This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Stress-Life & Strain-Life module (new features in 2011) • Updated GUI and environment • Repackaging of entry level product (with Stress-Life & Strain-Life modules). • New DTLIB based module in 2012
MSC Shaker (new in ver. 2012) • MSC Fatigue Shaker predicts the fatigue life of components subjected to a
single input random vibration load or sine sweep – a typical example of which would be a shaker table test. Shaker tables are routinely used to “proof test” components before sign-off. Typical input loads could be displacement, velocity or acceleration PSD’s.
• The module works for a single input loading only and the following methods are available,
FE Model of Car Body Component Location and Fatigue Life of Spot Welds
No attempt to directly calculate stresses in the spot weld, instead we use moments and forces in equivalent beams from which structural stresses are derived.
MSC Fatigue Spot Weld uses the Rupp, Storzel and Grubisic algorithms for computing stresses in each spot-weld nugget and in adjacent sheets.
MSC Fatigue Seam Welds (new solver in ver. 2011)MSC Fatigue includes, as standard, the traditional weld classification approach (BS5400/BS7608 etc) for the fatigue design of weldment details. Using this type of approach there is no attempt to model the weld detail. Instead, “component” S-N curves are used which have the weld classification detail (loading and geometry) built in to the S-N detail. This approach can, however, be awkward and time consuming to implement for thin sheet steels commonly used for automotive manufacturing because the level of integration with FE is minimal.
More recent work has focused on calculating the equivalent structural stress in the weld detail. The method implemented in MSC fatigue is based on the method developed by Fermer et al. Ref: SAE 982311
MSC Fracture Little known and under used crack growth tools embedded in MSC Fatigue which provide sophisticated crack growth modeling tools for estimating life to grow a crack through a structure. Features include:• Kitagawa minimum crack sizing• Fracture toughness failure criterion• Mean stress correction• User-defined life units• Rain flow cycle counting with cycle re-ordering• Initial and final crack length specifications• Plane stress correction• Notch effects modeling• Retardation and closure effects modeling• Modified Paris law modeling based on effective stress intensity range• Fracture mechanics triangle solutions (stress – stress intensity – crack length)• Graphical interface to NASA/FLAGRO 2.03 (via MSC Patran or MSC Fatigue Pre & Post)
Compliance Function (Y)LibraryStandard specimens
• Single edge crack in tension• Single edge crack in pure bending• Double edge crack in tension• Center cracked plate in tension• Center cracked square plate in tension• Three-point bend specimen• Compact tension specimen• Round compact tension specimen• Wedge opening load specimen• Quarter circular corner crack tension specimen
Cracks at holes•Single crack at a hole in tension•Double crack at a hole in tension•Surface crack at a hole in tension
Elliptical, semi-elliptical cracks in plates•Surface cracks in tension•Surface cracks in bending•Embedded cracks in tension•Embedded cracks in bending•Surface and embedded cracks in combined loading
Cracks at corners•Quarter elliptical corner crack in tension•Quarter elliptical corner crack at a hole in tension
Cracks in solid cylinders•Circumferential crack in tension•Straight crack in tension•Semi-circular crack in tension•Crack at thread in tension•Straight crack in bending•Semi-circular crack in bending
Cracks in hollow cylinders•Internal surface crack under a hoop stress•Circumferential crack in thin-walled tube in tension
Cracks in welded plate joints•Weld toe surface cracks in tension•Weld toe surface cracks in bending•Weld toe embedded cracks in tension•Weld toe embedded cracks in bending•Surface cracks in combined tension and bending
Cracks in welded tubular jointsCracks at spot welds in tensionUser parametric definitions
MSC Strain Gauge MSC Fatigue Strain Gauge allows the creation of virtual Software Strain Gauges within an MSC Nastran finite element (FE) model. These gauges can be used to produce analytical response time histories from the FE model under the effect of multiple time varying applied loads. Stress and strain time histories may be extracted at any point on the FE model surface, based on either standard or user-defined strain gauge definitions. The results obtained from the Software Strain Gauge may be based on static, transient, or quasi-static FE loading.
