Académique Documents
Professionnel Documents
Culture Documents
Plastics
Business Group
Increasing Productivity through Process Optimization
Contents Page
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
Fault diagnosis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
2
Increasing Productivity through Process Optimization
!
se
for plastics processors to adopt appropriate
rea
procedures to analyze and optimize the pro- inc
cessing operation and the factors that influ-
red
ence it [1].
uce
!
The development of improved, automated
quality assurance systems has brought ad-
vances in process monitoring for injection
molding machines, thanks to new open-loop Fig. 1: Increasing productivity
and closed-loop control facilities. This has
been made possible through rapid advances
in microelectronics and sensor technology. Figure 2 shows an example of a clear in- In order to achieve cost-optimized produc-
Today therefore, it is feasible to achieve crease in productivity. Continual improve- tion with a high level of productivity (cost-
comprehensive (100 %) monitoring of the ments to the different segments of the efficient production), the next step involves
injection molding process on the basis of process chain have made it possible to re- optimizing the processing operation from
target/actual comparisons. In order to keep duce the cycle time for CD production from the angle of quality and reliability. Process
the amount of data involved to a manage- more than 20 s at the start of the 1980s to optimization generally forms part of the
able level, however, it is wise to restrict this approximately 3 s today. The clear reduc- start-up process for a new mold.
monitoring to a number of process parame- tion in the price of CDs since then has been
ters that affect quality [2]. due not least to this development.
An increase in productivity can be achieved
by increasing the output-to-input ratio. If it 25
proves possible to produce a greater num-
22
ber of units (the output) while simultane-
20
ously lowering the cost of molded-part pro- Share of machine and mold technology
duction (the input), then this will have ful-
Cycle time (s)
15
∆ tz = 19 s
Share of material
5 Cycle time
3
0
1982 1984 1986 1988 1990 1992 1994 1996 1998 2000
Year
Fig. 2: Example of CD production
3
Increasing Productivity through Process Optimization
Molded part
4
Increasing Productivity through Process Optimization
In order to achieve cost-efficient produc- The correlation between the point of origin These individual points will now be looked
tion, it is necessary to take a close look at of a fault and the impact of this fault is also at, taking a number of practical examples.
each individual link in this process chain described by the 'rule of ten'. This states, on
and then to optimize it, making due al- the one hand, that the outlay required to rec-
lowance for the interaction between the dif- tify a fault increases tenfold with each step
ferent parameters. Figure 4 shows just how from the idea for the part through to its de-
important it is to observe and optimize each ployment and, on the other hand, that the
individual segment at the earliest possible potential for influencing the production costs
stage. falls to a level of 10 %. This shows just how
important it is to clarify in good time the in-
The majority of faults have their origins in
fluence that the individual segments of the
the development phase, but these faults are
process chain have on quality and produc-
generally not rectified until the parts are in-
tivity.
spected after production, since it is only then
that the faults are discovered.
40 Origin of fault
Frequency (%)
30
20
10
0
Definition Development Operational Production Inspection Deployment
planning
Fig. 4: Origin and rectification of faults [3]
5
Increasing Productivity through Process Optimization
Standard PA
Durethan® BKV 30 H2.0
Shear
Adjusting this product to give it more pro- Influence of the molded part The target set by the customer for achiev-
nounced non-Newtonian characteristics led geometry ing more economical production was a wall
to a broad processing window, less flash at thickness of between 3.0 and 3.5 mm and
the weld and a reduced tendency to adhere The design of the molded part marks a fur- hence a weight of approximately 1.6 kg. The
to the heating element during heated tool ther link in the overall process chain. Para- cycle time was not to exceed 80 s. Working
welding [4]. The bursting pressure was in- meters such as wall thickness, rib-to-wall- on the basis of rheological and mechanical
creased by some 40 % in the course of ma- thickness ratio, positioning of ribs, bosses calculations, a design suitably tailored to the
terial optimization, and it also proved pos- and perforations, surface requirements and stress pattern was established, which not on-
sible to further extend the processing win- permitted tolerances play a part in deter- ly permitted the target to be attained but al-
dow in response to frequent requests from mining the requisite component quality and so led to further improvements. The opti-
welding engineers. In addition, it was only hence the productivity. Figure 7 shows an mized wall thickness was ultimately 2.5 –
by virtue of this optimization that produc- example of how productivity can be clearly 3 mm, bringing the weight down to only
tion by the two-shell overmolding technique increased through the selective optimization 1.45 kg, while the cycle time was reduced
was possible at all. Further success factors of wall thickness. to 65 – 70 s. The additional 10 % reduction
included improved potential for rheological in weight thus produced significant materi-
and mechanical computation. In its original wall thickness of 4 mm, the al savings, and the reduction in cycle time
lawnmower chassis shown below weighed led to a further increase in productivity.
