Abstract
This study explored the effect of processing parameters on surface roughness as a result of aesthetic designs processed on walnut, chestnut, and beech wood edge-glued panels (EGPs) by CNC (computer numerical control) router. To accomplish this, the average roughness value (Ra) on an engraved surface in a Ying-Yang design treated on the material was measured. Using analysis of variance (ANOVA), the feed rate, spindle speed, step-over, and axial depth of cut factors; surface roughness factors; and the interactions between these factors were found to form significant differences at the level of 95%. At the end of the study, the Ra value was lower in walnut and beech EGPs (3.423 μm and 4.316 μm, respectively) and higher in chestnut EGPs (5.005 μm).
Download PDF
Full Article
The Influence of Process Parameters on the Surface Roughness in Aesthetic Machining of Wooden Edge-Glued Panels (EGPs)
Abdullah Sütçü a,* and Ümmü Karagöz b
This study explored the effect of processing parameters on surface roughness as a result of aesthetic designs processed on walnut, chestnut, and beech wood edge-glued panels (EGPs) by CNC (computer numerical control) router. To accomplish this, the average roughness value (Ra) on an engraved surface in a Ying-Yang design treated on the material was measured. Using analysis of variance (ANOVA), the feed rate, spindle speed, step-over, and axial depth of cut factors; surface roughness factors; and the interactions between these factors were found to form significant differences at the level of 95%. At the end of the study, the Ra value was lower in walnut and beech EGPs (3.423 μm and 4.316 μm, respectively) and higher in chestnut EGPs (5.005 μm).
Keywords: Surface roughness; Aesthetic machining; Wood edge-glued-panel (EGP); End-milling; Beech; Walnut; Chestnut
Contact information: a: Department of Forest Products Engineering, Suleyman Demirel University, Orman Fakultesi, Dogu Yerleskesi, 32260 Isparta, Turkey; b: Department of Forest Products Engineering, Kastamonu University, Orman Fakultesi, Kastamonu, Turkey;
* Corresponding author: abdullahsutcu@sdu.edu.tr
INTRODUCTION
As the furniture industry moves from mass production to customer-oriented production, the design and production of customized products needs to be fast, and the manufacturing lead time needs to be shortened. To respond to these needs and increase profitability under a global marketing environment, firms attach importance to new technologies such as computer-aided production (CAP), computer-aided design and manufacturing (CAD/CAM), and their direct relative, CNC machines, in the workshop. The art of engraving holds an important place in the furniture manufacturing industry and can be applied by hand by valuable craftsman; however, engravings today can be carried out in shorter times as a result of the facilities provided by CAD/CAM technologies. It is important to know the processing parameters of the relevant technologies for engravings to be produced at the desired quality and cost. This study explored the effect of different processing parameters on surface roughness in making engraved wood panels.
Wood is a natural polymeric material with a heterogeneous structure. However, surface irregularities on solid wood surfaces are normally not considered as they are in other materials, such as metals and plastics (Zhong et al. 2013). Sinn et al. (2009) reported that wood surface properties were the result of complex and time-dependent interactions between the material and machining. They supported this finding with an overall review of the characteristics of solid wood surfaces and characterization techniques. With regard to the processing of the wood, there are numerous important factors, such as wood species, anatomical characteristics, moisture content, grain direction, feed rate, spindle speed, cutting depth, and tool geometry (Kopac and Sali 2003). Further improvements in technology, the development of different variants of engineered wood materials, different technological machinery, and different tools have made it necessary to continuously investigate the processing parameters.
When wood species were evaluated in terms of processing properties, hardwood was found to have better processing performance than softwood (Malkoçoğlu and Özdemir 2006). Because the density of latewood is higher, it has a lower surface roughness value than earlywood (Malkoçoğlu 2007).
Iskra and Tanaka (2005) reported that surface roughness had a direct relation to the inclination angle of wood grain and feed rate. Mitchell and Lemaster (2002) reported that processing at grain surfaces and on the flat side relative to the grain provide better surface quality than processing perpendicular, curve, and transversal faces; it was also pointed out that the surface quality in perpendicular, curve, and transversal cut surfaces decreases with increasing feed rate. The surface formation mechanisms that occur when processing with a straight blade and up- and down-milling at various grain angles have been studied. The forces occurring in the cutting process have been reconsidered for application to wood (Goli et al. 2004).
Wooden EGPs are increasingly gaining importance as an alternative substitute to other wood-based products in the furniture manufacturing process. Currently, many furniture firms use EGPs in products such as tables, beds or chests, and doors (Mitchell et al. 2005). The total production capacity of EGPs in Europe is reported to be approximately 2 to 2.5 million m3/year (Dilik et al. 2012).
Wood EGPs are simply created from narrow strips of lumber that are glued together under pressure. Some of the advantages of edge-glued panel production are the relatively low cost of equipment, the smaller diameter and low-value grades of lumber, flexibility in panel product sizes, and opportunities to sell products within established local markets (Nicholls 2010).
