Failure investigation question 1

[bjpil]

Hello All,

I was talking to a metallurgist last night and he was trying to convince me that there is a process to determine the stress of a part after it has failed such as an airplane crash. He said that an acid and dyes are put on the specimen and then the specimen is examined through a microscope. From this the exact stress level can be determined. Is there any validity to his statements or parts of his statements?

BJP

 

[metengr]

BJP;
I have not heard of such an examination - this is not to say it does not exist. Most of the failure analysis investigations that I have been involved with and exposed to use a combination of scanning electron microscopy (SEM) for examination of the fracture surface features, and metallographic examination to evaluate the microstructure of the metal. Metallographic examination is typically performed by chemically etching a metal sample for viewing under the microscope.

I would be hard pressed to believe that one could determine the exposed stress level of the part other than to say the part failed from "mechanical overload" in service. This implies that the part had been exposed to a stress level that exceeded the tensile strength of the material, based on the appearance of the fracture surface.

 

[alexit]

Should the part have failed due to fatigue stress at levels within the plastic deformation range, you will see a wave-effect to the fracture surface features. If the stress level was in excess of the material tensile strenght, the fracture surface features will be more smooth/linear.

These are the only cases I've seen differentiated in SEM analysis, so you would have a stress range (XtoY, >Y) for the failure but not a specific value.

 

[hacksaw]

there are visual clues that can identify the class of failure and infer that the stresses may have been in excess of yield strength, etc., but none actually quantify the stresses.

studies of serious failures are far more comprehensive than a simple visual or metallographic inspection.

 

[TVP]

Quantitative fractography can provide information on fatigue crack growth and fracture toughness by measuring details like:

a. distance between striations
b. stretched zone width and/or height
c. crack tip opening displacement
d. proportion of different fracture modes over the entire fracture surface
e. average linear density of microvoids in ductile regions
f. average aspect ratio of microvoids
g. crack arrest & retardation due to corrosion products

Some good sources of informaation on this subject are the following:

K. Wolf, "Ejection Seat Quick-Release Fitting: Quantitative Fractography and Estimation of Local Toughness Using the Topography of the Fracture Surface", Practical Failure Analysis Vol 2 (3) June 2002.

Quantitative Methods in Fractography ASTM STP 1085

 

[metengr]

BJP;
You have some good comments here that show just how complex a failure investigation can become, and one of many approaches that a metallurgist can use to aid in determination of root cause.

If you are interested in more information as to what is involved with failure analysis, the ASM Metals Handbook Volume 11, Failure Analysis and Prevention" contains excellent information on this subject.

 

[bjpil]

Metengr (and others),

You are absolutly correct that these above statements are good and I appreciate everyones help on this. I personally could not believe that there was a technique using an acid and some dye to predict what the stress level is on the metal away from any fracture which this person claimed could be done. I just needed a sanity check or if what he claimed was true I just wanted to get more information bceause then I would have another method to check residual stresses.

The best part of this thread was all the good information and directions I can investigate to get a better understanding on failure analysis.

Again, thanks for everyones help.

BJP