Industry Analysis: The Fastener

Supply Chain in Aerospace Industry

Emad Zilouchian Moghaddam, Arturo Cardenas Martinez,

Saeed Koochak-Yazdi, Hani Murad 1

Desautels Faculty of Management,

McGill University, Montreal QC H3A 1G5 Canada

emad.zilouchianmoghaddam@mail.mcgill.ca, arturo.cardenas@mail.mcgill.ca,
saeed.koochakyazdi@mail.mcgill.ca, hani.murad@mail.mcgill.ca

Abstract

In a highly regulated industry such as the Aerospace it is impossible to have suppliers entering
and leaving as they please. Entry barriers raise the importance of understanding the behavior of
the entire supply chain system ranging from turbine engines to fasteners; any part can force the
plane production to stop when there is a supply shortage. Nowadays, every purchasing order
placed by Aerospace, from OEMs to high-tech suppliers, is followed thoroughly. In contrast,
fasteners fall in the category of nuts and bolts for their simplicity and high volumes and often
they are taken for granted. This paper focuses on the fastener supply chain, developing a system
dynamic model to show insights of its behavior and to set the base for a further improvement
analysis.

Keywords:
Fasteners, Supply Chain, Aerospace, System Dynamics, and Vendor-Managed Inventory

Introduction

The Aerospace industry is continuously facing risks and instability at every stage of its supply
network. Operating in the volatile environment of Original Equipment Manufacturers (OEMs)
such as Boeing, Airbus and Bombardier are directly affected by various factors including market
uncertainty, global economic challenges, fluctuating oil prices, fierce competition, and ever-
changing international politics. As a consequence of the recent global recession, many of the

1
The authors would like to thank Antoine Deshors (Bombardier Aerospace Senior Supply Chain Agent) and Ignacio
Gallo (Bombardier Aerospace Supply Chain Business Analyst) for hosting this research and making data available.
We also want to acknowledge Jeroen Struben (Assistant Professor at Desautels Faculty of Management, McGill
University) for his ongoing support and valuable comments to improve the quality of our works.
suppliers have exited the industry or have been forced to adjust their production capacity in
response to the sharp fluctuations in demand. As a result, the aerospace Supply Chain (SC) has
been tremendously fragmented.
However, over the past few years the industry started to bounce back and began a production
ramp up to cope with the projected growing trends in demand. In part, this recovery is the result
of new technology developments, lighter materials, and more fuel-efficient designs. This
situation has shifted the pressure upstream in the SC, pushing suppliers to expand production
capabilities and raising uncertainty whether the pipeline is prepared to keep pace with such a
rapid change.
When talking about airplane parts, what are the first things that people think of? Turbine engines,
electronic control devices, landing gear, seats, wings, to mention few in other words the big,
visible, high tech and expensive parts of the airplane. But even the production of a multimillion-
airplane can be stopped due to the most costly insignificant part: a fastener. One piece that cost
few cents can compromise the production of a several million dollars airplane. For instance, a
Boeing 747 has more than six million parts, from which more than half are fasteners2. Under
those circumstances, a widespread shortage of fasteners will disrupt the entire SC, increase lead
times, create order backlogs, and end up in frequent changes in production rates. Furthermore,
evolving industry practice and aerospace regulations are causing parts to become obsolete which
in turn creates tremendous strains on suppliers production capacities.
The recent economic downturn has forced fastener manufacturers in the market to consolidate.
As a result, OEMs are now relying on fewer and larger suppliers who not only serve the
aerospace industry but also supply to a wide range of sectors such as marine, construction, and
automotive. For that reason, there is an increasing concern that fastener suppliers may not be
able to accommodate their expected demand requirements for the next two years. Single source
suppliers that were able to survive the industrys slowdown only increase the complexity of the
supply base and intensify the fragility of the entire network.
Over the years, OEMs have been working closely with multiple distributors and third-party
logistics (3PL) providers in an effort to mitigate risks, optimize forecasting, improve inventory
management, and reduce costs. At the same time, industry-wide trends and polices such as
vendor-managed inventory (VMI), strategic contract agreements, and collaborative relationship
management are becoming instrumental in driving industrys dynamics and ensuring business
continuity across the SC structure.
This Industry Analysis is aimed to understand the behavior of the fastener SC system by taking
the interaction across different variables into consideration, identify scenarios that can create
production stoppage, and explore ways to reduce the probability production stoppage.

2
747 Fun Facts. Boeing. Internet. July 2011.
Overview of the Fastener Supply Chain System

OEMs have two types of fastener suppliers: manufacturers and distributers. Each supplier varies
in size and capability. Large fastener manufacturers such as Alcoa produce the majority of
fasteners in the world. They have the capability to quickly increase their production capacity;
however they serve multiple industries at the same time, therefore in case of demand increase, it
will be difficult for them to adapt effectively and simultaneously across all industries. On the
other hand, fastener distributors such as Anixter maintain inventory levels for multiple OEMs
giving them the advantage to add demand volumes and increasing their purchasing power. Figure
1 illustrates how players interact in the SC of fasteners. For instance, Manufacturer 1 not only
supplies to all distributors, it also supplies to OEM 1 directly, and the same behavior is shown
from Manufacturer 2 to OEM 2, while Distributor 3 only supplies to OEM 2. These are
few examples of the many forms that players connect among each other.
Information is shared across the SC by sending Purchasing Orders (POs) from OEMs to
distributors and manufacturers. Since distributors dont produce any material they only store it
once they receive a PO they forward it to manufacturers. In opposite direction, the material is
shipped in accordance to the POs from manufacturers to distributors and OEMs as well as
from distributors to OEMs. Fasteners are not exclusive to one industry; they are as much
required in Aerospace as in Automotive or Machinery markets. Moreover, these markets are
usually correlated thus when ones market demand improves, other industries behavior follow
the same pattern. This phenomenon raises the importance of the buying power of both the
company and the market or industry. A better idea for each market purchasing power can be seen
in figure 2.