Until recently, there were a limited number of studies of the processing of wooden EGPs and the aesthetic machining of wood separately. With regard to the aesthetic machining of wood materials, Negata et al. (2007) studied a robotic sanding system for attractively designed furniture with free-formed surfaces. Nagata et al. (2009) introduced an intelligent machining system based on a three-axis NC machine tool with a rotary unit for producing many kinds of specially designed wooden paint rollers. Also, Fujino et al. (2003) examined the influence of machining conditions such as feed rate and feed direction to the grain using two wood species. With respect to the machining of wooden EGPs, Sutcu (2013) reported that by making routing operations on pine, spruce, and beech EGPs both in the fiber direction and perpendicular to the fibers, the surface roughness with the relevant processing factors were found to be 34% for pine EGPs, 49% for spruce EGPs, and 27% for beech EGPs, respectively. Furthermore, the cutting direction is important for pine EGPs, the cutting depth and feed rate are important for spruce EGPs, and the cutting direction and feed rate are important for beech EGPs. Thus, the issue of aesthetic processing should be investigated further. The aim of this study was to determine how effective some processing parameters (e.g. feed rate, spindle speed, axial depth of cut, and step-over) are on the average roughness value (Ra), an important indicator of product quality, during the course of aesthetic processing of walnut, beech, and chestnut EGPs with a CNC router, to determine the level of effectiveness of the factors and interactions among them.
EXPERIMENTAL
Testing Material
A/B-class walnut, chestnut, and beech EGPs with 18-mm thicknesses were used (see Dilik et al. 2012 for the quality classes and standardization).
It is well known that walnut is a tight-grained and dense hardwood. Because of its interesting grain pattern, black walnut is of significant value for furniture, architectural woodwork, flooring, and decorative panels. Other important uses are gunstocks, cabinets, and interior woodwork (Miller 1999; Zhong et al. 2013). Beech is a dense, pale-colored hardwood. It has a fine structure with tight and large rays. Although beech wood is a hard and strong material, it does not have the endurance level of some other hardwoods (Zhong et al. 2013). Most beech is used for furniture, flooring material, brush blocks, handles, veneers, woodenware, and containers (Miller 1999). Chestnut wood is coarse in texture; annual rings are made conspicuous by several rows of large, distinct pores at the beginning of each year’s growth, and it has rich tannin content. It dries well and is easy to work with tools (Miller 1999). Because of its tannin content, it has a natural resistance to fungi, insects, and parasites. It can easily be polished and demonstrates good adhesion, wear resistance, hardness, strength, dimensional stability, and screw retention (Ay and Şahin 2002). Additionally, chestnut is widely used in underwater construction, ship and boat construction, parquet, and flooring (Gorisek and Strase 2011).
Before processing, the moisture contents of the wood panels were measured in accordance with TS 2471 (2005), and the densities of the wood panels were measured in accordance with TS 2472 (2005). The density and moisture contents for samples of wooden EGPs are summarized in Table 1.
Table 1. Values of Density and Moisture Contents for Samples of Wooden EGPs
Machining Treatments
Materials were processed with a modified Mekano P1500 model CNC router in our laboratory at the Suleyman Demirel University Faculty of Forestry. The table was fixed on the CNC router, and the spindle moved along the X, Y, and Z axes. The bench magazine capacity is limited to one set of coupling. The maximum spindle speed was set to 18,000 rpm. As a cutting tool, a 6-mm-diameter tungsten carbide-solid straight router bit with two blades was used (Fig. 1) (Leitz GmbH & Co. KG, 2012).
Fig. 1. Router cutter used on experiments
The experimental design was established according to the full factorial experimental design method. The main factors in the design of the experiment were feed rate, step-over, axial depth of cut, and spindle speed. Step-over can be defined as the “radial depth of cut” in this study. By carrying out a full factorial design with these factors, 3x3x4x4 = 144 samples were processed for each type of EGP. Factors taken into account and their levels are shown in Table 2. The test sequence was determined completely at random.
Table 2. End-Milling Parameters
The piece geometry and the cutting paths were defined by ArtCAM® (Artistic CAD/CAM) software. As an aesthetic design, a Ying-Yang model was processed, as it is a universal design and is preferred in freeform surface feature studies (Fig. 2) (see Sun et al. 2001).
After the design was created, the processing was carried out by transferring the CAM model to a computer connected to a CNC router (Fig. 3). The tool path used was a spiral tool path. The cutting tool path with the spiral tool processes the created model in an inward or outward circular motion. Sakarya and Goloğlu (2006) determined that a spiral tool path is the ideal tool path in finish processing for pocket milling.
