Question

A valve is assembled in two process cells by one worker each. The first worker inserts a spring and stem into the body with a tria(1,2,3) seconds process time, while the second worker closes the valve and tests for leaks in tria(5,10,15) seconds. The valve parts arrive into the system at the rate of expo(15) seconds. Approximately 3% of the assembled valve stems fail, requiring disassembly by a machine within tria(1,2,3) seconds. Disassembled valves are reworked in stage 1 with the same process time. Valves must be reworked only once, and any that fail after the one rework must be discarded.

Show the model and all relevant statistics about process wait times and resource availabilities. Run the simulation for one shift of 8 hours, estimating averages with a half-width of the 95% CI within about 10% of the mean.

Answer 1

The student's question deals with creating a simulation model for a valve assembly line in industrial engineering, which requires calculating process times.

Step-by-step explanation:

The question pertains to the field of industrial engineering and involves the creation of a simulation model for a valve assembly process. The first step involves a worker assembling a valve with a process time described by a triangular distribution (tria) with parameters (1,2,3) seconds. The second step has another worker closing and testing the valve, taking a process time with parameters (5,10,15) seconds. The rate of valve part arrival is given by an exponential distribution with a mean of 15 seconds (expo(15)). Additionally, there is a 3% failure rate for valve stems, which necessitates a rework process with a time distribution identical to the initial assembly (tria(1,2,3) seconds).

To create a simulation for one 8-hour shift, you would need to program these parameters into a discrete-event simulation software. The required statistics, such as the average wait time of parts, the utilization rate of workers, and the availability of resources, would be computed by the simulation. For confidence intervals, the simulation must be run multiple times to estimate the average performance metrics with the desired precision (half-width of the 95% CI) within 10% of the mean.