Methodology

The main objective of this analysis is to understand the dynamics of the fastener SC system when
shortage risk is involved and there is an abrupt demand growth. Likewise, it is important to
identify ways to improve the performance of the system through scenario evaluations that
consider the dynamic behavior caused by feedbacks and delays.
A system dynamics methodology was followed that enabled us to develop a model capable of
tracking the behavior over time in response to changes in the system. At first, a meeting with
Bombardier Aerospace (BA) supply chain experts was held to list and identify important
variables and come up with the respective reference modes3. Then, a casual loop diagram (CLD)
was created with the information gathered and follow-up meetings were arranged (feedback
elicitation) to include BAs contacts in the process and to validate the system captured in the
CLD. Finally a model was developed by taking into consideration the input from previous

3
David F Andersen and George Richardson, Scripts for group model building System Dynamics Review Volume
13 Number 2 Summer 1997
meetings, multiple interviews, workshops, and information obtained from other sources. Most of
the quantitative information collected throughout this analysis was confidential, therefore was
disguised.
The model is comprehensive in scope, capturing the critical relationship between parameters
such as production rates, supplier capacity, order demands, backlogs, inventory levels and
capacities, including links in the SC (fastener manufacturers, fastener distributors, and aerospace
OEMs). For analytical clarity the model is deliberately stylized in terms of the technical details
of the existing SC elements, for example aggregates over most SKUs and market players.
The underlying tool is a practical computer-based simulation developed in Vensim. This tool will
also be used to diagnose the various factors that are affecting the flow of fastener across the SC
and to explore potential outcomes for alternative inventory management decisions and polices as
well as other forms of coordination across the SC, under a variety of economic scenarios. The
model is designed to develop and explore what-if questions rather than serving as a black-
boxed forecasting tool. The current phase of analysis does not involve a detailed calibration to
the entire SC and set the bases for further analysis in the future.

Model Description

Since numerous variables affect the fastener SC system. Considering all of them would create a
complex model and complicate the analysis. Our model captures the essence of most
representative supply chain actors: OEM, distributors and manufacturers. There are two types of
manufacturer suppliers: major manufacturers who serve all market sizes/types and minor
manufacturers who only supply to the OEM. Distributors bridge the gap between major
manufacturers and the OEM by buying products from the former and supplying to the later. The
model is based on the material flow shown on figure 3.
Our model also considers the effect of the demand of other aerospace OEMs and the rest of the
industries requirements on common suppliers by calculating OEMs aerospace market share and
the aerospaces market share in the fastener industry. If the company has a 50% of aerospace
market share, according to figure 2 it will posses 7% share of the entire fastener industry. For the
purpose of analysis we assumed that airplanes are built only with fasteners. This will set the
boundaries to isolate the model to the fastener industry.
Our model represents each player as a subsystem except for Other OEMs. Subsystems are
connected by information flow upstream and material flow in the opposite direction. The OEM
sends POs to their minor manufacturer, their major manufacturers, and their distributers who will
forward the PO they just receive to their own suppliers major manufacturers. Processed
according to their capabilities, suppliers produce or receive the fasteners and then ship them to
the OEM.
Airplanes are very expensive therefore it is not viable for any company to have a stock of
airplanes waiting to be sold. This explains why aerospace is a Make to Order industry. A
backlog order is where a customer order received is placed until the plane production stars.
This is commonly used in this kind of industries. When a demand increase occurs, it affects the
OEMs backlog and triggers the reaction of the system to adapt to that change.
Purchasing orders are send from the OEM to its suppliers after considering its consumption rate
and the forecast for the expected future demand. In a similar way, the distributor calculates their
POs quantities based on the received order and their own forecasted values before sending their
POs to their suppliers major manufacturers.
Contrary to OEMs, distributors behave like a Make to Stock type of industry mainly because
they do not produce anything. To cover themselves from any demand volatility, they stock more
inventory than needed. At the short run this seems to be an advantage, but it is not sustainable
and later on the analysis will be demonstrated that produce noise in the system incurring in more
costs.
Manufacturers adopt their production rate according to the POs received from OEMs and
distributors. In addition, major manufacturers take into consideration demand from the other
industries by including their demand volumes to the production rate. Because all the industries
are correlated, the model considers that when there is a demand increase in one industry, the
other industries will follow the same path. As a result, there is a huge demand increase at the
major manufacturers order receiving rate that makes difficult to catch up with production for
everyone despite their unique capacity adaptation characteristic. On the other hand, there is a
delay for implementation of changes in manufacturing floor along with delay of realization of
need to make those changes.

