i e measured by a force measurement device 15 17 e if available the mass matrix of a FE model reduced to N MDOFs E BET wv xa x Sfructure rigid adap fer Shabel Sable pilot dot Fig 4 1 Typical test configuration used in base excitation testing ISSPAO2 User s Guide 4 ISSPA Input Feature
34. he modal damping matrix Output dampf aus listing file Limitations Position in ISSPA sequence post NOMOIS pre ISPRES ISSPAO2 User s Guide 5 ISSPA Modules 3 7 ISPRES Purpose Recalculation using identified data Input ispres ein ISPRES control parameters or interactively no of modes to be suppressed in the calculation include residual effects if previously calculated by FITKOR Output ispres aus listing file ispres_Ol mat recalculated response Limitations Position in ISSPA sequence post FITKOR NOMOIS pre 45 ISSPAO2 User s Guide 5 ISSPA Modules 46 3 8 FUELLEIS Purpose Assembly of identification results of different identification ranges Input fuelleis ein FUELLEIS control parameters Output fuelleis aus listing file fuel fr3 modal parameters and inverse system matrices Limitations Position in ISSPA sequence post FITKOR of single identification ranges NOMOIS pre ISPRES of combined identification range ISSPAO2 User s Guide 5 ISSPA Modules 3 9 ISPGEN Purpose Calculation of dynamic responses from given modal data Input ispgen ein ISPGEN control parameters modal parameters dynamic response Output ispgen aus listing file isspa2is ein ISSPA data file Limitations Position in ISSPA sequence post pre 47 ISSPAO2 User s Guide 5 ISSPA Modules 48 ISSPAO2 User s Guide 6 Using the ISSPA GUI 49 6 Using the ISSPA GUI 6 1 General A graphi
35. in general incomplete physical inverse stiffness matrix spectral synthesis in general incomplete If required this data file can be read directly into MATLAB by using the readfr3 command eigfreq mue xsi phi invM invK readfr3 fuel fr3 i B T Note The modal matrix is stored eigenvector by eigenvector i e is stored ISSPA99 User s Guide 7 ISSPA Main Results Files 84 example 6 File FUEL FR3 created from module FUELLEIS 1 4R 4 eigenfrequencies Hz 0 6496132E 01 0 1635492E 02 0 3342412E 02 0 3398502E 02 1 4R 4 modal masses 0 4151932E 01 0 4419946E 01 0 7181810E 00 0 6889081E 00 Hi 4R 4 diagonal of modal damping matrix Percent 0 9389533E 00 0 1211294E 01 0 6741044E 00 0 1190145E 01 4 24R 4 modal matrix transposed eigenvector by eigenvector 0 9961783E 00 0 9788201E 00 0 3004567E 02 0 9712480E 00 0 2164015E 01 0 3403399E 00 0 3887944E 01 0 9964801E 00 0 1000000E 01 0 1242326E 01 0 9955027E 00 0 1300680E 01 0 3474772E 00 0 3537678E 01 0 9509832E 02 0 1743245E 02 0 2338367E 00 0 1672707E 00 0 1111940E 02 0 6038030E 04 0 5121520E 01 0 6256211E 01 0 4697288E 01 0 6422953E 01 0 4433073E 00 0 4837220E 00 0 1392432E 00 0 5305442E 00 0 8225252E 01 0 1858084E 00 0 3741631E 00 0 4535891E 00 0 4970197E 00 0 1481616E 00 0 5321296E 00 0 7900893E 01 0 1779927E 00 0 3722551E 00 0 1768326E 05 0 1437991E 01 0 5258517E 02 0 2225824E 02 0 1869861E 00 0 1000000E 01 0 1225540E 01 0 7815332E 00 0 4
36. l fr3 file format options IDPFD INVMA INVSTEN active List Subdirectories ABO E oo E oo E oo E oos E oos E o E oos E 009 E oo e CLOSE Fig 6 2 14 The Create fuelleis ein window to create modify the fuelleis ein file The user must carefully fill out this template FUELLEIS EIN File press the New button to select the location of the fuelleis ein file if the fuelleis ein file already exists use the Browse button fuel fr3 format options IDPFD IDPFD 0 save complete modal damping matrix IDPFD 1 save diagonal of modal damping matrix INVMA INVMA 1 assemble and save inverse mass matrix INVSTEI INVSTEI 1 assemble and save inverse stiffness matrix active select deselect ISSPA identification subdirectory list list actual results of the ISSPA identification subdirectory prev next scroll ISSPA identification subdirectories After supplying all needed input the user must press write to store the actual settings ISSPAO2 User s Guide 6 Using the ISSPA GUI 75 Speicher in bemb al c Dateiname fuelteis ein Dateityp ein v Abbrechen ZL Fig 6 2 15 Select directory of fuelleis ein file Create FUELLEIS EIN d project isspa garteur b001 FUELLEIS EIN File spa garteur bemb fuelleis ein fuel fr3 file format options IDPFD INVMA INVSTEN nl List Subdirectories pOO1 E oo E oos E oo E oos E oos E oo E oos E oo E oo active WRITE CLOSE Fig
