**2. Background**

#### **2.1 Surface texture**

Surface texture, a three-dimensional measurement, has been described as the topography, roughness, or irregularity of the interface between a substance and its surroundings,

The Use of the Wavelet Transform to Extract

related causes.

dimensional variations, etc.

of surface roughness.

Additional Information on Surface Quality from Optical Profilometers 101

quantitative assessment of surface quality difficult. Wood materials exhibit a wide range of defects due to biological as well as machining-related causes. In some cases there is no clear distinction between biological causes of poor surface quality as opposed to machining

Monitoring the surface quality of a workpiece surface is a good indicator of the state of the machining process regardless of the workpiece material. It is common practice in wood product industries for lumber graders to check the quality of the surface visually, for composite panel manufacturers to use crayons to check for undesirable sanding marks, for planer operators to "feel" the depth and spacing of planer knife marks, and for saw operators to visually check the severity of saw marks. While these procedures are often used to attempt to determine if a process has varied with time, they are very inconsistent from day to day and do not permit the quantification of the defects. Monitoring the surface quality of a machining process is becoming increasingly important as the machining speed, the cost of raw material, and labor, all continue to increase. Any undetected changes in the quality of machining process can cause a significant impact on the economics of the process. Workpiece quality evaluation during the actual wood machining process (on–line surface evaluation) has been done using cameras, lasers, x-ray, and various combinations of these technologies. These systems are able to provide a relatively rapid scan of the wood material, usually while the sample is moving slowly (or temporarily stopped on a conveyor) prior to or after being sorted or machined. Such systems are in common use in industry and are aimed primarily at detecting biological defects such as rot, discoloration, knots, etc. These types of systems have also been used to detect simple geometry problems, such as gross

The work that this chapter is based on consisted of using a laser based position sensing device (PSD) to obtain a 2 dimensional surface profile of the surface. The signal processing techniques that are discussed is an attempt to extract more information as to the type and cause of the surface irregularity than simple measuring the magnitude of the irregularity as is normally done based on the U.S. (ASME B461-2009) and international (ISO 4287/1) standards. The utility of simple frequency analysis is demonstrated below, for several idealized (simulated)

As mentioned above, all examples of surfaces analyzed in this chapter were from wood or wood based products. It is beyond the scope of this chapter to go into detail about wood structure. If interested, the reader is referred to "Understanding Wood" by Hoadley, 2000. The surface texture that is generated when machining wood is very complex and has many factors that can contribute to the variations of the surface quality. Surface defects can be either biological or machining based defects. The fact that wood is an anisotropic and hygroscopic

Peripheral milling or planing (moulding) may be defined as the removal of wood in the form of single chips by intermittent engagement of the workpiece with knives carried on the periphery of a rotating cutterhead (Koch, 1955). The resulting surface on the workpiece of a peripheral milling operation consists of individual knife traces generated by successive engagements of each knife or cutting edge (Figure 1). In addition to the height of the ridges or scallops (t), the distance between successive ridges or **pitch (Sz)** is also an import feature

material can cause the surfaces generated by a machining process to vary greatly.

examples of surface quality issues relevant to wood machining.

generally air (Stumbo, 1963). Surface roughness and surface topography are properties of engineering materials that are important to functional performance and can be used as a measure of product quality and process performance. Surface texture can be caused by the nature of the material itself, a manufacturing process applied to the material, or a combination of both. The processing characteristics that affect the surface texture include: inaccuracy in the machine tool, deformation under cutting force, tool or workpiece vibration, geometry of the cutting action, material tearing during chip formation, and heat treatment effects. Wood characteristics that can affect surface texture include: wood species, density, moisture content, and cutting direction. In most instances, however, surface finish has not been fully exploited in the areas of process monitoring, quality and performance prediction. Today, new measurement techniques and signal processing methods make it feasible to take a new look at the ways available for measuring and evaluating surface texture.