Detail Analysis:

OEM:

The model builds upon policy structure diagrams of inventory and production4. At the top of
the system hierarchy, the OEM subsystem controls the input of demanded fasteners in the system
and the output of shipped airplanes delivered to the customer (Figure 4).
The customer behavior is reflected in the system via Customer Order Rate. When in
equilibrium, this flow depends entirely on the companys market share of the industry the
market percentage of the company from the industry percentage of the fastener industry. One of
the questions that the industry experts were curious about was to see if the industry would be
able to keep up with an unexpected demand increase. In order to do so, the model includes two
variables to create a step increase at specific time.
Every customer order received will pass through the Customer Order Rate and end up in the
OEM Backlog stock and every order processed will flow out by Order Start Rate. This
outflow depends directly on a parallel material flow called Production Start Rate that

4
John Sterman, Business Dynamics: Systems Thinking and Modeling for a Complex World. Boston:
Irwin/McGraw-Hill, 2000. Print. Pages 709-723.
represents the production capacity of the company in terms of inventory availability. Once the
production of an airplane is started, it flows to the WIP (Work In Progress) stock.
Simultaneously, the backlog orders flow at the same rate to the Started Backlog stock. A WIP
airplane is a Started Backlog Order. Finally, the Shipment Rate represents the airplanes
delivered to customers.
As mentioned before, Production Start Rate represents the production capacity of the OEM in
terms of inventory available and do not depend on the production capacity installed at their
facilities (Eq. 1). The larger the OEM Backlog stock is, the larger the PSR will be, as long as
there is available inventory. Additionally, if the OEM is able to reduce their Production Start
Time the PSR will also increase if and only if there is available inventory to do so.

Equation 1:

The importance of Production Start Rate resides in the inclusion of the Actual Stoppage
variable in its calculation as a result of the Actual Stoppage Loop. This loop certifies that there
are enough fasteners in stock to cover the Required Inventory for Expected Production Start in
the Critical Time Horizon (Required Inventory for EPS in CTH). The inventory validation loop
begins tracking any change in the Production Star Rate by comparing it with the Expected
Production Start Rate; if there is any discrepancy the stock will adapt in the ESPR Average
Time period through the Change in EPSR rate. Afterwards, the system adapts the expected
airplanes per month to cover the Expected Production Start Rate in the Critical Time Horizon
and then convert the units from airplanes to fasteners at the Required Inventory for EPS in
CTH variable. The next step in the loop is to calculate the Probability of Reaching a Critical
Point as shown on Eq. 2.

Equation 2:

It exist a major reliability to some fasteners more than others reason why by using a Table of
Stoppage it is possible to consider a reliable value to continue production. The Table of
Stoppage as shown on Figure 5 was validated by industry experts considering the system
major assumptions and proved through evaluation of past scenarios.

At the end of the loop Actual Stoppage compares the Probability of Reaching Critical Point
to the Actual Stoppage Point resulting in a Go-or-No Go value [1,0].
Equation 3:

The OEM subsystem sends their Purchasing Orders to its suppliers with the OEM Fastener
Order Rate variable. It has been said that the Fastener Inventory stock increases with the
Fastener Receiving Rate and decreases with the Consumption Rate. Similar to the Expected
PSR (Short term) stock mentioned in the Actual Stoppage loop, the OEM calculates the OEM
Expected Order Rate as some sort of forecast considering the current rate values and comparing
it to any changes in the Customer Order Rate. Then, the stock value plus a safety stock is
compared to the Fastener Inventory resulting in Adjustment for inventory. As shown on Eq.
4, the OEM will order the OEM Expected Fastener Consumption Rate in addition to any
required adjustment.

Equation 4:

Distributor:

Distributors play a vital role in the system by shipping 50% of the Fastener Receiving Rate to
the OEM and an even more critical role when there is a short-term market fluctuation.
Introducing distributor in the supply chain not only decreases the inventory holding cost of the
OEM but also gives the buying power to distributor among the different OEM demand to
decrease the cost of ownership of the fasteners from Manufacturer. Since they hold the inventory
for the OEM its model is similar to a general policy structure diagram model. Contrary to the
OEM, instead of Make to Order, distributors model is Make to Stock. Based on their
received orders, they place their orders to Major Manufacturers and ship to the OEM based on
current demand and expected growth in future.
Manufacturer:

