Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Computer simulation of overseas product manufacturing potentials of a redesigned multipass headrig mill… Orbay, Laszlo 1984

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-UBC_1984_A1 O73.pdf [ 34.86MB ]
Metadata
JSON: 831-1.0096564.json
JSON-LD: 831-1.0096564-ld.json
RDF/XML (Pretty): 831-1.0096564-rdf.xml
RDF/JSON: 831-1.0096564-rdf.json
Turtle: 831-1.0096564-turtle.txt
N-Triples: 831-1.0096564-rdf-ntriples.txt
Original Record: 831-1.0096564-source.json
Full Text
831-1.0096564-fulltext.txt
Citation
831-1.0096564.ris

Full Text

COMPUTER SIMULATION OF OVERSEAS PRODUCT MANUFACTURING POTENTIALS OF A REDESIGNED MULTIPASS HEADRIG MILL IN COASTAL BRITISH COLUMBIA by LASZLO ORBAY DIPL.WOOD IND.ENG.,UNIVERSITY OF SOPRON, 1966 DIPL.BUS.ADM.,UNIVERSITY OF ECONOMICS, BUDAPEST, 1974 A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF FORESTRY We accept t h i s t h e s i s as conforming to the r e q u i r e d standard THE UNIVERSITY OF BRITISH COLUMBIA JUNE 1984 © LASZLO ORBAY, 1984 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements f o r an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y a v a i l a b l e for reference and study. I further agree that permission for extensive copying of t h i s t h e s i s fo r s c h o l a r l y purposes may be granted by the head of my department or by h i s or her representatives. I t i s understood that copying or publication of t h i s t h e s i s for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department O f F o r e s t R e c o r c e s Management. F a c u l t y of F o r e s t r y The University of B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 'E-6 (3/81) ABSTRACT M o d i f i c a t i o n of the sawmill Woodroom #3, MacMillan B l o e d e l L t d . , at. Harmac became necessary because of i t s uneconomical o p e r a t i o n . The key concept of the m o d i f i c a t i o n p l a n was to gain f l e x i b i l i t y in s e r v i n g seven overseas and U.S.A. markets. S t o c h a s t i c components of s a w m i l l i n g and the high cost of r e c o n s t r u c t i o n make investment d e c i s i o n s complex and r i s k y . A new computer s i m u l a t i o n model was needed and developed to e v a l u a t e behaviour of the redesigned m i l l . In b u i l d i n g the model, two major f a c t o r s were taken i n t o c o n s i d e r a t i o n which a f f e c t m i l l performance, log breakdown and l o g throughput. Thus, the model i s a combination of FLOWSIM, a newly developed dynamic p i e c e flow s i m u l a t o r and SAWSIM, an e x i s t i n g sawing s i m u l a t i o n package. A s p e c i a l s i m u l a t i o n language GPSS/H, the most e f f e c t i v e v e r s i o n of GPSS, was employed to b u i l d FLOWSIM. SAWSIM, p r o v i d i n g a l o g breakdown l o g i c f o r FLOWSIM, simulates sawing p a t t e r n s more a c c u r a t e l y than any other model. Using SAWSIM with FLOWSIM reduced programing e f f o r t and improved accuracy of sawmill s i m u l a t i o n . I n t e r a c t i n g a u x i l i a r y programs were w r i t t e n and o r g a n i z e d i n t o one system which supported the o p e r a t i o n of the two major s i m u l a t o r s and the data t r a n s m i s s i o n between them. Six market runs were c a r r i e d out to t e s t market f l e x i b i l i t y of p r o d u c t i o n . S i m u l a t i o n r e s u l t s of l o g demand, lumber value i i i and volume of p r o d u c t i o n , and machine and t r a n s p o r t a t i o n equipment u t i l i z a t i o n p r ovided a b a s i s f o r the f i n a l investment d e c i s i o n s . Model output helped to l o c a t e flow problems and was used to recommend f u r t h e r m i l l design m o d i f i c a t i o n s . To estimate the e f f e c t of machinery breakdown on p r o d u c t i v i t y , three runs were accomplished. As an example of breakdown s e n s i t i v i t y a n a l y s e s , an a d d i t i o n a l run was made to estimate- the l o s t p r o d u c t i o n i n d o l l a r s per hour of machine "down" time. A l l runs simulated m i l l o p e r a t i o n f o r one s h i f t . S i m u l a t i o n of Woodroom #3 m o d i f i c a t i o n plans shows that i t i s p o s s i b l e to b u i l d a f l e x i b l e sawmill capable of producing for v a r i o u s overseas markets. Although, the s i m u l a t i o n model of t h i s t h e s i s was designed to analyze a p a r t i c u l a r sawmill of MacMillan B l o e d e l , the model b u i l d i n g concepts c o u l d be used to h e l p develop a general model for sawmill s i m u l a t i o n i n other m i l l s and other r e g i o n s . T a b l e of Contents A b s t r a c t i i Tab le of Content s i v L i s t of T a b l e s v i i i L i s t of F i g u r e s ix Acknowledgement x i 1 . INTRODUCTION 1 2. OBJECTIVES 10 3. METHODS AND PROCEDURES 12 3 . 1 . REASONS FOR COMPUTER SIMULATION 12 3 .2 . WHY EXISTING SIMULATION MODELS SHOULD BE IMPROVED 14 3 . 3 . MAJOR PHASES OF THE SIMULATION PROCEDURE AND THEIR RELATIONSHIP 22 3 .4 . DESIGN OF THE COMPUTER SIMULATION RUNS 27 4. LITERATURE REVIEW OF COMPUTER SIMULATION APPLICATIONS IN SAWMILLING 33 5. INPUT DATA NEEDED FOR THE MODEL 40 5 . 1 . ESTABLISHMENT OF THE SAMPLE LOG DATA BASE 40 5 . 1 . 1 . Two-dimens iona l d iameter and l e n g t h d i s t r i b u t i o n of logs i n Woodroom #3. Taper as an a d d i t i o n a l shape c h a r a c t e r i s t i c 42 V 5 . 1 . 2 . Assignment of sample boom logs to the c l a s s e s of the two-d imens iona l d i s t r i b u t i o n 44 5 . 1 . 3 . Buck ing p o l i c i e s of Woodroom #3 47 5 . 1 . 4 . Sawlog s e l e c t i o n for SAWSIM runs 48 5 . 2 . EXPORT MARKETS, LUMBER SIZES AND PRICES 57 5 . 3 . MACHINERY 60 5 . 4 . DESCRIPTION OF MILL OPERATION 66 6. APPROACH TAKEN IN BUILDING THE MODEL 69 6 . 1 . THE PROBLEM OF DATA TRANSMISSION 71 6 . 2 . THE CONCEPTS OF DYNAMIC MODEL CONSTRUCTION . . . 8 1 7. MODEL VALIDATION 108 7 . 1 . LOG BREAKDOWN VALIDATION 110 7 . 2 . PIECE FLOW VALIDATION 111 8. COMPUTER SIMULATION RUNS, RESULTS AND DISCUSSION .127 8 . 1 . STRUCTURE OF INTERACTING PROGRAMS 127 8 . 2 . RESULTS OF SIMULATION RUNS AND DISCUSSION . . . 1 3 2 9. SUMMARY AND CONCLUSIONS 157 10. REFERENCES 162 10.1 BIBLIOGRAPHY OF COMPUTER SIMULATION APPLICATIONS IN SAWMILLING 162 o 10.2 BIBLIOGRAPHY REFERENCED THROUGHOUT THE THESIS 1 67 Appendix 5 . 1 . Length and d iameter e m p i r i c a l f requency d i s t r i b u t i o n of boom logs 171 Appendix 5 . 2 . Boom log f requency d i s t r i b u t i o n 172 Appendix 5 . 3 . Taper and sweep c a l c u l a t i o n program 173 v i Appendix 5 . 4 . P r o p o r t i o n e s t i m a t i o n of B and N boom logs 176 Appendix 5 . 5 . Log data c a l c u l a t i o n s 177 Appendix 5 . 6 . Log poo l w i t h i t s sample boom l o g codes . . . 1 8 4 Appendix 5 . 7 . Examples of boom log measurements 185 Appendix 5 . 8 . Bucking s chedu le s 186 Appendix 5 . 9 . Expected cant widths as a f u n c t i o n of sma l l d iameter 188 Appendix 5 .10 .Sawlog s e l e c t i o n t a b l e 189 Appendix 5 . 1 1 . B u c k i n g program 190 Appendix 5 .12 .Sawlog s e l e c t i o n program 199 Appendix 6 . 1 . FLOWSIM l i s t i n g 207 Appendix 6 . 2 . Headr ig l o s t t ime data 279 Appendix 6 . 3 . F u n c t i o n f o l l o w e r c a r d w r i t e r program 280 Appendix 6 . 4 . T r a n s p o r t a t i o n equipment speeds and l eng ths 284 Appendix 7 . 1 . V a l i d a t i o n of m i l l dynamics by i n t e r a c t i v e GPSS/H run 286 Appendix 8 . 1 . Data Base #1 w r i t e r program 291 Appendix 8 . 2 . Log measurement w r i t e r program 296 Appendix 8 . 3 . DIA w r i t e r program 300 Appendix 8 . 4 . M a t r i x reader program 305 Appendix 8 . 5 . Log s e l e c t f i l e w r i t e r program 311 Appendix 8 . 6 . Condensed FLOWSIM o u t p u t - " T " market 313 Appendix 8 . 7 . P r o d u c t i o n r e s u l t s - " T " market 322 Appendix 8 . 8 . Condensed FLOWSIM o u t p u t - " K " market 324 Appendix 8 . 9 . P r o d u c t i o n r e s u l t s - " K " market 333 Appendix 8 .10 .Condensed FLOWSIM o u t p u t - " A " market 335 v i i Appendix 8.11.Product ion r e s u l t s - " A " market 344 Appendix 8.12.Condensed FLOWSIM output-"F" market 346 Appendix 8.13.Product ion r e s u l t s - " F " market 355 Appendix 8.14.Condensed FLOWSIM o u t p u t - " J " market 357 Appendix 8.15.Product ion r e s u l t s - " J " market 367 v i i i L i s t of Tables 3.1. Comparison of two log breakdown s i m u l a t o r s 17 5.1. Lumber s i z e s of v a r i o u s markets 58 5.2. Data to c a l c u l a t e maximum and minimum rough green s i z e s 58 5.3. Lumber and byproduct p r i c e s (Canadian $) 59 5.4. Lumber p r i c e s of the U.S.A. market (Canadian $) 60 5.5. Machine s p e c i f i c a t i o n data 62 5.6 Machinery breakdown data 64 5.7. Number of pi e c e s which can t r a v e l side by side and the average t h i c k n e s s of lumber l a y e r 65 6.1. T y p i c a l groups of GPSS cards and the corresponding p a r t s of FLOWSIM 82 6.2. Summary of observed downtime data of Headrig 9' 86 6.3. I n t e r p r e t a t i o n of t r a n s a c t i o n s i n d i f f e r e n t segments 89 7.1. Piece propagation at machines used to process the log 115 7.2. Movement of sideboard ( T r a n s a c t i o n #4) between Headrig 8' and Combined Machine #2 120 7.3. Comparison of a c t u a l and FLOWSIM piece s i z e s 122 8.1. Summary of market runs with simulated time of one s h i f t (460 minutes) 147 8.2. Condensed r e s u l t s of breakdown runs 153 8.3. P r o d u c t i o n l o s s e f f e c t of Combined Machine # 1 breakdown 155 9.1. Ranking of the s i x markets 157 Note: Tables are numbered i n accordance with Chapter numbers. E.g., Table 3.1 i s the f i r s t Table of Chapter 3. ix L i s t of F i g u r e s 3.1. Comparison of two l o g breakdown s i m u l a t o r by SAWSIM p l o t s 18 3.2. Flowchart of the s i m u l a t i o n procedure 23 5.1. Taper v a r i a t i o n along the length of stems 43 5.2. Flowchart of sawlog s e l e c t i o n procedure 50 5.3. Sawing i n s t r u c t i o n s i n the f i n a l boom l o g t a b l e 54 5.4. M i l l layout of the proposed m o d i f i c a t i o n plan ..61 6.1. Exogenous log breakdown l o g i c r e q u i r i n g data t r a n s m i s s i o n 70 6.2. SAWSIM output using machine codes 73 6.3. Network of p i e c e routes 74 6.4. SAWSIM output using route codes i n s t e a d of machine codes 76 6.5. DIA matrix of log U2/1 77 6.6. General DIA matrix s t r u c t u r e 80 6.7. S t r u c t u r e of Data Base #2 81 6.8. A l g o r i t h m of sawlog and the corresponding SAWSIM code s e l e c t i o n process 93 6.9. E x t r a c t of Data Base #1 94 6.10. T r a n s p o r t a t i o n time i s a f f e c t e d by p i e c e l e n g t h 98 6.11. Four d i f f e r e n t ways of a s s i g n i n g p i e c e width 101 7.1. SAWSIM output to v a l i d a t e FLOWSIM r e s u l t s 113 7.2. P r o c e s s i n g a s i n g l e l o g by FLOWSIM to compare with SAWSIM r e s u l t s 114 7.3. Correspondence between t r a n s a c t i o n s and p i e c e s 117 7.4. Time ch a r t of l o g and piec e p r o c e s s i n g 119 X 7.5. Log used to check machine u t i l i z a t i o n s t a t i s t i c s . . . 1 2 3 7.6. Predetermined breakdown events and the e q u i v a l e n t GPSS program segments causing these events to happen 124 7.7. D e t e r m i n i s t i c FLOWSIM run to check machine u t i l i z a t i o n s t a t i s t i c s ...125 8.1. S t r u c t u r e of i n t e r a c t i n g programs 129 8.2. Condensed FLOWSIM output 133 8.3. Production r e s u l t s of SAWRES '..142 8.4. Snap shots of t r a n s p o r t a t i o n equipment pie c e counts at d i f f e r e n t s i m u l a t i o n c l o c k times ...148 8.5. Use of Twin Saw to produce 4x4 lumber 150 8.6. The three bucking p o l i c i e s r e s u l t in d i f f e r e n t sawlog l e n g t h d i s t r i b u t i o n s 151 8.7. GPSS segment of machinery breakdown s e n s i t i v i t y run 154 9.1. Computer usage s t a t i s t i c s of a t y p i c a l FLOWSIM run 1 60 Note: F i g u r e s are numbered in accordance with Chapter numbers. E.g., F i g u r e 9.1 i s the f i r s t F i g u r e of Chapter 9. x i ACKNOWLEDGEMENT I wish to acknowledge the f i n a n c i a l support of the U n i v e r s i t y of B r i t i s h Columbia and MacMillan B l o e d e l L t d . , p r o v i d i n g f e l l o w s h i p s , and the resource support of the F a c u l t y of F o r e s t r y and the Computer Centre f o r the s u b s t a n t i a l l y extended computer usage. T h i s study would not have been p o s s i b l e without the h e l p of Howard Leach who made h i s sawing s i m u l a t i o n package (SAWSIM) a v a i l a b l e , shared h i s v a l u a b l e time and gave much u s e f u l a d v i c e . A s p e c i a l note of thanks to Jan Aune and C h r i s Boniface at MacMillan B l o e d e l L t d . , f o r t h e i r c o - o p e r a t i o n and guidance. Log measurement data p r o v i d e d by FORINTEK Canada Corp., Western Laboratory was i n v a l u a b l e . A p p r e c i a t i o n i s a l s o due to John Emanuel, A n t a l Kozak, L a s z l o Paszner, Har.ry Smith, Willem Vaessen, Douglas W i l l i a m s , Barry Wong, and Glendon Young who c o n t r i b u t e d t h e i r p r o f e s s i o n a l s k i l l s . 1 J_j_ INTRODUCTION The lumber i n d u s t r y p l a y s an important economic r o l e in B r i t i s h Columbia. The p r o v i n c e i s the l a r g e s t lumber manufacturer i n Canada, producing 63% of a l l lumber in 1982. The number of employees working i n B r i t i s h Columbia sawmills was 65,000-70,000 duri n g the l a s t decade. Sawmills employed 70% of the labour f o r c e i n the f o r e s t products i n d u s t r y . T h i s i n d u s t r y was the primary generator of employment in B r i t i s h Columbia (77). T h e r e f o r e , the lumber i n d u s t r y has a great impact on the whole p r o v i n c i a l economy. The p r o p o r t i o n of lumber exported during the past 15 years has v a r i e d between 70 and 80%. The share of the U.S.A. market, in comparison to other export markets, has a l s o been high. During the l a s t two decades t h i s share was 53-73% (77). T h i s heavy r e l i a n c e on the U.S.A. market leaves the p r o v i n c i a l lumber producers s u b s t a n t i a l l y exposed to d i f f e r e n t p r e s s u r e s . The c y c l i c a l l y changing demand i n U.S.A. lumber markets, c o r r e l a t e s h i g h l y with v a r i a t i o n s i n the housing i n d u s t r y , and has been causing severe s e l l i n g d i f f i c u l t i e s . Compounding the problem i s the competition from southern yellow pine producers as more plantation-grown southern pine reaches c u t t i n g age (86). With i n c r e a s i n g age the q u a l i t y and s e l e c t i o n of lumber from t h i s source i s improving (86). G e o g r a p h i c a l l y , B r i t i s h Columbia i n c l u d e s C o a s t a l and 2 I n t e r i o r lumber producing r e g i o n s . While t r a d i t i o n a l l y the coast m i l l s dominated lumber exports from Western Canada, c o n s i d e r a b l e competition has developed from i n t e r i o r producers due to t h e i r lower l o g g i n g c o s t s and much higher p r o d u c t i v i t y than i s p o s s i b l e i n a n t i q u a t e d m i l l s on the West Coast. T h i s change in export p o t e n t i a l f o r Canadian lumber has caused a severe s t r a i n i n the p r o f i t a b i l i t y and growth of the western lumber producing sawmills i n recent y e a r s . C o a s t a l m i l l s are f a c i n g some tough d e c i s i o n s with respect to market d i v e r s i f i c a t i o n and advancement in sawmill technology. A quote from the COFI ( C o u n c i l of F o r e s t I n d u s t r i e s of B i t i s h Columbia) report (54) submitted to the Royal Commission on Economic Prospects in 1983, summarizes the best f u t u r e s t r a t e g i e s to make the coast m i l l s p r o f i t a b l e a g a i n : "1. Maximize recovery of, and demand f o r , higher value upper grades and pursue those markets and end uses that o f f e r the best long term volume p o t e n t i a l . 2. Develop and market a new l i n e of higher value s p e c i a l t y products. 3. Concentrate more on overseas markets where they can compete best (e.g. Japan and North A f r i c a ) with t h e i r c o n s t r u c t i o n grades. 4. Take a p p r o p r i a t e a c t i o n to enhance the value of lumber products by making them more s u i t e d f o r the market and intended use." Indeed, a c c o r d i n g to the recent trends of the i n d u s t r y , the key r e v i t a l i z a t i o n of the c o m p e t i t i v e p o s i t i o n of c o a s t a l m i l l s appears to be i n production of " s e l e c t e d s t r u c t u r a l " lumber grades. Sales have to take i n t o account the needs of d i f f e r e n t overseas markets, while c o n t i n u i n g to compete in domestic and U.S.A. markets with lower grades. Changed marketing s t r a t e g i e s w i l l r e q u i r e a complete 3 modernization of the c o a s t a l m i l l s . The o f t e n s p e c i a l i z e d lumber s i z e s r e q u i r e w e l l d e f i n e d c u t t i n g s t r a t e g i e s , a degree of p r o d u c t i o n f l e x i b i l i t y and automation not a v a i l a b l e i n o l d c o a s t a l m i l l s at the present time. Large investments i n m i l l r e n ovation w i l l be i n e v i t a b l e . Before any f i n a n c i a l commitment, computer s i m u l a t i o n as an o b j e c t i v e design t o o l , a p p r o p r i a t e l y combined with the experience of sawmill s p e c i a l i s t s , w i l l be r e q u i r e d to evaluate a l t e r n a t i v e s i n the complex redesign process of o l d c o a s t a l m i l l s . The present t h e s i s was conceived with these c o n s i d e r a t i o n s and o p p o r t u n i t i e s i n mind. The study analyzes a change-over i n one of MacMillan B l o e d e l Ltd.'s sawmills with the concept of p r o d u c t i o n f l e x i b i l i t y i n accordance with v a r y i n g market demands. The sawmill s e l e c t e d was Woodroom #3 at Harmac, B r i t i s h Columbia, a l a r g e - l o g m u l t i p a s s - h e a d r i g m i l l . Overseas products of s p e c i a l s i z e s , corresponding c u t t i n g p a t t e r n s , p i e c e flow and p r o d u c t i o n w i l l be evaluated through a h y b r i d s i m u l a t i o n model of SAWSIM (property of H.A. Leach and Company L t d . ) , a s t a t i c model, and FLOWSIM, a dynamic model, developed s p e c i f i c a l l y f o r Woodroom #3 as p a r t of t h i s t h e s i s . During the years of 1975-83, the o p e r a t i o n of Woodroom #3 became uneconomical. The need f o r r e - d e s i g n i n g the e x i s t i n g sawmill was r e c o g n i z e d and a plan f o r m o d i f i c a t i o n was submitted by the management. While q u e s t i o n s of o p e r a t i n g e f f i c i e n c y and p r o f i t a b i l i t y c o u l d be answered by t r i a l and e r r o r i n time, a computer s i m u l a t i o n technique was deemed to be more e f f i c i e n t i n p r e d i c t i n g and e v a l u a t i n g the behaviour of the proposed new m i l l d e s i g n . 4 O r i g i n a l l y , Woodroom #3 was c o n s t r u c t e d to produce dimension lumber f o r the United S t a t e s market. Major reasons for the uneconomical o p e r a t i o n of the m i l l were as f o l l o w s : • The c o s t of e x t r a c t i n g logs on the Coast i s twice that of I n t e r i o r l o g s . • Competition from an i n c r e a s i n g number of I n t e r i o r dimension producers. • Decreasing markets because of g r e a t l y reduced housing s t a r t s i n the U.S.A du r i n g the l a s t f i v e y e a r s . • Disadvantageous g e o g r a p h i c a l circumstances of h a r v e s t i n g on the c o a s t : high e l e v a t i o n , steep slopes on mountainsides and hence high cost of road c o n s t r u c t i o n . • Expensive sawmill manpower - consequently high c o n v e r s i o n c o s t . The b a s i c idea of the m o d i f i c a t i o n was to c r e a t e products y i e l d i n g higher s e l l i n g p r i c e s . Lumber markets f o r overseas c o u n t r i e s appeared to be a t t r a c t i v e . The demand f o r v a r i o u s small c u t t i n g markets i s d i f f i c u l t to p r e d i c t . T h i s i s due to the extremely v a r i a b l e nature of economic s t r e n g t h of buyer and producer c o u n t r i e s ; the r e l a t i v e buying power of a p a r t i c u l a r market's currency to the currency of the producer; and supply c o m p e t i t i o n from c o u n t r i e s such as Sweden, Russia and the United S t a t e s (50). T h e r e f o r e , the key concept i n d e s i g n i n g the new 5 m i l l i s market f l e x i b i l i t y , i . e . , the m i l l i s being designed bearing i n mind that s e l l i n g o p p o r t u n i t i e s i n overseas markets are h i g h l y v o l a t i l e and v a r i a b l e . Thus the m i l l i s planned to be capable of producing up to 70% of i t s p r o d u c t i o n i n the form of small c u t t i n g s f o r export, with the balance as dimension lumber as one extreme of product mix, and 100% dimension lumber, as the other extreme of p r o d u c t i o n . The most v i a b l e export small c u t t i n g markets with the corresponding lumber t h i c k n e s s e s a r e : 1 9/16" Un i t e d S t a t e s dimension, 1 13/16" Japan, 1 7/8" U n i t e d Kingdom and A u s t r a l i a , 2 1/2" North A f r i c a and Belgium, 3" France and the United Kingdom, 4 1/8"X4 1/8" Japan ( l a t e r r e f e r r e d to as 4x4) A c l o s e i n t e r r e l a t i o n s h i p between the above lumber s i z e s and technology, as w e l l as the search f o r a l e s s expensive c o n v e r s i o n process, n e c e s s i t a t e m o d i f i c a t i o n of the c u r r e n t m i l l operat i o n . Conversion f a c i l i t i e s used f o r the primary breakdown of lo g s , up to and i n c l u d i n g the head r i g s , w i l l remain the same. It i s planned that a l l of the m i l l machinery from the headrigs to the back end of the m i l l , w i l l be removed and r e p l a c e d by new machinery. A d e t a i l e d d e s c r i p t i o n and layout of the mod i f i e d m i l l are presented in a l a t e r s e c t i o n . A c c o r d i n g l y , the proposed m i l l o p e r a t i o n had to be eva l u a t e d with s p e c i a l a t t e n t i o n to the redesigned p a r t of the 6 conversion p r o c e s s . Given the log mix, with the new machinery and technology as c o n s t r a i n t s , performance of the r e v i s e d sawmill had to be analyzed as to i t s response to v a r i o u s c u t t i n g p a t t e r n s r e q u i r e d f o r p r o d u c t i o n of export and dimensional lumber. T h i s turns out to be an extremely complex task. The raw m a t e r i a l i s c h a r a c t e r i z e d by randomly changing l o g shape. There are a l s o s t o c h a s t i c breakdowns of i n t e r a c t i n g machine c e n t r e s , and t r a n s p o r t a t i o n equipment of c r o s s c h a i n conveyors, b e l t s and r o l l c ases. D i f f e r e n t c u t t i n g p a t t e r n s , d i c t a t e d by the v a r i e t y of f i n a l product s i z e s f o r v a r i o u s markets, make the system of the sawmill very complicated. The p e r c e i v e d complexity of modern sa w m i l l i n g f o r the proposed m i l l design, and the high cost ($ 28 m i l l i o n ) of t o t a l investment, were the major reasons to use computer s i m u l a t i o n as a technique f o r p r e l i m i n a r y e v a l u a t i o n of i t s performance. In g e n e r a l , the p r o d u c t i o n e f f i c i e n c y i n a sawmill depends on l o g throughput and the way of l o g breakdown. T h e r e f o r e , i f the o p e r a t i o n of a m i l l i s to be simulated a c c u r a t e l y , the model must have a l o g breakdown l o g i c and must be dynamic. Dynamic i n the sense that i t simulates p i e c e flow as a f u n c t i o n of time. A drawback of e x i s t i n g dynamic models of sawmill s i m u l a t i o n i s that t h e i r l o g breakdown l o g i c i s l a r g e l y s i m p l i f i e d , i . e . , they c o n s i d e r the l o g as a t r u n c a t e d cone. The disadvantageous consequence of t h i s i s not only that i n a c c u r a t e lumber recovery estimates r e s u l t but a l s o that p i e c e flow s i m u l a t i o n becomes i n a c c u r a t e . To ensure accurate s i m u l a t i o n , the dynamic model developed in t h i s t h e s i s p r o j e c t has access to a h i g h l y s o p h i s t i c a t e d l o g 7 breakdown model. The i n t e n t i o n i s to simulate the a c t u a l m i l l behaviour more a c c u r a t e l y than other models d i d b e f o r e . Thereby, the model can be used to analyze and c o o r d i n a t e the m i l l o p e r a t i o n f o r p r o d u c t i o n f l e x i b i l i t y . The dynamic model i s c o n s t r u c t e d i n GPSS (General Purpose S i m u l a t i o n System) language and processed by the GPSS/H compiler. (GPSS/H i s the most e f f e c t i v e compiler f o r the GPSS language and was developed by J.O. Henriksen of Wolverine Software C o r p o r a t i o n ( 6 2 ) ) . T h i s dynamic model w i l l be r e f e r r e d to throughout the te x t as FLOWSIM (piece FLOW S I M u l a t i o n ) . SAWSIM (SAWmill SIMulation) i s one of the most comprehensive sawing s i m u l a t i o n packages (25). I t w i l l be used to inform FLOWSIM how the l o g i s sawn. In other words, i t serves as the l o g breakdown l o g i c of the dynamic model. An examination of the computer s i m u l a t i o n l i t e r a t u r e showed two commonly mentioned disadvantages of t h i s technique: 1. b u i l d i n g a s i m u l a t i o n model i s time demanding for the a n a l y s t and 2. running a s i m u l a t i o n model r e q u i r e s a l a r g e amount of computer time. Bearing i n mind the above two disadvantages of modelling i n g e n e r a l , and i t s a p p l i c a t i o n in sawmill design e v a l u a t i o n i n p a r t i c u l a r , one of the purposes of t h i s t h e s i s i s to r e v e a l the p o t e n t i a l s of a new sawmill design and e v a l u a t i o n technique. The a p p l i c a t i o n of GPSS/H in c o n s t r u c t i n g the dynamic model, and the usage of SAWSIM, as i t s a l r e a d y e x i s t i n g log breakdown l o g i c , were thought to l e s s e n programing e f f o r t and to improve the powers of sawmill s i m u l a t i o n . If sawmill modelling 8 p r i n c i p l e s brought toge ther i n t h i s t h e s i s serve as s t a r t i n g s teps of a c l o s e r e v a l u a t i o n of s awmi l l d e s i g n s , and i f the new model b u i l d i n g concepts would be accepted i n s awmi l l a n a l y s i s , the purpose of w r i t i n g t h i s t h e s i s would be f u l f i l l e d . The major a spec t s of r e d e s i g n i n g Woodroom #3, the l i m i t a t i o n s of the a n a l y s i s , the t echnique to be used to e v a l u a t e the m o d i f i c a t i o n p l a n , and the purposes of w r i t i n g t h i s t h e s i s were d i s c u s s e d i n the p r e v i o u s paragraphs . Based on t h i s i n t r o d u c t i o n Chapter 2 e n t e r s i n t o p a r t i c u l a r s and p r o v i d e s the d e t a i l s of the o b j e c t i v e s of t h i s r e s e a r c h . Chapter 3 dea l s wi th q u e s t i o n s of why computer s i m u l a t i o n was employed to a t t a c k the problem and why the p a r t i c u l a r model s t r u c t u r e i s a p p l i e d . Des ign of exper iments for the computer s i m u l a t i o n model i s a l s o d e s c r i b e d i n t h i s c h a p t e r . Chapter 4 g i v e s a l i t e r a t u r e review of the two major groups of computer s i m u l a t i o n a p p l i c a t i o n s i n s a w m i11 i n g . Chapter 5 surveys the major data groups of l o g s , machinery , t echno logy and f i n a l p r o d u c t s . These data are e i t h e r i n p u t s of the model or b u i l t i n t o the model . Chapter 6 d e s c r i b e s the major concepts of model b u i l d i n g , the i n t e r a c t i o n s among v a r i o u s p a r t s of the model , the a p p l i c a t i o n of SAWSIM as the l o g breakdown l o g i c of the dynamic model , the implementat ion of data t r a n s m i s s i o n between SAWSIM and the dynamic model . V a l i d a t i o n , the degree of accuracy to which the model s i m u l a t e s m i l l o p e r a t i o n , i s o u t l i n e d i n Chapter 7. Chapter s 8 and 9 d i s c u s s d e t a i l s of the r e s u l t s and c o n c l u s i o n s . 9 Chapter 10 l i s t s the l i t e r a t u r e r e f e r e n c e d throughout the t h e s i s . Computer s i m u l a t i o n l i t e r a t u r e i s grouped s e p a r a t e l y . 