Optimizing Thermoforming of High Impact Polystyrene (HIPS) Trays by Design of Experiments (DOE) Methodologies

Optimizing Thermoforming of High Impact Polystyrene (HIPS) Trays by Design of Experiments (DOE) Methodologies

Vishal M. Dhagat, Ravindra Thamma , University of Connecticut;

Central Connecticut State University

Abstract – The process of heating and reshaping plastic sheet and film materials has been in use since the beginning of the plastics industry better known as thermoforming [1]. Today this process is very ubiquitous for industrial products including signage, housings, and hot tubs [1]. It also produces much of the packaging in use today including blister packs, cartons, and food storage containers [1]. The process of thermoforming has many advantages over other methods for producing high-quality plastic products, with some limitations, which is resolved by implementing stringent quality control using scientific methods to improve process performance. Two areas of interest in todays industry of great concerns are lean manufacturing operations and environment [2]. Polystyrene scrap must be segregated from other materials in the waste stream before it can be recycled [3]. Thermoforming of high impact polystyrene sheets using Vacuum thermoforming technique requires technical knowledge on material behavior, mold type, mold material, and process variables. Research on various subjects documented but limited research focuses on the process optimization of HIPS (High Impact Polystyrene) [2]. The design of Experiments (DOE) approaches like those that the face-centered cubic central composite design used to refine the process to cut rejects [2]. This paper presents a case study on thermoforming of HIPS single use trays made on a semi-automatic machine using three criteria solely based on the FCC Design method. The optimization of HIPS tray forming using wall thickness distribution explored. Results show that ideal performance parameters achieved using Design Of Experiments [1].

INTRODUCTION

The development of educational and industrial software and Thermoforming is an industrial process in which thermoplastic sheet (or film) transforms into a new shape using heat and pressure. Thermoforming is one of the earliest processes in the plastic industry since the mid- 1800's beginning with the forming of cellulose nitrate sheet in [1]. The growth increased dramatically as new materials and applications developed [1]. For example, the need for aircraft canopies in World War II along with the development of poly-methyl-methacrylate (acrylic) created the perfect opportunity to advance thermoforming process technology. 5% to 6% growth rate of thermoforming process technology sustained over forty-five years.

Today this process produces many products from small blister packs to display AAA size batteries to large skylights and aircraft interior panels. End use of the manufactured products defines the market. "Industrial Products" include items with expected long life such as those used in the transportation and construction industries. "Disposable

Products" (non- packaging) include items that have a short life expectancy but are not on the packaging side of the business [1]. This market includes disposable plastic plates and drinking cups. "Packaging Products" is a huge, high volume, an industry devoted to providing manufacturers with low-cost packaging to display, protect, and/or extend the life of their products [2].

Various new research technologies define new and exiting thermoforming processes and products. Process simulations using novel computer based software like COMSOL developed and well studied. Detailed new studies on process optimization of PET using Taguchi Method proposed [1], [2]. Taguchi Method works with a single qualitative characteristic or response. However, most products have several qualitative characteristics or responses of interest [4]. Process conditions and plug materials in plug-assisted thermoforming are well investigated [1], [2], [5]. Most thermoforming processes the important practical operating consideration is the ability to control the wall thickness distribution, since this will largely determine the physical properties of the final product. An uneven thickness distribution the thinnest regions will dictate mechanical strength of the formed part. Industry is very well focusing in trying to balance the wall thickness of thermoformed parts [6]. Although many studies conducted, none have investigated the process about the ideal processing settings that can produce the high yield with consistent part thickness and least processing time [2]. The face-centered cubic central composite Design of Experiments is an all- inclusive method that can be used to optimize product quality

/ trays by implementing suggested processing parameters while minimizing waste and process iterations.

