1 “AN.” - 1 ’Ti‘tzgfifif" raw a A. p r..- ." .44. TI .22." «aflgaspt -.- ‘ at" Wm ..... ~32 ~K r v 7&5 ._ ‘9.“ .2!- “Jam-7;“ 1.... ‘3") i ‘ ‘ . ” r ‘2‘ ' ‘ ‘ , ‘ , 15L)" .... . ‘ ' '4’ ‘ - 4 ‘ p; . . 1 . ,7 ”pi 4. "' '5 a ' '2 3:3 ,': 3. . . _' - ‘ w.” '13,“:- "3.“ .,..:1 Wfig’tfi': ’ rm ‘ ' 3‘1 Fr n " . ‘ ' «.9 - ‘ ‘ “Zn:- . :: a~ ' vii-$55,“; ' ~u 5 1 1...... u... 0"“ .- ..,.~.. - 7 . _ a ,. ., A...J...~..::L..w.. , “ ' ' $9“ M », ll lcllllll'lHillllllllfillllllfllfillll LIBRARY Michigan State University This is to certify that the thesis entitled EFFECTS OF POLYMERIC AND GLASS MICROBUBBLE CONTAMINATION ON RECOVERED MATERIALS FROM A LIGHT MEDIUM HYDROCYCLONE SEPARATION OF HIGH DENSITY POLYEHTYLENE FROM POLYPROPYLENE presented by EDWARD ARAM AKKASHIAN has been accepted towards fulfillment of the requirements for 64% (61% Major professor Date MAY 18, 1994 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution PLACE It RETURN BOXtoromavothb Mention macaw. TO AVOID FINES Mum on or him dd.“ DATE DUE DATE DUE DATE DUE w“: 0% Isa ' 0 “—41 MSU ioAn Attirmdtvo Action/EM Opportunity "amnion ‘ ' Wan-pd EFFECTS OF POLYMERIC AND GLASS MICROBUBBLE CONTAMINATION ON RECOVERED MATERIALS FROM A LIGHT MEDIUM HYDROCYCLONE SEPARATION OF HIGH DENSITY POLYETHYLENE FROM POLYPROPYLENE BY Edward Aram Akkashian A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of ' MASTER OF SCIENCE School of Packaging 1994 ABSTRACT EFFECTS OF POLYMERIC AND GLASS MICROBUBBLE CONTAMINATION ON RECOVERED MATERIALS FROM A.LIGHT MEDIUM HYDROCYCLONE SEPARATION OF HIGH DENSITY POLYETHYLENE FROM POLYPROPYLENE BY Edward Aram Akkashian As the amount of plastic recycled continues to increase, more efficient methods must be developed to collect, separate, and reprocess this recyclable material. This study analyzes the effects that a light medium hydrocyclone separation of high-density polyethylene from polypropylene may have on the materials recovered for reprocessing. The light medium proposed consists of an aqueous suspension of tmdlow glass spheres used in conjunction with hydrocyclone technology. Tensile, flexural, and impact properties were determined for both pure polymers and for each polymer with low contamination levels by the other polymer and by the glass microbubbles. Also determined were the environmental stress-crack resistance (ESCR), densities, and flow rates of the polymers with and without contaminants. While some properties of each polymer were adversely affected by the contaminants (impact strength and ESCR), the other properties were either unaffected or enhanced by the various levels of contamination. ACKNOWLEDGMENTS I would like to express my sincere appreciation to my major professor, Dr. Susan Selke, Ph.D. (School of Packaging), for her assistance and support. I would also like to thank my committee members Dr. Charles Petty, Ph.D. (Department of Chemical Engineering), and Dr. Jack Gaicin, Ph.D. (School of Packaging). Special thanks is extended to Mike Rich and Brian Rock of the Composite Materials and Structures Center for their instruction and help with all of the equipment used there. Sincere appreciation is also extended to Mark Sanderson at Himont for his extreme generosity with materials and his time in helping to mold the materials. The generosity of 3M and Quantum for their contribution of materials is also very much appreciated. ifi TABLE OF CONTENTS List of Tables ............................................ vi List of Figures .......................................... vii Chapter 1. Introduction ............................................ l 2. Literature Review ....................................... 7 2.1 Effects of Glass Bubbles in the Polymer Matrices..8 2.1.1 Particle Description ....................... 8 2.1.2 Tensile Properties ......................... 8 2.1.3 Flexural Properties ....................... 12 2.1.4 Impact Strength .......................... 12 2.1.5 Environmental Stress-Crack Resistance ..... 12 2.1.6 Flow Rate ................................. 13 2.2 Effects of HDPE and PP Cross-Contamination ....... 15 2.2.1 PP Contamination of HDPE .................. 16 2.2.2 HDPE Contamination of PP .................. l6 3. Materials and Methods .................................. 18 3.1 Materials ........................................ 18 3.1.1 Polymer Matrices .......................... 18 3.1.2 Hollow Glass Microbubbles ................. 18 3.2 Methods .......................................... 19 3.2.1 Method of Sample Preparation .............. 19 3.2.2 Tensile Properties ........................ 19 3.2.3 Flexural Properties ....................... 20 3.2.4 Impact Strength .......................... 