Experimental Veriﬁcation of a Highly Simpliﬁed, Preliminary Machinability Test for Wood-Based Boards in the Case of Drilling

: In contrast to metalworking there are no standardized or (at least) generally accepted, relative machinability tests for innovative or less known wood-based panels. The most reliable testing procedures are based on the use of a specialized, accurate system for measuring cutting forces and on conducting all tests in conditions that are similar to real industrial conditions (machine tool, cutting parameters etc.). However, the need for a more simpliﬁed testing procedure has often been voiced— not all scientists specializing in wood-based materials development have a machine tool comparable to one that can be found in a real furniture factory and piezoelectric force sensors at their unlimited disposal. To meet this need, the highly simpliﬁed, preliminary machinability test for wood-based boards in the case of drilling was developed and tried. The results of experimental research suggest that the simpliﬁed way of testing of relative machinability of wood-based boards (i.e., testing based on the photoelectric measurement of the time needed to make a 10 mm deep hole under constant feed force) can be a useful substitute of standard machinability testing procedure (based on accurate cutting forces measurements carried out in the standard industrial conditions). When verifying the simpliﬁed testing procedure, samples from each of the three basic groups of wood-based materials of substantially different internal structures (ﬁberboard, particleboard, and veneer boards) were tested. The relationship between signiﬁcantly reliable and highly simpliﬁed machinability indexes turned out to be at a satisfactory level (R 2 = 0.97 for particleboards and R 2 = 0.95 for ﬁberboards or boards made of veneer or solid wood). The use of a simpliﬁed procedure can be especially pragmatic in case of any preliminary testing of innovative wood-based boards during the material development work.


Introduction
Wood-based boards are becoming more and more widely used in many areas of industrial production [1][2][3][4]. Their growing popularity has been determined by their numerous advantages over natural wood [5]. Wood-based materials are much cheaper, more homogeneous and isotropic, more resistant to fungi and insects, etc. [2]. Moreover, they allow the manufacturing of flat surfaces of any size-which is impossible with solid wood. Therefore, the fact that the ability of modern wood-based boards can be tailored to specific uses, together with their strength properties and affordability, makes them a viable solution to reducing the need for solid wood [5][6][7][8][9].
For all these aforementioned reasons, wood-based boards are more and more commonly machined with a variety of cutting tools-with a particular emphasis on drills. It is worth noting that the resistance to axial withdrawal of screw-mounted after drilling the adequate hole is one of the fundamental and standardized [10] properties of wood-based panels. Therefore, the machinability testing procedures for wood-based boards in the case of drilling seem to currently be the essential problem [11,12]. Only based on the (strictly defined and experimentally determined) relative machinability indexes can the machinability of different materials can be compared-not only in a qualitative (descriptive), but also in a quantitative (numerical) way. This kind of comparison, made for different types of wood-based boards, seems to be very important from the point of view of their designers, producers, and potential users, especially when an innovative wood-based material is to be developed [5,[7][8][9]13,14]. Simple, fast, and inexpensive testing methods are necessary to characterize their machinability properties.
The issue presented above becomes a key problem when one tries to design (or buy) a new (or not well-known to them) board and wants to know what its machinability is compared to some standard wood-based materials with much better-known characteristics. The key question is-which machinability test should be used (by the inventor or the potential consumer) to compare new, prototypical products to well known, presently used boards. In contrast to metalworking, there are no standardized or (at least) generally accepted, relative machinability tests for wood-based panels [15]. To make matters worse, there are not even many scientific publications on this topic. Most of the previous scientific publications on the machinability of wood-based materials focused on selected commonly used materials such as medium-density fiberboards (MDF) and/or standard particle boards, e.g., [16][17][18].
There are relatively few studies comparing the machinability of various materials with different properties and structures, e.g., [19,20]. In this situation, the manufacturers of wood-based materials only usually report on their mechanical properties. As a rule, there is no clear (numerically expressed) information on the machinability of these materials, whereas the machinability aspect of any construction or decorative material is an important factor that may affect the different manufacturing phases, including product design and planning of manufacturing processes.
One of the most reliable testing procedures (which can be used for wood-based boards in the case of drilling) has been suggested and tried by Podziewski et al. [12]. The procedure is based on the use of a specialized, accurate system for measuring cutting forces and on conducting all tests in conditions that are similar to real industrial conditions (machine tool, cutting parameters etc.). Therefore, this procedure has been generally well received by scientists known to us who work in the field of innovative wood-based materials. However, the need of a more simplified testing procedure has often been expressed-not all scientists specializing in wood-based materials development have a machine tool comparable to one that can be found in a real furniture factory and piezoelectric force sensors at their unlimited disposal. It turned out that many of them need an alternative, simplified, yet effective enough procedure for their preliminary ("internal") tests.
The study presented in this paper is focused on the idea of a highly simplified, preliminary machinability test for wood-based boards in the case of drilling, which comes down to a constant feed force machinability testing. This idea of experimental procedure has a decades-long tradition [21,22]. On the other hand, this concept is still valid and used [23][24][25].

