33 B. Neyses and O. Hagman: Identifying suitable wood species for wooden products with multivariate data analysis
holztechnologie 56 (2015) 2 © IHD, Dresden
Introduction
Despite or maybe even due to its long history as a construction
material wood is still very popular. Perhaps it will become
even more popular in the future. Among others, one of the
main reasons for this is the possibility to use the material in
a sustainable way, unlike metal or various types of plastics.
However, in many cases wood is harvested and used irres-
ponsibly. Relentless deforestation without subsequent refo-
restation is the standard in numerous countries around the
world (Laurance, 1999). This is partially caused by the high
demand for certain species with a limited availability. An
example is the guitar industry. Species like mahogany, rose-
wood or ebony were and are still used for guitar bodies, necks
and fretboards, even though they are vulnerable or critically
endangered. Even if endangered species are not used, the
wood choice is often not justiiable from an objective point
of view. In many cases only a few species are considered
for application. Alder for example, is a very popular wood
species for electric guitar bodies. If people are asked why,
some say because Leo Fender – who was the founder of the
Fender Musical Instruments Cooperation – used it already
decades ago. As a consequence, it must be good. Allegedly,
In many cases only few wood species are used or even considered for any given wooden product,
even though there are hundreds of wood species available. The objective of this project was the
development of a time efficient and structured method to identify the most suitable wood species
for wooden products, based on a set of required material properties. This goal was achieved by
applying multivariate data analysis. The method was based on a dataset consisting of commercial-
ly available wood species represented by many different properties. The scores and loadings of the
multivariate data analysis method Principal Component Analysis (PCA) were used to identify the
wood species with the most fitting property combinations for the product in question. Applying the
method to an example case resulted in several plausible alternatives to the commonly used wood
species. It is possible to apply the method to any wooden product by determining the set of
required properties.
Keywords: Wood properties, product development, wood choice
Identifying suitable wood species for wooden pro-
ducts with multivariate data analysis
Benedikt Neyses, Olle Hagman
he started using alder just because it was cheap and available
(Fender, 2014). Similar behaviour most likely prevails regar-
ding many other wooden products. This is contrasted by the
fact that hundreds of wood species from all around the world
are commercially available, many of them not endangered in
any way (Kribs, 1968).
The question is: How to provide people with a tool to identify
the most suitable wood species for any wooden product in
a quantiiable and comprehensive way that is also easy and
fast to apply. The objective of this research project was the
development of a wood identiication tool which is able to
achieve this.
The tool was based on the application of principal components
analysis (PCA), a multivariate data analysis (MVDA) method.
In general, PCA is utilized to extract and visualize useful in-
formation from large datasets, in particular those with many
observations and variables (Eriksson et al., 2006).
Giving people the means to make a profound material choices
in an easy and quick way has the potential to enhance the qua-
lity and variety of wooden products while supporting respon-
sible utilization of timber at the same time.
Review: