Home

A Strain-Life Module for CGAP - Theory, User Guide and Examples

1. Theinput fileis atext document with the extension input Table 3 gives an example of the input file and an associated list of the source code variables 21 DSTO TR 2392 Table3 An exampleoftheinput data fileusein CGA P Ge Example input enen Line description 1 generic test seq JTITLE Inputtitle 2 al IPT1 Material data output information 3 12 IPT2 Analysis Output Information 4 T IPT3 Loads Data Format ASCII or Binary 5 0 IPT4 Damage Source Indicator 6 0 IPT5 N umber of Passes 7 1 KN Notch Factor 8 D7075 T651 MATERIAL Material Identification 9 d MATF Material Data Type LO UNITCV Unit Conversion 1 0 RS Residual Stress 2 2 NMLT Number of Stress Multipliers 3 235 5 SREF Stress Multipliers 4 le 007 DCUT 9 Damage Truncation Level 5 2 Equivalent strain equation 6 Flight Hours 1 Spectrum file total flight hours 17 konya ES Ge Originally in a separate file ftn15 dat Here Flight Hours and Maximum Delta Stress are keywords and must be entered as shown Lines 1 14 must be entered in the sequence shown Lines 15 17 may be entered in any order or any line may be omitted When omitted the equivalent strain equation defaults to 2 the flight hours default to 1 and the maximum delta stress defaults to 60000 kk Input Title On thefirst line any information can be input up to a length of 80 characters Material D ata Output Information This line contains an integer number t
2. DSTO TR 2392 5 7 Analysis Using FAM SH Performing the analysis using CGAP is rather simple Oncethe relevant data has been input into CGAP all the user needs to do is to press the Build button to build the case and then press Run to executethe analysis see Figure 23 File Edit View Commands Tools Help a E x a Ak A ST op New Save Remove Import Export Save Out Build Run Configure Figure23 CGAP menu bar Active analysis is indicated by the progress bar at the bottom of the window see Figure 24 Progress MP SSGGUHHRRRRRRRREE Figure 24 Progress bar When the analysis is complete a message will be displayed in the messages pane at the bottom of the window 5 8 FAMSH output Several outputs are created by FAM SH A brief description of each of the output files is given in Table 9 Examples of the output are provided in Appendix C Table9 FAM SH output files Output File Description Example filename dmo The dmo file contains the damage matrix The damage matrix Table 23 provides the user with information relating to the distribution of damage based on the mission types This information is only available when the DSA format is used filename dsout This file records the damage density table calculated using the Table 24 maximum change in stress defined in the ftn15fileor input file The dsout file is equivalent to the ftn13 file produced by FA MS filename passes The first line of this file contain
3. flmatl dat tabulated data or the flmat2 dat equation specification files These files are generated automatically when the user selects a 6 Damage Source Assignments were developed for the P 3C SLAP See reference 2 for more information 7 The Damage Source Assignment was a modification made by Lockhead Martin to examine the proportions of damage associated with different in flight manoeuvres 20 DSTO TR 2392 material from the material databasein the CGAP GUI Thesefiles can be used as inputs using the original FAMSH executable The single number contained in ftn15 is used to generate the damage distribution table and does not affect the calculated fatigue life The stress levels are categorised into 20 bins using this value By default CGAP chooses 60000 if no number has been specified 5 2 1 Units To maintain consistency and backward compatibility with older versions of the FAMS and FAMSH codes it is recommended that the following units be used when entering data into CGAP Table2 Recommended input units for CGA P Input parameter Units Young s Modulus ksi Fracture Strength ksi Proportional Limit ksi Ultimate Strength ksi PTQI 1 ksi All stresses in the stress strain or stress life table ksi Residual Stress Resd ksi MAXRNG psi Spectrum file psi 5 2 2 Importing an Analysis File Importing data can be accomplished using the O pen button on the top tool bar or by clicking File gt 0 pen
4. 2 92 3 93 6 93 6 94 1 2 94 6 95 SKS 1 EEN 602 1100 1340 2980 3640 4450 6000 7330 9900 12100 14800 22000 32900 54200 109000 243000 540000 68 82 0 322 0 1573 214 w5 494 493 427 GG ECH 542 314 743 59 031 99 886 711 485 216 911 576 828 421 996 554 096 625 142 649 149 644 41 646 176 79 435 yo 1 DSTO Trendline 71 DSTO TR 2392 72 DSTO TR 2392 Appendix E FAMSH Test Cases To verify the FAMSH module a number of test cases were developed Some of thetest cases have been included here All the test cases were evaluated using the original FAMSH code and compared to the results produced by CGAP All cases returned consistent results N ote These models are only to be used for validation and not for any other purposes M aterial data and parameters should not be transferred to solve real engineering problems The attached CDROM contains all the CGAP strain life module validation cases Table 28 FAM SH Test Case descriptions Case No Test Description 1 LO O JO Un P N OH Hi Hi 0 SS Q N Pi CO o gt N N N N S N N N F LF DD JO Un Q N F O O ON Uses a moderate sized spectrum file in conjunction with two reference stress values This case similar to the first uses a smaller version of the same spectrum file this time only containing 2 flights This case uses four reference stresses and is designed to fail on the last refer
5. 23 DSTO TR 2392 Residual Stress This line contains the magnitude of the residual stress needed in the analysis If no residual stress is to be used then enter 0 0 The units of the residual stress must be consistent with the units of the proportional limit that is ksi Number of Multipliers This integer parameter is used to specify the number of Stress multipliers to use The multipliers are specified on the next line of the input data file Stress Multipliers An array isused to storea maximum of 10real numbers Each number in the array specifies a multiplication factor to usein conjunction with thenotch factor Thenotch factor is multiplied by one of the stress multipliers which in turn is used to scalethe applied spectrum A nalysis occurs for each of the stress multipliers If the notch factor is given a value of 1 then stress multipliers can be used to specify a series of different notch stress concentration factors Truncation Level This line contains areal number that is used to denotethe user specified truncation level This truncation level truncates the spectrum based on the associated damage calculated for each cycle Thenumber used hereisa damagelevel below which cycles with damagelessthan this value are ignored Equivalent Strain Equation An integer number is used to select the equivalent strain equation to use in the analysis Several equivalent strain equations are available for the user and are as follows Section 4 5
6. Australian Government Department of Defence Defence Science and Technology Organisation A Strain Life Module for CGA P Theory User Guide and Examples Chris Wallbrink and W eiping Hu Air Vehides Division Defence Sdence and Technology Organisation DSTO TR 2392 ABSTRACT Existing airworthiness standards require that all aspects of fatigue be considered in the design development and maintenance of aircraft structures To minimise costs and improve combat readiness requires accurate and efficient fatigue life evaluation This report details the introduction of a strain life algorithm known as FAMSH into the software tool suite known as CGAP developed and maintained at the DSTO Along with improvements to the original FA MSH code this most recent development introduces a materials database and dynamic memory allocation With other additional improvements the latest rel ease of the CGA P environment seeks to improvethe ease with which theengineer is ableto accurately and reliably conduct fatigue life analysis Contained within this report is an explanation of the strain life theory in conjunction with a detailed description of its implementation in the CGAP GUI environment Examples are provided to guide the user through its operation RELEASE LIMITATION Approved for public release Published by Air Vehicles Division DSTO Defence Science and Technology Organisation 506 Lorimer St Fishermans Bend Victoria 3207 Australia Teleph
7. Data Set 2 Region 2 Data Set 3 Region 3 600 0 0087 700 0 0083 900 0 0078 1500 0 007 2000 0 0067 3000 0 0063 5000 0 0056 40000 0 0032 1000000 0 0015 5000 0 0056 40000 0 0032 1000000 0 0015 In this example PLET 0 0105 and ELET 0 0063 There are 8 other unknown parameters to be determined and they are C1 C2 B1 G1 C3 C4 B2 and G2 Perhaps the easiest method to determinethese parametersisto usea non linear curvefitting program such asthat contained in Origin 6 0 42 If you use Origin 6 0 you will need to create the function used in the fitting process The inputs should look something like that shown in Figure 37 62 DSTO TR 2392 NonLinear Curve Fitting Edit Function DAR Function Action Options Scripts Name Strainlite Type UserDefned v Number of Parameters M User Defined Param Names Parameter Names c1 c2 b1 91 Independent Var fk Dependent Var ly Definition y 014 b1 c24x g1 File Form Y Script y Save Cancel Function formula user defined function description built in Basic Mode lor external DLL function Figure 37 User defined non linear curvefitting function Using the non linear curvefitting program determinethe parameters C1 C4 BI and G2 using the second data set As we have chosen PLET to be greater than ELET Region 2 is defined with the parameters C1 C4 Bl and G2 If PLET was less than ELET then Region 2
8. on the top menu and select Save As Save the input file in the same location as the spectrum file and give it the same name as the spectrum file Name the file Gen seq input see Figure 30 DSTO TR 2392 Save in FAMSH w My Recent Documents File name Gen seqinpul vi Save astype CGAP Input File vi Cancel Figure 30 The save as dialog box The input file looks similar to that presented in Table 18 Table18 Theinput file Gen seq input Crack Growth Analysis Program 11 0 7050 T7351 2 N NO 9 5 e 007 2 Flight Hours 100 7 3 Execution Once the files have been created click on Build to build the case see Figure 31 File Edit View Commands Tools Help a a x ea n W ST CH New Save Remove Import Export Save Out Build Run Configure Figure 31 CGAP menu bar You should see a new label underneath cases on the left hand of the screen see Figure 32 49 DSTO TR 2392 File Edit View Commands Tools Help a a B x h B gt New Open Save Remove Save Out Build Run Material Load Case Cor Den set Crack Growth Analysis lt Material Database 7075 1651 vi x Material Name DEC 1979 SAR 79 4 Static Properties Young s Modulus 10300 ksi r ssa Figure 32 A new case underneath the cases control box Multiple cases can be loaded into CGAP if need be After the problem has been successfully built the fina
9. 4 100 0 CYCLE COUNT amp FLIGHT DURATION 27001022 12000 12000 0 5 27002019 12000 6000 0 20 27003005 10000 8000 0 1000 27004019 12000 6000 0 20 As this problem uses DSA s it requires an extra file to describe the DSA codes This file is called ftnO7 and is also a text file For this problem create a file with the following information in Table 17 and save it with the name ftn07 Table17 Theftn07 file Ftn07 MISSIONS 27 DAMAGE SOURCES 3 005 R 0 8 019 R 0 5 022 R 1 To create the input file you can either create it using a text editor or using CGAP To use CGAP open CGAP and enter the FAMSH module as described in section 5 1 By default the materials tab should be displayed Under the materials database dropdown box select 7050 T7351 as shown in Figure 26 Material Load Case Contre Material Database z x Figure26 M aterials tab M aterials D atabase drop down menu The materials tab will look likeFigure27 with thetabular material description active Both the Cyclic Stress Strain Curve and Strain Life Curve figures should now be populated DSTO TR 2392 Material Load Case Control Output Matora Database Cyclic Stress Strain Curve 7050 17351 z x 145 Unit Convtr 1 poer Sirain for 7050 17351 Material Name AL ALLOY NORTHR PREC 116 Static Properties Young s Modulus foo ksi 2 e Fracture Strength S Mou Proportiona
10. An example of the damage density information contained in the filename dsout file Table25 An example of Example passes thefilename passes file Example dsout FAMSH PROGRAM INITIATED 20090604 Title generic test seq LOADS APPLICATION NUMBER TWO KN 1 0000 SREF 2 500 KN SREF 2 500 OCCURRENCES OCCURRENCES BIN RANGE 5 IN RANGE PCT DMG REMAINING J 0 5 14 5 0 0 441 5 2 5 10 54 0 1 7 3815 3 10 15 347 0 41 6 40 5 4 154 20 34 0 22 8 6 5 5 20 25 4 5 5 6 2 0 6 E e 1530 WE 1 8 135 7 30 35 0 0 0 0 les 8 35 40 0 0 0 0 15 9 40 45 0 5 GER 10 LO 45 50 1 0 S 23 0 0 11 50 55 0 0 0 0 0 0 12 55 60 0 0 0 0 0 0 13 60 65 0 0 0 0 0 0 14 65 70 0 0 0 0 0 0 15 70 75 0 0 0 0 0 0 16 75 80 0 0 0 0 0 0 17 80 85 0 0 0 0 0 0 18 85 90 0 0 0 0 0 0 19 90 95 0 0 0 0 0 0 20 95 100 0 0 0 0 0 0 21 100 0 0 0 0 0 0 TOTAL CYCLES 456 0 TOTAL DAMAGE 1 22661E 04 MAXIMUM DELTA STRESS 60000 00 C Cgap Example 1 Exampl e Flight Hours Kn 8152 52148437500 2 500000 313 114379882812 5 000000 Table26 An example of the filenametruncspc file Example truncspc 2 1 CYCLE 0 1 000 CYCLE COUNT FLIGHT DURATION 4391 38 FLIGHT NUMBER amp MISSION TYPE 224 000 CYCLE COUNT FLIGHT DURATION 0 0 00 38001022 1 00 38001022 38002023 9068 00 38003019 6891 00 38003019 38004004 5740 00 38004005 13942 00 38004005 38004004 7405 00 38005004 13167 00 38005004 38007008 10301 00 38007007 14201 0
11. The total fatigue life of a structure may be assessed by using one of four approaches 1 by assuming the total fatigue life of a structure is governed entirely by crack initiation 2 by assuming that the total fatigue life is governed entirely by crack growth 3 by treating the total fatigue life of astructureas a serial combination of crack initiation and crack growth and finally 4 by a parallel application of both crack initiation and crack growth analyses The first approach was typically used in theearly years of fatigue assessment and is still the design practice of someindustries whereno regular inspections of thestructures are carried out The second approach which is an emerging method relies on the accurate prediction of the growth of very small cracks that takes place in the early stages of fatigue damage With the gradual improvement in short crack growth theory this approach may gain more prominence The serial combination approach which has been adopted for the P 3C fleet assessment 2 treats the early stages of fatigue damage as a process of crack initiation and then treats the subsequent fatigue damage in terms of crack growth2 Thefourth approach is being used on some later generation combat aircraft particularly where low structural maintenance must be assured Different fatigue crack initiation models are routinely used in the safe life design and the development of aircraft structural integrity management plans One of t
12. UNIQUE CYCLES REMOVED 10 10 394 451 10 00 00 29 00 00 451 00 451 451 00 00 00 00 456 00 000 LEVELS 000 LEVELS NUMBER PASSES LEVELS LEVELS NUMBER PASSES NUMBER PASSES TION NUMBER TWO for damage and truncation CYCLES REMAINING 446 446 427 62 De 5 00 ou 5 00 446 00 00 00 00 00 00 00 295 0847 313 2757 313 1144 studies 2 6 REMOVED 2 193 6 86 98 98 98 2 98 2 193 360 404 904 904 904 904 1 93 Sample flmat1 dat file 651 AL ALLOY gt D7075 T 10300 32 100 005873 006358 006850 007464 008218 009060 009969 010935 011951 013017 014128 015286 016487 017732 019020 020350 021721 023134 024587 026081 027615 029189 030804 032462 034166 035921 037739 039641 041664 043871 053135 065403 o 018 013 012 0085 008 0075 007 0065 006 0055 005 0045 004 0035 003 0025 002 0 50 000 000 1 0 010 0 00 mM SSS Oe OS Oo 0 CODO 00 Y So EE ET CA E 00 00 00 0 30 Sample fl mat2 dat file SL DEC 1979 SAR 79 4 1 31575e 005 gt 7075 T6 10300 0 01103 0 0588 DSTO TR 2392 Appendix D Sample FAMSH Materials Files 68 82 0 1 2 60 6 65 4 70 6 74 1 76 8 78 7 79 80 9 7 82 82 9 9 83 2 84 8 85 7 86 4 86 3 87 7 88 2 88 5 89 2 89 2 90 6 9315 9 91 8 92