10 2_;_ OBJECTIVES In the p r e v i o u s chapter some reasons f o r the n o n - p r o f i t a b l e m i l l o p e r a t i o n of Woodroom #3 were e s t a b l i s h e d . The m o d i f i c a t i o n plans which were intended to put the m i l l revenue and expenditures i n balance, were a l s o d i s c u s s e d . The t o t a l cost of the recommended r e c o n s t r u c t i o n i s h i g h : $ 28 m i l l i o n . Hence the d e s i r e of management to decrease the r i s k of investment i s understandable. Before the f i n a n c i a l commitments are made, the management would l i k e answers to the f o l l o w i n g q u e s t i o n s : • How w i l l the redesigned m i l l react to d i f f e r e n t c u t t i n g s t r a t e g i e s ? • W i l l the p i e c e flow provide the r e q u i r e d l o g throughput? • W i l l the machine u t i l i z a t i o n be s a t i s f a c t o r y and can i t be f u r t h e r improved? What are the recommended m o d i f i c a t i o n s ? • How s e n s i t i v e i s the m i l l to a n t i c i p a t e d machine r e l i a b i 1 i t y ? A c c o r d i n g to the above qu e s t i o n s the r e s e a r c h o b j e c t i v e s are the f o l l o w i n g : 1. To examine how the m i l l w i l l react to d i f f e r e n t c u t t i n g programs. More s p e c i f i c a l l y , to evaluate p r o d u c t i v i t y when the m i l l produces f o r one of the s i x p o s s i b l e lumber markets: U.S.A. dimension, North A f r i c a , U n ited Kingdom, France, and the two Japanese markets, 4x4 and lumber f o r remanufacturing purposes. The e v a l u a t i o n of pr o d u c t i o n p r e d i c t s the volume and value of lumber p r o d u c t i o n and r e c o v e r i e s , estimates raw m a t e r i a l requirements, and determines the machine and t r a n s p o r t a t i o n equipment piece counts. 2 . To eva l u a t e how long and with what frequency v a r i o u s machines w i l l have to wait because downstream t r a n s p o r t a t i o n equipment i s f u l l y occupied and to recommend changes i n the design capable of e l i m i n a t i n g these blocked s t a t e s and improving machine u t i l i z a t i o n . 3 . To analyze the s e n s i t i v i t y of the flow plan to machinery breakdowns. 4. To b u i l d a s i m u l a t i o n model which enables i t s " user to c a r r y out experiments i n c o n j u n c t i o n with the fo r e g o i n g quest i o n s . 5. To w r i t e a u x i l i a r y programs • supp o r t i n g the data t r a n s m i s s i o n between SAWSIM and FLOWSIM the dynamic model, • to s e l e c t sample sawlogs bucked a c c o r d i n g to a given bucking p o l i c y from the company boom l o g d i s t r i b u t i o n . 6. To recommend methods of computer s i m u l a t i o n found to be a p p l i c a b l e i n sawmill design and a n a l y s i s i n the view of the new model b u i l d i n g approach taken i n t h i s r e s e a r c h . 1 2 3_;_ METHODS AND PROCEDURES In t h i s chapter, reasons are d i s c u s s e d why computer s i m u l a t i o n i s employed to analyze the m o d i f i c a t i o n redesign of Woodroom #3, why e x i s t i n g dynamic s i m u l a t i o n models should be improved and why the SAWSIM-FLOWSIM model combination i s used. Major phases of the s i m u l a t i o n procedure and the design of computer s i m u l a t i o n experiments are a l s o d i s c u s s e d . 3.1. REASONS FOR COMPUTER SIMULATION In g e n e r a l , there are three p o s s i b l e approaches to e v a l u a t i n g or p r e d i c t i n g the o p e r a t i o n of v a r i o u s systems: • performing experiments on the a c t u a l system, • f o r m u l a t i n g and s o l v i n g an a n a l y t i c a l model as r e p r e s e n t a t i v e of the r e a l system, • b u i l d i n g and experimenting on a s i m u l a t i o n model of the r e a l system. C o n c e n t r a t i n g only on sawmill systems, the high c o s t of sawmill c o n s t r u c t i o n and o p e r a t i o n excludes the f i r s t approach. Besides, s t o c h a s t i c l og shape c h a r a c t e r i s t i c s make i t impossible to repeat experiments under the same circumstances on the e f f e c t of d i f f e r e n t ways of sawing which are the most s e n s i t i v e f a c t o r s i n f l u e n c i n g lumber y i e l d . A n a l y t i c a l models are based on the u n d e r l y i n g s t r u c t u r e of 1 3 the system in the form of mathematical equations. A v a i l a b l e systematic methods to s o l v e them, and the r e l a t i v e l y low time-demand to b u i l d and analyze them, make t h i s approach s u p e r i o r to the other two, i f mathematical modeling i s p o s s i b l e . U n f o r t u n a t e l y , changing l o g mix, product demand of v a r i o u s markets, wide s e l e c t i o n of machinery, random components of l o g geometry and machinery breakdowns make sawmill systems too complex to approach a n a l y t i c a l l y . Hence, t h i s study r e s o r t s to the use of computer s i m u l a t i o n techniques. Computer s i m u l a t i o n i s one of the most widely used techniques to explore complex systems. The approach uses preconceived models in the form of computer programs. In other words, a computer i s programmed to simulate the i n t e r a c t i o n s among system elements. The p o p u l a r i t y of s i m u l a t i o n i s c o n s t a n t l y growing in sawmill a p p l i c a t i o n s . A q u o t a t i o n from Reynolds (35) e x p l a i n s why: "By using a computer to simulate the sawing, we have de v i s e d a way to saw the same l o g i n many d i f f e r e n t ways to f i n d out which way gets the best y i e l d from the l o g . Sawmill production managers o f t e n f i n d i t hard to t e l l what sawing p a t t e r n w i l l y i e l d the most. If sample logs are sawed in one sawing p a t t e r n , they can not be reassembled and sawed again i n a d i f f e r e n t p a t t e r n . And i f a new sample of l o g s i s used fo r each new sawing p a t t e r n , i t i s hard to t e l l i f d i f f e r e n c e s in y i e l d are due to the sawing p a t t e r n s or are due to n a t u r a l d i f f e r e n c e s between the logs i n the samples. But by using a computer to simulate the sawing, the same log sample can be sawed again and again, and y i e l d s of the d i f f e r e n t sawing p a t t e r n s can be compared." Notice that the above q u o t a t i o n from Reynolds g i v e s a r a t i o n a l e f o r computer s i m u l a t i o n a p p l i c a t i o n regarding only the l o g breakdown aspects of s a w m i l l i n g . E v a l u a t i o n of sawmill 1 4 behaviour r e q u i r e s f u r t h e r a n a l y s i s - and hence a p p l i c a t i o n of s i m u l a t i o n techniques - on m i l l dynamics l i k e machine and t r a n s p o r t a t i o n equipment i n t e r a c t i o n s , p o s s i b l e b o t t l e n e c k s of pi e c e flow caused by i n a p p r o p r i a t e speeds of machine and/or t r a n s p o r t a t i o n equipment, machinery breakdown times and fr e q u e n c i e s , and inadequate surge a r e a s . C o n s i d e r i n g the above f a c t s , t h i s study employs the technique of computer s i m u l a t i o n . Employment of computer s i m u l a t i o n models i s not new i n sawmill a n a l y s i s . The method of c o n s t r u c t i n g the model of t h i s study makes i t d i f f e r e n t from other a l r e a d y e x i s t i n g models and i s b e l i e v e d to improve s i m u l a t i o n accuracy. How t h i s new model provides improved ways of sawmill a n a l y s i s i s d e s c r i b e d in the next s e c t i o n . 3.2. WHY EXISTING DYNAMIC MODELS SHOULD BE IMPROVED Log breakdown and l o g throughput, as two major f a c t o r s a f f e c t i n g m i l l behaviour, must be d e a l t with i n any sawmill model i f one wants to simulate m i l l o p e r a t i o n . The e f f e c t of the f i r s t f a c t o r i s s t r a i g h t f o r w a r d ; the higher the lumber value recovery, the higher the revenue. The second f a c t o r - i n the case of a not w e l l organized m i l l - e x e r c i s e s i t s e f f e c t through poor machine time u t i l i z a t i o n caused by congestion i n b o t t l e n e c k s or machine breakdowns. Conversely, high throughput can s i g n i f i c a n t l y a f f e c t p r o f i t a b i l i t y of the m i l l through higher p r o d u c t i v i t y . The d e s i r e to improve m i l l performance by o p t i m i z a t i o n of the above two f a c t o r s has been the su b j e c t of sawmill system 1 5 a n a l y s i s f o r a long time. T h i s can be seen from the l a r g e number of computer s i m u l a t i o n a p p l i c a t i o n s in sawmilling (see Chapter 10.1). Among these a p p l i c a t i o n s , however, much more a t t e n t i o n has been p a i d to the lumber recovery issue (17,25,41,42,43) than to m i l l dynamics (6,36,65). There are two reasons f o r t h i s . F i r s t , the e f f e c t of l o g breakdown on m i l l performance i s grea t e r than that of m i l l dynamics. Thus r e s e a r c h on l o g breakdown s i m u l a t i o n had p r i o r i t y over r e s e a r c h on dynamic s i m u l a t i o n . The second reason i s the f o l l o w i n g . Models of log breakdown c a r r y out repeated g e o m e t r i c a l c a l c u l a t i o n s about the r e l a t i v e p o s i t i o n s of r e c t a n g u l a r p i e c e s of lumber i n t o which logs can be s u b d i v i d e d . A l a r g e amount of c a l c u l a t i o n i s needed for t h i s purpose although the scope of the undertaking i s r e l a t i v e l y s m a l l ; knowledge of geometry, sawmill technology and a computer language are s u f f i c i e n t . On the other hand, dynamic models r e q u i r e a d d i t i o n a l knowledge of s i m u l a t i o n techniques and s t a t i s t i c a l methods. A s p e c i a l purpose s i m u l a t i o n language, which makes the a n a l y s i s of s t o c h a s t i c systems e a s i e r , might a l s o be r e q u i r e d . Development of s i m u l a t i o n techniques and languages s u i t a b l e f o r analyses of d i s c r e t e , s t o c h a s t i c systems i s r e l a t i v e l y new. Consequently, a p p l i c a t i o n s of dynamic models in s a w m i l l i n g were int r o d u c e d l a t e r and they s t i l l have weaknesses and need improvement. One of these i s the accuracy of t h e i r l o g breakdown l o g i c . Piece flow i n a m i l l depends to a l a r g e degree on the way logs are sawn. The r e f o r e , i f we want to simulate p i e c e flow a c c u r a t e l y , dynamic models must have as accurate l o g breakdown 16 l o g i c as p o s s i b l e . A drawback of e x i s t i n g dynamic models of sawmill s i m u l a t i o n i s that t h e i r log breakdown l o g i c i s f a i r l y s i m p l i f i e d ; they c o n s i d e r the l o g as a t r u n c a t e d cone(6,36). In other words, the only l o g shape c h a r a c t e r i s t i c s c o n s i d e r e d are diameter, uniform taper and l e n g t h . The consequence of t h i s i s not only that i n a c c u r a t e lumber recovery c a l c u l a t i o n i s obtained but a l s o that the s i m u l a t i o n of the dynamic behaviour of the m i l l i s i n a c c u r a t e . Accuracy in log breakdown s i m u l a t i o n i s of utmost importance. A small e r r o r in the l o g breakdown l o g i c can l e a d to erroneous c o n c l u s i o n s r e g a r d i n g the dynamic behaviour of a m i l l . To i l l u s t r a t e the importance of t h i s , c o n s i d e r the f o l l o w i n g example. A western hemlock l o g i s c r o s s cut i n t o three sawlogs. These three sawlogs then are "sawn" by two d i f f e r e n t l o g breakdown s i m u l a t o r s . The f i r s t one c o n s i d e r s the l o g as a t r u n c a t e d cone, i . e . , i t takes i n t o account only top and butt end diameter and l o g l e n g t h . The second s i m u l a t o r , in a d d i t i o n to these three shape c h a r a c t e r i s t i c s , takes a l s o i n t o c o n s i d e r a t i o n sweep, v a r i a b l e taper along the l e n g t h , o v a l i t y and t w i s t . N o t i c e that e x a c t l y . t h e same l o g was sawn by the two s i m u l a t o r s . Table 3.1 and F i g u r e 3.1 compare the two cases. The l e s s a c curate s i m u l a t o r over or underestimates the number of lumber p i e c e s f o r a l l three sawlogs. C o n s i d e r i n g an average throughput of 1 boom log per 2 minutes, a dynamic model, having l e s s a c curate log breakdown l o g i c , c o u l d over or underestimate the number of p i e c e s flowing through the m i l l by 960 p i e c e s per s h i f t . Thus, i t c o u l d not represent the r e a l Table 3.1. Comparison of two log breakdown s i m u l a t o r s Boom lo g B228 S a w l o g B228/1 B228/2 B228/3 Case#1 ( S i m p l i f i e d 32 20 14 l o g shape) Case#2 (Accurate shape 33 18 13 c h a r a c t e -r i s t i c s ) m i l l dynamics. In a d d i t i o n , the d i f f e r e n c e i n lumber p i e c e s does not t e l l the whole"story of f a u l t y dynamic s i m u l a t i o n . For i n s t a n c e , the d i f f e r e n c e i s only one (33 vs. 32) in the case of sawlog B228/1. However, i t i s worth a n a l y s i n g in d e t a i l how f a l s e p i e c e flow p r e d i c t i o n s might be made based on the l e s s accurate l o g breakdown s i m u l a t o r . F i r s t , the number of sideboards f l o w i n g from headrig to edger would be overestimated by 1 (5 vs. 4). Second, the number of p i e c e s flowing from edger to trimmer would a l s o be overestimated by 1 (8 vs. 7). T h i r d , the number of p i e c e s produced from the center cant d i f f e r by 2 (26 vs. 24). Fourth, n o t i c e how the s i z e s of some p i e c e s d i f f e r i n the two cases. T r a n s p o r t a t i o n equipment occupancy c a l c u l a t i o n s , based on these erroneous s i z e s , would a l s o be f a l s e . S i m i l a r comparisons can be made in c o n j u n c t i o n with sawlog B228/2 and B228/3. These are the reasons why the dynamic model employed in t h i s r e s e a r c h t r i e s to use more acc u r a t e l o g breakdown l o g i c than others have done. However, to b u i l d an accurate l o g 18 SAt rs i t i 'LOT • 2 2 1 / 1 U H M 2 l 1 0 2 • *• * • • • • | 2 1 0 4 12 II — 1 2 1. 2 X 0 . 1 2 l l 12 1 2 I 0 C 1 2 II « . 1 2 | 3 «Ot 1 2 1 2 2 X 0 . 1 2 | | 2 X 0 * 1 2 2 X 0 . 1 2 | | 2 , 0 . 1 2 2 > 0 C 1 2 II 2 . 0 . 1 2 2 1 0 * 1 2 | | 2 . 0 , 1 2 2X0* 1 2 II 2 » 0 » 1 2 2>0« 1 2 II"" 1 2 2 X 0 * 1 2 II 2 * 0 * 1 2 2 x 1 0 2 CASE #1 - Number of lumber p i e c e s i s 32 SA»SIX PLOT B228/ 1 UH3« • • 2 X 0 4 1 2 ' 1 2 X 0 . 1 2 I 2 X 0 . 1 2 2 1 0 . 1 2 2 X 0 . 1 2 2 X 0 . 1 2 2 1 0 . 1 2 2 1 0 . 12 2 1 0 . 1 2 2 X 0 . 1 2 2 1 0 . 1 2 2 1 0 . 1 2 2 < 0 « 1 2 2 X 0 . 1 2 2 1 0 . 1 2 2 < 0 * 1 2 1 2 2 » 0 * 1 2 | 2 X 0 . 1 2 2 X 0 * 1 2 I 2 X 0 . 1 2 —« 2 X 0 . 1 2 1 2 X 0 . 1 2 • • • 1 1 2 X 0 4 - 0 1 2 X 1 O O ' 4» • • I F i g u r e 3 . 1 . C A S E #2 - Number o f l u m b e r p i e c e s i s 33 C o m p a r i s o n o f t w o l o g b r e a k d o w n s i m u l a t o r b y SAWSIM p l o t s 19 • 2 2 1 / 2 U M 3 4 2 I 3 0 2 • • -2 X 0 * • | • 1 X 1 0 1 1 2 X 1 0 2 2 1 2 1 1 0 2 2 | > H 0 2 2 | 2 1 1 0 2 2 1 2 > 1 0 2 2 1 2 X 1 0 2 2 1 2 X 1 0 2 2 | 2 X 1 0 2 2 1 2 X 1 0 2 2 1 .'I 2 I 2 0 2 c 2 X i 0 0 | 2 X 0 1 2 2 CASE #1 - Number of lumber p i e c e s i s 20 S U S t d P L O T • 2 2 * / 2 U M 3 4 | - 2 X 0 » 2 X 0 * 2 2 2 X 1 0 2 2 2 X 1 0 2 2 2 X 1 0 2 2 2 X 1 0 2 2 2 X 1 0 2 2 2 X 1 0 2 2 2 X 1 0 2 2 2 X 1 0 2 2 ' 2 X 2 | » -o-o| - | 2 I 0 ( 2 2 2 I 2 0 O 2 - J.-2 1 1 C A S E #2 - Number o f l u m b e r p i e c e s i s 18 F i g u r e 3 . 1 . C o m p a r i s o n o f t w o l o g b r e a k d o w n s i m u l a t o r by SAWSIM p l o t s - c o n ' d 20 S ' V S I B P L O T • 2 2 * / 2 U M 3 2 2 X 0 * 1 0 • • 2 > 1 2 I I 1 2 1 1 2 1 1 | 2 X 1 2 I I | 2 X 1 2 1 1 | I 1 2 2 B 1 0 I • I 2 X 1 0 I I CASE #1 - Number of lumber p i e c e s i s 14 S . V S I P P L O T 0 2 2 1 / 1 U M 3 3 l " 2 1 * 2 • I 2 I O I 1 2 • • 1 C A S E #2 - Number o f l u m b e r p i e c e s i s 13 F i g u r e 3 . 1 . C o m p a r i s o n o f t w o l o g b r e a k d o w n s i m u l a t o r by SAWSIM p l o t s - c o n ' d 21 breakdown s i m u l a t o r i s very time demanding and consequently expensive. To overcome t h i s t i r i n g programing a c t i v i t y the dynamic model uses SAWSIM, which i s one of the most s o p h i s t i c a t e d l o g breakdown models i n e x i s t e n c e today. SAWSIM and the problem of i t s a p p l i c a t i o n as an exogenous source of pi e c e flow s i m u l a t i o n w i l l be d i s c u s s e d l a t e r i n d e t a i l . As mentioned e a r l i e r , both b u i l d i n g and running a s i m u l a t i o n model are time demanding a c t i v i t i e s . As a r e c o g n i t i o n of these disadvantages, a great amount of e f f o r t has been expended to reduce the time consuming process of programming and consequently improve the cost e f f e c t i v e n e s s of model b u i l d i n g . One of the r e s u l t s of these e f f o r t s i s the GPSS/H compiler i t s e l f . T h i s new v e r s i o n of the GPSS f a m i l y helps to decrease the l a b o u r - i n t e n s i v e a c t i v i t y of model development. For in s t a n c e , i t s general-purpose input-output c a p a b i l i t i e s h elp the a n a l y s t produce s p e c i a l i z e d r e p o r t s ; the a n a l y s t can communicate with the o u t s i d e world e a s i l y , e.g., he can gain access to r o u t i n e s w r i t t e n i n FORTRAN; a great amount of time can be saved by i t s debugging f a c i l i t y , e t c . I t s s u p e r i o r run-time performance i s s i g n i f i c a n t from the computer economics p o i n t of view. "Models t y p i c a l l y compile 2.5 times f a s t e r with GPSS/H than with GPSSV. Some users have c o n s i s t e n t l y achieved 6:1 o v e r a l l improvements with t h e i r a p p l i c a t i o n s " , Vaessen (79). These s u p e r i o r c o m p i l a t i o n f e a t u r e s of GPSS/H can s u b s t a n t i a l l y decrease s i m u l a t i o n c o s t s . During the c o n s t r u c t i o n of my s i m u l a t i o n model the above advantages of GPSS/H were re c o g n i z e d . The time-saving and e f f o r t - l e s s e n i n g f e a t u r e s of GPSS/H were adapted to evaluate the 22 redesign of Woodroom #3. The usage of SAWSIM, as an a l r e a d y e x i s t i n g s o p h i s t i c a t e d s i m u l a t i o n package f o r log breakdown l o g i c of the dynamic model, was a l s o thought to l e s s e n programming e f f o r t and to improve the powers of the approach taken i n e v a l u a t i n g the m i l l . The f o r e g o i n g d i s c u s s i o n gave a r a t i o n a l e of why computer s i m u l a t i o n and why t h i s p a r t i c u l a r model composition of SAWSIM and FLOWSIM would be used i n t h i s r e s e a r c h . The f o l l o w i n g s e c t i o n s w i l l d i s c u s s the major phases of the s i m u l a t i o n procedure and t h e i r r e l a t i o n s h i p . 3.3. MAJOR PHASES OF THE SIMULATION PROCEDURE Th i s chapter g i v e s a gen e r a l o u t l i n e of the major a c t i v i t i e s to be done during the s i m u l a t i o n procedure. The r e l a t i o n s h i p s of d i f f e r e n t phases are d e p i c t e d by Fig u r e 3.2. Not i c e that v a r i o u s block shapes have d i f f e r e n t meaning i n the flowchart of F i g u r e 3.2. The upper p a r t of a block with o b l i q u e s i d e l i n e s c o n t a i n s the problem assignment of the s i m u l a t i o n phase i t belongs t o, whereas the lower part e x p l a i n s b r i e f l y how the problem i s sol v e d i n the s i m u l a t i o n procedure. Blocks with v e r t i c a l s i d e l i n e s are e i t h e r i n p u t s (shaded with l i n e s sloped down to the l e f t ) , i n t e r n a l r e s u l t s of s i m u l a t i o n phases (not shaded), or f i n a l s i m u l a t i o n r e s u l t s (shaded with l i n e s sloped down to the r i g h t ) . While reading the f o l l o w i n g paragraphs, there are two th i n g s to keep i n mind. F i r s t , u n d e r l i n e d t e x t in t h i s s e c t i o n always r e f e r s to p a r t i c u l a r blocks of the flowchart in F i g u r e 3.2. T h i s p r o v i d e s a s s i s t a n c e to keep tr a c k of what 23 (5. 1 > T . ( S O CROSSCUT BOOM LOGS INTO SAWLOGS ACCOROING TO/ BUCKING POLICY CALCULATE SAWLOG MEASUREMENTS ~ ^ r ~ — | SAWLOGS| (5.1) ECT SAMPLE SAWLOGS GROUP SAWLOGS OF SIMILAR SHAPE AND SE-LECT ONLY ONE FROM EACH GROUP <S.2)|255 (5.3)pfrPl (5.4) sm t (5.1) [SELECTED SAMPLE SAWLOGS}, SAW SAMPLE SAWLOGS MAKE SAWSIM RUNS 7 POOL OF SAMPLE SAWLOGS WITH THEIR SAWING INSTRUCTIONS (5.1) , — 2 . / REFORMAT SAWING INSTRUCTIONS / / WRITE SAWING INSTRUCTIONS / / INTO DIA MATRIX FORMAT / I 5 2 1 |01A MATRICES j(€.11 / ORGANIZE DIA MATRICES I N T O 7 / AN EFFICIENT DATA STRUCTURE / L WBITF i P^t»R4«:f TtR! r / (6.2) DATABASE '2 CONTAINING ALL DIA MATRIX ELEMENTS r SAVE SAWING RESULTS OF SAMPLE SAWLOGS SAVE BINARY INPUTS AND RESULTS OF SAWSIM RUNS  (6.3) S2-BINARY INPUT AND RESULT FILES 1 /CALCULATE FINAL 7 PRODUCTION FIGURES / RERUN SAWRES 1 >INA\>I RESULTS (6 21 7 _M FIND EOUIVALENT SAWLOGS / Y OF EACH BOOMLOG / / WRITE A DATABASE TABLE/ DATABASE «! CONTAINING ALL BOOM LOGS AND CORRESPONDING SAWLOGS (6 21 ^ ( 6 . 2 ) _S2_ REPEAT ACTIVITIES BELOW FOR A CERTAIN PERIOD OF TIME: - SELECT ONE BOOMLOG - FIND CORRESPONDING SAWLOGS AND THEIR DIA MATRICES - PROCESS SAWLOGS - COLLECT STATISTICS OF MILL BEHAVIOUR RUN FLOWSIM F L O W S I M R E S U L T S F i g u r e 3.2. Flowchart of the s i m u l a t i o n procedure 24 part i s being d i s c u s s e d w i t h i n the whole s i m u l a t i o n process. Second, the purpose of t h i s d i s c u s s i o n i s to give only a general overview about the s t r u c t u r e of i n t e r r e l a t e d phases of the s i m u l a t i o n p r o c e s s . D e t a i l e d d i s c u s s i o n of the components i s pro v i d e d l a t e r i n the corresponding chapters r e f e r r e d to by numbers in p a r e n t h e s i s beside b l o c k s . Given the d i s t r i b u t i o n of the a v a i l a b l e boom l o g mix in the log pond of the m i l l , they have to be c r o s s c u t a c c o r d i n g to the bucking p o l i c y set up by management. Log measurement data of SAWSIM input d e s c r i b e log shape c h a r a c t e r i s t i c s i n the form of data groups corresponding to l o g c r o s s s e c t i o n s . C r o s s c u t t i n g a boom l o g i n t o sawlogs produces a d d i t i o n a l c r o s s s e c t i o n s . If one wants to "saw" sawlogs s e p a r a t e l y by SAWSIM, these new c r o s s s e c t i o n s - as r e s u l t s of c r o s s c u t t i n g - must have log measurement data, too. These data are c a l c u l a t e d by using i n t e r p o l a t i o n . As a r e s u l t of these c a l c u l a t i o n s s i m u l a t i n g c r o s s c u t t i n g , the sawlogs and t h e i r l og measurement data c o n s t i t u t e the "raw m a t e r i a l " f o r f u r t h e r s i m u l a t i o n p r o c e s s . Because of the high computer cost (10-15 cents per sawlog as an average), not a l l of the sawlogs are to be sawn by SAWSIM. Instead, the sawlogs are grouped a c c o r d i n g to t h e i r l e n g t h and diameter. Assuming that sawlogs of the same group r e s u l t in approximately the same lumber y i e l d , only one sawlog w i l l be chosen f o r sawing as a r e p r e s e n t a t i v e of i t s group among the s e l e c t e d sample sawlogs. The group s i z e can be a l t e r e d f o r the r e q u i r e d s i m u l a t i o n accuracy. Taking i n t o account a d d i t i o n a l SAWSIM inputs of machine s p e c i f i c a t i o n s , sawing p a t t e r n s and f i n a l products, the s e l e c t e d 25 sawlogs are then sawn by SAWSIM, and put i n t o the pool of sample  sawlogs with t h e i r sawing i n s t r u c t i o n s . The purpose of sawing i n s t r u c t i o n s i s to command FLOWSIM to make p i e c e s flow through the dynamic model, s i m i l a r l y to the r e a l m i l l . These i n s t r u c t i o n s i n c l u d e the machines manufacturing a p a r t i c u l a r l o g , the p i e c e propagation of each machine ( i . e . , the number of a r r i v i n g vs. l e a v i n g p i e c e s ) , the addresses of the next machines, and the sequence of o p e r a t i o n s . These i n s t r u c t i o n s produced by SAWSIM are not i n the a p p r o p r i a t e format f o r FLOWSIM, hence they have to be reformatted i n t o dynamic i n f o r m a t i o n a r r a y s , DIA m a t r i c e s . Regarding the number of sample sawlogs, the number of ways these sawlogs are sawn, and the number of sawing i n s t r u c t i o n s w i t h i n each sawlog, the amount of data to which FLOWSIM must have access i s f a i r l y l a r g e . T h e r e f o r e , the data of dynamic i n f o r m a t i o n a r r a y s are organized i n t o an e f f i c i e n t Data Base #2  c o n t a i n i n g a l l DIA matrix elements. Having set up the DIA matrix data base, the remaining q u e s t i o n i s : which sawlog i s to be s e l e c t e d to flow through the model? The sawlog a r r i v a l to headrigs must correspond to boom lo g a r r i v a l from which they o r i g i n a t e . The boom l o g a r r i v a l i s simulated by randomly sampling from the boom l o g d i s t r i b u t i o n . Sawlogs, belonging to the boom log j u s t a r r i v e d , must then be found. The e q u i v a l e n t phase of s i m u l a t i o n i s accomplished by p i c k i n g up boom logs from Data Base #1 c o n t a i n i n g a l l boom logs  and corresponding sawlogs. N o t i c e , that there are two d i f f e r e n t data bases supporting FLOWSIM: 26 • Data Base #1 a s s i s t s FLOWSIM to simulate boom log s e l e c t i o n and to f i n d corresponding sawlogs, and • Data Base #2 i n s t r u c t s FLOWSIM how to saw these sawlogs. Beside these data bases, a d d i t i o n a l FLOWSIM inputs i n c l u d e : s i z e s and speeds of t r a n s p o r t a t i o n equipment, p o s s i b l e routes of piec e flow, and frequency d i s t r i b u t i o n s of machinery breakdown and inter-breakdown times. Having set up the two data bases and i n p u t s , FLOWSIM runs can be made to simulate m i l l dynamics. FLOWSIM r e s u l t s i n c l u d e u t i l i z a t i o n data and pie c e counts f o r both machines and t r a n s p o r t a t i o n equipment, as w e l l as sawlog l e n g t h and diameter d i s t r i b u t i o n s . FLOWSIM a l s o p r o v i d e s i n f o r m a t i o n on sample sawlog f r e q u e n c i e s . These f r e q u e n c i e s i n d i c a t e how many times each sample sawlog was processed dur i n g the s i m u l a t i o n . They a l s o c o n s t i t u t e the sample sawlog weights to c a l c u l a t e f i n a l p r o d u c t i o n d a t a . T h i s c a l c u l a t i o n i t s e l f i s c a r r i e d out by rerunning SAWRES ( a subprogram of SAWSIM) based on bin a r y input  and r e s u l t f i l e s saved from p r e v i o u s SAWSIM runs when the s e l e c t e d sample sawlogs were "sawn", and based on the sample sawlog weights provided by FLOWSIM. Having d i s c u s s e d the s t r u c t u r e and phases of s i m u l a t i o n process, the f o l l o w i n g s e c t i o n p r e s e n t s the experiments to be c a r r i e d out on the s i m u l a t i o n model. 27 3.4. DESIGN OF THE COMPUTER SIMULATION RUNS The s i m u l a t i o n runs of t h i s study were designed with two separate o b j e c t i v e s . F i r s t , there are to be v a l i d a t i o n runs to convince the user about the c o r r e c t n e s s of the model o p e r a t i o n . Second, experimental runs are set up to evaluate the behaviour of the redesigned m i l l while i t produces f o r d i f f e r e n t markets and to analyse how pie c e flow responds to breakdowns of v a r i o u s machines. A c c o r d i n g l y , the computer s i m u l a t i o n runs are c l a s s i f i e d as f o l l o w s : 1. V a l i d a t i o n runs. 1.1. Sawing p a t t e r n t e s t runs, using SAWPLOT (property of H. A. Leach Company L t d ) . 1.2. One FLOWSIM run to t e s t c o r r e c t p i e c e flow s i m u l a t i o n . 1.3. FLOWSIM runs to check c o r r e c t s i m u l a t i o n of p r o c e s s i n g times, the time sequence of o p e r a t i o n s and route d e c i s i o n s , by us i n g the i n t e r a c t i v e debugging f a c i l i t i e s of GPSS/H. T h i s i n t e r a c t i v e debugging run was a l s o used to check i f FLOWSIM p r e d i c t s c o r r e c t piece s i z e s . 1.4. One FLOWSIM run to check f o r c o r r e c t machine u t i l i z a t i o n s t a t i s t i c s and machinery i n t e r a c t i o n s . 1.5. S t o c h a s t i c , l a r g e s c a l e runs to t e s t the FLOWSIM op e r a t i o n under more r e a l i s t i c circumstances than p o s s i b l e with s i m p l i f i e d one-log runs and to estimate the computer cost of s i m u l a t i o n . T h i s c o s t estimate serves as a b a s i s to design the 28 scope of a c t u a l experiments on the model. 