Fatigue Testing

Dundalk Institute of Technology- Fatigue Test

Dundalk Institute of Technology Faculty Of Engineering Mechanical Engineering Department

, FATIGUE TEST

Importance and application

Submitted to:

Joe Byrne Mechanical Engineering Department Faculty of Engineering, Dundalk Institute of Technology Prepared by: James Gargan 29 February 2012

1


Table of Contents FATIGUE TEST, ....... .............. .............. .............. .............. .............. .............. .............. .............. .............. ............ ........ ... 1 Table of Contents ....... .............. .............. .............. .............. .............. ........... ........ ........ ........ ........ ........ ........ ....... ... 2 Abstract............ ........................ ........................ ......................... ......................... ........................ ......................... .............................. .................3 Introduction Introdu ction ............ ....................... ....................... ........................ ....................... ....................... ........................ .................... ............. ....... .. 4 Wire Failure Due to Fatigue – 0.0 ....... .............. .............. .............. .............. ............. .......... ........ ...... ..4 Aircraft fatigue Testing – 0.1....... .............. .............. .............. .............. ........... ........ ........ ........ ........ ...... ..5 Wind turbine fatigue - 0.2....... .............. .............. .............. .............. ........... ........ ........ ........ ........ ........ ...... ..6 Pin head fatigue 0.3 ........ ............... .............. .............. .............. .............. .............. .............. .............. ............. ...... 6 Case Study: High cycle fatigue of an un-cracked component- failure of a pipe organ mechanism........ .............. .............. ............ ......... ........ ........ ........ ........ ........ ........ ........ .... 8 Theory Theor y........... ....................... ........................ ....................... ....................... ........................ ....................... ..................... ............... .......... ......... .... 8 1.1 Metal Fatigue.................................................. ............................................................ .................... ............. ...8 Types of Fatigue ....... .............. .............. .............. .............. ............... ............... .............. .............. ........... ........ ....... ... 9 1.2 Fatigue loading....... .............. .............. .............. .............. .............. .............. .............. ............ ......... ........ ...... ..9 Example of fully reverse rev erse loading ....... .............. .............. .............. .............. ........... ........ ........ ....... ...10 1.3 Fatigue strength ...... .............. ............... .............. .............. ............ ......... ........ ........ ........ ........ ........ ....... ...11 1.4 Factors influencing fatigue ....... .............. .............. ............. .......... ........ ........ ........ ........ ........ ......11 1.5 Notch sensitivit sensitivity y factor ....... .............. ............. .......... ........ ........ ........ ........ ........ ........ ........ ........ ......12 1.6 S-N Graphs..................................................... .......... .................... ............ ..13 Definitions.................................... ................................................. ....................... .................... .................... ............ ..14 Results Result s........... ....................... ....................... ....................... ........................ ....................... ................. ........... .......... .......... .......... .......... ....... .. 14 Table of Experimental Exp erimental Observatio O bservations: ns:....... .............. .............. .............. .............. ............. ......... ... 14 Discussion Discussi on........... ....................... ....................... ....................... ........................ ....................... ................... ............. .......... .......... .......... ..... 17 Conclusion Concl usion.......... ...................... ........................ ....................... ....................... ........................ ....................... ..................... ............... ....... .. 19 References Refe rences ........... ....................... ........................ ....................... ....................... ........................ .................. ........... .......... .......... .......... ..... 20

