Dania -Academy of Higher Education 2016.04.10 . How t o assure a stable quality in a complex production system ? Supervis or: Mikael L. Nielsen… [601719]

2016
Áron Fekete
Automotive Technology
Dania -Academy of Higher Education
2016.04.10 . How t o assure a stable quality in
a complex production system ?
Supervis or: Mikael L. Nielsen
Number of characters: 44910

2

3

Table of content

Introduction ………………………….. ………………………….. ………………………….. ………………………….. ………………….. 5
Problem statement ………………………….. ………………………….. ………………………….. ………………………….. ………… 5
Methodology ………………………….. ………………………….. ………………………….. ………………………….. ………………… 5
Demarcation ………………………….. ………………………….. ………………………….. ………………………….. …………………. 6
Definitions ………………………….. ………………………….. ………………………….. ………………………….. ……………………. 6
Presentation of theory ………………………….. ………………………….. ………………………….. ………………………….. …… 8
The Diesel system ………………………….. ………………………….. ………………………….. ………………………….. ………. 8
Common rail ………………………….. ………………………….. ………………………….. ………………………….. ……………… 8
The performance line ………………………….. ………………………….. ………………………….. ………………………….. …. 9
The processes of the performance line ………………………….. ………………………….. ………………………….. ……… 9
Analysis ………………………….. ………………………….. ………………………….. ………………………….. ………………………. 15
Process Capability Test of G -type injector ………………………….. ………………………….. ………………………….. … 15
Comparison test of M -type injector ………………………….. ………………………….. ………………………….. ………… 30
Production bench levelling to quality ………………………….. ………………………….. ………………………….. ………. 38
Control items ………………………….. ………………………….. ………………………….. ………………………….. …………… 38
Management structure ………………………….. ………………………….. ………………………….. …………………………. 39
SWOT analysis ………………………….. ………………………….. ………………………….. ………………………….. …………. 41
Conclusion ………………………….. ………………………….. ………………………….. ………………………….. ………………….. 42
Putting into perspective ………………………….. ………………………….. ………………………….. ………………………….. .. 42
Bibliography ………………………….. ………………………….. ………………………….. ………………………….. ………………… 44

4

5

Introduction

In this project I am going to investigate how it is possible to have the same quality through a
complex production system. A huge manufacturer's like my internship company, its main task is t o
satisfy its buyer’s needs in the field of required quality, quantity and costs. In order to be able
satisfy the buyers the production system is requiring daily control and maintenance. I am going to
find out how is it managed in a huge plant.
The company where I spent three months as inter as engineer assistant in the Performance
engineering department is a Japanese company . The company was founded in the end of the 1940’s,
when they became independent from one of the biggest Japanese car producer. The co mpany has
188 locations globally, 61 in Japan, 28 in North America, 35 in Europe and 58 in Asia and 6 in
South America with more than 146000 employees. The company is being listed more than 20 years
on Fortune magazine’s top 500 companies globally. In 2015 the company was in the top 300 with
the revenue of more than 39 million dollars and in the top 5 largest auto parts suppliers.
The Hungarian plant was built in Székesfehérvár in the end of the 1990’s . It is situated
approximately less than one hour from the capital of Hungary, Budapest on the motorway. The
factory is situated in one of the industrial areas of Székesf ehérvár surrounded by other
manu facturers like Harman, Moñdelez, Loranger, Emerson, Jüllich Glass and Grundfos. According
to the newest emplo yment data the company has more than 5200 employees.
The factory produces diesel and gasoline car parts as well. The gasoline departments are producing
VVT systems, VCT systems, A/F sensors, EGTS sensors and EGR valves, while the main products
of the dies el department are injectors, common rail systems and fuel pumps. The main buyers of the
diesel products of the company are the leader companies of the automotive industry.
Probl em statement

How to assure the stable quality in a complex production system with several variations?
Method ol ogy

In order to get the best answer to my problem statement I will use the equipments of the company
and the knowledge of its employees from Japan and Hungary. I will use the performan ce line
number 5 and the final assemb ly line to answer my question. The stuff in the production area is

6

going to teach me how to use all the machines alone and my engineer colleagues are going to teach
me how to install a new product, in our case a new type of injector to the performance line and to
the final assembly line and how to get information from the histograms . I will get a better
knowledge how to use hand tools from Allen keys to the more complex oscilloscope. I am going to
compare the M -type injector results on two different lines t o see if the quality is the same or not.
The SWOT analysis of the company will be made. I will get more information how the everyday
life is going on in a factory.
Demarcation

The row materials from the suppliers are delivered to the location of the compa ny. After that come
the machining , the heat treatment, grinding and the assembly line . These are all processes before the
performance line and the final assembly line in the produc ing of an injector. These steps also have a
crucial effect on the final prod uct, but these steps are left out from this project.

Definitions

VVT – Variable Valve Timing
VCT – Valve controller timing
A/F – Air/Fuel
EGTS -Exhaust Gas Temperature Sensor
EGR -Exhaust Gas Recirculation
ECU – Engine Control Unit
NG- Not Good
Poke -yoke system – Process failure detection system with sensors RAW
MATERIAL MACHINING HEAT
TREATMENT GRINDING ASSEMBLY PERFORMANCE
LINE FINAL
ASSEMBLY

7

Nagara switch – Specially designed switch that can be activated with the swipe of the operator’s
hand to allow activating the production machines, while moving to the next station1.
BKD – Japanese abbr eviation for Bari Kirikuzu Dakon, which stands for burr, chip and nick
TST- Technical Support Team
Qm- injection quantity
Tq- opening time
PE- Production Engineering

1
http://www.larapedia.com/glossary_of_lean_six_sigma_terms/nagar a_switch_nagara_suicchi_meaning_and_definition_in_lean_six_sigma_termin
ology.html