2. Market response runs. 3. Machinery breakdown response runs. C o n s i d e r i n g the f i r s t group of s i m u l a t i o n runs, Chapter 7 i s devoted s o l e l y to v a l i d a t i o n a s p e c t s . I t d e s c r i b e s what approach was taken to v a l i d a t e the model, what kinds of v a l i d a t i o n runs were c a r r i e d out and what the r e s u l t s of these runs were. Hence, at t h i s p o i n t only a summary of these v a l i d a t i o n runs i s g i v e n . On the other hand, problems of de s i g n i n g market and machinery breakdown experiments are d i s c u s s e d here in d e t a i l . The problem of experimental design i s that the v a r i a t i o n of FLOWSIM s i m u l a t i o n r e s u l t s may be a t t r i b u t e d to d i f f e r e n t sources. These are random boom l o g a r r i v a l , random machinery breakdown, v a r y i n g s t a r t i n g c o n d i t i o n s and experimental changes in c e r t a i n model inputs of i n t e r e s t . Since the primary o b j e c t i v e of these experiments i s to evaluate the e f f e c t of producing f o r the v a r i o u s markets, c o n t r o l l i n g the e f f e c t s of other sources i s d e s i r a b l e . The f o l l o w i n g paragraphs d e s c r i b e how the experiments were designed i n view of the above problems. The f i r s t problem i s the d i s t o r t i o n e f f e c t of v a r y i n g s t a r t i n g s t a t e s of the m i l l which should be minimized. T r a n s i e n t vs. steady s t a t e should be d i s t i n g u i s h e d and employed c o n s i s t e n t l y at s i m u l a t i o n s t a r t s . Comparison of i d l e m i l l s t a r t s and lumber-saturated s t a r t s , might l e a d to erroneous c o n c l u s i o n s . The second problem i s the random f l u c t u a t i o n caused by random machinery breakdowns. C o n s i d e r i n g that brand new 29 machinery i s to be i n s t a l l e d , breakdowns are assumed to happen r e l a t i v e l y i n f r e q u e n t l y . Hence, FLOWSIM runs should be long enough to produce an adequate sample of machinery breakdown events. T h i r d , random f l u c t u a t i o n s c o u l d a l s o be caused by v a r y i n g sequences of boom l o g a r r i v a l times and log shape c h a r a c t e r i s t i c s . To reduce t h i s random f l u c t u a t i o n the same or at l e a s t s i m i l a r - boom log a r r i v a l sequences, both i n time and shape, should be ensured. To s o l v e the problems above, that i s to minimize the random f l u c t u a t i o n caused by undesired sources and to be able to concentrate only on the e f f e c t s of d i f f e r e n t export markets, the experiments are designed as f o l l o w s . S i m u l a t i o n of a l l the s i x market runs w i l l be s t a r t e d with an "empty m i l l " . The advantage of t h i s empty m i l l s t a r t i s that the computer-time demanding procedure of determining the steady s t a t e f o r a l l of the s i x markets can be saved. On the other hand, the disadvantage of s t a r t i n g the s i m u l a t i o n with an empty m i l l i s the o v e r e s t i m a t i o n of i d l e s t a t e s t a t i s t i c s of machinery u t i l i z a t i o n . To decrease the degree of t h i s b i a s , at the very beginning of each market run the f i r s t boom logs w i l l be d e l i v e r e d to the headrigs at an a r t i f i c i a l l y h igh r a t e . Hence, both headrigs can s t a r t o p e r a t i n g r i g h t at the beginning of the s i m u l a t i o n runs thereby e l i m i n a t i n g the s t a r t i n g i d l e times of head r i g s . As a r e s u l t , i d l e times of machines at the t a i l end of the m i l l are a l s o shortened. The l e n g t h of s i m u l a t i o n , i n the case of a l l s i x market runs, w i l l be one s h i f t (27,600 seconds). When t h i s l e n g t h of 30 s i m u l a t i o n was decided, computer c o s t vs. computer d o l l a r s a v a i l a b l e and run l e n g t h s u f f i c i e n t to decrease the b i a s i n g e f f e c t of empty m i l l s t a r t s , were c o n s i d e r e d . When c o n s i d e r i n g a v a i l a b i l i t y of computer d o l l a r s and t h e i r l i m i t i n g of experimental design, two c o s t s should be d i s t i n g u i s h e d , running and developing the model. Running the model with SAWSIM and FLOWSIM now r e q u i r e s about 200 seconds of CPU time. T r a n s l a t e d i n t o computer d o l l a r s , the cost of one t y p i c a l run i s about $50. However, the t o t a l c o n v e r s a t i o n a l t e r m i n a l connect time used f o r model b u i l d i n g was about 1500 hours. Thus the computer cost of model development was i n excess of about $1000. The same seed of a separate random number generator ensures that logs of v a r i o u s shapes a r r i v e i n a s i m i l a r sequence. To ensure s i m i l a r boom log a r r i v a l times, f o r a l l s i x markets, a constant i n t e r a r r i v a l time i s used. The l o g a r r i v a l sequences of d i f f e r e n t runs are s i m i l a r and not i d e n t i c a l in the sense that the number of boom logs a r r i v i n g at the m i l l i s not the same. The r a t e of l o g a r r i v a l , as i n r e a l l i f e , depends on machine times r e q u i r e d to process l o g s . The higher t h i s time demand, the fewer logs can enter the m i l l . Consequently, the number of a r r i v i n g boom logs might be d i f f e r e n t i n the case of d i f f e r e n t markets. Summarizing, the l e n g t h of s i m u l a t i o n runs, one s h i f t (460 minutes), w i l l be the same i n the case of a n a l y s i n g a l l s i x markets. A l l s i m u l a t i o n runs w i l l s t a r t with an empty m i l l s t a t e . S i m i l a r l o g a r r i v a l sequences w i l l be used. F i v e out of the s i x runs are the s o - c a l l e d export market runs. These runs 31 i n c l u d e the pr o d u c t i o n of both dimensional lumber and lumber f o r export. The p r o p o r t i o n of dimensional lumber p r o d u c t i o n - based on m i l l management recommendation - w i l l always be 1/3. F i n a l l y , the r e s u l t s of the s i x market runs w i l l be analysed on a "judgmental b a s i s " . At t h i s p o i n t the reader might wonder why the r e s u l t s of the simulated experimentation w i l l not be analysed on a s t a t i s t i c a l l y v a l i d b a s i s . Reasons f o r t h i s are the f o l l o w i n g . F i r s t , to produce s t a t i s t i c a l l y independent r e s u l t s would be d i f f i c u l t - i f not im p o s s i b l e . Second, too many a d d i t i o n a l computer runs would be r e q u i r e d without the b e n e f i t of g a i n i n g much in f o r m a t i o n about the behaviour of the redesigned m i l l . The e f f i c i e n c y of the experimental design would not be s a t i s f a c t o r y . The book of Meier et a l (68), d e a l i n g with s i m u l a t i o n theory, says that "... i t should be p o i n t e d out that i t i s p o s s i b l e to experiment with s i m u l a t i o n models and draw c o n c l u s i o n s without using any formal s t a t i s t i c a l a n a l y s i s . Since s i m u l a t i o n models are capable of p r o v i d i n g complete h i s t o r i e s of behaviour d u r i n g a run, the r e s u l t s can o f t e n be observed and eval u a t e d on a judgmental b a s i s without using formal s t a t i s t i c a l t e s t s " . C o n s i d e r i n g the output of FLOWSIM t h i s i s the case. FLOWSIM output provides comprehensive data d e s c r i b i n g m i l l behaviour d u r i n g the time of s i m u l a t i o n : l o g throughput, machine u t i l i z a t i o n s t a t i s t i c s , machine and t r a n s p o r t a t i o n equipment p i e c e counts, and w a i t i n g time s t a t i s t i c s of p i e c e s at t r a n s p o r t a t i o n equipment. Based on FLOWSIM runs, the pro d u c t i o n data: lumber mix, lumber volume and val u e , gross p r o d u c t i o n v a l u e , recovery data e t c . , can a l s o be 32 obtained by rerunning SAWRES (as a subprogram of SAWSIM). Thus, the a n a l y s i s w i l l be based on c a r e f u l examination of t h i s output information and not on a s t a t i s t i c a l a n a l y s i s . The o b j e c t i v e of machinery breakdown runs i s to estimate roughly the extent of p r o d u c t i o n o v e r e s t i m a t i o n i n the case of breakdown f r e e market runs. The runs estimate roughly, because the r e l a t i v e l y i n f r e q u e n t machinery breakdown events would r e q u i r e too many s i m u l a t i o n runs and consequently, a great amount of computer d o l l a r s . Thus, only three s t o c h a s t i c machinery breakdown runs of d i f f e r e n t random seeds are designed when the m i l l produces f o r the U.S.A. market. The average of these runs then w i l l be compared to the breakdown-free U.S.A. market run to c a l c u l a t e the extent of o v e r e s t i m a t i o n . An a d d i t i o n a l d e t e r m i n i s t i c breakdown run i s a l s o designed to present a technique a n a l y s i n g the machinery breakdown s e n s i t i v i t y of the m o d i f i c a t i o n p l a n . The breakdown run i s d e t e r m i n i s t i c i n the sense that a p a r t i c u l a r machine breakdown i s "caused to happen" f o r a predetermined p e r i o d of time and the production l o s s i n $/hour i s c a l c u l a t e d . 33 4.LITERATURE REVIEW OF COMPUTER SIMULATION APPLICATIONS IN SAWMILLING The number of computer s i m u l a t i o n a p p l i c a t i o n s i n research and p r o d u c t i o n management of sawmill o p e r a t i o n s i s l a r g e . To d e s c r i b e a l l of these, even b r i e f l y , i s beyond the scope of t h i s t h e s i s . T h e r e f o r e , the approach taken i n reviewing the l i t e r a t u r e i s as f o l l o w s . In Appendix 1., BIBLIOGRAPHY OF COMPUTER SIMULATION APPLICATIONS IN SAWMILLING, the l i t e r a t u r e of s a w m i l l i n g r e l a t e d to computer s i m u l a t i o n i s given. T h i s l i s t i s based p a r t l y on the b i b l i o g r a p h y of "Use of Computers and Computerized P r o c e s s i n g Systems" (51), p a r t l y on the resear c h paper of "A Survey of Operations Research A p p l i c a t i o n s in Wood Products I n d u s t r i e s " (73), and p a r t l y on my own l i t e r a t u r e survey. In a d d i t i o n to t h i s l i s t i n g , a b r i e f c h a r a c t e r i z a t i o n i s presented by major c a t e g o r i e s . The c a t e g o r i e s are d i s c u s s e d r e s p e c t i v e l y on the b a s i s of examples. (Note that r e f e r e n c e s , other than s i m u l a t i o n a p p l i c a t i o n s i n s a w m i l l i n g , are l i s t e d s e p a r a t e l y i n Chapter 10. A p p l i c a t i o n s of s i m u l a t i o n in sawmilling g e n e r a l l y f a l l i n t o one of the two f o l l o w i n g c a t e g o r i e s : A. S t a t i c s i m u l a t i o n models of log or cant breakdowns. These models perform c a l c u l a t i o n s repeatedly on a l o g or cant i n 34 order to see how v a r i o u s breakdown s t r a t e g i e s may r e s u l t i n d i f f e r e n t volume and value y i e l d s . The a p p l i c a t i o n s of t h i s group can be f u r t h e r c l a s s i f i e d a c c o r d i n g to whether both the shape and q u a l i t y c h a r a c t e r i s t i c s of the log (A1) or only the shape c h a r a c t e r i s t i c s (A2) are c o n s i d e r e d . B. Dynamic s i m u l a t i o n models of piece flow i n a m i l l . These models are a p p l i c a b l e to simulate the time sequence of o p e r a t i o n s on d i f f e r e n t machine c e n t e r s . C o n s i d e r i n g the above two c a t e g o r i e s , much more a t t e n t i o n has been p a i d to the lumber recovery issue (group "A") than to m i l l dynamics (group "B"). There are two reasons f o r t h i s . F i r s t , the e f f e c t of log breakdown on m i l l performance i s g r e a t e r than that of m i l l dynamics; consequently, m i l l managers and a n a l y s t s have been more i n t e r e s t e d i n t h i s i s s u e . Thus, research on s i m u l a t i o n of log breakdown s i m u l a t i o n had p r i o r i t y over res e a r c h on dynamic s i m u l a t i o n . The second reason i s the f o l l o w i n g . B u i l d i n g a dynamic model i s more complicated than developing a s t a t i c model. Knowledge of s i m u l a t i o n theory and s p e c i a l purpose s i m u l a t i o n computer languages i s necessary. T h e r e f o r e , i t r e q u i r e s a higher programmer-hour input and more s k i l l e d programmer a c t i v i t y . Perhaps these are the major reasons why much more research has been done on s t a t i c s i m u l a t i o n models than on the dynamic ones. S i m u l a t i o n models f a l l i n g i n category "A", i n one way or the other, analyze how changes of v a r i o u s sawing v a r i a b l e s a f f e c t lumber recovery. Models of group "A1", f o r instance (30, 35, 47), take i n t o c o n s i d e r a t i o n the log q u a l i t y through the degrading i n f l u e n c e of i n t e r n a l knot c h a r a c t e r s . In b u i l d i n g 35 these models, one of the problems i s to determine the s i z e , the l o c a t i o n and the nature of i n t e r n a l knots. For t h i s purpose a sample of log s must be s e l e c t e d and sawn i n t o boards. T h i s makes " l o o k i n g i n t o the l o g s " p o s s i b l e , and the inf o r m a t i o n on knots can be encoded i n t o the computer. In a l l cases the encoding process i s awkward. Once the knots are determined, t r i g o n o m e t r i c c a l c u l a t i o n s l e a d to accurate grading of t h e o r e t i c a l l y sawn lumber. A l l resea r c h i n t h i s category concludes that the r e l a t i v e p o s i t i o n of cu t s to knots in the l o g has a s u b s t a n t i a l impact on lumber value recovery. However, the que s t i o n s of "how can the i n t e r i o r l o g q u a l i t y c h a r a c t e r i s t i c s : knot p o s i t i o n and decay be l o c a t e d and coded i n t o the computer" and "how to saw as a f u n c t i o n of the i n t e r i o r l o g q u a l i t y " have not been answered y e t . T h i s i s why one of the expected improvements in the sawmills of futu r e i s a method of " l o o k i n g i n t o the l o g " not only by X-ray but by r e l a t i n g i n t e r i o r c o n d i t i o n s to e x t e r i o r geometry (85). As a proof of W i l l i s t o n ' s (85) p r e d i c t i o n regarding the sawmills of tomorrow, i n t e r e s t i n g r e s e a r c h i s being done by Funt (59, 81). His o b j e c t i v e i s to "look i n t o the l o g " by a s p e c i a l (medical-type) scanner and then to convert the output of the scanner by an i n t e r p r e t a t i o n a l g o r i t h m i n t o an a p p r o p r i a t e input form f o r a d e c i s i o n making program. The d e c i s i o n making program then decides how to cut the l o g to gain maximum lumber value from i t . However, the c u r r e n t l y r e q u i r e d high scanning time, high computer time r e q u i r e d for i n t e r p r e t a t i o n and d e c i s i o n making per l o g , as w e l l as the high scanner p r i c e ($ 1.25 m i l l i o n ) , make i t s p r a c t i c a l o n - l i n e a p p l i c a t i o n e c o nomically 36 im p o s s i b l e . One of the best known r e p r e s e n t a t i v e s of the s t a t i c group (A2) i s the Best Opening Face (BOF) computer sawing program developed by Hallock and Lewis at the F o r e s t Products Laboratory, F o r e s t S e r v i c e U.S. Department of A g r i c u l t u r e , i n 1971. To analyze lumber y i e l d , t h i s program, i n a d d i t i o n to determining the l o c a t i o n of the opening face, took i n t o account the f o l l o w i n g sawing v a r i a b l e s : the t h i c k n e s s and width of the dry f i n i s h e d lumber, p l a n i n g allowance, shrinkage d u r i n g d r y i n g , saw kerf width, log diameter and sawing method.. BOF made two bas i c assumptions: the lumber i n c l u d e s no wane and the l o g i s c y l i n d r i c a l (17). A l a t e r v e r s i o n of the BOF computer program, r e l e a s e d in 1976, made i t p o s s i b l e to simulate e f f e c t s of uniform taper along the len g t h of the l o g . One of the advantages of t h i s BOF program i s i t s a v a i l a b i l i t y f o r p u b l i c usage. However, a p p l i c a t i o n of BOF at d i f f e r e n t computer i n s t a l l a t i o n s might make some changes necessary. A l i s t of p o s s i b l e changes i s a v a i l a b l e from the author of the program (20) . S t i l l w i t h i n group "A2", another model named SIMSAW, was developed at the N a t i o n a l Timber Research I n s t i t u t e of the CSIR (C o u n c i l f o r S c i e n t i f i c and I n d u s t r i a l Research), i n South A f r i c a . Given the l o g data, the sawing p a t t e r n and p r i c e s of f i n a l products, SIMSAW determines the boards, c h i p s and sawdust that can be produced c o n s i d e r i n g e i t h e r maximum value or volume recovery o b j e c t i v e s . SIMSAW was s p e c i a l l y designed f o r sawing systems i n South A f r i c a (43). SAWSIM, (and not SIMSAW) another r e p r e s e n t a t i v e of group 37 "A2", w i l l be used in t h i s study, hence, i t i s d i s c u s s e d l a t e r . A prominent example of the dynamic models (group "B") i s MILLSIM developed by Aune, when at the Western F o r e s t Products Laboratory, Canadian F o r e s t r y S e r v i c e i n 1973 ( 6 ) . A f i r s t v e r s i o n of t h i s model was a p p l i c a b l e only to a given sawmill setup, but l a t e r i t was g e n e r a l i z e d to i n c l u d e a wide range of small l o g softwood dimension m i l l l a y o u t s . MILLSIM was w r i t t e n i n FORTRAN with a GASP s i m u l a t i o n package. T h i s d i s c r e t e event s i m u l a t i o n model of the sawmill enables the user to compare v a r i o u s designs f o r a given l o g p o p u l a t i o n , to c a r r y out s e n s i t i v i t y a n a l y s i s of e i t h e r m i l l design or l o g mix c h a r a c t e r i s t i c changes, and to detect flow problems. A d d i t i o n a l f e a t u r e s of Aune's model a r e : • It approximates l o g shape by t r u n c a t e d cones; • I t simulates headrig o p e r a t i o n i n d e t a i l ; • Time needed to process a p i e c e on a machine i s c a l c u l a t e d by r e g r e s s i o n equations; • The d u r a t i o n of machine breakdowns i s sampled from e m p i r i c a l d i s t r i b u t i o n s based on breakdown time s t u d i e s throughout B r i t i s h Columbia (52). • I t can be operated i n t e r a c t i v e l y , hence, from the user's p o i n t of view, i t does not need too much knowledge of s i m u l a t i o n techniques; • With some m o d i f i c a t i o n s i t can produce lumber other than 2 inch t h i c k n e s s ; • The model uses about 120 CPU seconds to simulate three hours of m i l l o p e r a t i o n . Under the commercial r a t e s of 1977 at UBC, t h i s run cost about $500 (36). 38 Based on the concepts of Aune's model, Richard (36) c o n s t r u c t e d a d i s c r e t e dynamic s i m u l a t i o n model. I t s s p e c i a l purpose was "to examine the p o t e n t i a l e f f e c t s on t o t a l m i l l p r o d u c t i o n of s o r t i n g logs by diameter c l a s s e s " at Warm Springs F o r e s t Products Pine Sawmill, Oregon, USA. Compared to Aune's model, the main i n n o v a t i o n s i n Richard's model were: "the i n c l u s i o n of grade i n the l o g breakdown; a h e a d r i g r o u t i n e c u t t i n g 4/4 and 5/4 from a l l four l o g f a c e s ; feedback mechanisms in the h e a d r i g r o u t i n e ; s i z e and shape dependent, s t o c h a s t i c elements i n the p r o c e s s i n g times; s e p a r a t i o n of p r o c e s s i n g time and i n t e r - e v e n t time; " [ s i c ] ... more r e a l i s t i c flow r e p r e s e n t a t i o n ; and output r o u t i n e extension (36). In a d d i t i o n to the above two models, in the l a t e 60's, M a r t i n (67) developed a dynamic model in FORTRAN to analyze a p a r t i c u l a r hardwood sawmill c o n f i g u r a t i o n in the Northeast. Log diameters, lengths and s e r v i c e times were simulated by predetermined t h e o r e t i c a l f u n c t i o n s . The model estimated s t a t i s t i c s of queues in f r o n t of machine c e n t r e s , p r o c e s s i n g times, machine u t i l i z a t i o n and s i z e s of boards produced. One weakness of the model was that p r o c e s s i n g times were taken i n t o account independently of p i e c e s i z e s . Another sawmill s i m u l a t o r , developed by the Jaakko Poyry Company, in F i n l a n d , was used to t e s t sawmill designs (65). The manner by which i t avoids the time demanding programming of l o g breakdown d i s t i n g u i s h e s i t from other models: the number of lumber p i e c e s produced from each diameter c l a s s of l o g i s predetermined by the user. Another study d e a l i n g with m i l l dynamics worth mentioning, 39 though i t does not use s i m u l a t i o n techniques was by Carino and Bowyer (53). I t used queuing theory combined with a d i r e c t search o p t i m i z i n g a l g o r i t h m to f i n d optimal s o l u t i o n s to m i l l o p e r a t i n g problems. The a r t i c l e does not show how s t o c h a s t i c elements of p i e c e flow, caused by i r r e g u l a r l o g shape c h a r a c t e r i s t i c s , are taken i n t o c o n s i d e r a t i o n . 40 5_;_ INPUT DATA REQUIRED FOR THE SIMULATION MODEL The major c h a r a c t e r i s t i c s of the o p e r a t i o n at any m i l l can be d e s c r i b e d by the logs a v a i l a b l e f o r p r o d u c t i o n , the products to be manufactured, the machinery and the technology. A c c o r d i n g l y , t h i s chapter d i s c u s s e s these major data groups and the r e l a t e d problems. 5.1 ESTABLISHMENT OF THE SAMPLE LOG DATA BASE One of the fundamental concepts of the dynamic model e x p l a i n s the importance of s e l e c t i n g the sample logs f o r s i m u l a t i o n runs. T h i s b a s i c concept i s d e s c r i b e d b r i e f l y below, but, to prevent misunderstanding some e x p l a n a t i o n of the terminology i s needed. Boom logs are logs which have been bucked i n the woods and s t o r e d i n the m i l l l o g yar d . They are g e n e r a l l y some m u l t i p l e of the sawlogs cut by the m i l l . Sawlogs are logs which have been bucked i n the m i l l to a p p r o p r i a t e l e n g t h f o r sawing. Sample sawlogs are s e l e c t e d and "sawn" by SAWSIM in advance. Then the sample sawlogs, together with the sawing in f o r m a t i o n produced by SAWSIM, (time needed to process a pie c e on a p a r t i c u l a r machine, how many pi e c e s are generated at a machine, where do p i e c e s go, etc.) are placed i n t o a " l o g p o o l " . From t h i s pool they are randomly p i c k e d up by the dynamic model 41 and processed through the m i l l , while s t a t i s t i c s are c o l l e c t e d about m i l l dynamics. Due to the f a c t that l o g shape a f f e c t s m i l l performance to a l a r g e degree, there are two major requirements when one i s s i m u l a t i n g the behaviour of a sawmill: a. Sample logs should be true r e p r e s e n t a t i v e s of r e a l l o g s a v a i l a b l e in the l o g pond (or l o g yard) e s p e c i a l l y as to t h e i r shape c h a r a c t e r i s t i c s . Length, diameter, taper, sweep, crook, o v a l i t y and t w i s t of logs should be c o n s i d e r e d to a reasonable degree a c c o r d i n g to t h e i r impact on m i l l performance, and b. Logs of c e r t a i n shape should be sampled a c c o r d i n g to r e a l shape f r e q u e n c i e s observable i n the l o g boom. To meet the above requirements, the sample l o g s e l e c t i o n process f o r t h i s study was as f o l l o w s : 1. Determine the diameter and l e n g t h f r e q u e n c i e s of logs a r r i v i n g i n Woodroom #3 to e s t a b l i s h a two-dimensional d i s t r ibut i o n . 2. Assign sample boom logs to each c l a s s of the two-dimensional d i s t r i b u t i o n . 3. Buck boom logs i n t o sawlogs a c c o r d i n g to the p r e v a i l i n g m i l l bucking p o l i c y . 4. S e l e c t sawlogs to be sawn by SAWSIM. Th i s s e l e c t i o n procedure must be f l e x i b l e enough to c o n s i d e r computer c o s t s of "sawing" and s i m u l a t i o n accuracy. The remaining part of t h i s chapter i s a d e t a i l e d d i s c u s s i o n of the p r e v i o u s sawlog s e l e c t i o n procedure. 42 5.1.1. Two-dimensional diameter and len g t h d i s t r i b u t i o n of logs  in Woodroom #3. Taper as an a d d i t i o n a l l og shape c h a r a c t e r i s t i c The two most important shape c h a r a c t e r i s t i c s , l e n g t h and diameter are given by a two-dimensional e m p i r i c a l frequency d i s t r i b u t i o n of boom logs coming i n t o the m i l l (Appendix 5.1). T h i s d i s t r i b u t i o n i s c a l c u l a t e d from the data made a v a i l a b l e by the company's log supply d i v i s i o n (Appendix 5.2). Taper, another important l og shape c h a r a c t e r i s t i c w i l l be taken i n t o c o n s i d e r a t i o n a c c o r d i n g to the f o l l o w i n g . S t u d i e s show that the taper v a r i e s along the len g t h of a stem. Stems have r e l a t i v e l y l a r g e r taper at both butt and top end. On the other hand l e s s taper i s observed i n the middle p o r t i o n of the stem. Demaerschalk and Kozak's taper model (56), p r e d i c t i n g diameter of a t r e e , r e f l e c t s t h i s . In a d d i t i o n to t h i s a b r i e f study was c a r r i e d out of how taper v a r i e s along the leng t h of western hemlock stems. The computer program to c a l c u l a t e taper f o r c r o s s - s e c t i o n s of logs can be seen i n Appendix 5.3. T h i s short study of 26 stems concluded that the butt end of the stems almost always had high t a p e r . F i g u r e 5.1 i l l u s t r a t e s t h i s by d i s p l a y i n g four t y p i c a l taper conf i g u r a t i o n s . Consequently, i f one wants to consid e r the e f f e c t of taper on m i l l performance, boom logs i n c l u d i n g the butt end of the stem with l a r g e r taper and boom logs without butt end, that i s with r e l a t i v e l y smaller taper should be d i s t i n g u i s h e d . ' (The need f o r d i s t i n g u i s h i n g among boom logs of v a r i a b l e taper w i l l F i g u r e 5 . 