October 2013 Volume 3, Issue 1
Measuring the Haptic Characteristics of Various Wood Finishes
Ryan Gott1* and Chris Byrne2
Student1, Teacher2: The Gatton Academy, Bowling Green, Kentucky
*Corresponding author: email@example.com
In the modern age of touch-screen technology, there has been an increased interest in the friction characteristics of skin. Using a custom-built friction device, the frictional properties of skin on different wood species and wood finishes were tested. The wood samples included Eastern White Pine, North American Cherry, and Birdseye Maple. The finishes included shellac, lacquer, and polyurethane. The friction coefficients ranged from kinetic values of 0.22 to 5.42 and static values of 0.46 to 4.80. Raw wood had the lowest friction coefficients, but it is more common to apply a finish in most applications. The results showed that the satin finishes had lower friction than the gloss finishes on each type of wood. On average, a satin finish has 300% less friction than a gloss finish. The increased and very high friction from skin on gloss finish suggests an adhesion mechanism is a strong factor in the friction system. The data also suggests that moisture in a hydrated finger increases friction in both satin and gloss finishes. This research is the first to quantitatively define the tactile feel of wood surfaces. This information could useful for those using and making products from wood, such as musical instruments, furniture, and tools.
The science of sensing and manipulation through touch, referred to as haptics, is a multidisciplinary science and includes such fields as biomechanics, psychophysics, neuroscience, as well as traditional fields of electrical and mechanical engineering. Recent developments in the technological aspect of the haptics world, such as touch screens, have spurred an interest in the friction behavior of human skin on different surfaces. Because of this interest, new methods have been developed to test friction on different surfaces that have led to an enhanced understanding of the haptic response.
Although the study of friction, known as tribology, has existed for hundreds of years, there has been a recent wave of interest in the scientific community. Modern knowledge of friction comes from Leonardo da Vinci’s extensive notes about friction from the 15th century, which were supplemented by Amonton in 1699. Da Vinci’s work laid the foundation for all friction studies, and Amontons’ recognized the difference between kinetic and static friction. Many modern science and engineering academic disciplines still adhere to the concise, simple relationships developed in those early years.
Today friction investigations are important to a variety of research and engineering developments. At first glance characterizing friction appears a straightforward task. In practice, the opposite is true. Many factors influence friction and make the comparison of test results from one lab to another challenging. Moreover, in the same lab, test results can show considerable variation despite attempts at controlling each factor. Nevertheless, the need for determining friction coefficients and wear resistance has led to the creation of some test standards. For example, the American Society for Testing and Materials (ASTM) has a compilation of standard ways to test different materials1. These materials range from hip replacements to industrial equipment. The ASTM G 115-93 standard provides some guidance in best practices for gaining reliable, useful data from friction experiments. However, none include testing the friction of skin or wood finishes.
It is important to note that friction is not an attribute of a material, but two materials that are in contact with the tendency to slide against each other. Factors such as surface finish, surface contaminates, atmosphere, temperature, sliding speed and contact pressure might all contribute to the friction of the material pair. As a result, many consider friction a system response and take great care to consider all relevant system variables while testing. The concept of static and kinetic friction coefficients are so important they are often incorporated in mechanical design. When depending on these it is often important to measure them for the system at hand, as opposed to relying on published values. In areas of design requiring natural materials such as skin, no applicable database exists to draw from. In addition, little is understood about friction systems consisting of skin, a material with viscoelastic properties highly dependent upon hydration and surface history.
The study of friction on skin was first extensively explored by Naylor in 19552. Naylor studied friction blistering, and the variation in friction between skin and different surfaces was not studied in great extent until Comaish and Bottoms experiments in 19713. Recently, Derler (2009) explored friction of human skin on different glass materials4. Atack and Tabor (1958) studied the frictional effects on wood and showed that friction on wood was very similar to friction on metal5, but their studies were limited to balsam wood on steel and did not include skin surfaces. They succeeded, however, in demonstrating that the cause of friction on these surfaces is adhesion and deformation in the area of contact. A study done in 1977 by Highley, Coomey, DenBeste, and Wolfram showed that different coatings and moisturizers on skin can affect these frictional properties6, validating the concepts of skin properties that are fundamental to this research. That study showed that moisture in a finger could increase friction. These studies have been driven by a desire to further the understanding of the haptic response on different materials. However, there have been no extensive studies on the frictional effects of skin against different finishes on woods. The published data regarding optimized, systematic test methods is incomplete. The most similar method to that used in this study was done by Latash, Savescu, and Zatsiorsky.7 They used a 3D force plate to measure the change in force of a fingertip as it was swiped across different surfaces. That study is one of very few that matches the flat surface on fingertip experiments done in this research. The methods used in the present research differ in that both the speed and normal load of the swipe were controlled. This research attempts to fill these gaps in literature.