1.83 kg and required a cycle time of 95 s.
7
Increasing Productivity through Process Optimization
Fig. 9:
Door liner with Contura Mold optimization –
cooling
8
Increasing Productivity through Process Optimization
9
Increasing Productivity through Process Optimization
The transverse bar fitted just upstream of To ensure that complex cavities can be uni- and positioned (Fig. 11). The outlay on sub-
the gate causes the melt flow to divide into formly filled via a number of different gates, sequent alterations to the mold will be kept
three part-flows at the start of the injection it is important for a rheological calculation to a minimum, series production can com-
phase. This then reduces the speed at the to be performed at the mold design stage. mence sooner, and component properties
gate. Once the transverse bar has been com- This will enable errors and problems to be will be improved and the cycle time reduced.
pletely filled, the compound continues to pinpointed right at the planning stage, and
flow at the higher injection speed set on the permit the gates to be optimally designed
machine. The lower speed of the melt at the
start of the injection phase serves to reduce
the shear and the temperature increase, and
the thicker layer of melt then freezes on the
mold wall and can no longer be torn apart
by the melt flowing in afterwards. A flaw-
less, glossy surface results.
Influence of the injection molding If the metering stroke is below this range the metering stroke is above this range
machine (<1D), the residence time will rise to inad- (>3D), then there is a danger that air will be
missibly high levels and there will be a con- introduced together with the molding com-
The injection molding machine must be ad- siderably increased danger of silver streaks pound, which will once again lead to streaks
equately dimensioned for the requirements. due to decomposition. The reproducibility on the surface of the molded part. The opti-
The chief selection criteria are: of the production process will also suffer on mum screw diameter for a specific part weight
account of the short screw displacement. If can be estimated with the aid of Fig. 13.
● clamping force
● metering volume
● permitted mold size
1D 2D 3D 4D
m rang
Optimu
100
10
1
10 20 40 60 80 100 200 mm
Screw diameter
Influence of processing These parameters must therefore be estab- process (Fig. 14). It is of decisive impor-
lished during the initial processing trials that tance to achieve the correct machine setting
Once all the requirements have been ful- are conducted. and to ensure that this setting is reproducible
filled in respect of the optimum material, [7]. The machine settings for the injection
Apart from determining which process pa-
the molded part design, the mold and ma- and compression phase have a very clear in-
rameters affect quality, it is also necessary to
chine, this still leaves the processing opera- fluence on viscosity, molecular weight re-
know how these parameters affect individ-
tion to be optimized in a bid to ensure cost- duction, crystallinity and state of orienta-
ual quality characteristics, such as appear-
efficient production. The first step here is to tion, and this essentially affects the me-
ance, mechanical properties, and dimen-
establish which process parameters have a chanical and physical properties of the in-
sional and shape accuracy. The only para-
bearing on quality. The processing parame- jection molded part. The surface finish,
meters that have a direct impact on molded
ters with the greatest influence on the prop- shaping and weight are also clearly influ-
part quality during production are the pres-
erties of the molded part and the production enced during this phase, as is flash forma-
sure level, the pressure profile, the temper-
costs are the mold and molding compound tion. The holding pressure phase essential-
atures, and the flow front velocity inside the
temperature, the injection rate or injection ly affects crystallinity, orientation inside the
mold. The remaining parameters have only
profile, the point of switchover to holding molded part, shrinkage and hence dimen-
an indirect influence. It is possible to at-
pressure, the level and duration of holding sional stability, weight, and also voids, sink
tribute key quality characteristics to the in-
pressure, the cooling time and the residence marks and demolding behavior.
dividual phases of the injection molding
time of the melt at high temperatures. The
process parameters that are of relevance will
depend on the geometry of the molded part
and the requirements placed on this part.