Major and minor manufacturers are the last subsystems in the supply chain model. They receive
raw material from the mills and produce fasteners for the industry. As mentioned, one of the
characteristics of manufacturers is their ability to adapt their capacity. Figure 8 shows the basic
dynamics of the fastener manufacturers in a casual loop of capacity.
In the CLD the Backlog stock has as an inflow the order rates and as an outflow the shipment
rate. This flows control the level at which the backlog stock can be found at any given time.
When backlog is built-up it creates delivery delays, which will generate pressure to expand the
manufacturers capacity in order to improve the shipment rate. New facilities are acquired to
adjust to the actual capacity with the desired shipment rate. This set of variables creates a
balancing feedback loop that tends to control the delivery delay. A weaker balancing loop
controls the utilization of the capacity.
Taking a step further away, a higher fastener order rate will not only affect the backlog but also
will increase the expect order rate, which in turn will exert more pressure on fasteners
manufacturers to expand capacity. The figure 9 illustrates the final casual loop for manufacturer.
Based on the concept of Economies of Scales, the manufacturer with higher capacity has higher
revenue and as a result has higher ability to expand its capacity in less time. With this in mind
the model differentiates major manufacturers from minor ones. In addition, learning capabilities
for new employee and new machinery should be taken to the account in the form of
delays.
Transferring the previous CLD to a stock and flow diagram, the Manufacturer WIP will have
Manufacturer Production Start Rate as the raw material inflow. This rate is driven by Capacity
and the Utilization, which is effected by Capacity and Desired Production Start Rate. The
Desired Production Start Rate can be found in typical manufacturing model as a result of
Desired Production and Adjustment for WIP. (Refer to Major Manufacturer Model in
Appendix).
A short-term solution for changes in the Desired Production Start Rate would modify the
Utilization that affects the Manufacturer Production Start Rate. At the long term, there
should be a change in capacity that will be triggered by Pressure to Expand Capacity. The
manufacturer will start to build up capacity as a consequence of the Capacity Adjustment at a
Capacity Acquisition Delay and later installed capacity at a Capacity Installation
delay.
Results

The model starts at equilibrium only to receive a shock in the customer order rate and analyze
the resulting behavior. For each of the 3 simulations the shock is represented by a step increase
of 10%, 20% and 30% respectively at the 24th month. The graphs in figure 11 show the resilience
experienced by the OEM subsystem after the shock demand at each simulation. The step increase
presented on the Customer Order Rate produces more backlog orders and consequently, a
goal-seek increase behavior on the work in progress (WIP). Since more production demands
more from the inventory the Expected Order Rate increases until reaching equilibrium.
During the 30% increase simulation the system presents stoppage months after the demand shock
was introduced. In order to analyze the details of this event we looked at different variables in
every subsystem of the model. Although the Inventory level of OEM should increase accordingly
to satisfy the new demand, information and material delay produce an over shoot and an under
shoot on the inventory levels.
Behind the production stoppage is the behavior of the suppliers and its reaction to demand
changes (Figure 12). On one hand, distributors are capable of supplying fasteners at the same
rate in every simulation despite the fact that their inventory levels fluctuate. Distributors
business model is to have material available every time is needed by customers and since they
dont produce anything, they need to assure they have enough inventory considering the
manufacturers delivery delay. This extra inventory consideration is what keeps distributors
shipping at the same rate without affecting its delivery delay. On the other hand, manufacturers
are not able to produce enough fasteners to keep shipment rate constant extending its delivery
delays across the system resulting on production stoppage.
Capacity represents a major difference between suppliers, while manufacturers are constrained to
their machines capacity; distributors only need to store the ordered product. This explains why
Distributor Inventory overshoots and Manufacturer Inventory behaves different.
Comparing the expected order rates across the supply chain explains the influence of information
delay on the system (figure 13). Contrary to material delays, there is no capacity constrain in the
information that can be shared. Therefore, manufacturers expected order rate can overshoot and
manufacturers inventory wont overshoot. Between Customer Order Rate and OEM
Expected Order rate there is a 6 months forecast delay explaining the gradual growth until
equilibrium. As mentioned before, distributor needs to counteract their lack of production
capability by storing inventory, making Distributor Expected Order Rate more sensible to
changes. Upstream, manufacturers have their own Manufacturer Expected Order Rate but by
the time they receive the information, it already has the influence of the expected order rate of
distributors and OEM adding noise to the system.
Vendor Managed Inventory Policy

Vendor Management Inventory (VMI) is a distribution channel operation system where the
manufacturer/vendor monitor and manage the inventory at distributor/retailer. In a VMI
partnership, the distributor makes main inventory replenishment decisions for the OEM. The
vendor monitors the buyers inventory levels physically or via electronic messaging and makes
periodic resupplies decisions regarding order quantities, shipping, and timing5. The inventory
policy demands suppliers to replenish inventory for OEM and supply in cycles. Suppliers set a
target inventory for manufacturers accompanied with more frequent inventory cycles of
complement and joint distribution to achieve economic benefits.
The VMI policy included in the model shifts the forecast information from OEM to the
distributor representing vendor access to the buyers inventory levels and reducing information
delays. Distributors will know exactly how many fasteners are on stock and how many are
needed to continue production. The information confidence reduces overshoot in distributor
order rate and is reflected even at the manufacturers expected order rate. The consolidation of
order rates help avoid the stoppage presented on previous simulations, at the same time lower the
inventory level required by the distributors. Figure 14 exhibit the distributors inventory levels
drop compared to the previous simulations. In addition, the manufacturer visibility improves
resulting in faster inventory adaptability. Manufacturers also acquire new capacity more
effectively reproducing less variation in delivery delays. Overall, we can see significant
improvement in the supply chain by comparing figure 13 with figure 15, where variability of the
bullwhip effect it is reduced.