13. 5 221042834 00 221042834 00 0 00 5 0 8424547 69 8424547 69 0 00 7 2 5 221042834 00 221042834 00 0 00 5 0 8424547 69 8424547 69 0 00 8 2 5 221042834 00 221042834 00 0 00 5 0 8424547 69 8424547 69 0 00 9 25 233007031 05 233007031 05 0 00 5 0 7181150 94 7181150 94 0 00 10 2 5 115019049 25 115019049 25 0 00 5 0 3946915 39 3946915 39 0 00 11 2 5 249151 21 249151 21 0 00 5 0 10864 74 10864 74 0 00 12 2 5 220312 25 220312 25 0 00 5 0 3457 82 3457 82 0 00 13 25 6 54 6 54 0 00 5 0 0 42 0 42 0 00 14 2 5 122287885 33 122287885 33 0 00 5 0 4696715 79 4696715 79 0 00 15 2 5 122287885 33 122287885 33 0 00 5 0 4696715 79 4696715 79 0 00 75 DSTO TR 2392 76 Case Kn CGAP1 8 FAM SH FAM SH 1 50 LINUX difference 16 2 5 122287885 33 122287885 33 0 00 5 0 4696715 79 4696715 79 0 00 17 25 122287885 33 122287885 33 0 00 4696715 79 4696715 79 0 00 5 0 18 2 5 122287885 33 122287885 33 0 00 5 0 4696715 79 4696715 79 0 00 19 2 5 122287885 33 122287885 33 0 00 5 0 4696715 79 4696715 79 0 00 20 2 5 48337584 70 48337584 70 0 00 5 0 4563759 61 4563759 61 0 00 21 2 5 122287885 33 122287885 33 0 00 5 0 4696715 79 4696715 79 0 00 22 2 5 40333647 35 40333647 35 0 00 5 0 2148989 19 2148989 19 0 00 23 2 5 316442565 90 316442565 90 0 00 5 0 10173569 74 10173569 74 0 00 24 2 5 28051631 66 28051631 66 0 00 5 0 2624609 12 2624609 12 0 00 25 2 5 22761393 09 22761393 09 0 00 5 0 2007716 34 2007716
14. 5 3 2 Creating a New Material Database Entry To create a new material enter a new name or modify an existing material ID in the Material Database dropdown box This will enabletwo dropdown menus Stress Strain 26 DSTO TR 2392 Data and D ata Type The user may now select either the tabular input or the equation input option from the Stress Strain D ata dropdown menu see Figure 14 Stress Strain Data Tabular w Tabular Figure14 Stress Strain D ata dropdown menu The Data Type dropdown menu see Figure 15 can be used to specify the form of the stress strain relationship This option only affects data entered in tabular form it does not affect the equation format The equation parameters always define the cyclic stress strain relationship Three D ata Type options are available e Hys sts v hys stn x hys sts H ysteresis stress versus hysteresis strain x hysteresis stress e Cyc stress v cyc strain Cyclic stress versus cyclic strain e Hys stress v hys strain hysteresis stress versus hysteresis strain Data Type Hys sts v hys stn x hys sts w Hus sts v hys stn x hys sts Cyc stress v cyc strain Hys stress v hys strain Figure 15 Data Type dropdown menu Data can be entered directly into the grids provided as shown in Figure 16 The number of data is specified for both the stress strain curves and the strain life curves Material Properties Number of Data Stress Strain Data Number
15. C1 B1 C4 G2 N A ELET gt PLET gt 0 f C1 B1 C2 G1 f C3 B2 C2 G1 f C3 B2 C4 G2 PLET gt ELET gt 0 f C1 B1 C2 G f C1 B1 C4 G2 f C3 B2 C4 G2 4 4 Rainflow Counting Rainflow counting is an internal function that is conducted by the FAMSH code It is performed on the local stress history derived from the applied load spectrum using the Neuber rule Theload history isinput into FAMSH as aseries of turning points Thestrain life algorithm used in FAMSH requires that this spectrum be Rainflow counted Rainflow counting has a physical basis and that is when metals are subjected to repeated loading the stress strain response forms hysteresis loops like those in Figure 8 If the loading remains in the elastic regime then the hysteresis loops remain closed such as that produced by thecycle G to H If the loads applied are large enough to cause plastic yielding then open hysteresis loops are formed It is useful to note that if the loading does not cause plasticity then any hysteresis loops formed degenerateto astraightline Any actual loop with non zero enclosed area indicates plasticity Thearea contained by thehysteresis loops represents the energy lost during that cycle In the case of a variable amplitude load sequence hysteresis loops can be drawn such that onefinds small loops contained within larger loops The Rainflow counting scheme developed by Matsuishi and Endo 14 may be used to extract the peak and valley of all the closed hysteresis loops from
16. General description not used in of TPPs Duration the analysis inthe flight 1 00 13408 00 1773 00 9068 00 Maximum Minimum Maximum Minimum ee inimu to enter up to 4TPPson TP stress TP stress TP stress TP stress E the same line 6891 00 1087 00 15104 00 5740 00 Each number should be l0spaces lOspaces lOspaces 10spaces contained within the 10 Real Real Real Real spaces 37 DSTO TR 2392 5 6 5 DBI SST input format The layout and format of the spectrum file that is required when using the DBI SST type inputs is presented in Table 8 Table8 Theformat of a FAM SH input file using DBI SST formats Example DBI_Cycle input 2 1 CYCLE Bi pu The Arbitrary Spectrum number number format of flights thatis not used 1 38 FLIGHT NUMBER amp MISSION TYPE A Flight Mission General description not used in number Type the analysis 114 1 000 TURNING POINT PAIRS FLIGHT DURATION AAA Number Flight General description not used in of TPPs Duration the analysis in the flight 0005001091 869 62 0005001091 295 63 0005004072 3884 06 0005004072 1068 51 DSA Maximum DSA Minimum psA Maximum DSA Minimum TP stress TP stress TP stress TP stress 0005004072 3884 06 0005004072 1068 51 0005004072 4819 06 0005004072 1690 94 10 spaces 10 spaces 10 spaces 10spaces 10 spaces 10 spaces 10 spaces 10 spaces Integer Real Integer Real Integer Real Integer Real OneSpace One SpaceOne Space One Space One Space One Space One Space
17. MPa a value of 6 895 should be used It does not affect the loads in the spectrum file 5 4 Load tab Theload tab see Figure 20 allows the user to select the spectrum file and to enter a number of load related parameters 29 DSTO TR 2392 CGAP1 8 beta Sample File Edit View Commands Tools Help a H x Ak kIB L e New Open Save Remove import Esport Save Out Build Rur Stop Material Load m Load Database CS XI Spectrum Binary Damage Source Assignme sel Tue MAXRNG 60000 ksi Scale 1 ksi Stress Levels No 0 Running FAMSH Databases Case Database E se amp gt Crach Configure Case Control Output x Hours per Sequence fi Block r MAXBLK H zi I Spectrum File estamp Spectrum Path Geometry Database Material Database Messages Input File Output File Plot File Spectrum File Load Database Figure 20 Theload tab used in theFAM SH CGAP module 5 4 1 Spectrum File Thespectrum fileis loaded using the Spectrum File button By default the spectrum file has the extension spectra 5 4 2 Spectrum The user must indicate if the spectrum file contains DSA information whether it isin binary or text format This is done by selecting the appropriate option under the load tab using the Spectrum dropdown menu There are four options available and shown in Figure 21 Spectrum Binary Damage S
18. Program A ustralian T est Interpretation Report for theU SN Wing Fuselage Landing Gear Test Articles DSTO TR 1929 M elbourne Defence Science and Technology Organisation Naval Air Warfare Center 1995 Fatigue Analysis of M etallic Structures FAMS A Computer Program to Calculate Fatigue D amage by Local Stress Strain A pproach Ayling J and M olent L 1998 A n Investigation intotheP rogram FAM S FatigueA nalysis of M etallic Structures DSTO TR 0681 DSTO Newman J C Jr 1992 FASTRAN II A fatigue crack growth structural analysis program NASA TM 104159 NASA Matricciani E 2005 P 3C SLAP Loads DataBase and Spectra Sequencing Tool DBI SST DSTO Validation DSTO TR 1720 Melbourne DSTO Neuber H 1961 Theory of stress concentration for shear strained prismatical bodies with arbitrary nonlinear stress strain law Transactions of the A SM E Journal of A pplied M echanics 544 550 Masing G 1926 Proceedings ofthe2nd International Congress for A pplied M echanics Zurich Switzerland September Stephens R I et al 2001 M etal Fatiguein Engineering John Wiley amp Sons Inc Basquin O H 1910 The exponential law of endurance tests Proc ASTM Vol 10 Part 11 ASTM West Conshohocken PA Coffin L F 1954 A study of the effects of cyclic thermal stresses on a ductile metal Trans ASM E 76 931 950 Manson S S 1965 Fatigue A complex subject Some simple approximations Experimental M echanic
19. Review DSTO TR 0381 Fishermans Bend Defence Science and Technology Organisation Zhao T and Jiang Y 2008 Fatigue of 7075 T651 aluminum alloy International Journal of Fatigue 30 5 834 849 Ghidella J R and Molent L 2003 A comprehensive guide to the fatigue damage assessment computer program C189 DSTO TR 1506 Defence Science and Technology Organisation Miner M A 1945 Cumulative damagein fatigue Journal of applied mechanics 12 A 159 A164 Palmgren A 1923 The endurance of ball bearings Z V er D eut Ing 68 No 14 339 41 Fatemi A and Yang L 1998 Cumulative fatigue damage and life prediction theories A survey of the state of the art for homogeneous materials International Journal of Fatigue 20 1 9 34 Veul R P G 2003 P 3C Orion Serive Life Assessment Program Final USN FMS Spectra Severity Comparison using Phase IIB Loads NLR CR 2003 Draft National Aerospace Laboratory NLR Bauschinger J 1881 Ueber die Veranderung der Elasticitatsgrenze und el astcitatsmodul verschiedener M etal Civiling 27 289 348 Morrow J 1965 Cyclic Plastic Strain Energy and Fatigue of Metals In Internal Friction Damping and Cyclic Plasticity Vol ASTM STP 378 West Conshohocken PA ASTM 45 87 57 DSTO TR 2392 31 ASTM 2003 Standard Practice for Strain Controlled Fatigue Testing ASTM E606 92 West Conshohocken ASTM 32 Shin C S Man K C and Wang C M 1994 A practical method to estimate the str
20. a complex spectrum It is based on the observed memory effect in many metals where a smaller load cycle appears to contribute an interruption to a larger load cycle In Figure 8 the load cycle B C serves as a small interruption to the larger 12 DSTO TR 2392 load cycle A D At point C as the load increases back to B the stress strain curve resumes its original path from A to D in effect remembering the previous load application from A to B TheRainflow counting method can be envisaged by rotating the strain timehistory so that the vertical axis is thetimeand the horizontal axis the strain amplitude Thestrain history can be viewed as a series of sloping roofs that overlap one another at different points The Rainflow cycles are then defined by the manner in which rain is allowed to drip down the roofs A number of rules are used to define the behaviour of the rain The process of Rainflow counting is as follows 14 e Allow therain to start from the largest peak or valley and run down the roof till it falls off completely e Start thenext droplet of water from the valley or peak at the point wherethedroplet fell from the roof and allow it to run in the opposite direction till it falls off the roof This constitutes one Rainflow counted cycle e Next begin again from the first load reversal that has not been fully touched by a droplet Allow the droplet to run till it falls off and terminate this path when either it falls off completel
21. describes these equations in greater detail 2 M odified M orrow equation The next options required extra information to be placed in the input file The extra information is entered on the next line following the equivalent strain equation option in the input file 3 Loopin The beta parameter in the Loopin equation is required on the next line underneath 3 4 M odified Loopin Two parameters used in the Modified Loopin equation beta and alpha are required on the next line underneath 4 5 Walker A parameter m used in the Walker equation is required on the next line underneath 5 6 Smith W atson T opper A parameter m used in the Smith Watson T opper equation is required on the next line underneath 6 24 DSTO TR 2392 7 M odified F 18 Two parameters used in the Modified F 18 equation beta and gamma are required on the next line underneath 7 Spectrum File Total Flight Hours This number specifies the number of flight hours the spectrum represents Thetext Flight Hours must be entered exactly as shown followed by the number of flight hours Maximum D elta Stress Thelargest changein stressin thespectrum is used in thecalculation of damagedensities The text Maximum Delta Stress must be entered exactly as shown followed by a number Thevaluehereis not critical to the final life estimates produced Thefunction of this variable influences the bin sizes used in determining the percentage of damage ca
22. is the sum of the elastic and plastic components Figure 5 shows the total strain behaviour in terms of the number of reversals to failurefor 4340 steel 13 The elastic and plastic components of equation 12 have been plotted in Figure 6 and the linear relationship with fatigue life is evident Similar relationships have been observed in many metals 4 e Experimental measurment FI 3 0 1 Ki A a Dls y Osp 3 2 e 0 01 8 A E 2 0 001 0 0001 1 10 100 1000 10000 100000 1000000 10000000 Reversals to Failure 2N P Figure5 Low cyclefatigue behaviour of annealed 4340 steel showing thetotal strain amplitude for constant strain amplitudeloading 13 10 DSTO TR 2392 Ga gia 2 0 1 0 1 2 E E E E lt lt 0 01 0 01 H H m N 3 2 E D 0 001 0 001 A 0 0001 0 0001 1 10 100 1000 10000 100000 1000000 1E 07 Reversals to Failure 2N p Figure6 Low cycle fatigue behaviour of annealed 4340 steel showing the elastic and plastic components of strain amplitude for constant strain amplitude loading 13 4 3 FAMSH Representation of the Strain Life Curve FAMS and FAMSH allow a more flexible representation of the strain life curve based on equation 12 Itis also worth noting that FA MSH also supports tabulated data to represent the strain life curve While the codes allow the user to implement equation 12 a more accurate representation of the stain life relationship over the whole strain range can be achieved by piece wise
23. life approach for crack initiation analysis 3 FAMSH Strain Life Algorithm FAMSH is a software tool based on FA MS 3 4 and developed at the DSTO It is used to perform fatigue life calculations based on a strain life methodology FAMSH was originally developed to support the technical interpretation and development of the structural management plan SM P for the P 3C It is currently being used in the certification process of the C 130 FAM SH is based on FAM S 3 4 with several added equivalent strain equations new outputs for damage analysis and support for the DBI SST spectrum format 6 The current release of FA MSH providesfully dynamic memory managenentto allow the program to be used on spectra of any size only limited by the size of the physical memory of the hardware without recompilation Furthermore thesame FA MSH source codeis now used to generate the executable file on both Windows and Linux platforms thus ensuring the consistency and currency of the results obtained from the code 4 The Strain Life Approach Traditionally most theories on fatigue relate the damage caused by cyclic loading to thelocal stress or strain ranges resulting in two similar but different approaches stress life approach and strain life approach The stress life approach is commonly used for low loads and high cycle fatigue when local plasticity is negligible while the strain life approach is commonly used in situations where local plasticity m
24. of Data Strain Life Data Cyclic Strain Cyclic Stress A Strain Amp 1 0 00587 60 5 A 1 92 0 0195759 SCH 0 00612 63 2 194 0 0144739 3 o 00636 655 3 296 0 0125531 4 0 0066 68 FC 398 0 0114951 E 0 00685 70 5 EN 500 0 0108079 lt 2 6 602 0 0103177 v Data Type Eq Stn Eq w Figure 16 Material tabular input in theFAMSH CGAP module 27 DSTO TR 2392 Data entered in thetable can be deleted by right clicking the specific data you wish to delete and selecting D elete see Figure 17 Insert Above Cyclic Stress Insert Below 20 Mi A Data Type Figure17 Deleting data in the material data tables Material data can also beinput in theform of equations Selecting Equation from the Stress Strain Data dropdown menu will allow the user to enter therelevant constants that specify the material equations used in FAMSH An example of the material property specification is shown in Figure 18 Material Properties Stress Strain Data Strain Life Data cl 0 01039 C2 0 17734 B1 0 07363 B2 0 1645 EI 0 02204 c3 5316 G1 0 57856 G2 2 1493 PTOI 1 16519e 0 1 ksi Q 3 5 ELET 0 005747 PLET 0 0103 Data Type Eq Stn Eq w Figure 18 M aterial equation input in the FAM SH CGAP module The stress strain equation used in FAMSH equation 8 is presented in section 4 1 2 In FA MSH theinput parameters used arePTQI and Q and relateto thecyclic strength coefficient K and cyclic strain hardeni