2


Abstract The following report details an experiment conducted by final year mechanical engineering students, at Dundalk institute of technology. This report gives emphasis on a fatigue test conducted during class hours at the institute. Fatigue is one of the fundamental components of mechanical failure that is due to due to cyclic loading on mechanical parts or components, this can be clearly seen in many real life examples, such as in the manoeuvrability of aircraft during flight, as air resistance and wind changes in hit the wings of the aircraft sending repetitive forces causing stress to the aircrafts wing. It is therefore an important consideration in the design of such components not only in such a large scale but also in smaller scale component applications. The report outlines all the factors that contribute and influence the mechanical phenomenon that is fatigue and include a host of other associated engineering terminology. This report is mostly of an analytical nature and shows tables and graphs constructed during the course of the experiment. The fatigue test conducted was that of a typical traditional fatigue test, where the specimen chosen is of a specific size diameter specification, as well as material type and notch size. The results for this experiment were already given to us during the following experiment. This is due to phenomenon that is fatigue; fatigue will only occur over repetitive forces on the specimen, which typically takes, a considerable time is not instant. Therefore is not possible to be present for in this case the maximum approximate number of repetitive cycle that it took for fatigue to occur in the specimen, which was 3,000,000 cycles. There is also a discussion and conclusion section that links in the understanding of the results, with other factors found during the course of the experiment that would influence fatigue in the specimen. Since the number of cycles influence the fatigue life of a material, it is assumed that no matter what the material the larger the force applied to the component the less amount of cycles it will take for that component to fail by fatigue, likewise if a smaller force is applied to the component the number of cycles needed until the material fails due to fatigue should substantially increase. This was the case during the course of this experiment over various forces. All of which can be easily seen in the S-N graph, which is used to visually illustrate the fatigue process. All experiences, which I have learned and wish to comment on, are available in the presented discussion and conclusion all of which are presented in the following pages of the report.

3


Introduction In engineering design there is many applications where parts of a machine or device designed will experience varying amounts of fluctuated loads at different times and at certain points through out the lifetime of the material. For example in the aeronautical industry, the wings of a plane constantly undergo varying loads when the plane maneuvers in the sky countless times over its lifetime. The countless forces the wings undergo over a period of time could cause weakness at certain points of the material. Fatigue Analysis tells us how the material reacts to these loads with respect to Time. When a material is loaded too much, it may break even if the force exerted on the material is not the maximum force the material can withstand. This is due to the repeated loading and unloading which weakens the material over the space of time and consequently makes it fail.

Figure 1. Aircraft wings that are undergoing a structural fatigue analysis.

Wire Failure Due to Fatigue – 0.0 If you try and pulled the wire it takes a lot of strength and effort in order for the wire to break. Yet when you grab the wire and twist it from side to side repeatable over a period of time the wire eventually breaks even thought the force you pulled the wire by was no where near the force it took to break the wire when you initial tried to snap it. This is because the wire has failed due to fatigue because the force applied to the wire weakened the wire of the same force over the time

4


Figure 3 (a) Showing an initially straight wire before force applied, (b) & (c) the motion causing bending & (C) fatigue failure over exerted force over time.

When you bend a paper clip back and forth until it breaks, you are demonstrating fatigue behavior. Fatigue was the name given to this behavior as it was originally thought the material got tired a subsequently failed.

Aircraft fatigue Testing – 0.1 Fatigue forms part of nearly all-engineering analysis and can be important in order that the material chosen confirms with its application and duties that it may be subjected to. Engineers at Boeing for example conduct fatigue testing in all areas of all their planes in order to confirm their planes as safe for their intended use before their first flight. They accomplished this by putting the plane under a 100,000 simulated flights to match the damage the plane would ever see over the course of its lifetime. To do this the plane is placed in a make shift frame which simulates the maximum forces the plane will undergo during the course of flight during its lifespan. Fatigue testing is not like static testing, which sees the plane pushed to its extreme limits to see how strong it is. Fatigue testing shows how the aircraft structurally holds up over time when loads are subjected to it. This is why fatigue analysis and testing forms the integral part of the design of the plane, so that it will be deemed safe for flight. Fatigue testing also corresponds with the validation process authorities need to deem the aircraft safe for use, without fatigue testing there is no way of knowing when or were the material or part of the plane could fail at any point during its life.

Figure 2. Showing the mainframe in which mechanical devices impart forces to subject the plane through its simulation process.

Fatigue testing is one of the most important factors to human safety before engineered products can be considered safe to use. This has been seen in many cases, as driven disasters have occurred throughout history due to

5

Dundalk Institute of Technology- Fatigue Test inferior engineered products. Such an incident occurred in may 1842, when a Versailles train crash attributed to metal fatigue in one of the axles of the engine. The metal fatigue was caused by repeated stress cycles caused by constant pressure on the rail axels (loading & unloading). Rail axles are structural elements that undergo constant repeated stresses of rotation and transmission of vertical loads that they have to be able to subdue. A further investigation showed a growth in cracks that lead to a build up of stress concentration and consequent failures. This proves the importance of fatigue testing, as if the axel was properly fatigue tested the analysis would have shown weather the material used or process of manufacture would have been deemed suitable for the application.