Conclusions and Future actions

The model represents the fastener supply chain dynamics, making possible to analyze its
behavior based on the information provided by industry experts. In addition, the robustness of
the model allows testing different policies and scenarios representing plausible outcomes. For
instance, increasing the fastener safety stock will save the company stoppages fees at expense of
adding more inventory carrying costs. Modifying the time to average order rate changed the
forecasted fasteners either increasing volatility or reducing response time presenting stoppage in
both scenarios. Alternative policies were discussed based on simulation results. The insights gain
on each simulation help identify what variables influence the system more than others.
As a result, information delays were identified as an important variable and selected to control
the performance of the system. The selection was made considering experts feedback and
implementation feasibility. This is shown in the model in the performance improvement of the
Vendor Management Inventory simulations.

5
Allan N. Portes, and Ghilherme E. Vieira, The Impact Of Vendor Managed Inventory (VMI) On The Bullwhip
Effect In Supply Chains.
In conclusion, distributors play a critical role in the supply chain and the OEM should establish
strategic sourcing relations to improve the confidence of the information sharing. Further steps
are needed to improve the sensitivity of the system. The current stage model was designed with
the intention of introducing executive decision-makers to System Dynamics and to give them the
capability to explore scenarios of their own, challenge the underlying assumptions, and examine
effectiveness of combining policies.
Figures:

Figure 1: Fastener Supply Chain interaction

All Other OEM US Fastener Industry Estimated Market Share, 2010

7%
Electrical/Electro
nic Motor Vehicles
8% 25%
Fabricated Metal
Product
10%
Industrial
Machinery
13%
MRO,
Construction,
Aerospace Other
14% 23%

Figure 2: Fastener Industry market segmentation6

6
Industrial Fastener Institute. Annual Report 2010. Printed. July 2011.
 Figure 3: Material Flow Model Overview

Demand Fractional
shock Demand
OEM Initial Time Shock
Market Share

OEM Initial Customer Order

Aircraft Demand Rate
OEM Backlog Started Backlog
Order start rate Order
Aircraft Demand Fullfilment rate

+ Shipment Rate
Expected PSR WIP
(short term) Production Production Start
change in EPSR rate
Start Time Rate

Production
Expected Production EPSR Average Time Actual Stopage Cycle time
Start CTH
Table of Stopage +

Actual Stoppage Loop Actual Stoppage

Critical Time Probability of Point
Horizon reaching critical
Required point
Inventory for
EPS in CTH +
Min Inventory Max Inventory
Depletion Time Consumption Rate
Fasteners Per
<Fasteners Per Airplane
Airplane>

<Small Manufacturers Consumption Rate

Fastener Inventory
Total Shipment Rate>
Fastener
<Distributor Receiving Rate
Shipment Rate>
+
<LM OEM
Shipment Rate>

Figure 4: The OEM Model Policy Structure Diagram

-
 1
0.9
0.8
0.7
0.6

Outout
0.5
0.4
0.3
0.2
0.1
0
0 0.5 1 1.5 2
Input

Figure 5: The Table of Stoppage

+

<Small Manufacturers
Total Shipment Rate> Consumption Rate
Fastener Inventory
<Distributor Fastener
Shipment Rate> Receiving Rate
+
<Large Manufacturers to
OEM Shipment Rate>

Fastener Inventory
OEM -
AdjExpected
OEM Time
Fastener Adjusment for
Fastener Consumption Order Inventory
Rate Rate
+ Normal Suppliers
Lead Time
<Fasteners Per
Airplane>
Desired Fastener +
Inventory
OEM Safety
Stock
OEM Expected
Order Rate
OEM Change in
Expected Order Rate
- <Customer
Order Rate>

Time to average
order rate

Figure 6: The OEM Model Order Structure Diagram

 <OEM Distribution Order
Fastener Percentage Distributor
Order + Delivery Delay -
Rate>

Distributor +
Order Rate
Distributor Order
Backlog
Distributor Order
Fulfillment Rate
+

Dis. Min. Inventory Dis. Maximum

Inventory
Shipment Rate
Consumption

Distributor
Shipment Rate
Distributor Inventory
<LM Dist Inventory
Shipment Rate> Receiving rate

-
Avg. order
Distributor
Distributor Fastener + processing time
Adjusment for
Order Rate
Inventory

Distributor Expected
Backlog

Figure 7: The Distributor Model Order Structure Diagram

s

-
Backlog

- Delivery +
+
Supplier Delay
Utillization Shipment Rate
-
+

Capacity Pressure to
expand capacity
+

Capacity +
Acquisition + Desired
Capacity

Figure 8: The Manufacturer First Causal Loop Diagram

 s

-
Backlog +
Fastener Order
Rate
Learning Curve

- +
+ Delivery Delay
++ Supplier
Utillization Shipment Rate
-
+
B

+
B
Capacity Pressure to
expand capacity
+

+
Capacity Expected
+
Acquisition Order Rate
+ Desired Capacity
-
+

Capacity
Acquisition Delay

+
R
Economy of Scale
Ability to Expand
+

Figure 9: The Manufacturer Second Causal Loop Diagram

Capacity Capacity Manufacturer WIP

Instalation Delay Obsolescence Time Manufacturer
Production Start Rate
++
Capacity
Instalation Rate Capacity
Capacity Obsolescence Utillization
Adquisition Capacity -
Change in Buffer
capacity
Desired Production
Capacity Start Rate
Capacity
Adjustment Acquisition Delay
Planned Pressure to
Capacity Expand Capacity

Effect of expansion Table of effect of

Desired
pressure on desired expansion pressure on
Capacity
capacity desired capacity

Figure 10: The Manufacturer Model Capacity Acquisition Diagram

 Customer Order
Customer Order Rate Rate WIP
WIP
140 100

125 90

Aircrafts
Aircrafts

Aircraft

Aircraft
110 80

95 70

80 60
0 24 48 72 96 120 144 168 192 216 240 0 24 48 72 96 120 144 168 192 216 240
Time (Month) Time (Month)
Demand Increase 10% Demand Increase 30% Demand Increase 10% Demand Increase 30%
Demand Increase 20% Inventory Level
Fastener Inventory BA of OEM
Demand Increase 20% Expected
BombardierOrder Rate
Expected Order Rate of OEM
Fastener Inventory BA
600 M 140
600 M

525 M 125
525 M

Aircraft/Month
Fasteners

Fastener

Fasteners

Aircraft
450 M 110
450 M

375 M 95
375 M

300 M 80
300 M
0 24 48 72 96 120 144 0 24 48 72 96 120 144 168 192 216 240
0 168 24 192 48 216 72 240 96 120 144 168 192 216 240
Time (Month) Time (Month)
Time (Month)
Demand Increase 10% Demand Increase 30% 10% Demand Increase 10% Demand Increase 30%
Demand Increase Demand Increase 30%
Demand Increase 20% Demand Increase 20%
Demand Increase 20%

Figure 11: The Simulation Results of Step Increase in Demand

Distributor Inventory
Distributor Inventory Manufacturer Inventory
Manufacturers Inventory
450 M 20 B

362.5 M 15 B
Fastener

Fastene
Fasteners

Fasteners

275 M 10 B
r

187.5 M 5B

100 M 0
0 24 48 72 96 120 144 168 192 216 240 0 24 48 72 96 120 144 168 192 216 240
Time (Month) Time (Month)
Demand Increase 10%
Demand Increase 20% Distributor Delay
Distributor Delay
Demand Increase 30%
DistributorDemand
DelayIncrease 20%
Demand Increase 10%
Manufacturer Delay
Demand Increase 30%
Manufacturer Delay
8 8 12

7.5 7.5 9
Months
Months

Months
Month

Month

7 7 6

6.5 6.5 3

6 6 0
0 24 48 72 96 120 144
0 168
24 192
48 216
72 240
96 120 0144 24168 48192 72
216 96
240 120 144 168 192 216 240
Time (Month) Time (Month) Time (Month)
Demand Increase 10% Demand
DemandIncrease
Increase30%
10% Demand
DemandIncrease
Increase 10%
30% Demand Increase 30%
Demand Increase 20% Demand Increase 20% Demand Increase 20%

Figure 12: The Simulation Results of Step Increase in Demand

 Customer Order Rate
Customer Order Rate
OEM Expected
Bombardier Order
Expected Order Rate Rate
140 140

125 125

Aircraft/Month
Aircraft
Aircrafts

Aircraft
110 110

95 95

80 80
0 24 48 72 96 120 144 168 192 216 240 0 24 48 72 96 120 144 168 192 216 240
Time (Month) Time (Month)
Demand Increase 10% Demand Increase 30% Demand Increase 10% Demand Increase 30%
Demand Increase 20%Distributor Expected Order Rate
Distributor Expected Order Rate
Demand Increase 20%
Manufacturer Expected Order Rate
Manufacturer Expected Order Rate
80
Distributor Expected Order Rate
15,000
80

70 11,250
70
Fastener

Fastener
Aircraft

Aircraft
60
Aircraft

7,500
60

50 3,750
50

40 0
0 24 48 72 96 120 40144 168 192 216 240 0 24 48 72 96 120 144 168 192 216 240
Time (Month) 0 24 48 72 96 120 144 168 192 216 240Time (Month)
Demand Increase 10% Demand Increase 30% Time (Month)
Demand Increase 10% Demand Increase 30%
Demand Increase 20% Demand Increase 10% Demand
Demand Increase
Increase 20%30%
Demand Increase 20%

Figure 13: The Simulation Results of Step Increase in Demand Bullwhip effect

Distributor Inventory
Distributor Inventory VMI VMI Manufacturer Inventory
Manufacturers Inventory VMI VMI
450 M 20 B

337.5 M 15 B
Fastener

Fastener
Fasteners

Fasteners

225 M 10 B

112.5 M 5B

0 Distributor Expected Order

0 Rate (VMI)
18 27 36 45 54 63 72 81 90 99 108 117 18 27 36 45 54 63 72 81 90 99 108 117
80Time (Month) Time (Month)
Demand increase 10% Demand Increase 20% VMI Demand Increase 10% Demand Increase 20% VMI
Demand increase 10% VMI
Demand Increase 20%
Manufacturer Delay
Manufacturer Delay VMI VMI
Demand Increase 30%
Demand Increase 30% VMI
Demand Increase 10% VMI
Demand Increase 20%
Manufacturer Capacity
Manufacturer Capacity VMI VMI Demand Increase 30%
Demand Increase 30% VMI

12 70 7B
Aircraft/Month

9 5.75 B
Fastener/Month

60
Fasteners/Month
Months

Month

6 4.5 B

50
3 3.25 B

0
40
2B
18 27 36 45 54 18
63 7227 81 36 90 45 99 54
108 63
117 1872 27 81 36 9045 99
54 63 108 72 11781 90 99 108 117
Time (Month) Time (Month) Time (Month)
Demand Increase 10% Demand Increase 20% VMI Demand Increase 10% Increase Demand Increase 20% VMI
Demand Increase 10% VMI
Demand Increase 10%
Demand Increase 30%
Demand 20% VMI
Demand Increase 10% VMI Demand Increase 30%
Demand Increase 20% Demand Increase 10%Increase
Demand VMI 30% VMI Demand Increase 30%
Demand Increase 20% Demand Increase 30% VMI
Demand Increase 20% Demand Increase 30% VMI

Figure 14: The Simulation Results of Step Increase in Demand VMI

 Customer Order
Customer Order RateRate VMI OEMBombardier
Expected Order
Expected Order Rate VMI
Rate (VMI)
140 140

125 125

Aircraft/Month
Aircraft
Aircrafts

Aircraft
110 110

95
Distributor Expected95Order Rate (VMI)
80
80 80
0 24 48 72 96 120 144 168 192 216 240 18 27 36 45 54 63 72 81 90 99 108 117
Time (Month) Time (Month)
Demand Increase 10% Demand Increase 30% Demand Increase 10% Demand Increase 20% VMI
70
Distributor
Demand Increase 20% Expected
Distributor Expected OrderOrder
Rate (VMI)Rate VMI Manufacturer
Demand Increase 10% VMI
Demand Increase 20% ManufacturerExpected Order
Expected Order Rate (VMI) Rate VMI
Demand Increase 30%
Demand Increase 30% VMI
Aircraft/Month

80 15,000

60
70 11,250
Aircraft/Month

Aircraft/Month
Fastener

Fastener
60 7,500
50

50 3,750

40
40 0
18 27 36 45 54 18 63 2772 8136 90 4599 54
108 117 63 72
18 27
8136 45
90 54
9963 72
108 81
117
90 99 108 117
Time (Month) Time (Month) Time (Month)
Demand Increase 10% Demand Increase 20% VMI Demand Increase 10% Demand Increase 20% VMI
Demand Increase 10% VMI Demand Increase 10%
Demand Increase 30% Demand
Demand Increase Increase
10% VMI 20% VMI Demand Increase 30%
Demand Increase 20% Demand Increase 10% VMI
Demand Increase 30% VMI Demand
Demand Increase 20% Increase 30% Demand Increase 30% VMI
Demand Increase 20% Demand Increase 30% VMI

Figure 15: The Simulation Results of Step Increase in Demand Bullwhip effect VMI
References

[1] John Sterman, Business Dynamics: Systems Thinking and Modeling for a Complex
World. (Boston: Irwin/McGraw-Hill, 2000. Print.)
[2] Paitoon Vashirawongpinyo, A System Dynamics Model to Analyze Behavior of
Manufacturing in Supply Chain. The 2nd RMUTP International Conference 2010:
Green Technology and Productivity. Bangkok, Thailand. 29-30 June 2010. 2010.
Internet. July 2011.
[3] James M. Lyneis, System Dynamics In Business Forecasting: A Case Study of the
Commercial Jet Aircraft Industry. The 16th International Conference of The System
Dynamics Society. The System Dynamic Society. Quebec City, Canada. July 20 - 23,
1998. Internet. July 2011.
[4] William Killingsworth with Chavez Regina and Martin Nelson, The Dynamics of
Multi-Tier, Multi-Channel Supply Chains for High-Value Government Aviation Parts.
The 2008 International Conference of the System Dynamics Society. Athens, Greece.
July 20 24, 2008. Internet. July 2011
[5] Tu Yi-Ming, Ho Chun Fu, Teng Rodney, Yi Chien Ching. VMI and international
Supply Chains: A Case of Fastener Industry. Proceedings of the 2003 System
Dynamics Conference. The System Dynamic Society. New York City, Usa. July 20-23,
2003. Internet. July 2011
[6] David F Andersen, and George Richardson, Scripts for group model building
System Dynamics Review Volume 13 Number 2 Summer 1997
[7] Allan N. Portes and Ghilherme E. Vieira, The Impact of Vendor Managed Inventory
(VMI) On The Bullwhip Effect In Supply Chains. Third International Conference on
Production Research Americas Region 2006. Internet. July 2011
Appendix

OEM Model:

OEM Initial
Market Share Demand shock Fractional
Time Demand Shock

OEM Initial Customer Order

Aircraft Demand Rate

OEM Backlog Order start rate Started Backlog

Order
Fullfilment rate
Aircraft Demand

Shipment Rate
Expected PSR +
(short term) WIP
change in EPSR rate Production Start
Production Start Rate
Time

Expected Production Production

EPSR Average Time
Start CTH Cycle time
Actual Stoppage
Table of Stopage Point
Critical Time
Horizon Required Inventory Probability of
reaching critical
for EPS in CTH point + Actual Stopage
+
Min Inventory Max Inventory
<Fasteners Per Depletion Time Consumption Rate Fasteners Per
Airplane> Airplane

