Master's Thesis in Structural Engineering

Friction of wood on steel

Authors: Radek Koubek, Karolina Dedicova

Surpervisor LNU: Michael Dorn, Erik Serrano

Examiner, LNU: Johan Vessby

Course Code: 4BY05E

Semester: Spring 2014, 15 credits

Linnaeus University, Faculty of Technology

I
Abstract
This thesis deals with the experimental description of friction between steel
and wood materials, specifically laminated veneer lumber (LVL) and pine
wood with two types of annual rings. It studies the influence of a number of
different parameters on the coefficient of friction such as contact pressure,
moisture content, fiber orientation in relation to the load direction, steel
surface roughness, and horizontal load rate. First, the theoretical mechanical
and physical properties as well as the coefficient of friction itself are
described. This is followed by the description of the test setup including the
test method and how the obtained data is exported, handled and processed
and how the coefficient of friction is determined.

The results study the influence of different parameters and show that the
coefficients of friction for the smooth sliding plate tests vary in between 0.1
and 0.3, whereas tests with the rough sliding plate vary around 0.7.

Factors influencing the coefficient of friction were found to be the different

moisture content under all tested pressures, the different fiber direction
under low contact pressure, the contact pressure itself, though under higher
pressures the influence was found to be low, and the horizontal load rate
under low pressures. The outcomes are further discussed in the discussion
chapter.

Key words: Friction, wood, steel, moisture, coefficient of friction, contact

pressure, LVL, laminated veneer lumber, pine, fiber direction

II
Acknowledgement
We would like to thank the supervisors of our thesis Michael Dorn and Erik
Serrano for providing us with time, guidance and assistance during the
conduction of the experiments and for contributing with valuable
information, comments and advice during analysis of the data and the actual
writing process. Furthermore we would like to thank to the personnel of the
laboratory of Linnaeus University for providing us with the necessary
equipment and machinery.

III
Table of contents
1. INTRODUCTION............................................................................................1
1.1 BACKGROUND ...................................................................................................................... 1
1.2 PURPOSE AND AIM ................................................................................................................ 2
1.3 HYPOTHESIS AND LIMITATIONS ............................................................................................ 2
1.4 RELIABILITY, VALIDITY AND OBJECTIVITY ........................................................................... 3
1.5 LITERATURE REVIEW ............................................................................................................ 4
1.5.1 History of friction determination.................................................................................. 4
1.5.2 Research on friction between wood and steel .............................................................. 5
1.5.3 Wood friction characteristics during exposure to high pressure ................................. 5
2. THEORY .......................................................................................................... 8
2.1 MECHANICAL PROPERTIES OF WOOD..................................................................................... 8
2.1.1 Strength and stiffness of wood...................................................................................... 8
2.1.2 Compression parallel to the fiber direction ................................................................. 8
2.1.3 Compression perpendicular to the fiber direction ....................................................... 9
2.1.4 Compression stresses at an angle to the grain ............................................................. 9
2.1.5 Orthotropic elasticity ................................................................................................. 10
2.1.6 Stress at an angle to the grain - Hankinson's formula ............................................... 11
2.2 PHYSICAL PROPERTIES OF WOOD ........................................................................................ 12
2.2.1 Moisture and wood..................................................................................................... 12
2.2.2 Density ....................................................................................................................... 13
2.2.3 Shrinkage and swelling .............................................................................................. 14
2.3 DESCRIPTION OF FRICTION AND THE COEFFICIENT OF FRICTION .......................................... 14
3. METHOD .......................................................................................................17
3.1 MATERIALS ..................................................................................................................... 17
3.2 SPECIMEN PREPARATION AND PHYSICAL PROPERTIES ...................................... 19
3.3 EXPERIMENTAL PART .................................................................................................. 24
3.3.1 Test procedure............................................................................................................ 20
3.3.2 Methodology of experiments ...................................................................................... 22
3.4 ANALYTICAL PART....................................................................................................... 24
4. RESULTS ....................................................................................................... 25
4.1 OBTAINING THE COEFFICIENT OF FRICTION OF A SINGLE SPECIMEN .................................... 25
4.2 STATISTICAL EVALUATION OF A SINGLE TEST SERIES .......................................................... 28
4.3 STATISTICAL EVALUATION FOR VARYING CONTACT PRESSURE ........................................... 28
5. ANALYSIS .....................................................................................................29
5.1 VARIATION OF CONTACT PRESSURE .................................................................................... 29
5.2 VARIATION OF MOISTURE CONTENT .................................................................................... 31
5.3 VARIATION OF FIBER DIRECTION......................................................................................... 33
5.4 VARIATION OF ROUGHNESS OF THE STEEL SLIDING SURFACE .............................................. 35
5.5 VARIATION OF LOAD RATE .................................................................................................. 36
6. DISCUSSION .................................................................................................38
7. CONCLUSIONS ............................................................................................39
REFERENCES ...................................................................................................40
APPENDICES ....................................................................................................42

IV
1. Introduction
Friction is everywhere where there is contact between two surfaces of
materials. Depending on the amount of friction, it plays an important role in
determining the behavior of the materials in contact. In timber applications,
contact between wood and steel appears nearly in every construction, mostly
in connections.