8

Presentation of theory

The Diesel system
The diesel engine is an internal combustion engine invented in the 1890’s. The main principal of the
engine was written down by Rudolf Diesel, a German inventor and mechanical engineer, in his
work titled, Method of and Apparatus for Converting Heat into Work in 1892. First they were use d
for replacing the old steam engines. The diesel engine also has a two -stroke and four -stroke
versions. It was used in submarines and later heavy vehicles like, trucks and locomotives.
The four -stroke diesel engine is very similar to the Otto -engine. The main difference between the
two kinds of engine is that the Diesel engine does not have spark plugs. The reason behind that is
the diesel engine uses high compression to ignite the air -fuel mixture.
The four strokes are totally the same as in the Otto -engine, that is: intake, compression, combustion
and exhaust.
In the intake stroke the clean air is drawn into the cylinder with the atmospheric pressure of 1 bar.
In the compression stroke the fuel is injected into the cylinder and compressed with the air. The
compression ratio is 18:1 and the pressure is between 35 -40 bars.
In the combustion stroke the air -fuel mixture is ignited due to the high -pressure, which is between
50-150 bars and the temperature is between 1800 -2400 degrees.
In the exhaust stroke the burnt air -fuel mixture is forced outside by the piston through the exhaust
port.
Common rail
The common rail system was developed by a Swiss engineer called Robert Huber in the late 1960’s,
however the first commercial common rail system was introd uced by DENSO in 19952.
The advantages of the common rail system are the following. It has a high injection pressure up to
250 MPa. It is capable of pre -, main – and after injection and can it has the possibility of combining
them. The injection pressure i s being adjusted to the car’s operating conditions. The control of the
injection timing, the injection quantity and the injection pressure is very precise.
How the fuel does reach the combustion chamber in the common rail system?
The fuel is stored in th e fuel tank from where a delivery pump transports the fuel to a fuel filter,
where the fuel is being filtered. Then the electrical fuel pump delivers the fuel to the high pressure
fuel pump, where the pressure of the fuel is increased and reaches the commo n rail. The common
rail has sensors in both ends and it is distributing the fuel to the injectors. The ECU tells the
injectors when to inject the required amount of fuel to the combustion chamber.

2 https://en.wikipedia.org/wiki/Common_rail#History

9

The performance line
The performance line, where the testi ng of the injectors is done is situated in the production area.
From the newest generation injectors witch is called 4th Generation there are 5 performance lines in
the factory.
A performance line is consists of a combination of different machines. These machines are the
moulding machine, the connector air leak tester, the pallet assembly and oil filling up machine, the
running machine, three performance benches and the laser marking and pallet disassembly machine.
These machines are being operated by fou r people. One person is responsible for the moulding
machine, one for the air leak tester and the pallet assembly, one for the oil fill -up, running and the
performance benches and one for the laser marking and disassembly machine.
When an injector is OK i t goes to the final assembly line, where it goes through a 3D dimension
test, if needed a gasket is fitted to the injector. After that comes the oiling and the injector gets the
different kind of plugs, like plug for nozzle, plug for inlet pipe and plug fo r outlet pipe. An electric
inspection is the next step in the process. In these process the resistance, inductance and insulation
resistance. It is still possible that an injector fails in the electric inspection in this case this injector is
not going to the last step before packaging, which is auto appearance check.
The final assembly line has 3 operators, one from the production and two from the quality
department. One operator is for the customer dimension check machine and the gasket press fitting
mach ine, when a gasket is required to fit to the injector. One operator is for the visual inspection,
oiling the nozzle and to put on the caps. The last operator operates the electric inspection machine
and the camera, which visualises the injector and he is a lso responsible for the packaging.
The processes of the performance line
The injectors leave the injector assembly line in boxes and handed over to the performance
production department through windows, because in the injector assembly lines cleanness i s the
most important that’s why there are shelves between two windows. In the assembly’s side the boxes
are put in to the conveyor and when everything is packed they are calling the production to bring
them to the performa nce lines and the performance test of the injectors is being started .
First the injector goes to a conveyor until it reaches the moulding machine. The main outcome of
this process is to plastic moulding terminal after welding. This process is made by an operator. The
machine has two mouldi ng dies so it is faster than it would be only with one die. The plastic is
made of resins of two types. One kinds of resin is the main component of the plastic while the other

10

kind of resin is just for the colouring. The proportion of the two k inds is 50:1 , which means for 5
kilogram s the employee needs to add 100 grams of colouring resin. This mixture goes to a drying
machine first for 240 minutes on 80 degrees to dry it and get all the humidity out of the resins
particles.
The process of the usage of the moulding machine is the following:
 First the operator takes out one injector from the box and set it to the moulding die
 The machine automatically starts after the operator used the Nagara switch
 When the other die returned to home position and the green lamp is flashing, he sets the
next injector to the empty die
 If the previous injector moulding finished and the die returns to home position the operator
picks it up from the die, removes remained materials and put it to the forward box
The cycle time of t he moulding of one injector is 24 seconds. 20 seconds is the moulding process
and the other 4 seconds is the movement of the injector by the operator.

From the forward box the injector goes to the connector leak check machine. The operator picks up
an injector from the incoming box and puts it into the work holder, then closes the chamber and the
test start automatically. When the test is finished and the injector is OK the operator sends it to the
next process. If the injector is NG he puts it into an NG holder.
From the NG holder it goes to an oil leak tester, where it can be found out where is the leakage on
the injector. To find the source of the leak the injectors are connected to the machine with the outlet
pipe and then a pipe is connected to the inlet pipe. Then the operator turns the Nagara switch and
the injectors sink into the oil bath. If a particle is missed out from the injector in the assembly line
than the injector has inner leakage and the machine show it automa tically with a lighting lamp.
Beside the inner leakage the injector can have nozzle and nut leakage. When the leakage occurs on
the nozzle it can be detected by the massive amount of bubble from the tip of the nozzle up to the

