A SIMULATION-BASED STUDY AND ANALYSIS OF THERMO-PLAST COATING PROCESS IN ALVENIUS COMPANY KPP319 INDUSTRIAL EXCELLENCE Authors Group 8 (Zeinab Sayahi, Umer shafique) School of Innovation, Design and Engineering – IDT Mälardalen University, Eskilstuna, Sweden
26
Embed
ANALYSIS OF THERMO -PLAST COATING PROCESSzoomin.idt.mdh.se/course/kpp319/HT2012/Report... · KPP319 Industrial Excellence ABSTRACT This paper represents a simulation based study of
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
A SIMULATION-BASED STUDY AND ANALYSIS OF
THERMO-PLAST COATING PROCESS
IN
ALVENIUS COMPANY
KPP319 INDUSTRIAL EXCELLENCE
Authors
Group 8
(Zeinab Sayahi, Umer shafique)
School of Innovation, Design and Engineering – IDT
Mälardalen University, Eskilstuna, Sweden
KPP319 Industrial Excellence
ABSTRACT This paper represents a simulation based study of “Thermo-Plast Coating process” in Alvenius
Company. Alvenius is a swedish leading manufacturer of steel and aluminum pipes. It was
founded in 1951 in Eskilstuna while they started the production with steel pipe systems used in
mining industry. Today, products of this company, based on their usages, are divided into five
main categories including: fire protection, industry, mining and tunneling, snowmaking and
Miscellaneous.
This project concentrates on the production line of the newest product from Alvenius, the inside
and outside thermoplastic coated pipe, called TP-pipe system. This system has within several
applications outperformed stainless steel, when it comes to corrosion protection and service life.
The aim of the project is to perform a deep study on the manufacturing process, visualize and
analyze the current process flow by the help of simulation softwares, find probable problems in
the process and finally come up with some recommendation for production improvement and
efficiency enhancement. It has been tried to provide the readers with a clear view of the
production line, a precise analysis on process flow and introduce potential areas of
improvement.
The process flow undergoes some major bottlenecks which decrease the efficiency in the entire
production line. In this project, by the help of two simulation softawres, Extendsim 7 and 3D-
Create, an accurate analysis has been done on the current process flow. Furthermore, extra
scenarios of potential improvements have been simulated to investigate the effects of changes in
BACKGROUND AND DESCRIPTION OF SIMULATION CASE..................................................................................................................... 1
OBJECTIVES AND GOALS ........................................................................................................................................................................... 1
Project plan ............................................................................................................................................................................................ 2
PROCESS DESCRIPTION .................................................................................................................................................... 4
PROBLEM AND REASON FOR SIMULATION ............................................................................................................................................ 7
DATA COLLECTION ................................................................................................................................................................................... 8
MODEL ..................................................................................................................................................................................... 8
SIMULATION MODEL CONSTRUCTION................................................................................................................................................... 8
LOGICAL MODEL DESCRIPTION ........................................................................................................................................................... 14
3D MODEL DESCRIPTION ...................................................................................................................................................................... 19
VALIDITY AND RELEVANCE OF THE MODEL ....................................................................................................................................... 23
DESCRIPTION OF SIMULATION ............................................................................................................................................................. 23
RESULTS AND CONCLUSIONS ....................................................................................................................................... 26
Subsequently the pipes are heated in the oven, because they have to be hot, when they are
dipped into the Thermoplasting powder. Otherwise this powder wouldn´t melt on the surface
correctly. The oven has a temperature about 250°C to 300°C. The capacity in the oven depends
on the size of pipes. It can vary between 8 and 18 pipes. For the diameter of 114,3mm the
capacity is 18 pipes. Inside the oven, pipes are fixed on a conveyor belt without rotating them.
The heat in the oven is distributed by two fans, so that the pipes are heated evenly.
Figure 4: Oven Entrance and Exit doors
After heating, one pipe comes out of the oven and it is picked by a robot. First, pipe receives
coating substance on its internal surface. Same robot runs the pipe through the Thermoplast-
coating. Pipe is dipped into the coating material which is floating as powder in the bath. The
material melts on the pipe´s external surface because of the heat. In the coating section of the
process, first, pipes are coated inside. Afterwards, the pipe is completely submerged into the
coating powder bath for the external coating. The robot, which holds pipe, rotates the pipe
during this process. Because of the fixture in robot, only one pipe at a time can run through this
step.
KPP319 Industrial Excellence
6
Figure 5: External Coating Bath
In the next steps, pipes run through a manual repair. At the fixture points, pipes are not covered
with powder. So two workers cover the pipes at both ends with powder again and reheat the
pipe ends.