Consumption Rate
Fastener Inventory
<Small Manufacturers Fastener
Total Shipment Rate> Receiving Rate
+

<Distributor
OEM -Adjusment for
Shipment Rate>
Fastener
Order Inventory
<LM OEM Rate + Normal Suppliers
Shipment Rate> Lead Time
Fastener Inventory
Adj Time
OEM Expected <Fasteners Per
Fastener Consumption Desired Fastener + OEM Safety Stock
Airplane> Inventory
Rate

OEM Expected
Order Rate
OEM Change in
Expected Order Rate
-
<Customer Order
Rate>
Time to average
order rate
Distributor model:

<OEM Distribution Order

Fastener Percentage Distributor
Order + Delivery Delay -
Rate>

Distributor +
Order Rate
Distributor Order
Backlog
Distributor Order
Fulfillment Rate
+
Dis. Min. Inventory Dis. Maximum
Shipment Rate Inventory
Consumption

Distributor
Shipment Rate
Distributor Inventory
<LM Dist Inventory
Shipment Rate> Receiving rate
Delivery Time
VMI Switch
OFF
-
Distributor Avg. order
+ Distributor processing time
Fastener Order Adjusment for
Rate Inventory Delivery Time
VMI Switch
Distributor Fastener + ON
VMI
Inventory Adjusment <Desired Production Switch
Time for OEM VMI>
<Fasteners Per
Airplane> Distributor Desired Desired Production <OEM Fastener
+ Fastener Inventory for OEM VMI Order Rate>
Distributor Expected Distributor Expected
Aircraft Order Rate Backlog
Distributor Change in <Distribution Order
Expected Order Rate Distributor Percentage>
Safety Stock
+

Distributor Time to
average order rate

<Distributor
Order Rate>
Major Manufacturer Model:

<OEM
Fastener
Order
Rate>
Manufacturers
Order Percentage

<Distributor Fastener OEM Total

Order Rate> Orders Rate Large Manufacturers
Order Percentage

+Manufacturer
Normal Aerospace
market share Delivery Delay
All Industry Large Manufacturers
Orders Rate -
Order Rate
Manufacturer Order
<OEM Initial Backlog
Market Share> Manufacturer Order
Fulfillment Rate
+
Manu. Min Manu. Max
Inventory shipment
inventory
rate consumption

Larger Manufacturers
Total Shipment Rate
+
Capacity Capacity Manufacturer WIP Manufacturer Inventory
Instalation Delay Obsolescence Time Manufacturer Manufacturer
Production Start Rate Production Rate 0
++
Capacity Manu. order
Instalation Capacity processing time
Capacity Rate Obsolescence Utillization Manu. Cycle Time Manufacturer
- - Safety Stock
Adquisition Capacity
Change in Buffer Adjusment of WIP Manufacturer
capacity Adjusment for
Desired Production +
Capacity Desired WIP Inventory +
Capacity Start Rate
Adjustment Acquisition Delay + Manufacturer Desired Inventory Cycle <Manu. Cycle
WIP adj. Time Fastener Inventory Time Time>
Planned Capacity Pressure to
Expand Capacity
Desired Production +
+
Effect of expansion Table of effect of
Manufacturer Expected
Desired Capacity pressure on desired expansion pressure on
Order Rate Manufacturer Change in
capacity desired capacity
Inventory Expected
- Order Rate +
<Fasteners Per
Adjustmen Time Airplane> <Large Manufacturers
Manufacturer Expected Manufacturer time to Order Rate>
Aircraft Order Rate average order rate
Minor Manufacturer Model:
<Manufacturers
<OEM Order Percentage>
Fastener <Large Manufacturers
Order Order Percentage>
Rate> Small Manufacturers
Order Percentage

SM Delivery Delay
+
-
+
Small Manufacturers
Order Rate
SM Order Backlog
SM Order
Fulfillment Rate +

SM Min
SM Max Inventory
Inventory
Consumption
Shipment Rate

Small Manufacturers Total

Shipment Rate
SM Capacity
Instalation Delay SM Capacity +
Obsolescence Time SM WIP SM Inventory
SM Production Start
SM Production Rate
Rate

SM Capacity SM Capacity +
Obsolescence SM Order
SM Capacity Instalation Rate
Adquisition SM Capacity SM Utillization SM Safety Processing Time
Small Mfg Cycle
Buffer Time - Stock
SM Change
in capacity SM Adjusment
of WIP SM Adjusment
for Inventory +
SM Capacity SM Desired Small Mfg +
Acquisition Delay Production Start Rate Desired WIP SM Desired
+ SM Inventory <Small Mfg
SM Fastener
SM WIP Cycle Time Cycle Time>
SM Capacity Planned Inventory
Adjustment Capacity Adjustmen Time
SM Pressure to
Expand Capacity
SM Desired Production +
SM Effect of expansion +
SM Desired
pressure on desired
Capacity capacity SM Expected Order Rate
SM Change in Expected
Order Rate +
SM Inventory -
SM Table of effect of Adjustmen Time
expansion pressure on <Small Manufacturers
desired capacity Order Rate>
SM Time to Average
Order Rate