11

surface of the oil. In the ot her case there are small bubbles coming out at the joint point of the nut
and the lower body.
The cycle time of this machine is 300 seconds. The operator does not have to watch throughout the
whole process. After the 300 seconds he needs to press a button for 20 seconds and during this
period he is examining the injectors and if he founds one injector which has see able leakage, he
presses the button of the injector.
After the injectors are tested in the oil leak tester they are re -adjusted and start the whole process
from the beginning.
The main point of the connector air leak check machine is to test the leakage of between the
moulding and the injector body based on pressure difference.
When the leak check machine is done with the measurement, the ope rator pulls a pallet to the
setting station, takes out the injector from the leak check machine and puts it to the ID auto jig. The
operator uses the Nagara switch and a camera checks the engraving of the body and compares it to
the data stored in the comp uter. The operator also compares the serial number and the customer
number on the monitor. When all the numbers are matching the injector is being put into the pallet.
The machine records the data to the pallet, while the operator assembles the inlet pipe, the work
clamps and outlet pipe of the pallet to the injector. After that with a torque wrench he tightens the
work clamp first and the inlet pipe after. When it is done he assembles the connector to the solenoid
and the outlet pipe to the injector. He fi xes the data written in the pallet by the Nagara switch, when
the OK lamp turns on he is forwarding it to the next work procedure.
The next step is the oil fill up machine. During this process oil is being injected to the injector
before the running machi ne to warm up a little bit.
The operator pushes the pallet into the oil fill up machine and closes the doors and with the Nagara
switch he turns on the machine. The pallet is being clamped into the machine and the air is being
vacuumed out of the injector and oil is being injected to the injector. This vacuuming and oil fill up
varies in each type of injector. When the required amount of vacuuming and oil fill up is done the
door opens automatically and the operator pulls it out and takes it to the next ma chine.
The next machine is the running machine where the injector is being heated to the required heat
level to be able to perform the best possible way in the performance bench.

12

The pallet is pushed into the running machine and the door is being closed by the operator. The
machine is starting after the operator uses the Nagara switch. The injector is in the running machine
until the camber reaches the temperature level needed to finish the process. In that case the machine
stops automatically, a green la mp turns on and the door opens. The operator pulls out the pallets and
forwards them to the next machine.
The next machine is the performance inspection bench. During this process the injection quantity of
the injector is being measured and compared to the calibrated values and a machine automatically
collects the values of the correction to the Data Collection Device.
The first process is when the operator pushes the pallet to the chamber of the machine, closes the
door and uses the Nagara switch and the machine starts automatically. The machine measures the
injection quantity and corrects it to the set value. The correction value is saved to the Data
Collection Device. A green lamp is on when the inspection is done and the door opens
automatically. The pa llet is taken out and being transferred to the next process. Throughout the
performance bench two kinds of NGs can be detected. These two kinds of NGs are High or Low
Qm NG.
The next step starts with the air blow machine, where the operator pushes the pall et into the
machine and closes the door, but before that in case of some injector types a QR plate is being
attached on the top of the solenoid connector. Right after that he uses the Nagara switch and the air
blow machine cleans the marking area on the to p of the solenoid connector, where in the next step
the QR code or Data matrix will be written with a laser beam depends on the type of the injector.
When the process is done the door opens automatically and the pallet is being pulled off by the
operator a nd pushes it to the laser marking machine, uses the Nagara switch and the machine starts.
In the laser marking machine first the machine makes the solenoid angle reading. When the angle is
correct it goes through the laser marking area, where the laser marking occurs and in the last step
there is a QR code or Datamatrix reading ability test. The door opens automatically when the
process is done; the operator pulls the pallet to the disassembly area.
There he checks the marked surface manually. If the marke d surface is not good for the standards at
the end of the disassembly process he puts it into an NG box.
If an injector was NG during the performance test it does not make it to the air blow machine,
instead it goes to the disassembly station right away. It is being disconnected from the pallet and the
injector is being put into a box based on if it is high or low NG and then it is going to be re -adjusted
in the assembly room.

13

The process of disassembling is the following:
 Loosen the inlet with torque wre nch
 Loosen the work clamp with torque wrench
 Disassemble the inlet pipe first and then the work clamp manually
 Take out the injector and compare the serial number on the injector with the one showed on
the monitor
Finally, when the injector is OK, he puts the injector in a green box where the OK injectors are put
and the injectors go to the next station via a conveyor to the final assembly line.
The first station of the final assembly line is the customer dimension check. In this process the
injector’s cri tical dimension are measured with function gauges.
The operator takes out one work piece from the box and tests the inlet thread with gauge, them
places the injector into the 3D check jig and checks the solenoid connector angle, the width of the
clamp fac e shoulder and the height of the injector. The NG identified injectors are put into the NG
chute with a pallet. The OK identified injectors get an O -ring on them and they are passed to the
gasket press fitting machine.
The purpose of the gasket press fitti ng machine is to press and fit seat nozzle on the nut sealing
surface.
The operator clamps the injector into the holder jig. Takes out a gasket from the holder and puts it
to the top of the nozzle and pushes it to the machine and the pressing start automa tically. When the
press fitting is done the jig comes out of the machine. The NG injectors are going down on the NG
chute with in a pallet. The OK identified injectors go to the next station.
During the next step a visual and electric inspection is being made by the operator. He picks up the
injector from the temporary holder and checks the critical items with magnifier lens. These critical
items are the following: inlet sealing surface and size, leaser markings on the solenoid connector,
solenoid connecto r and pins, inlet thread, engravings on the lower body, nut retaining and the gasket
and the last item is the nut nozzle.
NG identified injectors put into NG chute with a pallet. OK identified injectors go to the next step.
In that step the operator dump s it into an oiling jig which is filled with rustproof oil for the nozzl e.
After this process the injector gets the inlet cap, the outlet cap and the nozzle ca p as well from the

14

identified holder. Then places the injector into the resistance check machine, where resistance,
inductance and insulation resistance is being checked. This machine also starts when the operator
uses the Nagara switch. NG injectors go to the NG holder box via an NG chute in pallet. The OK
injector goes to the last operation before s hipping the product.
The last step is the auto appearance inspection, where a camera checks the final appearance of the
injector. The injector is being put into the camera jig from the resistance tester jig and uses the
Nagara switch to operate. If the inj ector is NG it slides down on the NG chute in a pallet and when
it is OK it goes to the package box.
The boxes are going to the warehouse, where a truck picks it up later and delivers it to the customer.

