CAPITULO II: MARCO TEÓRICO
2.7. MARCO CONCEPTUAL
To research whether the simulation is reliable, a validation study is done. The cycle time, lead time (time from the start of a new stack, until that stack is done and moved to packing), and the process time for each area are measured. The simulation is run with the same input, after which the times are compared. SeeTable 6.6for the information about the stacks for which
the process times were measured. Graphs of the comparison between the real situation and the simulation are given inFigure 6.4.
Test Thick- Stack Stack Rip Cross Sheets Parcels
ness length width cuts cuts
(mm) (mm) (mm) 1 2.50 4290 2550 4 3 360 6 2 30.00 7400 2510 3 4 30 6 3 7.00 4050 2550 3 2 80 4 4 12.00 4200 2550 3 2 59 5 5 12.00 3770 2550 3 3 75 6 6 4.00 3570 2550 3 3 230 6 7 3.00 4210 2550 4 3 200 4
Table 6.6:Stack information for which the process times were measured for the validation study.
There is a limited number of comparisons of the cycle time and lead time, since these must be measured at a time when the B-saw is running without stopping.
The infeed shows a lot of variation. This is explained by the fact that the operator has a lot of influence on what happens at the infeed. It is tried to simulate the decisions the operator makes, but it is of course not fully accurate. It is assumed that it will only have a limited effect on the results of the simulation. This is, however, not validated.
The cross cut process times are lower in every situation. This is caused by the fact that the cross cut pusher retreats all the way to the back. Only in the case when the cross cut area is the bottleneck, and a parcel is waiting on the white belts, the cross cut pusher moves to the right position. In the simulation, however, it is assumed that the cross cut pusher moves to the right position in every situation, even when it is not the bottleneck. This does not have any effect on the cycle time, and only limited effect on the lead time, since the cross cut area was not the bottleneck in these situations.
In conclusion, the validation shows that for a range of input stacks, the model provides an accurate simulation of the B-saw.
6.8
Implementation
It was expected to be able to implement the results of the simulation during the internship. However, during the last weeks of the internship, the electrician who is able to change the PLC did not have time.
However, a ten second improvement possibility is identified at ”Back cross cut”. Some smaller improvement possibilities were identified at the ”Rip cut”. These possibilities are communi- cated with the electrician. It is expected that the simulation results will soon be implemented.
(a) (b)
(c) (d)
(e) (f)
Figure 6.4:Process time comparison between the B-saw and the simulation of the B-saw. The cycle time, lead time, infeed time, rip cut time, cross cut time and stacking time are compared.
7 Conclusions
Chapter 7
Conclusions
This report started out with the goal of improving the sawing machines at Nelson Pine, to reduce the need for overtime. The overall equipment effectiveness was used as a framework. It was decided to focus on the machine with the highest downtime. This turned out to be the B-saw (with a downtime of 22%). The causes of the downtime of the B-saw were investigated, and a Pareto analysis was done on the causes. The top ten downtime reasons, which account for 40% of the downtime, were discussed with employees of Nelson Pine. To tackle the problems that were still relevant, the problems were divided over the employees.
An attempt was made to solve one of the major problems, the ”White belt broken”. Together with the relevant ”White belt off”, this problem accounts for 6% of the downtime. In the case the designed ”Infeed blocks” would work, it would not only reduce the downtime by solving the white belt problems, but also increase the quality of the boards (which is also a factor of the overall equipment effectiveness). Since the strapping line has to manually remove broken boards, a quality increase would lead to a lower downtime of the ”Strapping line down” (this problem currently accounts for 6% of the downtime).
Before this internship, it was only possible by observation to see whether these problems are indeed solved. To facilitate a culture of continuous improvement, an automatic downtime anal- ysis system is implemented in the B-saw. Every time the machine stops for a consecutive time of two minutes, the operator must assign a downtime reason associated with the stop. Next to the uptime and quality improvement stated above, the throughput, or theoretical output, of the B-saw was researched. This was done with the aid of a simulation model. The simulation model looked for the bottleneck area, and the bottleneck area that would appear when the process time of the previous bottleneck area would be reduced by one second. This was done for 30 times. Based on the simulation results, the B-saw can be improved by adjusting the programmable logic controller software. The improvements made with Nelson Pine, should result in an overall throughput improvement of almost 10%.