HIGH IMPACT POLYSTYRENE (HIPS)

Polystyrene is the fourth biggest polymer produced in the world after polyethylene, polyvinyl chloride, and polypropylene. General-purpose polystyrene (GPPS) is a glass like polymer with a high processability [7]. Polystyrene modified with rubber known as high impact polystyrene (HIPS) encompassing characteristics, like toughness, gloss, durability and an excellent processability [8]. Polystyrene is one of the most versatile plastics. Whether packaging for food products, office and information technology, refrigerators, all sectors place high demands on the properties of the materials used. In its diverse variants, HIPS offers extraordinary

property combinations, thus making a vital contribution to everyday life. High impact polystyrene is also used in many applications because of its excellent balance of properties and low-cost. HIPS have good impact resistance, dimensional stability, excellent aesthetic qualities, is easy to paint and glue, manufactured at a low-cost and approved by the U.S. Food and Drug Administration [8].

Face-Centered Cubic Central Composite Design

Design of experiments (DOE) is a statistical technique introduced by Sir R. A. Fisher in early 1920s [7]. Face- centered cubic Central Composite Design is a Design of Experiments (DOE) method [3]. FCC Design can have many factors that affect your process outcome simultaneously. Studying each factor one at a time would be very expensive and time-consuming, and you would not get any information about how different factors interact with each other [3]. That is where the design of experiments comes in. DOE turns the idea of needing to test only one factor at a time on its head by letting you change more than a single variable at a time [3]. This minimizes the number of experimental runs you need to make, so you can get meaningful results and reach conclusions about how factors affect a response as efficiently as possible.

Oven % On Time, Heating Time, and Vacuum Time, are the three process variables of interest [2]. Some variables have more importance than others and some show important interdependence or interactions with others [1]. A deep understanding of your current thermoforming process and equipment is essential prior conducting the experiments to get robust results [2]. A simple screening experiment is necessary to weed out which of these factors have the biggest effect on the part quality [1]. This method provides a robust combination of process variables that examines for best part quality and least deviation from the target [2].

There is three distinct steps in this method: preparation of the trials, the realization of the trials and analysis of results as shown in figure 1 [2]. 1. Preparation of Trials: The specific characteristic (response) analyzed [2]. Through some experimental runs and pior knowledge of equipment and processes, the most important variables identified and levels determined. 2. Then, the right Face-centered cubic central composite design selects levels in the Minitab software. Minitab software produces the orthogonal array table to create the trials. Additional tables created to ease analysis [1]. After performing all trials and recording all relevant data, results analyzed using adapted averages calculations and variance analysis. Minitab enabled us to gain a better graphical representation with contour plots of the results [2]. An optimal combination of different variables at the right condition obtained. Using the suggested optimal combination of process parameters, a final vacuum thermoformed part made for validation.

Figure 1. Design of Experiments process flow chart

Equipment Setup

Thermoforming of HIPS Trays performed on a MAAC thermoforming machine. Several adjustments were necessary in order to perform thermoforming of trays successfully which are listed below:

1. The vacuum connection was modified with a connector to facilitate easy disconnect of hose.

2. Pneumatic clamps that hold blanks adjusted for the size of 16in X 20in blanks.

3. Several dry runs helped remove any kinks in processing steps.

4. Software on controller did accommodate our processing parameters.

5. Several wet runs while changing parameters on the controller helped to get a good finished product.

All kinds of process variables from a cooling fan on and off time to vacuum and heating times analyzed to detect which factors have a direct relationship with part quality.

Processing of HIPS Tray

Thermoforming process involves a large and rapid deformation of a sheet and process is a common technique in forming of thermoplastic sheets. A pre-manufactured thermoplastic sheet is clamped in a place and is heated to its softening temperature. In the next step, the softened material is formed into a female/male mold by applying atmospheric pressure against a vacuum [9]. The heating stage is of primary importance in the thermoforming process [10] along with cycle time, because it has influence on the pieces final thickness distribution [11]. Thermoforming of the HIPS trays performed over a time of 4 weeks in the summer afternoons to minimize effects of the room temperature, humidity and other uncontrollable factors. The processing sequence followed as listed in Table 1. The equipment allowed warming up between the runs when temperature settings changed.