Aircraft wheels play a major role in the takeoffs and landings of an aircraft, whether it’s a 747 loaded with 568 passengers, the Space Shuttle, or an F-16 Fighting Falcon. Repetitive landings, takeoffs and associated taxi runs subject the wheels to a considerable spectrum of operational loads that the wheels must withstand time and again. Ensuring that a wheel meets stress and load criteria over time is an important part of the product development process and typically is accomplished by testing physical prototypes. However, building and testing a prototype is expensive and time consuming and wheel development programs often require several prototypes be evaluated before the production design is finalized.
Some components have multiaxial loads inputs, and some of those have multiaxial stresses and strains in critical locations. In these situation uniaxial methods may give poor answers needing bigger safety factors. MSC Fatigue includes sophisticated stress state assessment tools to test for the presence of secondary stresses and non-stationary stresses. MSC Fatigue then has available several methods for multiaxial fatigue calculations which include,
• 6 Critical Plane Methods & 1 Total Life Factor of Safety Method • Wang-Brown method, with and without mean stress correction • Normal Strain, Shear Strain, SWT-Bannantine and Fatemi-Socie critical plane
MSC Fatigue UtilitiesMSC Fatigue Utilities contains advanced and practical applications to help the MSC Fatigue user who has a need to collect, analyze, and post process measured data, such as stress or strain time histories, or to process such data to prepare for a subsequent MSC Fatigue analysis.
Advanced Loading Manipulation Arithmetic Manipulation Multi-Channel Editor Rainflow Cycle Counter Formula Processor File Cut and Paste Multi-File Peak-Valley Extraction Simultaneous Values Analysis Amplitude Distribution Auto Spectral Density Fast Fourier & Butterworth Filter Frequency Response Analysis Statistical Analysis
Advanced Fatigue Utilities Single Location Stress-Life & Strain-Life Analysis Single Location Vibration Fatigue Cycle and Damage Analysis Time Correlated Damage Multi-Axial Life Crack Growth Data Analysis Kt/Kf Evaluation
Graphical Display & Conversion Utilities Graphical Editing Single & Multi-File Display Two & Three Parameter Display Binary/ASCII Convertor Signal Regeneration RPC to DAC - DAC to RPC & Cross-Platform Conversion Waterfall File Create
MSC Fatigue Pre&Post provides the required graphical interface to efficiently and easily set up, run and post process an MSC Fatigue analysis. It is intended for the user who doesn’t need the full processing power of Patran to run MSC Fatigue or perhaps has an alternative pre & post for routine day to day work.
“The most important development in CAE Fatigue since the introduction of MSC Fatigue.”
• No large data files to transfer. • No complicated file management. • Significant reduction in CPU requirements. • Likely that whole fatigue calculation process can be done in memory. • Will open up opportunity to do full optimization for fatigue calculations. • Full Body Fatigue calculations, including dynamic behavior, will be
This defines a design optimization response. This entry already exists in Nastran but needs to have a few more responses defined as shown here for doing optimization taking fatigue life/damage into account.
Differentiating between a GUI Based Approach and Solver Embedded
MSC Fatigue (GUI driven)
When to use • Failure investigation. • Sensitivity studies (eg effect of load change). • Where stresses are from non Nastran solvers. • For highly interactive analysis. • Where a large amount of post processing is
anticipated.
Nastran Fatigue (Solver Embedded)
When to use • Well defined processes. • Large models. • Many load inputs. • For optimisation of parts or systems. • Simpler and more concise file management. • Extremely fast analysis. • BDF file auditable process.