Materials and Methods
When verifying the suggested simplified testing procedure, the samples were taken from three basic groups of wood-based materials (with substantially different internal structures-fiberboards, particleboards, and veneer boards). The first type of wood-based boards (fiberboards) was represented by the five following materials: raw and melamine faced medium-density fiberboards (MDF), raw and lacquered high-density fiberboard (HDF), and raw hard fiberboard. The second type was represented by four materials: raw three-layer particleboards P4 (load-bearing boards for use in dry conditions) and P5 (load-bearing boards for use in humid conditions) (symbols according to standard EN 312-2010 [26]), melamine faced particleboard P3 (non-load-bearing boards for use in humid conditions) (also according to EN 312-2010), and raw oriented strand board (OSB). The third type was also represented by the five following materials: compreg, raw transformer plywood, raw plywood, melamine faced plywood, and veneer faced blockboard. All samples came from wood-based boards produced on a mass scale and targeted on the European Union market. Moreover, all samples came from the same sheets of the fourteen boards that were previously used by Podziewski et al. [12] (the detailed characterization of the samples is shown in Table 1). Therefore the relative machinability indexes based on the simplified test can be directly compared with the earlier, significantly reliable results [12]. Table 1. Detailed information about the tested wood-based boards [12]. The schemes of experimental setups for both (significantly reliable and highly simplified) tests are shown in Figures 1 and 2 (relatively) for the general comparison. The fundamental difference is as follows. In significantly reliable tests the instantaneous cutting forces (feed force and drilling torque generated for seven different, constant, and predetermined feed rates: fz = 0.1-0.7 mm) were measured. In this experiment, the standard industrial CNC router (Busellatto Jet 130- Figure 3) was used. In this case, feed forces naturally varied depending on the properties of the processed material. Whereas, during the simplified test, only simple photoelectric measurement of time needed to make a 10 mm deep hole under three different constant feed forces was used. The tests were carried out for three different weights (W = 2.5 kg; 2.75 kg; 3 kg). In this case, the feed rates naturally varied depending on the properties of the processed material. The easiest way to carry out drilling with the constant feed force is by using a slight modified standard bench drill. Therefore, the typical drill (Bernardo BF 30, Linz, Austria- Figure 4) has been adapted. The special rope pulley and handle for a weight were used. The whole mechanism was assembled in a way that allowed the rope to move freely on the pulleys, moving the weight in the vertical direction ( Figure 2). In Figure 2, the enlarged drawing in the circle illustrates the principle of operation of the slotted optical switch, which was intended for the measurement of the amount of time needed to make a 10 mm deep hole under constant feed forces.    The scheme of the test stand used during the earlier, significantly reliable machinability testing [12].  The scheme of the test stand used during the earlier, significantly reliable machinability testing [12].