Wind turbine fatigue - 0.2 Fatigue testing can also be seen in energy renewal sector, as wind turbines have to undergo constant fatigue tests as the structure of the wind turbine are undergoing constant repeated loading from alterations of wind and other weather conditions load over time. Structural failures demonstrate this. Cracks frequently occur at welded joints, in most cases at fillet welds. But cracks are also found at notches in the base material or in bolts. Problems can arouse due to the fatigue strength of welded joints and bolts. This is due to the complexity of the wind turbine where it is hard to find the loads that a wind turbine will have to withstand over its lifetime.

Figure 3. Wind turbine that’s structural integrity failed due to metal fatigue.

Pin head fatigue 0.3 Metal fatigue however can occupy smaller components also over time such an example can be seen in a pin. If the pin experiences a change in geometry due to a weld or gash on the pin head for example, a deformity could eventually grown onto the pins head into the form of a crack. If the dimensions of this crack are known it is possible however to choose a load that wont cause the crack to grow. However the crack will grow even if the load is even lower than this, if it is regularly repeated in variance or if the area surrounding the crack is corrosive.

6


Figure 4. Microscopic view of crack present on the surface of a pin.

7


Case Study: High cycle fatigue of an uncracked component- failure of a pipe organ mechanism. An unusual case of fatigue failure occurred in a pipe mechanism of an organ two years after it had been re-built. The problem was caused by failed mechanical linkages in the between the pedal board and the pallets in the wind chest of the organ. These mechanical linkages were part of the re-build and so were not expected to fail. Further studies showed that the main cause of failure was by an inappropriate use of an aluminium-copper alloy in a small part of the pedal action. Once the location of the failure in the pedal action was marked. Further analysis showed that the problem arose in the pedal through a horizontal notch that was connected to a vertical tracker arm. The horizontal bar that’s diameter was 4.75mm had been turned down to 4mm diameter to fit in a hole in the roller and then was riveted to hold it place. The turning operation was conducted with a sharp-ended lathe tool, which gave a negligible fillet radius where the 4mm diameter part of the horizontal bar met the main body of the rod. The fatigue crack initiated at this fillet radius in the sharp change of the section. The horizontal bar containing the notch was put into tension when the vertical tracker arm was pulled down. The crack had initiated in the region where the maximum tensile stress was transmitted downward across the 4mm section. The overall result were design modifications on the organ to reduce the likelihood of the incident happening again. It was decided to eliminate the notch on the horizontal round bar and to use a material that had higher fatigue strength. Recreating replacement arm with these new modifications did this. The rivets were also made redundant and instead it was decided to fix the arms with an anaerobic adhesive instead of the formerly used rivets. Reasons for this were that removing the additional material of the rivets actually increased the strength. The Organ was then tested and the design life was taken to be 100 years.

Theory 1.1 Metal Fatigue Metal fatigue is caused by repeatedly applying a nominal load to and from a metal part. This is known as the cyclic load. After this cyclic load is applied over a period of time of load-unloaded cycles will cause the metal part to break. This will occurs even when the maximum cyclic stress is much lowers than the ultimate tensile stress or yield stress. These are the stress levels that categorised the strength of the material. Metal fatigue can be reduced if the consequent magnitude of the cyclic stress has been reduced, this will

8

Dundalk Institute of Technology- Fatigue Test result in the part surviving more cycles before breaking.

Types of Fatigue In components there are basically two types of component fatigue. Fatigue of un-cracked components was no cracks pre-exist on the component. Such examples where these may be seen are on small components such as on gudgeon pins, gear teeth, axels, crankshafts or drive shafts. Fatigue on cracked structures, were cracks are present in the structure, is normally found on large structure especially on welded joints such as on bridges, ships or pressure vessels.

1.2 Fatigue loading There are different types of fatigue loading. Zero-to-max-to zero – This is where a part, which is carrying no load, is then subjected to a load, then the load is removed, so the part goes back to the no-load condition. An example of this type of loading is a chain used to haul logs behind a tractor. When the driver drives the tractor pulling the logs along the ground, depending on terrain or in general maneuvering stops and moves off to avoid obstacles causing cyclic loading on the chains.