37. l parameters can be combined to an overall identification results file using the ISSPA module FUELLEIS This combination should also be performed in a separate directory e g bomb The ISSPA GUI provides on the JSSPA START and ISSPA COMBINE windows basic tools to properly perform these data management tasks ISSPAO2 User s Guide 3 Theoretical Background 3 Theoretical Background Note The ISSPA User s Guide extract does not comprise chapter 3 of the original ISSPA User s Guide For detailed information about the ISSPA theory refer to a b c d Link M Theory of a Method for Identifying Incomplete Matrices from Vibration Test Data Z d Flugwissenschaft und Weltraumforschung ZFW 9 1985 Heft 2 3 Kr tzig Meskouris Link Baudynamik und Systemidentifikation in Der Ingenieurbau Grundwissen Band 5 Baustatik Baudynamik Mehlhorn Hrsg Berlin 1995 Ernst amp Sohn ISBN 3 433 01571 6 Link M Qian G Identification of Dynamic Models for Substructure Synthesis Using Base Excitation and Measured Reaction Forces Revue Francaise de Mecanique No 1 1994 Kasai T and Link M Identification of Non Proportional Modal Damping Matrix and Real Normal Modes Mech Systems amp Signal processing Vol 16 No 6 921 934 2002 ISSPAO2 User s Guide 4 ISSPA Input Features 27 4 ISSPA Input Features The ISSPA identification procedure is based on the linear equation of motion in the fre
38. mpared to Phase Resonance Testing An Assessment Based on Practical Application International Modal Analysis Conference IMAC London 1987 Potter R Richardson M Identification of the Modal Properties of an Elastic Structure from Measured Transfer Function Data 20th I S A Albuquerque N M 1974 Goyder H G D Structural Modeling by the Curve Fitting of Measured Frequency Response Data Institute of Sound and Vibration Research Technical Report 87 1967 Link M Badenhausen K Identification and Dynamic Condensation of Physical System Matrices Using Incomplete Dynamic Response Data Second International Symposium on Aeroelasticity and Structural Dynamics Aachen 1985 DGLR Bericht 85 2 ISBN 3 922010 28 8 Caesar B Baier H Badenhausen K Link M H ners H Erben M Procedures for Updating Dynamic Mathematical Models Final Report ESTEC Contract No 5597 83 NL PB SC 1985 Link M On the Determination of the Number of Effective Modes from Vibration Test Data Structural Safety Evaluation based on System Identification Approaches H G Natke J T P Yao eds Vieweg Verlag Braunschweig 1988 MATLAB High Performance Numeric Computation and Visualization Software Reference Guide Mathworks Inc 24 Prime Park Way Natick Mass 1993 User s guide DYNAWORKS 4 0 Chapter 6 Modal analysis Issue 1 0 of 20 11 96 ITS ISSPA99 User s Guide 8 Reference and Related Literature 86 12 13 14 1
39. n DOFs the pseudo inversion UA T UP 4 2 ISSPAO2 User s Guide 4 ISSPA Input Features 31 where T pseudo inverse of TP relates the base excitation U and the measured pilot responses UP ina least squares sense Example Determination of base excitation in X Z plane at reference A from 4 pilot MDOFs Fig 4 2 Base excitation in X Z plane Eq 4 1 yields N 1 ab u O b u S ls 4 3 X 435 p 0 b Z2 TP The pseudo inverse of TP is given by x zj T TP TP 4 4 ISSPAO2 User s Guide 4 ISSPA Input Features From that the base excitation at reference A can be calculated Ur gt Ut UP a uP 2 Uy E v2 u In the special case when the base rotations can be neglected A ER A U U U 0 eq 4 2 yields U Xy UP Xj n px 1 Ursy U y n Yi Py 1 us S UP Zi n pz where Nox Npy Npz overall number of pilots in X Y Z 32 4 5a 4 5b 4 6 4 7a c In this case the base excitation in X Y Z is calculated from the average of the corresponding pilots Note The actual software implementation of ISSPA requires a constant base excitation level r with respect to the excitation frequencies Therefore UP should be constant within the selected frequency range i e only tests without notching should be used However if only data from tests are available where input notching was applied the response of the str