21 3.2.5 Environmental Stress-Crack Resistance ..... 21 3.2.6 Flow Rate ................................. 22 3.2.7 Differential Scanning Calorimetry ......... 24 3.2.8 Density ................................... 25 3.2.9 Scanning Electron Microscopy .............. 26 3.2.10 Data Analysis ............................ 26 iv Table of Contents (cont'd) 4. Results and Discussion ................................. 28 4.1 Density .......................................... 28 4.1.1 MB Contamination of the Polymer Matrices..28 4.1.2 HDPE/PP Cross-Contamination ............... 30 4.2 Differential Scanning Calorimetry ................ 32 4.2.1 MB Contamination of the Polymer Matrices..32 4.2.2 PP Contamination of HDPE .................. 32 4.2.3 HDPE Contamination of PP .................. 32 4.3 Scanning Electron Microscopy ..................... 34 4.3.1 MB Contamination of HDPE .................. 34 4.3.2 MB Contamination of PP .................... 34 4.3.3 HDPE/PP Cross-Contamination ............... 39 4.4 Tensile Properties ............................... 39 4.4.1 MB Contamination of HDPE .................. 39 4.4.2 MB Contamination of PP .................... 41 4.4.3 PP Contamination of HDPE .................. 44 4.4.4 HDPE Contamination of PP .................. 44 4.5 Flexural Properties .............................. 45 4.5.1 MB Contamination of HDPE .................. 45 4.5.2 MB Contamination of PP .................... 45 4.5.3 PP Contamination of HDPE .................. 45 4.5.4 HDPE Contamination of PP .................. 48 4.6 Impact Strength ................................. 48 4.6.1 Contaminated HDPE ......................... 48 4.6.2 MB Contamination of PP .................... 48 4.6.4 HDPE Contamination of PP .................. 50 4.7 Env vir ronmental Stress- Crack Resistance ............ 50 4.7.1 MB Contamination of HDPE .................. 50 4.7.2 PP Contamination of HDPE .................. 50 4.7.3 Contaminated PP ........................... 52 4.8 Flow Rate ........................................ 52 4.8.1 MB Contamination of the Polymer Matrices..52 4.8.2 HDPE/PP Cross-Contamination ............... 53 5. Summary and Conclusions ................................ 55 References ................................................ 57 Appendix A ................................................ 59 Appendix B ................................................ 70 Appendix C ................................................ 73 LIST OF TABLES Table 1. Characterization of Glass in the Polymer Matrices ...... 31 2. Effects of Contamination on Polymer Density ............ 59 3. Effects of Contamination on Heat of Melting ............ 6O 4. Effects of Contamination on Tensile Strength at Yield..61 5. Effects of Contamination on Elongation at Yield ........ 62 6. Effects of Contamination on Tensile Modulus of Elasticity ............................................. 63 7. Effects of Contamination on Flexural Strength at 5% Strain ................................................. 64 8. Effects of Contamination on Flexural Modulus of Elasticity ............................................. 65 9. Effects of Contamination on Izod Impact Strength ....... 66 10. Effects of Contamination on Environmental Stress-Crack Resistance ............................................ 67 11. Effects of Contamination on Flow Rate ................. 68 12. Calculated Microbubble Breakage in HDPE and PP ........ 69 13. Number of PP Samples to Break Upon Bending When Testing for Environmental Stress—Crack Resistance ............ 71 LIST OF FIGURES Figure 1. Effects of Contamination on Polymer Density ............ 29 2. Effects of Contamination on Heat of Melting ............ 33 3. Micrograph of HDPE with 5% Microbubble Contamination...35 4. Micrograph Showing Typical Amount of Microbubble Breakage in HDPE ....................................... 35 5. Typical HDPE/Microbubble Interface ..................... 36 6. Typical Surface of a Microbubble ....................... 36 7. Micrograph of PP with 5% Microbubble Contamination ..... 37 8. Typical PP/Microbubble Interface ....................... 37 9. Micrograph of PP with 5% HDPE Contamination ............ 38 10. Effects of Contamination on Tensile Strength at Yield.40 11. Effects of Contamination on Elongation at Yield ....... 42 12. Effects of Contamination on Tensile Modulus of Elasticity ............................................ 43 13. Effects of Contamination on Flexural Strength at 5% Strain ................................................ 46 14. Effects of Contamination on Flexural Modulus of Elasticity ............................................ 