Integration of ANN and RSM to Optimize the Sawing Process of Wood by Circular Saw Blades

At a Glance

Metadata	Details
Publication Date	2025-09-19
Journal	Applied Sciences
Authors	Mihai Ispas, Sergiu Răcășan, Bogdan Bedelean, A. Angelescu
Institutions	Transylvania University of Brașov
Analysis	Full AI Review Included

Executive Summary

Objective: To optimize the circular sawing process of solid wood (Beech and Spruce) by minimizing both cutting power (P_c) and surface roughness (R_a).
Methodology: A hybrid computational approach integrating Artificial Neural Networks (ANN) for robust predictive modeling and Response Surface Methodology (RSM) for process optimization.
Key Factors: Four independent variables were analyzed: Blade rotation speed (X1), Feed speed (X2), Wood species (X3), and Tool type (X4: teeth count/geometry).
Model Performance: The developed ANN models demonstrated high predictive accuracy, achieving a coefficient of determination (R²) of 0.9625 for cutting power and 0.8236 for roughness in the validation phase.
Critical Influence: Feed speed (X2) was identified as the most significant factor affecting both cutting power and surface roughness, followed by blade rotation speed (X1).
Optimal Regimes: Optimal conditions generally require minimum feed speed (3.5 m/min). For the 24-tooth blade, a high rotational speed (6000 rpm) is recommended. For the 54-tooth blade, a moderate speed (4500 rpm) is optimal.
Species Effect: Wood species (Beech vs. Spruce) significantly affected cutting power but did not show a statistically significant influence on surface roughness (R_a).

Technical Specifications

Parameter	Value	Unit	Context
Wood Species Tested	Beech (Fagus sylvatica), Spruce (Picea abies)	N/A	Experimental materials
Sample Dimensions	600 x 100 x 18	mm	Initial workpiece size
Average Moisture Content (Beech)	8.56	%	Wood condition
Average Moisture Content (Spruce)	8.11	%	Wood condition
Blade Diameter (D)	190	mm	Tool specification
Blade Bore Diameter	30	mm	Mandrel fitment
Kerf Width (b)	2.3	mm	Saw cut width
Blade Rotation Speed (X1) Range	3500 to 6000	rpm	Numerical factor range
Feed Speed (X2) Range	3.5 to 27	m/min	Numerical factor range
Roughness Stylus Tip Radius	2	µm	Mahr MarSurf XT20
Roughness Vertical Resolution	7	nm	Mahr MarSurf XT20
P_c ANN Model R² (Training)	0.987	N/A	Coefficient of correlation
R_a ANN Model R² (Training)	0.985	N/A	Coefficient of correlation
Optimal P_c (Beech, z=24)	0.364	kW	At 6000 rpm, 3.5 m/min
Optimal R_a (Beech, z=24)	8.41	µm	At 6000 rpm, 3.5 m/min

Key Methodologies

The study utilized a structured experimental design (Face Central Composite Design, FCCD) followed by dual computational modeling for prediction and optimization.

1. Experimental Setup and Measurement

Machine Tool: FELDER F 900 M spindle moulder, ensuring a 30 mm mandrel diameter constraint.
Tooling: Two distinct 190 mm diameter circular saw blades were used:
- Blade 1 (z=24): Wedge angle β = 60°, clearance angle α = 10°, hook angle γ = 20°, Alternate Top Bevel (ATB).
- Blade 2 (z=54): Wedge angle β = 65°, clearance angle α = 10°, hook angle γ = 15°, Flat Top Teeth.
Cutting Regimes: A total of 48 unique cutting regimes were tested (2 species x 2 blades x 3 rotational speeds x 4 feed rates).
Power Measurement: Active power (P_T) was measured using a Camille Bauer Sineax P530/Q531 transducer. Cutting power (P_c) was derived by subtracting the idle power (P₀).
Roughness Measurement: Surface roughness (R_a) was measured perpendicular to the grain using a Mahr MarSurf XT20 system. Data analysis included form error correction and Gaussian regression filtering (2.5 mm cut-off length).

2. Artificial Neural Network (ANN) Modeling

Purpose: To create highly accurate predictive models for the nonlinear relationships between input factors and output responses (P_c and R_a).
Software: NeuralWare’s Predict Software (v.3.24.1) utilizing a cascade-correlation learning algorithm.
Structure: The algorithm automatically determined the optimal number of hidden neurons (6 for P_c, 5 for R_a).
Validation: Models were validated against a subset of 20 experimental data points, confirming reliability for prediction.