25. of fatigue life during the second pass One pass will ignore any residual stress calculated after the first pass 0 Two passes required 1 One pass required Notch Factor The notch factor is the N euber notch factor K and not the stress concentration factor K It should be noted that scaling factors are applied to this number using a different input parameter to be discussed later Material Identification This variable is a character string of length 25 In earlier versions the character string was limited to 9 characters This variable specifies the material to be used and reflects the exact name used in the material database If the name does not exist in the database a warning will be displayed by the CGAP GUI Material D ata Type MATF isan integer that indicates whether material data is specified in table format or as an equation If 1 is selected then the program expects the material data to bein a table look up format in a file with the name flmatl dat If 2 is selected then the program expects the material data to be in the form of an equation specified in the flmat2 dat file This will be automatically selected by CGA P using the information in the database 1 Tablelook up 2 Equation Unit Conversion Thisline contains areal number which is used to convert the units of the material data file To convert from KSI to MPa use the conversion factor of 6 895 It does not affect the loads in the spectrum file
26. relationship can befound in the Appendix A DSTO TR 2392 Hysteresis i stress versus 9 hysteresis i strain curve Strain Cydlic stress strain curve Figure3 Stress strain relationship after the application of a load o and then followed by fully reversed loading The hysteresis stress versus hysteresis strain curve is evaluated assuming Masing type behaviour4 8 and using the following relations gt a Ca 6 S 7 Ag Ne E Ac 1 Ao n O a gt O gt 0 2 2 2 2E e K 2 8 A A fe DO O SO 2 2E P where Ag and Az are the elastic and plastic components of the total strain range Ae respectively o isthe proportional limit o isthetotal applied stressand K and n arethe cyclic strength coefficient and cyclic strain hardening coefficients respectively Equation 8 describes the form of the stabilised hysteresis loop representing the cyclic stress strain behaviour 4 1 3 Fatigue N otch Factor The previous discussion of the strain life method suggests that the fatigue life of a smooth un notched specimen can predict the fatigue behaviour at the notch root of a notched 4 Masing type behaviour assumes that the hysteresis loop is twice the cyclic stress strain curve DSTO TR 2392 structure Elastictheory can be used to relatetheremote stress to the local stress at theroot of anotch using the stress concentration factor K In the case of fatigue it was originally hoped that t
27. specific cases 8 3 Definition and Validity of M aterial Data Another important consideration when using the FAMSH codeis the origin of the material data used in theanalysis This can be demonstrated by considering theinput used to describe thestrain life curve This curve can be generated in several ways and itisimportant to noteon what basis these curves were created A key consideration is the definition of failure Failure could be considered to have occurred when the test specimen reaches a predefined level of compliance or when the specimen has completely failed A changein the failure definition can completely change the interpretation of the solution Another consideration is the surface condition of the test specimens used to generate the strain life curve Typically these tests are performed with smooth un notched specimens with certain amount of surface polishing but thesurface condition of thestructuremay not always match those of thespecimens Again this will affect interpretation of the results produced by FAMSH and the engineer must consider this when interpreting the results Itis worth highlighting the strain life techniques sensitivity to small strains where fatiguelife estimates are high To demonstrate this issue consider two models both using the same 53 DSTO TR 2392 material one using the equation description and the other using thetabular input Themodels areidentical and only differ in theway thenotionally identica
28. 0 38007007 38007007 6401 00 38007007 13380 00 38007007 38007007 7222 00 38007007 13380 00 38007007 13408 1087 10422 7246 6401 7222 1222 38002023 38004004 38004004 38007008 38007007 38007007 38007007 SE RES 15104 13439 13512 14201 13380 14714 DSTO TR 2392 Table27 An example of the filename output file Example output FAMSH Version 1 42 Modified FAMS FAMS PROGRAM INITIATED 20090714 13 24 17 RR Information SS Passes vs Kn data are appended to file C Program files DSTO CGAP Analysis Files Example passes FOR KKK Information Default Morrow equation selected KKK k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k k KKK KKK KK KKK KKK KKK KKK KKK KKK KKK KK KK KKK KKK FATIGUE ANALYSIS OF METALLIC STRUCTURES FAMS Developed by N R Krishnan Modified by K N Bailey KKK KKK KKK KKK KKK KKK Fe ke Fe KKK KKK KKK KKK KKK KKK MATERIAL DATA Material D7075 T651 AL ALLOY DSTO Trendline Young s Modulus psi 1 0300E 07 Yield Stress psi 68000 Ultimate Stress psi 82500 Fracture Stress psi 110000 Title generic test seq LOADS APPLICATION NUMBER TWO KN 1 0000 SREF 2 500 KN SREF 2 500 name 2 TIME REPRESENTED BY LOADING FOR BLOCK NO 1 IS 1 000 NOTCH STRESSES AND STRAINS COMPUTED FOR 458 UNIQUE LEVELS BLOCK DAMAGE 6 78602E 05 name 2 TIME REPRESENTED BY LOADING FOR BLOCK NO 2 IS 1 000 NOTCH STRESSES AND S
29. 07 Required only for DSA formats e Ftn15 Optionally the single value contained in this file may be entered in Filename input on the last line with the keyword Maximum Delta Stress followed by a value such as 60000 The output files that are created by the CGAP FAMSH module are e Filename input e Filename passes e Filename truncspc e Filename dsout e Filename dmo e Flmatl dat e Flmat2 dat See section 5 8 for more detail There aretwo ways to enter input data e Open an existing input file e g from previous analysis e Enter data directly through CGAP GUI In either case the main input file has the extension input and it contains all the control information for the analysis Thespectrum file has the extension spectra and it contains the load spectrum defined by turning points Thedamage source codefileis entitled ftn07 and it contains the codes for different missions comprising the spectrum see reference 2 These codes are called Damage Source Assignments or DSA s Optionally a file ftn15 containing a single number representing the maximum stress range in the spectrum can be used The input file can be created with CGAP by entering data into CGAP directly and pressing the Build button Thespectrum file and the damage source code file are normally generated by upstream software but can also be created using a text editor The material specifications are contained in either the
30. 075T651 1 1 0 2 255 LE 07 2 15000 YES CYCLE Multiple 17 o w 1 0 o 1 DNSTGI 1 1 0 2 255 LE 07 2 15000 YES CYCLE Multiple 18 o u 1 0 o 1 DNSTGI 1 1 0 2 255 LEI 2 15000 YES CYCLE Multiple 19 o 2 1 0 o 1 D775T61 1 1 0 2 255 LE 07 2 15000 YES CYCLE Multiple 20 1 2 1 0 o 1 D7075T651 1 1 200 2 255 LEI 2 15000 YES CYCLE Multiple 21 1 2 1 0 1 1 D7075T651 1 i 0 2 255 LE 07 2 15000 YES CYCLE Multiple 22 1 12 1 0 o 15 D7075T651 1 1 0 2 255 1E 07 2 15000 YES CYCLE Multiple 23 1 2 1 0 o 1 D7075T651 1 12 0 2 255 1E O7 2 15000 YES CYCLE Multiple 24 1 2 1 0 o 1 D7075T651 1 1 0 2 255 1E 07 3 054 15000 YES CYCLE Multiple 25 1 2 1 0 o 1 D7075T651 1 1 0 2 255 1E 07 4 05 06 15000 YES CYCLE Multiple 26 1 12 1 0 o 1 D7075T651 1 1 0 2 255 1E 07 5 oa 15000 YES CYCLE Multiple 27 1 2 1 0 o 1 D7075T651 1 1 0 2 255 1E 07 6 os 15000 YES CYCLE Multiple 28 1 2 1 0 o 1 D7075T651 1 1 0 2 255 LE O 7 0 432 12900 15000 YES CYCLE Multiple 74 Table30 FAM SH Test Case Fatigue Life Estimates DSTO TR 2392 Case Kn CGAP1 8 FAM SH FAM SH 1 50 LINUX difference 1 2 5 98082 37 98082 37 0 00 5 0 6341 44 6341 44 0 00 2 2 5 122287885 33 122287885 33 0 00 5 0 4696715 79 4696715 79 0 00 3 2 5 28781 00 28781 00 0 00 5 0 1860 58 1860 58 0 00 1 2145999 00 2145999 00 0 00 10 Notch Strain Exceeds Fracture Strain For Material 4 2 5 233007031 05 233007031 05 0 00 5 0 7181150 94 7181150 94 0 00 5 2 5 115019049 25 115019049 25 0 00 5 0 3946915 39 3946915 39 0 00 6 2
31. 34 0 00 26 2 5 32232623256 35 32232623256 35 0 00 5 0 968792708 66 968792708 66 0 00 27 2 5 32672040 32 32672040 32 0 00 5 0 3145626 73 3145626 73 0 00 28 2 5 16197755 53 16197755 53 0 00 5 0 2563178 05 2563178 05 0 00 Page classification UNCLASSIFIED DEFENCE SCIENCE AND TECHNOLOGY ORGANISATION DOCUMENT CONTROL DATA 1 PRIVACY MARKING CAVEAT OF DOCUMENT 2 TITLE 3 SECURITY CLASSIFICATION FOR UNCLASSIFIED REPORTS THAT ARE LIMITED RELEASE USE L NEXT TO DOCUMENT A Strain Life Module for CGAP CLASSIFICATION Theory User Guide and Examples Document Title Abstract 4 AUTHOR S 5 CORPORATE AUTHOR Chris Wallbrink and Weiping Hu DSTO Defence Science and Technology Organisation 506 Lorimer St Fishermans Bend Victoria 3207 A ustralia 6a DSTO NUMBER 6b ARNUMBER 6c TYPE OF REPORT 7 DOCUMENT DATE DSTO TR 2392 AR 014 729 Technical Report March 2010 8 FILENUMBER 9 TASK NUMBER 10 TASK SPONSOR 11 NO OF PAGES 12 NO OF REFERENCES 2009 1131037 AIR 07 283 DGTA OIC 76 42 13 URL on the World Wide Web 14 RELEASE AUTHORITY http www dsto defence gov au corporate reports DSTO TR 2392 pdf Chief Air Vehicles Division 15 SECONDARY RELEASE STATEMENT OF THISDOCUMENT Approved for public release OVERSEAS ENQUIRIES OUTSIDE STATED LIMITATIONS SHOULD BE REFERRED THROUGH DOCUMENT EXCHANGE PO BOX 1500 EDINBURGH SA 5111 16 DELIBERATE ANNOUNCEMENT No Limitations 17 CITATION IN OTHER DOCUMENTS 18 DSTO RESEARCH L
32. AP module 5 5 1 Problem Title Under this tab the user can add a title to the output using the Problem Title input box 5 5 2 Nosref vs passes sref vs passes The user can also choose to create a passes file by toggling between No sref vs passes and sref vs passes See section 5 8 for more information 5 5 3 Analysis Output Options Therearefour options under the Analysis O utput O ptions dropdown menu Theuser has the option to choosefrom Basic Detailed More D etailed and MostD etailed These options control the amount of information written to the output file during analysis Basic level information reports the total damage calculated and the total life as well as other low level model information Detailed information additionally reports on the damage each block in the spectrum produces More Detailed output also includes the Rainflow cycles in the 32 DSTO TR 2392 spectrum and the notch stress and strain calculated for each of thesecycles The most detailed option also gives information about the notch stress and strain the alternating strain mean stress equivalent stress failure cycles and the damage for each cycle 5 5 4 Material Data Output Under the Material D ata Output dropdown menu the user can choose what material data to output to the output file The user has the choice of No material output Cyclic Stress Strain and Hysteresis Stress St
33. D MAXIMUM NUMBER OF BLOCKS LOADS APPLICATIO NOTCH BLOCK NOTCH BLOCK TOTAL DAMAGE LIFE UNDER LOAD TRUNCATION CUTOFF 00E 08 00E 07 00E 06 00E 05 00E 04 00E 03 00E 02 00E 01 00E 07 TOTAL CYCLES TOTAL REAL C TOTAL VIRTUA TOTAL CYCLES RESIDUAL STRESS LAST LOAD CAUSIN Analysis complet 70 AT END OF LOADING G YIELD test seq N NUMBER TWO SREF 5 000 KN name 2 EPRESENTED BY LOADING STRESSES AND STRAINS DAMAGE 1 87001E 03 name 2 EPRESENTED BY LOADING STRESSES AND STRAINS DAMAGE 1 51884E 03 3 38886E 03 LOAD 16952 00 lt FLIGHTS gt ANALYZED NUMBER TWO name 2 STRESSES AND STRAINS DAMAGE 1 68650E 03 name 2 STRESSES AND STRAINS DAMAGE 1 50659E 03 3 19310E 03 BLOCKS REPEATED The option was chosen to use LOADS APPLICAT OF DAMAGE DAMA 0 00319 100 0 00319 100 0 00318 99 0 00169 52 0 00039 AR 0 00039 12 0 00039 12 0 00039 12 0 00319 100 YCLES L CYCLES RAINFLOW AT END OF LOADING G YIELD ed successfully SREF D0 FOR BLOCK NO COMPUTED FOR FOR BLOCK NO COMPUTED FOR LIFE MINIMU 2 LOAD COMPUTED FOR COMPUTED FOR 626 ORIG GE 00 00 68 91 11 TA 11 11 00 3561 16952 LIFE 229 NAL 455 00 1 00 456 00 05 00 00 1 ES 458 2 LS 456 590 169 13408 458 456 626 351 L UNIQUE 1 UNIQUE 00 UNIQUE
34. DSTO developed software tool that contains a set of algorithms commonly used to assess and analyse fatigue damage It is envisaged that CGAP will gradually become a one stop tool set with which the engineers can perform most of their required fatigue analyses CGAP isa Microsoft Windows based application with a GUI and an integrated database for the management of material properties geometry sets and load cases The GUI assists the user in entering input data inspecting spectra and checking for data consistency It also providesa simple plotting capability for crack growth outputs CGAP contains a native crack growth analysis module based on the concept of plasticity induced crack closure In addition to all the functionalities of FA STRAN 3 8 5 the native module also has the capability for crack growth analysis involving notch plasticity and probabilistic crack growth analysis based on the 3 The DSN comprises ASI DGTA DSTO SPOs and numerous Australian industry partners DSTO TR 2392 Monte Carlo method For probabilistic analysis the initial crack size crack growth rate parameters and the peak spectrum stress can be randomised using three distribution functions Importantly CGA P interfaces seamlessly with FASTRAN3 8 5 and the DKEFF 5 code A recent update to the program includes the incorporation of the FAMSH code described in this report The inclusion of FAMSH to the CGAP environment expands the functionality of CGAP to include a strain
35. GHT DURATION q Number Flight General description not used in of linesin Duration the analysis the flight 38001022 1 13408 4 DSA Maximum Minimum Number TP stress TP stress of TPPs 38003019 6891 1087 2 Eahnumbershauidiba 12 10 contained within the spaces spaces lOspaces 10spaces allocated spaces Integer Real Real Integer DSTO TR 2392 5 6 2 Block format without Damage Source A ssignment DSA Thelayout and format of the spectrumfilethat is required when using Block type inputs and no DSA is presented in Table 5 N ote that the arbitrary number seen in the DSA block format in the first line should be removed Also note that the maximum and minimum stresses should be expressed using integer numbers and not real numbers Table5 Theformat of a FAM SH input file using the Block format with no D SA Example no DSA Block input 1 BLOCK The Spectrum number format of flights Example Flight Spectra Title of the spectrum 2 1 000 TURNING POINT PAIRS FLIGHT DURATION banan ian manni Number Flight General description not used in of TPPs Duration the analysis in the flight 1 13408 4 mme anm emm Maximum Minimum Number TP stress TP stress of TPPs 6891 1087 2 Each number should be lOspaces lOspaces 10spaces contained within the 10 Integer Integer Integer spaces 35 DSTO TR 2392 5 6 3 Cycle format with Damage Source A ssignment DSA The layout and format of the spectrum file that is required when