47 15. Effects of Contamination on Izod Impact Strength ...... 49 16. Effects of Contamination on Environmental Stress-Crack Resistance ............................................ 51 17. Effects of Contamination on Flow Rate ................. 54 List of Figures (cont'd) 18. Graphical Method Used to Determine Environmental Stress- Crack Resistance ..................................... 7O 19. Equations Used to Convert Flow Rate to Viscosity ...... 72 ' CHAPTER I INTRODUCTION As consumer and legislative pressures build on producers of plastic packaging to incorporate more recycled content in their products, new, more efficient methods of recycling must be developed. Presently, most of the recycling infrastructure relies upon. manual sortation, of commingled containers either at the curbside or at a local materials recovery facility (mmJ). The high cost of this process and the low purity of the reclaimed plastic, however, are making apparent the fact that the most economically viable solutions lie in the separation of plastics through automated processes. With the throughput of recycled material increasing tremendously, economies of scale will soon necessitate automatic separation of plastics in replacement of the labor intensive processes employed at manual sorting facilities. Automatic sortation of plastics may be achieved through three routes - macro, micro, and molecular sortation. Macro sortation relies upon identification and separation of entire packages by various scanning techniques as the packages are passed by on conveyors or are dropped in front of a sensor. Micro sortation relies on fundamental properties of different plastics once they are granulated or otherwise made into small chips. Some of the properties currently exploited for 1 micro sortation are density, elecrostatic charge, tacking temperature, and behavior upon cryogenic impacting. Molecular sortation separates plastics through time use of different solvents or one solvent at varying temperatures. In the long run, the most viable solution involves the use of micro sortation of collected plastics. Molecular sortation provides a high purity product but at a relatively high cost (1) and solvent retention by the recovered plastics can be a concern. Also, these solvents have their own environmental impacts and it is sometimes hard to justify using these chemicals in order to "save the environment" from plastic waste. Macro sorting has limitations also. Multi-component packages may be classified by just one component causing high levels of contamination, or the package may be diverted from the system entirely, requiring manual sortation or landfilling. Crushed bottles may also cause difficulties in macro sorting facilities. Perhaps the largest advantage of micro separation over macro separation, however, lies not in the separation process, but in the collection of the recyclable plastic. In. the curbside collection. programs currently employed, a large economic setback is the low weight:volume ratio of the plastic containers collected. This problem cannot be dealt with if macro separation processes are used. The flex memory of plastics inhibits a high degree of densification and may also complicate the separation process. With micro separation, however, granulation at the curbside would result in a high degree of densification, allowing collection trucks to go on longer routes before returning to the local MRF. This would result in a large reduction in energy and labor costs. Coupled with the generation of a less contaminated (higher value) post consumer resin (PCR) stream, this translates into an opportunity for great economic advantage over other methods of separation. In light of the numerous advantages of micro sorting over other sorting processes, many efforts have been focused on new methods of micro sortation. The project of which this study is a part focuses specifically on the micro separation of blow molding grade high-density polyethylene (HDPE) copolymer from injection molding grade polypropylene (PP) homopolymer. HDPE and PP can be separated from most of the other commonly used packaging plastics using conventional hydrocyclone technology. The HDPE and PP (along with low density polyethylene), being less dense than water, go to the over flow, while the other plastics (such as PET, PVC, and PS) go to the underflow. In order to generate a valuable stream of resin from the overflow (in the absence of LDPE), HDPE and PP must then be separated from one another. This is being attempted through the use of a light medium hydrocyclone. The densities of blow molding grade HDPE copolymer and injection molding grade PP homopolymer