The improvement in throughput is amplified by a higher uptime, so this research leads to a direct increase in overall equipment effectiveness of the B-saw of about 16%.
Chapter 8
Recommendations
During the weekly ”Sanders and saws meetings”, currently, no information is given about the downtime of the sander and sawing machines. Only a subjective comment is given by the work supervisor whether there were any major issues last week. To create awareness of the signif- icant amount of downtime at the sawing machines (potentially also at the sander machines), the writer recommends that the downtime of the last week of the B-saw will be inspected every week during the ”Sanders and saws meetings”. Major problems can be discussed shortly, after which action points could be given.
Next to the weekly analysis of downtime problems, it is recommended to inspect the downtime reasons over a longer period of time, say three months. Problems that are not significant during a week’s period of time, might have a big impact on a longer period of time when they are reoccurring frequently. Also, a comparison between different periods of time, might work very stimulating when it can be seen that problems are solved, and downtime decreases.
This way of working should result in a culture of continuous improvement. When the major prob- lems are solved, the downtime should decrease, after which the maintenance staff is called less frequently. The maintenance staff should now have more time to focus on different machines, or on improving the uptime of the machine even further, by solving less significant problems. It is advised to not implement the downtime system on other machines yet, but rather use the system on the B-saw only and prove its worth. In that case, supervisors and operators of other machines might see the added value of such a system and use it properly.
In case the downtime of one or multiple machines has been reduced significantly, and the ma- chines should be able to process even more board, simulation studies can be done to further increase the capacity of the machines. The simulation made during this study, can be used to further improve the B-saw. New simulation studies can be made to analyze and improve other machines.
A Infeed block results
Appendix A. Infeed block results
Original situation Optimal 2 blocks (64 & 133mm) Length (mm) Block (mm) Cross cut Block (mm) Cross cut rollers
rollers overhang (mm) overhang (mm) 3100 100 70 64 34 3150 50 70 64 84 3200 0 70 0 70 3560 0 86 0 86 3750 50 75 64 89 3770 0 45 0 45 3780 0 55 0 55 4050 0 80 133 33 4060 100 10 133 43 4100 100 50 64 14 4110 50 10 64 24 4150 50 50 64 64 4200 0 50 0 50 4210 0 60 0 60 4250 0 100 133 53 4280 100 50 133 83 4290 100 60 64 24 4350 50 70 64 84 4360 0 30 0 30 4460 0 130 133 23 4560 100 90 133 3 4760 0 70 0 70 4770 0 80 0 80 4800 0 110 133 3 4950 0 20 0 20 4960 0 30 0 30 4990 0 60 0 60 5160 50 40 64 54 5180 0 10 0 10 5200 0 30 0 30 5480 100 50 64 14 5555 0 25 0 25 5570 0 40 0 40
5800 0 35 0 35 6310 0 185 64 4 6560 0 70 0 70 6720 100 95 64 59 6970 0 5 0 5 7280 50 5 64 19 7400 50 5 64 19 7410 50 15 64 29 7460 0 15 0 15
Table A.1:Table with the blocks that should be used for a certain length of an input stack. The results are given for the original situation, and the situation with 2 optimal blocks.
B PLC flowcharts
Appendix B. PLC flowcharts
B PLC flowcharts
B PLC flowcharts
Appendix C. Simulation model
The model is developed in Technomatix Plant Simulation 11. The model consists of a main screen, shown inFigure C.1, from which the simulation is controlled. One can change the vari-
ables seen in “Simulation control” to alter the simulation.
Figure C.1:Main screen of the simulation. From here the simulation can be controlled. One can zoom in on the areas Infeed, Rip cut, Crosscut and Stacking.
C Simulation model