1 . T a p e r v a r i a t i o n a l o n g t h e l e n g t h o f s t e m s 44 become c l e a r when s u b s t i t u t e sample sawlogs are d i s c u s s e d . Two sawlogs of d i f f e r e n t taper do not produce s i m i l a r sawing r e s u l t s even i f a l l other shape c h a r a c t e r i s t i c s are the same. T h i s i s why taper has to be d i s t i n g u i s h e d . ) However, boom logs i n c l u d i n g the top end of the o r i g i n a l stem, where the taper i s a l s o l a r g e , do not r e c e i v e s p e c i a l a t t e n t i o n . The reason for t h i s i s that Woodroom #3 cuts only logs of 14" minimum top diameter. Logs which do not meet t h i s requirement are t r a n s p o r t e d to e i t h e r Woodroom #4, the small l og sawmill at the Harmac m i l l complex, or to the Pulp m i l l f o r c h i p product i o n . To determine the p r o p o r t i o n of boom logs with butt end (denoted by "B") and boom logs with no butt end ("N"), a sample of incoming boom logs was taken at the logdeck of Woodroom #3 (Appendix 5.4). R e s u l t s show that the p r o p o r t i o n of "B" boom logs i n c r e a s e s with top diameter. A f t e r t h i s d i s c u s s i o n , the " l o g p o o l " from which the dynamic model samples up boom log s f o r flow through the m i l l , can be thought of as a two dimensional array (Appendix 5.6). Each c e l l r e p r e s e n t s a c e r t a i n length/diameter combination and may c o n t a i n two types of boom l o g s : B and N. 5.1.2. Assignment of sample boom logs to the c l a s s e s of the  two-dimensional d i s t r i b u t i o n Now the q u e s t i o n s are what and how many a c t u a l boom logs w i l l be as s i g n e d to each p a r t i c u l a r c e l l ? Woodroom #3, at Harmac, c u t s western hemlock l o g s . To prov i d e l o g measurements, a group of hemlock stem data, made 45 a v a i l a b l e by FORINTER CANADA CORP., w i l l be used as sample l o g data. For SAWSIM runs the shape of a l o g i s d e s c r i b e d by data r e c o r d s . Each data r e c o r d d e s c r i b e s one log c r o s s s e c t i o n with the z, x, y c o o r d i n a t e s of l o g ce n t e r , diameter, o v a l i t y and angle of the l a r g e s t diameter of the c r o s s s e c t i o n to the x-a x i s . Thus the shape of a l o g i s d e s c r i b e d by as many data records as c r o s s s e c t i o n s c o n s i d e r e d and r e q u i r e s s i x data e n t r i e s f o r each c r o s s s e c t i o n . U n f o r t u n a t e l y , the hemlock stem data of FORINTEK were not i n the r e q u i r e d format f o r SAWSIM runs. T h e r e f o r e , r e f o r m a t t i n g was needed. The procedure i s d e s c r i b e d i n Appendix 5.5. As i s g e n e r a l l y true i n c o n j u n c t i o n with sampling, a c o n f l i c t a r i s e s from the s e l e c t i o n of sample l o g s . The sample s i z e needed i n c r e a s e s with accuracy r e q u i r e d and so i n c r e a s e s the cost of sampling and computer s i m u l a t i o n . Consequently, compromise i s needed. Based on the f o r e g o i n g data and p r i n c i p l e s , the number of sample boom logs f o r each c e l l of the " l o g p o o l " i s assigned p r o p o r t i o n a l l y to the two dimensional e m p i r i c a l d i s t r i b u t i o n of boom logs (Appendix 5.1), and to the p r o p o r t i o n of B boom log s (Appendix 5.4). Bearing i n mind, that the number of boom log s w i t h i n the "l o g p o o l " should not be high, as a rough r u l e , one boom log per 50 f r e q u e n c i e s was ass i g n e d . The p r o p o r t i o n s of "B" boom log s i n rows 1 to 4 of the " l o g p o o l " (Appendix 5.6) are e i t h e r 1:1, 2:2, or 3:2, in accordance with the p r o p o r t i o n of "B" boom logs which i s 56-61 % w i t h i n the diameter range 15-22" (Appendix 5.4). Logs of 23" or grea t e r diameter always i n c l u d e the butt of the t r e e , thus rows 5 to 6 of the log pool 46 c o n t a i n only "B" l o g s . The r u l e s r e p r e s e n t i n g these p r i n c i p l e s are the f o l l o w i n g : In c e l l s of row 1,2,3 and 4, as w e l l as with f r e q u e n c i e s of 149 or l e s s two sample boom l o g s : B1, N1; with f r e q u e n c i e s between 150 and 549 four sample boom l o g s : B1, B2, N1, N2; otherwise f i v e sample boom logs B1, B2, B3, N1, N2 w i l l be chosen. If the d i s t r i b u t i o n c e l l i n q u e s t i o n i s i n row 5,6, that i s only a butt end log i s i n v o l v e d , then only one sample boom log B1 w i l l be chosen f o r c e l l s with f r e q u e n c i e s of l e s s than 150; otherwise three sample boom logs B1, B2, B3 w i l l be s e l e c t e d . To summarize, each c e l l of the boom log pool c o n t a i n s a c e r t a i n number of boom logs p r o p o r t i o n a l to the o r i g i n a l f r e q u e n c i e s given by m i l l management. The diameter and length of a boom l o g , must be w i t h i n the corresponding diameter and le n g t h c e l l . Taper as another important shape c h a r a c t e r i s t i c i s taken i n t o account by d i s t i n g u i s h i n g between boom log s with butt end and without butt end. Other shape c h a r a c t e r i s t i c s such as sweep, v a r i a b l e taper, o v a l i t y and t w i s t were sampled randomly. The "boom log p o o l " , formed a c c o r d i n g to the above r u l e s , with i t s sample boom log codes content of 84 c e l l s , i s d i s p l a y e d in Appendix 5.6. To designate boom l o g data s u i t a b l e for the two-dimensional d i s t r i b u t i o n of Woodroom #3 FORINTEK stem data were s t u d i e d . For a p a r t i c u l a r boom l o g , stems were s e l e c t e d and one of these was designated randomly. I f no stem c o u l d be found from the FORINTEK data for a p a r t i c u l a r c e l l , then boom l o g data had to be "manufactured" by e x t r a p o l a t i n g the diameter while other 47 shape c h a r a c t e r i s t i c s sweep (x, y, z c o o r d i n a t e s ) , o v a l i t y , t w i s t remained unchanged. Examples of boom l o g measurements are l i s t e d in Appendix 5.7. 5.1.3. Bucking p o l i c i e s of Woodroom #3 Boom l o g s are c r o s s cut a c c o r d i n g to the m i l l management bucking p o l i c i e s . The b a s i c concept of bucking i s to c r o s s cut boom logs i n t o sawlogs a s s u r i n g lumber pr o d u c t i o n of maximum val u e . Taking i n t o c o n s i d e r a t i o n lumber p r i c e s , minimum sawlog l e n g t h (determined by knee d i s t a n c e s on the h e a d r i g c a r r i a g e ) , three d i f f e r e n t bucking schedules (Appendix 5.8) were set up f o r the three market groups below. 1. U.S.A., U n i t e d Kingdom, Japan (lumber of 1 13/16" t h i c k n e s s ) , 2. France and North A f r i c a and 3. Japan (4 X 4). These bucking schedules are given to the bucking operator who based on the l e n g t h of the l o g (and expected cant width i n the case of U.S.A. market), looks up bucking lengths and c a r r i e s out the c r o s s c u t t i n g . The expected cant widths as a f u n c t i o n of diameters are summarized in Appendix 5.9. In t h i s Apppendix f a c t o r s which were c o n s i d e r e d when determining the expected cant width are a l s o d e s c r i b e d . Despite the f a c t that the number of boom logs was minimized, the l o g pool c o n t a i n s 174 boom l o g s . Obviously the number of sawlogs i s even l a r g e r . In a d d i t i o n to t h i s , i f we t h i n k of the number of lumber markets, t h e i r d i f f e r e n t bucking schedules and t h e i r d i f f e r e n t sawing p a t t e r n s , we end up with a l a r g e number of sawlogs to be sawn by SAWSIM. Consequently, the 48 computer cost of sawing a l l these sawlogs would be high. This, r a i s e s the need f o r d e c r e a s i n g the number of sawlogs to be sawn by SAWSIM. 5.1.4. Sawlog s e l e c t i o n f o r SAWSIM runs The problem of sawlog s e l e c t i o n must be s o l v e d by keeping in mind that the fewer the sample logs the l e s s accurate i s the s i m u l a t i o n . In other words we have to decrease the number of sawlogs but at the same time we must t r y to minimize the l o s s of accuracy. The f o l l o w i n g procedure o f f e r s a p o s s i b l e s o l u t i o n to t h i s problem. A f t e r having bucked the boom logs i n t o sawlogs, i f there are sawlogs having s i m i l a r major shape c h a r a c t e r i s t i c s , only one of these sawlogs should be sawn. The sawing info r m a t i o n of others can be s u b s t i t u t e d by the one sawn a l r e a d y . Thus, sawing of sawlogs by SAWSIM can be reduced. It i s up to us what we mean by s i m i l a r sawlogs. For in s t a n c e , we can say that sawlogs f a l l i n g w i t h i n a c e r t a i n l e n g t h and diameter i n t e r v a l and o r i g i n a t i n g from a "N" boom l o g are s i m i l a r . The degree of s i m i l a r i t y can be c o n t r o l l e d by s e t t i n g up c e l l s f o r the s e l e c t i o n process and a s s i g n i n g a p p r o p r i a t e boundaries to these c e l l s . Within these c e l l s sawlogs can be c o n s i d e r e d s i m i l a r . The narrower the c e l l width, the l e s s l o s t i n s i m u l a t i o n accuracy, but the fewer the sawlogs whose sawing by SAWSIM can be saved. The degree of s i m i l a r i t y i s e v e n t u a l l y set up by the sawlog s e l e c t i o n t a b l e and i t s c e l l s . As an example, a sawlog s e l e c t i o n t a b l e i s d i s p l a y e d in Appendix 5.10. I t has ten columns f o r lumber lengths and 12 49 rows of diameter c l a s s e s . By grouping s i m i l a r sawlogs, and then s e l e c t i n g only one from each group, the number of sawlogs to be sawn by SAWSIM can be reduced s u b s t a n t i a l l y . A flowchart d e p i c t i n g the o u t l i n e of t h i s sawlog s e l e c t i o n a l g o r i t h m can be seen as F i g u r e 5.2. T h i s a l g o r i t h m c o n s i s t s of two major p a r t s . The f i r s t , d e p i c t e d by Block 1-4, i s the bucking p a r t . T h i s s e c t i o n c r o s s - c u t s boom logs i n t o sawlogs of r e q u i r e d number and at bucking d i s t a n c e s measured from the l a r g e end of the boom l o g . I t a l s o c a l c u l a t e s sawlog measurement data of SAWSIM f o r the bucking c r o s s s e c t i o n ( s ) . T h i s i s c a r r i e d out by three p o i n t curve l i n e a r i n t e r p o l a t i o n (Kozak and Young p r i v a t e communications). The corresponding computer program i s in Appendix 5.11. Bucking d i s t a n c e ( s ) from the butt end of the boom lo g must be s u p p l i e d as input. Bucking d i s t a n c e s must be determined in accordance with the bucking p o l i c y of the m i l l ; otherwise the m i l l performance p r e d i c t e d by the model w i l l be i n c o r r e c t . T h i s p o l i c y was d i s c u s s e d i n s e c t i o n 5.1.3. T h i s bucking program, i s a more accurate accomplishment of bucking p o l i c y because of the f o l l o w i n g . I t assumes that the c u t - o f f saw operator never misjudges the length and small end diameter of boom l o g s ; that he always judges the expected cant width c o r r e c t l y ( i . e . , there i s an agreement between the h e a d r i g and c r o s s cut saw operators and as a consequence, the cant which i s cut at the h e a d r i g i s always the same as expected by the c r o s s cut saw o p e r a t o r ) . I t a l s o assumes that he always c r o s s c u t s boom logs to sawlogs of the proper l e n g t h . Obviously, t h i s i s not always the case. 50 S t a r t No -t>(Take one boom l o g Buck boom l o g i n t o as many sawlogs as r e q u i r e d by sawing p o l i c y . Yes Q Take one sawlog Yes 2± Based on the diameter and length put t h i s " B / r sawlog i n t o the approp-r i a t e c e l l of " B / r sawlog s e l e c t i o n t a b l e . Based on the diameter and l e n g t h put t h i s sawlog i n t o the approp-r i a t e c e l l of sawlog s e l e c t i o n t a b l e . F i g u r e 5.2. Flowchart of sawlog s e l e c t i o n procedure 1 3 . Put t h i s sawlog i n t o f i n a l boom l o g t a b l e with sawing i n s t r u c t i o n . S e l e c t randomly one sawlog w i t h i n t h i s c e l l and put i t i n t o f i n a l boom l o g t a b l e with sawing i n s t r u c t i o n . S u b s t i t u t e a l l other sawlogs with the s e l e c t e d one. Yes F i g u r e 5.2. Flowchart of sawlog s e l e c t i o n procedure - con'd. 52 B 15 Take one c e l l of " B / l " sawlog s e l e c t i o n t a b l e 18 Put t h i s sawlog i n t o f i n a l boom log t a b l e with sawing i n s t r u c t i o n . S e l e c t r a n d o m l y one sawlog w i t h i n t h i s c e l l and put i t i n t o f i n a l boom l o g t a b l e with sawing i n s t r u c t i o n . S u b s t i t u t e a l l other sawlogs with the s e l e c t e d one. F i g u r e 5.2. Flowchart of sawlog s e l e c t i o n procedure - con'd. 53 The second p o r t i o n of the flowchart ( F i g u r e 5.2), Blocks 5-19, i s the sawlog s e l e c t i o n p a r t . The corresponding d e t a i l e d computer program i s i n Appendix 5.12. Based on length, diameter and taper of the sawlog, t h i s program s e l e c t s sawlogs f o r SAWSIM runs. Assuming that sawlogs of l a r g e taper and sawlogs of small taper r e s u l t i n d i f f e r e n t SAWSIM runs, the program groups the sawlogs i n t o two groups (see d e c i s i o n block 6 of fl o w c h a r t , F i g u r e 5.2). In the f i r s t group (Block 7) there are sawlogs o r i g i n a t i n g from the extreme butt end of a stem, i . e . sawlogs of l a r g e taper and r e f e r r e d to as "B/1" sawlogs. In the second group there are sawlogs of low taper (Block 8 ) . The sawlogs of these two groups, based on t h e i r diameter and len g t h shape c h a r a c t e r i s t i c s , are put i n t o the a p p r o p r i a t e c e l l of one of the two separate sawlog s e l e c t i o n t a b l e s . Sawlogs of l a r g e taper and sawlogs of low taper are handled s e p a r a t e l y , although the p r i n c i p l e s of s e l e c t i o n procedure are the same. The diameter and len g t h of sawlogs belonging to a p a r t i c u l a r c e l l of the sawlog s e l e c t i o n t a b l e are w i t h i n predetermined l i m i t s . T h e r e f o r e , w i t h i n a p a r t i c u l a r c e l l , sawlogs are assumed to produce the same SAWSIM in f o r m a t i o n , and can s u b s t i t u t e f o r each other as f a r as SAWSIM r e s u l t s are concerned. T h i s i s why Block 12 and 17 s e l e c t only one sawlog randomly from a c e l l of the sawlog s e l e c t i o n t a b l e s and a t t a c h the "saw i t " i n s t r u c t i o n to i t . A l l other sawlogs from t h i s c e l l are branded with the " s u b s t i t u t e i t with s e l e c t e d sawlog of t h i s c e l l " i n s t r u c t i o n . F i n a l l y , sawlogs with t h e i r sawing i n s t r u c t i o n s are put i n t o the f i n a l boom log t a b l e (Figure 5.3). The f i n a l boom log t a b l e with the sawlogs and t h e i r sawing 54 •»»•• • Ct • — — ft • - • • Z • Z 2 1 Z Z • _••»••- mm n n n «•«-•»» « n r i c i • « • fp Cf r-t • g):l • ~ — — • «-••» n • B z z « z z «J — r » n f t c » — * J UI M*- P* f- K K r- r» UI <J U — ftp* — n — ft<-B B O I Z Z Z • 8 Z CD Z •u n rt ft ft n n n n n —• ft •"•CM ft — — rt B s z » z a n z z . MP* — ' • 00« tp O n i f i n - ~" " ' 8 B Z Z - j • • • a • • u» — — — — — — —DM ci rt rt rt rt rt rt :B B * B Z Z Z Z • • Cl • o • • •*• • n • • a • j n r t -u O O O O UN c. n rt > • 'B O B » * z z z J • : • I I '.J — — o r> ft :W B • B B B : U ft rt f* rt n — ct n — f t B a z a z -« n o n 8 • • — N t t a « • ft — — «- • zzzz • • I •• « • « n ft — — : a a a a a a r* rt r% r* n t% rt ~ «- « rt — B Z 0 = 8 Z Z a B z z .. - - ! • - n n •» • • <• . • f i n t • • rt — a • — « • • Z Z n • e> 8 z -i • • • • _1 • • • • W _l ft ft— ui r» P» »*• fc u; r> (T) m kf) v rt rt r* rt u r> n rt r> a z B z mm z z • — • • • rt rt rt • a • • • rt rt rt-. • a a a a • — « • ^ « — — — rt . z • • n • a a z «•> • • • » J — w w . j — ft ft — w a a a a in -« tr U ft ft ft ft u n n n n a z B Z a a z z i • • a r» r» i n — « - • •» • * Z 8 2 Z i • 4«r« rt • — • •»••*-•»• o • a « • a a z • z . j • • * • • a _j«~ n n n PI — 41 — — u •» «r •»•»•»» •» a a z e Z z • a a •» o • - - -v » Z Z Z J • • • • _i — ft c« — lu o O O O o ^ * •» •« • z • z w a a a a • • • rt rt rt • n * • • i n n c i • •>*• n • • • — — — • — • c i • • z z z • a * z _j — — — r» c i n n n — •»* — « r» •»» r» n o •<-«•»» 8 a a a a z a z Z z * • a o • • n i * • • 8 Z J • • • • • i— rt rt — -Cb a a a • Z Z z z z • z • • • - • • • : • a • — n rtn rt n —- — > a a a a a a a a i w r» r» r» n n r u n 1 8 8 Z 8 B z z z r* rt o n r> a : m n n n n u o n n r> • • o n n n ci • •• r* a a a «r a • • «- « n « « Z Z Z Z Z ^ • • • • a * B ^ ~ rt rt rt n — w a a a a a a a " PI - ft - ft N| a s 8 a Z z z • • • z _) — PI ft mj •»••»» tr •« 8 8 Z Z O • •* • w - • z • z J • • • • _< — ft ft — tu O O O O u — -' 8 a Z z a • « Z ft « Z - Z • ..J • • • • J ft ft -l§* •»•>» •» O ft ft ft ft • ft ft — • r»- n t-• : n a <v . «a • — — — n • a z z . j • • • t -J — ft c« ui a a a a u n n n n a Z a z j * • • • UJ n n n n • ft ft CI ft ft • :• • r i ft r* o a • • « —— *- •» n • a • • • ••- -» • • * zz z za• j : • • • • . • « — ft ft CI ft— uj a a a a a a a a • : • ft •« z _ - r i r« — 1*1 n n n n o r i r i r< n u n n a z • • n • • r* • « •* n • n m m 49 • -I • • • • _i — •« rt -f • h i •>•!». f -u n n n n ft • • — ft • • z a . j • • * • -i — — — ui ft ft ft ft • rtin n ft — w • ft r* ca n ; a a • — ft • zzz zz a a a 8 - n n : ' « a » z •« a a f r - a a o l> 8 : 8 • » • i d • —•-•»- • i n • Z Z • _ ft ft — kw ft c a n n u r t f t f t n 8 Z 8 Z •»- ft n >r n i n • z -j • • i - a a o • r t r* • • rt n • • a a • a ••*-••- • n • a B • : -!•••« J — C* Ct o n o n n a a z z •» n r> <* a i B Z — ft n •* a t * a a a z : • c i n •» a i U ft ft a - Z i g u r e 5.3. Sawing i n s t r u c t i o n s i n the f i n a l boom l o g t a b l e c to cn « CO CO D> * 3 LQ rf »-•• Ol 3 t r t n to n I o o 3 c o rt O 3 - 0) 3 * 3 cr o o 9 O vO CELL43 CELL44 1 BMSI****** B1441****** 3 N143I-N1903 N1441****** 3 4 CELLBO I BiSOl****** 3 B1SO3*N1S03 WBOS* * * * * * N|B01****** CELL4B BMBI****** N14S1*N1471 CELLBt B1BI N1S13-B1B33 B1512-N1502 N151I-CELL4S 81461 N146I*N1471 CELLB3 B1531****** B1933****** N1533****** NI52I...... NI39I-N1J81 CE1147 BU7|...... B1472****** NI471****** CELLS3 BIS3i****** B3331****** N1533****** B1B33*N1B03 B2832*NtB02 N2S32*N1S02 N1531'N243t N2S3I*N3431 CELLS? * B1S71**** a CELLSS BI981-CELLSO B1S91**** CELL60 B1601*"* CELL81 B1611**** BI612**** CELL64 1 B1641 • a B1642*8165* 3 4 i 6 CELL6B B16S1**** B16S2**** CELL66 B1661 • Bi<662****** CELLIt 1 B171I**** a CELL73 B1721**** CELL73 B1731**** CEIL74 B1741****** 81743*81783 CELL78 1 BI7B1****** a Bi7ea«'«*** a 4 • CELL79 B 1 7 0 , . . . . . . B i793* * * * * * CELLBO 81801 • «11802 -81783 CELL67 81671 B267i•••••• B3671* 81673*81693 83672-B1652 83673*81663 CELL79 B , 7 5 ) . . . . . . S17S3-***** CELLS 1 8)811 Basii**»*»» B3B1I****** 83813*81793 B38i3*B1793 B18I3*B17S3 CELL49 B148|.««««* B1482*N1413 N1482-NI413 N14BI-N1922 CELLS4 8194 1.«»».* N1S43****** BIS43*NI441 N1941-NI281 CELL63 B 1 6 3 1 * " * B1633**** CELL68 B168I • 'Bissa**** '**" CELL7C B176I B)762*BI792 CELL82 B1821****** Bi822*B17B2 N143I* C E L L 4 9 B 1 4 9 1 * * * * * * B I 4 9 3 * B 1 B 3 3 N1492*N1902 N1491-N1B22 C E L L 9 8 B I B B i * * * * * * N 1 S 9 3 * * * * * * B 1 5 5 3 * * * * * * N t s s i * N a 4 3 i CILLB8 SiBSI****** NIB63*N12«1 81862*81983 NI9«1*N1281 CELL63 81631 B1632*81843 CELL69 CELLTO B1691-**'** B1101****** Bi69a****** BiTda**''*** B170a CELL77 BI771****** B1773-B1793 CELLS? CILL94 BI831****** B1S41****** 81B32-B1843BJS43* ' * * * * * " 81833*81842 Note: I f there are f i v e a s t e r i s k s on r i g h t hand side of equation s i g n , then the sawlog on l e f t hand aide i s to be sawn. If there i s a sawlog code on RHS then t h i s i s the sawlog to be sawn instead of sawlog on LHS. cn cn 56 i n s t r u c t i o n s p r o v i d e s the base to the "sawing" procedure simulated by SAWSIM. A f t e r having sawn the sample sawlogs by SAWSIM the dynamic s i m u l a t i o n may s t a r t . The dynamic model must pic k the boom logs at random and corresponding to the freq u e n c i e s noted i n the boom log d i s t r i b u t i o n t a b l e (Appendix 5.1). A d d i t i o n a l l y , sawlogs o r i g i n a t i n g from the same boom l o g must flow through the model a group. This "pick-up" procedure i s d e s c r i b e d i n a step by step example as f o l l o w s : 1. S e l e c t one c e l l of the boom l o g pool (Appendix 5.6) by generating a unif o r m l y d i s t r i b u t e d random number between 1 and 10000. Say t h i s random number i s 200. Therefore, C e l l #2 i s the one from which the boom log w i l l be picked up. 2. S e l e c t one boom log randomly w i t h i n the c e l l s e l e c t e d in Step 1. Say t h i s i s N102 (Appendix 5.6). 3. Look up the sawing i n s t r u c t i o n s of boom l o g , s e l e c t e d in Step 2, in f i n a l boom l o g t a b l e (Figure 5.3). This i n s t r u c t s that only one sawlog N1021 i s cut from boom l o g N102 which i s s u b s t i t u t e d with sawlog B2122. (If the number of sawlogs to be cut from the same boom l o g i s l a r g e r than one, the dynamic model provi d e s that sawlogs w i l l flow in a c l u s t e r ) . 4. Look up the sawing inf o r m a t i o n of B2122 and use as input f o r the dynamic model. The f i n a l boom l o g t a b l e c o n t a i n s 316 sawlogs. Among these only 169 sawlogs have the "saw i t " i n s t r u c t i o n . T h i s means a s u b s t a n t i a l sawlog r e d u c t i o n of 47%. But here the f l e x i b i l i t y of t h i s sawlog s e l e c t i o n method has to be emphasized a g a i n . Varying the diameter and length c e l l boundaries as input of the s e l e c t i o n program makes i t p o s s i b l e to decrease (or in c r e a s e ) 57 the number of sample sawlogs to be sawn by SAWSIM and to decrease (or increase) s i m u l a t i o n accuracy. For i n s t a n c e , d e c r e a s i n g the number of diameter c e l l s in the sawlog s e l e c t i o n t a b l e (Appendix 5.10) from the o r i g i n a l 12 to 9, the r e d u c t i o n of sawlogs with "saw i t " i n s t r u c t i o n c o u l d be f u r t h e r decreased to 90. 5.2. EXPORT MARKETS, LUMBER SIZES AND PRICES Base lumber markets f o r the c u t t i n g program of Woodroom #3 are d e f i n e d by the marketing department (50). These are the f o l l o w i n g : • United S t a t e s - (U), • North A f r i c a and Belgium - (A), • Japan - (J and T ) , « United Kingdom and A u s t r a l i a - (K), and • France and United Kingdom - ( F ) . Among the above market codes i n p a r a n t h e s i s , (A), (K), and (F) r e f e r to two markets, si m u l t a n e o u s l y . Lumber s i z e s on the markets of North A f r i c a and Belgium, U n i t e d Kingdom and A u s t r a l i a , France and United Kingdom are the same, hence these markets are t r e a t e d a l i k e . On the other hand, there are two market codes i n connection with Japan, i n d i c a t i n g that two d i f f e r e n t s i z e s of lumber are demanded by the Japanese market: lumber f o r remanufacturing purposes (T) and lumber f o r the c o n s t r u c t i o n i n d u s t r y ( J ) . Table 5.1 i n c l u d e s lumber t h i c k n e s s e s , width and len g t h , r e f e r r i n g to markets by t h e i r codes. In t h i s t a b l e , data on 58 Table 5.1. Lumber s i z e s of v a r i o u s markets T a r g e t Market Thickness Width codes (inches) ( i n c h e s ) u 1 9/16 3 9/16, 5 5/8, 7 1/2, 9 1/2, 11 1/2 A 2 1/2 6, 7, 8, 9 4 1/8 / J 4 1/8 T 1 13/16 4, 6, 8, 10, 12 K 1 7/8 4, 6, 7, 8, 9 F 3 6, 8, 9, 10, 12 L e n g t h s ( f e e t ) u 8, 10, 12, 14, 16, 18 , 20, 22, * 24 * A 8, 10, 12, 14, 16, 18, 20, 22, 24, 26 J 10, 12, 13, 20 T 8, 10, 12, 14, 16, 18, 20, 22, 24, 26 K 8, 10, 12, 14, 16, 18, 20, 22, 24, 26 F 8, 10, 12, 14, 16, 18 , 20, 22, 24, 26 * Note: not f o r 2x4 t h i c k n e s s and width are always rough green t a r g e t s i z e s . The s i z e range, w i t h i n which a p i e c e of lumber can be accepted, i s give n by the maximum and minimum rough green s i z e . These were c a l c u l a t e d by a d d i n g / s u b t r a c t i n g v a l u e s to/from the rough green t a r g e t s i z e s , Table 5.2. Tabl e 5.2. Data to c a l c u l a t e maximum and minimum rough green s i z e s Thickness Width Maximum + 1/8" +1/8" Minimum - 1/16" -1/16" Among the export products o n l y the U.S.A. market r e q u i r e s s u r f a c e d lumber. In t h i s case, to c a l c u l a t e the rough green t a r g e t s i z e , the p l a n i n g allowance must be taken i n t o account. 59 T h e p l a n i n g a l l o w a n c e s c o n s i d e r e d f o r t h i c k n e s s a n d w i d t h a r e 2 / 1 6 " a n d 1 / 4 " , r e s p e c t i v e l y . Wane a l l o w a n c e , i . e . , t h e a l l o w a b l e wane f r a c t i o n o f t h e f i n i s h e d s i z e , f o r a l l m a r k e t s i s 1/3 f o r b o t h t h i c k n e s s a n d w i d t h . L u m b e r p r i c e s c o n s i d e r e d i n t h i s s t u d y a r e c o n s e r v a t i v e o n e s a n d c o r r e s p o n d t o t h e p r e v a i l i n g p r i c e s d u r i n g t h e f a l l o f 1 9 8 2 , when t h e l u m b e r m a r k e t f e l l t o i t s r e l a t i v e l o w e s t l e v e l o f t h e l a s t f o u r d e c a d e s . O n l y o n e s e t o f l u m b e r p r i c e s p e r m a r k e t w e r e r e c o m m e n d e d by t h e m a r k e t i n g d e p a r t m e n t f o r u s e i n t h e a n a l y s i s . T h e s e p r i c e s , w i t h t h e e x c e p t i o n o f t h e U . S . A . m a r k e t , p l u s t h e b y p r o d u c t p r i c e s , a r e i n T a b l e 5 . 3 ? T a b l e 5 . 3 . L u m b e r a n d b y p r o d u c t p r i c e s ( C a n a d i a n $) M a r k e t c o d e s A J T K F P r i c e s ( $ / M F B M ) 350 340 245 300 390 C h i p S a w d u s t P l a n e r s h a v i n g s D r y t r i m m i n g s S h o r t R e s i d u a l ( * ) l o q r o u n d w o o d 82 28 v a l u e 28 ( $ / c u n i t o f 28 s o l i d wood) 82 82 ( * ) H e a d i n g s c o r r e s p o n d t o SAWSIM t e r m i n o l o g y . F o r t h e U . S . A . m a r k e t l u m b e r p r i c e s b a s e d o n p r o d u c t i o n d a t a o f M a y / 1 9 8 3 p r o v i d e d by m a n a g e m e n t , w e r e c a l c u l a t e d a s a w e i g h t e d a v e r a g e f o r e a c h s i z e c o m b i n a t i o n . B a l a n c e a n d s u r p l u s p r i c e s w e r e c o n s i d e r e d when c a l c u l a t i n g a v e r a g e s w e i g h t e d by t h e c o r r e s p o n d i n g p r o d u c t i o n d a t a . The a v e r a g e l u m b e r p r i c e s ( n e t m i l l p r i c e s ) f o r t h e U . S . A . m a r k e t i n $ / M F B M , a r e i n T a b l e 5 . 4 . 60 Table 5.4. Lumber p r i c e s of the U . S . A . market ( C A N $) 8 L 10 E 12 N 14 G T H ( f t ) 16 18 20 22 24 dimen-t i o n a l s i z e s (nomi-n a l ) 2 x 04 2 x 06 2 x 08 2 x 10 2 x 12 185 187 175 163 199 179 145 176 151 199 163 169 193 165 187 145 157 229 241 241 205 217 263 225 266 175 172 164 210 174 197 199 226 203 199 205 229 228 260 261 277 183 187 5.3. MACHINERY A c c o r d i n g to the m o d i f i c a t i o n design of Woodroom #3, the c o n v e r s i o n f a c i l i t i e s used f o r c r o s s c u t t i n g and f o r the primary breakdown w i l l remain the same as ones used i n the o l d m i l l . Other components of m i l l machinery from the h e a d r i g s to the back end of the m i l l , w i l l be r e p l a c e d by new ones. T h i s new machinery c o n s i s t s of the f o l l o w i n g u n i t s : • two I2"x66" combination machines operable i n both "gang" ( f o r cen t e r c a n t s ) and "edger" ( f o r sideboards) mode, • one I2"x66" r o t a r y gang edger, • one twin band v e r t i c a l resaw, • one 4" edger, and • two 26' multi-saw mechanical l i f t trimmers. The m o d i f i c a t i o n p l a n a l s o i n c l u d e s an i n s t a l l a t i o n of new t r a n s p o r t a t i o n equipment, c r o s s c h a i n conveyors, b e l t s and r o l l c a s e s . The l a y o u t of m i l l machinery i s d i s p l a y e d i n F i g u r e 5.4. The p r e v i o u s l y mentioned machines and t r a n s p o r t a t i o n equipment, between the headrigs and trimmers, c o n s t i t u t e the c r i t i c a l p a r t of m i l l r edesign on which the s i m u l a t i o n 91.25' "3 (-•• iQ C «-t a> cn 63.75' 0* *< O c CD • i o •d o in a. s o a o 0) cr o 3 •o M D> 3 85' 1 1 47.5' . I M 42.5' • 1 J • »' .1 "1 62 experiments are to be focused. Thus, s p e c i f i c a t i o n s of only these u n i t s w i l l be d i s c u s s e d below. However, f o r the sake of a complete l i s t i n g of the new co n v e r s i o n f a c i l i t i e s , a d d i t i o n a l equipment such as one new l o g feeder crane at the j a c k l a d d e r i n f e e d , two 60-bin J-bar s o r t e r s with s t a c k e r s , packaging f a c i l i t i e s , one drum-type h o r i z o n t a l c h i p p e r , and one h y d r a u l i c p l a n e r have to be mentioned. Table 5.5. Machine s p e c i f i c a t i o n data Machines gap lo a d 1 ines kerf speed time t ime /pass (inch) (ft/min) (sec) (sec) 9' Headrig 8* Headrig 0.250 400 5 1 0 1 0.250 400 5 10 1 Comb.machine #1 0.160 350 5 5 (as edger) Comb.machine #1 0. 160 125 6 (as gang-edger) Comb.machine #2 0. 160 350 5 5 (as edger) Comb.machine #2 0. 160 125 6 (as gang-edger) Gang 0. 160 1 25 10 19 Twin 0. 