36. IBRARY THESAURUS http web vic dsto defence gov au workareas library resources dsto_thesaurus shtml Strain Life FAM SH CGAP Fatigue 19 ABSTRACT Existing airworthiness standards require that all aspects of fatigue be considered in the design development and maintenance of aircraft structures To minimise costs and improve combat readiness requires accurate and efficient fatigue life evaluation This report details the introduction of a strain life algorithm known as FA MSH nto the software tool suite known as CGAP developed and maintained at the DSTO Along with improvements to the original FAM SH codethis most recent development introduces a materials database and dynamic memory allocation With other additional improvements the latest release of the CGAP environment seeks to improvethe ease with which the engineer is able to accurately and reliably conduct fatigue life analysis Contained within this report is an explanation of the strain life theory in conjunction with a detailed description of its implementation in the CGA P GUI environment Examples are provided to guidethe user through its operation Page classification UNCLASSIFIED
37. L2 1 SNL1 2 SNL2 2 SNLI NPTS SNL2 NPTS Table12 File parameter descriptions flmat1 dat File Parameter Description of the file parameter P asingle character used by FA M SH to distinguish between a comment represented by a or a material header line represented by a MTITLE material title E Young s Modulus PLMT proportional limit FULT ultimate strength SF fracture strength NPTH the number of cyclic hysteresis data pairs FCTR the factor required to convert data from ksi into psi IM a flag that specifies the type of cyclic hysteresis data IM 1 cyclic strain versus stress data IM 2 hysteresis strain versus stress data IM 3 hysteresis stress strain versus stress data HYSI NPTH array with either cyclic strain hysteresis strain or hysteresis stress strain data Forms data pairs with HYS2 NPTH HYS2 NPTH array with stress data Forms data pairs with HYS1 N PTH NPTS the number of material strain life data pairs SNL1 NPTS array with material strain amplitude data Forms data pairs with SN L2 N PTS SNL2 NPTS array with material life data in number of cycles Forms data pairs with SNL1 N PTS 43 DSTO TR 2392 Table13 Format of the flmat2 dat file Line File Parameters 1 P MTITLE 2 E PLMT FULT SF PTQI Q 3 C1 C2 C3 C4 4 B1 B2 G1 G2 PLET ELET Table14 File parameter descriptions flmat2 dat File Parameter Description of the file parameter P asingl
38. S 1 vVa r where a is the characteristic length dependant on the material and r isthenotch radius Some values for the characteristic length a can befound in 9 The N euber notch factor Ky isonly appropriate for unloaded holes or notches in individual specimens It does not take into account any effects the material itself might introduce In FAM SH the fatigue notch factor is used in place of the theoretical stress concentration factor to calculate the local stress and strain Ku s 10 4 2 Stress and Strain Related to Fatigue Life Unlikethestress life approach thestrain life approach does not require test data for a variety of notch types Rather the strain life methodology reduces the required input data by providing a more general means of estimating fatigue damage and by considering prior knowledge about stress strain relationships i e Neuber K Thisis achieved by assuming that the fatigue behaviour at the notch root is equivalent to that of uni axial test specimens see Figure 4 In other words the fatigue life of material at the root of a notch is equivalent to the fatigue life of an un notched specimen of the same material subjected to the same cyclic straining DSTO TR 2392 AS Ae N otched Aer As Specimen Notch A Un notched smooth specimen Ao NE AS Ae Figure4 Diagram demonstrating the rationale for using strain life data obtained on smooth specimens to predict the fatigue life of notched specimens The basic
39. SION MISSION MISSION TOTAL AVG DMG MISSION NUMBER UTILIZ DAMAGE PER FLT DMG OF TOTAL 1 121 0 1345E 02 0 1111E 04 0 88 2 65 0 1545E 02 0 2377E 04 E 3 63 0 2486E 02 0 3946E 04 1 63 4 132 0 1991E 02 0 1508E 04 1 30 5 94 0 2596E 02 0 2762E 04 1 70 6 139 0 6220E 02 0 4475E 04 4 07 7 75 0 1867E 02 0 2489E 04 1 22 8 193 0 5910E 02 0 3062E 04 3 87 9 96 0 3295E 02 0 3433E 04 216 LO 95 0 2777E 02 0 2923E 04 1 82 11 114 0 4070E 02 0 3570E 04 2 66 12 33 0 6235E 03 0 1889E 04 0 41 13 91 0 3849E 02 0 4230E 04 dwsIZ L4 126 0 4775E 02 0 3789E 04 3 12 15 111 0 2781E 02 0 2505E 04 1 82 16 175 0 6870E 02 0 3926E 04 4 49 17 216 0 1363E 01 0 6309E 04 8 91 18 206 0 1511E 01 0 7336E 04 9 88 19 77 0 9120E 02 0 1184E 03 5 96 20 102 0 4437E 02 0 4350E 04 2 90 21 66 0 4884E 02 0 7399E 04 3419 22 82 0 5841E 02 0 7124E 04 3 82 23 153 0 1823E 02 0 1192E 04 1 19 24 107 0 2016E 02 0 1884E 04 1 32 25 87 0 1791E 02 0 2059E 04 LE 26 90 0 2324E 02 0 2582E 04 1552 27 120 0 3334E 02 0 2778E 04 2 1 8 28 109 0 4402E 02 0 4038E 04 2 88 29 102 0 4158E 02 0 4077E 04 Serie 20 191 0 3698E 02 0 1936E 04 2 42 Sch 143 0 1063E 02 0 7436E 05 0 70 32 94 0 1027E 02 0 1093E 04 0 67 33 290 0 2573E 02 0 8874E 05 1 68 34 150 0 5210E 02 0 3473E 04 3 41 35 127 0 1492E 02 0 1175E 04 0 98 36 100 0 1847E 02 0 1847E 04 Lt Sd 59 0 9397E 02 0 1709E 03 PJ 38 11 0 7360E 03 0 6691E 04 0 48 TOTAL 4401 0 1529 0 3475E 04 100 0 67 DSTO TR 2392 Table24
40. TRAINS COMPUTED FOR 456 UNIQUE LEVELS BLOCK DAMAGE 5 48013E 05 TOTAL DAMAGE 22661E 04 LIFE 16305 0 NUMBER PASSES 8152 5215 APPLIED MAXIMUM LOAD 16952 00 MINIMUM LOAD 13408 00 NUMBER OF BLOCKS lt FLIGHTS gt ANALYZED 2 LOADS APPLICATION NUMBER TWO name 2 NOTCH STRESSES AND STRAINS COMPUTED FOR 458 UNIQUE LEVELS BLOCK DAMAGE 6 78602E 05 name 2 NOTCH STRESSES AND STRAINS COMPUTED FOR 456 UNIQUE LEVELS BLOCK DAMAGE 5 48013E 05 TOTAL DAMAGE 1 22661E 04 LIFE 16305 0 NUMBER PASSES 8152 5215 LIFE UNDER LOAD BLOCKS REPEATED 16305 0 NUMBER PASSES 8152 5215 The option was chosen to use LOADS APPLICATION NUMBER TWO for damage and truncation studies TRUNCATION OF ORIGINAL CYCLES CYCLES CUTOFF DAMAGE DAMAGE REMOVED REMAINING REMOVED 00E 08 0 00012 100 00 30 00 426 00 6 579 00E 07 0 00011 89 38 92 00 264 00 42 105 00E 06 0 00005 39 07 442 00 14 00 96 930 00E 05 0 00004 28 54 451 00 5 00 98 904 00E 04 0 00004 28 54 451 00 5 00 98 904 00E 03 0 00004 28 54 451 00 5 00 98 904 00E 02 0 00004 28 54 451 00 5 00 98 904 00E 01 0 00004 28 54 451 00 5 00 98 904 00E 07 0 00011 89 38 92 00 264 00 42 105 TOTAL CYCLES 456 00 TOTAL REAL CYCLES 455 00 TOTAL VIRTUAL CYCLES RAINFLOW 1 00 TOTAL CYCLES 456 00 69 DSTO TR 2392 RESIDUAL STRESS LAST LOAD CAUSIN Title generic LOADS APPLICATIO KN 1 0000 TIME R NOTCH BLOC TIME R NOTCH BLOC TOTAL DAMAGE APPLIE
41. ain rangelevel Ag read off the strain life curve for the given stress range level If the spectrum consists of repeated cycles at one constant strain amplitude then failure is assumed to occur when n N and thus the damage D 1 For a variable amplitudeload history containing morethan onestrain range fatigue damage can be calculated as the sum of the damage attributable to each individual strain range level within the spectrum That is n YD E 23 1 Failure is assumed to occur when gt Diels 24 This linear summation of damage caused by different strain ranges is Known asthe Palmgren Miner rule 25 To apply this summation rule thespectrum must first be analysed to identify the dosed hysteresis loops using a scheme such as Rainflow counting Each hysteresis loop is converted to an equivalent strain cyclewith zero mean stress using an appropriate equivalent strain equation Following this step the Palmgren M ner rule equation 23 can be applied to obtain the total damage resulting from the spectrum Once the damage calculated using equation 23 exceeds 1 failure of the structure is assumed to occur In the FAMSH implementation theinverse of thetotal damagein equation 23is used to calculatethe number of times passes the spectrum is applied such that the total damage is 1 0 This number is then used to determine the total fatigue life in flight hours using the flight hour s one spectrum represents Presently FA MSH only imp
42. al D atabase 43 Te EXAMPLE iy A AR 44 TAL aA 44 RT ER 45 Tid E EE 49 L I MT 50 8 DISCUSSION 51 81 Cyclic Stress Strain Behaviour U U U uu uuu rr 51 82 Notch Root CC seers 53 83 Definition and Validity of Material Data 53 8 4 Damage SuMMAation cin o 54 9 FUTURE IMPROV EM ENTS iria 55 10 CONCLUSION iii 55 11 ACKNOWLEDGEMEN T 5 enki ie I nanana aani u a asus sata diia sa n an daian 56 12 REFERENCES U U U U U U uu u uu Qu uu Q 56 APPENDIX A APPENDIX B APPENDIX C APPENDIXD APPENDIX E THEBASISOFTHEFAMSH STRESS STRAIN RELATIONSHIP I U I rr 59 DERIVING MATERIAL PARAMETERS FOR THEFAMSH STRAIN LIFE EQUATION usecase 61 FAM SH OUTPUT FILES U U rr 67 SAMPLE FAMSH MATERIALS FILES sssi 71 FAM SH TEST CASES snm 73 AAP ADF ASI CGAP CI DBI SST DGTA DSA DSN DSTO FAMS FAMSH GUI NDT OEM SLAP SMP SPO TP TPP Abbreviations Australian Air Publication Australian Defence Force Aircraft Structural Integrity Crack Growth Analysis Program Crack Initiation DataBase Interface Spectra Sequencing Director General of the Technical Airworthiness Damage Source A ssignment Defence Support N etwork Defence Science and Technology Organisatio
43. atiguecrack growth in aircraft structures constitutive models and plasticity and numerical methods in engineering Contents ABBREVIATIONS E INTRODUCTION EE 1 2 CGAP CRACK GROWTH ANALYSIS PROGRAM u uu 2 3 FAMSH STRAIN LIFE ALGORITHM I I U U U U U Uu Q uu u 3 4 THE STRAIN LIFE APPROACH iii 3 41 Notch Stresses and Strains U u U u U rr 4 4 1 1 Neuber s Method I dd 5 4 1 2 FAMSH Representation of Greg Gran 6 4 1 3 Fatigue Notch Factor cti nn sns 7 42 Stress and Strain Related to Fatigue Life 8 4 3 FAMSH Representation of the Strain Life C urve 11 AA Raintlow C Outta WEE 12 4 5 Mean Stress Effects and Equivalent Strain Equations 14 4 6 Fatigue Damage Accumulation U U U u u u uuu u Q uu 16 5 USING FAMSH IN CGAP vaina 18 5 1 e estate HE WE 18 52 Datalnputsand OU PU inci 20 5 2 1 KEE a a ak n a tree 21 5 2 2 Importing an Analysis File 21 53 Material tab 25 5 3 1 Loading an Existing Material a 26 5 3 2 Creating a New Material Database Entry 26 5 3 3 Eq Stn Eq Equivalent Strain Equation 29 5 3 4 Strain Life EQUAI N vasito 29 5 3 5 UnitCv Un
44. ay occur For the stress life approach the main material data required is the stress life curve for the stress concentrator concerned Thus different stress life data need to be generated for different notch features The local stresses and strains are calculated using the stress concentration factor of the notch feature For the strain life approach it is assumed that as far as fatigue damage is concerned the material at the notch root behaves like as a smooth un notched specimen Thus only one set of strain life data obtained from a smooth specimen is required for the basic material data However validation against other notched datais usually required for ASI purposes When local plastic deformation occurs the local stresses and strains are calculated using a method such as N euber s equation When the applied stress is low thelocal stress will simply bea product of the applied stress and the elastic stress concentration factor Therefore theoretically for low stress and high cycle fatigue the strain life approach is identical to the stress life approach Hence the strain life approach may be considered a more general approach for modelling fatigue damage The block diagram in Figure 1 illustrates the process of the strain life approach The basic input data are the stress concentration representing the geometry of the specimen the cyclic DSTO TR 2392 stress strain relation the strain life relation of the material and the applied load
45. cond number is a dummy variable which is not used in the analysis Enter 1 as the second number Finally the key word BLOCK should beentered to indicatethat thespectrum fileis defined in block format Line2 containstwo numbers theflight number and the mission type Theflight number is not used in the analysis but is used to indicate which flight the analysis refers to Enter 101 for theflight number Themission type identifies the mission using a number In thisexamplewe are identifying the mission as mission 27 so enter 27 as the second number on line 2 Any text that is added following these numbers will not be used by FAMSH in the analysis Line 3 also contains two numbers the first number is the total number of lines below Line 3 that definethe spectrum and the second number defines the flight duration Sincefour lines are needed in this example to describe the flight we enter 4 for the first number For this problem let us further assume that the spectrum represents 100 flight hours so enter 100 for the second number Although 100 is entered here it is not used in the analysis The actual flight hours represented by the spectrum is input separately in the input file Thenext four lines contain the information shown in Table 15 For this problem namethefile Gen_seq spectra and save it 45 DSTO TR 2392 Table16 The spectrum file Gen_spec spectra 1 Bl BLOCK DOU 27 FLIGHT NUMBER 6 MISSION TYPE
46. contains either cyclic strain hysteresis strain or hysteresis stress cydlic hysteresis strain data stress strain curves StressData HYS2 N PTH contains stress data represented by discrete data points from flrrat1 dat file Links with Material_Cl and Material StrainLife Cl tables 41 DSTO TR 2392 Table Name Table Description alas Seas FAMSH Variable Description Material_StrainLife CI Contains crack id MTITLE 1 9 Material specification initiation data for StrainData SNL1 NPTS contains material strain amplitude data n LifeDat SNL2 NPTS tains material life data in number of cydl curves represented by ifeData contains material life data in number of cycles discrete data points from flmet1 dat file Links with the Material_CycHys Cl table and Material CI tables Material_Equation_Cl Contains crack id MTITLE 1 9 Material specification initiation datafor the PTQI Q PTQI Q Parameters used to describe the cyclic stress strain curve see n aik section 4 1 2 for more information relationship and the Ss strain life de C1 C2 C3 C4 Cl C2 C3 C4 Parameters related to the fatigue strength and ductility represented by coefficients and exponents See equation 12 and equation 13 Also see Appendix B for a suggested method to evaluate these equations from efficient flmet2 dat file Links panais with the Material CI B1 B2 G1 G2 B1 B2 G1 G2 Exponents related to the fat
47. der strain controlled cyclic loading wherethe mean stress of the spectrum relaxes towards zero as observed in Figure 35 a When a non symmetric constant amplitudeload is applied under stress control another effect known as ratchetting may occur as in Figure 35 b At structural details such as notches the local load is neither under strain control nor under stress control irrespective of the mode of control of the remote loading Hence we may expect to see a combination of mean stress relaxation and ratchetting as in Figure 35 c Norminal stress s G G NW NW G G N lo ye f Figure35 Elastic plastic deformation behaviour subjected to a constant cyclic strains b constant cyclic stresses and c remote constant nominal cyclic stresses 52 DSTO TR 2392 In FAMSH none of the above complex stress strain behaviours are considered Instead the materials are assumed to be of M asing type i e the hysteresis loops can be reconstructed by magnifying the stead y state cyclic stress strain curve by afactor of two The steady state cyclic stress strain curve is obtained by conducting tests on smooth un notched specimens subjected to fully reversed constant amplitude loading 311 Other effects such as cyclic softening and hardening ratchetting etc which effect the shape of the hysteresis loops have not as yet been considered in the code These effects are hard too quantify in situations of variable amplitud
