Failure Analysis & Durability Improvement

	Failure Analysis & Durability Improvement

Home Page	Bearings	Gears	Electronics	Contact Us

Welcome to VEXTEC's website on Failure Analysis & Durability Improvement. The purpose of this site is to educate potential VEXTEC clients about analyzing failed components, understanding the mechanisms which cause damage, and strategies to improve life. Please visit the main VEXTEC site for informationa about our products and services.

Virtually any manufacturing company will at one point or another have a problem with their components lasting as long as they should. This happens in a wide variety of industries, including:

Aerospace
Automotive
Bearings - Please visit our Bearing Failure Analysis web page for more information
Gears - Please visit our Gear Failure Analysis web page for more information
Manufacturing
Medical Devices
Electronics - Please visit our Electronics Failure Analysis page

In failure analysis, it is important to look for a variety of possible causes, including overload, fatigue, creep, corrosion & erosion. Each of these mechanisms may leave tell-tale characteristics on the fracture surfaces of the failed components that can be seen with a scanning electron microscope. It is also important to examine the microstructure of the component. Two components can have the same chemical composition, but very different properties if the grain size or microstructure are different. Consider the two titanium microstructures below.

Different Heat Treatment Yields Different Microstructures of the Same Material

Both of these have the same chemical composition, but different microstructures & therefore different properties. Microstructure can be affected by cool down rates in the casting process and by the heat treatment process. If either of these processes are not well controlled, the microstructure may not be what it should be. It is also important to look for voids, inclusions or other imperfections which may become crack origination sites.

Damage Mechanisms

Fatigue
	Fatigue is the failure of a component due to repeated loading & unloading. Each loading & unloading is called a cycle. The simplest demonstration of fatigue is take a paper clip, straighten it out, and then bend it back & forth. You'll probaly cause failure fairly quickly. It has been estimated that 90% of engineering failures are due to fatigue. Factors influencing fatigue life include the stress range and the R ratio. The R ratio is defined as the minimum stress divided by the maximum stress. For exmaple, if the stress is cycled between -50 and 50 ksi, the stress range is 100 ksi and the R ratio is -1 (-50/50). Half of the stress range is called the alternating stress, which in this case would be 50 ksi. Two common R ratios are -1, which is often referred to as "Fully reversed" and 0, often called "Zero to Max". Not surprisingly, the greater the stress range, the shorter the fatigue life. If fatigue life is plotted against stress range, we get something like the curve to the left. This is called an S-N curve, since it relates stress range (S) to cycles (N). Typically, the Y axis is linear while the X-axis is logrithmic. Note that the slope of the curve decreases as life increases. Some materials have what is called an "Endurance limit" or "Fatigue limit". In these cases, the curve goes flat, and there is a stress range below which the material will never fail, no matter how many load cycles are applied. S-N curves are also influence by R ratio. The higher the mean stress, the lower the fatigue life. If we plot S-N curves with different R ratios, we get something like this:
Even when designed by top-flight engineers, fatigue failures may come about for a number of reasons: Unrecognized Stress Concentration Factor - Most fatigue cracks begin at stress concentrators. These can be fillets, where two right angle surfaces come together, holes drilled for attachment purposess, or any other discontinuity in the component. A good example of this type of failure is the de Havilland Comet. It was the world's first commercial passenger jet, and a commercial success until a couple were lost due to fatigue that started at the window corners. With a pressurized cabin, the fuselage was stressed every time the plane reached cruising altitude, where the exterior pressure was much lower. The stress concentration at the window corners was greater than anticpated, and fatigue cracks initiated there. Unexpected Loads - Loads that engineers think a component will experience in the field are not always what it actually experiences. If this is the case, no amount of testing will reveal the problem. This could be due to the way the operator uses the component, or it may be due to unforseen forces. An example of this is the failure of the Tacoma Narrows Bridge in 1940. Unexpected aerodynamic loads caused it to sway, earning it the nickname "Galloping Gertie", and eventually causing it to fail. Microscopic defects: Typically, be voids or contaminants. These defects act as stress concentrators and cause small cracks to initiate very quickly, diminishing the overall fatigue life. Incorrect microstructure - Microstructure is very important and often overlooked. Two components with exactly the same alloy can have vastly different mechanical propoerties. This is because of different microstructure, which can be influenced by heat treatment. If a part is cast and cooled very slowly, large grains can form. If it is cooled quickly, there is not enough time to form large grains, so the grains will be small.
Overload
	Overload occurs when a section of a component is so highly loaded that the materials ultimate tensile strength (UTS) is exceeded. The fracture surface exhibits what is commonly referred to as "cup and cone" appearance, as shown in this photograph. This type of failure is not very commom, since it only happens when something goes terribly wrong, such as: A component experiences a load far greater than it be expected to see in normal usage - Think of a house hit by a tornado, or a harbor in the path of a tsunami. An earlier failure significantly changes the loading - Let's say that there are 2 bars that hold a load, and the stress in each is 75 percent of the material's UTS. Everything is fine, but if one bar should break, the second will instantly be overloaded. The final product is not built like it was analyzed. This was a factor in the 1981 Kansas City Hyatt Regency walkway collapse. Changes to the original design were made by people who did not realize the consequences.
Corrosion
	Corrosion is the attack on a component by a chemical reaction with its environment. One of the most common examples is rusting of iron or steel, which is caused by iron chemically reacting with oxygen, and can be significantly increased by the presence of water. Corrosion occurs at the surface, where both the component and its environment come into contact. The simplest way of combatting corrosion is to put a barrier in place to separate the two, such as paint or a coating of zinc. Some metals are more resistant to corrosion than others. Aluminum, like iron, reacts with oxygen to form an oxide. Unlike iron, the oxide layer on aluminum adheres tightly to the surface & inhibits further oxidation, instead of flaking off and allowing corrosion to penetrate deeper into the component. Since corrosion is a chemical process, it attacks the different constituents of the microstructure at different rates. Corrosion damage will show some microstructural features (Those more resistant to corrosion) remaining while those around it are gone. This is shown in the pit on the left of the photo below, where there are some white ridges remaining while the material around it has been removed. Contrast that with the pit on the right which due to erosion, a mechanical process.
Erosion
	Erosion is a mechanical process where small particles are blasted against a component removing material. Rather than selectively removing portions of the microstructure as in corrosion, it removes whatever it hits, leaving a smoother surface, like the one shown to the left. A good example of a component subjected to erosion would be blades on a helicopter operating in the desert. When the chopper comes in to land, the downdraft kicks up a lot of sand particles. The blades, which are rotating rapidly, have many high speed impacts with these particles.

Possible Ways to Increase Life

Once the damage mechanism is determined, we start looking at what we might do to increase life. If the failure was due to fatigue, the following questions need to be asked:

Can the nominal stress can be lowered by simply increasing the component's cross section?
Can the stress concentration factor be reduced by increasing fillet size or shaping holes?
Will stiffeners such as ribs reduce stress?
Would inducing residual stresses by shot-peening or other processes solve the problem?
Will surface treatments such as superfinishing increase life to an acceptable level?

For overloads:

Is the loading on the component greater than what was assumed in the design process?
Does the manufactured part match what was analyzed during design?

For corrosion:

Can a protective coating be applied?
Is a change to corrosion resistant material (such as going to stainless steel instead of regular steel) possible?
Can the operating environment be changed? Perhaps an enclosure can be added to protect the component

For erosion:

Can anti-erosion coatings be applied?
Can wear strips be applied to the most erosion prone areas?

To view a couple videos of spectacular engineering failures, please visit our Links Page.