Figure 6: Manual Repair
Then the pipes are cooled down in a waiting queue and then tested on defects. With proof
equipment they generate a voltage. One clip is connected to metallic material. With the other end
they investigate the surface of the pipe. If there is any failure in the coating material a current
will flow and a sound signal will give a warning.
Figure 7: Test stage
KPP319 Industrial Excellence
7
Followed by another manual repair step, where the touch-points with the conveyor belt during
the cooling process are fixed, pipes` ends are closed by a plastic cap. Also a label is set on the
pipe which includes the manufacturer name and part name, sizes and ID number. At last they are
packed in groups of fourteen and loaded.
Figure 8: Packed Pipes, ready to leave
PROBLEM AND REASON FOR SIMULATION
By studying the production unit, it is obvious that there are some bottlenecks in material flow.
By using some inefficient machines, there are waste of time and energy. One of main problems in
the production process is the step between oven and Thermoplastic coating. The oven can hold
18 pipes at one time. After the heating, which lasts for 62,1 minutes for one pipe, the oven door
opens and three pipes come out, but only one pipe is picked up by the robot. During this time the
other two pipes lose some of their heat and have to go back to the oven to be heated again.
Beside this problem which pipes become colder, there are two other major disadvantages. The
first big disadvantage of this step is that a lot of heating energy gets lost, because the oven door
is opened for quite a long time. So the oven has to be reheated after every 8th pipe for about 15
minutes. Secondly the reheating takes not only energy and time in the production line, but also
turns the oven into a bottleneck in the process. A big amount of pipes accumulates in front of the
oven and has to wait. It makes the entire production inefficient and increases the costs.
Simulation is done to study current circumstance of production and investigate influencing
factors on material flow.
The objectives of simulating the process line of the Thermoplast-coating are:
• Simulate the current process of production in Extendsim and 3D-Create softwares
• Specify where there are bottlenecks in the process and find how to eliminate them
• Find which parts of the production can be improved and try to come up with some
solutions to improve those parts and consequently improve the whole production line
• Define an efficient production by minimizing bottle necks, cost and waste
• implement the improvements in the simulation softwares and get the result
• compare the verified model with the current one
KPP319 Industrial Excellence
8
DATA COLLECTION
Data related to production process are gathered directly from the production line. At the first
visit, the technical manager prepared the group by a comprehensive explanation about the
company, as well as thermo-plastic coating process, its material flow, layout of the production
facilities, and product specification. After first visit of the factory, a primary model of process
was prepared. And within the second visit, with the cooperation of production manager, more
data related to details of production line was collected. All involved equipments in the
production were investigated. Cycle time for all operations in the process, were collected.
Furthermore, details about machines and their characteristics were studied. Also, some data
related to company, its activities and products were extracted from Alvenius webpage.
MODEL
SIMULATION MODEL CONSTRUCTION
The entire simulation has been implemented with the Simulation Software ExtendSim7. The
production of the pipes can be described in a serial manufacturing process. Discrete event model
is used to simulate the process. Figure 9 illustrates the final simulation model.
Figure 9: final simulation model
For the Modeling we used blocks from Item, Value and Plotter libraries. Because a working day
last for 8 hours a day the running time (End Time) for our simulation is set to 8 hours.
As shown in figure 10, at the beginning, we prepared a simple simulation of the whole process
without going through the details. After model completion, we started to pay more attention to
details and make the process as it is in the real life.
KPP319 Industrial Excellence
9
Figure 10: Initial simulation model
The first version of Simulation is a good working but very abstract version of the production line
of TP-pipes. To adapt the simulation more to the real process, we had to do some improvements.
The simulation model starts with the outdoor storage which contains a large number of steel
pipes. Pipes are carried into indoor storage which includes 30 pipes. In the rout between
Outdoor and indoor storages, there is a Gate block beside a Decision block. During the process,
whenever the number of pipes in the Queue block, after the “Select Item In” block, falls under a
minimum of 5, the gate opens and more pipes are transferred from outdoor to fill the capacity of
30 pipes.
From this block, items are lead to the first processing step- the outside blasting. A simple
“Activity” blocks is suitable for this step. The outside blasting only takes 2.5 minutes. One item at
a time can pass through this activity block.
After the outside blasting, pipes go to the inside blasting. But we put a queue block before Inside
Blasting to observe the queue length before this step. Since inside blasting takes about 7
KPP319 Industrial Excellence
10
minutes, much more than outside blasting, it is a probable bottleneck in the process. Even
though two pipes are handled parallel in this step, pipes can be accumulated in front of this
machine due to its long cycle time. The same queue block should be between inside blasting and
the oven, because heating the pipes in the oven takes about 62 minutes. So we placed a queue in
front of the oven as well. A “Plotter” block will show us a record of queue length before inside
blasting and oven. Pipes enter inside-blasting machine 2 by 2.