15

Analysis

In the analysis section I am going to describe how a process capability test is being made.
To be able to perform the process capability test, first all the program of the injector needs to be
installed on each machine. When all the required programs are installed the test can be p erformed.
The injector has tolerances and all the measured data needs to be within that tolerance. This test was
performed with the G -type injectors on Line 5.
The machines are arriving from Japan and usually the programs are installed to the machine. Mos t
of the time the program is the same, but there can be a difference between the Japanese serial
number and the Hungarian serial number for example: in Japan the injector’s serial number is 0001,
while in the Hungarian plant it is produced as 0031. In that case the process is to copy all the data
from the existing program and create a new one with the Hungarian serial number.
Process Capability Test of G -type injector
In case of the moulding machine the critical items are oil temperature, die working tempe rature, die
inner pressure, 3D measuring check and a BKD check.
First I checked the oil temperature. The temperature of t he oil needs to be between 75 -85 degrees
and it needs to be checked in every 10th minute 10 times.

From the diagram it can be s een all the measured data is around 80 degrees, which is within the
tolerance that’s why the oiling function works correctly.
During this I was able to check the inner temperature of t he die. The tolerance here is 75 ± 5
degrees. This measurement is carri ed out with a thermometer in every 10th minute 10 times. On the
diagram below, it can be seen that everything was in tolerance, so it works correctly.
Temperature
74,075,076,077,078,079,080,081,082,083,084,085,086,0
12345678910(℃)

16

Temperature
68,069,070,071,072,073,074,075,076,077,078,079,080,081,082,0
12345678910(℃)
To be able to measure the inner pressure of the die, the Marshal system needs to be connected to the
moulding machine. In this case the tolerance is 35± 3 degrees. It is measured 10 times when an
injector is in the machine. On the diagram below can be seen that all the data is within the tolerance.
Pressure
30,031,032,033,034,035,036,037,038,039,040,0
1 2 3 4 5 6 7 8 910(MPa)

For the 3D dimension test 2 injectors were moulded wi th the two difference dies. These two
injectors were taken to the 3D measuring laboratory. The injector has 176 measuring points, but all
of these points are checked only at the beginning of the production or when a die is chanced. On a
daily bases only 20 measuring points are measured.
Last check on the moulding machine is the BKD check. BKD is a Japanese abbreviation. The
purpose of this check is to visually check if there is any damage on the injector or not, 5 -5 injectors
are being checked from each die .
The next machine is the connector leak tester. The inspection pressure, the zero flow master, the
0,108 NG flow master and the BKD check are the key measurements.
The inspection pressure is 0.5 ± 0.1 MPa. During the measurements the inspection pressure was 0.5
MPa all the time as it can be seen on the diagram below.

17

Pressure
0,300,350,400,450,500,550,600,650,70
1 2 3 4 5 6 7 8 9 10(MPa)

The zero flow master is checked 10 times and the flow can be maximum 0.002 ml/min. As you can
see on the diagram and the picture the value was always 0,000 ml/min. The other flow m aster is the
0.108 ml/min where the tolerance is from 0.106 -0.110 ml/min. From the diagram you can see it was
0.110 ml/min in every case, which is in the tolerance.

The last check here is also the BKD check. 5 injectors were checked and all of them were without
any defects.
The next series of measurements were made on the pallet assembly and oil fill -up machine.
First I checked the serial number reading ability with the camera. During this test 10 injectors were
checked and there was no misjudgemen t. During this check the injector is in a holder jig, that’s why
it has to go through a BKD check to see if the jig made any defect on the injector. None of the
injectors had defects after this test.
Flow measuring
-0,001 – 0,001 0,002 0,003 0,004 0,005
12345678910(ml/min)
Flow measuring
0,1050,1060,1070,1080,1090,1100,111
1 2 3 4 5 6 7 8 9 10(ml/min)

18

In the next test I put a G -type injector to the holder j ig, but changed the program of the camera to all
the other kinds of injector programs. The camera was able to detect that the injector in the jig is not
the correct one.
In the next step I check the torque wrench for the work clamp with a digital torque c hecker. The
tolerance here is 24 -26 Nm. As the diagram shows below, the measurements were always within the
standard.
Torque wrench [Work clamp]
2323,52424,52525,52626,527
1 2 3 4 5 6 7 8 9 10SampleTorque [Nm]

The torque wrench for the inlet pipe was measured next. The tolerance in this case is the same as for
the work clamp, which is 25 ± 1 Nm.
Torque wrench [Inlet pipe]
2323,52424,52525,52626,527
1 2 3 4 5 6 7 8 9 10SampleTorque [Nm]

As you can see all the data is in the tolerance that is the torque wrenches are working correctly in
both cases.
The degassing pattern was the next checkable data on the check list. The degassing pattern of this
injector can be seen below.

19

From the diagram of the degassing pattern you can see that the degassing lasts for 15 seconds and
the oil filling up for 8 seconds. This is one cycle. This cycle repeats 4 times in case of this kind of
injector.
BKD check is the last measurable detail here as well. 3 injectors needed to be check and all of them
were without any visible defects.
The next machine was the running machine, where there were 8 measurable items. The first item
was the motor revolution accuracy. The motor revolution needs to be 2000 rpm and the tolerance is
±2 revolutions per minute. The running machine has 4 stations and there is one common rail system
for two cells. The 1st station and the 2nd station counts one, and the 3rd and 4th cell counts one as
well. Both two parts were chec ked 5 times and the revolution of the motor was always 2000
revolutions per minute.
The temperature of the oil was the next item. The temperature need to be 60 degrees and the
tolerance is ±5 degrees. It needed to be checked 10 times in every 10th minute as you can see below
on the diagram. As you can see the measured value was always between 62 and 63 degrees. 15 secs 15 secs 15 secs 15 secs 8 secs 8 secs 8 secs 8 secs

20

The next item was the common rail pressure indication precision. In this case I needed to check the
difference written on the screen of the machine and the pre -set value. The difference between the
set value and the measured value cannot be more the 2 MPa. From the table below you can see that
the difference was ranging from -0.2-0.8 MPa.
1 / 2 ST
Monitor Measured IX-YI ≦ 2.0 Mpa Judge
19MPa 19 19,2 -0,2 O
35MPa 35,1 34,8 0,3 O
160MPa 160 159,2 0,8 O

The difference between the Tq command value and the measured value was the next check. It
needed to be check in all stations with the same injector. The difference cannot be more than 0.01
milliseconds. As the chart shows below there was no problem with it.
G-TYPE
Set Measured IX-YI ≦ 0,01 msec Judge
1 ST 1,7 1,70255 0,00255 O
2 ST 1,7 1,700913 0,000913 O
3 ST 1,7 1,700928 0,000928 O
4 ST 1,7 1,700018 0,000018 O

The injection quantity measurement precision was the next test. This measurement is carried out on
160 MPa for 1700 mi lliseconds. The standard injection quantity is 135 ± 10 mm3. It was tested with
one injector 5 times in all the 4 stations as you can see on the diagram below. The stations marked
with different symbols.