Materials Characterization and Fatigue Testing Service – including Strain-Life and Stress-life, for Metallic, Composite and Advanced Materials
• Fatigue life testing, stress-life and strain-life• Single and multi axis, static and dynamic specimen
and component testing.• Full specimen preparation from stock materials.• Extraction of specimens from real components.• Strain-life fatigue testing in tension/compression,
fully-reversed bending and torsion.• Load controlled fatigue testing for the derivation of
stress-life or load-life parameters.• Static and cyclic thermal loading from -40°C to
1050°C.• Tension/compression from 5kN to 150kN. Torsion up
MSC Fatigue is an excellent platform on which to build company processes – it can also be linked with SimManager
How can MSC Fatigue be used in a large company structure
Auditable and Repeatable Processes MSC Fatigue is configured to run in a similar way to Nastran using an editable text batch file (cf BDF file). This leaves a useful auditable trail as well as a convenient means to run in batch mode.
Incorporate More Reliable Material Properties Because reliable materials data is an essential part of any fatigue and durability process MSC Software is pleased to now offer a comprehensive materials testing service (see separate brochure).
Comprehensive Solution MSC Software now offer a complete solution including advanced fatigue software (MSC Fatigue), advance materials testing facilities and, new with this release, a comprehensive engineering analysis(for fatigue) consulting service.
#3.1ANALYSIS TYPE = VIBRATION FEA RESULTS LOCATION = NODE AVERAGING = GLOBAL TITLE = P3DATABASE = VS_FRA.dbS-N DATA SET 1 = 817M40 S-N TYPE = Material M_DIRECTORY = CENTRAL FINISH = No Finish TREATMENT = No Treatment KF = 1 REGION = skin MULTIPLIER = 1 OFFSET = 0 WELD = N/A TENSOR TYPE = STRESS STRESS UNITS = MPA STRESS COMBINATION = MAX ABS PRINCIPAL MEAN STRESS CORRECTION = NONE VIBRATION METHOD = DIRLIK DESIGN CRITERION = 50. FEA ANALYSIS TYPE = TRANSFER FUNCTION FEA RESULTS TYPE = DATABASE TRANSFORMATION = BASIC EQUIVALENT UNITS = 1. FEU UNITS = SECONDS DATABASE = nmatsmas.dbNUMBER FREQUENCY STEPS = 45 NUMBER PSD INPUT = 3 LOAD FILENAME = FSFDS.PMX PSD_DIRECTORY = INPUT ID = 1 FREQUENCY ID 1 = 2.1-1.1-1- FREQUENCY ID 2 = 2.2-1.1-1- FREQUENCY ID 3 = 2.3-1.1-1- FREQUENCY ID 4 = 2.4-1.1-1-
Modal Participation Factors The basis of the Modal Participation Factor (MPF) method is that each loading input can be split up into the contribution factors associated with each mode shape for the structure. In Nastran both the modal stresses and modal participation factors can be extracted from a single sol 112 analysis. The modal superposition is then calculated as follows:
where σ(t) is the output stress tensor
σi is the stress tensor for mode i
φi(t) is the modal participation factor for mode i
How Testing Supports AnalysisProvision of material fatigue properties Verification of stress/strain analysis results Correlation of life predictions Provision of load data Final sign-off
How Analysis Supports TestingEliminating unnecessary tests Test acceleration Gauge type selection and positioning Test design
Fatigue Analysis vs. Fatigue Testing Testing is not a good way to optimise designs, but is always required for sign-off. Useful fatigue analysis requires verification and good test-based information. Neither Testing nor Analysis have exclusively the “right” fatigue answer; Best results are obtained when an integrated approach is adopted
Between 1852 and 1870, August Wöhler studied the progressive failure of railway axles. He plotted nominal stress Vs. cycles to failure on what has become known as the S-N diagram. Each curve is still referred to as a Wöhler line.
1.45 x 10-3 300 5001.21 x 10-3 250 25000.98 x 10-3 203 150000.76 x 10-3 157 1203000.68 x 10-3 140 4000000.60 x 10-3 124 10000000.54 x 10-3 112 30000000.45 x 10-3 93 5000000
Strain range recorded Derived stress range (MPa) Number of occurrences
Part of a steel structure, located permanently in the sea, is known to be susceptible to fatigue damage. Strain gauges attached to this part were monitored continuously during the first year of service producing the following information.