Other elements were further studied in this research. Pasumarty, Johnson, Walker, and Adams8 explored the effects of moisture on the friction of the human finger pad. They showed that even on dry surfaces, a moist finger could cause high friction coefficients. Sivamani and Maibach9 define “normal” coefficients of friction for human skin on any surface to be between 0.12 and 3.25, while anything outside of this range would be caused by other factors like moisture.
In this study, some traditional finishes were studied due to their availability and frequent use by wood finishers. These finishes are being used in many applications, one of which is fine woodworking. Wood finishers have a plethora of finishes from which to choose, including traditional oils and waxes as well as newer products like water-based polyurethanes. The finishes are chosen for durability, appearance, and tactile feel. It has often been noted that a wood masterpiece invites the viewer to touch it, and that action has a significant impact on perceived quality. Wood-based products like furniture, baseball bats, and musical instruments can also be more attractive because of their tactile quality. Finishes can change the wood by either decreasing friction or increasing friction. Different finishes provide different textures (roughness) and surface chemistry both of which can influence adhesion and friction phenomena. As a representation of how surface chemistry influences friction one can consider a study by Zhang and Mak10. In that work, it was shown how cosmetic products smooth out skin by depositing a set amount of ingredients that change adhesion and friction properties. Since different wood finish formulations result in different surface chemistries one can hypothesize they would exhibit different friction characteristics.
In this research, different wood finishes were evaluated by testing the friction behavior of a finger passed across the surface. This was done using a custom designed test device that is also being used to evaluate touchscreen materials. By using this very sensitive device, it is possible to identify differences between wood finishes and provide a greater understanding of this aspect of each product’s quality. Quantitative friction data will be of value to those making fine woodwork pieces, individuals using wooden tools, and those who manufacture and supply finishing products.
Materials and Methods
Using a custom-built device, the friction of a human finger on twenty-four different wood samples was tested. Three different wood species and seven different finishes were used, in addition to a sample of each wood with no finish. The woods were Eastern White Pine (Pinus Strobus), North American Cherry (Prunus Serotina), and Birdseye Maple (Acer Suchrum). The finishes were Zinsser® shellac (product number 00408), Deft® lacquer (both clear gloss and clear satin), and Minwax® polyurethane (both clear gloss and clear satin). The shellac was applied three different ways. The first was sprayed from the aerosol can. The second was a spray-on coat dulled by 0000 steel wool. The third was a spray-on coat dulled by 0000 steel wool and completed with paste wax.
The wood samples were all prepared by hand. Each piece was cut to the intended size of 4” x 4” x 0.25”. Surface sanding with a random-orbital sander began with 80-grit aluminum oxide sandpaper followed by 150-grit and 220-grit. Then each wood sample was sanded by hand with a rubber block using 240-grit paper and stroked with the grain. Each piece was then coated with a layer of Zinsser® SealCoat (product number 00854) applied by a paintbrush, then sanded with 220-grit Silicon Carbide sandpaper. The samples were subsequently coated by another layer of SealCoat, this time applied by rubbing on with cheesecloth. After another round of sanding with 320-grit, the respective finishes were applied to each piece by following the manufacturers’ instructions.
After all the wood samples and finishes were completed, each was evaluated using a Mitutoyo SJ-201P stylus profilometer. Surface roughness parameters Ra and Rq were determined using a 1-inch pass across the surface. Each sample was tested three times, in different locations, and the results were averaged. Attempts to find differences between passes with the grain and passes against the grain gave inconclusive results. This suggested little anisotropy, or lay, was present on the finished surface despite the grain of the underlying wood.
