Comparison of Switchover Methods for Injection Molding

1
Comparison of Switchover Methods for Injection
Molding

• David O. Kazmer, Sugany Velusamy, Sarah
  Westerdale, and Stephen Johnston
• Plastics Engineering Department
• University of Massachusetts, Lowell
• Priamus Users Group Meeting
• September 30th, 2008

2
Agenda

• Motivation
• Manufacturing competitiveness
• Characteristics of highly productive molders
• Switchover Methods
• Overview
• Experimental Setup
• Results
• Conclusions

3
Is U.S. Manufacturing in Decline?
4
Is U.S. Manufacturing in Decline?
5
U.S. Manufacturing Productivity
6
U.S. Manufacturing Productivity

• Manufacturers need 1.5 annual productivity gains
  to remain competitive

7
Characteristics of Highly Competitive Molders

• Highly systematized
• Excellent layout
• Consistent and often uni-directional flow of
  materials
• Uniform internal planning processes
• Uniform quality control processes.
• Many highly productive facilities use only one
  primary supplier of plastics machinery.

8
Characteristics of Highly Competitive Molders

• Highly utilized
• 24 x 7 operation
• 90 plus machine utilization
• Steady state strategy
• Use fewer and better machines running
  continuously rather than more machines running
  fewer shifts

9
Characteristics of Highly Competitive Molders

• High yields
• 95 typical
• 99.8 not necessary
• High quality assurance
• Automatic in-mold systems, vision, poka-yoke
• Conservative rules to contain defects
• Better to automatically reject 10 good parts than
  accept one bad part

10
Characteristics of Highly Competitive Molders

• Industry sector andapplication focus
• Connectors
• Gears
• Syringes
• Focus provides
• Advanced application-specific knowledge
• Market commitment and technology investment

11
Obsolete vs. Competitive

• Number of machines

Obsolete Competitive
12
Obsolete vs. Competitive

• Number of workers

Obsolete Competitive
13
Obsolete vs. Competitive

• Number of supervisors

Obsolete Competitive
14
Obsolete vs. Competitive

• Plant size

Obsolete Competitive
15
Obsolete vs. Competitive

• Energy usage

Obsolete Competitive
16
U.S. Manufacturing Productivity

• Manufacturers need 1.5 annual productivity gains
  to remain competitive

17
Agenda

• Motivation
• Manufacturing competitiveness
• Attributes of highly productive molders
• Switchover Methods
• Overview
• Experimental Setup
• Results
• Conclusions

18
Overview Switchover Concept

• Switchover is the point at which the filling
  phase ends and packing phase starts
• From a controls perspective, there is a switch in
  the systems boundary conditions and stiffness
• Variances cause
• Dimensional errors
• Part weightvariations
• Back flow

19
OverviewSwitchover Methods

• Various methods for switchover
• Screw Position
• Injection Time
• Injection Pressure
• Cavity Pressure
• Cavity Temperature
• Nozzle Pressure
• Tie Bar Deflection
• Other studies have been conducted.
• This study is more comprehensive with respect to
  number of methods and also long term variation.

20
Experimental Setup

• Molding Machine
• 50 metric ton All Electric Machine
• Make Ferromatik Milacron
• Model Electra 50 Evolution
• Plastic Material
• AMOCO Polypropylene
• Grade 10-3434

21
Process Monitoring Control

• Extremely well instrumented machine mold
• Screw position transducer
• Nozzle pressure transducer
• Ram load transducer
• 3 barrel thermocouples
• 4 in-mold pressure transducers
• 2 in-mold temperature sensors
• Nozzle infrared pyrometer
• In-mold infrared pyrometer
• PRIAMUS DAQ8102 acquisition
• Custom machine override circuit
• Internal or external voltage signal triggers the
  machine for switchover

22
Switchover Methods Measured Attributes

• Seven Switchover Methods
• Machine Controlled
• Screw Position
• Injection Pressure
• Injection Time
• Externally Controlled
• Nozzle pressure
• Runner Pressure
• Tensile Cavity Pressure
• Cavity Temperature

• Six Measured Attributes
• Impact Thickness (mm)
• Impact Weight (g)
• Impact Width (mm)
• Tensile Thickness (mm)
• Tensile Weight (g)
• Tensile Width (mm)

23
Single Cycle Screw Position, Nozzle Pressure,
Cavity Pressure
24
10 Consecutive Cycles
25
Molding Machine Statistical Characterization

• 100 consecutive molding cycles were monitored
  data acquired
• The average standard deviation was calculated
  to measure of short term variation

26
Switchover Settings

• Switchover values for each method were determined
  to provide same part weight

27
Design of Experiments (DOE)

• DOE performed to impose long term variation

28
Analysis

• The 90 cycle DOE was repeated for each of the
  seven switchover conditions
• Parts weighed dimensions measured
• The data was analyzed in Matlab to provide
• Individual traces for each of 630 cycles
• Overlaid traces for all cycles in a DOE run
• Overlaid traces for all cycles in a switchover
  method
• Regression coefficients main effects plots

29
90 Cycles across the DOE for Ram Position
(Conventional) Switchover
30
Main Effects on Impact Thicknessfor Ram Position
Switchover
Good process robustness
31
90 Cycles across the DOE for Filling Time
Switchover
32
Main Effects on Impact Thicknessfor Filling Time
Switchover
Very poor process robustness
33
90 Cycles across the DOE for Cavity Pressure
Switchover
34
Main Effects on Impact Thicknessfor Cavity
Pressure Switchover
Good process robustness
35
90 Cycles across the DOE for Cavity Temperature
Switchover
36
Main Effects on Impact Thickness Cavity
Temperature Switchover
Best process robustness
37
Coefficient of Variation COV s / µ
Different switchovers are best for different
attributes
38
Switchover PerformanceShort vs. Long Run
Variation
Short Run Variation ()
More robust
Long Run Variation ()
39
Switchover PerformanceLong-Run Variation
Screw position
Injection time Machine
pressure Nozzle pressure
Runner pressure Cavity pressure Cavity
temperature
40
Conclusions

• Cavity temperature provided the most robustness
  against changes the process settings.
• Place the sensor near but not at the very end of
  flow due to small control system delays (speed
  matters)
• Cavity pressure provided reasonable switchover
  control but had susceptibility to changes in melt
  temperature and velocity.
• Position control provided reasonable control but
  roughly twice the variation of cavity
  temperature.
• Injection time is the least reproducible method
  for the transfer from fill to pack, with
  literally 10 times the variation of temperature
  control.

41
Conclusions

• Measured consistency is much better than SPI
  guidelines of ?0.2
• Response time of the molding machine, controller
  and ram velocity are important to process
  repeatability.
• Weight and thickness show higher COV than length
  and should be used for QC

In-mold instrumentation is vital to achieving
process robustness, automatic quality control,
and competitiveness.
42
Acknowledgements

• National Science Foundation grant
  numberDMI-0428366/0428669
• Priamus System Technologies