180 250 4 2 Edger 0.250 300 4 5 Tr immers 2 Machine s p e c i f i c a t i o n data were given by management and summarized i n Table 5.5. The meaning of kerf and machine speed data are s t r a i g h t f o r w a r d . Gap and l o a d time, as w e l l as maximum number of sawlines per pass are used to c a l c u l a t e the p r o c e s s i n g time on a machine a c c o r d i n g to the SAWSIM Manual (25): 63 no.of sawlines a c t u a l l y made No.of passes = — max.no.of sawlines per pass length of l o g Time per pass = + gap time l i n e a l machine speed Machine time = time per pass * no.of passes + load time Gap time i s the time between each pass of a p i e c e through a machine or the time between p i e c e s i f only one pass i s allowed per p i e c e . Load time i s the a d d i t i o n a l time r e q u i r e d between p i e c e s f o r a primary breakdown machine. For the purpose of a c o n c i s e overview, machine breakdown data are d i s p l a y e d here i n Table 5.6 but d i s c u s s e d l a t e r i n d e t a i l , among the model b u i l d i n g concepts, i n Chapter 6. T r a n s p o r t a t i o n equipment s i z e data are summarized i n Appendix 6.4. Note that some t r a n s p o r t a t i o n equipment have data with more than one l e n g t h data, because p i e c e s can be conveyed at d i f f e r e n t lengths along t r a n s p o r t a t i o n equipment. For i n s t a n c e , the f i v e lengths f o r Chain 9 are 8.75', 42.5', 80', 100' or 115' depending on whether a p i e c e i s conveyed by Chain 9 to Unscrambler 2 from e i t h e r R o l l 13, B e l t 2, 3, or 1. The maximum number of lumber p i e c e s that can t r a v e l s i d e by side and the average t h i c k n e s s of the lumber " l a y e r " , as data a f f e c t i n g t r a n s p o r t a t i o n equipment c a p a c i t y , must a l s o be given. Table 5.7 c o n t a i n s these data. Turning to speeds of the t r a n s p o r t a t i o n equipment, the new u n i t s , on request of m i l l management have designed speeds 25 fe e t per minute higher than the preceding machine's i n f e e d speed given by the machine manufacturing companies. Common . to the 64 Table 5.6. Machinery breakdown data H e a d r i q s 9' 8' Down time Working time Down time Working time per f a i l u r e per f a i l u r e E m p i r i c a l E x p o n e n t i a l E m p i r i c a l / Exp o n e n t i a l d i s t r i b u t i o n d i s t r i b u t i o n d i s t r i b u t i o n d i s t r i b u t i o n Average:54.7min Average: Average:54.1min Average: Standard 13523 sec Standard 13491 sec deviation:37,9min deviation:38.2min Numb.of observ ' s : (225.4 min) Numb.of observ's: (224.8 min) 376 376 O t h e r m a c h i n e s Down Work E m p i r i c a l d i s t r i b u t i o n Average:57.4min Standard deviation:37.9min Exponent i a l d i s t r i b u t i o n Average: 185471 sec (3091.2 min) Note: These data are based on the recommendation of m i l l management that the average amount of down time i s 2 hours per 15 s h i f t s and that the down-time d i s t r i b u t i o n s are the same as that of Headrig 9'. T r a n s p o r t a t i o n e q u i p m e n t B e l t Down time per f a i l u r e Working time U n i f o r m l y E x p o n e n t i a l d i s t r i b u t e d d i s t r i b u t i o n over the range of Average: 413100 sec 15 ± 1.5 min (6885 min) Chains/Rolls/Unscramblers Down time per f a i l u r e U n i f o r m l y d i s t r i b u t e d over the range of 30 ± 3 min Working time E x p o n e n t i a l d i s t r i b u t i o n Average: 412200 sec (6870 min) Note: These data are based on the recommendation of m i l l management that the average amount of down-times are 15 min ± 10%/15 s h i f t s and 30 min ± 10%/15 s h i f t s f o r b e l t s and other t r a n s p o r t a t i o n equipment, r e s p e c t i v e l y . 65 Table 5.7. Number of p i e c e s which can t r a v e l s i d e by s i d e and the average t h i c k n e s s of lumber l a y e r T r a n s p o r t a t i o n e q u i p m e n t B i t ! B l t 2 B l t 3 B l t 4 B l t 5 Rol5 Rol7 R 0 I 8 RollO Rol12 R o l l 3 Max.no. of 20 20 20 3 3 20 20 20 5 3 4 p i e c e s that may / be conveyed s i d e by s i d e . T r a n s p o r t a t i o n e q u i p m e n t Chain8 Chain9 Chain12 C h a i n l 3 Chain14 Chain!5 Average t h i c k n e s s of 5 5 5 5 5 5 lumber l a y e r (inches) p i e c e flow design, c r o s s chain speeds are 10 inches / s e c . The speed data are summarized i n Appendix 6.4. Unscrambler s p e c i f i c a t i o n s are d e f i n e d by the maximum number of p i e c e s which can be conveyed simultaneously and the time r e q u i r e d f o r them to flow through. These data are 8 p i e c e s and 6 seconds. F i n a l l y , there i s one more type of t r a n s p o r t a t i o n equipment l e f t which needs s p e c i a l a t t e n t i o n . The s k i d t r a n s f e r s can be thought of as common areas of equipment t r a n s p o r t i n g p i e c e s i n two d i f f e r e n t d i r e c t i o n s p e r p e n d i c u l a r to each other. In the m i l l l a y o u t , i n Fi g u r e 5.4, s k i d t r a n s f e r s are coded by "COM1" and "COM2". T h e i r s i z e , i . e . , the length to t r a v e l along, and t h e i r conveying speed depend on whether p i e c e s are conveyed p a r a l l e l to p i e c e l e n g t h (towards r o l l s ) or p e r p e n d i c u l a r (towards c h a i n s ) . T h e i r v a r y i n g s i z e and speed data are i n Appendix 6.5. 66 5.4. DESCRIPTION OF MILL OPERATION A f t e r s a l t water storage, boom logs of western hemlock are e l e v a t e d to the m i l l f l o o r by a jack ladder where they are bucked by two c u t - o f f saws to a p p r o p r i a t e lengths f o r f u r t h e r manufacturing. A f t e r the bark has been removed by one of the two b a r k e r s , sawlogs a r r i v e at the h e a d r i g logdeck from where they are d i s c r i m i n a t e l y sent to one of the two band-carriage h e a d r i g s . Primary breakdown of r e l a t i v e l y s m a l l e r logs i s c a r r i e d out by the Headrig 8', whereas logs of l a r g e r diameter go to the Headrig 9'. However, t h i s r u l e can be o v e r r i d d e n as a f u n c t i o n of queue length b u i l d i n g up i n f r o n t of h e a d r i g s . At the two headrigs logs are o r i e n t e d and t r a n s p o r t e d by b a l l screw c a r r i a g e s while they are sawn by the c u t t i n g system. T h i s c u t t i n g system c o n s i s t s of a s i n g l e h e a d r i g s l a b chipper to perform s l a b removal and a s i n g l e cut band saw to cut only i n the forward d i r e c t i o n . Both q u a r t e r and l i v e sawing can be c a r r i e d out. For a l l the markets of t h i s study, logs are s p l i t -taper-cant-sawn, in "horns-up" p o s i t i o n . When the m i l l produces f o r the Japanese 4x4 market, logs are q u a r t e r sawn. The combined machines can be operated i n e i t h e r gang edging mode to break down center c a n t s , or edging mode to process r e l a t i v e l y small sideboards before trimming. On the combination machines, l i g h t l i n e s h e l p p i e c e p o s i t i o n i n g . The pre-p o s i t i o n i n g f e a t u r e of the combined machines ensures that both cants and sideboards are o r i e n t a t e d i n such a way that t h e i r 67 edge i s p a r a l l e l to the sawing plane. Using sawmill terminology, cants are f u l l taper sawn. The i n s t a l l a t i o n of the Gang Edger was designed with two o b j e c t i v e s . F i r s t , i t was to process r e l a t i v e l y low grade cants i n t o dimension s i z e lumber when the m i l l i s i n export mode, that i s i t produces f o r both overseas and domestic markets. Second, in the dimension mode i t i s to operate as a "lo a d l e s s e n i n g " and back up machine to ensure smooth, u n i n t e r r u p t e d o p e r a t i o n when the combined machines are overloaded or down. Hence i t serves the same purpose as the combined machines i n gang edging mode. Cant p o s i t i o n i n g i s the same as on combination machines. The primary task of the twin band resaw i s to process p i e c e s developing at headrigs or combined machines r e q u i r i n g s p e c i a l i z e d c u t t i n g s . I t a l s o serves as a back up machine f o r edging purposes when combined machines are down. The edger i s used to r i p p i e c e s lengthwise i n t o f i n a l width before trimming. Pieces from any of the preceding machines can be processed to remove waned edges. Sideboards, on both twin saw and edger, are p o s i t i o n e d p a r a l l e l to t h e i r center l i n e . To square the ends and form the f i n a l l e n g t h of lumber they are c r o s s c u t at one of the two m u l t i p l e saw trimmers depending on the workload of the bin s o r t e r s behind them. Three major groups of t r a n s p o r t a t i o n equipment are used to f a c i l i t a t e the movement of lumber w i t h i n the m i l l : • r o l l e r s , • b e l t s , a n d • c h a i n s . The f i r s t two types of t r a n s p o r t a t i o n equipment are used to 68 move lumber p a r a l l e l to i t s length whereas chains serve to move lumber i n t r a n s v e r s e d i r e c t i o n s . 69 6^ APPROACH TAKEN IN BUILDING THE MODEL There are two major f a c t o r s a f f e c t i n g m i l l performance: how logs are cut, and how many logs can be cut over a c e r t a i n p e r i o d of time. Thus, the computer s i m u l a t i o n of a sawmill must mimic • l o g breakdown and • pi e c e flow as a f u n c t i o n of time. Furthermore, these two f a c t o r s are s t r o n g l y interdependent. The way a l o g i s cut a f f e c t s p i e c e flow. Hence, the dynamic model, s i m u l a t i n g p i e c e flow, must e i t h e r i n c l u d e a l o g breakdown l o g i c , or have access to one. B u i l d i n g a log breakdown l o g i c i s d i f f i c u l t . The grea t e r the requirement f o r accuracy; the more time i s needed to write the program. Consequently, e x i s t i n g dynamic models have s i m p l i f i e d l o g breakdown l o g i c and simulate the log as a tr u n c a t e d cone. To a v o i d the time consuming procedure of w r i t i n g a l o g breakdown program, and to ensure access to a more accurate l o g breakdown l o g i c , an e x i s t i n g sawing s i m u l a t i o n package (SAWSIM) i s used i n t h i s study. SAWSIM, designed by H.A. Leach and Company L t d (25), simulates l o g breakdown more a c c u r a t e l y than any other model w r i t t e n i n the p a s t . In a d d i t i o n to the usual diameter, uniform taper and len g t h l o g shape c h a r a c t e r i s t i c s , SAWSIM takes i n t o account e l l i p t i c a l c r o s s s e c t i o n s , sweep and v a r i a b l e taper along the l e n g t h , and hence p r o v i d e s the most 70 accurate l o g d e s c r i p t i o n f o r the computer s i m u l a t i o n . SAWSIM was o r i g i n a l l y w r i t t e n i n CDC ( C o n t r o l Data C o r p o r a t i o n ) FORTRAN f o r l a r g e computers. I t c o n s i s t s of over 25,000 executable FORTRAN statements. U n f o r t u n a t e l y , there are i n c o m p a t i b i l i t i e s between IBM and CDC FORTRAN. ' Because of these f a c t s , to run SAWSIM on the UBC AMDAHL V6-II computer (IBM/370-e q u i v a l e n t machine), c o n v e r t i n g the o r i g i n a l FORTRAN d i a l e c t was necessary. T h i s i n v o l v e d a great amount of work. Log breakdown l o g i c SAWSIM SAWSIM output Problem t>|of data t r a n s m i s s i o n FLOWSIM input Dynamic p i e c e flow $| s i m u l a t i o n model FLOWSIM Fi g u r e 6.1. Exogenous l o g breakdown l o g i c r e q u i r i n g data t r a n s m i s s i o n In order to use SAWSIM r e s u l t s as an exogenous source of in f o r m a t i o n f o r the dynamic model, the problem of data t r a n s m i s s i o n between the two models had to be so l v e d ( F i g u r e 6.1). The dynamic model needs s p e c i a l data i n p u t , i n a s p e c i a l s t r u c t u r e . SAWSIM, based on inputs commonly prepared, does not prov i d e the data i n the format r e q u i r e d . Hence, some t r i c k y SAWSIM input p r e p a r a t i o n was needed. S p e c i f i c a l l y , the machine data and sawing p a t t e r n codes r e q u i r e d s p e c i a l a t t e n t i o n d u r i n g input p r e p a r a t i o n . A c c o r d i n g l y , t h i s chapter i s d i v i d e d i n t o two p a r t s . The f i r S t d e s c r i b e s the problem of data t r a n s m i s s i o n and i t s s o l u t i o n , and the second d e s c r i b e s the design concepts 71 of the dynamic model . 6.1 THE PROBLEM OF DATA TRANSMISSION The data t r a n s m i s s i o n between SAWSIM and FLOWSIM r a i s e s the f o l l o w i n g q u e s t i o n : What i n f o r m a t i o n i s needed by FLOWSIM to c a r r y out the p i e c e flow s i m u l a t i o n ? The input da ta of FLOWSIM must i n c l u d e : • the codes of machine c e n t r e s needed to p roce s s a p a r t i c u l a r l o g , • the number of " o f f s p r i n g " p i e c e s at each machine c e n t r e , • the address of the next machine c e n t r e i . e . , route dec i s i o n s . • In format ion which d e f i n e s the unambiguous sequence of o p e r a t i o n s in t i m e . T h i s i s to prevent absurd cases such as the p r o c e s s i n g of a l og on the h e a d r i g f o l l o w i n g the p r o c e s s i n g of the board o r i g i n a t i n g from t h i s l o g . How can SAWSIM p r o v i d e the above i n f o r m a t i o n ? The way SAWSIM s i m u l a t e s sawing, i s s i m i l a r to the p roce s s of sawing a l o g fo r lumber, and can be thought of as a s e r i e s of machine o p e r a t i o n s . In SAWSIM t h i s p roce s s i s d e s c r i b e d by a s e r i e s of data l i n e s . Each group of data l i n e s , c o r r e s p o n d i n g to a machine o p e r a t i o n , i s c a l l e d a sawing p a t t e r n . Ju s t as a sawing proces s s t a r t s w i t h the o p e r a t i o n on the l o g , the f i r s t p a t t e r n i s a l o g p a t t e r n , d e s c r i b i n g the pr imary breakdown of a l o g . Subsequent p a t t e r n s are cant p a t t e r n s d e s c r i b i n g o p e r a t i o n s l e a d i n g to f i n a l p r o d u c t s . Thus a l o g p a t t e r n and s e v e r a l cant p a t t e r n s d e s c r i b e the complete sawing p r o c e s s . 72 Going i n t o f u r t h e r d e t a i l , a sawing p a t t e r n c o n s i s t s of a s e r i e s of type 1, 2 and 3 records ( l i n e s ) . A type 1 record i d e n t i f i e s the p a t t e r n , the machine used, and the p o s i t i o n of the l o g (or cant) r e l a t i v e to the sawlines. A type 2 rec o r d s p e c i f i e s the saw spacing i . e . , the d i s t a n c e between two p a r a l l e l sawlines. A type 3 rec o r d d e s c r i b e s what w i l l happen to the p i e c e s generated by t h i s p a t t e r n . These p i e c e s are e i t h e r manufactured i n t o f i n a l products by edging, resawing and trimming, or pi e c e s are generated from the cant on an a p p r o p r i a t e machine. According to t h i s a type 3 rec o r d e i t h e r c a l l s a "standard p a t t e r n " which i s always the l a s t one i n the p a t t e r n "chain", or a cant p a t t e r n which can c a l l another standard p a t t e r n . In f i e l d 3 of type 1 and f i e l d s 7-11 of a type 3 rec o r d the codes f o r the machine c a r r y i n g out the r e q u i r e d o p e r a t i o n are w r i t t e n . During program execution the number of passes on each machine i s counted and the o p e r a t i o n time i s c a l c u l a t e d . F i n a l l y , SAWSIM pr o v i d e s data which d e s c r i b e the machines which are needed to process the l o g , and the number of passes necessary on each machine. I t a l s o i d e n t i f i e s how long the op e r a t i o n takes. If the r e a l machines were coded i n the sawing p a t t e r n i . e . , 9' Headrig, 8' Headrig, Combined Machines #1 and #2, Gang Edger, Twin Saw, Edger and Trimmer, then the SAWSIM output would not provide the r e q u i r e d i n f o r m a t i o n f o r the dynamic model. To i l l u s t r a t e t h i s , a SAWSIM run was made, the output of which i s d i s p l a y e d in F i g u r e 6.2. T h i s SAWSIM output does not provide i n f o r m a t i o n about the number of a d d i t i o n a l p i e c e s generated at each machine c e n t e r , 73 2 , O ? ! T lirlTL -2 "'"-.I °i* l , M W € * "*»««," <**•< «s * • » « • / CHIP LIME* • m o COST MOSS * / VL ul T, ~ " i 0 C f C»L ACTUAL LM LM COFT COFT I • « * CUNIT «%» ua / i uMoa O.BOO o o i«.oo aa.ao o. 110 o.o o o * r . i u u> i o . a a • • toa <s T.JJ O.O IW.IOSHJ 3 ME0DG9I HECWeai C0WLV.1 COMQN01 CC-EOCJ CtMCMSa QAMCCOC TWIMSAW EDOER TRIMME* CHIPPER «% PASSES 3 •Ot SECONDS is. <X * » ' a 13 a «• IT. IT. ac. SAWSIM PLOT ua / i uHoa axc« i o ax t o i s a xio aa •I-a xio aa ax t o aa ONE INCH I axoa aa ax i o aa ax i o aa ax i o aa a X»1 0 4 4 I a • x 1 | 0 « • 4 • F i g u r e 6 . 2 . SAWSIM o u t p u t u s i n g m a c h i n e c o d e s t h e a d d r e s s o f t h e n e x t m a c h i n e c e n t e r , o r t h e t i m e s e q u e n c e o f o p e r a t i o n s . A s o l u t i o n t o t h i s p r o b l e m i s n e e d e d . T h e k e y c o n c e p t l e a d i n g t o t h e s o l u t i o n o f t h e a b o v e p r o b l e m i s t h e f o l l o w i n g . T h e m a c h i n e c o d e s i n t h e i n p u t f i l e o f s a w i n g p a t t e r n s a n d m a c h i n e d a t a , a r e t o be r e p l a c e d by r o u t e c o d e s . S u c h r o u t e c o d e s c o n s i s t o f t h e c o d e o f t h e " f r o m " m a c h i n e a n d 74 " t o " machine. F i g u r e 6 . 3 . Network of p i e c e r o u t e s Obviously there must be as many route codes as there are routes e x i s t i n g i n the m i l l l a y o u t . To v i s u a l i z e these routes, the network c h a r t of F i g u r e 6 .3 can h e l p . On t h i s c h a r t the 75 number of machines i s 13, whereas the a c t u a l number of machines in the redesigned m i l l segment i s only 8. The f i r s t reason f o r t h i s d i f f e r e n c e of 5 machines i s that the edger and gang components of the combined machines are d i s t i n q u i s h e d (CE1, CG1, CE2, CG2). Another reason i s that i n order to simulate the d i f f e r e n t routes by which p i e c e s can be taken to the Twin Saw, the i n s t a l l a t i o n of dummy twin saw machines i s necessary. Thus, the twin (TWN), and " v i a d r o p s o r t e r " twin (DTW) simulate the resaw o p e r a t i o n depending on whether p i e c e s flow d i r e c t l y or through the d r o p s o r t e r to the Twin Saw (F i g u r e 3.1). Pi e c e s which are r e d i r e c t e d to the Twin Saw f o r an a d d i t i o n a l o p e r a t i o n , are manufactured on the dummy twin (TWD) and dummy twin#1 (TD1) machines. In the example there are 34 p o s s i b l e routes represented by arrows. These 34 route codes were taken i n t o c o n s i d e r a t i o n when the input f i l e of sawing p a t t e r n s and machine data were set up. The r e s u l t s from sawing the same l o g by the same technology as was used b e f o r e , but a c c o r d i n g to new route codes, are presented in F i g u r e 6.4. Based on the SAWSIM output of Fi g u r e 6.4, a matrix (DIA -Dynamic Information Array) can be b u i l t i n con n e c t i o n with each sample l o g . T h i s matrix i s a r e c t a n g u l a r matrix having as many columns as there are machines i n v o l v e d i n the network c h a r t . Thus, the number of columns i s 13. The number of rows i s the number of columns plu s one. The a d d i t i o n a l l a s t row c o n t a i n s the machine time data as c a l c u l a t e d by SAWSIM. The matrix corresponding to the above SAWSIM output i s presented in Fi g u r e 6.5. 76 <X LOG SAWING — OFFSET — DIA LEN TAPER —SWEEP— CUFT PCS FBM F M / CHIP LUMBER BYPRD COST CROSS XI <X I Of NT PATTERN HOR VtR IN FT SN/FT EDGE CYL ACTUAL LBS LBR CUFT CUFT S S t S CUNIT <X> U2 / I UM02 O.5O0 0.0 IB .0 0 32.BO 0-110 0.0 0.0 BT.1 IS 383 10.S3 C .B 103.45 7.31 0.0 110.75 398 8 <% HCDRG91 HE DOG* 1 HR8-CE1 HR8-C01 HR9-0N6 HR9-TWN HRS-CE3 HR8-CS3 HRS-ONG HRB-TwW CE1-TRM CGI "TWN C01-DTW CG1-EDG CS1-TRM <% PASSES 3 3 1 <X SECONDS 3 3 . 48. tT. <% S <% CE3-TRM C62-TWN CC3-0TW CQ3-EDG CG3-TRM ONC-TWN 8NG-0TW BNC-E06 OMG-TRH TWN-TWD TWN-EDG TWN-TRM DTW-TWO OTW-EOG OTW-TRM <* PASSES 4 2 7 <% SECONDS 8 . IT. 14. <% S <Z TW0-TD1 TWO-EOG TWD-TRM TD1-EDG TO1-TRM EOO-TRM CHIPPER <S PASSES a 2 <X SECONDS 4 <% t SAWSIM PLOT U3 / 1 UH03 ONE INCH 2X08 10 3X10 16 2X10 22 2X10 22 3X10 22 2X10 22 2X10 22 2X10 22 • I 2X0S 23 2 X ' l O 4 4 I 2 • X 1 I 0 6 * 4 • -« • • * F i g u r e 6.4. SAWSIM output using route codes i n s t e a d of machine codes 77 1 HR9 2 HR8 3 CE1 4 CG1 5 CE2 6 CG2 7 GNG 8 TWN 9 DTW 10 TWD 1 1 TD1 12 EDG 13 TRM 1 HR9 2 HR8 3 1 3 CE1 4 CGI 5 CE2 4 6 CG2 2 7 7 GNG 8 TWN 9 DTW 1 0 TWD 1 1 TD1 12 EDG 2 13 TRM Machine t ime | 33 16 17 9 2 F i g u r e 6.5. DIA matrix of l o g U2/1 From t h i s matrix the necessary i n f o r m a t i o n can be read as f o l l o w s . The codes of a l l machines, corre s p o n d i n g t o e i t h e r a row or a column having at l e a s t one element not equal to zero, d e f i n e the machines needed t o process the l o g . In the example, the machines t h a t p a r t i c i p a t e i n p r o c e s s i n g l o g U2/1 are 8' Headrig (HR8), Combined Machine #2 i n edger mode (CE2) and i n gang mode (CG2), Edger (EDG) and Trimmer (TRM). The number of a d d i t i o n a l p i e c e s generated at a p a r t i c u l a r machine can be c a l c u l a t e d by adding the elements of a row i n con n e c t i o n with t h i s machine. For i n s t a n c e , t h i s number i s 4 (3+1)in the HR8 row, counting the 3 sideboards and the cant ( F i g u r e 6.4). The non-zero elements of the DIA matrix d e f i n e the route of a p i e c e i n such a manner t h a t the address of the next machine i s the code of the column where the number i n q u e s t i o n i s l o c a t e d . 78 For i n s t a n c e , the number 2 i n the row CG2 and i n the column EDG means that 2 p i e c e s take the route from Combined Machine #2 to the Edger. F i g u r e 6.4 shows that there are only two p i e c e s produced from the center cant, needing edging o p e r a t i o n . The l a s t important i n f o r m a t i o n which can be gained from the DIA matrix i s the time sequence of o p e r a t i o n s . In order to get the r i g h t time sequence, data must be read from the DIA matrix a c c o r d i n g to the f o l l o w i n g r u l e s . I t i s always one of the he a d r i g s which c a r r i e s out the f i r s t o p e r a t i o n of p r o c e s s i n g . In t h i s example, i t i s the 8' Headrig. The machine codes f o r the next o p e r a t i o n s are d e f i n e d by columns having non-zero elements i n the corresponding row. In DIA matrix of log U2/1, these machines are the Combined Machine #2 i n edging and gang mode. The machines of these columns (CE2, CG2) are succeeded by those machines whose codes can be found i n the columns of non-zero elements of the rows of these (CE2, CG2) machines. In the row of CE2 the only non-zero element i s 4 i n column of TRM, whereas i n the row of CG2 there are 2 and 7 i n columns of EDG and TRM, r e s p e c t i v e l y . Thus, p i e c e s from CE2 go to Trimmer only, whereas from CG2 2 p i e c e s go to Edger and 7 to Trimmer. F i n a l l y , t h i s reading process i s terminated by the l a s t machine which i s the trimmer in t h i s example. The number of DIA ma t r i c e s to which FLOWSIM must have access d u r i n g s i m u l a t i o n i s f a i r l y l a r g e . The l a r g e r the number of sample logs and the number of ways of sawing these l o g s , the l a r g e r the number of DIA ma t r i c e s which have to be s t o r e d i n computer memory. In a d d i t i o n , t a k i n g i n t o account the number of 79 r e f e r e n c e s to DIA matrices throughout the model of FLOWSIM, and that these r e f e r e n c e s are invoked whenever a pie c e of lumber ( t r a n s a c t i o n ) flows through the model, the problem of data t r a n s m i s s i o n i s obvious. FLOWSIM must have an e f f i c i e n t access to DIA ma t r i c e s and to t h e i r elements d u r i n g the dynamic s i m u l a t i o n . Before c o n t i n u i n g the d i s c u s s i o n on the problem of data t r a n s m i s s i o n between FLOWSIM and data base #2, the general DIA matrix content must be e x p l a i n e d . F i g u r e 6.6 d i s p l a y s a l l the p o s s i b l e l o c a t i o n s where non-zero e n t r i e s may occur. In the f i r s t 12 rows, these p o s s i b l e l o c a t i o n s are marked by "a", meaning the number of p i e c e s flowing from machine " i " to machine " j " . Since p i e c e s from Trimmer go nowhere, row 13 (TRM) i s used to s t o r e s p e c i a l i n f o r m a t i o n on l o g diameter and le n g t h , market code, number of p i e c e s to be cut from the sawlog, e t c . These are r e l e v a n t to c o n t r o l FLOWSIM o p e r a t i o n and are marked by " s " . The l a s t row c o n t a i n s data of the machine time r e q u i r e d to process one p i e c e . These are marked by " t " . Turning back to the problem of data t r a n s m i s s i o n again, to manage communication between Data Base #2 and FLOWSIM e f f i c i e n t l y , there are two p o i n t s to c o n s i d e r . F i r s t , the d e n s i t y of DIA ma t r i c e s , i . e . , the r a t i o of non-zero e n t r i e s to the t o t a l number of matrix elements i s low. The average d e n s i t y i s 9-10 % (17/182). Thus Data Base #2 should c o n t a i n only non-zero elements of DIA m a t r i c e s . Second, Data Base #2 must be s t r u c t u r e d i n such a way that only a subset of Data Base #2 need be examined to f i n d the r e q u i r e d element of a p a r t i c u l a r matrix. In accordance with the above c o n s i d e r a t i o n s the approach 80 \ " t o " \ ( j ) ( i ) \ 1 HR9 2 HR8 3 CE1 4 CG1 5 CE2 6 CG2 7 GNG 8 TWN 9 DTW 10 TWD 1 1 TD1 12 EDG 13 TRM 1 HR9 a a a a 2 HR8 a a a a 3 CE1 a 4 CG1 a a a a 5 CE2 a 6 CG2 a a a a 7 GNG a a a a 8 TWN a a a 9 DTW a a a 10 TWD a a a 11 TD1 a a 12 EDG a 13 TRM s s s s s s s s Machine t ime t t t t t t t t t t t t t F i g u r e 6.6. General DIA matrix s t r u c t u r e taken i n d e s i g n i n g the data t r a n s m i s s i o n i s the f o l l o w i n g . Whenever FLOWSIM needs the value of a p a r t i c u l a r element of a p a r t i c u l a r DIA matrix, a subroutine i s invoked by a BCALL block which r e t u r n s the r e q u i r e d v a l u e . The name of the subroutine i s the e x t e r n a l ampervariable &DIA. I t s 4 arguments are the r e t u r n e d value (&VALUE), row number of the DIA matrix, column number of the DIA matrix, and the DIA matrix number.(Appendix 6.1). To understand the s t r u c t u r e of Data Base #2 from which the DIA s u b r o u t i n e p i c k s up the r e q u i r e d matrix element, the f o l l o w i n g example might be h e l p f u l : Assume that FLOWSIM at a c e r t a i n p o i n t needs to know the value of the DIA matrix #2 i n i t s Row 5 and Column 11. Based on these arguments "DIA" sub r o u t i n e f i n d s the p o i n t e r of DIA matrix #2 which i s LOCATN(2)=12, and the p o i n t e r of the next DIA 81 i LOCATN ( i ) k ROWIND(k) COLIND(k) VALUE(k) 1 1 1 ... • • 2 • • • ... • • • • • 3 • 2 • 1 2 • 12 • ... • • 19 5 1 1 4 • 3 • • 26 • • 26 • • • • • • • • • • • • • • • • • • • * • « •> • • • • • F i g u r e 6.7. S t r u c t u r e of Data Base #2 ma t r i x : LOCATN(3)=26. The range of k, i n which the wanted VALUE(k) must be, i s [LOCATN(2),(LOCATN(3)-1) ]=[12,25]. So. the DIA s u b r o u t i n e looks up only t h i s range, stops when both row and column numbers match, f i n d s k=l9, then p i c k s up VALUE(k)=4 and r e t u r n s i t f o r FLOWSIM as an e x t e r n a l v a r i a b l e (&VALUE). As a r e s u l t of the above data s t r u c t u r e both the computer memory demand and the CPU time needed to f i n d a given entry are low and p r o v i d e e f f i c i e n t data t r a n s m i s s i o n . 6.2. THE CONCEPTS OF DYNAMIC MODEL CONSTRUCTION The a c t u a l s i m u l a t i o n model, and the computer program are w r i t t e n i n GPSS/H, the newest and most developed d i a l e c t of the GPSS language. GPSS i s an acronym f o r "General Purpose S i m u l a t i o n System". "H" stands f o r J.O. Henriksen, the developer of the compiler at Wolvarine Software C o r p o r a t i o n . G e n e r a l l y , the way the model i s s t r u c t u r e d f o l l o w s the t y p i c a l input deck format of GPSS (62), Table 6 . 