21

110115120125130135140145150
1 2 3 4 5Injection qty(mm3/st)1ST 2ST 3ST 4ST
The tolerance for the leak back pressure is 6 ± 5 kPa. This value can be read from a meter on the
back of the machine.

From this diagram it can be seen that the leak back pressure was 1 kPa an d 1.5 kPa during the
measuring.
The most difficult measurement was the EDU drive current value check me asuring. For this
measurement I used an oscilloscope. The EDU has 3 values to be measured and the tolerance for
these values is ±0.8 Amperes. These set values are 21.2 Amperes, 7.5 Amperes and 5.5 Amperes as
in the picture below.

-11357911
1 2 3 4 STATIONLeak back pressure (kpa)1ST 2ST 3ST 4ST

22

First I connected the oscilloscope to the running machine with a wire. Then I connected the
oscilloscope to the wiring of pallet with a clamp and pushed the pallet inside the machine. I set the
oscilloscope to be able to see the required data and to save the picture of the osci lloscope, than I
started the machine with the Nagara switch. During the measuring the oscilloscope automatically
showed me the graph I wanted to see, what you can see below.
G-TYPE
Spec. Measured IX-YI ≦ 0,8 A Judgement
1 ST 21,2 21,05 0,15 O
2 ST 21,2 21 0,2 O
3 ST 21,2 21,1 0,1 O
4 ST 21,2 21,1 0,1 O
1 ST 7,5 7,65 -0,15 O
2 ST 7,5 7,55 -0,05 O
3 ST 7,5 7,55 -0,05 O
4 ST 7,5 7,55 -0,05 O
1 ST 5,5 5,85 -0,35 O
2 ST 5,5 5,95 -0,45 O
3 ST 5,5 5,8 -0,3 O
4 ST 5,5 5,8 -0,3 O

23

24

From the t able you can the values were within the tolerance in all three cases, so the EDU drive
current is good in the running machine.
The last checking process was the BKD check here as well. Three different injectors were checked
and all of them as without def ects.
The last machine on the performance line is the Leaser marking and pallet removal machine. This
machine has 5 items to be checked. These are the marking position, marking character, data matrix
readability, data matrix readability decision function a nd BKD check.
First I checked marking position. In the beginning we used a special paper, which was attached to
the solenoid connector with a dual layer glue stripe. When these test showed that the position and
the laser marking character are also good, then it was possible to do the laser marking to the
solenoid connector.

25

The marking position was check 10 times on 10 injectors and all the positions were on the correct
spot.
Then the marking characters were checked 10 times on 10 different injectors. The characters were
the same as the drawing content, so they were OK.
The next test was the Data matrix readability decision function. For this test three injectors were
used. There were two OK injectors and one NG injector, on which the Data matrix was r uined a
little bit, so the machine was not able to ready the required amount of data out of it. The readability
needs to be at least 95 per cent to be able to read out the required data from the data matrix. Below
95 per cent the injector become an NG inje ctor.
From the table below you can see that injector number 1 and 2 are OK injectors and injector number
3 is an NG injector.
G-type
INJ 1 INJ2 INJ3
1 OK OK NG
2 OK OK NG
3 OK OK NG
4 OK OK NG
5 OK OK NG
6 OK OK NG
7 OK OK NG
8 OK OK NG
9 OK OK NG
10 OK OK NG
Judge O O O

The last check is the BKD check. Three different injectors were checked for defects, but none was
found with a defect.
After I finished with the process capability check of the performance line I moved to the final
assembly line. There all the four machines were tested as well.
The first station here is the customer dimension check station. In this station there is a jig for the
check. n
r

26

On this station first I needed to check all the masters in the jig. There are seven NG masters. The
NG masters have problems with width, height, clamp and nut. When the three main dimensions are
being checked, which are the solenoid connector angle, the width of the clamp face shoulder and the
height of the injector these masters are not able to go through the checking process. All the NG
masters were tested 10 times. None of them passed this check, so the measuring jig works correctly.
After the NG masters the next step was to test the jig with an OK master. The OK master was an
injec tor produced during production. This injector was tested 10 times and all the dimensions were
OK all the time.
In the next measurement one of the NG injectors were tested again 10 times and it was NG 10
times.
In the next process the function of the NG ch ute was tested. When one of the dimensions is not
good the injector is being put to a pallet and slides down on the chute. It was also tested 10 times
and the function of the NG chute was OK.
The thread of the injector is also tested with a gauge. This gau ge is placed on a sensor and the OK
lamp does not turn on until the gauge was lifted up from the sensor. During the thread poke -yoke
function test, it is being test if it is possible to miss out this step. I tried 10 times, but it was not
possible the chea t the system.
The last test here like in all other tests is the BKD check. 5 injectors were tested after being in the
3D dimension jig and there was no damage on the injectors.
On the gasket press fitting machine the first test reliability of load measure ment. This test is to
check the difference between the set value and the actual value of the pressure during the gasket
fitting. The set value is 2.75 K N. I measured 2.85 KN which was within the ±3 per cent.
The next test was the press fitting load check. Gasket was fitted 10 tim es to the injector and the load
was measured.

27

From the graph it can be seen that the load of the press fitting was 2.85 KN in all the 10
measurements, which is a constant value so the machine works as it should be.
The next test was made by the measuring room. During this test they used a special tool to perform
the pilling out load measurement. In this test with the help of that tool the gasket is removed from
the nozzle. The value of the required pressure needs to be more than 1 00 N.