are approximately 0.95 g/cm3 and 0.91 g/cm3 respectively. To achieve a medium of the optimal density for separation, an aqueous suspension of hollow glass microbubbles is being used. The HDPE being used in this study is typical of the resin used to produce opaque blow molded bottles for products such as laundry detergents, juices, motor oils, and household chemicals. The PP is typical of some of the injection molded spouts and closures that are affixed to or snapped or screwed onto the HDPE blow molded bottles. In 1993, 2,518 million pounds of HDPE was blow molded into containers of up to two gallons in America (2). That is over one fourth of the total HDPE produced domestically that same year. Because this is such a large, visible, easily identified and collected portion of all the HDPE produced, it is only logical to tap this valuable source of plastic - reducing natural resource consumption while simultaneously diverting this plastic from costly landfills. Of this 2,518 million pounds of HDPE that is blow molded, 1102 pounds are HDPE homopolymer (the type of resin used in milk bottles)(2). Because of their similar densities (HDPE hompolymer density is approximately 0.96 g/cm3), HDPE homopolymer and HDPE copolymer would both exit through the underflow in the system proposed here. In 1993, 130 million pounds of the copolymer was recycled. while 245 million pounds of the homopolymer was recycled. This results in recycling rates of about 9.2 percent and 22.2 percent, respectively (2). Because in some applications the mixing of these two resins may be unacceptable (especially when environmental stress-crack resistance is necessary), further study of the separation of HDPE homopolymer from HDPE copolymer is suggested. This problem is not addressed here. 528 million pounds of PP were used domestically to produce closures in 1993. This is equal to 6.2 percent of all the PP produced that year (2). Although, presently, emphasis is not placed on PP reclamation, interest is beginning to rise. In 1993 PP recycling doubled to 10 million pounds from 5 million pounds in 1992 (2). Also, if a relatively pure stream of PP is produced as a by-product of HDPE reclaim (through spouts and caps on HDPE and other bottles), this by-product can be sold to help offset the cost of processing the HDPE. This is analogous to the sale of gold, silver, and platinum to help offset the cost of purifying copper after it is removed from its ore. This process could be applied to the light fraction recovered from the separation of PET from the other constituents of soft drink bottles. The labels and caps on these bottles are PP, whereas the base cups are HDPE. This may soon lose its importance, however, as the trend turns towards self- supportive bottles with a petaloid base to replace the HDPE base cups in soft drink bottles. As a result of this trend away from base cups, the amount of HDPE recovered from PET soft drink bottle recycling has dropped from 45 million pounds in 1992 to 35 million pounds in 1993 (2). The role that this study plays in the overall project is to determine whether or not the properties of products fabricated from the reclaim exiting the overflow and underflow are significantly reduced from products formed from pure, uncontaminated resin. Also, increased strength, reduced density, or any other value adding properties discovered will be noted in an attempt to offset the cost of processing a recycled resin stream. In order ix) attribute property differences solely to contamination from the hydrocyclone separation employed (and not to reprocessing techniques or other forms of contamination) virgin resins were used in all the tests conducted. Contamination of each polymer by the other was studied in addition to studying contamination by the hollow glass microbubbles in each of the polymer matrices. CHAPTER II LITERATURE REVI EW Because the method of separation proposed here is a unique idea, no empirical research pertaining to the effects that low levels of spherical glass contaminants may have on polymer properties was found. There have, however, been many models proposed to estimate the effects that spherical fillers have on the properties of plastics. There is a subtle, yet important, distinction between these two classifications. When considering fillers, a larger percent loading is usually assumed. At the lower levels typical of contamination, these models may not hold up well, and may even predict a deterioration in polymer properties when an increase is actually observed (3). Also, the models based on spherical fillers do not address the problem of breakage. As we will see later, breakage — in some processes - is a very realistic concern. Although the effects of spherical inclusions as contaminants in polymers have not been addressed in the literature, PP and HDPE blends have been studied both as a contaminant in each other and also as blends from the viewpoint of enhancing polymer properties. SDI the latter case, however, most of the studies include a third agent used as a compatibilizer. 