48. e character used by FA M SH to distinguish between a comment represented by a or a material header line represented by a MTITLE material title E Young s Modulus PLMT proportional limit FULT ultimate strength SF fracture strength PTQI Q Parameters used to describe the cyclic stress strain curve see section 4 1 2 for more information C1 C2 C3 C4 Parameters related to the fatigue strength and ductility coefficients and exponents See equation 12 and equation 13 B1 B2 G1 G2 Exponents related to the fatigue strength and ductility exponents via equation 13 PLET ELET are strains chosen to define the transitions between region 1 2 and 3 N otethat FA MSH haslimits on thesizeof thearrays associated with theflmat1 dat file Arrays HYS1 and HYS2 can contain no more than 75 elements while arrays SNL1 and SN L2 can contain no more than 65 elements CGAP has a limit of 50 elements 7 Example This section uses an example to illustrate the use of the FAMSH code It is helpful to remember that the FAMSH module operates in a very similar manner to the FAMS code Thus we can expect similar outputs to be produced For a detailed description with regard to the strain lifetheory on which FA MSH is based refer to section 4 The example presented in this section will becovered in four sections which will explain the Problem Inputs Execution and Results 7 1 Problem Consider a problem in which the same spectrum is ap
49. e damage cut off threshold DCUT This specifies the truncation level That is it truncates loads in the spectrum with damage less than DCUT 5 4 7 Hours per sequence This value is used to convert the life in cycles to flight hours This number appears in the input file after the key phrase Flight H ours 8 The DBI SSI format was developed to supporttheP 3C servicelife assessment program and was used to examinethe damage resulting from different flight conditions 31 DSTO TR 2392 5 5 Case Control tab The Case Control tab see Figure 22 allows the user to select appropriate output options CGAP1 7f Sample fel File Edit View Commands Tools Help a m B CS New Open Save Save Out Build Configure cases I Material Load Case Contol Output Case Database lt D Deseription Problem Tile wed A Crack Growth Analysis Add your own notes here Input Unit Output Unit Imperial Uni gt EI Output Options e Nosref vs passes Material Data Output C sref vs passes No Material Output gt Running FAMSH Databases Case Database Analysis Output Options Basic y Geometry Database Number of Pass to Analyse Material Database Load Database Crack Length ae 0 0 Number of Cycles Messages Input File Output Fie Plot File Spectrum Fie 0 0 Ready Figure 22 The Case Control tab used in the FAM SH CG
50. e loading and require a more sophisticated approach when modelling The assumption used when applying the FAMSH code is that these second order effects have minimal impact on thefinal solution However itis up to the user to ensure the validity of these assumptions 8 2 Notch Root Stress FAMSH uses the N euber method to estimate the stress at the root of a notch The N euber method while assumed to beapplicableto a widerange of geometries does havelimitations The method does not hold truein situations where the notch root stress is not in phase with the remote load It does not account for time dependant phenomenon such as creep and stress relaxation It also does not account for cyclic stress relaxation at the notch root While the Neuber method has limitations it is still widely used In most cases the Neuber method slightly overestimates the notch tip stresses and strains 32 This may result in a conservative estimate of life Other methods of estimating the notch root stress are availablesuch as that by Glinka 33 which has been shown to provide better results in conditions of plane strain and torsional loading 34 35 The inclusion of other methods into CGAP to estimate the notch root stress could improve the accuracy of predictions for certain cases where the N euber method has been shown to be deficient At present the N euber method is the only method available and it must be remembered that the method may lead to erroneous solutions in
51. ed The document also provides a thorough explanation of the operation of the software tool Finally a step by step problem is worked through for the users benefit Accurate fatigue life assessment and life extension of metallic aircraft structures directly benefits defence These benefits through improvements in prediction reliability can be realised in terms of reduced operational and maintenance costs improved performance and combat readiness Authors Chris Wallbrink Air Vehicles Division Chris Wallbrink graduated in 2005 with a PhD from Monash University in the department of Mechanical Engineering H e then held the position of research fellow at M onash U niversity till theend of 2006 where his research interests included modelling of fatigue crack growth infrared NDT technologies and fibre optic corrosion detection devices Chris joined DSTO in 2007 in the Air V ehicles Division and is currently conducting research into methods for fatigue crack growth modelling Weiping Hu Air Vehicles Division Weiping Hu joined DSTO in 1998 as a research scientist He is currently a senior research scientist leading the development of modelling capabilities for the analysis of structural integrity of aircraft structures A fter obtaining his P hD degreein 1993 at D ublin City University Ireland he has held various academic positions at Dublin City U niversity M onash U niversity and D eakin U niversity His current research interests includef
52. educed after the application of a load of opposite sign The application of one single load is enough to change the stress strain behaviour of a material The way in which the behaviour changes after one single load is also affected by the temper of the material A dramatic example of how the temper of a material can alter the stress strain relationship with repeated loads has been presented by Morrow 30 He observed the continuous stress strain behaviour of copper for conditions where the copper was a fully annealed b partially annealed and c cold worked The results of this investigation are presented in Figure 34 For this particular material the curves eventually stabilise after enough repeated loads 51 DSTO TR 2392 h Ae 0 0084 Ae 0 0078 2N 8060 reversals 2Nr 4400 reversals S Ni A fh ons AEE reversal ig bb fe de 0 0099 2N 2000 reversals Figure 34 Stress strain behaviour of copper subjected to cyclic strain controlled axial loads a Fully annealed showing cyclic hardening b Partially annealed showing small cyclic hardening and softening c Cold worked showing cyclic softening 30 Other effects such as mean stress relaxation and ratchetting have also been observed with varying degrees in different materials Some examples of the how these effects change the stress strain response with repeated load applications are provided in Figure 35 Mean stress relaxation can be observed un
53. ence stress analysed A spectrum file containing 2 flights is defined in block format using DSA A spectrum file containing 1 flight is defined in block format using DSA A spectrum file containing 1 flight is defined in cycle format using DSA A spectrum file containing 1 flight is defined in cyde format with no DSA A spectrum file containing 2 flights is defined in cycle format with no DSA A spectrum file containing 1 flight is defined in block format with no DSA A spectrum file containing 2 flights is defined in block format with no DSA A large spectrum file is used An equation material description is used 7075 T651 with the case 1 spectrum Removed the Flight hours and maximum delta stress option from the input file No stress strain information is output and Basic analysis output is requested Cyclic stress strain information is output and Basic analysis output is requested hysteresis stress versus hysteresis stress strain information is output and Basic analysis output is requested No stress strain information is output and detailed analysis output is requested No stress strain information is output and more detailed analysis output is requested No stress strain information is output and most detailed analysis output is requested Residual stress is added to the analysis One pass is used in the analysis A Kn of 1 5 is used in conjunction with the reference stresses A conversion factor of 1 2 is used in the analysis The loopin equivale
54. ess concentration of notches International Journal of Fatigue 16 4 242 256 33 Glinka G 1985 Energy density approach to calculation of inelastic strain stress near notches and cracks Engineering Fracture M echanics 22 3 485 508 34 Knop M etal 2000 On theGlinka and N euber methods for calculating notch tip strains under cyclic load spectra International Journal of Fatigue 22 9 2000 10 743 755 35 Jones R et al 1998 Stress and Strain Estimation at N otches in Aircraft Structures DSTO GD 0196 Fishermans Bend Defence Science and Technology Organisation 36 Cheng G and Plumtree A 1998 A fatigue damage accumulation model based on continuum damage mechanics and ductility exhaustion International Journal of Fatigue 20 7 495 501 37 Bilir G 1991 Experimental investigation of fatigue damage accumulation in 1100 A alloy International Journal of Fatigue 13 1 3 6 38 deCastro J T P Meggiolaro M A and Miranda A C d O 2009 Fatigue crack growth predictions based on damage accumulation calculations ahead of the crack tip Computational M aterials Science 46 1 115 123 39 Dur n J A R Castro J T P and Filho J C P 2003 Fatigue crack propagation prediction by cyclic plasticity damage accumulation models Fatigue and Fracture of Engineering M aterials and Structures 26 2 SPEC 137 150 40 Qiu B X et al 2008 Advanced M aterials Research Vol 44 46 41 Ellyin F and Fa
55. fatigue life data needed for the strain life approach are thefatigue lives at various strain amplitudes These data are acquired by conducting tests on smooth cylindrical specimens subjected to fully reversed constant amplitude loading under strain control Similar to Basquin s observation on stress life data 10 Coffin 11 and M anson 12 found that the plastic strain life data could be linearised on a log log scale and be expressed as AE p E QN 11 whereAc isthe plastic strain range e isthefatigue ductility coefficient 2N isthenumber of reversals to failureand c istheductility exponent Combining the above with the Basquin equation the fatiguelife may berelated to the total strain range experienced by the material In FAMSH this relationship is described in the following form E J A Az E 2 g Ae SCH SEN A a 12 SS A N NE N NG where the parametersa 8 b andc are related to the fatigue strength and ductility coefficients and exponents via DSTO TR 2392 c c For a particular material the strain life N data must first be obtained through a series of tests Thesetestsinvolvetheapplication of fully reversed loads under strain control to smooth un notched specimens Oncethe N curveisknown itisthen possibleto estimatethenumber of cycles to failure for a given strain history It is useful to recall that the N diagram deals with strain amplitude and as such the total strain amplitude
56. fitting equation 14 to experimental data This process is depicted in Figure 7 The strain life data are divided into three segments dominated by plastic strain elastic strain or by a balanced combination of elastic and plastic strain Here the elastic component of strain can be represented with Ae a Cl C3 3 N NE of Nee 14 f f f and the plastic component can be represented with A f C C4 or 15 2 N N NP In Figure 7ELET and PLET arestrain ranges chosen to definethetransitions between regions 1 2and 3 Depending on the values of ELET and PLET different combinations of equations are used to represent the entire strain life curve as detailed in Table 1 The constants used in Table 1 and Figure 7 can be found by curve fitting equation 12 to experimental data See Appendix B for an example of determining the input parameters for the FA MSH strain life equation 11 DSTO TR 2392 Sha E ar Region 1 f A Ag 2 Zen ZO us lt a Region 2 2 N 2 N PLET ere Ae C3 AE p SE R ion 3 E ELET les NN l Ko 8 E Reversals to Failure 2N Figure7 Strain lifecurverepresentation in FAM S and FAM SH Tablel Parameters used in FAM SH to defined each region of the strain life curve in Figure 7 Total strain Ac Region 1 Region 2 Region 3 PLET ELET 0 f C1 B C2 G1 N A N A ELET gt PLET 0 f C1 B1 C2 G1 f C3 B2 C2 G1 N A PLET gt ELET 0 f C1 B1 C2 G f
57. fixed parametersto determinethefinal two unknown parameters C3 and B2 using data set 3 The curve fit is presented in Figure 40 Data set3_Strain Model user21 Chi 2 DoF 2 6671E 9 R 2 0 99969 0 04192 0 00232 1 37415 0 0 24271 0 00591 0 99716 0 Strain 10000 100000 1000000 Life Figure 40 Curvefit for data set 3 using Origin 6 0 After fitting thefinal set of data wenow havedetermined all the parameters needed to define the strain life equation These parameters are presented in Table 22 DSTO TR 2392 Table22 Fitting parameters for the FAM SH strain life equation for 7075 T7351 aluminium Region 1 C2 G1 C1 B1 0 81999 0 00907 C3 B2 0 24271 CA G2 1 37415 In this example when compared to the original 7075 T7351 aluminium data the FAMSH strain life equation produces excellent agreement see Figure 41 1 o 7075 T7351 Data FAMSH Strain Life Equation 0 1 0 01 o 0 001 0 0001 1 10 100 1000 10000 100000 1000000 1000000 1E 08 0 Life Figure41 TheFAM SH strain life equation compared to the original 7075 T 7351 aluminium data 65 DSTO TR 2392 Appendix C FAMSH Output Files Table23 An example of thefilename dmo file Example dmo DSTO TR 2392 FAMSH PROGRAM INITIATED 20090731 10 45 06 Title generic test seq LOADS APPLICATION NUMBER TWO KN 1 0000 SREF 2 500 KN SREF 2 500 DAMAGE DISTRIBUTION BY MIS
58. hat controls the type of material data written to the output data file The following options are available 0 No material output No stress strain data is output 1 Cyclic Strain Stress A strain vs stress table is printed to the output file 2 Hysteresis Stress Strain A stresststrain vs stress table is printed to the output file Analysis O utput Information Another integer number is used to control the amount of information that is output to the output data file The following options are available 11 or 1 Basic 10 or 0 Detailed 11 or 1 More Detailed 12 or 2 Most Detailed If the second number above is used then no passes file is produced For more information see section 5 5 3 22 DSTO TR 2392 Loads D ata Format ASCII or Binary This line contains an integer number that tells the program whether the spectrum file is in Binary or A SCII format 0 Binary format 1 ASCII format D amage Source Indicator This is an integer number that tells the program whether the Damage Source Assignment DSA code is used The DSA is a nine digit number that is used in the spectrum file and is used in the subsequent analysis and output 2 For further information and an exampleof the application of damage source assignments refer to 28 0 DSA is used 1 DSA isnot used Number of Passes An integer number is used to switch between one pass and two passes A residual stress is calculated after one pass and is used in the calculation
59. he most commonly used fatigue crack initiation models is the so called strain life method The theory that underpins the strain life method forms the basis of the fatigue life tool FAMS 3 4 and FAMSH FAMSH sa modified version of FAM S with additional expanded functionality with fewer limitations imposed However both FAM Sand FA MSH lack an intuitive user interface and a method by which to manage material properties Recently the FA M SH sourcecodehas been incorporated as a module into the Windows GUI known as CGAP CGAP provides a common user friendly utility that can be used to control the material property definitions 1 The particulars of the combination account for the limitations of each model assessed against part inspectability damage susceptibility NDI capability criticality of thestructure new aircraft materials and cost to repair replace 2Thecrack initiation and growth stages are delineated by a crack size relating to the capability of non destructive inspection technique and the limitations of the crack growth models DSTO TR 2392 which are important in achieving consistency in input data that underpins the advice provided to the A DF The introduction of error checking consistent material database and intuitive interface will ultimately minimise mistakes and improve the accuracy and assuredness of advice The aims of this document are to e summarise the salient points of the strain life theory e explain how the strai
