Published March 17, 2023 | Version 1.0.0

Vibration Sensor and Process Data from Ferrosilicon Production

Description

Elkem Facility

The facility specializes in producing ferrosilicon (FeSi) and ferrosilicon magnesium (FSM) master alloys. Elkem Bjølvefossen is among the world’s largest producers of FSM. Three reduction furnaces deliver the base metal which is then alloyed and refined to the right quality of FeSi or FSM. These alloys are important additives in the manufacturing of steel products. Silicon in the form of FeSi is used to remove oxygen from the steel and as an alloying element to improve the final quality of the steel. Silicon increases strength and wear resistance, elasticity, i.e., spring steels, scale resistance, and heat resistant steels and lowers electrical conductivity and magnetostriction.

After tapping and refining, the ferro-alloys are crushed to grains ranging from 1 mm to 25 mm in size. Consumers of FeSi and FSM have strict requirements for particle size, related mainly to the chemical kinetics of their refining and alloying processes. For this reason, the crushed material is separated in sieves and packaged by particle size before shipment. Two lattice gratings inside the Mogensen shaker separate the material according to required particle size. 

The subject of the present study is a mechanical shaker platform containing one or more such sieves. The shaker is a Mogensen S0556  that was installed in 1996 in Bjølvefossen and is no longer produced in this type. This device is powered by two counter-rotating 1.2-horsepower AC motors operating at 960 RPM. Together with the spring suspension, these cause an elliptical motion that both transports and scatters the incoming material across the sieve. The shaker is engineered so that the motion transitions from a slanted ellipse at the in-feed to nearly linear at the output.

Vibration Data

Vibration data from two sensors. Each sensor measures acceleration in three axes with three different ADCs. 

ERP and MES data

Manufacturing Execution System (MES) data as well as process data from the Enterprise Resource Planning (ERP) is given. It is providing information about the material that is currently being produced as well as the data from the scales from the material packing station where the bags with completed production were packed. MES data has a resolution of 5 seconds and the process data from ERP has a resolution of roughly 10 minutes. This operational data is meant to provide insight into the current state and throughput of the facility and will serve as labels for the correlation analysis with the vibration data. 

Notes

This work was partly funded through the projects BigDataMine (NFR 309691), DataCloud (H2020 101016835), enRichMyData (HE 101070284), and Graph-Massivizer (HE 101093202).

Files

vibration_data_sample.zip

Files (6.0 GB)

Name Size
md5:ea5c27565b5055a7c348e8201714630f
6.0 GB Preview Download

Additional details

Funding

European Commission
DataCloud - ENABLING THE BIG DATA PIPELINE LIFECYCLE ON THE COMPUTING CONTINUUM 101016835
European Commission
enRichMyData - Enabling Data Enrichment Pipelines for AI-driven Business Products and Services 101070284
European Commission
Graph-Massivizer - Extreme and Sustainable Graph Processing for Urgent Societal Challenges in Europe 101093202