3. Response Surface Methodology (RSM) Optimization

Purpose: To identify the optimal combination of numerical and categorical factors that simultaneously minimize P_c and R_a.
Design: Face Central Composite Design (FCCD) was used to generate quadratic polynomial equations describing the process behavior.
ANOVA Analysis: Statistical analysis confirmed that feed speed (X2) was the most influential factor for both responses.
Optimal Results: The optimization algorithm recommended specific low feed speeds (3.5 m/min) and high/medium rotational speeds, depending on the blade type, to achieve the lowest power consumption and best surface quality.

Commercial Applications

The integration of ANN and RSM provides a robust framework for optimizing industrial wood machining, offering significant benefits in efficiency and quality control.

Sawmill Operations: Directly applicable to optimizing ripping and sizing processes to maximize material yield and minimize energy consumption per unit volume of wood processed.
Tool Design and Manufacturing: Provides quantitative data on how specific blade geometries (tooth count, rake angle, bevel) interact with machining parameters, guiding the development of specialized, high-efficiency circular saw blades.
Quality Control and Predictive Modeling: The validated ANN models can be integrated into manufacturing execution systems (MES) to predict surface quality (R_a) and power consumption (P_c) in real-time, allowing for immediate parameter adjustments.
Energy Efficiency Programs: Enables wood processing facilities to select and implement machining regimes that are specifically optimized for low energy use, supporting sustainability goals and reducing operational costs.
Advanced Wood Machining Research: The methodology (ANN + RSM) is transferable to other complex mechanical processes involving anisotropic materials, such as routing, planing, or drilling of wood composites.

View Original Abstract

Various parameters, like blade design, rotational speed, feed speed, tooth geometry, wood moisture content, and wood species, influence the efficiency and quality of sawing processes. Knowing the optimal combination of these factors could lead to lower power consumption and high surface quality during wood processing. Therefore, in this study, we applied a novel method that could be used to optimize the cutting of wood with circular saw blades. The analyzed factors included rotational speed, feed speed, blade type (the number of cutting teeth and blade geometries), and two wood species, such as beech and spruce. The samples were cut longitudinally using two circular saw blades. The power consumption and the roughness of the processed surfaces were experimentally measured using an active/reactive electrical power transducer and a DAQ connected to a computer and a diamond stylus roughness meter, respectively. Once the data were gathered and processed, an artificial neural network modeling technique was involved in designing two models: one model for the cutting power and the other for surface roughness. Both models are characterized by high values of performance indicators. Therefore, the models could be considered a reliable tool that could be used to predict the cutting power and the surface roughness for the cutting of wood with circular saw blades. Next, response surface methodology was used to identify how each factor affects the cutting power and the surface quality, and to find the optimal values for both. The results showed that the most important factor that influences the roughness of the processed surfaces is the feed speed; the second factor is the blade rotation speed; the third factor is the tool type (the number of cutting teeth combined with their geometry). The optimal machining conditions recommended by the optimization algorithm (low power consumption and low roughness) imply minimum feed speed values (3.5 m/min) and medium (4500 rpm for 54-tooth blade) or high (6000 rpm for 24-tooth blade) blade rotation speeds. A further study will be conducted to consider the behavior of wood species during the circular sawing of wood and to clarify the influence of the different constructive parameters of the blades (number of teeth, tooth geometry) on their performance.

Tech Support

Original Source

DOI: https://doi.org/10.3390/app151810206

References

1971 - On the Behaviour of Circular Sawblades during Cutting—Part II: Effect of the Cutting Conditions on the Quality of Sawn Wood Surfaces [Crossref]
1992 - Effect of tooth front bevel angle on cutting accuracy and chip formation for circular rip saws [Crossref]
2011 - Specific cutting energy consumption in a circular saw for Eucalyptus stands VM01 and MN463 [Crossref]
2012 - Economical Wood Sawing with Circular Saw Blades of a New Design [Crossref]
2013 - Industrial Sawing of Pinus sylvestris L. Power Consumption
2014 - Effect of the saw blade construction on the surface quality when transverse sawing spruce lumber on crosscut miter saw
2015 - The Dependence of Surface Quality on Tool Wear of Circular Saw Blades during Transversal Sawing of Beech Wood [Crossref]
2017 - The influences of circular saws with sawteeth of mic-zero-degree radial clearance angles on surface roughness in wood rip sawing [Crossref]
2018 - Quality of machined surfaces and specific cutting energy in wood of two African mahogany species
2019 - Practical Guideline in the Design and use of Woodworking Tools [Crossref]