The graph shows that the pilling out load was always abo ve 100 N during the 10 measurement , so
the machine presses the gasket to the injector with a good pressure.
When the gasket is not pressed with a good value or the gasket was not put perfect ly, the gasket is
deformed or the gasket was not put on, the NG lamps turns on. In that case the injector is being put
into a pallet and goes down on an NG chute.
The function of this chute was tested 10 times and it was working without a problem.
The BK D check showed no damages on the injectors in all 5 cases.
The next machine is the electric inspection machine, where 4 injectors are tested 10 times. The first
inspection is the resistance check. The injector is placed in the jig and with the Nagara switc h the
machine starts. The test was performed 10 times continuously.
Resistance
500,000520,000540,000560,000580,000600,000620,000640,000660,000680,000700,000
1 2 3 4 5 6 7 8 9 10(KOhm)

28

From the graph it can be seen that the result were close to each other in case of the four injectors.
The same process was carried out with the insulation resistance and the inductance as well.
Inductance
270,00290,00310,00330,00350,00370,00390,00410,00430,00450,00470,00
1 2 3 4 5 6 7 8 9 10(uH)

In case of the inductance there was a slight difference between the highest and lowest measurement,
but all the injectors had similar values during the 10 measurements.
Insulation
0,0002000,0004000,0006000,0008000,00010000,00012000,00014000,00016000,00018000,00020000,000
1 2 3 4 5 6 7 8 9 10(MOhm)

This graph shows the results of the insulation resistance test. 3 form 4 i njectors had very similar
values, but the 4th injector there were two lower measurements. In case of the insulation resistance it
is not a problem because the tolerance says it needs to be more than 20Mohm.
The next step was the function of the oiling jig . The purpose of this test was to find out if it is
possible to cheat the system and leave out the oiling of the injector. It was tested 10 times but it was
unable to miss the oiling function, so the sensors are working correctly.

29

The poke -yoke functio n of the plugs was tested next. As you can see in the picture there are 3 kinds
of plugs. It has an order in which the operator needs to attach them. This test is to find out if it is
possible to cheat the system by changing the order of the plugs. This wa s tested 10 times as well
and was working as it should be.
When the injector fails on the electric inspection the NG lamp turns on and the injector goes down
on an NG chute in a pallet. This test was carried out 10 times and the function the NG chute was
working.
Finally a BKD visual check was performed on 5 injectors and there were no defects at all.
The final machine in the process is the auto appearance check machine. The first test is the camera
judgement function check. For this test 5 injectors were u sed. 4 of them were NG s and 1 OK. The
NG injectors were missing one particle, like inlet plug, outlet plug, nozzle plug and the gasket. The
OK injector was attached with all the plugs and the gasket. Each master was tested 5 times.
Master 1 Master 2 Master 3 Master 4 OK Master
NG NG NG NG OK
NG NG NG NG OK
NG NG NG NG OK
NG NG NG NG OK
NG NG NG NG OK

From the table it can be seen that all the NG masters injector was marked NG by the camera and
only the OK master was marked OK.
The next step was the serial number reading function check. It was tested with 10 injectors 5 times
with each. All the time the reading ability was OK so the camera works as it should be.
The next was the data matrix reading ability. The grade of it should be A or B. This is te sted with
one injector 10 times. The results were Bs all the time, which is good according to the tolerance.
Camera grade level checking was tested with two master injectors. A and B ranks considered OK
and C and D ranks considered NG. There was one OK and one NG master. The grade of the OK
injector was A, and the NG was C. The test was repeated 10 times and A was A 10 times, while C
was C 10 times as well.
The final test of the whole process is a last BKD visual check. It was performed 5 times and there
were no defects on the injectors.

30

Comparison test of M-type injector
Previously we made the process capability test of the M-type injector on Line 4 and later we
performed it on Line 5 as well. The processes in both cases were the same as the one des cribed
previously with the G -type injector . In case of the performance line the results were similar to each
other and all the values were within the tolerance.
In the following I am going to compare the results of the performance benches, because that is the
most important in the comparison. Totally there were 200 OK measurements. There were 100
injectors on the 4th line and exactly the same 100 injector on the 5th line.
On the diagrams below one dot shows one injector and the average of 50 injections. Co lumn
number 1 indicates the results of the 4th line and column number three shows the values from the 5th
line.
The first diagram shows the values without correction, the second one shows the correction value
and the main point is the third one, which is the result after the correction. One injector can be tried
to be corrected maximum three times. If it is not in the tolerance after the third correction it will be
an NG injector.
The M -type injector has 15 measuring points; some of the points have the sam e pressure value, but
with different opening times measured in microseconds.

31

From the diagrams above we can see that the not all the injectors are within the tolerance, that’s
why the correction values are being used. These correction values a re stored in the chip of the pallet
and in the laser marking machine, these data are being written to the solenoid connector as a QR
code with a laser beam.
From the third column it can be seen that all the values are within tolerance and they are having
very similar shapes, which means that the same injectors had almost the same correction values.

32

The above showed measuring points are tested on 120 M Pa pressure. In point 6 it can be seen that
before the correction one is slightly above the rest and one is lower than most of the injectors, but
after the correction all the injectors are within tolerance. The injectors had the same results after the
correction, because the outlines of the two kinds of diagrams are almost the same.
In the other cases th e results are similar everywhere and after the correction they are between the
upper and the lower tolerance.

In both cases the above showed diagrams there are a few injectors, which is not close to the other
injectors, but thanks to the correction p rocess the after it the results are the same in both lines.

33

The last measuring point is the 12th one. From the graph it can be seen that after the correction all
the injectors are within the tolerance and the two graphs have the same profile.
The injecto r also has 3 check points shown below. In the first case there is only a lower tolerance
and both of the injectors are above it and the figur e of their shape is almost the same except one
injector on the 5th line’s case.
In the 14th check point there is onl y an upper tolerance. All the results are below the tolerance or on
the edge of the tolerance, so it is ok.
The last check point has only an upper limit as well, but all the data is 0, which is under the
tolerance.