The friction testing was conducted using a custom-built system. The system utilizes a National Instruments® compactRIO and associated modules for data acquisition and control. Force is measured with an AMTI HE 6x6 3D force plate with resolution of 0.01N in each of 3 primary directions (x, y, and z) used in testing normal and frictional forces during finger swiping. The hand is placed on a movable plate such that the finger can be pressed upon a test surface and be moved in a predetermined and controlled manner as indicated in Figure 1.
The plate, and thus finger motion, is controlled by a 3-axis translation stage capable of speeds from 0.1mm/s to 200mm/s. A Labview software virtual instrument is used to control motion and select data sampling rates during a friction test. During a test, the plate is moved towards the test surface until the desired normal load is achieved. Lateral motion is then conducted while force data is collected for the swiping action.
A series of tests were performed under highly controlled conditions. Each wood finish sample was tested three times using swipes 50 mm in length, 20 mm apart, and a velocity of 50 mm/s. The data sets discussed in this research were found by averaging the three swipes for each sample. The fingers used in each experiment were washed thoroughly with soap, rinsed with copious amounts of water, and then rinsed with denatured alcohol. The finger was then dried in a clean paper towel. Each sample tested was completed within two minutes of cleaning to ensure consistency. Another small set of data was taken to compare a moist finger to a dry finger. A finger was cleaned and then allowed to dry for three minutes. Three swipes were taken on each sample. Tests were conducted in a room with temperature and relative humidity ranging from 68oF to 72oF and 40% to 65%, respectively.
During friction testing, the z direction force, or normal load, was targeted to be 1N. By dividing the friction force (Newton) in the x direction by the normal load (Newton), the coefficients of friction (µ) were found. These values were graphed versus the changing position of the finger. Most graphs displayed an initial peak at the onset of the finger swipe. This peak value is the coefficient of static friction (µs). The average of the coefficients from 10 mm to 40 mm of the swipe produced the coefficient of kinetic friction (µk). For swipes with µs values lower than the µk value, there was no initial peak. The µs was then found at the point where motion of the finger relative to the test surface began. These approaches are outlined in Figure 2.
Figure 2. Typical test results of coefficient of friction during swipe. On top, the third swipe on a sample of pine treated with a satin lacquer gave an initial peak which shows a static coefficient, µs, of 0.78. On bottom, the second swipe on a sample of cherry with shellac and wax has a µs that was found where the motion starts, in this case at a value of 0.75. Averaging values between 10mm to 40mm yields a kinetic coefficient, µk.
Summaries of the measured friction and roughness values are provided in Table 1. It was found that the unfinished wood had the lowest friction of all samples. The unfinished maple gave the lowest µk and µs of 0.22 and 0.46 respectively. In great contrast, the finished wood samples had much higher values, although the values varied for each wood species and finish. With the lacquer, steel wool, wax, and polyurethane gloss finishes, the maple samples had the highest average coefficient of kinetic friction, getting as high as 4.81 with the lacquer gloss. However, with the raw and polyurethane satin finishes, the maple samples had the lowest average µk, getting as low as 0.514 with the polyurethane satin. In most experiments, the pine samples had the lowest average friction coefficients, getting as low as 0.445 with the lacquer satin. The pine sample with shellac had the highest average µk at 5.42. Figures 3 and 4 show a direct comparison of all friction values.
Table 1. Friction coefficient and roughness results for each sample type. The data shows the average of three swipes for each sample.
Figure 3. The average µk values for each sample is shown.
Figure 4. The average µs values for each sample is shown.
There are also significant differences between the finishes. With both the static and kinetic coefficients of friction, the satin finishes had less friction than the gloss finishes. The samples finished with shellac produced high friction values as well.
With the exception of the maple samples, the finish with the least amount of friction was the lacquer satin with a µk of 0.445, followed very closely by the polyurethane satin with 0.514. With the maple samples, the polyurethane satin had the lower µk at 0.514, and the lacquer satin was second at 0.841. The finish with the most friction for the cherry samples was the polyurethane gloss at 2.66. For the maple samples, it was the lacquer gloss at 4.81. For the pine samples, it was the regular shellac at 5.42. This shows that all three of these finishes caused high amounts of friction.
Marring the shellac with steel wool did not affect the cherry samples, and neither did adding a layer of wax. Additionally, the maple seemed to be affected only by the wax, while the pine was affected by both methods.