Figure 2.
The scheme of the test stand used during the highly simplified machinability testing. Figure 2. The scheme of the test stand used during the highly simplified machinability testing.   In both studies, drilling was carried out with a brand new single polycrystalline diamond cutting edge Leitz drill bit with a diameter of 10 mm (Leitz ID 091193). However, the holes were drilled with different spindle speeds (6000 rpm for significantly reliable test and 2000 rpm for simplified test).
The time needed to make a 10 mm-deep hole under constant feed force was measured using a standard slotted optical switch. The time mentioned is the amount of time during which the light beam is being interrupted by the 10 mm wide nontransparent obstacle, which was attached to the drill spindle housing and moved down with it. The principle of operation of the slotted optical switch is shown on the enlarged part of the drawing shown in Figure 2. The general view of the slotted optical switch and the obstacle is shown in Figure 5.   In both studies, drilling was carried out with a brand new single polycrystalline diamond cutting edge Leitz drill bit with a diameter of 10 mm (Leitz ID 091193). However, the holes were drilled with different spindle speeds (6000 rpm for significantly reliable test and 2000 rpm for simplified test).
The time needed to make a 10 mm-deep hole under constant feed force was measured using a standard slotted optical switch. The time mentioned is the amount of time during which the light beam is being interrupted by the 10 mm wide nontransparent obstacle, which was attached to the drill spindle housing and moved down with it. The principle of operation of the slotted optical switch is shown on the enlarged part of the drawing shown in Figure 2. The general view of the slotted optical switch and the obstacle is shown in Figure 5. In both studies, drilling was carried out with a brand new single polycrystalline diamond cutting edge Leitz drill bit with a diameter of 10 mm (Leitz ID 091193). However, the holes were drilled with different spindle speeds (6000 rpm for significantly reliable test and 2000 rpm for simplified test).
The time needed to make a 10 mm-deep hole under constant feed force was measured using a standard slotted optical switch. The time mentioned is the amount of time during which the light beam is being interrupted by the 10 mm wide nontransparent obstacle, which was attached to the drill spindle housing and moved down with it. The principle of operation of the slotted optical switch is shown on the enlarged part of the drawing shown in Figure 2. The general view of the slotted optical switch and the obstacle is shown in Figure 5. During the significantly reliable test, the cutting force problem index (CFPI) was defined and calculated for each material according to the following equation [12]: where FX and TX denote the mean feeding force and mean torque for the X material, and FMDF and TMDF describe the mean feeding force and torque of the reference material, for which the raw MDF was taken. In each material, 15 holes for each of the 7 feed rates were made.
During the highly simplified test, an alternative index called the cutting time problem index (CTPI) was defined and calculated for each material according to the following equation: where tX and tMDF denote the mean time needed to make a 10 mm deep hole in the X material and in the raw MDF, respectively (in each material, 20 holes for each of the 3 weights were performed).

Results and Discussion
The results of the measurements of the time needed to make a 10 mm deep hole (in short-the drilling time) for all tested materials are presented in Figure 6. In any case, the drilling time was reduced when a heavier weight was used. For most materials, the averaged drilling time was less than 0.5 s. A longer time was observed only for two materials (compreg and transformer plywood). During the significantly reliable test, the cutting force problem index (CFPI) was defined and calculated for each material according to the following equation [12]: where F X and T X denote the mean feeding force and mean torque for the X material, and F MDF and T MDF describe the mean feeding force and torque of the reference material, for which the raw MDF was taken. In each material, 15 holes for each of the 7 feed rates were made.
During the highly simplified test, an alternative index called the cutting time problem index (CTPI) was defined and calculated for each material according to the following equation: where t X and t MDF denote the mean time needed to make a 10 mm deep hole in the X material and in the raw MDF, respectively (in each material, 20 holes for each of the 3 weights were performed).