Specimen tests on the same material showed that the fatigue limit in air was 156 MPa and that in seawater (no corrosion protection) it was 110 N/mm2. In addition it was found that stresses above either of these levels produced failure according to the following relationship.
3333.825.10196.2 −= SxNAfter 7 years the corrosion protection fails. Determine the expected fatigue life of this structure after failure of the corrosion protection.
Would the original total life of 20 years be achieved?
Some software products allows fatigue analysis with multi-mean stress curves. These curves (S-N curves for different R-Ratios) which account for mean stress effects in S-N method by Goodman, Gerber or other empirical methods.
A component undergoes an operating cyclic stress with a maximum value of 759 MPa and a minimum value of 69 MPa. The component is made from a steel with an ultimate strength Su of 1035 MPa, an endurance limit Se (at 106) of 414 MPa and a fully reversed stress at 1000 cycles, S1000 of 759 MPa.
Plot a Goodman diagram with 2 constant life lines on it corresponding to 103 and 106
cycles. These must both go through Su on the zero mean stress amplitude (x) axis and the appropriate points on the stress amplitude (y) stress axis which are the endurance limit, Se, and S1000 values (see Figure below).
When the stress conditions for the component (Sa = 345 MPa, Sm = 414 MPa) are plotted on the Goodman diagram, the point falls between the 103 and 106 life lines. This indicates that the component will have a finite life, but the life is greater than 1000 cycles. This 3rd line intersects the fully reversed alternating stress axis at a value of 573 MPa.
By taking one vertical (zero mean stress) slice through this diagram the equivalent (zero mean stress) S-N diagram can be envisaged.
A 3rd line can be drawn through Su (on the x axis) and another point defined by the operating stress.
The value for Sn can now be entered on the S-N diagram to determine the life of the component Nf. (Recall that the S-N diagram represents fully reversed loading). When a value of 573 MPa is entered on the S-N diagram for the material used for the component, the resulting life to failure can be obtained graphically as Nf = 2.4 x 104 cycles. Alternatively, this figure can be obtained using the S-N equation stated earlier.
• Some software products allows fatigue analysis with multi-mean stress curves. These curves (S-N curves for different R-Ratios) which account for mean stress effects in S-N method by Goodman, Gerber or other empirical methods.
[1]. Extract peaks and troughs from the time signal so that all points between adjacent peaks and troughs are discarded.
[2]. Make the beginning, and end, of the sequence have the same level. This can be done in a number of ways but the simplest is to add an additional point at the end of the signal to match the beginning.
[3]. Find the highest peak and reorder the signal so that this becomes the beginning and the end. The beginning and end of the original signal have to be joined together.
[4]. Start at the beginning of the sequence and pick consecutive sets of 4 peaks and troughs. Apply a rule that states,
If the second segment is shorter (vertically) than the first, and the third is longer then the second, the middle segment can be extracted and recorded as a Rainflow cycle. In this case, B and C are completely enclosed by A and D.
[5]. If no cycle is counted then a check is made on the next set of 4 peaks, ie peaks 2 to 5, and so on until a Rainflow cycle is counted. Every time a Rainflow cycle is counted the procedure is started from the beginning of the sequence again.
Eventually all segments will be counted as cycles and so for every peak in the original sequence there should be a corresponding Rainflow cycle counted. There will be 5 cycles obtained from 10 peaks and troughs.
The Strain-Life (Local Strain, ε-N, or “Crack Initiation”) Method
• Relates local strain to fatigue damage at that point • Useful when cycles have some plastic strain component • Suitable for predicting life in components which are supposed to be
defect free, i.e., not structural joints, sharp features, etc.
• MSC Fatigue Features – Based on Local Strain Concepts – Mean Stress Correction – Elastic-Plastic Conversion – Statistical Confidence Parameters – Palmgren-Miner Linear Damage – User Defined Life – Cyclic Stress-Strain Modeling – Surface Conditions – Factor of Safety Analysis – Biaxiality Indicators – Multiple Loads
and the stress concentration factor,after plastic yielding.