Table 1. Process variables and their levels

Process Variables	Letter	Levels
		L1	L2	L3
% Oven On (Temp. Setting)	A	30 (310°F)	35 (335°F)	40 (370°F)
Heating Time (Sec.)	B	30	40	50
Vacuum Time (sec.)	C	2	4	6

Note: % Oven On process variable results in corresponding even oven temperatures that resulted in HIPS sheet temperatures shown in parenthesis.

Processing steps as listed below:

1. Turn On main power

2. Turn On power to equipment and vacuum pump

3. Turn On shop air half way and make sure the pressure gauge reads 15 PSI on the equipment (used for pneumatic clamps)

4. Turn control panel power on by releasing the red button

5. Go to main settings and change temperature and time as per your requirement

6. Turn oven On and wait till oven reaches the required set point and room temperature stays uniform

7. Put the HIPS blank in the holder and adjust to keep it in the center

8. Start the thermoforming process by pressing 2 green buttons on control panel

9. The clamp will close and the pneumatic lift will move the clamped HIPS blank into the oven

10. Blank will start to get hot and will reach a molding temperature as set by user

11. Measure sheet temperature right before it is moved down to thermoforming station

12. Now the bottom mold will get lifted to the desired stop and vacuum will turn on

13. After the sheet is molded it will get ejected by air pressure and clamp will release while the fan will turn on to cool down mold and formed parts

14. The thermoformed HIPS tray removed gently and a label applied to match a sequence number, temperature settings and forming time along with sheet temperature.

15. Process will be repeated for the next sequence

A reasonable delay time of 5 minutes was added to get the oven back to the set point temperature

Process Variables: HIPS Tray

For this research, the equipment used is a MAAC Thermoforming System Single Station Model # ASP, Serial # 03904, having a total molding and clamping area of 30 x 36 and an oven of about 36 x 48 in size. The HIPS tray made from a sheet stock of High Impact Polystyrene (HIPS) 16 x 20 and 0.040 thick. Figure 2 and 3 shows the CAD model and thermoformed tray respectively. A female mold made of aluminum with four cavities mounted with clamps on the bottom pneumatic holder. Vacuum channels in the bottom female mold assist in proper part formation. The equipment is a semi-automatic laboratory use machine. The main quality specifications (response) selected are even wall thickness with least variance from one cavity to

the next. Figure 4 and 5 shows MAAC thermoforming machine and aluminum mold with vacuum holes.

Figure 2. CAD model of HIPS tray

Figure 3. Thermoformed HIPS tray

Figure 4. MAAC Thermoforming Machine

Figure 5. Aluminum mold with vacuum holes

After some brainstorming and some trial runs, the optimization study would consider the effect of three process variables on quality specifications [12]. Each variable tested at three different levels [12]. Selected variables and levels as listed in Table 1.

Optimization Plot Overlays of Results

Residual plots for thickness variance and part quality of all thermoformed trays. Results plots as in figure 8 and 9.

Residual Plots for T Var

Normal Probabilit y Plot Versus Fit s

The 40 trials completed consecutively for about 5 minutes of production time for each trial, which includes loading of blank sheets, clamping of blank sheets, processing, measuring the temperature of HIPS sheet with an infrared thermometer,

99

P e r c e n t

90

50

10

1

-0.00001

0.00000

Residual

0.000090

0. 000095

0.000100

0.000105

0.000110

0.00001

0.00001

R e s id u a l

0.00000

-0.00001

Fitted Value

0000

removal of the finished tray and changing settings on the controller of the MAAC machine [2]. Wall thickness of the bottom, sides and corners were measured. Total of 11 measurements per tray recorded in Minitab. Wall thickness

Hist ogram Versus Order

F r e q u e n c y

R e s id u a l

16 0.00001

12

0.00000

8

4 -0.00001

0

1 5 10 15 20 25 30 35 40

-0.000016

-0. 000012

– 0. 000008

-0.000004

0.00

0.000004

0.000008

0.000012

measurements analyzed to get mean thickness per tray; the standard deviation of each tray and the variance analysis/p>

Residual

Observation Order

performed for each tray. There were 10 instances where the processed sheet did not result in a part that analyzed for the thickness measurement so a zero part quality and a max

Figure 8. Residual plots for thickness variance data

Residual Plots for Part Quality, %

variance of 0.00011 assigned as shown in figure 7.