The degree of roughness of a surface often affects the way the material itself is used. In general, surface irregularities can cause misalignment and part malfunctions, excessive loading over small areas, friction and lubrication problems, general finish and reflectivity problems, as well as catastrophic failures. Although surface quality for wood products has been a key issue since woodworking first began, the level of precision required does not approach that found in the metal working industry. This has been due, in part, to wood's inherent dimensional instabilities. The other main reason was that many common uses for wood did not require exceptional surface finishes as compared to many metal applications. The monitoring of surface irregularities in wood is, however, important to assure proper fit of machined parts for gluing, acceptable surface finish for furniture, and as a methodology to monitor the accuracy of the manufacturing process. The last reason has become even more important in recent years due to the increased cost of raw materials, the increased production costs, and the higher production speeds available. Any deviation in expected product quality can quickly cause significant economic losses. There has also been a trend toward tighter tolerances for many forest products industries. An example of this would be the lamination of wood or wood-based products with plastic films or ultra-thin veneers. Even the slightest irregularity in the surface will show through the top laminate.

Usually, wood machining processes are heavily influenced by workpiece surface quality considerations. Tool sharpness requirements as well as machine feed and speed decisions are often based on workpiece surface quality. Research in surface measurement technology was aimed at identifying and quantifying defects associated with a variety of machining processes. Surface waviness is often introduced by the machining process or by the vibration of the tool or workpiece, whereas surface roughness is often introduced by the detachment of material from the workpiece. Of particular interest in this research was the use of frequency domain analysis to separate the random from the periodic components of the surface. The optical profilometer surface measurement system discussed in this chapter has been found to be effective for identifying surface defects including surface waviness, torn grain, fuzzy grain, and abrasive (sanding) grit marks.

Though beyond the scope of this chapter, methods of assessment have ranged from entirely subjective methods (simply feeling the wood surface) to modern day computerized threedimensional (3D scans) assessments of the surface. The very nature of wood has made the

generally air (Stumbo, 1963). Surface roughness and surface topography are properties of engineering materials that are important to functional performance and can be used as a measure of product quality and process performance. Surface texture can be caused by the nature of the material itself, a manufacturing process applied to the material, or a combination of both. The processing characteristics that affect the surface texture include: inaccuracy in the machine tool, deformation under cutting force, tool or workpiece vibration, geometry of the cutting action, material tearing during chip formation, and heat treatment effects. Wood characteristics that can affect surface texture include: wood species, density, moisture content, and cutting direction. In most instances, however, surface finish has not been fully exploited in the areas of process monitoring, quality and performance prediction. Today, new measurement techniques and signal processing methods make it feasible to take a new look at the ways available for measuring and evaluating surface

The degree of roughness of a surface often affects the way the material itself is used. In general, surface irregularities can cause misalignment and part malfunctions, excessive loading over small areas, friction and lubrication problems, general finish and reflectivity problems, as well as catastrophic failures. Although surface quality for wood products has been a key issue since woodworking first began, the level of precision required does not approach that found in the metal working industry. This has been due, in part, to wood's inherent dimensional instabilities. The other main reason was that many common uses for wood did not require exceptional surface finishes as compared to many metal applications. The monitoring of surface irregularities in wood is, however, important to assure proper fit of machined parts for gluing, acceptable surface finish for furniture, and as a methodology to monitor the accuracy of the manufacturing process. The last reason has become even more important in recent years due to the increased cost of raw materials, the increased production costs, and the higher production speeds available. Any deviation in expected product quality can quickly cause significant economic losses. There has also been a trend toward tighter tolerances for many forest products industries. An example of this would be the lamination of wood or wood-based products with plastic films or ultra-thin veneers. Even the slightest irregularity in the surface will show through

Usually, wood machining processes are heavily influenced by workpiece surface quality considerations. Tool sharpness requirements as well as machine feed and speed decisions are often based on workpiece surface quality. Research in surface measurement technology was aimed at identifying and quantifying defects associated with a variety of machining processes. Surface waviness is often introduced by the machining process or by the vibration of the tool or workpiece, whereas surface roughness is often introduced by the detachment of material from the workpiece. Of particular interest in this research was the use of frequency domain analysis to separate the random from the periodic components of the surface. The optical profilometer surface measurement system discussed in this chapter has been found to be effective for identifying surface defects including surface waviness,

Though beyond the scope of this chapter, methods of assessment have ranged from entirely subjective methods (simply feeling the wood surface) to modern day computerized threedimensional (3D scans) assessments of the surface. The very nature of wood has made the

torn grain, fuzzy grain, and abrasive (sanding) grit marks.

texture.

the top laminate.

quantitative assessment of surface quality difficult. Wood materials exhibit a wide range of defects due to biological as well as machining-related causes. In some cases there is no clear distinction between biological causes of poor surface quality as opposed to machining related causes.

