maintenance man-hour/1000hours. Relation that MI＝λ× mh holds among
three quantities.
This data requires that maintenance and repair record of the ship is
accurately observed. In other words, records of maintenance and repair are valuable at the
time of work such as alteration, damage, inspection and exchange of spare parts. They
include date and time, repair man-hour, repair time, service parts used, type of work
(corrective maintenance or preventive maintenance). Late years, such source materials are
made ready for use. One example is a reliability database of ships which is prepared by
Ship Research Institute (Japan).
Ship is firstly planned, then designed and constructed. The plan
influences both the number of workers and repair hours which are required in ship's
operation and maintenance work on board. It is extremely important to make a plan of
workers and tools from a logistics point of view. FIG. 5 is a typical curve of work loads
input through life cycle (Blanchard, 1974). The FIG. 5 shows that logistics support
program starts in the beginning of product planning, and is continued to the operation
step where cost performance of the system is evaluated. Support of maintenance activity of
machinery is performed through entire life cycle of activity beginning from latter term of
production.
Actual demand of service parts and consumable supplies will depend on
the level od difficulty of OBM activity. Maintenance can be classified roughly into the
following:
(1) Serious failures which induces obstruction of seaworthiness,
operational balk in main engine are supported by an extraordinary examination followed by
serious repair.
(2) Corrective maintenance (CM) which includes recovery from general
failures, and mend of troubles.
(3) Preventive maintenance (PM) which includes routine maintenance of
servicing, plan of replacement, inspection, rounds, etc.
(4) Enforcement maintenance such as regular inspection.
In order to find out the optimum maintenance of ship, a simulation
model was built and some experimental results are given. SLAM II is the simulation
language which was developed in 1979 by Dr. A.A.B. Pritsker of American Pritsker &
Associate company (Pritsker, 1986). It is commonly used as simulation of discrete type and
continuous type. This model assumes following scheme (Miki, 1996): As preventive
maintenance is held daily, serious repair followed by corrective maintenance rarely
occurs. It takes + hours for personnel of preventive maintenance to make rounds and check,
and some preventive maintenance are done with service time of log-normal distribution β
hours which is equal to the average 0.8 and standard deviation 0.6. If a serious repair
happens while preventive maintenance is going on, corrective maintenance starts at once
and it takes time to repair of log-normal distribution with average 2.5 and standard
deviation 1.7.
We consider two cases of the number of maintenance workers required: In
case 1, both of preventive maintenance and corrective maintenance are performed by one
person. In case 2, the personnel specializes preventive maintenance. When a failure
arises, another personnel takes up corrective maintenance.
As for simulation analysis for case 1, simulation was carried out for
8,760 units of simulation time, during which 1,094 spare parts were consumed in preventive
maintenance. Workers are on duty for 8 hours. The mean time of preventive maintenance is
7.11 hours with standard deviation of 3.385 hours. Time for mending is maximum of 28.01
hours and minimum of 1.963 hours. 668 service parts was consumed for corrective
maintenance, which is on the average 2.563 hours. Availability of maintenance workers is
79.61%. FIG. 6 shows the preventive maintenance time distribution for case 1, where
maintenance service was performed by one person. FIG. 7 is corrective maintenance time
distribution. As a result, it is too much for one worker to engage in both of preventive
maintenance and corrective maintenance. Secondly, as for a simulation analysis for case 2,
(one person specializes preventive maintenance. When a failure arises, another person
takes up corrective maintenance. The number of workers which is engaged in preventive
maintenance is indicated by CREW1, and the number which performs corrective maintenance is
expressed by CREW2
Since CREW1 takes up only preventive maintenance, work burden is
reduced, and availability
amounts to 60.07%. The average preventive maintenance time is 4.811
hours, with maximum of 10.78 hours and minimum of 1.683 hour. CREW2 is supposed to be
hastened to the spot, if an alarm sounds.