A commonly used connection is the dowel-type steel-to-timber joint in

structural timber engineering. Joints are often considered as the weaker part
of the structure. Therefore estimating the load-bearing capacity and stiffness
of the connection should be done in an accurate and reliable manner
(Dorn, 2012). Previous researches on the influence of friction between wood
and steel in connections has proved that using different values of the
coefficient of friction can give a significant difference in the final results
(Sjödin et al. 2008; Dorn, 2012).

Nowadays, Eurocode 5 (EC5) is used for designing timber constructions.

EC5 does not include the coefficient of friction as a parameter in the
designing. Friction, of course, exists and should be applied during the
designing process, but the coefficient of friction changes depending on
various parameters. But also the properties of wood widely vary from one
tree to another and from sawn wood to engineered wood products, which
again affect frictional properties.

Whether and how the coefficient of friction changes in dependence on

different parameters can bring very important knowledge for timber design.

1.1 Background
Broadly speaking, the current calculations and literature for timber
constructions provide relatively poor information about the coefficient of
friction and its values are seldom precisely determined. For contact between
wood and a steel surface, coefficients of friction between 0 and 1 (and
above) are commonly found in literature (American Forest & Paper
Association, 1997; Residential structural design guide, 2010), but the values
given are seldom rationalized. In general, frictional resistance to slipping of
connection members is conservatively ignored in design equations, although
in some cases coefficient of friction is taken into account (American Forest
& Paper Association, 1997).

Friction is usually not accounted for in wood connection design because the
amount of frictional force is difficult to predict and in many instances may
not exist, if a wood member shrinks or a connection relaxes. (American
Forest & Paper Association, 1997). The effect of friction between wood and
steel is supposed to be dependent on a variety of parameters such as surface

1
 texture, roughness, wood density, moisture content, applied pressure,
orientation of annual rings in respect to the sliding plane and the direction of
sliding, etc. (Sjödin et al., 2008) show, in their experimental and numerical
study of effect of friction in single dowel joints, how the coefficient of
friction is changing by using different surfaces of dowels. From the work of
(Sjödin et al., 2008) it is obvious, that there, in some cases, can be a
substantial effect of friction that could be taken into account in calculations.
Nevertheless, the current version of Eurocode 5, does not explicitly involve
friction in design as a parameter.

1.2 Purpose and Aim

The purpose of this thesis is laying down a range of coefficients of friction
for different situations involving pine wood and laminated veneer lumber
(LVL) coming into contact with steel. For obtaining results it is also
necessary to design the experiments including the MTS machine used in
order to obtain reliable data and being able to evaluate the coefficients of
friction.

The aim of this work can be split into three sub-goals:

1) Carry out experiments on two wooden materials with different
combinations of selected parameters (moisture content, surface of the
steel sliding plane, applied pressure, orientation of annual rings and load
rate).

2) Calculate the coefficient of friction for different types of wood species

and combinations of parameters.

3) Generate data from the measurements (forces) and evaluate the data.

1.3 Hypothesis and Limitations

Hypotheses
1) The machine setup and the way the specimens are mounted will provide
data that makes it possible to obtain the necessary information regarding
differences between e.g. wood species and different steel surfaces.

2) There is a correlation between all the studied parameters – contact

pressures, fiber directions, moisture contents of pine wood and LVL
materials - and the coefficient of friction measured.

2
Limitations and assumptions

In regard to the hypotheses above, there are several limitations and

assumptions that have to be considered in this work. For instance:

1) There is only one machine setup.

2) A limited amount of tests for each parameter variation can be conducted

3) There is no possibility to measure roughness of the used materials.

4) Some mechanical properties are taken from literature.

1.4 Reliability, validity and objectivity

The experiments have been performed using a MTS Frame machine 322,
which has high reliability in functions provided by a hydraulic system. The
load cells and actuators of the machine are designed for laboratory testing of
materials and every test is controlled by a computer. The setup of the MTS
Frame machine 322 has been done by professional employees of Linnaeus
University laboratory.