1 . The f o l l o w i n g s e c t i o n s d e s c r i b e the concepts of model 82 Table 6.1. T y p i c a l groups of GPSS cards and the corresponding p a r t s of FLOWSIM Part •Compiler d i r e c t i v e s • E n t i t y d e f i n i t i o n and i n i t i a l c o n t r o l statements •Block statements •Report statements 2,3,4,5.1-5.5 5.6,6,8,9 10 b u i l d i n g . To make t h i s d e s c r i p t i o n e a s i e r and improve program r e a d a b i l i t y , major p a r t s of the program l i s t i n g (Appendix 6.1) are numbered w i t h i n comments, s t a r t i n g with "*". They should not be confused with numbers under the headings of LINE#, STMT# and BLOCKS. Throughout the f o l l o w i n g s e c t i o n s r e f e r e n c e i s made to p a r t s of the program l i s t i n g i n b r a c k e t s , so the reader can study the model d e s c r i p t i o n and model l i s t i n g t o g e t h e r . To show the correspondence between t y p i c a l groups of GPSS cards and p a r t s of FLOWSIM, numbers r e f e r r i n g to these p a r t s are d i s p l a y e d on the r i g h t hand s i d e of Table 6.1. The compiler d i r e c t i v e s , i n Part 1, a s s i g n numeric values to symbolic names of machines, t r a n s p o r t a t i o n equipment and queues. Hence, referen c e to these e n t i t i e s throughout the program becomes e a s i e r . The e x t e r n a l d i r e c t i v e s (Part 1.2) assure communication with e x t e r n a l programs and v a r i a b l e s supporting FLOWSIM. In P a r t s 2 to 5.5 of the model, m a t r i c e s , savevalues, f u n c t i o n s , storage c a p a c i t i e s , v a r i a b l e s and t a b l e s are d e f i n e d and i n i t i a l i z e d . These e n t i t i e s c o n t r o l computations d u r i n g the a c t u a l s i m u l a t i o n . Part 5.5 to 9 represent a quadruple sequence of b l o c k s . Each block sequence c o n s t i t u t e s a separate model segment. The three r e l a t i v e l y small segments, #1 (Part 5.6), #2 (Part 6) and #4 (Part 9), are a u x i l i a r y segments to the main segment #3 (Part 83 8) which i s the bulk of the model. Part 5.6 i s numbered among the computational e n t i t i e s d e s p i t e the f a c t that t h i s segment #1 c o n s i s t s of block statements. The reason for t h i s i s that t h i s segment c a l c u l a t e s time i n t e r v a l s needed to flow through the t r a n s p o r t a t i o n equipment as a f u n c t i o n of t h e i r s i z e and speed. T h i s i s done by a t r a n s a c t i o n moving along a loop and a s s i g n i n g time data to MSAVEVALUE "TIME". Part 6, model segment #2, i s where the a c t u a l m i l l s i m u l a t i o n s t a r t s . Before d e s c r i b i n g t h i s p a r t of the model i t w i l l be u s e f u l to e x p l a i n the correspondence between GPSS e n t i t i e s and m i l l components. B a s i c a l l y , the components of a sawmill are machines, t r a n s p o r t a t i o n equipment, and raw m a t e r i a l to be manufactured, pi e c e s of l o g s , cants, e t c . These components f a l l i n t o one of the two f o l l o w i n g groups: • Machines and t r a n s p o r t a t i o n equipment are " s t a t i c " components in the sense that they are present dur i n g the whole manufacturing process. • Lumber pi e c e s are "dynamic" components, which enter the system, flow through i t , are processed, and then leave the system. Consequently, f a c i l i t i e s and storages, as " s t a t i c " e n t i t i e s of GPSS, represent machines and t r a n s p o r t a t i o n equipment, whereas t r a n s a c t i o n s , as "dynamic" e n t i t i e s of GPSS, represent lumber p i e c e s i n the model. Rev e r t i n g to Part 6 of the model, separate segment #2 simulates the machinery breakdown. The f o l l o w i n g paragraphs d e s c r i b e the s i m u l a t i o n technique of machinery breakdown in 84 FLOWSIM and the data made a v a i l a b l e as w e l l as recommended by the m i l l management. During the design of the breakdown s i m u l a t i o n three major groups of the machinery were c o n s i d e r e d : • headrigs • other machines, and • t r a n s p o r t a t i o n equipment. D i f f e r e n t i a t i o n between these three groups was necessary i n accordance with the three breakdown s i m u l a t i o n techniques. The b a s i c concept of breakdown s i m u l a t i o n i s to a l t e r the s t a t e of a p a r t i c u l a r u n i t of machinery from one to the other of i t s two p o s s i b l e s t a t e s : work and down. The two h e a d r i g s , i n the f i r s t group are the only e x i s t i n g and o p e r a t i n g machines in the c u r r e n t m i l l . Down-time data of the two headrigs were made a v a i l a b l e by the m i l l management for June, J u l y , August and September of 1983. (Appendix 6.2). These down-time data were c o l l e c t e d corresponding to s i x causes: e l e c t r i c a l , o p e r a t i o n a l , power f a i l u r e , maintenance, saw/knife change and other. U n f o r t u n a t e l y , the a c t u a l times of the s h i f t s when breakdowns occured were not recorded. Thus the i n t e r breakdown times (work periods) are not known but are assumed to be e x p o n e n t i a l l y d i s t r i b u t e d . The only parameter determining the e x p o n e n t i a l d i s t r i b u t i o n , the average inter-breakdown time, was c a l c u l a t e d as f o l l o w s : AIBT = (T - D) / (N + 1) where AIBT stands f o r the Average Inter-Breakdown Time; T i s the 85 t o t a l time p e r i o d during data c o l l e c t i o n ; D i s the sum of down-times; and N i s the number of down-time data c o l l e c t e d . Based on the above assumption the sim u l a t o r p r e d i c t s "work" times by sampling from the e x p o n e n t i a l d i s t r i b u t i o n with the average inter-breakdown time parameter of the he a d r i g in quest i o n . At the end of a work p e r i o d , the leng t h of down-time must be p r e d i c t e d . T h i s i s done by sampling from e m p i r i c a l d i s t r i b u t i o n s which are i n c o r p o r a t e d in the model by d e f i n i n g the two GPSS continous f u n c t i o n s DWNH8 fo r Headrig 8' and DWNH9 for Headrig 9' (Part 4). These two f u n c t i o n s were c o n s t r u c t e d by the cumulative f r e q u e n c i e s which were c a l c u l a t e d with the help of FREQ program a v a i l a b l e at UBC software l i b r a r y . Based on the FREQ output the observed down times of Headrig 9', as an example, are summarized in Table 6.2. The r e l a t i o n s h i p between the f u n c t i o n f o l l o w e r cards and t h i s t a b l e i s s t r a i g h t f o r w a r d . Function DWNH8 f o l l o w e r cards and data are co n s t r u c t e d e x a c t l y the same way as that of DWNH9, and hence are not summarized here. According to the fo r e g o i n g d i s c u s s i o n , the p r i n c i p l e of machine breakdown s i m u l a t i o n f o l l o w s : a breakdown timer t r a n s a c t i o n c y c l e s along an i n f i n i t e loop, a c t i v a t i n g two ADVANCE blocks r e p r e s e n t i n g inter-breakdown times, i . e . , work peri o d s and down times (Part 6.1). The second group c o n s i s t s of brand new machines f o r which down-time data are not a v a i l a b l e . M i l l management recommended that the down-time d i s t r i b u t i o n s of the two combined machines, Gang Edger, Twin Saw, Edger and trimmers be assumed to be the 86 Table 6.2. Summary of observed downtime data of Headrig 9' Down-time* i n t e r v a l s (min) Frequency with which Cumulative down-time f a l l s i n the i n t e r v a l frequency Absolute Re l a t ive 5 - 10 1 4 .037 .037 10 - 20 68 . 180 .218 20 - 30 32 .085 .303 30 - 40 44 .117 .420 40 - 50 33 .087 .507 50 - 60 33 .087 .595 60 - 70 29 .077 .672 70 - 80 23 .061 .734 80 - 90 1 4 .037 .771 90 -1 00 1 4 .037 .808 100 -110 7 .018 .827 1 1 0 -120 9 .023 .851 1 20 - + 56 . 1 48 1 .000 t o t a l : 376 * Upper l i m i t s are excluded and lower l i m i t s i n c l u d e d . same as that of Headrig 9'. An a d d i t i o n a l assumption i s that the average amount of down-time i s 2 hours per 15 s h i f t s . Consequently, sampling from down-time d i s t r i b u t i o n happens l e s s f r e q u e n t l y than i n the case of h e a d r i g s . The working time data of machines in t h i s group are a l s o assumed to be e x p o n e n t i a l l y d i s t r i b u t e d . The breakdown s i m u l a t i o n concept i s the same as that of the headrigs (Part 6.1). Part 6.2 of the model simulates breakdowns of the t r a n s p o r t a t i o n equipment belonging to Group 3, with the f o l l o w i n g assumptions: • Average down-times are u n i f o r m l y d i s t r i b u t e d over the range of 15±1.5 min f o r b e l t s and 30±3 min f o r c h a i n s , r o l l s and unscramblers (based on m i l l management e x p e r i e n c e s ) . • The inter-breakdown times are e x p o n e n t i a l l y d i s t r i b u t e d : X (i=1,2,...,38)=(15*460*60 s e c ) " 1 i The X v a l u e s are c a l c u l a t e d f o r one complete c y c l e of work 87 and down time. T h i s X c a l c u l a t i o n d i f f e r s from the formula used to c a l c u l a t e the average inter-breakdown time f o r machines because d i f f e r e n t breakdown s i m u l a t i o n techniques are employed for the two types of machinery. (When the a c t u a l breakdown runs were made, the AIBT data f o r t r a n s p o r t a t i o n equipment were c a l c u l a t e d e r r oneously by the formula on page 84. T h i s r e s u l t e d in a very small d e v i a t i o n from what would be expected above.) There are two u s e f u l p r o p e r t i e s of the e x p o n e n t i a l d i s t r i b u t i o n which s i m p l i f y the t r a n s p o r t a t i o n equipment breakdown s i m u l a t i o n . F i r s t , the combinative p r o p e r t y of the e x p o n e n t i a l d i s t r i b u t i o n : the inter-breakdown time between two c o n s e c u t i v e breakdowns of any t r a n s p o r t a t i o n equipment i s a l s o e x p o n e n t i a l y 38 d i s t r i b u t e d with r a t e of X= L X i = 1 i Secondly, the s e l e c t i v e p r o p e r t y : s e l e c t i n g events with p r o b a b i l i t y of p from a Poisson stream of r a t e X, c r e a t e s a Poisson stream of events at rate p*X. Based on the two p r o p e r t i e s above, the technique of sampling t r a n s p o r t a t i o n equipment inter-breakdown times i s the f o l l o w i n g : 1. Sample from the e x p o n e n t i a l d i s t r i b u t i o n of X breakdown rate (X=IX ), and then i 2. F i n d out which machine breaks down, i . e . , f i n d " i " . In terms of GPSS, step 1 i s done by generating a breakdown timer t r a n s a c t i o n a c c o r d i n g to the e x p o n e n t i a l d i s t r i b u t i o n with 88 r a t e X. Step 2 i s c a r r i e d out by an ASSIGN block with TRCOD ( t r a n s p o r a t i o n equipment code) f u n c t i o n as "B" operand. T h i s f u n c t i o n a s s i g n s codes to Parameter 1 ac c o r d i n g to p =X /X i i p r o b a b i l i t i e s . Corresponding to the code i n Parameter 1 the t r a n s a c t i o n then "breaks down" the s e l e c t e d t r a n s p o r t a t i o n equipment with the a p p l i c a t i o n of two SUNAVAIL-ADVANCE-SAVAIL block sequences. The a p p r o p r i a t e block sequence, e i t h e r i n Part 6.2.1 or 6.2.2, i s s e l e c t e d on the b a s i s of whether the t r a n s p o r t a t i o n equipment i s a b e l t or not. In summary, comparing the two breakdown s i m u l a t i o n techniques used f o r the machines and t r a n s p o r t a t i o n equipment, the machine breakdown s i m u l a t i o n i s more r e a l i s t i c but needs more GPSS b l o c k s . Each machine needs i t s own separate breakdown segment. On the other hand the t r a n s p o r t a t i o n equipment breakdown s i m u l a t i o n i s simpler but l e s s a ccurate because of the f o l l o w i n g . P r e d i c t i n g the next s t a r t of down-time p e r i o d of a p a r t i c u l a r p i e c e of t r a n s p o r t a t i o n equipment should be delayed f o r the p e r i o d of down-time. T h i s does not happen because the GENERATE block of Part 6.2 does not wait f o r the next "down" p r e d i c t i o n . Consequently, i t may happen that the next "down" s t a r t of a p a r t i c u l a r machine i s p r e d i c t e d by the s i m u l a t o r , d e s p i t e the f a c t , that i t i s s t i l l down. In the GPSS model t h i s breakdown i s then ignored and, consequently, the s i m u l a t i o n i s i n a c c u r a t e . F o r t u n a t e l y , the p r o b a b i l i t y that t h i s happens i s very low and i t s e f f e c t can be ignored. Part 7 s o l e l y c o n s i s t s of comments ( s t a r t i n g with "*") d e s c r i b i n g the contents of the t r a n s a c t i o n parameters at d i f f e r e n t segments. Preceding the main segment, i n many cases, 89 i t i s handy to r e c a l l the content of v a r i o u s parameters r e l e v a n t to model behaviour. Parameters as i n f o r m a t i o n c a r r i e r s can be thought of as small l a b e l s attached to the t r a n s a c t i o n s to r e c o r d and c a r r y i n f o r m a t i o n along with them. Parameters used i n t h i s model with t h e i r content, are d e s c r i b e d i n Part 7 of the extended program l i s t i n g (Appendix 6.1). T r a n s a c t i o n s and t h e i r d i f f e r e n t i n t e r p r e t a t i o n s i n d i f f e r e n t segments are summarized in Table 6.3. Table 6.3. I n t e r p r e t a t i o n of t r a n s a c t i o n s i n d i f f e r e n t segments T r a n s a c t i o n s of d i f f e r e n t segments I n t e r p r e t a t i o n Model Segment 1 Model Segment 2 Model Segment 3 (Main segment) Loop-index i n t r a n s p o r t a t i o n time c a l c u l a t i o n . Column index of matrix savevalue "TIME". It assumes as many values as t r a n s p o r t a t i o n equipment time c a l c u l a t i o n i s needed. The elements of "TIME" matrix are used in ADVANCE blocks which simulate t r a n s p o r t a t i o n time passage. T r a n s p o r t a t i o n equipment i d e n t i f i e r . E i t h e r boom l o g , sawlog, cant, s i d e board, or lumber depending on the phase of the s i m u l a t i o n p r o c e s s . Model Segment 4 The timer Part 8 i s where the a c t u a l p i e c e flow s i m u l a t i o n takes p l a c e . In GPSS, the flow of t r a n s a c t i o n s through the " s t a t i c " e n t i t i e s i s c o n t r o l l e d by b l o c k s . These GPSS blocks are the e q u i v a l e n t s of statements of other computer languages. Many times, p a r t i c u l a r l y i n academic e x e r c i s e s , a primary goal of a 9 0 good computer programmer i s to minimize the number of statements, i . e . , b l o c k s i n GPSS. In b u i l d i n g t h i s sawmill model, some a s p i r a t i o n s were i n c o n s i s t e n t with the above g o a l , hence some e x p l a n a t i o n i s r e q u i r e d . One of the major ambitions was to keep block sequence as s i m i l a r to the a c t u a l sequence of machines and t r a n s p o r t a t i o n equipment as p o s s i b l e . One co u l d say that t h i s i s obvious, because otherwise the v a l i d i t y of the model c o u l d be questi o n e d . But t h i s i s not q u i t e t r u e . There are techniques i n GPSS (e.g. i n d i r e c t a d d r e s s i n g , i n d i r e c t s p e c i f i c a t i o n ) which can reduce s i g n i f i c a n t l y the number of bl o c k s d e s c r i b i n g the system. At the same time, these techniques a l s o reduce resemblance of the model to the r e a l d e s i g n . A sawmill model w r i t t e n i n such a co n c i s e f a s h i o n , would be d i f f i c u l t to debug, to modify and to v a l i d a t e . Consequently, when the model was c o n s t r u c t e d , the primary o b j e c t i v e was not to write the program with as few GPSS blocks as was p o s s i b l e . The order i n which the bl o c k s appear i n the main segment (Part 8) corresponds to the sequence of machine c e n t e r s and t r a n s p o r t a t i o n equipment through which lumber p i e c e s move acc o r d i n g to the r e a l m i l l d e s i g n . The boom log a r r i v a l s i m u l a t i o n (Part 8.1) sends boom logs to the "entry" of the model. According to the two-dimensional boom l o g d i s t r i b u t i o n (Appendix 5.1), boom logs are pi c k e d up by "BLGEN" f u n c t i o n . The purpose of t h i s f u n c t i o n i s to ensure that the boom l o g s e l e c t i o n procedure corresponds to the two dimensional boom l o g d i s t r i b u t i o n (Appendix 5.1) and with the sample boom l o g "p o o l " 91 (Appendix 5.6). The boom l o g r e l a t i v e f r e q u e n c i e s are c a l c u l a t e d by the formula below: R e l a t i v e frequency of = ((ABSFR(i)/SUMAF)*0.999999)/NBLOG(i) boom logs i n i - t h c e l l where ABSFR(i) i s the absolute frequency of i - t h c e l l , NBLOG(i) i s the number of boom log s i n the i - t h c e l l , SUMOF i s the sum of ab s o l u t e f r e q u e n c i e s , and 0.999999 i s the upper end value of the c l o s e d i n t e r v a l from which GPSS s e l e c t s random arguments f o r f u n c t i o n s . For i n s t a n c e , the l i k e l i h o o d that B101 (cell#1) boom l o g code w i l l be s e l e c t e d i s (102/10000)*0.999999)/2=0.0051. S i m i l a r l y , the r e l a t i v e frequency of the second boom l o g (N101) in cell#1 i s a l s o 0.0051. Hence, the cumulated r e l a t i v e f r e q u e n c i e s on the "BLGEN" f u n c t i o n f o l l o w e r cards, f o r the f i r s t two boom logs are 0.0051 and 0.0102 (Part 4.3). The f o l l o w e r cards of t h i s f u n c t i o n c o n t a i n the boom log cumulative r e l a t i v e f r e q u e n c i e s . To a v o i d e r r o r s and tedious work a FORTRAN program (Appendix 6.3) computes and w r i t e s these cumulative f r e q u e n c i e s i n the r e q u i r e d format of GPSS. A f t e r boom logs have been s e l e c t e d , they are s p l i t i n t o sawlogs a c c o r d i n g to the m i l l bucking i n s t r u c t i o n s and go to one of the two p o s s i b l e headrigs i n c l u s t e r s . Thus the proper t i m i n g and sequence of sawlog a r r i v a l i s b u i l t i n t o the model. The number of sawlogs to be bucked from a given boom l o g , the log diameter of sawlogs and the sawing i n s t r u c t i o n s of SAWSIM ( i n the form of dynamic i n f o r m a t i o n a r r a y , DIA), are organized in Data Base #1. T h i s Data Base #1 ensures the economical 92 s e l e c t i o n of a r e l a t i v e l y l a r g e number of boom logs (about 4 5 0 / s h i f t ) , from the computer time stand p o i n t . The a l g o r i t h m of the boom l o g , sawlog and SAWSIM code s e l e c t i o n process i s de p i c t e d i n F i g u r e 6.6. Studying t h i s a l g o r i t h m , the reader should take a simultaneous look at the s t r u c t u r e of Data Base #1. Thus, f o r convenience, a t r u n c a t e d segment of Data Base #1 i s d i s p l a y e d i n F i g u r e 6.9. To understand t h i s s e l e c t i o n process an example might be h e l p f u l . Consider the f o l l o w i n g case. Assume that the boom log generator f u n c t i o n of FLOWSIM re t u r n s random number 21. In row.21 of Data Base #1 (Figure 6.9), NSL(21)=2, FIRST(21)=37 can be found. These numbers mean that boom log B113 i s to be c r o s s cut i n t o two sawlogs. The p o s s i b l e ways of sawing are i n row 37, f o r the f i r s t sawlog B1131, and i n row 38, for the second sawlog B1132. Assuming that the sawing mode s e l e c t i o n mechanism of FLOWSIM p i c k s up MODE 1 (k=l) and MODE 5 (k=5) f o r the f i r s t and second sawlog, the DIA matrix codes, assigned to Parameter 1, are U011 and U206. Information governing p i e c e f l o w , has to be a s s i g n e d to parameters i n the f r o n t - p a r t of the model. DIA matrix and headrig codes are assign e d to Parameter 1 and 2. Length, market code and diameter are assigned to Parameters 7, 10 and 11, r e s p e c t i v e l y , and i n i t i a l i s e d i n Part 8.2. Parameter 7 and 11 are s e l f - e x p l a n a t o r y , however Parameters 1, 2 and 10 need explanat i o n . DIA matrices serve as important inf o r m a t i o n to c o n t r o l p i e c e flow both in time and space, throughout the presence of t r a n s a c t i o n s i n the model. Thus the DIA matrix code i s attac h e d 93 GENERATE A RANDOM NUMBER, i=1,2,..,NO.OF BOOM LOGS, WITH APPROPRIATE PROBABILITIES P1 <-- i ASSIGNED BY "BLGEN" (BOOM LOG GENERATOR FUNCTION). _^ LOOK UP THE CORRESPONDING NUMBER OF SAWLOGS TO BE PRODUCED FROM THIS BOOM LOG: NSL(i) P4 < — NSL(i)  INITIALISE SAWLOG COUNTER COUNT=1 LOOK UP THE POINTER OF THE FIRST (OR NEXT) SAWLOG ORIGINATING FROM BOOM LOG ( i ) : j=FIRST(i) I SELECT MODE OF SAWING U ) k =1,2,...,8 (COLUMN OF MODE 1,2,...,8) ASSIGN DIA CODE TO PARAMETER 1 P1 <-- DIA CODE(j,k) I COUNT SAWLOGS COUNT= COUNT+1 YES ML TAKE NEXT SAWLOG F i g u r e 6.8. A l g o r i t h m of sawlog and the corresponding SAWSIM code s e l e c t i o n process c <D P i X f t n r t o r t 0) W 0> VI (0 BOOM BOOM LOG NUMBER L I N E * SAWLOG SAWLOG SMALL E Q U I V A - D I A M . O F D IA MATR IX CODES LOG CODE OF OF NUMBER CODE END LENT E Q U I V A - OF NUMBER SAWLOGS F I R S T O IAM. SAWLOG LENT MODE 1 MODE 2 MODE 3 MODE 4 MODE 5 M0DE6 MODE 7 MOOES W I T H I N SAWLOG CODE SAWLOG BOOM L O G ( B L N ) ( B L C O D ) ( N S L ) ( F I R S T ) ( S L N ) ( S L C ) ( TO IA ) ( E S L C ) ( D I A E S ) (MOD I ) ( M 0 0 2 ) ( M 0 D 3 ) ( M 0 D 4 ) ( M 0 D 5 ) ( M 0 D 6 ) ( M 0 D 7 ) ( M 0 0 8 ) 1 B 101 1 1 1 B101 1 14 . 9 5 B101 1 14 . 9 5 U 1 U 48 U 95 U 1 4 2 U 1 8 9 U 2 3 6 U 2 8 3 U 3 3 0 2 B 102 1 2 2 B 1 0 2 1 1 6 . 0 4 B1 161 1 6 . 7 5 U 2 U 4 9 U 9 6 U 1 4 3 U 1 9 0 U 2 3 7 U 2 8 4 U331 3 B 2 0 2 1 3 3 B 2 0 2 1 1 4 . 6 7 B 2 0 2 1 14 . 6 7 U 3 U 5 0 U 9 7 U144 U191 U 2 3 8 U 2 8 5 U 3 3 2 2 0 B2 12 2 35 35 B 2 1 2 1 16 55 B2121 16 55 U 1 1 U 58 U 1 0 5 U 1 5 2 U 1 9 9 U 2 4 6 U 2 9 3 U 3 4 0 36 B 2 1 2 2 15 06 B 2 1 2 2 15 06 U 17 U 64 U 1 11 U 1 5 8 U 2 0 5 U 2 5 2 U 2 9 9 U 3 4 6 21 B1 13 2 37 37 B1 131 17 1 1 B 2 1 2 1 16 55 U 1 1 U 58 U 1 0 5 U 1 5 2 U 1 9 9 U 2 4 6 U 2 9 3 U 3 4 0 38 B1 132 15 27 B1 132 15 27 U 18 U 6 5 U l 12 U 1 5 9 U 2 0 6 U 2 5 3 U 3 0 0 U 3 4 7 22 B 1 14 3 3 9 3 9 B1 141 20 85 B137 1 2 0 96 U 12 U 5 9 U 1 0 6 U 1 5 3 U 2 0 0 U 2 4 7 U 2 9 4 U341 4 0 B1 142 18 64 B 1 2 4 2 18 7 9 U 19 U 6 6 U 1 1 3 U 1 6 0 U 2 0 7 U 2 5 4 U301 U 3 4 8 41 B1 143 16 38 N2282 17 92 U 15 U 62 U 1 0 9 U 1 5 6 U 2 0 3 U 2 5 0 U 2 9 7 U 3 4 4 23 B 2 14 3 42 42 B 2 1 4 1 19 5 0 B 1 3 7 1 2 0 96 U 12 U 5 9 U 1 0 6 U 1 5 3 U 2 0 0 U 2 4 7 U 2 9 4 U34 1 4 3 B 2 1 4 2 17 82 N 2 2 8 2 17 92 U 15 U 62 U 1 0 9 U 1 5 6 U 2 0 3 U 2 5 0 U 2 9 7 U 3 4 4 44 B 2 1 4 3 16 33 N 2 2 8 2 17 92 U 15 U 62 U 1 0 9 U 1 5 6 U 2 0 3 U 2 5 0 U 2 9 7 U 3 4 4 VO 95 to Parameter 1. Parameter 2 c a r r i e s the " t o " machine code, that i s the answer to the q u e s t i o n "where am I going ?". Hence, i t s value changes as the p i e c e flows through the system from machine to machine. Parameter 10 c a r r i e s market codes of 1, 2, 3, 4, 5 and 6 depending on whether the p i e c e s go to USA, North A f r i c a n , Japanese #1, Japanese #2, UK or the French market. Parts 8.3 and 8.4 simulate the primary breakdown of l o g s . Since the number of sawlogs has to be known l a t e r to c a l c u l a t e t o t a l p r o d u c t i o n , a one-dimensional a r r a y serves as sawlog counter v i a a matrix savevalue. Whenever a sawlog r e p r e s e n t a t i v e t r a n s a c t i o n moves i n t o one of the two MSAVEVALUE bl o c k s , the corresponding matrix savevalue element whose number i s c a r r i e d by Parameter 1, i s incremented by 1. The elements of t h i s a r r a y serve as weights when s t a t i s t i c s of i n t e r e s t are c a l c u l a t e d : product mix value and volume recovery. A d d i t i o n a l p a r t s of main segment, from 8.5 to 8.15, simulate the lumber p r o c e s s i n g at each machine center and the o p e r a t i o n of v a r i o u s t r a n s p o r t a t i o n equipment that p i e c e s flow through before a r r i v i n g at the machine. Both manufacturing a p i e c e at a machine and t r a n s p o r t i n g i t to the machine r e q u i r e time. Consequently, p r o v i s i o n s must be made to ensure that p i e c e s move in time, a c c o r d i n g to the designed system. The time r e q u i r e d to process a p i e c e at a machine i s a u t o m a t i c a l l y c a l c u l a t e d by SAWSIM based on l o g l e n g t h and other machine data provided by the s p e c i f i c a t i o n data input f i l e of SAWSIM. These s p e c i f i c a t i o n s are l i n e a l machine speed, maximum number of l i n e s per pass, gap- and load time. The c a l c u l a t e d machine time data are then t r a n s m i t t e d to the dynamic model v i a 96 DIA m a t r i c e s . In GPSS, passage of time i s p r o v i d e d by the ADVANCE block. The A operand of t h i s block " t e l l s " the s i m u l a t o r how long a t r a n s a c t i o n i s to be h e l d up to simulate the r e q u i r e d p e r i o d of time. Hence ADVANCE b l o c k s , i n c o n j u n c t i o n with machines, simulate the passage of time r e q u i r e d to process a p i e c e . T h e i r "A" operands are ampervariables with a value returned by subroutine &DIA. The r e t u r n e d value v a r i e s as DIA matrix code of P1 does, and the matrix column number denoted by P2 changes as p i e c e s are on t h e i r way to d i f f e r e n t machines. Parameter #2 c a r r i e s the " t o " machine code (Part 7.1.2). The time r e q u i r e d to flow through a given t r a n s p o r t a t i o n equipment depends on: • the speed of the t r a n s p o r t a t i o n equipment, • the s i z e of the t r a n s p o r t a t i o n equipment, and • p i e c e l e n g t h . The e f f e c t of t r a n s p o r t a t i o n equipment speed i s obvious: the higher the speed, the l e s s the time needed. However, the s i z e of the t r a n s p o r t a t i o n equipment v a r i e s as a f u n c t i o n of the route a p a r t i c u l a r p i e c e flows through in the m i l l . Thus, to put the e f f e c t of t r a n s p o r t a t i o n equipment s i z e i n t o proper l i g h t , some e x p l a n a t i o n i s needed. As an example, c o n s i d e r Chain #7 (Figure 5.4). P i e c e s going to the Twin Saw, come from e i t h e r Combined Machine #1, #2 or Gang Edger. Thus, depending on t h e i r route, p i e c e s f l o w i n g through Chain #7 t r a v e l d i f f e r e n t d i s t a n c e s along i t . T h i s i s how p i e c e route a f f e c t s t r a n s p o r t a t i o n time by equipment s i z e . Consequently, conveying times of t r a n s p o r t a t i o n equipment have 97 to be taken i n t o account a c c o r d i n g to t h i s v a r y i n g l e n g t h . T h i s i s accomplished by matrix savevalues as ADVANCE block "A" operands whose valu e s are p r e c a l c u l a t e d as q u o t i e n t s of length and speed: MH$TIME(P1,1) = MH$LNGTH(P1,1)/MH$SPEED(P1,1) Separate Segment #1, i n Part 5.3, c a r r i e s out these c a l c u l a t i o n s f o r a l l (P1 = 1, 2, 57) l e n g t h and speed combinations. Note that there are only 14 r o l l s + 5 belts+17 chains=36 items of t r a n s p o r t a t i o n equipment. The reason f o r the 57 c a l c u l a t i o n s of time data i s that pieces can be conveyed f o r va r i o u s lengths along the t r a n s p o r t a t i o n equipment. The length and speed data of the t r a n s p o r t a t i o n equipment are i n i t i a l i s e d f o r the model in Part 2.1 and summarized i n Appendix 6.4. The t h i r d f a c t o r a f f e c t i n g t r a n s p o r t a t i o n times i s pie c e l e n g t h . From F i g u r e 6.10, i t i s easy to see that the time p a s s i n g between a p i e c e a r r i v i n g at R o l l 9 and i t s p r o c e s s i n g s t a r t on Twin Saw, can be c a l c u l a t e d a c c o r d i n g to the formula of (TEL-PL)/TES. The t r a n s p o r t a t i o n equipment with i t s conveying time c a l c u l a t e d by t h i s formula has ADVANCE blo c k s with "A" operand of V$TMM1, V$TMM2, ...V$TMM6. The a c t u a l c a l c u l a t i o n s are performed by v a r i a b l e s of TMM1, TMM2, ...TMM6 i n Part 5.2.1. An important component of t r a n s p o r t a t i o n equipment in the m i l l i s the unscrambler. Major c h a r a c t e r i s t i c s of i t s o p e r a t i o n a r e : i t separates p i e c e s one by one; i t has a l i m i t e d c a p a c i t y ; i t takes time while p i e c e s flow through i t ; and p i e c e s have to wait f o r t h e i r turn u n t i l an empty unscrambler u n i t a r r i v e s at 98 r-PL - p i e c e l e n g t h TEL - t r a n s p o r t a t i o n equipment le n g t h TES - t r a n s p o r t a t i o n equipment speed F i g u r e 6.10. T r a n s p o r t a t i o n time i s a f f e c t e d by p i e c e l e n g t h the bottom p o s i t i o n . The model has to operate a c c o r d i n g l y . As an example, one of the block sequences s i m u l a t i n g unscrambler o p e r a t i o n can be seen i n the segment of Part 8.9 which d e s c r i b e s p i e c e flow between Combined Machine #1 and " v i a d r o p s o r t e r " Twin Saw. The f i r s t (lowest) and l a s t (topmost) unscrambler u n i t s are represented by f a c i l i t i e s w i t h i n the usual ENTER and LEAVE storage b l o c k s . The l a s t f a c i l i t y ensures that p i e c e s leave the unscrambler one by one, whereas the f i r s t f a c i l i t y s erves as an entrance l e t t i n g p i e c e s i n one by one and ca u s i n g p i e c e s to queue f o r p r o c e s s i n g . Passage of time while a p i e c e flows through the unscrambler i s the sum of the two f a c i l i t i e s r e l a t e d and the one storage r e l a t e d ADVANCE b l o c k s . In P a r t 3, among other storage d e f i n i t i o n s , the unscrambler c a p a c i t i e s are d e f i n e d as the maximum number of p i e c e s that the unscrambler can convey at a time. In c o n j u n c t i o n with t r a n s p o r t a t i o n equipment another problem a r i s e s when p i e c e s f l o w i n g through Chain 1 can e i t h e r go to Chain 2 or R o l l 4. In the r e a l system, movement