Simulations of rolling steel shapes reflect reality on the mill floor
As the steel sector moves towards Industry 4.0 and smart steel plants, the use of models and simulations is becoming more prevalent. Models of steel as it is being transformed, the production process from beginning to end, and the complete plant including auxiliary equipment, are all are contributing towards the creation of a virtual steel plant. But one question still remains. Will the models reflect reality?
Russula and Scientific Forming Technologies Corporation have proven they already do. The two companies collaborated in a project to eliminate product defect and rolling instability occurring at a section mill. UPN 30 structural beams were simulated using the DEFORM® engineering software (manufactured by SFTC), which identified voids and overfilling at various passes.
3D simulations revealed voids and overfills while rolling UPN 30 with the current pass design.
In order to compare the accuracy of the simulations, samples from the section mill were taken from every forming pass by operating, Russula-loppers’ snap shears. The predicted dimensions of the final section were almost the same as the dimensions measured while rolling UPN 30 in the mill. The Russula team concluded that the DEFORM® simulations were very accurate.
| UPN 30 mm | DEFORM mm | MEASUREMENTS mm | DIFFERENCE mm | |
|---|---|---|---|---|
| h | 30.00 | 30.00 | 30.00 | 0.00 |
| b | 15.00 | 15.10 | 14.80 | 0.30 |
| tw | 4.00 | 4.00 | 3.90 | 0.10 |
| tF | 4.50 | 4.50 | 4.30 | 0.20 |
| r1 | 4.50 | 4.50 | 4.50 | 0.00 |
| r2 | 2.50 | 2.50 | 2.50 | 0.00 |
Predicted dimensions compared favorably to measured dimensions.
Because DEFORM® can accurately predict the results of a process prior to implementation, users can routinely test variations prior to committing to a design, resulting in more optimal and robust processes. The collaboration was able to accurately simulate rolling conditions on a computer, rather than on the rolling mill.
The software works really well. It never crashes, it’s very stable, extremely reliable, and it makes modeling easier because we can skip set-up time, and get to market faster.
