Why do some tensile test samples fail before the expected final elongation

Question

We tested several 1020 steel samples during my class' lab session last semester. during the test, one of the samples failed early, breaking at a strain of around 0.3 or 0.2, while the other samples are similar to the stress-strain curve below. the sample that failed early had the same Ultimate tensile stress, although lower uniform elongation (i forgot the value), the curve just curved down faster and failed early like I said.

The TA said the sample looked a little corroded and that might be the reason for it, but I don't understand why it failed early like that. If it is only slightly corroded why doesn't it behave like a slightly smaller sample, in which case shouldn't it give a barely different curve?

Thank you in advance

Answer 1

Every chunk of engineering material ever put onto an Instron tester and pulled to failure contains manufacturing flaws inside it. The smallest of these do not affect the test result, but the biggest ones will. This is why you have to run multiple tests on samples cut from the same batch to find the average and the standard deviation for that specific batch.

If you have a low result from one particular sample, you then perform failure analysis on it to determine root cause. In the case of an Instron tensile test you observe the fracture surfaces under a microscope to find the flaw (a void, a slag inclusion, a pre-existing crack, etc.) that triggered it.

If the sample was corroded, you will look for the presence of corrosion product on the fracture surfaces which will indicate the existence of a pre-existing crevice that was corroding and which reduced the load-bearing cross-section of the sample.

Answer 2

Likely because the corrosion helped accelerate a microscopic fracture and so the sample failed early.

Answer 3

Elastic strain goes up to about 0.2 % , then plastic strain begins and goes to about 25% for a typical 1020 depending on the nature of the rolling. Your stress /strain curve is not correct.

Answer 4

Your TA is likely correct. Elongation can be greatly affected by surface finish in some materials. It's not about the loss of cross-sectional area, it's about stress concentrations and nucleation sites for crack initiation. In our test lab we can sometimes polish a sample up 10 or 15 percentage points of elongation vs an as-machined specimen.