t0 t1 t2 t3
Injection phase Holding pressure phase
Packing phase
Time
13
Increasing Productivity through Process Optimization
tN Recommended
Cavity pressure
Time
Hydraulic pressure
Fig. 17: Sealing point and the set holding pressure time
14
Increasing Productivity through Process Optimization
The melt and mold temperature affect the Figure 18 shows the pronounced differences file then adopts a saw-tooth profile due to
quality of the molded part, the amount of that can emerge here. While the inflow tem- the heating during the injection phase and
waste that results, the cooling time and perature can be kept constant over the en- the cooling during the holding pressure and
hence the productivity. The melt tempera- tire processing operation, the actual mold cooling phase.
ture should be established from the pro- wall temperature increases as production
cessing data provided by the raw materials progresses on account of the hot melt, and
producer and kept as constant as possible settles at a high level. The temperature pro-
during processing, as well as during opti-
mization. The mold temperature has a con-
siderable influence on both the attainable
cycle time and the quality of the molded
part. A higher mold temperature will gen- 70
erally be required in cases where stringent 1
requirements are to be met in terms of the
surface finish, the inherent stress level, and 2
Temperature (°C)
60
the dimensional accuracy and stability of
the molded part. It should be borne in mind
that all material-specific figures and rec- 3
ommendations given for the cavity temper- 50
ature relate to the cavity wall temperature 4
and not to the melt inflow temperature or
the temperature set on the cooling unit.
40
0 5 10 15
Time (min)
15
Increasing Productivity through Process Optimization
It is a good idea to measure the actual cav- again, and it may be necessary to optimize Environmental influences
ity wall temperature (via the surface tem- other parameters once more in order to make
perature of the molded part immediately af- corrections to dimensions and weight. This The environmental influences acting on the
ter demolding, for example) and to have this will ultimately take a great deal of time and injection molding process include
temperature monitored during optimization incur costs that have a negative impact on
and processing. At all events, it is important productivity. This makes it clear just how ● drying and material conveyance
to be aware of the correlation that exists be- important it is to maintain the optimum op- ● the machine location
tween the inflow temperature and the actu- erating point during the start-up phase, and
al cavity temperature. If malfunctions oc- to keep to this during production. Interrup- ● cleanliness
cur, for instance (causing interruptions to tions to production and subsequent correc- ● ambient temperatures and relative hu-
production), or if the mold wall temperature tions to processing parameters should be midity
is altered by a change in the inflow temper- avoided as far as possible, and the melt and
ature or the cycle time, then this can result cavity temperature should be taken as the ● the energy, air and water supply
in dimensional and weight changes (Fig. basis for all optimization steps. ● the handling of the injection molded
19). It is then essential to wait until the equi- parts
librium temperature has become established
● the skill level of the work force.
16
Increasing Productivity through Process Optimization
To do this, the parameters are varied in the There are different methods that can be It is important for the processing parame-
following order (using the topmost value as adopted for optimization. The chief targets ters and all the measured data to be docu-
seldom as possible in a test series and the of process optimization are: mented for each of the parameter variations.
lower values more frequently): The setpoint values and the actual values
● production with a "robust" process set-
obtained with these should be recorded for
● melt temperature (by changing the cylin- ting. Although this will not smooth out
the individual processing parameters, since
der temperatures); since changing the inevitable disturbance factors such as
these two values can deviate from each oth-
melt temperature is a highly time-con- temperature fluctuations, control toler-
er. Each time the setting is modified, it is
suming process, this should be changed ances for velocities and pressures, or
essential to wait for a state of equilibrium
as rarely as possible batch-conditioned changes in molecular
to be established again. Only then will the
weight, additives and filler contents, etc.,
● mold temperature – this is similarly a reproducible molded part quality be
it will nonetheless minimize the impact
highly time-consuming change achieved that forms the basis for stable pro-
that they have on molded part quality [8]
duction. The dimensions of the molded part
● injection velocity
● improved cost efficiency through a re- should be checked both immediately after
● holding pressure level duced reject rate and a shorter cycle time. cooling and after the time stipulated in DIN
16901 (24 to 98 h after production), since
● holding pressure time (to establish the
Precisely this requirement for a high level the dimensions can still change a long time
sealing time)
of process reliability can make it necessary after the injection molding process has end-
● cooling time. to perform subsequent corrections to the in- ed. A relationship can then be established
jection mold. The most appropriate way of between the molded part dimensions directly
Once a reliable setting has been attained establishing a robust process setting and de- after cooling and the definitive dimensions
which will give flawless parts, the next step termining the changes required to the mold following an appropriate storage period,
involves checking the remaining quality cri- is to use systematic optimization tools, such which can be used in future start-up process-
teria, such as as statistical test planning. The optimum op- es. Following this, all the setpoint values
erating point can be established by means and resultant actual values for the optimum
● dimensional and shape accuracy
of regression calculations or an assessment operating point should be recorded so that
● stress status aided by neural networks. These studies al- they can be reproduced as rapidly as possi-
so provide information on the potential cen- ble for future start-up processes. This will
● material degradation
tering parameters. A number of manufac- keep downtime and waste to a minimum and
● mechanical and physical properties turers [9] offer predictive models based on considerably improve productivity.