Fig. 2. Aesthetic Ying-Yang design prepared in ArtCAM
Fig. 3. Operand (a) and processed (b) test samples
Surface Roughness Measurements
The first evaluations of surface roughness were made using visual observation and feel. This approach is very effective, but is also highly subjective. Currently, various sophisticated methods, such as the multi-element array diffuse reflection laser displacement sensor (CCD LDS) and camera-based vision systems (Sandak et al. 2004; Sinn et al. 2009), are available. The stylus technique is popular for evaluation of wood surface smoothness and is successfully employed due to its simplicity and its provision of accepted standard numerical values (Kilic et al. 2006; Zhong et al. 2013).
In this study, a Mitutoyo SJ 201 stylus-type surface roughness measuring device was used. This device operates on the inductive principle to measure the surface roughness. The instruments’ measurement head fits a diamond tracer tip (5-µm radius), the measurement range is up to 350 µm, and the measuring force is 4 mN. The surface roughness parameter was measured over a traverse length of 5 mm and cut-off length of 0.8 mm using a Pc50 (Gaussian) filter. The traverse speed was set at 0.5 mm/s. The measuring parameter (Ra) is described in TS971 (1988) (adapted from ISO468-’82). Ra represents the average surface roughness, which is very useful in understanding the quality of wood surfaces (Khazaeian et al. 2004).
The device gave all the relevant parameter results automatically with a computer connection and the correct software standards.
Three different surfaces were formed on the machined material. As shown in the Ying-Yang symbol (see Fig. 3b), milling operations were applied on the downmill side during climb cutting, on the opposite side during conventional cutting, and on the burr surface end during milling. In this study, the burr surface was chosen for the location of measuring surface roughness on the Ying-Yang symbol. The surface profile was engraved and measured along and across the grain on the wood surface. Five measurements were taken from the engraved surface in each test specimen.
Statistical Evaluation
A full factorial experimental design (four factors) provided a complete trial in each replicate of the experiment, and the factors provided all possible combinations of the levels utilized (Montgomery and Runger 2003). These factors are as follows: Factors A, B, C, and D with 4 levels of factor A, 4 levels of factor B, 3 levels of factor C, and 3 levels of factor D; each replicate contains all ab, ac, ad, bc, bd, cd, abc, abd, acd, bcd, and abcd treatment combinations.
SPSS statistical software was utilized for the evaluation of the experimental results and the effect of dependent variables on surface roughness (Ra). The interactions between these factors were used for univariate analysis.
Prior to analysis, all of the measured data were tested for normality using the Kolmogorov–Smirnov normality test, and the requirement that the error variance of the dependent variable should be equal in groups was checked by the Levene test and approved (Levene test p > 0.05).
For the determination of different groups after the variance analysis (ANOVA), the Duncan test, frequently used in similar studies, was used. The results of the multiple comparison tests conducted at 5% significance level were expressed in the form of a letter. The difference between the groups with the same letter was statistically insignificant.
RESULTS AND DISCUSSION
The results of the univariate analysis (UNIANOVA) of the surface roughness measurements for the wood EGPs processed by different machining conditionings in the CNC router are displayed in Table 3. The three wood EGP interactions between factors of interest and all factors were significant (P <5%). There was a significant effect of the processing parameters on the Ra value (Table 3).
For this calculation, the change in the single and multiple parameter interactions’ average roughness value (Ra) is proportional to the R2 value. As shown in the ANOVA table, the effects of part of these changes, e.g., walnut EGPs: 71.9%, beech EGPs: 62.2%, and chestnut EGPs: 88.4%, on the change in the value of average roughness can be explained by main effects and interactions (see Table 3).
The effect of progress parameters on homogeneous materials processes like metals can be explained by the higher R2 values (see Kirby et al. 2004). However, there are numerous factors that affect the machining of wood, including wood density, wood porosity, moisture content, extractives, grain figure, kinematics of the cutting process, and machine conditions (Kopac and Sali 2003; Sin et al. 2009). It is difficult to provide higher R2 values because of the difficulty of forming a model with consideration of all the interactions in the model.
The heterogeneous structure is increasing by virtue of how the wooden EGP structure of the solid material is processed, as well as tangential or radial irregular laths placed side by side or end to end in the same plate (Sütçü 2013).
Interactions and levels of individual factor effects can be elucidated easily by the partial eta-squared value. The partial eta-squared (η2p) describes the proportion of variability associated with an effect when the variability associated with all other effects identified in the analysis has been removed from consideration (Fig. 4). It is commonly used to report the effect size estimate for ANOVA. Then, one can easily calculate η2p from the ANOVA output using a statistical package like SPSS (Fritz et al. 2012). Cohen (1992) proposed a conversion table for η2, where 0.0099 constitutes a small effect, 0.0588 is a medium effect, and 0.1379 is a large effect.
Table 3. The ANOVA Table for Walnut, Beech, and Chestnut EGPs with Respect to Average Roughness (Ra)