The profilometry data showed only small differences. The raw samples were not tested to avoid damaging the stylus of the profilometer. The shellac and wax samples did not give reliable data due to the state of their surfaces. Out of the samples that were tested, the steel wool was the roughest. The polyurethane gloss was the smoothest. In the cases of the lacquer gloss and shellac with steel wool, the cherry sample was rougher perhaps due to inconsistent application of sandpaper and steel wool, respectively.
One set of data was also taken with a dry finger. The cherry wood with lacquer finish (both gloss and satin) gave a different perspective. Three swipes were done on each of the samples, giving an average µk of .508 for satin and .928 for gloss. This shows that moisture in a finger increases friction in gloss by 238% and friction in satin by 31%. A representative comparison is shown in Figure 5.
Figure 5. A comparison of the data for a wet finger versus a dry finger is shown. Both data sets are from the second swipe on a sample of cherry treated with a lacquer gloss. It is clear that the dry finger led to data that was more consistent, while the µk values for the wet finger varied more. Also, the µk values for the dry samples were significantly lower than those of the wet finger.
The data showed that the unfinished wood had the least amount of friction. However, wood finish is a necessity to prevent wood from swelling and cracking. The finishes with the least amount of friction were lacquer satin and polyurethane satin. These finishes would be best used when grip is not important. The wood sample with the least kinetic friction was the pine with the lacquer satin finish with a µk of 0.445. The finishes with the most kinetic friction were Shellac, polyurethane gloss, and lacquer gloss. The wood sample with the highest average kinetic friction was the pine with the shellac with a µk of 5.42.
For static friction, the results were similar. The main difference was that the sample with the lowest µs was the pine with polyurethane satin at 0.712. The sample with the highest µs was still maple with the lacquer gloss with a µs of 4.80. This could be great for products like baseball bats and instruments that need a firm grip. High static friction means that it takes a large amount of force to move against a surface, which leads to a decreased tendency to slide.
In comparison, the profilometry data gave some interesting insight. The rougher surfaces were the mostly satins, while the glosses tended to have a lower roughness. This was interesting because the satins had less friction than the glosses. Friction is caused by adhesion and deformation of a surface. If a surface is rough, it has asperities on the surface that can cause friction. However, in this case, the gloss had lower roughness and more friction. The cause of this could be due to the lack of roughness itself. The other factor in friction, adhesion, is how one surface sticks to another. With less severe asperities, the surface is flatter. This means that the area of contact is increased. This causes the gloss to have higher adhesion than the satin, causing more friction.
Also, the samples with the lacquer gloss have significantly higher roughness values than the polyurethane gloss, with average roughness values of 0.401 µm and 0.154 µm, respectively. However, they have very close coefficients of friction. This may be due to chemistry influences, with the molecules of the polyurethane gloss having stronger adhesion to the skin than those of the lacquer.
The majority of the tests gave coefficients within the normal range defined by Sivamani and Maibach9. However, the tests above the 3.25 limit could have been affected by moisture in the finger. When the finger was washed for each set of tests, the skin remained hydrated during swipes resulting in higher friction.
Another interesting phenomenon occurred with the repetitive testing. The finger used in testing was washed after each sample was tested, or roughly every two minutes. Because three swipes occurred on each sample, there was time in between each swipe for the finger to dry out. A study done by Adams, Briscoe, and Johnson10 showed that moist skin has more friction than dry skin. This is reinforced in our work by the decline in friction with each subsequent swipe. Each sample was tested with three swipes. The overall average of µk on the first swipe was 2.05. On the second swipe, this decreased to 1.68, and on the third swipe, the average was 1.35. This suggests a significant decrease in friction as the finger dried out.
The dry test results showed that the dry finger had results that were more repeatable. The data with the wet finger varied as much as 42% from the average value, while each swipe from the dry finger varied less the 15% from the average. This difference may have been caused by an occasional increase in adhesion due to the moisture in the hydrated finger. The moisture in the finger could have caused the finger to stick to some of the wood surface in some areas more than others. In some cases, this would cause abnormally high friction values. Also, the average friction values for a dry finger were significantly lower than those values for a hydrated finger. This is consistent with previous studies.