60. he peak stress and m is a material parameter v Smith Watson Topper SWT equation 22 o m max gt og E J O max Es 0 20 0 Omax lt max The Smith Watson and Topper mean stress equation is a general formulation that has been shown to correlate well with the fatigue experiments of aluminium alloys 23 vi Modified F 18 equation 22 24 Thereport by Ghidella et al 24 provides a good background to the development of the F 18 equivalent strain equation An important point to note is that the equation was developed using data that did not include negative mean stresses yE B O mean 2 Eeq Ea T s O mean lt 0 yE 1 R where y sa material parameter B F O mean Eq 8 S O mean Z 4 6 Fatigue Damage A ccumulation The fatigue damage caused by one cycle is defined by the reciprocal of the number of repetitions of that cycle needed to cause failure The loss of energy per hysteresis loop is 16 DSTO TR 2392 additive which leads to the assumption that fatigue damageisalso additive i e the Palmgren Miner rule 25 This linear damage accumulation concept was first proposed by Palmgren in 1923 26 and Miner provided the mathematics to describe it in 1945 25 Consequently the fatigue damage produced by n cycles at one strain level is defined as n D ee 22 where n is the number of applied cycles at strain range level Ag N isthetotal number of cycles to causefailureatthestr
61. his would hold true for thefatiguestrength the amplitude of the applied fully reversed cyclic stress that will cause failure after a series of repeated cycles However it has been found that the ratio of the fatigue strengths of the un notched and notched specimens at the same number of cycles is not equal to K This is because in reality the notched fatigue strength is also affected by thenotch radius material strength material properties etc 9 To correct this behaviour a new parameter the fatigue notch factor K was introduced to replace K The fatigue notch factor is not a theoretical parameter like the stress concentration factor and isnot just a function of the geometry and remote loading The fatigue notch factor relates the fatigue strength of a smooth specimen to that of a notched specimen using the following equation Smooth fatigue strength 9 K f Notched fatigue strength at equal life If the notched fatigue strength was just a function of the geometry and the load then K would be equivalent to the elastic stress concentration factor Thus the fatigue notch factor is essentially an empirical parameter used to account for these other factors An empirical relation was developed by N euber to evaluate the fatigue notch factor N euber developed what is known as the N euber notch factor based on the hypothesis that stresses may be assumed constant over small distances 9 This factor is defined as Get e
62. igue strength and ductility exponents table via equation 13 Also see A ppendix B for a suggested method to evaluate these exponents PLET ELET PLET ELET are strains chosen to define the transitions between region 1 2 42 and 3 DSTO TR 2392 6 2 FAMSH interaction with the M aterial D atabase The material properties stored in the database are used by FA MSH indirectly through one of two material files flmat1 dat and flmat2 dat automatically generated by the CGAP GUI A sample of each of these files is provided in Appendix D Both files contain basic material property data such asthe Young s M odulus yield and ultimate strengths They also contain material cyclic hysteresis curves and strain life curves They differ in that flmatl dat represents the material curves via discrete data points while flmat2 dat uses up to three equations to represent thecyclic hysteresis and strain lifecurves A flag variable MA TF within the FAMSH input file tells the program which material data file to utilise The parameters and format of flmatl dat and flmat2 dat are presented in Table 11 and Table 13 respectively The definitions of the variables contained in flmatl dat and flmat2 dat are presented in Table 12 and Table 14 respectively Table 11 Format of the flmat1 dat file Line File Parameters 1 P MTITLE 2 E PLMT FULT SF 3 NPTH FCTR IM 4 HYS1 1 HYS2 1 HYS1 2 HYS2 2 HYS1 NPTH HYS2 NPTH 5 NPTS 6 SNL 1 SN
63. ion of the program It should be noted that they are not a validation of the capability of FAM SH in predicting fatigue initiation lives All test cases produced consistent results with those produced by the FAM SH executable It is important to consider the limitations with the current strain life methodology An understanding of the deficiencies and limitations will help the engineer to better interpret results in a more meaningful manner Following is a discussion of some of these deficiencies and limitations 8 1 Cyclic Stress Strain Behaviour Materials under cyclic loading can display various phenomenathat arenot as yet modelled in the current strain life algorithm included in CGAP FAMSH TheFA MSH module evaluates damage based on stabilised stress strain hysteresis loops But it is well known that this stabilised stress strain behaviour observed under cyclicloading may be quite different to that observed under monotonic loading Obviously this has significant implications in the application of the strain life approach For accurate estimates of strain it is important to use the correct stress strain relationship The difference in stress strain behaviour between monotonic loading and cyclic loading clearly demonstrates a dependence on load history This material dependence on the prior loading was first observed by Bauschinger 29 and is known as the Bauschinger effect Bauschinger observed that the yield strength in tension or compression was r
64. is Material Data Material StrainLife CI Strain Life Data Material Equation CI Equation inputs describing the Stress Strain and Strain Life Figure25 CGAP Material D ata Structure Diagram Crack Growth Parameters Table Material_FCGRTable Data material properties used for fatigue crack growth analysis Tadle10 CGAP Materials Data Structure Description CGAP Database DSTO TR 2392 Table Name Table Description Field FAMSH Variable Description Material CI Contains unique id MTITLE 1 9 Material specification material identifieras name MTITLE 10 45 Material description well as common A A MatFormat MATF The material format material properties Mat t 1 Tabul Links with all the s GE Se ES YO other crack initiation arene 2 Equation parameter tables E E Young s Modulus Sproportional PLMT proportional limit Sultimate FULT ultimate strength NPTH Sfracture fracture strength N_pts CycHys SF the number of cyclic hysteresis data pairs N_pts EN NPTS the number of material strain life data pairs CycHysType IM a flag that specifies the type of cyclic hysteresis data IM 1 cyclic strain versus stress data IM 2 hysteresis strain versus stress data IM 3 hysteresis stress strain versus stress data factor FCTR the factor required to convert data from ksi into psi Material_CycHys Cl Contains crack id MTITLE 1 9 Material specification initiation data for CycHysData HYS1 NPTH
65. is is Neuber s equation which relates the remote elastic stress range to the local elastic plastic stress strain range at the notch Figure 2 shows graphically how the N euber rule is applied to the stress strain relationship Neuber s rule for a case where K 1 suggests that the product of the applied elastic stress and strain is equivalent to the product of the elastic plastic stress and strain Geometrically this means that the area of A is assumed to be equal to the area of B ta Stress Q o yield e E Strain Figure2 A graphical representation of the N euber rule assuming a stress concentration factor of K 1 4 1 2 FAMSH Representation of Stress Strain The stress strain relationship in Figure 3 can be represented in various ways The Ramberg Osgood equation is commonly used to represent the monotonic stress strain curve but can also be used to fit the cyclic stress strain data In FA MSH the cyclic stress strain curve or hysteresis stress versus hysteresis strain curve as opposed to the monotonic stress strain curve can be represented in one of two ways either in tabular or equation format In tabular format the data may be cyclic stress versus cyclic strain or hysteresis stress versus hysteresis strain In equation format parameters are used to represent the cyclic stress strain curve Further information on creating material definitions can befound in section 5 3 2 Moredetail on the origins of the FAMSH stress strain
66. it Conversion I L a 29 5 4 Load tab 29 5 4 1 A u 30 5 4 2 A maaa tails n amaya awa E 30 5 4 3 MAXRNG Maximum Delta Stress a 31 5 4 4 SON on ne lord eer ee eile aos el 31 5 4 5 Resd Residual rees U na 31 5 4 6 Dmg Damage cut off sealed 31 5 4 7 Hours p r sequia aa ok kk a ki oi pe pan 31 5 5 Case Control tab iii aia ii 32 5 5 1 Proba Te odo O 32 5 5 2 No sref vs passes sref VS passesS I u 32 5 5 3 Analysis Output OPWONS iveco teen 32 5 5 4 Material Data Output I I L creen 33 5 5 5 PASS EIA Po E E asad a kuyaq ees epa tini kan Ghats 33 56 Spectrum file formats U U iia agendas 33 5 6 1 Block format with Damage Source Assignment DSA 34 5 6 2 Block format without Damage Source A ssignment DSA 35 5 6 3 Cycle format with Damage Source Assignment DSA 36 5 6 4 Cycle format without Damage Source A ssignment DSA 37 5 6 5 DBI SST input format eege I do 38 5 7 Analysis USING FAM SH ina ii 39 58 FAMSH output U U U U U Q u u uu u u uuu 39 6 MATERIAL DATA BASE cima 40 61 Materials D ata Structure 40 62 FAMSH interaction with the M ateri
67. kinlede C O A 1988 Probabilistic crack growth by nonlinear damage accumulation International Journal of Fracture 36 2 137 149 42 Origin 6 0 1991 2000 Northampton Microcal Software Inc 58 DSTO TR 2392 Appendix A The Basis of the FAM SH Stress Strain Relationship In FAMSH the following equation is used to relate the plastic stress and the plastic strain amplitude O O be O pi K e y Oa gt O pi 27 where o istheplastic stress o isthe proportional limit o isthetotal applied stress and K and n are the cyclic strength coefficient and cyclic strain hardening coefficients respectively Using similar assumptions to that of the Ramberg Osgood relation we can express the total strain as the sum of the elastic and plastic components of strain such that Z 28 Rearranging equation 27 we get Ws E ER _ ell gt G Oy 29 E E E lo O J o gt o a e p aa pl a pl E 30 ESS gt O lt O a E a Equation 30 can be used to represent the stabilised hysteresis loops of thestress strain relation ship using a Masing type material approximation pl 59 DSTO TR 2392 DSTO TR 2392 Appendix B Deriving M aterial Parameters for the FA M SH Strain Life Equation Thestrain lifeequation used in FAMSH hasa high degree of flexibility however determining material parameters for this equation is not immediately apparent This section provides a brief overview of one technique tha
68. l Limit 40 ksi ss Ultimate Strength 72 ksi 29 Material Properties Number of Data Stress Strain Data Number of Data Strain Life Data D El 0 368 736 1104 1472 184 Se 2xStress x 2xStrain Teak Cyclic Strain Cyclic Stress A Life Strain Amp 1 jo 0 1 Im 0 085 2 Joe 80 2 20 0 051 3 1 09 100 3 Ja 0 039 Strain Life Equation z 4 his 110 4 a 0 0325 EE s 182 118 3 s Je 0 025 11 2 J g 6 88 0 0 g Data Type Z EqStnEq Modified Morrow vi Strain Amplitude R 1 1E 3 181 182 183 184 1E5 16 17 Life in Cycles Figure27 Materials tab with the tabular material description active Now click on the Eq Stn Eq dropdown menu Select M odified M orrow for equivalent strain equation Enter lin the UnitCv box Now select the load tab Enter the following information refer to section 5 4 for more information on what each input is In the Spectrum drop down box select ASCII Damage Source Assignment Under Hours per Sequence enter 100 Under M AXRN G enter 30000 Under Scale enter 1 Under Resd enter 0 Under D mg enter le 7 Inthesection containing thetable with thetitle Stress Levels therearetwo columns titled No and Kn Inthefirstrow under Kn enter the K of location 1 which is 2 5 In the next row enter the K of location 2 which is 5 You will note that the number next to the box titled Stress Levels has incremented to 2indicati
69. l material properties are entered into the program After running the same problem using both the material inputs the following output is produced see Table 20 Table20 Total life predicted for various K values K pais Krea Difference 3 153341 8 155716 2 1 52 3 5 77144 6 78133 6 1 27 4 40037 6 40235 6 0 49 4 5 21161 3 21903 5 3 39 5 13673 3 14215 3 3 81 5 5 9718 0 9004 4 7 92 6 5632 5 5748 5 2 02 6 5 3786 7 3693 3 2 53 7 2449 8 2321 4 5 53 Although the inputs for the tabular definition match the equation definition exactly the results in Table 20 indicate that errors introduced during interpolation have contributed to discrepancies in thefinal results The associated sensitivity to input data particularly for small strains and long lives should betaken into consideration when interpreting strain life results using this technique 8 4 Damage Summation TheFAMSH code uses Miner s 25 rule to determine the level of damage Miner s rule uses the linear accumulation concept proposed by Palmgren 26 as a measure of damage This method of evaluating the amount of damaged incurred by each cycle does not account for load history effects In other words Miner s rule does not consider the effect the sequence of loads has on the accumulation of damagein a structure Ignoring these effects can resultin an order of magnitude difference in the final result 36 37 Improvements to the code could be implemented via the in
70. l step is to execute the analysis by clicking on the Run button After the analysis has been completed the following message should be displayed in the message window see Figure 33 ile Output File Plot File Spectrum File Problem loaded successfully Click Run to analyse it Casel running Analysis Gen_seq input completed successfully Figure 33 M essages returned after analysis 7 4 Results After the analysis has been completed successfully several output files are created Thesefiles are located in the same directory as the input file Thefiles created are Gen_seq output Gen seg passes Gen seg dmo Gen seg dsout Gen seg truncspc Gen seg passes contains the notch concentration factors and their associated fatigue life to failure measured in flight hours For this example we expect that the output will return the values presented in Table 19 Table19 The estimated fatigue life at each notch feature using CGAP Notch concentration factor Estimated fatigue life flight hours 2 5 1 040 481 5 37 952 50 DSTO TR 2392 8 Discussion Through the development and verification of the CGA P FA MSH module a series of test cases were developed These cases were designed to test certain aspects of the CGAP FAMSH module Some of these test cases have been included in Appendix E for the users benefit A description of each caseis provided in Table 28 Thesetest cases were primarily used to verify the correct execut
71. lements the Pal mgren M ner ruleto calculate fatigue damage but it must be remembered that there are other techniques of fatigue damage calculation Since 1945 a plethora of methods have been proposed to evaluate fatiguedamage mainly to account for the observed effect of the load sequence on the total life of structures 27 17 DSTO TR 2392 5 Using FAMSH in CGAP 5 1 Starting FAM SH When CGAP islaunched fromthe Microsoft Windows Start menu the user is presented with the screen as shown in Figure 9 By default the first module displayed isthe CGAP Solver for crack growth analysis Ble bit Ven Commands Tools Help g Now open Save import sport Cri Q Sexe 1 Gacewoy Metid Leed Care Control On Geometry Database Crach Configuration ER Xx IN TYE 1 Cortes cack benson under S i asa ss ss sss ss Running CGAP Solver Elastic Swett Fi Figure9 Default Start up mode for CGAP To use the strain life model the user must first switch to the FAMSH module in the CGAP GUI This can be achieved in two ways Either by pressing the Configure button located in thetop tool bar or by selecting tools gt configuration on the menu bar The following Settings dialog box should appear Figure 10 18 DSTO TR 2392 Settings Solver Module Name FAMSH A CGAP Solver Fastran3 8 FCGR Converter AMSH Graphical user interface for the crack i