Figure 5. Tractor pulling logs causing cyclic loading on chains.

Another type of fatigue loading is a varying load superimposed on a constant load. The suspension wires in a railroad bridge are an example of this type. The wires have a constant static tensile load from the weight of the bridge, and an additional tensile load when a train is on the bridge. Fully-reversing load – This case of fatigue loading is known for being the worse case. It occurs when a tensile stress of a certain value is applied to a part that is unloaded and then is released, and a compressive stress of the same value is applied and released. An example of where this type of fatigue could occur is on a rotating shaft when a bending load is applied to it. When

9

Dundalk Institute of Technology- Fatigue Test the shaft is in a fixed position and not rotating but subject to a bending load. The upper face of the shaft due to the bending load will be loaded in tension; whereas the opposite bottom face of the shaft will be loaded under compression. If the shaft was then further twisted and rotated by 180° in a bearing and with the loads remaining the same, the shaft would experience stress levels that are the same due to the bottom face of the shaft that was loaded by compression before it was twisted is now loaded under tension and vice-versa. This is the type of loading was the loading conducted in this experiment.

Vice

Example of

Shaft

fully loading

reverse

Figure 6. (a) Shaft positioned in a fixed position in a vice.

Shaft under tension

Shaft Under compression Figure 7. (b) Shaft under bending loads before rotation has taken place.

Shaft under compression

Direction of rotation Shaft under Tension

10


1.3 Fatigue strength A single test consists of applying a known, constant bending stress to a round sample of the material, and rotating the sample around the bending stress axis until it fails. As the sample rotates, the stress applied to any fiber on the outside surface of the sample varies from maximum-tensile to zero to maximum-compressive and back. The test mechanism counts the number of rotations (cycles) until the specimen fails. A large number of tests are run at each stress levels, and the results are statistically managed to determine the expected number of cycles to failure at that stress level. The cyclic stress level of the first set of tests is some large percentage of the Ultimate Tensile Stress (UTS), which produces failure in a relatively small number of cycles. Subsequent tests are run at lower cyclic stress values until a level is found at which the samples will survive 10 million cycles without failure. The cyclic stress level that the material can sustain for 10 million cycles is called the Endurance Limit (EL).

1.4 Factors influencing fatigue Internal defects- Material that are internally damaged or defects such as trapped air or shrinkage will cause the material to have a lower fatigue life. Grain structure- The presence of scratches or deformations on the surface of the material will have a greater influence in a material where the grain structure is larger. In general materials with smaller grain structure have longer fatigue lives. Reduction in cross sectional area – reduction in diameter of a loaded component causes the local stress to increase above the of the background stress. The ratio of the maximum local stress to the background stress is called the stress concentration factor (SCF). SCF is highly dependent on the fillet radius to the shaft diameter this is because any sharp change in cross section of a loaded component causes the local stress to increase above that of the background stress. Fatigue failure will occur at the site of local stress concentration. Temperature - The presence of extreme hot or cold temperature decrease material fatigue strength. Material type – Some material are more resistant to cyclic loading than others, depending on their composition and some material have their composition adapted resist cyclic stress. Environment – Any environment that promotes weakness in material such as corrosion, erosion will reduce fatigue life of a material. Direction of loading – Fatigue strength can depend on the direction of the direction of the principal stresses. Residual stress – Stress that remains after the original cause of the stress, will weaken the cross section diameter of the material and so will decrease fatigue strength. Such an example of this residual stress may be seen on

11

Dundalk Institute of Technology- Fatigue Test welds, casting or cutting.

1.5 Notch sensitivity factor The notch sensitivity factor is a term used to describe how good a notch in a round bar is at resisting fatigue loading. In theory it is taken that any component with a sharp corner will always fail due to fatigue loading no matter how low the background stress. This is not always correct, as there are many real life examples of components that are used successfully in fatigue loading even though in theory they should not be used. Such an example of a component used to resist fatigue loading that has a sharp corner present is a chain. The fatigue loading is the same seen in previous figure. 5 of the zero-to-max-zero loading as stated in a previous section of this report. The relation to how the notch sensitivity influences the fatigue loading is basically due to the material at the fillet radius which in turn influences the background stress to the yield stress. A sharp notch produces a small process zone, which is where the crack starts, and this makes it harder for a fatigue crack to grow. The size of the angle also will affect how the crack forms and in general the smaller the zone where the crack starts the larger the fatigue strength of the component will be. Notch sensitivity curves are used in design when designing components with small fillet radiuses.

