In-situ identification of material batches using machine learning for machining operations

Abstract

In subtractive manufacturing, differences in machinability among batches of the same material can be observed. Ignoring these deviations can potentially reduce product quality and increase manufacturing costs. To consider the influence of the material batch in process optimization models, the batch needs to be efficiently identified. Thus, a smart service is proposed for in-situ material batch identification. This service is driven by a supervised machine learning model, which analyzes the signals of the machine’s control, especially torque data, for batch classification. The proposed approach is validated by cutting experiments with five different batches of the same specified material at various cutting conditions. Using this data, multiple classification models are trained and optimized. It is shown that the investigated batches can be correctly identified with close to 90% prediction accuracy using machine learning. Out of all the investigated algorithms, the best results are achieved using a Support Vector Machine with 89.0% prediction accuracy for individual batches and 98.9% while combining batches of similar machinability.

Towards Material-Batch-Aware Tool Condition Monitoring

“…This agrees with ground-truth data acquired from the process planning, as only one lot of material was used during the investigated time. In contrast, Figure 5b shows the cluster map of the dataset investigated in [10]. Here, two clusters can be clearly distinguished, which align well with the different machineabilities found in the related study.…”

“…According to the review in [12], most approaches in the area of automated material identification use sophisticated feature engineering, which so far has not been investigated for material batch identification in subtractive manufacturing. Thus, this study extends the previous work from [10] by investigating a variety of features for a new validation scenario.…”

“…One such influence are variations between material batches, which were shown to have significant influence on the material and its machinability [10]. The root cause for these deviations can be found in the underlying material manufacturing processes, between different suppliers and material lots, such as variations in the material and its chemical composition, fabrication procedure, or heat treatment.…”

“…It is shown that there are several material identification approaches researched in subtractive manufacturing, which aim at identifying various major materials, material sub-classes, specific materials, material grades, and material batches. While there are several approaches for identifying different materials during machining, such as aluminum and steel [13][14][15], aluminum and carbon-fiber-reinforced plastics (CFRP) [16,17], titanium and CFRP [18,19], and steel and ceramic [20], there are only a few approaches investigating batch-specific material identification [10,21].…”

Predicting Nonlinear and Anisotropic Mechanics of Metal Rubber Using a Combination of Constitutive Modeling, Machine Learning, and Finite Element Analysis

“…Artificial neural networks (ANNs) have many advantages, such as efficiency and ease of implementation. ANNs are convenient for accurate prediction and numerical simulation because they ignore all the processes and theories; thus, they have been widely used to study material properties [35][36][37][38][39][40][41][42][43][44][45][46][47][48]. The combination of machine learning and FEM has been used to solve complex engineering problems [49][50][51][52].…”