of lumber i n the r e q u i r e d d i r e c t i o n i s accomplished by s k i d t r a n s f e r s that 99 e i t h e r r a i s e the chai n above the l e v e l of the r o l l or lower the chain under r o l l l e v e l . T h i s causes lumber to move in e i t h e r chain or r o l l d i r e c t i o n . Note that i f a pi e c e of lumber moves to r o l l d i r e c t i o n , the succeeding lumber can not occupy the l a s t s e c t i o n of Chain 1 fo r a c e r t a i n time p e r i o d . T h i s a l t e r n a t i n g movement i s simulated by a s i n g l e s e r v e r f a c i l i t y named COM 1 as a common area of Chain 1 and R o l l 4. The conveying time i s a fu n c t i o n of speed and d i s t a n c e to cover. U n f o r t u n a t e l y both speed and d i s t a n c e depend on route d i r e c t i o n . T h i s i s why the "A" operand of Advance block r e l a t e d to COM 1, can be the v a r i a b l e of e i t h e r TMC1W or TMC1L. These two v a r i a b l e s c a l c u l a t e the conveying time when a pi e c e i s t r a n s p o r t e d to e i t h e r p e r p e n d i c u l a r to len g t h (TMC1W) or l o n g i t u d i o n a l (TMC1L) d i r e c t i o n s . The v a r i a b l e s of TMC2W and TMC2L have e x a c t l y the same meaning except that they c a l c u l a t e the conveying time of COM2 as common t r a n s p o r t a t i o n areas of Chain 4 and R o l l 6. The d e f i n i t i o n s of these four a r i t h m e t i c v a r i a b l e s are in Part 5.2.2 of the model. Another important t r a n s p o r t a t i o n equipment c h a r a c t e r i s t i c governing m i l l dynamics i s c a p a c i t y . When the occupancy of a t r a n s p o r t a t i o n d e vice i s so high that the remaining c a p a c i t y i s not enough f o r the a r r i v i n g lumber, i t denies e n t r y . Thus t r a n s p o r t a t i o n equipment that serve as surge areas, a f f e c t to a la r g e degree the o p e r a t i o n of the preceding m i l l components and can cause c o n g e s t i o n . C a p a c i t i e s of b e l t s and r o l l s are determined by l i n e a r measures, because t h e i r occupancy then can be c a l c u l a t e d as the d i f f e r e n c e between b e l t or r o l l c a p a c i t y and the sum of lumber 100 lengths occupying the t r a n s p o r t a t i o n equipment. On the other hand, c a p a c i t i e s of chains p r o v i d i n g lumber movement in tr a n s v e r s e d i r e c t i o n s , are determined by e i t h e r l i n e a r measures or square measures depending on whether only one l a y e r of lumber or more are allowed to be conveyed on the c h a i n . Cross Chains #1,2,...,6 t r a n s p o r t i n g cants only s i d e by s i d e , have c a p a c i t i e s determined i n l i n e a r measures. S i m i l a r l y , Chains #16 and #17 moving lumber to trimmers a f t e r unscramblers, a l s o have c a p a c i t i e s given i n l i n e a r measure. In view of the f a c t that Chains #7,8,...,15 t r a n s f e r lumbers on the top of each other (more l a y e r s ) , these have c a p a c i t i e s d e f i n e d i n square measures. The c h a i n c a p a c i t y occupancy then can be c a l c u l a t e d as the d i f f e r e n c e between chain c a p a c i t y and the sum of lumber c r o s s s e c t i o n areas. The maximum c a p a c i t i e s of t r a n s p o r t a t i o n equipment are given by storage d e f i n i t i o n cards i n Part 3 of the model. As can be seen from the forego i n g d i s c u s s i o n , to make a d e c i s i o n whether or not a p i e c e of lumber w i l l be allowed or refu s e d to enter to a t r a n s p o r t a t i o n u n i t the simulator must keep up-to-date records of a v a i l a b l e c a p a c i t i e s . To c a l c u l a t e these a v a i l a b l e c a p a c i t i e s the le n g t h , width, and t h i c k n e s s dimensions of p i e c e s , at any stage of p r o c e s s i n g , must be known. The c u r r e n t v e r s i o n of SAWSIM mo d i f i e d a c c o r d i n g to the IBM F o r t r a n d i a l e c t , can not provide such lumber s i z e i n f o r m a t i o n . To overcome t h i s problem the dynamic model generates these lumber dimensions as f o l l o w s . The l e n g t h of any pie c e of lumber i s equal to the l e n g t h of sawlog i t i s produced from. 101 Thicknesses are a f u n c t i o n of market and are always as s i g n e d to Parameter 9 by the f u n c t i o n named THICK i n Part 4.1.2 of the model. Widths are randomly a s s i g n e d t o Parameter 8 i n four ways, a c c o r d i n g to the piec e i n q u e s t i o n . A pie c e can be e i t h e r c a n t / s i d e board(a), only c a n t ( b ) , c e n t e r p i e c e s o r i g i n a t i n g from a cant that need f u r t h e r breakdown(c), or a l r e a d y edged lumber with f i n a l c r o s s s e c t i o n dimensions(d), F i g u r e 6.11. ( a ) (b) ( c j ( (d) F i g u r e 6.11. Four d i f f e r e n t ways of a s s i g n i n g p i e c e width Method (a) a s s i g n s random width to sideboards w i t h i n the i n t e r v a l of [8,d], where d i s l o g diameter. The lower end po i n t of 8 i s based on the assumption that the width of the s m a l l e s t sideboard i s opening faces (6") + wane (2*1")=8". The c a l c u l a t i o n i t s e l f i s performed by f l o a t i n g - p o i n t v a r i a b l e CANTW (Part 5.1). As an example, see Part 8.6 where p i e c e s on t h e i r way to Combined Machine #2 are a s s i g n e d cant or side b o a r d widths v i a block ASSIGN 8,$CANTW before e n t e r i n g Chain 4. 1 02 Method (b) simply c o p i e s the l o g diameter to Parameter 8 as cant width. As an example see the block of ASSIGN 8,&VALUE (Part 8.7). Method (c) assigns cant t h i c k n e s s as p i e c e width. Since these p i e c e s can only be e i t h e r 8 or 12 inches wide, function#3 has only two d i s c r e t e p o i n t s of v a l u e s : 8 or 12. (Part 4.1.1). Method (d) as s i g n s lumber widths depending on market codes. Functions a s s i g n i n g these widths, are i n d i r e c t l y addressed v i a Parameter 10 which c a r r i e s a market code. As an example, see p i e c e s from Combined Edger #1 to Trimmer. The block ASSIGN 8,FN*10 performs t h i s t a s k . (Part 8.13). Previous paragraphs d e s c r i b e d how l i m i t e d c a p a c i t i e s of t r a n s p o r t a t i o n equipment are taken i n t o account by the model. However, i n order to simulate piece flow a c c u r a t e l y , c o n s i d e r a t i o n of these c a p a c i t y l i m i t s i s not s u f f i c i e n t . GPSS a u t o m a t i c a l l y stops t r a n s a c t i o n flow i f a storage space or a f a c i l i t y i s f u l l although t h i s i s not a s u f f i c i e n t l y a c c urate c o u n t e r f e i t of r e a l m i l l behaviour. The problem i s that using the simple SEIZE-ADVANCE-RELEASE sandwich f o r f a c i l i t i e s and the ENTER-ADVANCE-LEAVE f o r storages, t r a n s a c t i o n s are always allowed to r e l e a s e or leave the l a s t block of the sandwich even i f the succeeding u n i t i s f u l l . When the " c a p a c i t y f u l l " case occurs, t r a n s a c t i o n s are w a i t i n g i n f r o n t of the storage or f a c i l i t y "entrance" b l o c k s , ENTER or SEIZE, having l e f t the preceding b l o c k . Consequently, the preceding u n i t i s not a f f e c t e d by t h i s flow i n t e r r u p t i o n and the so c a l l e d "chain r e a c t i o n " does not e x e r c i s e i t s e f f e c t . For s i m i l a r reasons, machinery breakdown, as another cause of stopping piece flow, 103 should a l s o be c o n s i d e r e d . The term "chain r e a c t i o n " r e f e r s to the phenomenon in a sawmill when f u l l c a p a c i t y or breakdown of a downstream machine or t r a n s p o r t a t i o n equipment causes f i l l i n g up and then stopping of p r e c e d i n g u n i t s of m i l l flow. To simulate t h i s the dynamic model must have a mechanism that does not allow p i e c e s to leave a machine or t r a n s p o r t a t i o n equipment unless p i e c e s can enter i n t o a succeeding u n i t . Among the p o s s i b l e permutations of two not n e c e s s a r i l y d i s t i n c t o b j e c t s , t r a n s p o r t a t i o n equipment and machine, there are three p o s s i b l e cases to be c o n s i d e r e d : • t r a n s p o r t a t i o n equipment - t r a n s p o r t a t i o n equipment, • t r a n s p o r t a t i o n equipment - machine • machine - t r a n s p o r t a t i o n equipment. (Notice that machine - machine permutation i s impossible because t r a n s p o r t a t i o n equipment i s always i n s t a l l e d between machines.) The s i m u l a t i o n of "chain r e a c t i o n " i s simply ensured by ...ENTER(TE1), ,ENTER(TE2),LEAVE(TE1), , L E A V E ( T E 2 ) , . . . , block sequence f o r the f i r s t case, and ...ENTER(TE), ,SEIZE(MAC),LEAVE(TE), ,RELEASE(MAC) block sequence f o r the second case. The a b b r e v i a t i o n s i n the above block sequences have the meaning as f o l l o w s : T E 1 T r a n s p o r t a t i o n Equipment # 1 , T E 2 - T r a n s p o r t a t i o n Equipment # 2 , TE - T r a n s p o r t a t i o n Equipment, MAC - Machine. D e s p i t e the f a c t that at f i r s t glance the sequence might seem to be strange, the GPSS block sequence ensures that a pi e c e i s allowed to leave a u n i t of machinery i f , and only i f , the succeeding u n i t i s ready to r e c e i v e i t . In the t h i r d case, a TEST block i n r e f u s a l mode, i s used to l e t p i e c e s go from the machine i f and only i f three 104 c o d i t i o n s are met si m u l t a n e o u s l y : t r a n s p o r t a t i o n equipment i s a v a i l a b l e , the remainder c a p a c i t y of t r a n s p o r t a t i o n equipment i s l a r g e enough to r e c e i v e the p i e c e ( s ) , and no p i e c e s are w a i t i n g in an imaginary queue between machine and t r a n s p o r t a t i o n equipment. The a c t u a l t e s t i s c a r r i e d out on Boolean v a r i a b l e s whose value i s 1 whenever a l l three c o n d i t i o n s are t r u e . Boolean v a r i a b l e s in c o n j u n c t i o n with a l l t r a n s p o r t a t i o n equipment are d e f i n e d i n Part 5.3. Among the block statements r e p r e s e n t i n g machine o p e r a t i o n , there are two a d d i t i o n a l b l o c k s , QUEUE and DEPART to mention. The purpose of these two bl o c k s i s to c o l l e c t s t a t i s t i c s about the length of time d u r i n g which the machine was blocked. Pieces going to the Twin Saw v i a the d r o p s o r t e r , to the Edger and to the Trimmers, have two routes to take, through e i t h e r Chain #8 (Trimmer l i n e #1) or Chain #9 (Trimmer l i n e #2). These two l i n e s are chosen at random by using TRANSFER blocks i n s t a t i s t i c a l t r a n s f e r mode. Since, the model simulates flow u n t i l the lumber leaves the trimmer, and the above route d e c i s i o n i n the r e a l m i l l depends on workloads at bin s o r t e r s a f t e r trimmers, the a p p l i e d 50-50% p r o b a b i l i t i e s at TRANSFER blocks are approximations. However, t h i s r u l e of route d e c i s i o n i s o v e r r i d d e n when one of the two trimmers i s down. In t h i s case p i e c e s go only to the trimmer i n o p e r a t i o n . Whenever an o p e r a t i o n at a machine has been f i n i s h e d , the p i e c e , v i a an u n c o n d i t i o n a l TRANSFER block, s k i p s to the block l a b e l e d SKIP (Part 8.15). T h i s Part of the model prevents erroneous p i e c e / t r a n s a c t i o n m u l t i p l i c a t i o n by assembling the t r a n s a c t i o n s i n t o one before they enter the SPLIT block and the 1 05 r i g h t number of t r a n s a c t i o n s i s propagated. To understand t h i s , c o n s i d e r the f o l l o w i n g example. As was d i s c u s s e d e a r l i e r , the data t r a n s m i s s i o n between SAWSIM and the dynamic model i s accomplished by DIA m a t r i c e s . I t s general a ( i , j ) element rep r e s e n t s the t o t a l number of p i e c e s going to the j - t h machine from the i - t h one. (Since the l a s t two rows of DIA matrix are used f o r t r a n s m i t t i n g other i n f o r m a t i o n than p i e c e number, the above d e f i n i t i o n i s v a l i d i f the s u b s c r i p t i < ( m a x ( i ) - 2 ) ) . I t i s very important to emphasize the word " t o t a l " i n the above d e f i n i t i o n . To see why, l e t us co n s i d e r the DIA matrix of F i g u r e 6.5. I t s a(5,13)=4 element does not mean that f o r each of the 3 p i e c e s (a(2,5)=3) 4 p i e c e s per p i e c e , that i s 3*4=12 p i e c e s go to machine TRM. I t means that the t o t a l number of p i e c e s l e a v i n g Combined Machine #2 i n edging mode (CE2) i s 4. In GPSS, i t i s the SPLIT block which can b r i n g an " o f f s p r i n g " t r a n s a c t i o n i n t o the model. Hence, t h i s block i s used to simulate the m u l t i p l i c a t i o n f e a t u r e of sawing i n the sense that a p i e c e i s being cut i n t o more than one p i e c e . Since the SPLIT block of the model, from the execution order p o i n t of view, i s a f t e r machine o p e r a t i o n s and takes p l a c e as o f t e n as a t r a n s a c t i o n e n t e r s i t , the model would operate a c c o r d i n g to the above f a u l t y p i e c e m u l t i p l i c a t i o n . To a v o i d t h i s , a c l u s t e r of ASSEMBLE blocks i s introduced with the purpose of assembling p i e c e s which leave a p a r t i c u l a r machine. There are separate ASSEMBLE blocks i n the program to assemble p i e c e s o r i g i n a t i n g from one l o g . Just before the ASSEMBLE set, a group of TEST blo c k s i n c o n d i t i o n a l t r a n s f e r mode, sends t r a n s a c t i o n s to the 106 one of the ASSEMBLE blocks which, i s the e q u i v a l e n t of the machine whose number i s i n Parameter 2. At the ASSEMBLE block the f i r s t t r a n s a c t i o n i s delayed u n t i l other t r a n s a c t i o n s from the same machine a l s o enter. Once a l l other t r a n s a c t i o n s have a r r i v e d , they are removed from the model and only one t r a n s a c t i o n continues i t s flow to SPLIT block. The number of t r a n s a c t i o n s to be assembled i s a s s i g n e d to Parameter 5, hence t h i s Parameter i s used as "A" operand of the ASSEMBLE b l o c k s . A f t e r having assembled i n t o one, the t r a n s a c t i o n goes to SPLIT block to t r i g g e r i t s propagation a c t i v i t y (Part 8.15). Then the r i g h t number of p i e c e s i s sent to block l a b e l e d SLCT to s e l e c t the machine of next o p e r a t i o n and i t s preceding t r a n s p o r t a t i o n equipment. Summarizing, the whole process of sawmilling can be thought of as a s e r i e s of t r a n s p o r t a t i o n / m a c h i n e o p e r a t i o n , transportation/machine o p e r a t i o n , . . . , e t c . A c c o r d i n g l y , the piece flow s i m u l a t i o n i s accomplished by a s e r i e s of c y c l e s . Each c y c l e simulates one transportation/machine o p e r a t i o n u n i t of the whole p r o c e s s . The s i m u l a t i o n of each c y c l e c o n s i s t s of 1. s p l i t segment (Part 8.15), 2. t r a n s p o r t a t i o n equipment segment, 3. machine o p e r a t i o n segment, and 4. assemble segment (Part 8.14). The flow s i m u l a t i o n of a p a r t i c u l a r sawlog and the p i e c e s o r i g i n a t i n g from i t ends at the trimmers. However, the computer program i t s e l f does not end with the s i m u l a t i o n of trimmer o p e r a t i o n . As the group of r e p o r t statements i n Table 6.1 suggests, the computer program ends with i t s report w r i t e r (Part 107 10). These r e p o r t statements ensure a " t a i l o r made" summary about r e l e v a n t s t a t i s t i c s of m i l l behaviour: • a b s o l u t e and r e l a t i v e (%) time s t a t i s t i c s of i d l e , busy, blocked and down s t a t e s of machines, • bar c h a r t s of the above u t i l i z a t i o n s t a t i s t i c s , • p i e c e counts of both machines and t r a n s p o r t a t i o n equipment, • maximum content and average u t i l i z a t i o n of t r a n s p o r t a t i o n equipment, • w a i t i n g s t a t i s t i c s at v a r i o u s p i e c e flow p o i n t s , « number of boom logs processed d u r i n g s i m u l a t i o n e t c . A l l these outputs w i l l be d i s p l a y e d and d i s c u s s e d i n Chapter 8 d e a l i n g with the r e s u l t s of s i m u l a t i o n runs. 1 08 7. MODEL VALIDATION V a l i d a t i o n i s one of the most important and d i f f i c u l t problems of computer s i m u l a t i o n . Important, because without showing that the model mimics the r e a l system as i t i s supposed to, one can not t r u s t the r e s u l t s produced by the model and can not make d e c i s i o n s , say, on the c o n s t r u c t i o n of a sawmill worth m i l l i o n s of d o l l a r s . V a l i d a t i o n i s d i f f i c u l t , because v a l i d a t i n g a s i m u l a t i o n model " . . . i n v o l v e s a host of p r a c t i c a l , t h e o r e t i c a l , s t a t i s t i c a l , and even p h i l o s o p h i c a l c o m p l e x i t i e s " (68). In the stage of v a l i d a t i o n , many quest i o n s a r i s e . What i s v a l i d i t y ? What i s a v a l i d s i m u l a t i o n model? What does v a l i d a t i n g a model mean? In the l i t e r a t u r e of computer s i m u l a t i o n theory, answers to these q u e s t i o n s can e a s i l y be found. For i n s t a n c e : "The v a l i d i t y of a s i m u l a t i o n i s a measure of the extent to which i t s a t i s f i e s i t s design o b j e c t i v e s " (66). "A v a l i d s i m u l a t i o n model should behave in a manner s i m i l a r to the u n d e r l y i n g phenomena" (68). "To v e r i f y or v a l i d a t e any kind of model means to prove the model to be t r u e " (69). However, these answers are too general and they generate a d d i t i o n a l q u e s t i o n s . When can we say that the model behaviour i s " s i m i l a r " to the system i t mimics? Can we recognize " t r u t h " ? Answers to these q u e s t i o n s - i f they e x i s t at a l l - take the a n a l y s t to the f i e l d of philosophy, but the most important 109 q u e s t i o n i s s t i l l unanswered. How to go about v a l i d a t i n g a model? In g e n e r a l , v a l i d a t i n g a model i s a two-step process. In the f i r s t step the a n a l y s t has to make sure that "... the model i s i n t e r n a l l y c o r r e c t i n a l o g i c a l and programming sense." (68) In the second stage the task i s to check i f the r e s u l t s of i n t e r e s t , produced by the s i m u l a t i o n model, are b e l i e v a b l e . T h i s i s u s u a l l y done by comparing model outputs with known r e s u l t s of the r e a l world. Since the model of t h i s t h e s i s r e p r e s e n t s a n o n - e x i s t i n g m i l l r e d e s i g n , d i f f i c u l t i e s i n c o n j u n c t i o n with the second step a r i s e . "When a model i s intended to simulate a new or proposed system f o r which no a c t u a l data are a v a i l a b l e , there i s no good way to v e r i f y that the model, i n f a c t , r e p r e s e n t s the system. Under these circumstances, there i s l i t t l e a l t e r n a t i v e but to t e s t the model thoroughly f o r l o g i c a l or programing e r r o r s and to be a l e r t f o r any d i s c r e p a n c i e s or unusual c h a r a c t e r i s t i c s i n the r e s u l t s obtained from the model", Meier (68) . How then can the model used i n t h i s t h e s i s be v a l i d a t e d ? R e f e r r i n g to M a i s e l and Gnugnoli d e f i n i t i o n of v a l i d i t y (66), the model has to be checked i f i t s a t i s f i e s i t s design o b j e c t i v e s . T h i s d e f i n i t i o n w i l l be the s t a r t i n g p o i n t of v a l i d a t i n g FLOWSIM. The b a s i c o b j e c t i v e of c o n s t r u c t i n g FLOWSIM i s to p r e d i c t lumber p r o d u c t i o n as a f u n c t i o n of l o g breakdown and l o g throughput. Thus, the accuracy of production e s t i m a t i o n depends to a l a r g e degree on the accuracy of log breakdown and pie c e flow s i m u l a t i o n . T h e r e f o r e , the model of t h i s t h e s i s can be v a l i d a t e d by t e s t i n g i f l o g i c of both l o g breakdown and p i e c e flow operate a c c u r a t e l y enough to f u l f i l l the design o b j e c t i v e s . 1 10 As d e s c r i b e d e a r l i e r , the dynamic model uses SAWSIM as i t s log breakdown l o g i c . Based on the SAWSIM sawing i n f o r m a t i o n , FLOWSIM simulates p i e c e flow. A c c o r d i n g l y , v a l i d a t i n g the model of t h i s study i n c l u d e s two stages: • Log breakdown v a l i d a t i o n • Piece flow v a l i d a t i o n 7.1. LOG BREAKDOWN VALIDATION SAWSIM has been i n use f o r more than seven y e a r s . I t s r e p u t a t i o n i s r e f l e c t e d by the names of companies a p p l y i n g i t . Weyerhaeuser Company has employed SAWSIM on a m u l t i - m i l l i o n d o l l a r research program aimed at improving lumber recovery in i t s sawmills. MacMillan B l o e d e l L t d . , used i t to improve c o a s t a l m i l l o p e r a t i o n s and f o r design of new c o a s t a l m i l l s . Crown Z e l l e r b a c h Canada L t d . , used i t f o r s e t - t a b l e g e n e r a t i o n . C a r o l l - H a t c h ( I n t e r n a t i o n a l ) L t d . , has used the software f o r sawmill design p r o j e c t s . In the view of the above f a c t s , t h i s t h e s i s r e s e a r c h t r u s t s i n SAWSIM and takes the accuracy of data t r a n s m i t t e d to the dynamic model f o r granted. However, n o t i c e that c o r r e c t n e s s of SAWSIM does not mean that i t cannot be used i n c o r r e c t l y . T h e r e f o r e , inputs of SAWSIM corresponding to lumber s i z e s , machine s p e c i f i c a t i o n , and l o g shape c h a r a c t e r i s t i c s were d i s p l a y e d f o r and d i s c u s s e d with the m i l l management. SAWPLOT, an o p t i o n of SAWSIM was used to t e s t each sawing p a t t e r n s f o r a l l markets. P l o t t e d outputs of SAWSIM were examined and checked to determine i f the logs were sawn as they were 111 intended. 7.2. PIECE FLOW VALIDATION The v a l i d i t y of p i e c e flow s i m u l a t i o n w i l l be accepted i f : 1. FLOWSIM simulates p i e c e flow a c c o r d i n g to SAWSIM sawing i n s t r u c t i o n s , i . e . , 1.1. i f the number of p i e c e s produced from a given l og correspond to the number of p i e c e s cut by SAWSIM ; 1.2. i f the machines used to process a l o g correspond to machines used by SAWSIM to "saw" the same l o g ; 1.3. i f the pi e c e propagation at machines corresponds to the number of " o f f s p r i n g " p i e c e s c a l c u l a t e d by SAWSIM ; 1.4. i f time p e r i o d s needed to process p i e c e s correspond to p r o c e s s i n g times c a l c u l a t e d by SAWSIM; 2. The time sequence of manufacturing o p e r a t i o n s i s c o r r e c t , i . e . , the p i e c e flow l o g i c of the model prevents absurd cases such as p r o c e s s i n g a log on the headr i g being preceded by manufacturing the board o r i g i n a t i n g from t h i s l o g ; 3. P i e c e s flow through the r i g h t t r a n s p o r t a t i o n equipment between machines i n both d e s t i n a t i o n and time; 4. The pie c e s i z e s , taken i n t o account when t r a n s p o r t a t i o n equipment c a p a c i t y u t i l i z a t i o n i s c a l c u l a t e d , are s a t i s f a c t o r y ; 5. Machine u t i l i z a t i o n s t a t i s t i c s are c o r r e c t ; 6. I n t e r a c t i o n of m i l l machinery u n i t s i s b u i l t in the model, i . e . , the cha i n r e a c t i o n i s simulated. (The term "chain r e a c t i o n " r e f e r s to the phenomenon i n a sawmill when the breakdown of a downstream machine or t r a n s p o r t a t i o n equipment 1 1 2 causes f i l l i n g up and stops preceding u n i t s of m i l l flow.) To t e s t model v a l i d i t y of the above aspects the f o l l o w i n g a c t i o n s were taken. D e t e r m i n i s t i c , monitored and short t e s t runs were c a r r i e d out to v e r i f y the model. The runs are d e t e r m i n i s t i c i n the sense that random sources are r e p l a c e d by known valu e s determined i n advance, hence s t o c h a s t i c elements are excluded. The runs were monitored in the sense that r e l e v a n t event h i s t o r y i s p r i n t e d out i n d e t a i l . F i n a l l y , runs are short i n the sense that only a few but important events of the t o t a l process are checked because of the time consuming and tedious nature of t h i s checking procedure. To check piece flow c o r r e c t n e s s corresponding to the above aspec t s , a s i n g l e l og was "sawn", p r i n t e d and p l o t t e d by SAWSIM (Figure 7.1). The same log was processed through FLOWSIM (Figure 7.2). The number of pi e c e s produced from the log [1.1] can be checked by adding up the pie c e counts of the two trimmers where pi e c e s leave the m i l l ( F igure 7.2). Throughout the f o l l o w i n g paragraphs r e f e r e n c e w i l l be made to the above aspects by numbers i n square b r a c k e t s . The sum of trimmer p i e c e counts i s 8, two at Trimmerl and s i x at Trimmer2, which corresponds to the SAWSIM p i e c e count of 8. The machines used to process the l o g [1.2] and the pie c e propagation [1.3], based on SAWSIM output, are summarized i n Table 7.1. The machines p r o c e s s i n g the l o g can be checked i n FLOWSIM output by the column of "piece count". Non-zero p i e c e counts i n d i c a t e the machines used to process the l o g . These non-zero 113 <% <x LOC SAVING IDENT PATTERN — O F F S E T — v c l 01* IN tEN TAPER --SWEEP-- CUFT PCS FBH FB«V CHIP LUMBER BrPRD COST CROSS M FT IN/FT E06E CYL ACTUAL LBR LBR CUFT CUFT * $ • f CUNIT <%> F3 /1 FM03 1.OOO 0.0 It) CO 31 10 O 123 0.0 0 0 •! I • 155 B.2T 10 5 1M.4S B BO 0 0 14*05 3*4 9 HEDBCB1 HEDBGB1 HB9-CE1 HM-CG1 M**~eNG HM-TW HBB-CE2 HB8-C02 MR8-0NG HRB-TtM CE1-TM CG1-TMN C61-0TW CCi-EDC CGi-'R <% PASSES <% SECONDS *% t as JO <X CE2-TRK C62-THN CC2-DTW CC2-EDG C63-TMI 0N6-TMN ONB-OT* 0N6-EDG ONG-TRH TMN-TMD TBN-EDC TtM-TBM OTW-TWD DTW-ED6 DTW-TB *t PASSES 2 1 9 <* SECONDS 4 B. 10 < X * <% TH0-T01 TWD-EDG TMD-TBM T01-EDG TD1-TM EDO-TRM CHIPPER <X PASSES 1 2 <% SECONDS 2. < X * SAMSIK PLOT F3 / 1 FM03 3X09 20 3X09 20 3X0B 20 3X09 20 ONE INCH 3X09 20 F i g u r e 7.1. SAWSIM output to v a l i d a t e FLOWSIM r e s u l t s SIMULATION RESULTS OF MACMILLAN BLOEDEL WR3 SAWMILL DESIGN. 08:39 P.M. MAR. 04, 1984 Sawmill operation was simulated for 431 sec. MACHINE UTILIZATIONS AND PIECE COUNTS. Name Idle Busy Blocked Down Piece Count T ime % T Ime % T ime X Time X HRG9 431.0 100.0 0.0 0.0 0.0 0.0 0.0 0. .0 O HRG8 405.0 93.9 25.9 6.0 0.0 0.0 0.0 0 .0 1 CMB 1 431 .0 100.0 0.0 0.0 0.0 o.b 0.0 0 0 o CMB2 385.0 89.3 45.9 10.6 0.0 0.0 0.0 0 .0 3 GANG 431.0 100.0 0.0 0.0 0.0 0.0 0.0 0 .0 O TWIN 431.0 100.0 0.0 0.0 0.0 0.0 0.0 0 .0 0 EDGER 423.0 98. 1 7.9 1.8 0.0 0.0 0.0 0 0 1 TRMR1 427.0 99.0 3.9 0.9 0.0 0.0 0.0 0. 0 2 TRMR2 419.0 97.2 11.9 2.7 0.0 0.0 0.0 0. .0 6 TRANSPORTATION EQUIPMENT PIECE COUNTS. TRANSPORTATION EQUIPMENTS: PIECE COUNTS: BLT01 0 BLT02 8 BLT03 0 BLT04 1 BLT05 0 CHN01 0 CHN02 0 CHN03 0 CHN04 3 CHNOB 3 TRANSPORTATION EQUIPMENTS: PIECE COUNTS: CHN06 0 CHN07 0 CHN08 3 CHN09 6 CHN10 0 CHN11 0 CHN12 1 CHN13 0 CHN14 0 CHN1B 0 TRANSPORTATION EQUIPMENTS: PIECE COUNTS: CHN16 2 CHN17 6 R0L01 0 ROL02 3 R0L03 0 ROL04 0 ROLOS 0 ROLOS 0 R0L07 8 R0LO8 0 TRANSPORTATION EQUIPMENTS: PIECE COUNTS: R0L09 0 R0L10 0 R0L11 1 R0L12 0 R0L13 1 ROL 14 0 UNSC1 3 UNSC2 6 115 Table 7.1. Piece propagation at machines used to process the l o g FLOWSIM SAWSIM Number of p i e c e counts* Machines used a r r i v i n g l e a v i n g to process p i e c e s "from" " t o " the log Headrig 8' 1 3 HR8 -> CE2 = 2 HR8 -> CG2 = 1 Comb.mach#2 3 8 CE2 -> TRM = 2 CG2 -> EDG = 1 CG2 -> TRM = 5 Edger 1 1 EDG .