Neither are known but Neuber suggested that the square root of the product of the stress and strain concentration factors was equal to Kt .Hence Neuber’s Rule is simply:
Re-arrangement of this Rule gives a useful equation:
• Tabular Results of: – Individual Nodes/Elements – Most Damaged Nodes/Elements – Statistical Summary of Damage Distribution – Interactive Results Interrogation of All Life and
• Relatively new field of engineering developed since start of 20th century. • 1913, Inglis, ‘Stresses in a Plate Due to the Presence of Cracks and Sharp
Corners’ introduction of infinite stress concentration. – Study of plate with circular hole & elliptical hole.
• 1921, Griffith elastic energy balance approach. (Tests on glass rods). – Crack propagation will occur if energy released on crack growth is sufficient
to provide all energy required for crack growth, i.e. energy cannot be created or destroyed.
• 1957, Irwin developed stress intensity factor approach. – Introduction of the term ‘fracture toughness, Kc’– Fracture process at crack tip cannot be related to local stress due to
singularity (see later notes). – SIF is used based on stress remote from crack tip.
• Increasing the yield strength of a metal by processes such as cold work, precipitation strengthening and solution strengthening generally decreases the ductility.
• Reducing temperature reduces toughness and ductility.
Ductility is the ability to deform irreversibly without fracture.
Definition: A structure is said to be damage tolerant when in a damaged state it can still sustain acceptably high loads. This term was introduced after numerous unexpected incidents involving military aircraft.
Note: The civil (commercial) aircraft manufacturers can seldom justify the enormous costs required to pursue an innovative research program. The military usually can.
[The accepted worldwide certification standards for aircraft civil certification are JAR/FAA Chapter 25.571.]
In 1969, a wing pivot fitting of an F-111aircraft failed under a steady 4g manoeuvre after 105FH. The failure resulted in loss of aircraft. The fitting was designed for 11g!
In 1970, the USAF started to develop a Damage Tolerance Philosophy in order to eliminate the type of structural failures and cracking problems encountered on many aircraft.
In 1974 a document entitled ‘Airplane Damage Tolerance Requirements’ MIL-A-83444 was issued.
In 1996, ‘Aircraft Structural Integrity Program’ MIL-HDBK-1530 – Approved for public release.
Satisfaction of the design guidelines can be achieved by consideration of what are now standard philosophy.
• Proper material selection and control. • Control of stress level. • Use of fracture resistant design concepts. • Manufacturing process control. • Use of qualified inspection procedures.
• ac This is the ‘critical’ or permissible length at which the structure is considered failed. (Kcexceedence or NSY).
• ad This is the ‘detectable’ crack length at which the chosen method of inspection is likely to identify.
• ai This is the ‘initial’ crack length as is a function of several factors including inspection technique employed, quality of access, available lighting etc.
As the majority of crack growth occurs during the early stages of propagation, the choice of aihas a profound effect on life and inspection interval.
• Visual. Feature must be easily accessible. Inspectable size is large (>50mm) resulting short (and expensive) intervals.
• Penetrant. Coloured, fluorescent liquid is sprayed onto surface, which penerates crack. Surface then washed carefully and a developer applied. UV light is then shone onto component to reveal any surface cracking.
• Magnetic particle. Similar to above except the liquid contains magnetic particles which when placed in a magnetic field and observed under a UV indicate any cracking. Component must be removed from structure and inspected in a special cabin. Components must be magnetic!
• X-Ray. X-rays pass through structure and are caught on film. Sensitive but not reliable for surface flaws. Component must be examined in laboratory.
• Ultrasonic. Probe transmits high frequency wave into material. The wave is reflected by the crack. The time taken between pulse and reflection indicates position of crack.
