PDF  | Print |

 

A Case Study

In this short Note we will reference some results from our first installations of Warp-I, an installation that now has been up running for three years. This very first installation was done at Smurfit Kappa Sweden, Eslöv, and with the full support and encouragegement from the Site Manager, Peeter Saarnak.

For more details of our measuring principles see this article.

interface_warpi
Figure 1. A Warp-I installation.

Figure 1. shows the installation in the upper stacker used in this case study. Thre cameras are used to cover the full with of the corrugator.

In Figure 2. the interface of the computer in the control room of the corrugator is shown. Besides the images of the stacks (just under the stop bar) and the corresponding profiles there are some calculated values for warp. For each “Out” a value of the warp in % is given together with the mean warp for the upper stacker and lower stacker respectively. And finally an overall mean value for the warp over the full width of the corrugator is calculated. In addition the degree of warp is also expressed as the inverted radius (1/R), a property that is independent of the dimensions of the sheets. The full width of the corrugator is analyzed for warp every 3:rd second.

In this way the operators used our system for two years to carefully, in real-time follow, the warp performance of the produced corrugated board. Any deviation from specifications was noticed and actions were taken to make corrections. The stored warp-data was also used for traceability and follow-up of produced board.

interface_warpi
Figure 2. The Warp-I Interface

In addition the plant started to compare their own stored data of variables (speed, temperature, moisture, paper grade etc.) with the continuously monitored warp information. This gave them the opportunity to increase the understanding and fine tuning of the process, not only with respect to productivity but also to board quality.

In the autumn 2008 the Warp-I System was integrated with a quality control system from Qualitek, USA. The overall mean warp-value on the corrugator was used as actual value.

Since all historical data are saved in a data base it is possible to recall any time period in the past. To demonstrate how the control loop works we have chosen a period after a maintenance stop on the corrugator when two "outs" were run in the upper stacker. In Figure 3. the profiles of the "outs", calculated from one single scan across the machine, are illustrated.

warp_profile_2
Figure 3. Warp profiles

Figure 4. illustrates the average warp across the machine versus time for the chosen time period. Initially the desired value for the warp was set to be within 0 and -0.8 % (down warp). As can clearly be seen an increase in warp was observed under the first 12 minutes at which point an automatic correction of the wrap arms on the double backer took place. After a total time of 20 minutes it was decided to set the desired values to be within 0 and -0.4 %. After 25, 35 and 45 minutes automatic correction of wrap arms brought down the warp-value to 0 %. At 32 minutes the DB-liner was spliced.

warp_curve
Figure 4. Warp measurements with contol loop active.

The diagram shows both the accuracy and consistency in the measurements and the possibility to use this information to set up a closed loop control for warp performance.