What is the difference between following a recipe or a procedure and using a methodology?

A recipe implies that there is a fixed set of steps that anyone can follow. There are few things that are so simple in life that a recipe could be all inclusive of everything you need to know. In order to bake a cake, for example, the recipe writer assumes, in most cases, that you know how to grease and flour the pan. They assume that you know which pan to grease and flour. They assume that you know how to measure flour, either using a cup (volume) or a scale (weight). They assume you know how to “cream butter.” But once you have these basic skills, you can usually reliably bake a cake assuming your oven is flat and the temperature is calibrated and assuming you know how to read the test stick that the baker is usually instructed to insert into the cake after a certain time to “see if it is done.”

A procedure is usually more detailed than a recipe. There might be a procedure, following the cake recipe example above, to grease and flour the baking pan. There might be a procedure to cream the butter, which details how to tell when it has been softened enough and how to add the sugar bit by bit rather than all at once. So there might be 10 procedures needed by someone who has never baked a cake or watched anyone else bake a cake that need to be learned before the cake is actually successfully baked by the recipe instructions.

A methodology is yet another variety of “instruction set.” A methodology may be thought of as a mental toolkit based on a structured knowledge foundation.

Let’s start by examining the widely used 13-step recipe for failure analysis in the popular book by Don Wulpi, Understanding How Components Fail. It starts with collection of background information and ends with analysis, formulation of conclusions and writing the report. This “procedure” (a recipe by my definition above) is taken from the Metals Handbook Volume 10 (of the edition current in 1985), and it fits on half of a printed page.

I was surprised to see that Wulpi admonished the reader in the paragraph leading up to the list of steps that the most important first step in any failure analysis investigation is to do NOTHING (his emphasis!) except careful visual inspection, thinking and asking detailed questions. I was pleasantly surprised to see this paragraph. The problem is that most people turn first to the list of steps and don’t review the paragraph preceding it. So many people are left without the knowledge being reinforced about the importance of doing NOTHING first.

Furthermore, the admonishment to do NOTHING is followed by the word “except.” To me, the first step truly must be to do nothing. Maybe even close your eyes for a few minutes to clear your visual field. Maybe count backwards from 10, 20 or higher if you are patient! This will help you to see what is present more clearly, without being primed to see something specific based on whatever you were thinking about previously.     

The first step in the Handbook article procedure is collection of background data and selection of samples. What is the best way to select samples? For routine industrial work involving components rather than, for example, a bridge collapse, selection of “control” or “reference” samples is a science and an art form unto itself.

The second step in the Handbook article is visual examination and record keeping. Future posts will describe key aspects of doing a visual examination, and there are also some posts in the archives of this blog on this subject.

Using a methodology requires more than knowledge of the steps to take during (i.e., the stages of) the investigation. It requires an extensive knowledge of the object being investigated, its use, its design methodology, the methods of manufacture, the raw materials used and perhaps those passed over.

Details of any inspection techniques used prior to placing the object/assembly into service may also be relevant to some investigations. Effective use of a methodology in failure analysis requires a lot of knowledge – both broad and deep – in addition to knowledge of how to do the right steps in the right order.

This is meant to be encouraging, not discouraging. The point is that no matter how long you have been doing failure analysis work, you have more to learn. But as I occasionally have the chance to look back on the work I did early in my career, I have not often been horrified about big mistakes. That was because I was fairly meticulous to follow the procedure that I knew, which meant going from less to more destructive and to Document! Document! Document! That documentation is a lot easier now than it was in 1986, when I started doing failure analysis before the days of digital photography.