FFS Forum: Using Fracture Mechanics to Lower Risk and Improve Plant Reliability - Part 1, Introduction to Fracture Mechanics

Editor's Note: This article is the first in a 3-part series on the application of fracture mechanics in a plant environment. This first article is an introduction to the subject and will focus mostly on the basic principles.

What is Fracture Mechanics?

Fracture mechanics is a branch of engineering that describes the behavior of cracks. Specifically, how they grow and when they fracture.

It is a relatively new area with its origins in the early 1900s. In 1913, C.E. Inglis published the seminal work in the field. In 1920, A.A. Griffith published another important paper including studies of scratches on steel wires and failure of glass rods. But research on fracture was very slow until WWII when massive fractures of several “Liberty ships” stimulated research interest. This was followed in the 1950s by the catastrophic failure of three de Havilland Comet aircraft, resulting in dozens of deaths, eventually traced to cracking at the square corners of the aircraft’s windows.

These conspicuous fracture-mode failures resulted in more aggressive research, and progress, in fracture mechanics. Nonetheless, it wasn’t until the 1980’s when the importance of fracture-mode failures became understood in the chemical and refining industry. Toughness requirements were not introduced into the ASME Boiler & Pressure Vessel Code (BPVC) until 1987.

Fracture of cracks is particularly dangerous because it can happen without warning under normal operating conditions.

Today, in the fitness-for-service (FFS) world, Part 9 of API 579-1 / ASME FFS-1 (API 579) deals explicitly with the behavior of crack-like flaws (i.e., fracture mechanics).

The Basics

What is a crack?

Most fundamentally, a crack is a flaw that has a very small (near zero) crack tip radius on each end. Typically, a crack has length and width, but virtually no depth. It’s thought of as a 2-dimensional defect. Cracks are sometimes referred to as “linear defects” because on an X-ray, they just look like a line. Other flaws, like slag inclusions are 3-dimensional. They are sometimes referred to as “volumetric defects.”

A “crack-like flaw” is a defect that may not, strictly speaking, be a crack, but is treated like a crack for the purposes of fitness-for-service (FFS) assessment. Flaws may be treated as crack-like for numerous reasons, such as: they have relatively small tip radii, NDE methods show them to be relatively long and slender, or perhaps even if they seem principally volumetric in nature, they have been subject to cyclic loading and may have small cracks emanating from their tips. In any event, they are defects that are going to be treated as a crack for the purpose of FFS assessment. API 579 includes a helpful discussion on this point in paragraph 9.1.4, and I think it’s worth quoting directly.