Figure 8. Showing values for stress concentration factor for different materials.

12


1.6 S-N Graphs S-N graphs are an empirical way for an engineer to illustrate visually the fatigue process so he may design against it. The s-n diagram works when a test specimen is exposed to a constant cyclic stress (N-m^2) and a number of loading cycles (N) until the specimen fails are determined. If the specimen is exposed to lower levels of load the time until the specimen fails might take million of cycles of cyclic loading. Due to this cyclic loading at lower loadings it could take a lot of time for the specimen to break so the time in the s-n graph is usually illustrated in logarithmical form, this is normally plotted on the (x-axis). The loading is then converted into the bending stress measured in (N/M^2) or Pascal’s (Pa).

Figure 9. A graph showing the ultimate Strength against the number of cycles until the wing failed (cyclic loading time)

. Some materials noticeably ferrous metals the S-N curve on the graph can flatten out eventually so that below the endurance limit failure does not occur no matter how long the loads are cycled. The engineer will specify a safety factor in the design stage in away that the stress should never reach the level to cause fatigue for a far distant number of cycles the parts life expectancy is expected. In fatigue testing it is commonplace for to test the life expectancy of up to twenty parts at ten or so different levels of loading and define an s-n graph with statistical confidence.

13


Definitions Fatigue: Fatigue is a term used to describe the failure of a material as a result of frequently varying or cyclic loading.

Yield Stress: The stress level at which a material no longer behave elastically but instead experiences a small permanent deformation. This is the limit at which the stress level at which the elastic limit of the material has been exceeded. Ultimate tensile strength: Ultimate tensile strength is the stress value at which the material will break under the influence of pure tensile stress. Second moment of area: Property of a cross section used to predict the resistance of a beam to bending and deflection around an axis that lies in the cross-sectional plane.

Results Table of Experimental Observations: Table 1. Showing all gathered experimental value and measurements.

Column 1

Load (N) 65 60 55 50 45 40 35

Column 3

Column2 Cycles to Failure 8264 8038 15980 13519 49322 32322 99720 106976 162790 413715 336971 344912 2876618 \

Figure 10. Measurements of Specimen

14

Column 4

7690 12003 64789 113640 376194 548700 3458797

Dundalk Institute of Technology- Fatigue Test Bending Stress =

= 8816.2 MN/m^2 Where: Y = Distance from neutral Axis. (9.5mm/2) I = Second moment of area m^4 M = Bending Moment (NM)

Table 2. Showing the calculated experimental values of stress.

Load (N) 65 60 55 50 45 40 35

Stress (MN/m^2) 1057.09 975.77 894.46 813.12 731.18 650.5 569.2

Cycles Failure 8264 15980 49322 99720 162790 336971 2876618

to Column1 8038 13519 32322 106976 413715 344912 \

Column 2 7690 12003 64789 113640 376194 548700 3458797

15


Figure 11. S-N Fatigue Testing also showing an illustration of simultaneous mean and cyclic loading.

Figure 12. S-N Graph of mild steel fatigue test

16


Discussion Overall fatigue test results show an S-N graph showing stress against time values of three tests on a mild steel specimen. As seen in the results section in above (table 2), all the stress values on each specimen are different at the same no of cycles during the test. This can be expected as not two or three specimens can be deemed 100% identical even though they could have come from the same batch or length of material. Each specimen could vary in grain structure especially on where the load was applied or in internal defects either when the specimen was being manufactured. Residual stress all though may be deemed unlikely the form of fatigue could have contributed in different specimens, as we don’t really know if the test specimens suffered previous tests since the day we conducted the experiment and has to be considered. In order to expect a correct statistical data for a fatigue analysis on a mild steel specimen to aid accuracy the test would have to be conducted with many specimens at specific varying loads of the specimen.

Figure 13. Illustration of bending stress as a bar is being rotated.