This data is the first quantitative classification of the tactile feel of wood. The feel of wood has long been a factor in the perception of quality of a wood species. However, all measurements of the feel of wood in the past have been subjective. This research shows the numerical comparisons of wood finishes. For example, on average, a satin finish has 300% less friction than a gloss finish. However, this conclusion was reached with a limited data set. Conducting more swipes could strengthen the data set and clarify this argument.
Manufacturers of wood finishes like Zinnser and Deft could use these results for developing and marketing new products. Because the tactile feel can impart a sense of quality in wood, the lower friction finishes such as satins would be more appealing. Friction studies have been done for marketable reasons before. After noticing an increase in drops in rugby matches, Thomlinson decided to study effective gripping techniques and materials for rugby balls12. Her studies led to a change in the sport. With this study, the results could lead to a change in the market, both in the way products are marketed and in how consumers search for the product. Wood finishes could be marketed toward specific applications, with satin finishes being used for a better tactile feel and gloss finishes being used for a better grip.
These results show that there is a tremendous difference in friction between wood finishes. This addresses the gap in literature by quantitatively describing the tactile feel of a wood surface and finish. These consumer-friendly results will be of use not only in the fine woodworking world, but also to anyone purchasing furniture or finishing wood. Also, finishes with a higher µs, such as lacquer gloss and shellac, will now be known for having a better grip. One example of how this could be used is for cellists. Cellists must be able to slide up and down the neck of their instrument with ease, but they must also have some grip. This is true for many string instruments. The finish they use could affect their ability to play.
Future work includes more tests to complete the body of work and reinforce the validity of the data. The data is currently limited by the fact that each value is an average of only three swipes. If more each sample was tested further, the data would be statistically stronger. A target aspect to focus on is the difference between dry and hydrated fingers. The limited data set gave promising results. Further tests could confirm that they have different frictional properties or provide evidence that there is a negligible difference caused by hydration. Other factors may be tested as well, such as the direction of the swipe (with the grain vs. against the grain), the time between each hand wash, the speed of each swipe, and different normal loads. Changing normal loads and the amount of moisture in the finger have the potential to make a significant difference in frictional properties. Also, recent environmental concerns have led to new kinds of finishes and paints with low volatile compound emissions. As those products are being developed, friction testing could be used to enhance product quality for a particular application.
1. Friction and Wear Testing, ASTM Standards and ASTM Handbooks. (1997)
2. Naylor P F D. (1955). The skin surface and friction. Br J Dermatol 67, 239-248.
3. Comaish S, Bottoms E. (1971). The skin and friction: deviations from Amontons’ Laws and the effects of hydration and lubrication. Br J Dermatol 84, 37-43.
4. Derler S, Gerhardt L-C, Lenz A, Bertaux E, Hadad M. (2009). Friction of human skin against smooth and rough glass as a function of the contact pressure. Tribology International 42, 1565-1574.
5. Atack D, Tabor D. (1958). The friction of wood. Royal Society of London 246, 539-555
6. Highley D, Coomey M, DenBeste M, Wolfram L. (1977). Frictional properties of skin. The Journal of Investigative Dermatology 69, 303-305, 1977.
7. Savescu A V, Latash M L, Zatsiorsky V M. (2008). A technique to determine friction at the finger tips. J Appl Biomech 24, 43-50
8. Pasumarty S M, Johnson S A, Watson S A, Adams M J. (2011). Friction of the Human Finger Pad: Influence of Moisture, Occlusion and Velocity. Tribology Letters 56.
9. Sivamani R K, Maibach H I. (2006). Tribology of skin. Proc. IMechE Part J: J. Eng. Tribol. 220, 729–737
10. Zhang M, Mak A F T. (1999). In vivo friction properties of human skin. Prosthetics and Orthotics International 23, 135-141
11. Adams M J, Briscoe B J, and Johnson S A. (2007). Friction and lubrication of human skin. Tribology Letters 26, 239-45.
12. Thomlinson S E. (2009). Understanding the friction between human fingers and contacting surfaces. 1-8.
The author wishes to acknowledge support from the WKU Department of Engineering as well as the Gatton Academy. There was also a considerable amount of guidance from Ron Rizzo and Dr. Doug Harper. The work was made possible by through the efforts of a WKU senior project conducted by Michael Goatley, Seth Blankenship, Travis Poulton, and Onur Demir. The sponsorship of Korry Electronics and Mr. Timothy Robinson allowed all of this to happen.