2.1 Effects of Glass Bubbles in the Polymer Matrices 2.1.1 Particle_uescriptign - In this literature review the microbubble contamination. will be referred to as a "filler", as all the prior research done on spherical inclusions in a polymer matrix has assumed their presence to be as such. In addition to the properties of the two components of a filled polymer and the volume fraction of the filler, ¢, other factors that affect the properties of the composite include the size, shape, and dispersion of the filler and also the degree of adhesion between the filler and the polymer matrix (4). Particle shape and size is often reflected through the maximum packing fraction, ¢m, of a filleru For random loose 'packing of spheres, ¢m is approximately 0.60 (3). In this review of the literature spherical inclusions with good dispersion and little or no adhesion to a non-polar crystalline matrix of high elongation will be considered whenever possible. When only more general models are available, these will be examined. Zhi all models the subscript "p" will denote "polymer" and the subscript "R" will denote the ratio of the composite property to the pure polymer property. 4) will represent volume fraction filler which will vary, while ¢m will represent the maximum packing fraction of the filler, which we approximated as 0.60 for random loose packing of spheres. 2.1.2 T n ' r er ' - Quantitative models for tensile strength at yield predict a decrease in strength with an increase in volume load of spherical fillers with little or no adhesion to the matrtx. This is because the continuity of the polymer matrix is displaced by what are essentially voids caused by a lack of adhesion. Several proposed models include: GR=(l—¢/¢m)A (3) o, =(1-1.21¢2’3) (5) where: 5R = the relative tensile strength at yield A = adhesion factor However, at low volume loadings, fillers may have a very different effect which, at present, cannot be quantified. Specifically, Katz and Milewski stated that non—polar polymers that orient with strain tend to have great improvements 'with. low' adhesion. particulate fillers (3). Because the filler exerts a 'viscous drag on the polymer it helps the polymer in molecular orientation, this is seen as an increase in yield strength at concentrations of approximately 0.05 relative filler volume (3). As with the tensile strength, elongation models predict a decrease in elongation with increased filler volumes. As the long flexible polymer chains capable of slipping past each other are replaced by a rigid filler, lower elongations are expected. In a model presented by Katz and Milewski, 10 relative elongation of the composite to the pure polymer is predicted by: ”3 ¢ R=1__ 3 8 [¢.) () This equation has limitations in that it.:hs not specified whether the elongation represents the elongation for tensile strength at yield or the elongation for tensile strength at break. Also, the degree of adhesion between the filler and the matrix is not incorporated into the equation nor is the equation assumed to be used for only good adhesion or only poor adhesion. As noted by Katz and Milewski, however, a high degree of adhesion between the filler and the polymer matrix will result in a greater loss of elongation then with lower levels of adhesion. Also noted is that at low volume loadings fillers may increase elongation at break (increased elongation at yield is rare). This phenomenon is explained by the orientation of the polymer molecules induced by the drag of the filler as explained above (3). Because an increase in modulus is the most characteristic effect of adding fillers (6), many models have been proposed to predict their effect upon addition to a polymer matrix. Qualitatively, the small inclusions restrict movement and deformation of the polymer molecules. This puts a mechanical restraint on the composite. Bigg (6) presented a variety of models, as follows: ll ER=1+26¢ ER =1+2.5q>+14.1¢2 G,¢/[(7— 5v)Gp +(8— 10v)G,]+¢/[15(1 —v)] R — Gp¢/[(7-5v)Gp +(8-10v)G,]+¢/[15(1-v)] 113 ER=1+_9_ (¢l¢"‘) In] 8 1-(¢/¢,.) 