• Eddy current. Coil induces eddy current in the metal, which induces a current in the coil. Under the presence of a crack, the induction changes. Ideal for bores and holes. Cheap but sensitive technique.
• None! This ‘method’ assumes total failure, hence, adjacent structure must withstand redistributed loads. Results in a severe weight penalty in order to achieve a multi-load path ‘Fail Safe’ design. No inspection cost. Consider safe life approach.
Inspection intervals. Two types of inspection interval require calculating;
• Threshold
• Repeat
Generally the inspection requirement is governed by the end user. The engineering must comply…
The threshold inspection is the first time an aircraft has to be inspected. Usually this is taken as the fatigue life divided by 5. Or, what is more likely, the operator will determine when the first inspection will take place, eg 5 yrs, 15000 FH. This then sets the required fatigue life. i.e. Fatigue life = 15000 x 5 = 75000 Flying Hours.
The repeat inspection interval is the time from detectable to critical divided by a suitable factor, between 2 & 4 (dependant on confidence of data, test backup etc.) This then allows the inspector more than ‘1 bite of the cherry’ to detect the pre-critical defect.
Most Aerospace materials exhibit some ductility, but are brittle (or semi-brittle) from an engineering point of view.
• Overall behaviour to failure is elastic.
• Hence behaviour is termed linear elastic.
• Analysis using linear elastic fracture mechanics (LEFM).
LEFM assumes that the body deforms as a linear elastic material except in a small region near the crack tip. This requires that the global stresses are below the yield stress, typically σσapplied < 0.8σσyield for LEFM to be valid.
Not valid when macroscopic yielding occurs prior to fracture, this is covered by Elastic Plastic Fracture Mechanics, EPFM. (Beyond the scope of this course!)
ββ is a dimensionless factor dependent on geometry which is sometimes called a Compliance Function (Y). The factor includes crack length and crack geometry amongst other things
σσ is the stress which occurs remote from crack tip. MPa
a is the crack length. m – do not forget this!
If two cracked components exist each of same material, but differing crack lengths, they will respond in the same manner if they have the same SIF (K).
• Sometimes referred to as ‘dimensionless stress intensity factor’. • Independent of material properties & loading magnitude. • Accounts for geometric effects (boundaries). • SIFC solutions are generally plotted in a non-dimensional form, by dividing
through by a suitable parameter. • Can be considered analogous to the stress concentration used in fatigue
Consider a small crack in a large body subjected to a remote stress, σfield. The crack acts as a stress concentration – infact, theoretically, the stress tends to infinity as we get close to the crack tip and for an elastic material the stress distribution is given by the ‘square root singularity’. (A singularity stress field is a stress field that tends to infinity at some point in the body).
Thickness effects • Thin material does not permit a state of tri-axial tension to
exist, i.e. there is no through-the-thickness stress, σz = 0. This is the definition of ‘bi-axial’ stress.
• For plane stress, yielding takes place when the maximum principal is equal to the yield stress. For plane strain, much higher stresses are required.
• Yielding is plastic deformation which takes place by slip, it is therefore caused by shear stresses.
• Thick materials (> 6mm approx) plane strain, prohibit the ‘thinning’ of the plate local to the crack and if the stresses in all three planes are equal, ‘tri-axial’, then there is NO SHEAR.
(v) Integrate this equation between a = ai and a = af to give the number of cycles needed to grow a crack from ai to af. This is the predicted life of the component. A classical integration is adequate if the Compliance Function is constant. Other Compliance Functions will require a numerical integration, as will any non-constant amplitude signal. Check if ai is critical (Kmax vs Fracture Toughness).
dNda
dNda
dNda 21 ××==
(i) Estimate the size of crack likely to be present in the component when it is first put into service, ai.
(ii) From a measured value of KIc, estimate the maximum crack length, ac, which the component will tolerate when the applied stress reaches maximum tension. An expression for the crack-tip stress intensity factor will be needed.
(iii) Using the same expression for crack-tip stress intensity factor and a value for ββ, then calculate Kmax & Kmin at ai, and hence ΔΔK.
