Flow Lines with Regular Service Times: Evolution of Delay, State Dependent Failures and Semiconductor Wafer Fabrication James R. Morrison KAIST (Korea Advanced Institute of Science and Technology) IEEE CASE 2008 – August 23, 2008
Presentation Overview Motivation Description of flow lines Successive bottlenecks Resettable monotone channel (RMC) Delay in an RMC Delay in a flow line Application to semiconductor wafer fabrication Setups Batches of wafers (i.e., lots) Concluding remarks August 24, 2008 2 IEEE CASE 2008 – Washington, DC
Motivation Semiconductor wafer fabrication Many cluster tools: Multiple process chambers clustered into a single chassis Photolithography tools Models for simulation and throughput analyses Expressive (not just a throughput model, allows lot-to-lot interaction) Practical (e.g., setups, first wafer effect, …) Want to describe transient and lot-to-lot behavior GOAL: Develop expressive and computationally tractable flow line models August 24, 2008 3 IEEE CASE 2008 – Washington, DC
System Description: Flow Lines . . . Wafers Wafers m 1 m 2 m 3 m 4 m 5 m 6 m 7 m 8 m 9 m 10 m 11 m 12 Enter Exit t 1 t 2 t 3 t 4 t 5 t 6 t 7 t 8 t 9 t 10 t 11 t 12 M distinct modules: Wafers must receive service from all modules in order, then the wafer exits the system Deterministic (regular) process times: t i in module m i Manufacturing blocking At most one wafer may occupy a module at a given time If the downstream module is occupied, a wafer must wait to advance August 24, 2008 4 IEEE CASE 2008 – Washington, DC
System Description: Successive Bottlenecks m 1 m 2 m 3 m 4 m 5 m 6 m 7 m 8 m 9 30 s 10 s 20 s 0 s 40 s 35 s 28 s 60 s 18 s Definition: The modules with process time t i > t j , j < i are termed successive bottlenecks Decomposition in terms of these successive bottlenecks Definition: M consecutive modules are termed a resettable monotone channel (RMC) if t M (w+1) > t M (w) for all wafers until empty, then t M (w+1) > t M (1) t 1 < t M (k) and t 1 > t j , j = 2, …, M -1 m 1 m 2 m 3 m 4 m 5 30 s 10 s 20 s 0 s 40 s August 24, 2008 5 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation BEGIN SIMULATION = Processing = Delayed (blocked) Time: 0 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 1 August 24, 2008 6 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 10 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 1 August 24, 2008 7 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 20 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 1 August 24, 2008 8 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 30 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 2 1 August 24, 2008 9 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 40 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 2 1 August 24, 2008 10 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 50 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 2 1 August 24, 2008 11 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 60 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 3 2 1 August 24, 2008 12 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 70 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 3 2 1 August 24, 2008 13 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 80 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 3 2 1 August 24, 2008 14 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 90 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 4 3 2 1 August 24, 2008 15 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 100 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 4 3 2 August 24, 2008 16 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 110 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 4 3 2 August 24, 2008 17 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 120 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 5 4 3 2 August 24, 2008 18 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 130 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 5 4 3 2 August 24, 2008 19 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 140 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 5 4 3 August 24, 2008 20 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 150 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 6 5 4 3 August 24, 2008 21 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 160 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 6 5 4 3 August 24, 2008 22 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 170 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 6 5 4 3 August 24, 2008 23 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 180 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 7 6 5 4 August 24, 2008 24 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 190 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 7 6 5 4 August 24, 2008 25 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 200 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 7 6 5 4 August 24, 2008 26 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 210 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 8 7 6 5 4 August 24, 2008 27 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 220 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 8 7 6 5 August 24, 2008 28 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 230 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 8 7 6 5 August 24, 2008 29 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 240 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 9 8 7 6 5 August 24, 2008 30 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 250 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 9 8 7 6 5 August 24, 2008 31 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 260 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 9 8 7 6 August 24, 2008 32 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 270 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 10 9 8 7 6 August 24, 2008 33 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 280 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 10 9 8 7 6 August 24, 2008 34 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 290 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 10 9 8 7 6 August 24, 2008 35 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 300 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 11 10 9 8 7 August 24, 2008 36 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 310 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 11 10 9 8 7 August 24, 2008 37 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 320 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 11 10 9 8 7 August 24, 2008 38 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation = Processing = Delayed (blocked) Time: 330 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 11 10 9 8 7 August 24, 2008 39 IEEE CASE 2008 – Washington, DC
Wafer Delay in an RMC: Simulation END SIMULATION = Processing = Delayed (blocked) Time: 340 s t 1 = 30 s t 2 = 10 s t 3 = 20 s t 4 = 0 s t 5 = 40 s 12 11 10 9 8 August 24, 2008 40 IEEE CASE 2008 – Washington, DC
Recommend
More recommend