1 E = R (1-1.25¢)2 l A 1- BR- + Q) w=l+[7?l)¢ _1-w¢ m A = f (geometry) where: ER = relative modulus Gf = shear modulus of filler GP 2 shear modulus of polymer v = Poisson ratio of polymer 12 2.1.3 F l i - No quantitative description. of filler effects on flexural strength. were found, but in general fillers reduce the flexural strength of a composite as the percent filler increases (3). However, it is also noted that non-polar, rigid polymers with moderate elongations may show an initial increase in flexural strength before a drop occurs at approximately 0.35 relative filler volume (3). No model was found specifically for flexural modulus but, because of the large overlap in the literature of models for "tensile modulus", "modulus", and "any modulus", it seems that the effects of fillers on flexural modulus of the resulting composite should be similar to the effects on tensile modulus, and the same equations should be representative of this effect. 2.1.4 lmpag;_§t;enggn — No quantitative models exist to correlate the impact strength of a polymer with the inclusion of a filler (6). However, it can be expected that as the volume fraction of a rigid filler is increased, the impact strength of the composite will decrease (3). The rigid inclusions act as stress concentrators and also reduce the continuity of the polymer matrix. 2.1.5 v' - R ' - When in contact with certain chemicals (not solvents of the plastic) and while under stress, some plastics may exhibit failure through cracking when the presence of the stress alone or the chemical alone would not result in failure of 13 the specimen. Weak areas in the polymer due to amorphous regions, internal stresses, microvoids, or scratches are amplified when a stress is applied (7). This results in a strain in the area of amplified stress causing further reduction of density. The presence of the chemical can swell the amorphous regions and act as a plasticizer, allowing the plastic to deform more. As the plastic deforms its density lowers permitting greater attack by the chemical” The process is self accelerating, eventually resulting in brittle fracture of the polymer (7). Because it is an empirically defined parameter, no model is available to quantitatively relate the ESCR of a composite to the presence of the filler. However, because the presence of microbubbles may create microvoids and built-in stresses (through differences in the coefficient of thermal expansion), it is assumed that the ESCR would decrease with an increase in filler concentration (8). Also, the greater degree of irmerfacial contact between the polymer and the chemical (through the presence of microvoids) should help accelerate the growth of the crack. 2.1.6 w Ra - The flow rate of a polymer is an empirical property that tells the amount of a pmdymer (in grams per 10 minutes) that is extruded through a capillary at a given temperature and under a given load. If flow rate can be converted to viscosity (a fundamental polymer property), then there are models available to approximate the effect of fillers on the flow rate and viscosity of the l4 polymer (see Appendix B for the equations used to convert flow rate to viscosity). The properties of importance when considering filler effects on viscosity include filler concentration, size, aspect ratio and shape, stiffness, strength, and interaction between the filler and the polymer matrix (9). However, the models available simplify the situation and reduce the number of variables. Sheldon (9) presents two models to predict the relative viscosity of a polymer filled with spherical inclusions to that of the unfilled polymer: nR=l+2.5¢ r},,=1+2.5<1>-i-14.1(p2 Aspect ratio and shape for the spherical inclusions are accounted for by the constant 2.5 (Einstein's coefficient, kE) and relative filler volume is represented by ¢, but the other variables are not accounted for in these equations. The second equation accounts for particle interaction, while the first equation assumes no such interaction. Using the equation to first obtain the predicted viscosity, flow rate can then be calculated. Another point of importance is the increased wear that fillers may have upon processing equipment. 15 2.2 Effects of HDPE and PP Cross-Contamination In addition to the properties of the individual components of a polyblend and the volume fraction of each component, the most significant factors affecting polyblend properties are phase morphology and interphase adhesion (10). Phase morphology deals with the size and shape distribution of the disperse phase as well as the degree of crystallization for both phases. Interphase adhesion is affected by the miscibility of the components present in the blend. HDPE and PP are often considered immiscible when blended together (11). This is seen as definite phase separation between the two polymers. Most often a compatibilizer is added to blends of HDPE and PP to help reduce the size of the disperse phase and increase interphase adhesion or the components are copolymerized creating block or graft c0polymers which will have parts of each chain miscible in each polymer. The discussion here will be limited to blends of HDPE and PP with no compatibilizer added. cm: copolymerization intentionally induced. However, during the process of blending, the HDPE and PP may undergo free radical initiation causing grafting of some HDPE and PP molecules. This may be the result of thermal or mechanical degradation. 