34

The difference between the two l ines in case of the Tis value cannot be more than 0.01
microseconds.
The Tis is the difference between the drive current and the delay of the starting of the injection as
the picture shows it below.

From the graph below it can be seen that the average o f the opening delay is exactly the same, so
there is no difference between the two lines.

The leakage in case of the M -type injector can be maximum 4 mm3. From the graph below it can be
seen that it never reaches the 4 mm3, so there is not much retur ning oil from the injector.

The kisadory activity is also calculated based on these kinds of histograms.

35

In case of the final assembly line most important comparison able data are the resistance and the
inductance during the electric inspection and the decision making ability of the camera. The tests
were carried out with the same 10 injectors in all the stations.
The electric inspection was done in two ways. Once it was done with 11 different injectors in both
lines and after that on both lines with 1 injector 11 times.
The difference between the two measurements in case of the resistance cannot be more than 4
percent, while the inductance cannot be more than 1 percent.
510520530540550560570580
03T76962 03T76959 03T77007 03T76966 03T77016 03T76965 03T77009 03T77012 03T77014 03T76963 03T76964L4
L5

The graph shows the results resistance test of the 11 different injectors. The dif ference between the
two values is ranging 2.8 -3.4 percent, which is in the tolerance so it is OK.
350360370380390400410420
03T76962 03T76959 03T77007 03T76966 03T77016 03T76965 03T77009 03T77012 03T77014 03T76963 03T76964L4
L5

36

This graph shows the results inductance test of the 11 different injectors. The difference between
the two values is always less than 1 percent, so it is wo rking right.
525530535540545550555
03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964L4
L5

On the graph the results of the resistance test can be seen. All the differences are around 3 percent,
which is less than the 4 percent tolerance. That means that the machine can work fine.
The results of the inductance test can be seen bel ow on the graph. It shows the same injector tested
11 times. The difference between the two measured data is 0.8 -0.9 percent. It is less than the
tolerance, which is 1 percent, so it is OK.
380382384386388390392394396398400
03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964 03T76964L4
L5

37

The camera’s decision making function was tested in 4 different ways repeated 10 times . First all
the plugs were attached to the injector, in the second case they were without nozzle ca p, in the third
test they were without inlet plug and in the last they were not equipped with the outlet plug.
S/N Camera function S/N without inlet cap
L4 L5 L4 L5
03T00000 OK OK 03T00000 OK OK
03T00001 OK OK 03T00001 OK OK
03T00002 OK OK 03T00002 OK OK
03T00003 OK OK 03T00003 OK OK
03T00004 OK OK 03T00004 OK OK
03T00005 OK OK 03T00005 OK OK
03T00006 OK OK 03T00006 OK OK
03T00007 OK OK 03T00007 OK OK
03T00008 OK OK 03T00008 OK OK
03T00009 OK OK 03T00009 OK OK
03T00010 OK OK 03T00010 OK OK

S/N without nozzle
cap S/N without ou tlet cap
L4 L5 L4 L5
03T00000 OK OK 03T00000 OK OK
03T00001 OK OK 03T00001 OK OK
03T00002 OK OK 03T00002 OK OK
03T00003 OK OK 03T00003 OK OK
03T00004 OK OK 03T00004 OK OK
03T00005 OK OK 03T00005 OK OK
03T00006 OK OK 03T00006 OK OK
03T00007 OK OK 03T00007 OK OK
03T00008 OK OK 03T00008 OK OK
03T00009 OK OK 03T00009 OK OK
03T00010 OK OK 03T00010 OK OK

From the table we can see that the camera’s decision making function is working correctly, because
it detected the injector OK when all the plugs were attached and d etected them NG when one of the
plugs were not put on.

38

Production bench levelling to quality
Within the company this is called the kisadory activity , which means the correc tion of the
performance benches to the quality benches.
The performance benches ar e sending the information to the data collecting device, where all the
data from each station are stored and TST & PE gets the required information from the data
collecting device. At the TST the responsible person checks the results and calculates the cha nges to
be made. He writes the results of the calculation to a printed paper with the histograms, then an
engineer in the main office sets the values to an excel table and prints it out. When he is done with
it, he goes to the production area are where he finds the correct host PC. On the host PC, he opens
an excel table to write in the calculated data. Every station has an own excel table so there are 6
tables in total to be changed. When he is done with the correction of the data he saves it and sends
the infor mation to each bench station via the company’s network. The transferring of the
information is not recommended until the performance bench is running.
Control items
There are items, which needs to be controlled. For the controlling of these items there are
measurements to carry out regularly. These measurements are called cleanness measurements.
From the moulding machine an injector needs to be hand in to the measuring laboratory to check
the injector’s critical dimensions. These critical dimensions are the HP connection position , the
clamping position , the u pper body position , the n ozzle tip high , the HP connector position and the
inlet diameter position . These items are being checked once a day, but from the machine there are
two injectors put away for the measurement. One is being brought to the measuring laboratory and
the other is the back -up injector if the first fails on the test.
The connector leak tester needs mastering with the zero flow master and the NG flow masters in the
beginning of every s hift.
The oil from the oil fill -up machine needs to go through cleanliness measurements. The number of
the particles is checked once a week, while the weight and the size of the particles are checked once
a month.
The oil from the running machine also ne eds to be checked for the number of the particles once in
every two weeks.

39

From the performance benches injectors need to be delivered to the quality measuring laboratory.
From every cell of the performance bench three injectors needed. In total it means 18 injectors,
because there are 6 cells all together in the 3 performance benches.
The cleanliness check of the inspection oil is checked once a month. The final product is also
inspected in the cleanliness laboratory. The allowable particles, the size an d the weight of the
injector are being checked.
Management structure
The company is a Japanese company, that’s why the management structure is very hierarchical.
The hierarchy can be divided to two parts. One is the indirect departments and the other is the
production.
In the production the lowest employee level is the Associate 1, the next in the row is the Associate
2. These two levels are the operators.
The next level is the Set Up; they are mainly responsible for the running of 1 or 2 assembly lines .
The Team Leaders are the next in the hierarchy. They are responsible for one type of product like
the 4th Generation injectors.
The highest level in the production is the Supervisor . The supervisor is responsible for the schedule
of the production. He d ecides what type and how many injectors need to be produced based on the
buyer’s needs.
From the indirect productions side the lowest level is the Associate 2 level. He does not have a
higher education or a diploma. He is mainly working an assistant.
The Specialist 2 is the next rank. He is an experienced engineer, able to work alone without
guidance and able to run projects.
The Expert 1 is the next level. An expert 1 colleague is an expert in one kind of technology or in a
few technologies. He is able to work with one kind of machine, for example a moulding machine in
all areas of the factory and does not have any employees.
The Section Leader is also an expert in one or a few kind of technologies. He is able to show the
way to his employees to be abl e to run an assembly line or a whole generation of assembly lines.
On the next rank the Managers and Coordinators can be found.