40. ools for SUN SOLARIS 2 sgi_unix ISSPA implementation tools for SILICON GRAPHICS In addition the media includes the ukl tbox a special software package which includes general data handle tools which are commonly used by several UKL software packages To run ISSPA within MATLAB the MATLABPATH has to be extended to include the following directories e ukl isspa src or ukl isspa pcode matlab e ukl tbox Note If only the pcode edition or the limited edition of ISSPA is available the user must link the corresponding subdirectory matlab51 matlab65 of the pcode directory to the MATLABPATH The user must also set the operating system program search path to include ukl isspa exe and ukl tbox Refer to the User s Manual of your computer to properly set the program search path Note ISSPAO2 User s Guide 1 ISSPA Installation 6 Your implementation of ISSPA may differ from this general directory structure due to individual requirements To check the ISSPA installation it is recommend to copy the demo directory to a temporary working directory and start the auto sequence of demonstration examples by calling demo_all within MATLAB from the actual demo directory ISSPAO2 User s Guide 2 General 7 2 General ISSPA is a software package for the identification of linear structural system parameters It is developed by the Department of Lightweight Structures and Structural Mechanics at the Department of Civil Engineering
41. quency domain Therefore the basic ISSPA input features can best be explained by the frequency response relation U jo H jo F jo 4 01 n 1 nng ngl where n no of measured degrees of freedom MDOF nr no of exciter forces nr lt n excitation frequency j vl imaginary unit H o oM K joD 4 02 nn nf columns of matrix of frequency response functions inertance w r to the complex exciter force vector F F Fy Fl M K D physical mass stiffness and damping matrix T U U at U measured complex frequency acceleration response at n MDOF The user must supply measured frequency acceleration response U jo and forcing function F j o as input to the identification procedure The type and amount of input depends on the type of the identification problem Refer to app A C for detailed input and output file specifications ISSPAO2 User s Guide 4 ISSPA Input Features 28 Casel Constant harmonic sinusoidal force excitation with n exciter forces ISSPA control parameters IHARMO 1 IFUSS 0 Input a force vector F F Fx Foe ik real part F F const for all excitation frequencies imaginary part F 0 b response j equivalent to H jo F Output a eigenfrequencies eigenvectors modal masses modal damping factors b inverse physical stiffness and mass matrix incomplete Special case FRF data provided by a standard data acquisition unit usin
42. s 30 There are three basic steps necessary to prepare the measured data from a sine base excitation test for the ISSPA identification procedure 1 Calculation of the base excitation at reference A from the pilot pickups 2 Calculation of the relative response at the structural MDOFs with respect to the base A 3 Definition of excitation forces Note All steps must be performed outside ISSPA prior to an ISSPA identification 1 Calculation of the base excitation at reference A from the pilot pickups In base excitation testing there are a number of pilot pickups mounted on the shaker table During test they are used as a part of the shaker control system For the subsequent identification process they provide all necessary information to calculate the applied base excitation with respect to the mounting point of the test structure point A in figs 4 1 and 4 2 If the adapter and shaker table is rigid the relation between the pilot responses UP lo and the base excitation U lo is given for each excitation frequency by A uy g A P p p y u SPO Oe Or eee 4 1 1 u p _yP XX uP 001 yi x 0 A 1 u yy A UL u T uA where ub response of the i th pilot measured in x direction 1 xP 3 P zP distance of the i th pilot location and reference A ie Yi oi p u excitation at reference A in x direction i number of pilot pickup i 1 2 np Since the number of pilot DOFs can be much larger than the number of excitatio