72. life approach From a continuum mechanics perspective thestrain life approach empirically relatesthelocal elastic plastic behaviour of a material to fatigue damage Thus understanding the elastic plastic behaviour of the material is fundamental to the application of the strain lifeapproach Given acyclicload history a methodology is required to accurately determinethestress and strain at the notch root FAMSH uses the N euber s method which will be discussed in the following section 4 1 Notch Stresses and Strains To apply thestrain life approach a means to accurately determinethestrain at the location of interest is required usually theroot of anotch on astructure Dueto stress concentrations that occur at such locations local stresses may exceed the yield stress or proportional limit of the DSTO TR 2392 material Under such conditions the local stresses and strains can no longer be determined using the stress concentration factor alone A relationship such as N euber s equation is needed to supplement the elastic equilibrium equations and the stress strain relationships This simpleapproach providesa relatively sound prediction of plastic response and issuitablefor dealing with fatigueloading The discussions presented hereare confined to planestress and uniaxial stress only which are the cases dealt with in FAMS and FAMSH 4 1 1 Neuber s Method Several methods exist that may be used to estimate the notch root stress and strain bu
73. n Fatigue Analysis of M etallic Structures Fatigue Analysis of M etallic Structures Hu Graphical U ser Interface Non Destructive Testing Original Equipment Manufacturer Structural Life A ssessment Program Structural M anagement Plan System Program Office Turning Point Turning Point Pair DSTO TR 2392 1 Introduction The assessment and assuranceof fatiguedamage predictionsin structures is an essential part of the management of air vehicles within the Australian Defence Force The AAP 7001 053 1 states that to assure structural airworthiness management of Aircraft Structural Integrity ASI is essential from the acquisition stage when the specification is developed through to fleet retirement The through life support of aircraft necessitates the assessment of fatigue damage caused by serviceloads and its effects on structural airworthiness As such methods of fatigue evaluation areimportant in theoverall safe and economical management of aircraft In practicethe analytical assessment of fatigue damage substantiated by coupon component and full scale tests form the basis of technical data for the design of airworthiness structural integrity management plans Given that accurate fatigue life assessment is critical to the safe and economical through life support of aircraft it is important that we develop algorithms and tools that areintuitiveand user friendly to support the structural engineer in the process of quality assurance
74. n life approach has been implemented in CGAP e provide guidance for the new module e provide examples demonstrating the use of the new module Accurate fatigue life assessment and life extension of metallic aircraft structures directly benefits defence These benefits can be realised in terms of reduced operational and maintenance costs improved performance and combat readiness Improvements in these methods and tools can be achieved in several ways for example by providing consistent material data built in error checking and a familiar graphical user interface GUI All these and others will lead to improved robustnessin predictions which ultimately contributes to the more efficient and reliable management of ASI fatigue advice The advice guidance and developments contained in this report contribute directly and or indirectly to the following important high level issues 1 Quality assurance is an integral part of ASI management 2 A widerangeof design philosophies ASI management philosophies fatiguemodals material types and analysis options must be addressed to support A DF aircraft both in being and in acquisition 3 Australian defence aircraft are sustained by multiple partners namely theindividual OEMs and the Australian Defence Support Network DSN 3 4 A common flexible and transparent fatigue prediction tool set is required to assureall of the above can be supported 2 CGAP Crack Growth Analysis Program CGAP isa
75. ng coefficients n via the following equations 1 PTOI ES 25 Q A 26 n 28 DSTO TR 2392 Refer to section 4 3 for a detailed description of the constants used in FAMSH to define the strain life equation Appendix B explains how to determine these parameters from experimental data Oncethe data has been entered the new material specification must be saved to the database To do this press the save button see Figure 19 next to the Material N ame input box Material Database 7075 T651 Test L Figure19 Click the save button to savethe material to the database Toremovethe material from the database usethe button marked with a cross nexttothesave button in Figure 19 5 3 3 Eq Stn Eq Equivalent Strain Equation The equivalent strain equation is also selected under the materials tab The available options are detailed in section 4 5 5 3 4 Strain Life Equation This menu is for information only and is not selectable by the user Its function is to indicate which strain life equation is being used in conjunction with the equivalent strain equation When this box either contains the word equation or remains empty then the default stain life equation equation 12 is used If it shows Tabular then the tabular strain life data is used 5 3 5 UnitCv Unit Conversion The variable UnitCv is used to convert the units of the material properties For example to convert from the Imperial unit ksi to the SI unit
76. ng thereare two cases to analyse When selecting the spectrum filethe user can chooseto sel ect the nameand path using the Spectrum file button see Figure 28 If the spectrum file has the same name and exists in the same directory as the input file then a path and name do not need to be specified In this case just delete the default name cstamp and leave it blank otherwise click on Spectrum File to browsethefilesystem and select thespectrum file 47 DSTO TR 2392 Spectrum File cstamp Spectrum Path Figure 28 Spectrum fileinput Now select the Case Control tab and enter the following e Select serf vs passes and under the Number of Passes to Analyse dropdown menu select First Pass e Under the Material Data Output dropdown menu select Cyclic Stress Strain This option will output cyclic stress strain data to the output file e Leaveall other options as default After you have entered all the above data the Case Control tab should look like Figure 29 Material Load Case Control Dutput Case Database z x r E Problem Title Description Add your own notes here Crack Initiation Life Analysis Input Unit Dutput Unit Imperial Unil w Dutput Options Ze No sref vs passes Material Data Output sref vs passes No Material Output vi Analysis Output Options Basic X Number of Pass to Analyse w Figure29 The Case Control tab Now select File
77. nitiation program FAMSH Cancel Figure10 TheCGAP settings dialog box In this dialog box click the drop down box titled Solver Module Name and select FAMSH Then click OK FA MSH should automatically link to the database file on installation however the database can be selected or reselected at a later stage by using the O pen Database File option under the File menu TheFAMSH module should now be loaded and you are ready to begin The module should look like Figure 11 The words Running FAM SH should appear in the bottom left panel to indicate that the FAMSH module has loaded properly CGAP1 beta Sample Cyclic Stress Strain Curve an yok streap strain tor 2024 1351 Si e Top Left Panel Stee EES a fe s Main Edit Window Ultimate Shang 77 Matena Properties 16 24 32 2xStreas x 2xStrain ol across aif ans af awa P O 165120 In DEI manm 1 Swan Ampituda 16 A VI mp au ma us ws wr yon Life in Cyclos Bottom Left Panel Main View Window Figure 11 FAM SH moduleloaded in CGAP 19 DSTO TR 2392 5 2 Data Inputs and Outputs TheFilename input file can be created manually and used as an input but CGA P will create this file for the user when the user input their data into the GUI The input files used to specify a problem prior to analysis are as follows e Filename input e Filenamespectra e Ftn
78. nt strain equation is requested The modified loopin equivalent strain equation is requested The Walker equivalent strain equation is requested The Smith W atson Topper equivalent strain equation is requested The F 18 equivalent strain equation is requested 73 DSTO TR 2392 Tadle29 FAMSH Test Caselnputs Flight exce Multiple CaseNo IPT1 IPT2 IPT3 IPT4IPT5 Kn Material MATFUNITCV RS NMLT SREF DCUT ESE m o PB Y A PSABock Se 1 1 2 1 0 o 1 D7075T651 1 1 0 2 255 LE O 2 15000 YES CYCLE Multiple 2 1 2 1 0 o 1 D7075T651 1 1 0 2 255 LE O 2 15000 YES CYCLE Multiple 3 1 12 1 0 o 1 D7075T651 1 1 0 4 255110 1E 07 2 15000 YES CYCLE Multiple 4 1 12 1 0 0 1 D7075T651 1 1 0 2 255 LEI 2 15000 YES BLOCK Single 5 1 2 1 0 o 1 D7075T651 1 1 0 2 255 LEI 2 15000 YES BLOCK Multiple 6 1 12 1 0 o 1 D7075T651 1 1 0 2 255 LE O 2 15000 YES CYCLE Single 7 1 2 1 a o 1 D7075T651 1 1 0 2 255 1E 07 2 15000 NO CYCLE Single 8 1 12 1 a o 1 D7075T651 1 1 0 2 255 1E 07 2 15000 NO CYCLE Multiple 9 1 12 1 1 0 1 D7075T651 1 1 0 2 255 1E O7 2 15000 NO BLOCK Single 10 1 2 1 a o 1 D7075T651 1 1 0 2 255 LE 07 2 15000 NO BLOCK Multiple 11 o m 1 0 o 1 DNSTGI 1 1 0 2 255 1E 07 2 15000 YES CYCLE Multiple 12 1 1 0 0 1 ste 2 1 0 2 255 1E 07 2 15000 YES CYCLE Multiple 13 1 1 0 o 1 D7075T651 1 1 0 2 255 1E 07 2 1 YES CYCLE Multiple 14 o m 1 0 o 1 DNSTGI 1 1 0 2 255 1E 07 2 15000 YES CYCLE Multiple 15 1 u 1 0 o 1 D7075T651 1 1 0 2 255 1E 07 2 15000 YES CYCLE Multiple 16 2 m 1 0 0 1 D7
79. one 03 9626 7000 Fax 03 9626 7999 Commonwealth of A ustralia 2010 AR 014 729 M arch 2010 APPROVED FOR PUBLIC RELEASE A Strain Life Module for CGA P Theory User Guide and Examples Executive Summary The assessment of fatigue damage in structures is an essential and an integral part of the management of air vehicles within the Australian Defence Force ADF The ADF Australian Air Publication 7001 053 states to assure structural airworthiness management of Aircraft Structural Integrity ASI is essential from the acquisition stage when the specification is developed through to fleet retirement As such the aircraft certification process requires that fatigue damage be assessed and critical areas susceptible to fatigue damage identified This highlights the importance of methods to evaluatefatiguedamageto the overall safe and economical management of aircraft Hence the need for robust user friendly software which the engineer can useto assess the fatigue life of critical structural elements Improvements in these methods and tools can be achieved in several ways for example by providing consistent material data the use of built in error checking and the use of a familiar graphical user interface This document discusses the incorporation of astrain life methodology into an existing softwaretool CGAP A detailed description of thestrain life theory and how it has been implemented into the CGAP environment is provid
80. ource Assignme D J DBI SST Format Binary Damage Source Assiqnment Binary No Damage Source Assignme ASCII Damage Source Assignment ASCII No Damage Source Assignme Figure21 Spectrum dropdown menu DSTO TR 2392 A fifth option is available DBI SST format8 This can be selected by choosing ASCII Damage SourceA ssignment and checking the D BI SST Format box next to the Spectrum dropdown menu For further information on DBI SST format refer to 6 5 433 MAXRNG Maximum Delta Stress The number entered here is the maximum applied stress range which is used to calculate damage distribution The maximum stress range can be calculated by subtracting the minimum spectrum stress from the maximum spectrum stress but currently a value of 60000 is used by default It should be noted that this value does not affect the calculated fatiguelife This value will be calculated automatically in the next version of FAMSH 5 4 4 Scale This was originally referred to as Kn in the FAMS code but has been renamed to Scale in CGAP to better reflect its functionality This value is essentially a scaling factor to the spectrum The actual applied stress is the product of this value the level of loading and Kn 5 4 5 Red Residual Stress Residual stress can be introduced using the Resd input under the load tab The residual stress should use the same units as the proportional limit 5 4 6 Dmg Damage cut off This is th
81. plied at two locations on a structure manufactured from 7050 T 7351 aluminium The locations on the structure will be denoted location 1 and location 2 Both location 1 and location 2 have geometric features that resultin a stress concentration Assume that we know the N euber stress concentration factors Ky in these locations are 2 5 and 5 respectively For the purposes of this example consider a spectrum composed of the following sub blocks of constant amplitude loading shown in Table 15 DSTO TR 2392 Table 15 Example spectrum loading Constant amplitude sub DSA R ratio S max psi Number of Cycles block number 1 22 1 12000 5 2 19 0 5 12000 20 3 5 0 8 10000 100 4 19 0 5 12000 20 The aim of the analysis is to determine the estimated time to failure at location 1 and location 2 7 2 Inputs Wewill call this example Gen_seq and nameall our input files accordingly To perform the analysis we need to create 3 input files These files are the input spectra and ftn07 files Thefirst step is to create the spectrum file For this examplethe spectrum filewill bein ASCII format in blocks and using Damage Source A ssignments which can be created using any text editor Section 5 6 1 presents the description of this format for the spectrum file Table 16 shows the complete spectrum file Line 1 contains three numbers Thefirst number 1 indicates oneflight has been defined in thespectrum file The se
82. r evaluation of this technique 10 Conclusion This document summarises the important aspects of the strain life theory and explains how this method isimplemented into the CGA P environment using thestrain life moduleFAMSH The document also provides the reader with guidance on and presents examples demonstrating the use of the new module Theaddition of the FA MSH strain lifealgorithm into the CGAP tool set introduces consistent material data handling built in error checking and afamiliar graphical user interface aiding quality management processes These improvements will lead to improved robustnessin the predictions which ultimately contributes to more efficient and reliable management of aircraft structural integrity 55 DSTO TR 2392 11 Acknowledgements Theauthors would liketo acknowledge and thank David Mongru for hisinvaluable assistance and contribution in the design of the material database for FAMSH and creating the preliminary graphical user interfaceelementsrelatingto FAMSH Theauthorswould alsolike to acknowledge the editorial contributions of David M ongru Marcus McDonald Manfred Heller and David Saunders who all provided their time to review and provided invaluable feedback in the course of production of this document 10 11 12 13 56 12 References Technical Airworthiness M anagement M anual 2005 7001 053 A M 1 Australian Defence Force Teunisse B et al 2006 P 3C ServiceLifeA ssessment