these data which use the data documented
The associated parameters must then be cor-
during the process to determine whether the
rected, where necessary, making sure that
molded part will probably be of good qual-
the optical properties are retained and that
ity, poor quality, or still needs to be assessed.
production is not run with settings right on
The subsequent outlay on measurement and
their upper or lower limits.
documentation can be reduced considerably
in this way.
18
Increasing Productivity through Process Optimization
Different concepts can be employed for pur- The evolution strategy, for instance, is based characteristics. The computer can then main-
poses of establishing the optimum operat- on the evolution theory (Fig. 21). This in- tain the optimum operating point established
ing point as rapidly and reliably as possible volves the optimum operating point being in this way within pre-set tolerance limits
and hence for achieving optimum molded established automatically with the aid of [7]. If this approach is compared with the
parts. A number of these are listed here by computer programs when a new mold is run heuristic method in which the parameters
way of example: in or when production is already running. are modified on an individual basis and the
The computer modifies the process para- impact of the modification assessed (trial
● the evolution strategy (with the appro-
meters within meaningful limits to this end. and error), then the evolution strategy has
priate hardware and software)
The quality results obtained in this way are clear advantages in that it is quicker to use,
- intelligent trial-and-error rated either automatically or manually and since more than one parameter can be varied
notified to the computer. Drawing a com- at a time and the impact assessed automati-
- use of the evolution mechanisms
parison with the preceding settings and re- cally [2]. A further advantage compared
● statistical test planning (different methods) sults, the computer repeatedly selects new with the single-point method is that the in-
settings, thereby approaching the point of teraction effects of the individual parame-
- design of experiments
operation that offers the most favorable ters can be detected and assessed.
- planning by the Shainin method compromise between the individual quality
- neural networks (for assessments)
- regression analyses (for assessments)
● start-up aids from machine builders or
raw material producers, some of whom
employ the concepts listed above.
Parent
Offspring
Quality
er 1
amet
Par
Parameter 2
20
Increasing Productivity through Process Optimization
Which material was used? ● Classification of these additional faults Fault-diagnosis aids
- semi-crystalline or amorphous
The questions set out above regarding the
- unfilled, fiber-reinforced, mineral-filled Machine operators can eliminate a large
boundary conditions should then be an-
- hygroscopic number of faults very rapidly by virtue of
swered for these additional faults that have
- especially sensitive to processing their general experience of injection mold-
been established either on the molded part or
- ... ing or their knowledge of the particular
during the production process, and the po-
What mold structure was employed? tential causes of the faults determined. process employed. When it comes to recti-
- with a hot runner fying more complex faults, the operator can
- a three-platen mold ● Determination of the most probable call upon advice from experienced col-
- with direct gating causes of the faults leagues or from experts at the machine man-
- ... ufacturer's or raw material producer's.