40

A manager is giving guidance to at least two areas of the factory, for example the diesel injector
performance and the diesel injector assembly.
A coordinator is a Japanese expert. They are coming to Hungary for a 3 -5 years period. During this
period they are coordinating projects.
Senior Managers are the next. A senior manager is the boss of the manager or we can say that he is
the manager of the managers. He is giving assistance to a lot of departments of the factory to be
able to achieve the goals set by the top management.
In the next level there are the general managers , directors, executive coordinators .
An executive coord inator is a Japanese professional controller. He does not have any employees,
but he is guiding and giving advices about the actual ongoing projects.
There are two general managers. One is responsible for the administration and the other one is in
charge for the technical departments. They are guiding there departments to the way set by the top
management.
The director is the boss of the general manager. He is also giving guidance to the general managers
to achieve the goals.
The next person is the execut ive vice president. He can decide in smaller subject areas and he is in
charge when the president is not present.
On the top of the hierarchy there is the president of the company. He has the final decision in all the
bigger decisions, which can affect the future of the factory.
Within the company the hierarchy can be seen by the colour of the stripes on the caps. The wearing
of the cap in the production area is obligatory to everyone.
The workers with less than 1 year experience have a red strip on their caps. The yellow stripe shows
that they are Set Up or Specialist 1 associates. The Team Leaders and Specialist 2 employees have
green stripe on their caps. A Supervisor or an Expert 2 colleague has a blue stripe on his or her cap.
The purple stripe shows t hat the person who is wearing it is an interpreter. From Section leader and
above the cap has a special logo on it.
All together there are 159 people from the Section Leader level and above. 38 people form the 159
is Japanese.

41

The top management consists of 12 people and only 3 people are Hungarian and the rest 9 people
are Japanese. They are setting the goals of the company for the next financial year, planning the
upgrades of the company. They are leading the company and they are the one, who decides in the
critical questions.
From the Senior Manager level this trends is changing and there are more Hungarians than Japanese
leaders. There are 2 Japanese and 6 Hungarian Senior Managers.
There are only 4 Japanese between the Managers and 29 Hungarians.
There are 13 Japanese coordinators and 10 Japanese trainees.
The section leaders are only Hungarians and there are 83 of them.
SWOT analysis

The strengths of the company are the newest technologies from Japan, which results in a constant
development. The Japanese top management guides the company to the goal as it is happening in
Japan. The company has long -term financial and production plans thanks to the loyal buyers, whom
are focusing on the company as main supplier.
One of the main weaknesses of the c ompany is the limited production area. The company needs to
follow Japan in every aspect and if they want to change something in a production method Japan as
the last word that is the right of veto. The work force is less and less in the area, that’s why t he
company does not have as much operators as it would be ideal.
However there are opportunities as well. The company is able to produce newer generations of
older products or a totally new product. The human resource department is recruiting operator form
Eastern Hungary and Transylvania to solve the workforce problem.

42

The competitor’s new products are threads for the company. There is a competition for the left
workforce with the surrounding companies. If the suppliers are late and the raw materials are n ot in
the factory in time the whole production is in danger thanks to the just -in-time system.
Conclusion

There are a few factors which makes it possible to be able to deliver the same quality to the
customers all the time even if the production is hap pening on different assembly lines.
Firstly, the programs of the different injectors need to be installed correctly and the setup data needs
to be the same in every machine on every performance line for the same kind of injector.
The regular correction of the performance benches based on the histograms is another key factor to
have the same quality. These correction calculations made by the Technical Support Team and the
changes are being made by the Production Engineering.
The daily cleanliness test is a lso important to check if there is any defect or contamination in the
injector. The quality checks of the injectors are also a factor to see if the performance benches are
working as they should.
The whole energy investment of the performance engineering and the workers of the logistics
department do not worth anything, if the operators are not doing their job, as it is described in this
project earlier.
From this it can be seen that the most important factor is the workers, whom working on the
production lines.
Putting into perspective

The injector assembly lines and the performance lines are further from each other as it should be. If
someone is thinking in a LEAN way, every kind of transportation is a waste of time. If the two lines
were closer to each other it would be better, but the best scenario is when the two lines are next to
each other. In that case there is no transportation, so there is no waste of time during the production.
As I see it, the easier way would be to move the performance l ines next to the assembly area, but
next to the assembly area there are the pump performance lines. To achieve this, the whole
production area should go through a complete reformation. During this transformation all the
machinery should have a better place ment. When this change should happen, there would not be

43

any production at all, which results in no income for a few months, but the employees needs to be
paid during this time as well.
In this situation the top management would decide if it is beneficial or not for the company to do it.
They need to consider if the loss in short term due to the shutting down of the production will be an
advantage in the future or not.

44

Bibliography
http://www.larapedia.com/glossary_of_lean_six_sigma_terms/nagara_switch_nagara_suicchi_mean
ing_and_definition_in_lean_six_sigma_terminology.h tml
https://en.wikipedia.org/wiki/Common_rail
https://en.wikipedia.org/wiki/Diesel_engine
Diesel English.ppt made by Mercant ec
Automotive Technology – Principles, Diagnosis and Service 4th edition – James D. Halderman
(Pearson, 2012) BBS

AutomotiveTech – A systems approach 5th edition – Jack Erjavec (Delmar Cengage Learining,
2009)

http://www.free -power -point -templates.com/articles/how -to-create -a-swot -analysis/

45