83. rain The Cyclic Stress Strain option outputs a table of data describing thecyclic stress strain curve The H ysteresis Stress Strain option outputs a table of data that describes the hysteresis stress strain curve 5 5 5 Passes The Passes dropdown menu provides the user the option to use one or two passes in the analysis process 5 6 Spectrum file formats There are essentially five spectrum file formats available and they are e Block format with Designated e Damage Source Assignment DSA e Block format without Damage Source Assignment DSA e Cydeformat with Damage Source Assignment DSA e Cycleformat without Damage Source A ssignment DSA e DBI SST input format The format of the spectrum files is important if the analysis is to run correctly Thefollowing sections describe in detail the format to use for each of the spectrum file formats 33 DSTO TR 2392 5 6 1 Block format with Damage Source A ssignment DSA The layout and format of the spectrum file that is required when using the Damage Source Assignment and Block type inputs is presented in Table 4 Table4 Theformat of a FAM SH input file using DSA and Block formats fixed format Example DSA Block input 2 1 BLOCK Keng The Arbitrary Spectrum number number format of flights thatis not used 1 38 FLIGHT NUMBER MISSION TYPE i K Flight Mission General description not used in number Type the analysis 2 1 000 TURNING POINT PAIRS amp FLI
84. re strength is known e The Smith Watson and Topper equation provides good results in most cases and for aluminium alloys produces somewhat more accurate results when compared to the Morrow equation e In cases where there is enough data to determine the exponent min the Walker equation the Walker equation gives superior results e For high strength aluminium alloys the exponent min the Walker equation is approximately 0 5 reducing the Walker equation back to the Smith Watson Topper equation For lower strength aluminium alloys the exponent is higher FAMSH implements the following eight equivalent strain equations i Modified Morrow equation 19 where oean isthemean stress and o isthetrue fracture strength the engineering stress at mean the beginning of fracture during a tensile test ii Loopin equation 20 Both the origins of the Loopin and modified loopin equations are described in 20 B O nale O mean ZO mean a 17 cg Ea 3 Ze O mean lt 0 where o isthe amplitude of the local stress and 8 is a material parameter 5 True fracture strength is defined as the load at failure divided by the actual cross sectional area at failure 15 DSTO TR 2392 iii Modified Loopin equation 20 B O mean O mean Z Oa W 18 Eeq Ea z zl O mean lt 0 Oa where c and 8 are material parameters iv Walker equation 21 Ss l m EV s l 19 E where Cna is t
85. s 5 7 193 226 Fatigue Design Handbook 1968 SA E Graham J A ed Warrendale PA 14 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 DSTO TR 2392 Matsuishi M and Endo T 1968 Fatigue of Metals Subjected to Varying Stress In Japan Society of Mechanical Engineers Fukuoka Japan Fuchs H O and Stephens RI 1980 M etal Fatigue in engineering John Wiley and Sons USA Ellyin F 1997 Fatigue D amage Crack growth and Life Prediction London Chapman and Hall Dowling N E 2009 Mean stress effects in strain life fatigue Fatigue and Fracture of Engineering M aterials and Structures 32 12 1004 1019 Dowling N E Calhoun C A and Arcari A 2009 Mean stress effects in stress life fatigue and the Walker equation Fatigue and Fracture of Engineering M aterials and Structure 32 3 163 179 Dowling N E 1999 M echanical Behavior of M aterials Engineering M ethods for D eformation Fracture and Fatigue Upper Saddle River Prentice Hall Inc Porter P G and Liu A F 1983 A Rapid M ethod to predict FatigueCrack Initiation V olume1 Technical Summary N A DC 81010 60 H awthorn CA Northrop Corporation Shah B 2004 P 3C SLAP Program Phase II Il1 Slides Reno N evada Lockheed Martin Aeronautics Co M olent L and Ogden R 1998 Review of theRAAF F A 18 Structural A ppraisal of Fatigue Effects SAFE Version 112 Volume 1 Detailed
86. s the path of the input file The Table 25 following lines contain the Kn values specified in theinputfileusing the source variable SREF along with the evaluated fatigue life in flight hours to failure filename truncspc This file contains the damage based truncated spectrum This file is Table 26 equivalent to the ftn11 file produced by FAMS filename output This file contains all the output produced by FAMSH The Table 27 information in the output file is controlled by the case control tab in CGAP 39 DSTO TR 2392 CGAP incorporates a new material database that links seamlessly with the CGAP environment for crack growth and fatigue analysis The database allows the user to select retrieve and store material properties to perform analyses The use of an integrated material database in comparison to individual material data files provides greater consistency This thereby minimises errors associated with incorrect material input files and improves traceability of the analysis results The database also provides support to quality control of outputs of the analyses for certification and validation purposes 6 1 Materials D ata Structure A diagram of the CGA P materials data structure is provided in Figure 25 The fields in the 6 Material D atabase four data tables pertaining to crack initiation properties are described in Table 10 Material_Cl material properties for strain life analysis Material CycHys Cl Cydic Hysteres
87. sequence The load sequence the stress concentration factor the stress strain relationship and the Neuber s ruleareused to determinethelocal stress and strain history Thelocal stress history is then processed using the Rainflow counting method to form pairs of peak and valley stresses or stress cycles that relates to fatigue damage The counted stress cycles are converted to strain cycles which are in turn converted to equivalent fully reversed strain cycles based on the selected equivalent strain equation Finally the converted equivalent strain amplitudeisused to determinethenumber of cyclesthe material would haveendured under this strain amplitude to cause failure The inverse of this number gives the fatigue damage attributable to this strain cycle By summing up the damages caused by all the Rainflow counted strain cycles thetotal damage by the spectrum is obtained Theinverse of the total damage gives the fatigue life in terms of the number of passes of the spectrum It should be noted that this approach does not account for the history effect of the load sequence It uses the Miner linear summation rule The details of the main blocksin Figure 1 are discussed in the following sections Cyclic O Strain Life NES Load spectrum Z N Cyclic Neuber s rule counting I Predicted Ae f AE On HS SE Aed 2 tress strain response Figure1 Block diagram illustrating the process of strain
88. t can be used to determine the input parameters for the FAMSH strain life equation The method presented here is one method but the reader is reminded that there maybe other better methods of evaluating these parameters For the purposes of this guide we have chosen to use the tabulated data for 7050 T 7351 from theCGAP material database By using theterms PLET and ELET wecan dividethestrain life curveinto three sections see Figure 36 The parameters PLET and ELET are determined by the user and the available tabulated data Ae Cl C2 W Te en la 2 Net SE SE 3 j 5 Ae Cl C4 2 Region 1 S a tie Region 2 e OPLET 222 2 Ko lt e Ae C3 C4 E Region 2 A Region 3 E ELET eee y 2 NP N UN Ke Region 3 Reversals to Failure 2N Figure36 Strain life curve representation in FAM SH Using the tabulated data for the material 7075 T 7351 we can create three data sets The first data set contains all the data with a strain above PLET that is all the data in Region 1 The second data set contains the data in Region 2 between PLET and ELET and thethird data set contains all the data with a strain below ELET Region 3 After separating the 7075 T7351 data into these three sets the tabulated data should look something like that in Table 21 depending on your choice of PLET and ELET 61 DSTO TR 2392 Table21 7075 T7351 material data separated into three sets Original D ata Set Data Set 1 Region 1
89. t the most commonly used one is that by Neuber 7 The strain life module FAMSH in CGA P utilises the N euber method for evaluating notch root stresses and strains For elastic deformation thetheoretical stress concentration factor K can be used to evaluate the local stress According to the definition of K Ao K 1 1 where AS istheremoteelastic stress rangeand K isthestress concentration factor Similarly in the elastic regime we have NE K K 2 ae 2 where Ae is the remote elastic strain range and K is the strain concentration factor In the elastic regime both thestress concentration factor and thestrain concentration factor are equal to the theoretical concentration factor i e K K K N euber noted that in the post yield range these relations would no longer hold That is the stress concentration factor K would decrease and the strain concentration factor K would increase relative to the elastic stress concentration factor K Neuber hypothesised that the elastic stress concentration factor would be equal to the geometric mean of K and K i e K K K 3 H e Therefore using equation 1 and 2 equation 3 can be rewritten to give AAs K ASK Ae K AS Ae 4 As the remote stress is assumed to be within the elastic range we have AS EAe otherwise the whole cross section will yield Therefore 2 2 hes s gt l 5 DSTO TR 2392 where E isthe modulus of elasticity Th
90. troduction of more sophisticated non linear damage accumulation algorithms A number of such models can be found in the literature 27 DSTO TR 2392 9 Future Improvements The following modifications to the FAMSH code are under consideration for future CGA P releases e Theincorporation of a cyclic strain hardening model e Ability to queue and run multiple jobs similar to a batch file process e Additional and improved equivalent strain equations e Moreadvanced damage accumulation models to account for sequence effects e Theindlusion of other techniques to estimate the notch root stress e Seamless interaction between strain lifeand crack growth algorithms to providetotal life estimates e Inclusion or emulation of other strain life codes Medium to long term research activities that may benefit from the CGAP analysis environment and improvethe analysis capability include a study of the effect of net section size and surface condition on the fatigue limit Improved damage models may help to disassociate the portion of damage due to crack growth and the component due to strain hardening of the material This may lead to more robust material properties that may lead to a generalised predictive capability M ore recent developments in fatigue life prediction have utilised thestrain lifeapproach to predict crack growth directly 38 41 Further research and development of this approach could be introduced to the CGAP environment and allow furthe
91. used by different loads This information is recorded in the filename dsout ouput There arethree tabs that are used to enter data They are the M aterial Load and Case Control tabs 5 3 Material tab The material tab see Figure 12 allows the user to create a new material or load or update an existing material in the database The material properties including Young s modulus the proportional limit the fracture strength the ultimate strength the stress strain relationship and thestrain life relationship are availableto all the cases using the same database This tab also allows the user to select the strain life equation equivalent strain equation and to enter related parameters These data are not stored in the database they aresaved in the input file and hence only available for the current case being analysed 25 DSTO TR 2392 CGAP1 8 beta Sample ol onn e nl ane B f 0105 af au a el 05 ef 2158 Pra TO IA Q 5 ELET Tome Pef oma z EqSmEqf Sii I H UK Ge N HM HHIH li wo um 1 D EI Life in Cyclos Figure 12 The materials tab used in theFAMSH CGAP module 5 3 1 Loading an Existing Material Materials from the database can be selected using theM aterial Database dropdown menu see Figure 13 Material Load Case Contro Material Database 7075 1651 ksi Figure13 The material drop down menu
92. using the Damage Source Assignment and Cycletype inputs is presented in Table 6 Table6 Theformat of a FAM SH input fileusing DSA and Cycle formats Example DSA Cycle input 2 1 CYCLE Bi The Arbitrary Spectrum number number format of flights that is not used 1 38 FLIGHT NUMBER MISSION TYPE td K gt y Flight Mission General description not used in number Type the analysis 114 1 000 TURNING POINT PAIRS amp FLIGHT DURATION YN Number Flight General description not used in of TPPs Duration the analysis inthe flight 38001022 1 00 38001022 13408 00 Se The user has the option DSA disip kk DSA Minimum to enter another TPP on eae TP stress the same line 38002023 1773 00 38002023 9068 00 Bache sung et be contained within the 12 spaces 10spaces 10spaces 10spaces allocated spaces Integer Real Integer Real OneSpace OneSpace OneSpace 36 DSTO TR 2392 5 6 4 Cycle format without Damage Source A ssignment DSA Thelayout and format of the spectrum file that is required when using Cycletype inputs and no DSA ispresented in Table 7 N otethat the arbitrary number seen in the DSA block format in the first line should be removed Table7 Theformat of a FAM SH input file using the Cycle format with no D SA Example no DSA Cycle input 1 CYCLE A The Spectrum number format of flights Example Flight Spectra Title of the spectrum 114 1 000 TURNING POINT PAIRS FLIGHT DURATION Number Flight
93. would be defined with C2 C3 Gland B2 The fit should look something like that shown in Figure 38 Strain Data set2_Strain Model user21 Chi 2 DoF 1 3282E 9 R 2 0 99955 ci 0 00907 0 00596 c4 1 37415 2 00187 b1 0 05491 0 07232 g2 0 99716 0 31632 1000 Life Figure 38 Curvefit for data set 2 using Origin 6 0 When interpreting the results of the non linear curve fit it is necessary to identify the terms relating to the elastic strain and the plastic strain Remember that the strain life equation equation 12 isthesum of the plastic and elastic components of strain The plastic component of the equation is identified by the larger gradient on a log log plot Thus the larger of the two exponents in equation 12 identifies the plastic component 63 DSTO TR 2392 By fitting Region 2 first we now have values for C1 C4 B1 and G2 which we can use in determining the parameters in theother regions N ow using thefirst data set input C1and B1 as fixed parameters and use the non linear curve fitting algorithm to find C2 and G1 The result of the fitting should look like that shown in Figure 39 Data set1_Strain Model user21 Chi 2 DoF 7 0651E 8 R 2 0 99989 ci 0 00907 0 c2 0 5087 0 00599 bi 0 05491 0 gi 0 81999 0 00419 Strain 10 100 Life Figure 39 Curvefit for data set 1 using Origin 6 0 Similarly we now use C4 and G2 determined from our non linear curve fit of Region2 as
94. y or hits a previous path created by a previous droplet e Fromthisreversal allow adroplet to flow in the oppositedirection and terminatein a similar manner This now becomes the next Rainflow counted cycle Continue the process until all the cycles within the load spectra have been counted In Figure 8 thereare 4 full Rainflow cycles A D B C E F and G H which correspond to the 4 closed stress strain hysteresis loops For further information refer to Fuchs Fatemi and Stephens 15 or Ellyin 16 13 DSTO TR 2392 Ed Strain Stress 0 Strain Figure8 Pictorial Representation of the Rainflow Counting M ethod 4 5 Mean Stress Effects and Equivalent Strain Equations For fully reversed strain cycles i e R minimum stress maximum stress 1 the strain range can be directly used to obtain the corresponding fatigue life from the N curve For a variable load history the strain cycles are usually not fully reversed henceit is important to account for the effect of non zero mean stresses on fatigue life An equation that converts a general strain cycle e where R 1 into an equivalent cycle e where R 1 isknown as an equivalent strain equation and several models have been developed over the years 14 DSTO TR 2392 In general Dowling 17 18 provides some guidanceto the use of equivalent strain equations Heconcludes the following e The Morrow equation is reasonably accurate in most cases when the true fractu

A Strain-Life Module for CGAP - Theory, User Guide and Examples

Contents

Download Pdf Manuals

Related Search

Related Contents