OREGON STATE UNIVERSITY

Bonding Wood and Family Trees

by Bryan Bernart 

Professor John Nairn studies bonds that hold wood composites together—while also creating apps that help genealogy enthusiasts find family connections and shared histories.

For anyone who’s met the multi-talented professor in the Department of Wood Science & Engineering, these seemingly divergent research interests are no surprise. Already well known for his work in composites research, Nairn arrived at Oregon State University in 2006 as the Richardson Chair in Wood Science and Forest Products, following a long stint at the University of Utah. During his years living in Utah—“a hotbed of genealogy research”—and working as a professor of materials science and engineering, Nairn developed a keen interest in genealogy himself. Nearly 20 years ago, he embarked on a project to write his own genealogy software to fill a previously unrealized niche for a kind of user-customizable genealogy data reader and editor. In so doing, Nairn drew on his experience with computer modeling. Now, his materials modeling software and his genealogy software share some of the same code. Although the fields of genealogy and composites science are entirely unrelated, Nairn’s work demonstrates that some computing tools are extremely valuable across disciplines.

Mild mannered and quietly funny, Nairn replies to the question, “how do you talk to a complete lay person about materials science and computer modeling?” with a laugh and an admission: “Not too easily.” 

At its core, though, it’s not incredibly difficult to understand—especially with Nairn’s guidance. As he explains, much of his work is, in essence, about running experiments virtually instead of in a lab. “Right now we’re looking at wood adhesive bonds in a collaborative project with Professor Fred Kamke,” he says. “Fred’s students are studying the x-ray structure of wood adhesive bonds down to the cellular level—this is, of course, the study of the bond between pieces of wood separated by an adhesive. What we do is put the wood in an x-ray machine, pull on it, and observe how it deforms under load, which is a means of gathering data about how a given adhesive performs.” 

While running lab-based experiments is one useful way of obtaining data about wood adhesives, it could be more efficient, and allow for a greater range of experimentation, if the process could be virtualized. So, how does one create a model of an experiment—the data from which is accurate enough to be applicable to the real world? 

“Two important concepts in computer modeling are verification and validation,” Nairn explains. “To say a model is verified is to say you’re written the software correctly, and, depending on what you think you’ve implemented, it’s giving you the right answer. A model is validated when the data it produces is the same as data gathered in an actual experiment with the same parameters.” 

The process of validating a model requires a huge number of calculations, however. “You have to run a lot of them, and so what I’ve done in my software is write a high-level scripting language that allows you to run simulations over and over again,” he explains. “This is critical, because without automation, you’d be running 50,000 calculations one at a time, and of course, you’d soon give up!” 

In Nairn’s genealogy program, scripting fulfills a completely different task. “In most genealogy software, a user has limited control over how the data is displayed. The programmer, and not the user, has decided, for instance, that a page will contain a person’s name, his wife’s name, and his parents’ names. The user must then fill those things in,” he says. “Well, what if you didn’t want to see all those things? What if you want to see just a part of that data, or something else entirely? You’re out of luck.” 

But Nairn’s program utilizes scripting in order to give users control over what data they see. “In this instance, when you write a script, the script will read your data and display it in any way you’ve told it to. This is much like how two different websites can each read the same data, but show it to you two different ways, based on their different code.” 

He is now working to get his genealogy software, GEDitCOM, on mobile devices. “I want to give users the chance to share their family trees with anyone —this would enable more people to participate and also take an interest in their shared histories.”