Tony Baker and Andrew Pelcin

November 14, 1998
Modified September 30, 1999

Over the past nine (9) months, I have attempted to replicate controlled, experimentally produced flakes with a type of engineering software call Finite Element Analysis (FEA). I have done this to understand the mechanics of the formation of flakes. At this time the research is far from finished but it is far enough along that I want to share some preliminary results. I would like to begin with an analogy to give the reader a view of the evolution of this work.

The Analogy -- Many years ago when I was a rookie engineer in the oil fields, I was transferred to a location where my responsibilities were associated with the drilling of wells. This was a totally new experience since I never had been exposed to drilling in the classroom, nor had I ever worked around a drilling rig. I had studied civil engineering, so drilling was totally new to me.

I remember during those early months standing on the derrick floor, observing the drilling operation and talking to the Driller (person in charge of the operation). I was trying to understand and learn the mechanics of the drilling process. I would ask questions similar to: "Does the bit drill faster if you rotate it faster? Does the bit drill faster if more weight (push) is placed on it?"

Always the response to these type questions was: "No, it drills best at this speed and/or weight. If it would drill better under different conditions I would be using them."

I would ask the Driller how they knew this and he would say: "After years of working on a drilling rig and being a Driller, one just knows. It takes experience."

As a young engineer this was very frustrating. I was from a world where mechanical things could be explained by physical models and mathematical terms. One just had to read a book or attend a class and the mathematics of the process could be understood. However, drilling seemed to be different. It appeared that one might have to spend many years on a drilling rig before the process could be understood and then the understanding might be more Zen than mechanics. Drilling appeared to be black magic.

After six (6) months of struggling to understand drilling, I was asked to attend a drilling school taught by a professor of petroleum engineering. After two weeks of all-day classes, I had learned drilling was not black magic. There were physical models and mathematical equations that explained the process. The difficulty I had been having for the last six (6) months was I was being taught by individuals who didn't know the mechanics themselves. They were drilling each well the same way they had drilled the previous well. They were doing what they knew would work. Their knowledge was truly founded on experience.

I offered this analogy because it parallels my learning process of flake formation. As I started to become interested in flake formation, I found that most archaeologists had what they believed to be a rudimentary understanding of flint knapping. When I would ask a technical question, they would direct me to The Formation of Flakes by Cotterell and Kamminga (1987). I would attempt to read this, but it was so complex I would get nothing out of it. Plus, I could find no one who could answer my questions about the article. It was opaque to me.

Next I found myself turning to the flint knappers. These were people doing the real thing. But like the drillers in my analogy they soon convinced me that flint knapping was black magic. They convinced me that if I was ever going to understand the subject, I was going to have to become a good flint knapper.

In 1997 I heard a paper by Pelcin at the SAA conference in Nashville.1 His paper was just a quick glimpse of his Ph.D. research which was Controlled Experiments in the Production of Flake Attributes ( 1996). But it was enough of a glimpse that I thought that maybe he could help me understand this black magic process of flint knapping. He provided me with a copy of his dissertation and later all his data in electronic form. We became good friends and colleagues.

In the fall of 1997, it occurred to me that I might be able to replicate Pelcin's controlled experiments with FEA software. I reviewed the available packages on the market and finally settle on a package offered by Algor. In February 1998, I purchased portions of the entire product and began attempting to replicate Pelcin's work.

The replication work with the software was a slow and tedious process as I attempted to match Pelcin's real world flakes. Throughout this time I was constantly conferring with Dennis Gray, a Texaco engineer and my running partner for the last 15+ years. As we would run along at lunch, many times stopping to draw on the sidewalk with a rock, we would discuss my most recent efforts. I can remember more than once Dennis would say something similar to: "well, I understand what you are saying, but I don't think that is the solution." Dennis is a wonderful sounding board and I would still be mired in my unsuccessful efforts if it was not for his insight.

After eight (8) months of intense replication work (remember, I work full time for Texaco) I believed I understood enough about flake formation to write this status report. In the early days of writing this report I attended a five (5) day, knapping school (October 1998) taught by Tim Dillard at Kampsville, Illinois. Since I am not a knapper, I did this to see if my understanding of flake formation would be consistent with the real world. I am excited to report that it is.

I finally completed the rough draft, uploaded it to the internet, and started inviting other people to read it. One of these individuals was Bill Watts, another Texaco engineer. Well, Bill found a major mistake in my potential energy equation. I had omitted multiplying my numbers by 0.5. This mistake didn't alter my findings, but it did cause me to re-do several graphs and other numbers. Bill saved me from publishing the errors and ultimately being beat up by an internet "smart ass."

Finally, I want to thank Dr. Bruce Bradley. Bruce was not directly involved in this FEA simulation process, but his seeds are definitely in the work. I don't believe my interest in flake formation would have lasted if I had not had Bruce to discuss this subject with over the years.

Proceed to:

Return to Paleoindian & Other... Home Page


1 I had actually met Andy Pelcin several years earlier at the SAA in St. Louis. As a graduate student he presented a poster paper on his experimental work of flake formation. I saw this paper and was attracted to his engineering approach to the problem of flake formation. However, I had no interest in the subject at the time and so we exchanged a few emails and our association died until the Nashville SAAs.