Project Detail

A major steel company was adding new equipment to an existing sheet coating line to improve the line speed. The line coats steel sheet for the automotive and distribution industries. The equipment had already been purchased, but since the technology had not been used before in the US, only the most basic process settings were known. The client knew that experimentation would be needed to optimize these settings, but the coating line was overbooked hence no activities were being performed prior to the scheduled installation.

  • Solution:Statistical Start-Up
  • Savings:$23 million annually


This equipment was designed by the inventor and promised to change the rules in sheet coating. The problem was that no one in the US, and few across the world, had ever used the equipment itself. It sat for a long time in storage since production was unwilling to set aside time on a highly constrained piece of equipment for something that had been perceived as a low probability of success.

Our Approach

When we were brought in, we set up a timeline of tasks to commission and tune the equipment according to our Statistical Start-Up methodology. Many of the tasks could be performed before installation. Area housekeeping was improved, and FMEAs, control plans, pre-installation baselining, gauge studies, and QFD tables were compiled, and training in problem solving, SPC, and equipment were performed over the four months before installation. The foundation of a Daily Management system was started as well. A statistically designed experiment was set up to study the effect of critical new process variables on different alloys and allow process optimization.

When the new equipment was available for testing, the experiment was performed on test coils and analyzed while the final touches were applied to the Daily Management system.

Project Scope

This work was done in two main episodes. The first was learning about the current process and putting in process controls to stabilize the line as much as possible. It still fall well short of production requirements. The second phase was setting up and then performing a fractional factorial design to identify optimal settings for speed and coating quality.
  • Speeds were less than expected
  • Coating adhesion was sometimes not meeting requirements
  • Dross (expensive waste) production was high
  • Process controls were lacking

End Results

The results of the gauge study showed that the usual maintenance recalibration of the coating thickness gauge actually introduced extra variability in the coating. A control chart-based calibration system was proposed to reduce coating thickness variability and save on wasted coating stock.

The results of the pre-installation baseline showed that in many cases the line actually exceeded temperature variability limits set up nearly a decade earlier. Equipment degradation and a lack of monitoring allowed for this condition, which might have been a cause of some surface defects.

The results of the experiment showed that the equipment could be run at faster speeds that even the manufacturer had guessed would be possible while actually reducing the number of coils scrapped for coating defects from about 2% to nearly none. All of the experimental coils were able to be sold. Due to production constraints, the line began producing coils at a higher inspection frequency with the new equipment immediately after the experimental results were completed. The next six months showed that the experimental results were confirmed, and that real production using normal coils at a conservative speed yielded a large increase in efficiency and maintained the nearly defect-free surface we had seen in the experiment. The Daily Management system was completed to monitor critical process and product characteristics to ensure that this profitable state was maintained and improved over the coming years.

The increase in speed and quality equaled an increase in throughput, and so a line that had been constrained and was behind schedule was able to catch up to production demands. The increase in productivity and surface quality is worth $23.6 million every year that the client fills that line to capacity.