Results and Discussion
The results of the measurements of the time needed to make a 10 mm deep hole (in short-the drilling time) for all tested materials are presented in Figure 6. In any case, the drilling time was reduced when a heavier weight was used. For most materials, the averaged drilling time was less than 0.5 s. A longer time was observed only for two materials (compreg and transformer plywood).
The averaged time was considered a good indicator of ease of machining and therefore was used to calculate the cutting time problem index (CTPI) which was defined by Formula (2). The numerical values of this index, which were determined for all tested materials, are shown in Table 2.
In order to initially check whether the simple time measurement can be a useful (in terms of the machinability testing) substitute for cutting forces measurements, the graphs shown in Figures 7-9 were created. The first one (Figure 7) presents the relationship between the averaged time needed to make a 10 mm deep hole (which was measured for all fourteen boards during highly simplified machinability test) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12]. The correlation turned out relatively high (R 2 = 0.97). It is worth noting that this correlation remained significant (R 2 = 0.75) even when we limited the analysis to a group of relatively less diverse materials (we did not include compreg and transformer plywood- Figure 8). The averaged time was considered a good indicator of ease of machining and therefore was used to calculate the cutting time problem index (CTPI) which was defined by Formula (2). The numerical values of this index, which were determined for all tested materials, are shown in Table 2. In order to initially check whether the simple time measurement can be a useful (in terms of the machinability testing) substitute for cutting forces measurements, the graphs shown in Figures 7-9 were created. The first one (Figure 7) presents the relationship between the averaged time needed to make a 10 mm deep hole (which was measured for all fourteen boards during highly simplified machinability test) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12]. The correlation turned out relatively high (R 2 = 0.97). It is worth noting that this correlation remained significant (R 2 = 0.75) even when we limited the analysis to a group of relatively less diverse materials (we did not include compreg and transformer plywood- Figure 8). Figure 7. The relationship between the averaged time needed to make a 10 mm deep hole (drilling time) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12] for all fourteen tested materials. Figure 7. The relationship between the averaged time needed to make a 10 mm deep hole (drilling time) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12] for all fourteen tested materials. tween the averaged time needed to make a 10 mm deep hole (which was measured for all fourteen boards during highly simplified machinability test) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12]. The correlation turned out relatively high (R 2 = 0.97). It is worth noting that this correlation remained significant (R 2 = 0.75) even when we limited the analysis to a group of relatively less diverse materials (we did not include compreg and transformer plywood- Figure 8). Figure 7. The relationship between the averaged time needed to make a 10 mm deep hole (drilling time) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12] for all fourteen tested materials. Figure 8. The relationship between the averaged time needed to make a 10 mm deep hole (drilling time) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12] without taking into account the compreg and transformer plywood. Figure 8. The relationship between the averaged time needed to make a 10 mm deep hole (drilling time) and the averaged feed forces, measured during the earlier, significantly reliable machinability testing [12] without taking into account the compreg and transformer plywood.
Forests 2021, 12, x FOR PEER REVIEW 9 of 12 Figure 9. The relationship between the averaged time needed to make a 10 mm deep hole (drilling time) and the averaged drilling torque measured during the earlier, significantly reliable machinability testing [12] for all fourteen tested materials.
Unfortunately, the correlation between the results of the drilling time measurement and the results of the drilling torque measurement [12] was much smaller than for feed force (R 2 = 0.77), as shown in Figure 9.
The general and the most relevant (to the purpose of this paper) relationship between significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes, which were determined for all the fourteen materials tested, is shown in Figure 10. The distribu- Figure 9. The relationship between the averaged time needed to make a 10 mm deep hole (drilling time) and the averaged drilling torque measured during the earlier, significantly reliable machinability testing [12] for all fourteen tested materials.
Unfortunately, the correlation between the results of the drilling time measurement and the results of the drilling torque measurement [12] was much smaller than for feed force (R 2 = 0.77), as shown in Figure 9.
The general and the most relevant (to the purpose of this paper) relationship between significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes, which were determined for all the fourteen materials tested, is shown in Figure 10. The distribution of points had a quite strong linear character (R 2 = 0.92). Analogous diagrams, but separately for each of the 3 types of wood-based boards, are shown in Figures 11-13. The highest level of correlation (R 2 = 0.97) was found for particleboards ( Figure 10). For fiberboards and boards made of veneer or solid wood, the correlation was slightly smaller (R 2 = 0.95). Therefore, the division of the results into individual groups of wood-based boards can increase the degree of correlation. This fact is rather promising from a practical point of view-when selecting or developing innovative wood-based boards, materials with a similar internal structure are compared with each other most often. Figure 9. The relationship between the averaged time needed to make a 10 mm deep hole (drilli time) and the averaged drilling torque measured during the earlier, significantly reliable machin bility testing [12] for all fourteen tested materials.
Unfortunately, the correlation between the results of the drilling time measureme and the results of the drilling torque measurement [12] was much smaller than for fe force (R 2 = 0.77), as shown in Figure 9.
The general and the most relevant (to the purpose of this paper) relationship betwe significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes, whi were determined for all the fourteen materials tested, is shown in Figure 10. The distrib tion of points had a quite strong linear character (R 2 = 0.92). Analogous diagrams, b separately for each of the 3 types of wood-based boards, are shown in Figures 11-13. T highest level of correlation (R 2 = 0.97) was found for particleboards ( Figure 10). For fibe boards and boards made of veneer or solid wood, the correlation was slightly smaller ( = 0.95). Therefore, the division of the results into individual groups of wood-based boar can increase the degree of correlation. This fact is rather promising from a practical poi of view-when selecting or developing innovative wood-based boards, materials with similar internal structure are compared with each other most often. Figure 10. The relationship between significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for all the fourteen tested materials. Figure 10. The relationship between significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for all the fourteen tested materials.
Forests 2021, 12, x FOR PEER REVIEW 10 of 12 Figure 11. The relationship between the significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for particleboards. Figure 11. The relationship between the significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for particleboards. Figure 11. The relationship between the significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for particleboards. Figure 12. The relationship between the significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for fiberboards. Figure 13. The relationship between significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for boards made of veneer or solid wood.