Monitoring the surface quality of a workpiece surface is a good indicator of the state of the machining process regardless of the workpiece material. It is common practice in wood product industries for lumber graders to check the quality of the surface visually, for composite panel manufacturers to use crayons to check for undesirable sanding marks, for planer operators to "feel" the depth and spacing of planer knife marks, and for saw operators to visually check the severity of saw marks. While these procedures are often used to attempt to determine if a process has varied with time, they are very inconsistent from day to day and do not permit the quantification of the defects. Monitoring the surface quality of a machining process is becoming increasingly important as the machining speed, the cost of raw material, and labor, all continue to increase. Any undetected changes in the quality of machining process can cause a significant impact on the economics of the process.

Workpiece quality evaluation during the actual wood machining process (on–line surface evaluation) has been done using cameras, lasers, x-ray, and various combinations of these technologies. These systems are able to provide a relatively rapid scan of the wood material, usually while the sample is moving slowly (or temporarily stopped on a conveyor) prior to or after being sorted or machined. Such systems are in common use in industry and are aimed primarily at detecting biological defects such as rot, discoloration, knots, etc. These types of systems have also been used to detect simple geometry problems, such as gross dimensional variations, etc.

The work that this chapter is based on consisted of using a laser based position sensing device (PSD) to obtain a 2 dimensional surface profile of the surface. The signal processing techniques that are discussed is an attempt to extract more information as to the type and cause of the surface irregularity than simple measuring the magnitude of the irregularity as is normally done based on the U.S. (ASME B461-2009) and international (ISO 4287/1) standards. The utility of simple frequency analysis is demonstrated below, for several idealized (simulated) examples of surface quality issues relevant to wood machining.

As mentioned above, all examples of surfaces analyzed in this chapter were from wood or wood based products. It is beyond the scope of this chapter to go into detail about wood structure. If interested, the reader is referred to "Understanding Wood" by Hoadley, 2000. The surface texture that is generated when machining wood is very complex and has many factors that can contribute to the variations of the surface quality. Surface defects can be either biological or machining based defects. The fact that wood is an anisotropic and hygroscopic material can cause the surfaces generated by a machining process to vary greatly.

Peripheral milling or planing (moulding) may be defined as the removal of wood in the form of single chips by intermittent engagement of the workpiece with knives carried on the periphery of a rotating cutterhead (Koch, 1955). The resulting surface on the workpiece of a peripheral milling operation consists of individual knife traces generated by successive engagements of each knife or cutting edge (Figure 1). In addition to the height of the ridges or scallops (t), the distance between successive ridges or **pitch (Sz)** is also an import feature of surface roughness.

The Use of the Wavelet Transform to Extract

and DeVries and Lemaster, 1992).

Fig. 2. Schematic of optical profilometer

Additional Information on Surface Quality from Optical Profilometers 103

Vast amounts of work have been conducted in attempts to develop techniques to measure and evaluate surface texture in materials. These techniques generally fall into two distinct groups. The first is the hardware or method to measure surface texture data. The second is the analysis procedure to evaluate the surface texture. Numerous methods have been developed and researched for both the measurement and evaluation techniques. Measurement techniques normally fall into two distinct categories: contact and non-contact methods. It is beyond the scope of this chapter to discuss the surface measurement techniques that have been investigated in the past. The reader is referred to Lemaster (2004) for an overview of the various works on this topic. A general review of the optical techniques (and surface roughness techniques in general) is provided in several comprehensive reference works (Thomas, 1999; Whitehouse, 2011; Whitehouse, 1994; Thomas and King, 1977; and Riegel, 1993). The work conducted by the author on optical profilometry of wood and wood-based products can also be found in the literature (Lemaster, 2010, Lemaster, 2004; Lemaster 1997a, 1997b; Lemaster and Beall, 1996; Lemaster and DeVries, 1992; Lemaster and Dornfeld, 1983; Jouaneh, Lemaster, and Dornfeld, 1987;