As results in table.2 show the experiment may be deemed a success as they show in theory what is expected to happen to a material in fatigue testing or in a material that fails by fatigue. In theory the specimen should fail over a long period of time even though the specimen is exposed to a load that is constant and under the amount needed to make the specimen fail instantly. The more the load is increased on the test specimen the more the stress is applied to the component, and so the likelihood of the specimen failing by fatigue increases. This can be seen in the results as the higher stress of 1057 MN/M^2 took a less number of cycles over the same logarithmic time (approx. 7600 – 8300) cycles until the specimen failed at this stress over time. When you then look at the lower loads that were applied to the specimen resulted in what is expected less stress as the force applied over the same area is reduced. Due to this stress being subjected to the material being smaller the number of cycles needed for the specimen to fail over time also consequently increased.

17

Dundalk Institute of Technology- Fatigue Test The fully reverse loading is also something that I would question during this experiment as the bar rotating may have a residual stress applied as the stress transformed when the bar was being rotated at 180°. I think a residual stress would remain when the bar was subjected to a loading, which would result in creating a compressive stress on top and a tensile stress on the bottom of the clerical test specimen. When the specimen was rotated by 180 degrees the type of stress due to the fully reverse loading would change so that the tensile stress would now be present on top of the cylindrical test specimen and the compressive stress present at the bottom and so residual stress may remain prior from before the bar was rotated and so could have influenced the change in the number of cycles needed in each of the specimens to fail by fatigue. This could also explain the result of the S-N graph 2 result as it can be seen that the test piece suddenly failed in the logarithmic time well before the other two specimens even though the composition and cross sectional area are the same throughout the three specimens. A gradual uniform curve that is expected to flatten out towards the end was present in the results of this experiment this also confirms the theory that the experiment was conducted correctly. In theory a ferrous metal such as mild steel should potentially flatten out in as SN graph as theoretically so that below the endurance limit failure does not occur no matter how long the period. The notch on the specimen piece also would not give a true representation on the how resistant the cylindrical bar is to fatigue failure. This is due to the reduction of cross-sectional area of the bar giving an unnecessary stress on the bending stress of the bar as the background stress is increased beyond the yield stress and so increasing the risk of fatigue failure. Making the notch redundant on the bar will increase the cross sectional area of the specimen, and so should reduce this error and give a truer representation of the correct fatigue for the diameter of the specimen. The notch being removed from the test piece could also alter the bending stresses on the test piece and so the s-n graph curves could alter as a result. The notch sensitivity factor could also be taken under consideration as there is no given angle of the notched specimen so there is no way of knowing how good the fatigue strength of the bar really is at with standing fatigue loading. Other noticeable outcomes of the test are the fact that the stresses at 731.18 MN/M^2 failed at 162920 cycles and yet this number of cycles for the same stress double when the test was conducted for the other two specimens. There is no analytical evidence that I can see for this occurring and I believe this may have been due to this material having internal defects present or the grain structure present within the material.

18


Conclusion If the notch was removed from the specimen will give a more accurate bending stress that will give alternative values for stress. This will give different results for the S-N curves on the graph. In order for the more precise analytical results the experiment would create more statistical information if it was conducted with more test specimens who would give a better mean evaluation on how well the specimen will be at able to withstand the fatigue loading at certain stress points.

19


References Versailles train crashhttp://www.brighthub.com/engineering/civil/articles/120966.aspx# accessed 20/2/12 Wind turbine fatiguehttp://www.germanlloyd.org/pdf/nafat.pdf - accessed 21-2-12 Fatigue informationhttp://www.engineersedge.com/material_science/fatique_failure.htm accesses 21-2-12 Fatigue informationhttp://www.epieng.com/mechanical_engineering_basics/fatigue_in_metals.htm accessed 21-2-12 Yield strength & Ultimate tensile strength- http://www.epieng.com/mechanical_engineering_basics/stress_and_strain.htm accessed 21-2-12 Engineering Materials 1 by David HR Jones, Published 1980, Rewritten 2006, Fatigue Failure- Pages 224-230. – Accessed 23-2-12 http://www.engineersedge.com/material_science/fatique_failure.htm

20

Fatigue Testing

Recommend Documents