-> TRM = 1 Trimmers 8 8 *Note: the number of p i e c e s processed at a machine i s the sum of numbers at the end of those l i n e s i n which the " t o " machine code i s the machine i n q u e s t i o n ; the number of p i e c e s l e a v i n g a machine i s the sum of numbers at the end of those l i n e s i n which the "from" machine code i s the machine in q u e s t i o n . e n t r i e s a l s o give the number of a r r i v i n g p i e c e s which are 1, 3, 1 and 8 corresponding to Headrig 8', Combined machine #2, Edger and Trimmers, r e s p e c t i v e l y . P i e c e propagation, i . e . , the number of p i e c e s l e a v i n g machines can best be checked by the t r a n s p o r t a t i o n equipment p i e c e counts a f t e r the machines. The p i e c e counts of Chain#4, Roll#7 and Roll#13 are 3, 8 and 1, r e s p e c t i v e l y , i n d i c a t i n g t h a t FLOWSIM accomplished p i e c e p r o p a g a t i o n a c c o r d i n g to the i n s t r u c t i o n s of SAWSIM. The c o r r e c t n e s s of machine times needed to process a l o g [1.4], the time sequence of o p e r a t i o n s [ 2 ] , and i f p i e c e s flow through the r i g h t t r a n s p o r t a t i o n equipment [3] were checked by the i n t e r a c t i v e debugging f a c i l i t i e s of GPSS/H. The "TRAP SYSTEM" command was used to t r a c k p i e c e s , be i t l o g , center cant, s i d e b o a r d or a p i e c e of lumber, while they were fl o w i n g through the m i l l . T h i s command t r a c k s t r a n s a c t i o n s , 116 r e p r e s e n t i n g p i e c e s , through the system i n a d e t a i l e d step by step f a s h i o n , by stopping t h e i r movement at d i s c r e t e event times. In t h i s mannner the whole l o g manufacturing process c o u l d be monitored, and t r a n s a c t i o n s of i n t e r e s t c o u l d be d i s p l a y e d and checked. The output of t h i s i n t e r a c t i v e run i s i n Appendix 7.1. T h i s appendix a c t u a l l y presents the i n t e r a c t i v e run output i n a reformatted way i n the sense that the output f i l e was e d i t e d and events belonging to one t r a n s a c t i o n were grouped together. Thus, the movement of a p a r t i c u l a r p i e c e can e a s i l y be monitored by observing the corresponding t r a n s a c t i o n . To show the correspondence between t r a n s a c t i o n s and pi e c e s F i g u r e 7.3 summarizes what p i e c e s are represented by what t r a n s a c t i o n s . To monitor pi e c e movement both i n time and space, the i n t e r a c t i v e debugging output p r o v i d e s two important p i e c e s of i n f o r m a t i o n : the s i m u l a t i o n c l o c k time and the block number. Sim u l a t i o n c l o c k time t e l l s when a p a r t i c u l a r event happens. Block number i n d i r e c t l y t e l l s the l o c a t i o n of the pie c e w i t h i n the m i l l . T h i s i n f o r m a t i o n i s i n d i r e c t , because the GPSS block number t e l l s only the t r a n s a c t i o n l o c a t i o n w i t h i n the GPSS program. The a c t u a l m i l l machinery u n i t , machine or t r a n s p o r t a t i o n equipment, corresponding to the GPSS block can then be looked up i n the program. Based on F i g u r e 7.3 and the i n t e r a c t i v e output (Appendix 7.1) a machine time c h a r t of p r o c e s s i n g can be c o n s t r u c t e d (Figure 7.4). T h i s time c h a r t p r o v i d e s the b a s i s to check model v a l i d i t y from machine time [1.4] and o p e r a t i o n sequence [2] p o i n t of view. At the same time t h i s c h a r t makes i t e a s i e r to 1 17 xact#14 F i g u r e 7.3. Correspondence between t r a n s a c t i o n s and p i e c e s . 118 survey the p i e c e movement i n time. Machine times on t h i s c h a r t are 27-1=26 seconds to process the log on Headrig 8'; 78-63=93-78=15 seconds to process the two sideboards on Combined Machine #2 i n edging mode; 109-93=16 seconds to process the center cant on Combined Machine #2 i n gang edging mode; 367-359=8 seconds to process the only p i e c e needing edging from c e n t e r cant on Edger; and 305-303=...=431-429=2 seconds to process the f i n a l lumber p i e c e s on Trimmers. These time data agree with those of SAWSIM (Figure 7.1). No t i c e that machine o p e r a t i o n time data of SAWSIM output are t o t a l times needed to process a l l p i e c e s at a machine. Consequently, the t o t a l time must be d i v i d e d by the number of p i e c e s ( i n "PASSES" l i n e s of SAWSIM output) to get the p r o c e s s i n g time per p i e c e . For i n s t a n c e , the manufacturing time of one sideboard on Combined Machine #2 i n edging mode i s 30/2=15 seconds. The manufacturing time of headrigs always means the t o t a l time of primary breakdown. Fi g u r e 7.4 can a l s o be used to check that the time sequence of o p e r a t i o n [2] i s c o r r e c t . A f t e r having f i n i s h e d primary l o g breakdown on Headrig 8' at s i m u l a t i o n c l o c k time 27, manufacturing of the two sideboards and the center cant on Combined machine #2 can s t a r t at 63 and f i n i s h at 109 s i m u l a t i o n time. A l l of the p i e c e s can only be processed on Trimmers a f t e r t h i s time. There i s only one p i e c e produced from the cente r cant which needs edging. Obviously, t h i s edging o p e r a t i o n must be c a r r i e d out before the f i n a l trimming. T h i s time sequence of ope r a t i o n shows that the s i m u l a t i o n agrees with the r e a l o p e r a t i o n sequence. To check i f p i e c e s flow through the r i g h t t r a n s p o r t a t i o n 119 HEADRIG 8' COMB.MACH. J2 EDGER TRIMMER 01 TRIMMER «2 LOG(XACT3) 27 SIDEBOARD SIDEBOARD CANT (XACT4) (XACT5) (XACT6) . 63 78 93 109 ., -Af^i t I I „i 359 PFCC TO BE EDGED (XACT9) SIDEBOARD TO BE TRIMMED (XACT10) SIDEBOARD TO BE TRIMMED (XACT8) SIDEBOARD TO BE TRIMMED (XACT7) 321 325 319 323 327 «U I I I I I W-429 431 PFCC TO BE TRIMMED (XACT15) PFCC TO BE TRIMMED (XACT11)»-PFCC TO BE TRIMMED (XACT12) — PFCC TO BE TRIMMED (XACT13)*— PFCC TO BE TRIMMED (XACT14).-SIMULATION TIME (SEC) 10 SEC PFCC - PIECE FROM CENTER CANT Figure 7.4. Time chart of log and piece processing equipment [3] in both space and time, let us consider the interactive GPSSH run output again (Appendix 7.1). All of the pieces cut from the log above (Figure 7.1) can be checked by looking up the corresponding transactions. The route of their movement can be checked by finding out the correspondence between GPSS blocks and transportation equipment (Appendix 7.1). As an example Table 7.2 summarizes the movement of one of the 120 sideboards, represented by t r a n s a c t i o n #4. Table 7.2. Movement of sideboard ( T r a n s a c t i o n #4) between Headrig 8' and Combined Machine #2 T r a n s p o r t a t i o n GPSS block S i m u l a t i o n time Time needed equipment i d e n t i f i c a t i o n of t o flow A r r i v a l Leave through Chain #4 (COMB2+5) 27 36 9 Com 2 (COMB2+10) 36 40 4 Chain #5 (COMB2+15) 40 58 18 R o l l #2 (LBL49+2) 58 63 5 No t i c e that Table 7.2 d i s p l a y s four important f a c t s suggesting v a l i d p i e c e flow s i m u l a t i o n . F i r s t , the times needed to flow through t r a n s p o r t a t i o n equipment are c o r r e c t . For i n s t a n c e , the speed and len g t h of Chain #5 are 10 inch/second and 180" ( 1 5 1 ) , r e s p e c t i v e l y . Thus, the I8sec p e r i o d of time needed to flow through Chain#5 i s c o r r e c t . Second, the a r r i v a l time to Chain #4 and the time when l o g p r o c e s s i n g ends at Headrig 8' (Figure 7.4) are the same. T h i r d , the time of l e a v i n g R o l l #2 corresponds to the time when si d e b o a r d ( t r a n s a c t i o n #4) p r o c e s s i n g s t a r t s on Combined machine #2 (F i g u r e 7.4). F i n a l l y , t r a n s p o r t a t i o n equipment p a r t i c i p a t i n g i n moving the sideboard are those which are i n s t a l l e d between Headrig 8' and Combined machine #2; hence they a re c o r r e c t . To see i f p i e c e s i z e s , taken i n t o account t o c a l c u l a t e t r a n s p o r t a t i o n equipment u t i l i z a t i o n , are s a t i s f a c t o r y [4] a monitored s i m u l a t i o n run was c a r r i e d out. The word " s a t i s f a c t o r y " vs. c o r r e c t must be emphasized because SAWSIM does not provide p i e c e s i z e s i n i t s c u r r e n t s t a t e . Thus, the s i z e s , sampled from p r o b a b i l i t y d i s t r i b u t i o n s by FLOWSIM, are 121 sometimes not a c c u r a t e . However, i n the long run t h e i r averages are c l o s e to the a c t u a l p i e c e s i z e s . For monitoring purposes at a p p r o p r i a t e p o i n t s of the model the GPSS processor was stopped by s e t t i n g up break p o i n t s , and t r a n s a c t i o n s were d i s p l a y e d by the "DISPLAY XACT=n" command. The second p a r t of the i n t e r a c t i v e GPSS output (Appendix 7.1), under the t i t l e "CHECKING PIECE SIZES", d i s p l a y s t r a n s a c t i o n s with t h e i r parameters. Here again F i g u r e 7.3 might be h e l p f u l to a p p r e c i a t i o n of the correspondence between t r a n s a c t i o n s and p i e c e s . For convenient comparison, Table 7.3 l i s t s a c t u a l p i e c e s i z e s based on SAWPLOT output and s i z e s c o n s i d e r e d by FLOWSIM. Cor r e c t n e s s of machine u t i l i z a t i o n s t a t i s t i c s [5] and i n t e r a c t i o n between u n i t s of m i l l machinery [6] were checked by a d e t e r m i n i s t i c run. A s i n g l e l o g was made to flow through the m i l l and i t s SAWSIM r e s u l t s are d i s p l a y e d i n Fig u r e 7.5. By c o n t r o l l i n g l og a r r i v a l to Headrig 9', breakdowns of Chain #1 and the preceding Headrig 9' machine, the h i s t o r y of events becomes known and the time p e r i o d s of Headrig 9' being i n one of i t s four p o s s i b l e s t a t e s ( i d l e , busy, blocked, and down) can be c a l c u l a t e d and compared to FLOWSIM r e s u l t s . For s i m p l i c i t y , d u r i n g the dynamic s i m u l a t i o n of manufacturing only Headrig 9' was c o n t r o l l e d and caused to breakdown and be i n a blocked s t a t e " a r t i f i c i a l l y " . T h i s " a r t i f i c i a l " c o n t r o l was c a r r i e d out by r e p l a c i n g s t o c h a s t i c a l machinery breakdowns and l o g a r r i v a l with predetermined timing of events, a c c o r d i n g to the event schedule in F i g u r e 7.6 by which the d e t e r m i n i s t i c machinery breakdown was c o n t r o l l e d . F i g u r e 7.6 a l s o c o n t a i n s the time i n t e r v a l s of the 122 T a b l e 7 . 3 . C o m p a r i s o n o f a c t u a l a n d FLOWSIM p i e c e s i z e s Piece d e s c r i p t i o n (» )Ac tua l s i ze of Equivalent d1am length i th lckn. width xact lon inch f t inch inch number (* )S lzes taken into account by FLOWSIM dlam length th lckn. width inch f t Inch inch p a r a m e t e r 11 7 9 8 Log 18 21. 1 XACT3 18 21 12 S1deboard 15 XACT4 a f t e r headrig 13 Sideboard 15 XACT5 a f t e r headrig 14 Center cant 18 XACT6 Sideboard 21. 1 3 12 XACT7 21 3 9 to be trimmed S1deboard 21 . 1 3 12 XACT8 21 3 9 to be trimmed Pfcc to be edged 21 . 1 3 9 XACT9 21 3 9 Pfcc to be trimmed 21 . 1 3 9 XACTIO 21 3 10 Pfcc to be trimmed 21. 1 3 9 XACT11 21 3 8 Pfcc to be trimmed 21? 1 3 9 XACT12 21 3 8 Pfcc to be trimmed 21 < 1 3 9 XACT13 21 3 8 Pfcc to be trimmed 21. 1 3 9 XACT14 21 3 9 Pfcc to be trimmed 21. 1 3 9 XACT15 21 3 12 Pfcc - P iece from center cant (*)Note: Table contains only relevant s i zes to c a l c u l a t e t ransportat ion equipment occupancy. f o u r s t a t e s o f H e a d r i g 9 * . T h e s e t i m e i n t e r v a l s a r e 333 ( 5 + 2 6 + 3 0 2 ) , 2 9 , 20 a n d 100 s e c o n d s f o r i d l e , b u s y , b l o c k e d , a n d down s t a t e s , r e s p e c t i v e l y . T h e s e d a t a c o r r e s p o n d t o t h e d a t a o f FLOWSIM o u t p u t ( F i g u r e 7 . 7 ) a n d i n d i c a t e v a l i d m a c h i n e u t i l i z a t i o n s t a t i s t i c s . I n t e r a c t i o n b e t w e e n C h a i n #1 a n d H e a d r i g 9 ' was a l s o c o r r e c t s i n c e t h e e n d o f b l o c k s t a t e a n d t i m e when C h a i n #1 b e c o m e s a v a i l a b l e a g a i n c o r r e s p o n d . T h e f o r e g o i n g d i s c u s s i o n on v a l i d a t i o n i n many c a s e s i s s i m p l i f i e d a n d d o e s n o t g i v e t h e d e t a i l s o f t h e w h o l e w o r k w h i c h was a c c o m p l i s h e d t o p r o v i d e c o n v i n c i n g e v i d e n c e a b o u t t h e v a l i d i t y o f F L O W S I M . T h e o n l y p u r p o s e o f t h i s r e l a t i v e l y s h o r t 123 <% LOG SAWING -- OFFSET -- OIA LEN TAPES --SWEEP" CUFT PCS FBN FBM/ CHIP LUMBER BVPRO COST GROSS * <X I DENT PATTERN HOB VER IN FT IN/FT EDGE CVL ACTUAL LBR LBR CUFT CUFT % % % t CUNIT <X> Ul /I UH01 0.500 O.O 15 00 1.30 O 109 0.0 0.0 13.1 13 117 9.73 2.6 20.53 3 81 0 0 33.34 193 5 <X HEDRG9' HE0RGB1 HS9-CE1 HR9-CG1 HR9-GM6 MB9-TWN HR8-CE3 HR8-CG2 HR8-GNG HR8-TWN CE1-TRM CG1-TWN CG1-0TW CG1-E0G C0< <X PASSES 3 3 1 5 2 <% SECONDS 39. SO. 10. 10. 12 <X t <X CE2-TRM CG2-TWN CG2-0TW CG3-EOG CC3-TM GNG-TWN GNG-OTW GNG-EOG GNG-TRW TWN-TWO TWN-EOG TWN-TRM DTK-TWO DTW-EDG P'w <X PASSES <X SECOkOS <x t <X TWO-TOT TWO-EOG TWD-TRM T01-E0G TO1-TRM EDG-TRM CHIPPER <X PASSES 2 2 / <X SECONDS 4 <X S SAWSIM PLOT Ul / 1 UH01 2X06 8 2X08 a 3X0B 8 2X08 8 2X08 8 2X08 8 ONE INCH • • | 2X06 8 I " F i g u r e 7.5. Log used to check machine u t i l i z a t i o n s t a t i s t i c s d i s c u s s i o n was to present the p r i n c i p l e s of v a l i d a t i n g FLOWSIM. To d i s c u s s a l l of the s t e p s taken i n the v a l i d a t i o n p r o c e s s goes beyond the scope of t h i s t h e s i s . However, c e r t a i n a d d i t i o n a l a c t i o n s taken i n c o n n e c t i o n with v a l i d model behaviour must be mentioned. FLOWSIM w r i t e s e r r o r messages on the output f i l e i n the 124 SIMULATION STARTS CHAIN #1 BREAKS DOWN CHAIN #1 IS AVAILABLE AGAIN LOG ARRIVAL; PROCESSING STARTS 1 LOG PROCESSING ENDS 20 1 34 54 LAST PIECE LEAVES MILL; SIMULATION HEADRIG 9' ENDS IS AVAILABLE AGAIN ' HEADRIG 9' BREAKS DOWN 80 180 482 SIMULATION TIME (SEC) BUSY 29 BLOCKED 20 IDLE 26 DOWN IDLE 100 302 . GPSS SEGMENT . OF HEADRIG 9' . BREAK DOWN . GENERATE ...1..1 . . ADVANCE 80 . FUNAVAIL HRG9 . SEIZE DWNH9 . ADVANCE 100 . FAVAIL HRG9 . RELEASE DWNH9 TERMINATE . GPSS SEGMENT . OF CHAIN * 1 . BREAK DOWN . GENERATE . ASSIGN 1.6 . ADVANCE 20 . SUNAVAIL PI . ADVANCE 34 . SAVAIL PI . TERMINATE F i g u r e 7.6. Predetermined breakdown events and the e q u i v a l e n t GPSS program segments causing these events to happen case of erroneous h e a d r i g s e l e c t i o n segment o p e r a t i o n and p i e c e flow l o g i c i s checked at c r i t i c a l p o i n t s of the program. A u x i l i a r y programs d e t e c t e r r o r s caused by mistakenly prepared bucking program input and format disagreement between SAWSIM output and DIA w r i t e r program. Erroneous model o p e r a t i o n can a l s o be d e t e c t e d by mo n i t o r i n g and checking data t r a n s m i t t e d SIMULATION RESULTS OF MACMILLAN BLOEDEL WR3 SAWMILL DESIGN. 02:12 P.M. MAR. 19. 1984 Sawmill operation was simulated for 482 sec. MACHINE UTILIZATIONS AND PIECE COUNTS. Name Idle Busy Blocked Down Piece Count T ime % T ime % Time % T ime y. HRG9 333 .0 69 .0 28.9 6.0 20.0 4 . 1 99.9 20 .7 1 HRG8 482 .0 100 O 0.0 0.0 0.0 0.0 0.0 O .0 o CMB1 454 . 0 94 . 1 27.9 5.6 0.0 0.0 0.0 0 .0 4 CMB2 482. 0 ioo 6 0.0 0.0 0.0 0.0 0.0 b .0 0 GANG 482. 0 100 0 0.0 0.0 0.0 0.0 0.0 0 0 0 TWIN 482 . 0 1CO o O.O 0.0 O.O 0.0 0.0 0. 0 o EDGER 470. 0 97 .5 11.9 2.4 0.0 0.0 0.0 0 0 2 TRMR1 468. .0 97 .o 13.9 2.9 0.0 0.0 0.0 0 0 7 TRMR2 470 0 97 .5 11.9 2.4 0.0 0.0 0.0 0 .0 6 TRANSPORTATION EQUIPMENT PIECE COUNTS. TRANSPORTATION EQUIPMENTS: PIECE COUNTS: BLT01 13 BLT02 0 BLT03 0 BLT04 2 BLT05 O TRANSPORTATION EQUIPMENTS: PIECE COUNTS: CHN06 0 CHN07 0 CHN08 8 CHN09 7 TRANSPORTATION EQUIPMENTS: PIECE COUNTS: TRANSPORTATION EQUIPMENTS: PIECE COUNTS: CHN16 7 R0L09 O CHN17 6 R0L1O 0 R0L01 4 ROLII 2 R0L02 0 R0L12 0 CHN10 O R0L03 0 R0L13 2 CHN01 4 CHN1 1 0 R0L04 0 R0L14 0 CHN02 4 CHN12 2 R0L05 13 UNSC1 8 CHN03 O CHN04 O CHN05 0 CHN13 O CHN14 O CHN15 O R0L06 0 UNSC2 7 R0L07 O R0L08 O 1 26 between a u x i l i a r y programs. For i n s t a n c e , database of boom lo g s e l e c t i o n , DIA matrices i n r e c t a n g u l a r form e t c . , can be monitored. The importance of m i l l management c o o p e r a t i o n , in both the model b u i l d i n g and v a l i d a t i o n ( t e s t run) stages of model development, was recognized. T h i s management involvement c o n t r i b u t e d to a l a r g e degree to s i m u l a t i o n v a l i d i t y . To avoid the "garbage in garbage out" case, managers were c o n t i n u o u s l y being asked to d i s c u s s and check model i n p u t s . "Good sense" of t e s t run r e s u l t s were a l s o reviewed by subject-matter m i l l s p e c i a l i s t s . Sometimes, as a feed back of these d i s c u s s i o n s , model l o g i c and input data had to be changed u n t i l managers had found the r e s u l t s c r e d i b l e . 1 27 COMPUTER SIMULATION RUNS, RESULTS AND DISCUSSION In Chapter 3, the flowchart of F i g u r e 3.2 d e p i c t s the i n t e r r e l a t e d phases of the s i m u l a t i o n procedure. One phase can be thought of as an a c t i v i t y , a block with o b l i q u e s i d e l i n e s , producing r e s u l t s in b locks with v e r t i c a l s i d e l i n e s . These a c t i v i t i e s r e q u i r e a l a r g e amount of c a l c u l a t i o n s , and data manipulation and hence are c a r r i e d out by corresponding computer programs. The time sequence of how these a c t i v i t i e s are performed i s d i c t a t e d by the l o g i c of the whole procedure of F i g u r e 3.2. S i m i l a r l y , the execution of corresponding computer programs must f o l l o w a p r e c i s e l y determined order. In other words, when making a s i m u l a t i o n run, l o g i c a l l i n k a g e s between programs must be kept in mind which i s not an easy task. Hence, to help the program execution procedure, a l l of the programs were o r g a n i z e d i n t o one system of i n t e r a c t i n g programs. T h i s program system i s a r e s u l t i t s e l f , i n the sense that i t f u l f i l l s one of the o b j e c t i v e s of the t h e s i s r e s e a r c h . Thus, before d i s p l a y i n g and d i s c u s s i n g the s i m u l a t i o n r e s u l t s , t h i s chapter g i v e s an overview about the s t r u c t u r e of t h i s program system, and a c c o r d i n g l y c o n s i s t s of two p a r t s . 8.1. STRUCTURE OF INTERACTING PROGRAMS The best way to summarize the s i m u l a t i o n program system and i t s i n t e r a c t i n g elements i s to d e p i c t i t in the form of a 128 f l o w c h a r t . When runs were made, the flowchart i n F i g u r e 8.1, was found to be extremely u s e f u l p r o v i d i n g guidance to what programs to execute during the experimentation, and what sequence and what I/O l o g i c a l u n i t numbers to use to t r a n s f e r data between programs. There are two kinds of blocks on F i g u r e 8.1, blocks with o b l i q u e s i d e l i n e s and blocks with v e r t i c a l s i d e l i n e s . The former ones represent computer programs, whereas the l a t t e r ones represent e i t h e r input or output f o r these programs. Numbers along arrows are l o g i c a l input/output u n i t numbers. The flowchart i s d i v i d e d i n t o two by a h o r i z o n t a l dotted l i n e . The purpose of t h i s p a r t i t i o n i n g i s to separate the a c t u a l sawing and pi e c e f l o w s i m u l a t i o n programs (lower part) from programs d e a l i n g with sample l o g p r e p a r a t i o n (upper p a r t ) . Hence, i t i s p o s s i b l e to make p e r c e p t i b l e how changes of the upper p a r t input a f f e c t i t s output, which i s the input for the lower p a r t , the a c t u a l sawing and p i e c e flow s i m u l a t i o n . Seeing these i n t e r a c t i o n s between the two p a r t s we know what programs of the computer s i m u l a t i o n procedure must be rerun when experimenting with the model. There are only two input f i l e s of the upper p a r t : boom l o g measurements with bucking i n s t r u c t i o n s f o r the bucking program and c e l l boundaries f o r the sample sawlog s e l e c t i o n program. The other two programs, Data Base #1 w r i t e r (Appendix 8.1) and measurement w r i t e r (Appendix 8.2) do not have e x t e r n a l input. Other f i l e s are b i n a r y f i l e s of communication between programs and output f i l e s . The bin a r y f i l e s are not d i s p l a y e d on the flow c h a r t and are simply r e f e r r e d to by arrows having two 129 CELL BOUNDARIES OF SAWLOG SELECTION BOOM LOG BUCKING MEASURMENTS • INSTRUCTIONS SCRATCH FILE TO WRITE OUT AND READ IN FROM '' BUCKING PROGRAM SCRATCH FILE •JJQ_A*MPLE SAWLOG/ i t : (/SELECTION / 6 1.JWI 7 PROGRAM / y ERROR MESSAGES ON TEfWlfWl, DATA OUTPUT FOR I MONITORING PURPOSE51 ATA BASE «1 RITER PROGRAM i 5 7 2 3 10 X 11 |B DATA OUTPUT FOR MONITORING PURPOSES /UBC COMBINE/ /SUBROUTINE / MEASURMENT PROGRAM 3SCRATCH FILE! WRITER/ 17 SAWING PATTERNS OUTPUTS FOR SAWSIM RUNSJ — — ^- — =-L SAWS I M J I SAWSIM OUTPUTS (SAWING INSTRUCTIONS FOR FLOWSIM RUNS) DIA MATRICES IN RECTANGULAR|0>-FORM FOR MONITORING PURPOSES 4 DIA WRITER/ PROGRAM / A2_ ' MATRIX READER PROGRAM •er DATA BASE "2 OF NON-ZERO DIA MATRIX ENTRIES 7^ TABLE OF DATA BASE »1 FOR MONITORING PURPOSES /FOLLOWER CARD 'WRITER FOR BOOM LOG GENERATOR FUNCTION DATA BASE »1 FOR BOOM LOG SELECTION.BUCKING DECISION.SAWING MODE SELECTION •as FLOWSIM SAWLOG WEIGHTS BINARY INPUT AND RESULT FILES (LOG SELECT / 'FILE WRITER/ PROGRAM / LOG SELECT FILE  S A W R E S RESULTS OF MILL DYNAMICS PRODUCTION RESULTS F i g u r e 8.1. S t r u c t u r e of i n t e r a c t i n g programs 1 30 numbers as input/output l o g i c a l u n i t numbers. The output f i l e s are p a r t l y f o r v a l i d a t i o n purposes, and p a r t l y are r e l e v a n t input f i l e s f o r the computer s i m u l a t i o n procedure of lower p a r t . There are two arrows c r o s s i n g the dotted b o r d e r l i n e , i n d i c a t i n g two kinds of input f o r the lower computer s i m u l a t i o n procedure. F i r s t , are input f i l e s f o r SAWSIM runs which are the so c a l l e d l o g measurement f i l e s . Second, i s Data Base #1 which i s the input of FLOWSIM v i a i t s matrix reader program. These two input f i l e s of the lower p a r t change i f , and only i f , the sawlog lengths are changed by using d i f f e r e n t bucking i n s t r u c t i o n input and/or by v a r y i n g the accuracy of the sawlog s e l e c t i o n program by changing the c e l l boundaries. Consequently, rerunning the upper part i s only r e q u i r e d i n the case of changing the above in p u t . As the flowchart i n d i c a t e s , beside l og measurement f i l e s , sawing p a t t e r n f i l e s must a l s o be set up before running SAWSIM. The sawing p a t t e r n data f i l e s d e s c r i b e the complete sawing process i n terms of the number of sawlines and the d i s t a n c e s between them, and the r e l a t i v e p o s i t i o n of the l o g or piec e to the saws as w e l l as the p r o c e s s i n g machines. C a r e f u l reading of the SAWSIM manual and d e t a i l e d d i s c u s s i o n with m i l l managers on sawing p o l i c y were found extremely important to c r e a t e proper sawing p a t t e r n f i l e s . A f t e r having set up the two input f i l e s , SAWSIM runs can be made. For each p a i r of sawing p a t t e r n and log measurement f i l e s , separate SAWSIM runs must be made. Each output of these separate SAWSIM runs corresponds to one p a r t i c u l a r way of sawing (mode 1, mode 2, mode 8 columns on the r i g h t hand s i d e of Data Base #1, F i g u r e 6.7). Regarding 131 t h i s correspondence between SAWSIM output and "mode" columns of Data Base #1, there are two c r u c i a l p o i n t s to keep i n mind to run FLOWSIM f r e e from e r r o r . F i r s t , sawlogs belonging to d i f f e r e n t "mode" columns of Data Base #1 are r e p r e s e n t a t i v e s of d i f f e r e n t ways of sawing s u p p l i e d to the computer program v i a sawing p a t t e r n s . Thus, the r i g h t correspondence between sawing p a t t e r n s and "mode" columns of Data Base #1 i s c r i t i c a l . Second, the numbers in columns of mode 1, mode 2, ... of Data Base #1 are a c t u a l l y s e r i a l numbers r e f e r r i n g to the l o c a t i o n s of DIA matrices i n Data Base #2. Since these l o c a t i o n s are determined by how SAWSIM output i s l i n e d up as input f o r the DIA w r i t e r program, the sequence of t h i s input i s c r u c i a l for e r r o r f r e e s i m u l a t i o n . Reformatting SAWSIM output i n t o the format r e q u i r e d by FLOWSIM i s c a r r i e d out by two programs, the DIA w r i t e r program (Appendix 8.3) and the matrix reader program (Appendix 8.4). The DIA w r i t e r program has two outputs. One c o n t a i n s the dynamic i n f o r m a t i o n a r r a y s in r e c t a n g u l a r and hence convenient form to make monitoring p o s s i b l e . The other output c o n t a i n s only non-zero e n t r i e s of DIA matrices i n the r e q u i r e d and e f f i c i e n t format f o r the matrix reader program. The matrix reader program i s run together with FLOWSIM and hence ensures the presence of Data Base #1 f o r FLOWSIM to make p o s s i b l e boom log s e l e c t i o n s , bucking d e c i s i o n s and sawing mode s e l e c t i o n s . Data Base #2 i s a l s o set up by the matrix reader program. Whenever i n f o r m a t i o n c o n t a i n e d by Data Base #2 i s needed, the DIA f i n d e r subroutine i s c a l l e d by FLOWSIM to l o c a t e and return the r e q u i r e d i n f o r m a t i o n . 1 32 G e t t i n g the two data bases ready , i t se t s up the stage of FLOWSIM run which produces two k inds of o u t p u t : r e s u l t s of m i l l dynamics and sawlog w e i g h t s . Sawlog weights are f r e q u e n c i e s showing how many t imes the sample sawlogs were s e l e c t e d by FLOWSIM d u r i n g the s i m u l a t i o n r u n . Sawlog weights are input to the l o g s e l e c t f i l e w r i t e r program (Appendix 8.5) whose task i s to produce input for SAWRES r e r u n s . In a d d i t i o n to the l o g s e l e c t f i l e , the b i n a r y input and r e s u l t f i l e s c o n t a i n i n g the sawing r e s u l t s of sample sawlogs from p r e v i o u s l y performed SAWSIM runs , serve as the base fo r t o t a l p r o d u c t i o n c a l c u l a t i o n s c a r r i e d out by SAWRES. 8 . 2 . RESULTS OF SIMULATION RUNS AND DISCUSSION S i m u l a t i o n runs were grouped a c c o r d i n g to the o b j e c t i v e s of a n a l y s i n g the e f f e c t of v a r i o u s overseas markets and machinery breakdown. A c c o r d i n g l y , t h i s segment d i s c u s s e s the r e s u l t s of market and machinery breakdown runs , s e p a r a t e l y . Among the market runs o n l y the U . S . A . market run i s d i s c u s s e d in d e t a i l to demonstrate the k i n d and source of i n f o r m a t i o n that can be ga ined by r u n n i n g the s i m u l a t i o n programs. Other market run r e s u l t s are not d i s c u s s e d i n s i m i l a r d e t a i l , but they are r e p r e s e n t e d and compared i n a summarizing t a b l e . C o n c e n t r a t i n g now o n l y on the U . S . A . market , to get p r o d u c t i o n and m i l l dynamics r e s u l t s , the m i l l o p e r a t i o n was s i m u l a t e d for one s h i f t - 460 minutes = 27600 seconds [ 1 ] . (References to s i m u l a t i o n output are g i v e n i n square b r a c k e t s throughout t h i s s e c t i o n ) . Numbers on the condensed output ( F i g u r e 8 . 2 - 8 . 3 ) , c o r r e s p o n d i n g to numbers i n square b racke t s of SIMULATION RESULTS Of MACMILLAN BLOEDEL WR3 SAWMILL DESIGN. 01 . 35 P M MAY 06. 1984 Sawmill operation was simulated for 27600 nnc. H* C l-f ro n o 3 ro 3 w ro o 0 C rt C rt MACHINE UTILIZATIONS AND PIECE COUNTS. Name HRG9 HRG8 CMB1 CMB2 GANG TWIN EDGER TRMR1 TRMR2 Idle T Ime Busy T Ime Blocked 0.0 11.0 395 460 18295 27600 1632 4462 4673 0.0 0.0 , ,1.4 100.0 5.9 16. 1 16.9 26069.0 27102.9 27204.9 27139.9 9304.9 0.0 25905.9 23137.9 22926.9 -\13 TRANSPORTATION EQUIPMENT PIECE COUNTS.7 Piece Count & 2842 2927 474 0 3385 11569 1 1464 TRANSPORTATION PIECE COUNTS: EQUIPMENT: BLT01 9428 BLT02 8487 BLT03 5294 BLT04 3394 BLT05 0 CHN01 3055 CHN02 3050 CHN03 195 CHN04 3222 CHN05 3221 TRANSPORTATION PIECE COUNTS: EQUIPMENT: CHN06 280 CHN07 0 CHN08 13346 CHN09 13231 CHN10 0 CHN1 1 0 CHN12 3394 CHN13 0 CHN14 0 CHN15 0 TRANSPORTATION PIECE COUNTS: EQUIPMENT: CHN16 1 1593 CHN17 11482 R0L01 2843 R0L02 2928 R0L03 474 RDL04 0 R0L05 9457 R0L06 0 R0L07 8490 R0LO8 5305 TRANSPORTATION PIECE COUNTS: EQUIPMENT: R0L09 O ROL10 O R0L11 3385 R0L12 O R0L13 3384 R0L14 0 UNSC1 13300 UNSC2 13175 co U) MACHINE UTILIZATION BAR CHARTS. IDLE STATE . (%) • • * » « * » • » « • • * • • • * * * HRIG9 HRIG8 COMB 1 COMB2 GANGE TWIN EOGER TRIM! TRIM2 BUSY STATE {%) HRIG9 HRIG8 COMB 1 COMB2 GANGE TWIN EOGER TRIM1 TRIM2 r