To simulate the fact that inside blasting is done parallel for two pipes, we inserted a batch block
before inside blasting and an unbatch block after this activity, so that both pipes go inside the
activity exactly at the same time.
Oven is warmed up before the production is started and is only switched on when there are
enough pipes to fill the oven full – otherwise this would be waste of energy. In the real
production line, just in the beginning of the day, there are a lot of pipes stored as the queue
before the oven; which are produced the day before. To simulate same matter, we added an
extra “Resource item” block to fill the oven at the beginning of the production. Now we pretend
that the simulation starts in this moment the oven is switched on. In other words, when the
process starts, oven has a full capacity for 18 pipes. The oven is represented by a queue block
and the oven door is the actual activity block.
When the oven door opens a lot of heat comes out and heating energy gets lost, because the oven
door stays opened for quite a long time. Therefore the oven has to be reheated after every 8th
pipe. Reheating takes about 15min. For a shutdown of 15 minutes, for each 8-pipes, we have
used scheduled shutdown for the activity block of oven to adjust the time for every 8 pipes. After
27,6 minutes (the time for 8 pipes) a Create block creates an item that initiate a shutdown so
that no pipe can pass this activity during the next 15 minutes.
When the oven is under shutdown, there are no pipes go in or out. The model which we have
used in our simulation makes it possible to close the doors. It also enables the oven to release
only one pipe in every 3.45 minutes. The whole time for a pipe which stays inside the oven is
(18*3,45= )62.1 minutes.
KPP319 Industrial Excellence
11
The “information” block, beside the oven, records released pipes from oven with the time of
their exit. The table below, is a part of its record witch is reported by “Plotter, Discrete Event”
block. As highlighted in the table, between pipe no. 8 and 9 there is a shutdown.
Time No.
0 0
3,45 1
6,9 2
10,35 3
13,8 4
17,25 5
20,7 6
24,15 7
27,6 8
42,6 9
46,05 10
49,5 11
52,95 12
56,4 13
Table 1: released pipes from oven
The heating in the oven is followed by three short steps –inside Thermoplasting, outside
Thermoplasting and manual repair of the pipes. Both inside Thermoplasting and outside
Thermoplasting take not more time than one minute. The manual repair of the pipes is
necessary because the Thermoplasting material can´t reach the pipe´s surface at the fixtures
touch-points. So two workers have to cover the pipes at this point again with Thermoplasting
powder and reheat the pipe ends to melt the material.
KPP319 Industrial Excellence
12
Thereafter the pipes are stored to cool down. We implemented this part with another Queue,
because in the reality it is like a queue. Pipes stay in the queue until operators come to them for
testing, usually in the oven shutdown time.
One in ten pipes is tested for any probable failures in the coating. So in our model, we have used
a ”set” block with a “Random Number” block, to divide pipes by 0.9 and o.1 probabilities. Then
by using “Select Item Out” and “Select Item In” blocks we defined the Rout for pipes with the
probabilities of 90% and 10%. There is an activity block in the testing rout which will do testing
on the passing pipes. So, 90% of items pass the first line and go right to the next step and 10%
(one in ten) are leaded to the other line and pass test Activity block that has 1 minute delay.
This step is followed by two other Activity blocks- one for another manual repair and one for
equip the pipes with caps at the ends and a label introducing the manufacturer and pipe ID
number. Both steps take only ca. 1 minute. As the final step the pipes are packed in groups of 14
and loaded to a storage place. For packing the pipes together we used a Batch block in the
simulation. The process ends with the Exit block which represents the storing.
To measure the important information for our simulation we added also two Plotter blocks. One
measures the amount of items in the two queue blocks in front of the inside blasting and the
oven and compares both values. So we can observe which part of the production turns into a
bottleneck, when we change different parameters.
KPP319 Industrial Excellence
13
The other plotter measures the numbers and time of items passing the oven, the time and
numbers of shutdowns in the oven and also counts the produced packages of pipes at the end of
production line.
After running the simulation, we have following results from two plotters.
Figure 11: Queue Lengths before Inside Blasting and Oven
After Oven shut down Packages Products
0 0 0 0 0 0 0 0
3,45 1 27,6 1 71,95 1 12,1 1
6,9 2 42,6 0 150,25 2 15,55 2
10,35 3 70,2 1 236,25 3 19 3
13,8 4 85,2 0 320,25 4 22,45 4
17,25 5 112,8 1 405,25 5 25,9 5
20,7 6 127,8 0 471,9 6 30,35 6
24,15 7 155,3 1 480 6 32,8 7
27,6 8 170,3 0 51,25 8
42,6 9 197,8 1 55,7 9
46,05 10 212,8 0 58,15 10
49,5 11 240,3 1 61,6 11
Table 2: data from running the simulation
Queue before
Inside Blasting
Queue before Oven
KPP319 Industrial Excellence
14
LOGICAL MODEL DESCRIPTION
Description of used blocks in the model (Extendsim)
Resource <Item>: This item stores and prepares items which are used as product in the
production process. It is like a storage room which the user can define the initial numbers of
items.
