Related Experiment Video
Updated: Sep 16, 2025

Determination of Aggregate Surface Morphology at the Interfacial Transition Zone ITZ
Published on: December 16, 2019
Surface Topography-Based Classification of Coefficient of Friction in Strip-Drawing Test Using Kohonen
Krzysztof Szwajka1, Tomasz Trzepieciński2, Marek Szewczyk1
1Department of Integrated Design and Tribology Systems, Faculty of Mechanics and Technology, Rzeszów University of Technology, ul. Kwiatkowskiego 4, 37-450 Stalowa Wola, Poland.
This study introduces an AI method to classify the coefficient of friction (CoF) in sheet metal forming using surface roughness data. The unsupervised approach achieved 98% accuracy, improving quality and reducing defects.
Area of Science:
- Materials Science and Engineering
- Artificial Intelligence and Machine Learning
Background:
- Accurate monitoring of the coefficient of friction (CoF) is crucial in sheet metal forming (SMF) for product quality and process efficiency.
- Traditional CoF monitoring methods struggle to identify force signal features sensitive to friction variations.
- Changes in CoF during SMF are influenced by evolving contact conditions, surface topography, contact loads, and material behavior.
Purpose of the Study:
- To develop a novel approach for classifying the CoF in DP780 steel sheets during strip-drawing tests.
- To apply unsupervised artificial intelligence (AI) techniques to handle unbalanced data for CoF classification.
- To identify disturbances in the friction process by analyzing friction parameters and sheet metal surface roughness.
Main Methods:
- Utilized unsupervised AI, specifically the Kohonen self-organizing map (SOM), for CoF classification.
- Collected surface topography parameters, including the Sq roughness parameter, and friction test parameters.
- Conducted strip-drawing tests under varying lubrication conditions, contact pressures, and sliding speeds.
Main Results:
- Achieved a high classification accuracy of up to 98% for the coefficient of friction (CoF).
- Demonstrated the effectiveness of using the SOM for classifying CoF with unbalanced data.
- Showcased that CoF classification is feasible using only the Sq surface roughness parameter and selected friction test parameters.
Conclusions:
- The proposed unsupervised AI approach effectively classifies the CoF in sheet metal forming processes.
- Surface topography parameters, particularly Sq, combined with friction test data, are reliable indicators for CoF classification.
- This method offers a promising solution for monitoring and controlling friction in SMF, leading to improved product quality and process optimization.
Related Concept Videos
Characteristics of Dry Friction
Before the wheelbarrow starts moving, the static frictional force acts tangentially to the contact surface, opposing the force that is about to induce the motion. This frictional force prevents the...
Frictional Force
Types of Friction Problems
The first type of dry friction problem involves situations where there is no apparent impending motion....
Friction: Problem Solving
Initially, a visual representation of the...
Rolling Resistance
For instance, imagine a hard cylinder rolling on a comparatively soft surface. The cylinder's weight compresses the surface beneath it. As the cylinder moves, the material in front of it slows down...
Methods of Obtaining Topography

