CTOD Size Matters
By Paul Zelenak on Friday, October 28th, 2011
In the past several months there has been a significant increase of crack tip opening displacement (CTOD) testing with requirement to use samples sizes that resemble the actual thickness of the finished product. The size of samples that I am referring to are generally have a W dimension of 4 inches or greater which have begun to be commonplace at EWI. It seems that in the past smaller sample sizes were tested and those COTD results were used to represent thicker finished dimension products. This was done in part because of the lack of testing equipment available for testing of large toughness samples. There has been considerable distress when the larger CTOD samples produce results that are lower than experienced with smaller samples.
Here are a couple of reasons why the larger samples are producing lower results. First, the larger samples put more material in the plane-strain region of the sample. This means that as the sample is loaded the material at the center of the fatigue precrack is stressed and unable to neck down because it is restrained by the surrounding material. Typically on samples with a significant amount of ductility the sides of the sample are drawn in similar to the necking down of a tensile sample. On larger samples there is more material between the edge and the center which increases the quantity of material that is not allowed to neck down.
Secondly, as more material is tested there is a greater probability that you will find the weakest link in the material. Since fracture initiation takes more energy that fracture propagation, a small area that is near the tip of the fatigue precrack can fracture and the sample can fail. In a larger test sample more material is being sampled.
As with all testing, closer simulation of conditions to which the final product will be exposed produces more accurate predictions of behavior. In the case of large samples CTOD testing use of samples sizes that resemble the thickness of the finished product is proving to be important in the prediction of behavior.
Click this link to view the steps in testing large CTOD samples – CTOD Testing Process
Paul Zelenak – Applications Engineer – firstname.lastname@example.org