Decision <Value>: This block makes a decision based on the inputs and defined logic between
two parameters A and B. The block lets the user to select a comparison between A and B values.
A is an input parameter; there is a connection on the top of block which can be connected to
another block to define the parameter of A. Parameter of B can be either defined by the user
within the block or be as input (if the B connection is connected to another block). It has two
outputs N and Y. If the condition is satisfied, output in Y is 1 and in N is 0.
KPP319 Industrial Excellence
15
Gate <Item>: It controls the passing of items through a portion of the model. The function of
this block can be defined under 3 types of condition: Conditional gating with items, conditional
gating with values, area gating. In this model “conditional gating with values” is used. In the
conditional gating with values, items will pass through gate, as long as the value at demand is
greater than 0.5. The value enters the block from “output Y” in “Decision” block. So, as long as the
condition in block decision is satisfied, it sends a value of 1 to the Gate Block. Since gate block
receives a value more than 0.5, it opens and lets the items to pass through.
Queue <Item>: It queues and releases items based on the conditions that are defined by the
user. In the “Queue Blocks” used in the model of this project, “minimum queue length” and
“minimum wait time” are defined in accordance with production specifications.
KPP319 Industrial Excellence
16
Activity <Item>: Holds a number of items and passes them out in accordance with the
parameters specified in Process tab. On the process tab user defines the capacity of the block
which is the number of items under operation in the block at once; and also how long each item
will take to process. In addition, shutdown can be defined for every activity block, by using
shutdown tab in this block. In the prepared model for this project, shutdown schedule is
received from an “Activity Block”. As shown in the picture below, “SD (shutdown) input from” is
adjusted to “value connection”. So when Activity block receives a value of 0 from Create block,
shutdown starts. And when it receives a value of 1 from create block, shut down finishes.
Create <Item>: In this model, create block prepares and sends values of 1 and 0 to SD
connection of Oven activity block, in accordance with a defined schedule. The schedule is
specified so that there is a 15-minutes shutdown after every 8 pipes in the oven.
KPP319 Industrial Excellence
17
Batch <Item>: This block joins several items as a single item. In this model it is used to
facilitate the “Inside Blasting” machine to process 2pipes at once.
Unbatch <Item>: Produces several items from a single input item. In this model, it is used
after Inside Blasting activity, to unbatch the 2-pipe batch which was created before.
Set <Item>: Assigns the properties of items passing through the block. In our prepared model,
the property values are input from “Random Number” block connector.
Random Number <Value>: As shown in the below figure, the first column values are defined
and in the second column the probability of that value is specified. The value column contains
“with test” and “without test” values that will be output. Probability describes the chance that
value will occur. In current project, 1 pipe in every 10 pipes will be tested. So, the probability for
“with test” is 0.1.
KPP319 Industrial Excellence
18
Select Item Out <Item>: This block determines the routs for input items with different
identities of “with test” and “without test”. It defines different values of 0 and 1 for “Select
Output”. And then allocates top output to items with “Select Output” value of 0.
Select Item In <Item>: In here, it merges all items together, either with test or without test.
Exit <Item>: Passes items out of the simulation. By the connection to a plotter, produced
packages in the model are recorded.
Plotter: In this simulation model, plotters record and illustrate following factors:
• Queue length before “Oven“ and “Inside Blasting“
• Time and Qty. of produced pipes and finished packages
• Time and Qty. of Shut-downs
KPP319 Industrial Excellence
19
3D MODEL DESCRIPTION
Following figures illustrate a view of components which are used in creating 3D-create simulation model. Components act as different operation steps. Figures also show the related
tables with used data within the components.
Inserting Pipes: Creator
The model starts to work with Creator component which inserts one pipe to the system every
150 seconds. 150 seconds is adjusted in accordance with outside blasting machine.
As explained in Extendsim model, in order to simulate the process as real as possible, and
considering that the oven is full in its capacity just when the process starts, a second creator is
used before oven to full fill this requirement.
KPP319 Industrial Excellence
20
Outside Blasting: manual Work Station
Inside Blasting: manual Work Station
Since Inside Blasting machine has a capacity of 2pipes, components Bundler and Debundler are
used to facilitate this purpose.
Oven: Process Point conveyor
KPP319 Industrial Excellence
21
Inside Coating: Component wrapper
Outside Coating: Process Machine Worktable and Robot “MD400N” with Machine Tending Robot