The heart of any surface quality assessment system is the detector. The optical method used for the detector in this research is a variation of the reflectance method, whereby the positional change of the reflected laser light into the detector is correlated to changes in the test surface height. In this method, a laser spot is projected on the workpiece surface and the reflected light is focused on the surface of a lateral-effect photodiode. The change of the position of the reflected laser spot on the surface of the detector, a' is correlated to the vertical height change of the workpiece, a. By moving a workpiece beneath the detector and recording the change in the position of the laser spot, a two-dimensional surface profile is obtained that is very similar to that obtained by the traditional stylus system (Figure 2). The resulting surface profile can then be analyzed according to traditional U.S. (ASME B461- 2009) and international standards (ISO 4287/1). This method is non-contact and capable of detection at high speed, and since it measures position changes of the reflected light and not

a

Laser

spot intensity, it is relatively insensitive to color changes of the workpiece.

a'

Wood Surface

Detector

**2.2 Conventional surface quality measurement and analysis techniques** 

Fig. 1. Definition of pitch and depth of cutter or tool marks (Weinig USA, training manual, www.weinigusa.com).

As the pitch increases the surface appears more "wavy" for a given cutter diameter and depth of cut. Many manufacturers specify the accepted or desirable pitch of a surface while others may specify the "knife marks per inch". The smaller the pitch the "smoother" the resulting surface will be, however, this is sometimes at the expense of quicker tool dulling. Experience (Effner, 1992) has shown that a good surface finish will have a pitch mark of approximately 1.5 – 1.7 mm (0.06 – 0.07 inches). For knife marks per inch this translates to 15-17 marks per inch for a high quality surface. Many moulder manufacturers recommend that the peak-to-valley height of the marks be kept below 0.005 mm (0.02 in) for fine furniture and between 0.005 and 0.017 mm (0.02 – 0.07 in.) for average quality building moulding.

Another type of machining of interest is abrasive machining. Abrasive machining includes **abrasive planing** the workpiece to a desired thickness or **sanding** a workpiece to achieve the desired level of smoothness. The surfaces that are generated from this type of machining process is complex in that they often include non-periodic abrasive grit marks running parallel to the feed direction (wide belt sanding) as well as regular periodic "tooling" marks running perpendicular to the feed direction. These "tooling" marks are caused by either the motion of the sanding head, the motion of the workpiece, or a combination of both.

In addition to the surface texture variation that may be caused by machining processes there are other surface defects that are caused by the manufacturing process of wood-based composites. A condition, called **pitting** is where wood fiber or fiber bundles are pulled out of the surface of the wood panel product during panel manufacturing. This can be caused by improper press times, resin content or blending, or the lack of release agents on the platens of the press.

Fig. 1. Definition of pitch and depth of cutter or tool marks (Weinig USA, training manual,

As the pitch increases the surface appears more "wavy" for a given cutter diameter and depth of cut. Many manufacturers specify the accepted or desirable pitch of a surface while others may specify the "knife marks per inch". The smaller the pitch the "smoother" the resulting surface will be, however, this is sometimes at the expense of quicker tool dulling. Experience (Effner, 1992) has shown that a good surface finish will have a pitch mark of approximately 1.5 – 1.7 mm (0.06 – 0.07 inches). For knife marks per inch this translates to 15-17 marks per inch for a high quality surface. Many moulder manufacturers recommend that the peak-to-valley height of the marks be kept below 0.005 mm (0.02 in) for fine furniture and between 0.005 and 0.017 mm (0.02 – 0.07 in.) for average quality building

Another type of machining of interest is abrasive machining. Abrasive machining includes **abrasive planing** the workpiece to a desired thickness or **sanding** a workpiece to achieve the desired level of smoothness. The surfaces that are generated from this type of machining process is complex in that they often include non-periodic abrasive grit marks running parallel to the feed direction (wide belt sanding) as well as regular periodic "tooling" marks running perpendicular to the feed direction. These "tooling" marks are caused by either the

In addition to the surface texture variation that may be caused by machining processes there are other surface defects that are caused by the manufacturing process of wood-based composites. A condition, called **pitting** is where wood fiber or fiber bundles are pulled out of the surface of the wood panel product during panel manufacturing. This can be caused by improper press times, resin content or blending, or the lack of release agents on the

motion of the sanding head, the motion of the workpiece, or a combination of both.

www.weinigusa.com).

moulding.

platens of the press.