Determining the causes that most faults
What measurements were recorded for the would tend to suggest. More detailed information on troubleshoot-
actual processing conditions? ing can be found in both the printed litera-
- melt temperature ● Establishment of the best remedial ture [10, 11, 12, 13] and the electronic lit-
- mold surface temperature measures erature [14]. Where several faults have oc-
- injection time (injection velocity) curred simultaneously, the most probable
Allocation of remedial measures to the most cause can readily be established with com-
- fluctuations in the processing conditions
probable causes of the fault, selection of puter-based aids. Background information,
- ...
promising remedial measures on the basis such as integrated "Application Technolo-
of the following criteria: gy Information" can then be used to select
● Determination of potential causes of
- efficiency the best corrective action.
the fault
- implementation outlay
The information that is established should - economic impact (cycle time, energy re-
be compared with recommendations from quirement, finishing work).
troubleshooting guides, start-up aids from
The best solution is always to eliminate the
the machine builders or raw material pro-
actual cause of the fault (due to the machine
ducers, or with the appropriate computer
or the mold, for example). Only in excep-
programs or expert systems.
tional cases should the symptoms be elimi-
nated by adjusting the processing parame-
● Search for further detectable faults
ters.
In many cases, a closer inspection will re-
veal further faults on the reject parts. These ● Rectification of the fault and continu-
additional faults, known as symptoms as in ation of production
medicine, will provide further information
on the most probable causes of the fault and
thus permit a more reliable diagnosis.
23
Increasing Productivity through Process Optimization
24
Increasing Productivity through Process Optimization
Processing
References
[1] B. M. Bauer Investigations into the use of an automatic optimization
system for material assessment in injection molding,
Diploma Thesis (in German),
University of Erlangen 1996
[2] H. Offergeld, J. Lochner Darwinian optimization of operating points (in German),
Kunststoffe 85, Carl Hanser Verlag 1995
[3] H. Jahn Production quality, the logical consequence of the
quality of workmanship (in German), VDI-Z,
130 (1988) 4, pp. 4-12
[4] Anon Durethan® KU 2-2140/30 H2.0
Application Technology Information 0953e
Bayer AG 1996
[5] Anon CONTURA System, Presentation documents from
Innova Engineering GmbH
[6] A. J. Kaminski, Relationship between screw diameter,
F. W. Lambeck metered volume, density and shot weight
ATI 1103e, Bayer AG 1997
[7] S. Joisten Injection molding: process optimization and control
PI 061e, Bayer AG
[8] R. Bourdon Quality optimization in injection molding through
robust process settings,
Seminar on optimum injection molds (German),
Süddeutsches Kunststoff-Zentrum, Würzburg 1999
[9] Hohenauer, Michaeli, Injection molding: reliably predicting the quality of each
Vaculik, Savoie dividual part (in German), QZ 39, Carl Hanser Verlag 1
Rapperswil, Wybitul
[10] K. Niemann (Ed.) Machine-setting strategy for thermoplastic injection
molding machines (in German),
Hüthig Verlag, Heidelberg 1992
[11] M. Bichler Injection molding flawless plastic parts (in German),
Hüthig Verlag, Heidelberg 1999
[12] P. Thienel, R. Sass, Troubleshooting guide for surface defects on
C. Vitz, J. Wahle thermoplastic molded parts (in German),
Kunststoffinstitut, Lüdenscheid 1992
[13] A. J. Kaminski, Spritzgießen – Fehler, Ursachen, Abhilfen
F. W. Lambeck Injection moulding – Faults, causes, remedies
Moulage par injection – Défauts, causes, remèdes
CD-ROM, Bayer AG 1999
[14] Anon Guide to Injection Moulding, CD-ROM,
Bayer AG 1999
27
Bayer AG
Bayer plastics on the Internet: Bayer Polymers
www.plastics.bayer.com D-51368 Leverkusen
This information and our technical advice – whether verbal, in writing Unless specified to the contrary, the values given have been established
or by way of trials – are given in good faith butwithout warranty, and on standardized test specimens at room temperature. The figures
this also applies where proprietary rights of third parties are involved. should be regarded as guide values only and not as binding minimum
Our advice does not release you from the obligation to verify the in- values. Please note that, under certain conditions, the properties can be
formation currently provided – especially that contained in our safety affected to a considerable extent by the design of the mold/die, the
data and technical information sheets – and to test our products as to processing conditions and the coloring.
their suitability for the intended processes and uses. The application,
use and processing of our products and the products manufactured by
you on the basis of our technical advice are beyond our control and, 2002-09
therefore, entirely your own responsibility. Our products are sold in ac-
cordance with the current version of our General Conditions of Sale
KU 11829en
and Delivery.