(iv) Substitute ΔK into the Paris equation to obtain a crack propagation rate. This will put da/dN in terms of crack length a.Calculate geometric mean if necessary,
• Kt is the theoretical stress concentration factor – Kt is geometry dependant – Value of Kt is seldom reached in ductile materials – Material yields and the loading (stress) redistributes
• Kf is the fatigue stress concentration factor – Kf is dependant on both geometry and material – Kf also known as fatigue notch factor
– Performed using plain (unnotched) S-N data – Theoretical stress concentration factor, Kt used – Fatigue calculation will be pessimistic
• IMPORTANT NOTE – Kt is always related to a reference stress. Some data sheets use net stress
and others gross stress. Care must be exercised in applying the Kt to the appropriate reference stress. American methodology often relate Kt to bearing stress.
• The prime function of a joint is to transmit load from one part to another. They include: – Lugs – Rivets– Assemblies of bolts and rivets – Bolts – Bondings– Welds
• Fatigue problems often occur within joints. • The concept of similitude is doubtful for this structure.
• A lug is a simple joint with a single pin or bolt. Often single load path. • Kt for lugs are generally high. • Fatigue limits can be very low due to secondary effects such as fretting and
pin-bending. • Size effect significantly influences fatigue life. • Life prediction excludes methods based on basic material fatigue properties. • Lugs are rarely critical statically.
• Bolts can be loaded in either shear or tension or both.
• If the ‘fit’ of the bolt is tight, then fatigue benefits can be gained. With an interference fit bush, the benefit is greater
• Significant ‘clamp-up’ can be generated, improving life, (friction load-path). In the railway industry this is often the only loadpath in many applications. Tight fit is expensive.
• Tensile fatigue spectra can be very damaging to under-head and to the threads. High Kt’s exist here.
• Threads and under head radii must be rolled in fatigue critical areas if subject to tension.
• Tension bolts can be pre-loaded to significantly improve life.
• The fatigue properties of a welded joint are completely different from those of the parent plate because of: – Fairly sharp and ill-controlled geometric features – Defects such as slag inclusions – Residual stresses (usually unknown) – Heat affected zone
• Fatigue properties of welds in a range of steels have much less variation than in the parent plate
● Sheets and welds modelled predominantly with 4-noded shells ● Sheets described by mean surfaces ● Thickness of weld elements equals effective throat, or about 2x sheet thickness ● Element length of about 5 mm ● Small radii not modelled
● Bending ratio r is calculated based on top and bottom surface stresses ● Flexible joints have more bending at weld toe, and high values of r ● S-N curve is selected based on values of r ● Stiff joints have lower fatigue strength
– Combined action of repeated loading and corrosive environment.
– Presence of corrosive medium results in essentially two phenomena. • Weakening of interatomic link in surface layer producing minute cracks. • Accumulation of corrosion products in microcracks further reducing fatigue
strength.
– A corrosive fatigue failure has a rough, craggy appearance.
The image cannot be displayed. Your computer may not have enough memory to open the image, or the image may have been corrupted. Restart your computer, and then open the file again. If the red x still appears, you may have to delete the image and then insert it again.
Principal stresses from transfer function analysis
In pure form (fibre) the carbon can have immense tensile strength (>5000MPa!). When made into a component, however, many factors reduce this strength significantly. They are:
• Pre-impregnation. • Variability. • Environmental Degradation. • Lay-up. • Notch sensitivity (a fatique concept but required for static analysis for CFC
Computer based example no 6: A Stress-Life fatigue analysis
Try running mSLF
Use mART to make a sensible time history – you will need to pick a material S-N curve first in order to decide on a sensible overall stress range. Don t forget to change the units to stress (MPa).
Use PFMAT to pick a material where the maximum range of stress in your time signal corresponds to about 1000 cycles.
Then run mSLF. Put any job name in and fill in all the required fields.
Experiment with the various options, including the stress multiplier.