During preparation of the experiments, there was an effort to obtain as much

numerical data as possible. Stress was put also on the selection of specimen
and their preparation. All specimens were cut by professional personnel of
Linnaeus University Växjö, measured by appropriate tools and marked and
placed in a climate chamber with constant temperature and moisture content.
All measured values were noted to excel files.

The specimens are made from two materials (pine wood and LVL) and they
have approximately the same dimensions (30 x 30 x 10 mm3). Each
specimen dimensions has been measured before the actual testing using a
digital sliding gauge and all data are noted in an Excel document.

For each combination of parameters used in this work, five repetitions have
usually been made. All data has been recorded by the machine computer and
afterwards evaluated with help of suitable software (Matlab, Excel).

All the above described procedures aimed at minimizing variability and

assuring the methodology used was objective.

3
1.5 Literature review
This section addresses previous research relating to measuring coefficients
of friction and its dependency on various parameters. It includes a short
history of determination and establishment of friction in a general way and
later research about friction between wood and steel.

1.5.1 History of friction determination

Looking back into history, (Jost, P., 1966) found that huge financial losses
were occurring as a result of wear, friction and corrosion. The ancients as far
back as Paleolithic times understood the need to control these forces. Also
drawings from ancient Egypt, 2400 years BC, show that lowering friction by
using grease in transporting heavy statues and building the pyramids
facilitated the works.

However, defining friction as an actual science started 500 years ago.

(Beek, A., 1995) describes that the first deep study of friction was conducted
in the 15th century by Leonardo Da Vinci (1452-1518) who realized that it
is important to take friction into account, while he was designing his
machines (Figure 1.1). He stated two basic laws of friction - “1st the size of
the areas in contact has no effect on friction and 2nd if the load of an object is
doubled, the frictional force will also be doubled”. Also other scientists who
were concerned with the definition of friction, observing its values and its
dependency on various parameters; the most famous are Guillaume
Amontos (1663-1705) whose work was based on the theory of friction as a
result of roughness of two surfaces and he rediscovered the basic laws of
friction of Da Vinci. Charles August Colomb (1736-1806) who expanded
Amontons’ work by stating “strength due to friction is proportional to a
compressive force” so the “Amontons-Coulomb Law” was established
(Beek, A., 1995).

Figure 1.1: Sketches from da Vinci's notebook, ca. 1480 demonstrating

some of his coefficient of friction experiments (Hart, 2011).

4
 In the late 1930s, F. Philip Bowden and David Tabor found that the true area
of contact is formed by asperities and with the increasing normal force the
area of contact increases. They also gave the researched discipline about
friction the name tribophysics, but that term was never generally accepted,
leaving the way clear for H. Peter Jost, author of the eponymous report, to
give the science its name: tribology (Hart, 2011).

Tribology is defined as the science and engineering of surface phenomena

such as friction, wear, lubrication, adhesion, surface fatigue and erosion
(Sinha S.K., 2010). Today, friction is mainly studied in scientific tribology.

1.5.2 Research on friction between wood and steel

In regard to friction between wood and steel, experiments mainly addressing

issues related to the mechanics of cutting wood have been done.

(Murase et al., 1980) conducted extensive evaluation of the influence of

change in various factors such as surface roughness, sliding speed, moisture
content on the friction characteristics during processing (cutting/shaving)
wood with tools during exposure to relatively low pressure (up to 0.1 MPa).

(McKenzie and Karpovich, 1968) worked on determining the more

important variables affecting friction between wood and steel by using a
milling machine for cutting wood; they found that sliding speed as well as
moisture content and roughness of steel have a significant influence on the
value of the coefficient of friction.

(Ning, et al., 1982) studied the friction between Swedish wood and steel by
using the same set-up as (McKenzie and Karpovich, 1968) and they obtained
the same results, in addition they obtained evidence for two basic relations:
a) the softer the wood surface, the higher the coefficient of friction and
b) the higher the surface roughness of steel, the higher the coefficient.

1.5.3 Wood friction characteristics during exposure to high pressure

The friction between wood and steel is becoming a widely discussed

phenomenon, especially due to the increased use of wooden structures with
e.g. dowel type joints, where high pressure occurs. Proof about the fact, that
friction is an important factor that should be taken into account has been
verified for example in (Seki et al., 2012). (Sjödin et al., 2008) showed by
experimental and numerical methods that the value of the coefficient of
friction has a significant effect on the load-bearing capacity of single dowel
connections.

(Seki et al., 2012) published a study on friction of wood and steel during
exposure to high pressure, where contact pressures 1, 5 and 10 MPa were
used. The effect of metal surface (polished and grinded) and effect of wood