Normal ProbabRileitsyiPdluotal Plots for Part Quality, %

99 Normal Probability Plot

99 10

Versus Fits Versus Fits

Preparing HIPS Tray for Analysis

Before any analysis can begin, the processed trays cut in

90

tn 90

e

etnc

rce 50

Pre 50

P

10

10

1

1

-20 0

al adl

u

i

u

e

e

ds Ris R

20

10

0

0

-10

-10

-20

-20

-30

-30 0

25 50

75 100

half and edges sanded to make it measurement ready. The reason to cut the trays was to expose the profile as shown in figure 6, which will make it easier to measure thicknesses with a vernier caliper.

20

20

c

15

y 15

ny cenu

eque10

Fr

qre10

F 5

5

0

-20

Residual 0

Residual

Histogram Histogram

20

edi

al adul uis

e

Rs

R

0

10

10

0

0

-10

-10

-20

-20

-30

25 Fitted Va5l0ue 75

Fitted Value

Versus Order Versus Order

100

0 -30

-20

-10

0 10

-30 1 5

10 15

20 25

30 35 40

-30

-20

R-e10sidual 0 10

Residual

1 5 10 Obs1e5rvatio2n0 Order25

Observation Order

30 35 40

T7

T6

CT4

T5 CT3

T4

CT2

T1

T2

T3 CT1

Figure 9. Residual plots for Part Quality

Contour plots as shown in figure 10 and 11 indicate that a vacuum time of approximately 4.75 seconds is optimal.

d Part 50

45

%

Contour Plot of T Var vs Heating Time, s, On Time, %

T Var

< 0.000086

Figure 6. HIPS tray profile top and side view

The measurements were performed at 11 different

ality

0.000086

0.000088

0.000090

0.000092

>

0.000088

0.000090

0.000092

0.000094

0.000094

H e a t in g Tim e , s

locations. A visual part quality assigned as shown in figure 7.

40

Hold Values Vacuum Time, s 4.75

Good Part 100%

Quality

Run # 40

90% Quality

Run # 1

60% Quality

35

30

30 32

34 36

38 40

Run # 20

90% Quality

Run # 2

0% Quality

On Time, %

Figure 10. Contour plot of thickness variance vs. heating time and % on time

Figure 7. Visual part quality rating of thermoformed trays

Contour Plot of Part Quality, % vs Heating Time, s, On Time, %

Part
Q uality , %
< 85
85 90
90 95
95 100
100 105
105 110
> 110
Hold Values Vacuum Time, s 4.75

50

H e a t in g Tim e , s

45

40

Table 2. Recommended vacuum thermoforming process parameters

Process Variables	Letter	Recommended Thermoforming Parameters
% Oven On (Temp. Setting)	A	37
Heating Time (Sec.)	B	50
Vacuum Time (sec.)	C	4.75

35

30

30 32

34 36

38 40

CONCLUSION

A Face-centered cubic central composite design used to optimize vacuum thermoforming of HIPS trays [2]. The MAAC thermoforming equipment used and optimal

On Time, %

Figure 11. Contour plot of part quality vs. heating time and % on time

Contour plots reformulated using vacuum time value of

H e a t in g Tim e , s

4.75 seconds as recommended by the analysis. Contour plots overlaid to minimize thickness variance and maximize part quality as shown in figure 12. Finally, considering extreme (robust) possibilities the optimized results plot is shown in figure 13 below.