Conclusions
The results of experimental research suggest that a highly simplified way of testing of relative machinability of wood-based boards (i.e., testing based on the photoelectric measurement of the time needed to make a 10 mm deep hole under constant feed force) can be a useful substitute for significantly reliable machinability testing procedure (based on accurate cutting forces measuring carried out in the real industrial conditions). When verifying the simplified testing procedure, samples from each of the three basic groups of Figure 12. The relationship between the significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for fiberboards. Figure 11. The relationship between the significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for particleboards. ure 12. The relationship between the significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes ich were determined for fiberboards. Figure 13. The relationship between significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for boards made of veneer or solid wood.

Conclusions
The results of experimental research suggest that a highly simplified way of testing of relative machinability of wood-based boards (i.e., testing based on the photoelectric measurement of the time needed to make a 10 mm deep hole under constant feed force) can be a useful substitute for significantly reliable machinability testing procedure (based on accurate cutting forces measuring carried out in the real industrial conditions). When verifying the simplified testing procedure, samples from each of the three basic groups of Figure 13. The relationship between significantly reliable (CFPI) and highly simplified (CTPI) machinability indexes which were determined for boards made of veneer or solid wood.

Conclusions
The results of experimental research suggest that a highly simplified way of testing of relative machinability of wood-based boards (i.e., testing based on the photoelectric measurement of the time needed to make a 10 mm deep hole under constant feed force) can be a useful substitute for significantly reliable machinability testing procedure (based on accurate cutting forces measuring carried out in the real industrial conditions). When verifying the simplified testing procedure, samples from each of the three basic groups of wood-based materials of substantially different internal structures (fiberboard, particleboard, and veneer boards) were tested. The relationship between highly reliable and significantly simplified machinability indexes turned out to be at a satisfactory level (R 2 = 0.97 for particleboards and R 2 = 0.95 for fiberboards or boards made of veneer or solid wood). The use of a simplified procedure can be especially pragmatic in case of any preliminary ("internal") testing of innovative wood-based boards during the material development work.
Funding: This research was funded by Polish Ministry of Science and Higher Education, project number: NN309007537.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.