43. sion black white plot to printer black white plot to paste buffer black white plot to bitmap file plot to paste buffer Post script file Bitmap file UKL Logo ISSPAO2 User s Guide 6 Using the ISSPA GUI 52 Generally there are 9 ISSPA windows which will be used in an identification process ISSPA Main ISSPA Start ISSPA Tools ISSPA Control ISSPA Linear Identification ISSPA Combine ISSPA Plot ISSPA Identification Process ISSPA Identification Results The first six windows are used to control the identification run whereas the later three are used to present the identification results or to log the identification process To switch among those windows each window provides the pulldown menu Window In addition the upper six windows provide buttons in the lower part of the window to easily switch to a different window If a certain window consists of several subwindows e g ISSPA Linear Identification additional buttons are also provided in the lower part of the window to switch the subwindows The user may close or resize and reposition a window at any time The upper six windows provide a close button to close the respective window 4 ISSPA Plot c user weiland m lab isspa demo fhg5 b003 Manage Plot ISSPA Tag Window Print Help ISSPA02_01_b001 UKL Root Edition Select Working Directory ISSPA Release Buttons Select Working Directory Fig 6 1 1 ISSPA Plot window ISSPAO2 User s Guide 6 Using the ISS
44. ts are listed below ISSPA Pulldown Menus Manage Select Working Directory Exit ISSPA Plot Close Figure Keep Figure keep actual ISSPA plot as separate figure Menubar on off MATLAB Menubar Select Zoom Zoom frequency axis graphically Zoom off Select MDOF MDOF_SET_1 user defined MDOF set MDOF_SET_2 MDOF_SET_3 all MDOF Scan MDOF Display MDOF one by one Scales View Real Imaginary Amplitude Phase Nyquist Waterfall Waterfall of abs Uim Parameter log x logy log z Rotate rotate subplot graphically all subplot display all subplots of actual figure upper subplot only lower subplot only ISSPAO2 User s Guide ISSPA Select all available frequencies Select identification ranges data reduction Pick eigenfrequencies ISSPA Control Run FITKOR Run ISSPALL Run DAMPF Run ISPRES EF Blowup Tag Tag estimated eigenfrequencies Tag identified eigenfrequencies Window ISSPA Main ISSPA Start ISSPA Tools ISSPA Control ISSPA Linear Identification ISSPA Nonlinear Identification ISSPA Combine ISSPA Plot ISSPA Identification Process ISSPA Identification Results Print b w Printer b w Paste b w bmp Paste DS bmp Help About 6 Using the ISSPA GUI 51 specify identification range perform automatic data reduction view ISSPA Control window Mode shape animation using a MATFEM FE model Only if UKL MATFEM is installed view ISSPA window not activated in the actual ISSPA ver
45. uctural MDOFs U must be related to the base excitation by ISSPAO2 User s Guide 4 ISSPA Input Features 33 U Uy Ur A U Uy U A Uyy Ny UR Un N u m U The transfer function matrix H contains all structural information and therefore can be used for modal identification It can be calculated from H UU 4 9 where 7 u pseudo inverse of u In the general case of a base excitation in 3 D space six independent data sets test runs are required to calculate Ur However if the base excitation can be considered as a single axis RE A A A A _ A _ A _ excitation e g U ZU Le U U De U Q a column of the transfer function matrix can be calculated simply from the complex division of the structural MDOFs by the corresponding base excitation e g Ulo J Hea o A 4 10 z U o where Ho A column of H with respect to a base excitation in Z direction at reference A j number of frequency point j 1 2 m This column represents the response of the structure with respect to a unit single axis base excitation e g in Z Hence all structural MDOFs which direction coincide with the base excitation direction must exhibit a transfer function of 1 in the low frequency range 0 lt lt first elastic structural mode This is due to the rigid body movement of the structure in this frequency range In the most practical case where a pure single axis base excitation is not achieved eq 4 10