Contour Plot of T Var, Part Quality, %

50

T Var

7e-005

8.8e-005

Part

45 Q uality , %

100

120

Hold Values

40 Vacuum Time, s 4.75

35

30

30

32

34

36

38

40

On Time, %

H e a t in g Tim e , s

Figure 12. Contour plot of thickness variance vs. part quality

Contour Plot of T Var, Part Quality, %

50

T Var

7e-005

8.8e-005

Part

45 Q uality , %

110

120

Hold Values

40 Vacuum Time, s 4.75

35

30

30 32 34 36 38 40

On Time, %

Figure 13. Optimized Contour plot of thickness variance vs. part quality

Recommended Vacuum Thermoforming Process Parameters The recommended vacuum thermoforming parameters to

improve part quality by minimizing thickness variance are

shown in table 2.

combination of process variables for the tray obtained [2]. This method is a simple and efficient approach if adopted can yield significant process improvements in an industrial production setting [2]. It resulted in short production times (< 120 seconds) and yields a robust product quality, minimizing waste and reprocessing [2]. A well-prepared test will bring relevant and useful results for economical production cycle. Systematic process optimization by the DOE enabled defect- free and uniform wall thickness and radii.

REFERENCES

1. Klein, P. (2009). Fundamentals of plastics thermoforming. Synthesis Lectures on Materials Engineering, Morgan & Claypool Publishers.

2. Labonte, M., & Dubois, C. (2011). Optimization of molding conditions of a plug-assisted thermoformed thin containers in a high speed and volume production context. Thermoforming Quarterly, 30(3), 16-17, 20-22.

3. Expanded polystyrene (EPS). (n.d.). Retrieved August 31, 2016, from http://www.plasticseurope.org/what-is- plastic/types-of-plastics-11148/expanded-polystyrene.aspx.

4. Yang, C., & Hung, S. W. (2004). Optimizing the thermoforming process of polymeric foams: an approach by using the Taguchi method and the utility concept. The International Journal of Advanced Manufacturing Technology, 24(5-6), 353-360.

5. Martz, E. (2013, March 25). Getting Started with Factorial Design of Experiments (DOE). Retrieved August 31, 2016, from http://blog.minitab.com/blog/understanding- statistics/getting-started-with-factorial-design-of- experiments-doe

6. P.J. Martin, H.L.C., C.Y. Cheong, E. Harkin-Jones (2009). Plug materials for thermoforming: the effects of non- isothermal plug contact. ANTEC 2009: The Plastics Technology Conference. 812-816.

7. Roy, R. K. (2001). Design of experiments using the Taguchi approach: 16 steps to product and process improvement. John Wiley & Sons.

8. Strong, A. B. (2006). Plastics: materials and processing. Prentice Hall.

9. H. Hosseini, B. V. Berdyshev, & A ehrabani-Zeinabad (2009). Dynamic characteristics of plug assist thermoforming process. Polymer Engineering and Science, 49(2), 240-243. Doi:10.1002/pen.21245

10. Z. Benrabah, P. Debergue, and A. Haurani (2005). Modeling the infrared sheet heating in roll-fed thermoforming. ANTEC 2005: The Plastics Technology Conference. 1197-1201.

11. Morales, R. A., & Candal, M. V. (2006). Diseño y fabricación de un molde de termoformado utilizando herramientas CAD/CAE. Revista de la Facultad de Ingeniería Universidad Central de Venezuela, 21(1), 83-99.

12. C.S. Härter, N.Tessier, and K. Kouba (2009). The dependence of wall thickness on changes in material and process conditions in plug assist thermoforming. ANTEC 2009: The Plastics Technology Conference. 817-821.

Biographies

VISHAL DHAGAT received his M.S. in Mechanical engineering technology from the Central Connecticut State University in 2013 and is currently a Ph.D. candidate in the Electrical and Computer Engineering Department at the University of Connecticut, Storrs, CT. His main areas of research include the synthesis and fabrication of surface acoustic wave sensors and semiconductor-based biosensors.

RAVINDRA THAMMA is currently a Professor of Manufacturing and Construction Management at the Department Engineering at Central Connecticut State University. Dr. Thamma received his Ph.D. from Iowa State University. His teaching and Research interests are robotics, linear control systems, and intelligent systems.