46. ust write a special program to convert the actual data format to a type which is supported by ISSPA e g 12 It is strongly recommended to use MATLAB m file programming for this task Special MATLAB m files are available to easily write and check a proper UKL FR3 file writefr3 readfr3 which then can be converted from formatted to unformatted UKL UR3 file formum Best performance can be expected if the data is available as a UKL UR3 file Most internal data flow is performed using this file format But care should be taken with this file format if the data is to be transferred between different operating systems like WINDOWS and UNIX because unformatted files are not exchangeable due to the different data representation The UR3 files can be converted to FR3 at any time using formum example data directory D project isspa garteur data data file D project isspa garteur data man_12z unv Type SDRC Universal files Type 58 EN Explorer D project isspa garteur data Of x Datei Bearbeiten Ansicht Wechselnzu Favoriten Extras Kal a gt a x 2 Zur ck Vorw rts Aufw rts Ausschneiden Kopieren Einf gen R ckg ngig L schen Eigenschaften Adresse D project isspa garteur data Ordner x asare Mentel Typ erinran A Desktop Bl 2587 KB Datei UNV 20 07 99 20 02 S E Arbeitsplatz H E 3 5 Diskette A Festplatte C o Wechseldatentrager D a project B L isspa EI Bu garteur C boo1 data 1
47. vailable However this mass matrix must represent sufficiently the inertia properties of the structure The excitation forces are calculated by F MT 4 16 where M analytical mass matrix with respect to the MDOFs F excitation force vector T rigid body transformation matrix acc eq 4 13 Example The excitation forces for the structural MDOFs in X Y and Z at the first measurement point with an analytical lumped mass m at that point are calculated from A FEN A Fy My is T AU ze yt F m us et UA 4 17a c N BEN u Fa m E yiUxx UA Z ISSPAO2 User s Guide 4 ISSPA Input Features 37 After data preparation of the raw base excitation test data ISSPA can be applied to identify the modal parameters ISSPA control parameters IHARMO 1 IFUSS 0 Input al force vector fasts F MT or a2 if analytical mass matrix M is not available F 1 dummy unit force applied at an arbitrary MDOF p rel b relative response with respect to the base U Output a eigenfrequencies eigenvectors modal damping factors if al is available modal masses Special case Reference Link M Ulrich H Weiland M ISSPA Guidelines Modal Identification using Base Excitation Test Data Laboratory of Lightweight Structures and Structural Mechanics UKL Kassel April 1999 ISSPAO2 User s Guide 4 ISSPA Input Features 38 ISSPAO2 User s Guide 5 ISSPA Modules 39 5 ISSPA Modules 5 1
48. window ISSPAO2 User s Guide 6 Using the ISSPA GUI 56 ISSPA Control i l E3 Manage Help Buttons ise Ate an close Close ISSPA Control window Command Window DISP 2 all ISSPA exe control DOS Window Display channel of run time messages during an ISSPA module run IDISP Switch to control the amount of run AUTORUN time a during an ISSPA SPMOD module run AUTORUN ISPMOD Start ISPMOD and plot data if aurorior oo EBENE FITKOR Start FITKOR after Pick Eigenfrequencies was called from ISSPA pulldown menu AUTOPLOT Plot after termination of module ISPMOD measured response FITKOR comparison of measured and recalculated response ISSPA XL for development use only main lin cmb id log main Start tools ctr lin cmb plot id log list Fig 6 1 5 ISSPA Control window Switch ISSPA window ISSPAO2 User s Guide 6 Using the ISSPA GUI 57 ISSPA Linear Id RIES Manage ISSPA Linear Id RITES Manage Help ISPMOD ISPMOD MIF ISSPALL ge NOMOIS DAMPF DEIN N create delete FITKOR NP id log ist close subwindow control subwindow edit Buttons ctr edit Switch ISSPA Linear Identification control or edit subwindow main Start tools ctr Switch ISSPA window lin cmb plot id log list close Close ISSPA Linear Identification window ISSPAO2 User s Guide cir ISPMOD Plot MIF FITKOR Cmp ISSPALL ISPKON Plot Cmp ISSPA NOMOIS DAMPF ISPRES mo

ISSPA User`s Guide

Contents

Download Pdf Manuals

Related Search

Related Contents