2.2.1 ££_§gn;aminatign_ofi_figfifi - Because PP tends to reduce the amount of crystallinity in HDPE, physical properties enhanced by crystallinity would be expected to drop with low amounts of PP. The incompatibility of the two l6 polymers would also be expected to deteriorate the mechanical properties of the HDPE. Several studies of blow molding grade HDPE contaminated with low levels of injection molding grade PP showed that levels of up to 5 weight percent PP in HDPE could be tolerated without significant reduction of properties (12,13). In the study by Christensen, et a1. (13), ESCR and flow rate were found to increase while density decreased with increasing PP as expected. Most other studies included a compatibilizer as a factor of the study, tested different properties, or used blow molded containers and cannot be compared to the results obtained in this study. 2.2.2 HDPE_Qontamination_of_PB - Lovinger and Williams (14) showed that when HDPE is present in PP it may induce a higher degree of crystallinity in the PP. This is because the HDPE nucleates more readily than the PP, causing more sites for the PP spherulites to grow from. This results in a larger number of spherulites with smaller diameters and an increase in overall crystallinity along with an increase in intercrystalline links. This increase in crystallinity results in a maximum observed tensile modulus at a composition of 80 percent PP and 20 percent HDPE (14). Other reports produced similar results at 90 and 75 percent PP (14). For tensile strength and elongation, however, the incompatibility between the polymers seems to dominate over the increased. crystallinity resulting' in. a jpolymer with inferior properties (14). The deterioration.ixijproperties 17 is much more pronounced for tensile strength and elongation at break than for these same properties at yield. The degree to which each phenomenon (increased crystallinity or incompatibility) will influence different polymer properties may be somewhat dependent upon the degree of strain involved in. determining that particular“ property; Typically, at higher strains incompatibility is dominant reducing the polymer properties, while at low strains increased crystallinity' dominates, enhancing‘ the jproperties of the polyblend (l4). HDPE is often added to PP to improve PP's low temperature impact resistance (15,16). However, no studies could be obtained which had empirical data showing the effects that HDPE has on Izod impact strength of PP. The model found to predict the viscosity of dilute polyblends, included, as a variable, the viscosity of the dispersed phase(in this case HDPE). Because the viscosity of HDPE could not be determined the polyblend viscosity could not be predicted (9). CHAPTER III MATERIALS AND METHODS 3.1 MATERIALS 3.1.1 Polymer Matrices - Blow molding grade high density polyethylene copolymer in powder form was obtained from Quantum - USI Division. The product code was LX 0100- 00 and it is typical of the HDPE used in bottles for detergents, motor oils, household chemicals, some juices, and other consumer products. Polypropylene was obtained from Himont. This injection molding grade Polypropylene homopolymer was also in powder form (product code 6501), and is typical of the PP used as spouts and closures for HDPE bottles. The small particle sizes of the powdered resins helped to ensure good mixing of the polymer with the fine microbubbles. 3.1.2 Hollow Glass Microbubbles - Scotchlite K20 glass bubbles (product code 70-0704-2956-1) were provided by 3M. These bubbles were composed of 97-100 percent soda lime borosilicate glass and less tflun1 3 percent amorphous silica. The average diameter CHE these ndcrobubbles (MB) was 31.11 microns with a wall thickness of 1.34 microns and a density of 0.260 g/cm3. l8 19 3 . 2 METHODS 3.2.1 Method of Sample Preparation - All materials were first dry mixed and then molded into tensile bars on a 75 ton Van Doren Injection Molder. Screw length and diameter were 28 in. and 1.5 in. respectively, with a screw speed of approximately 200 rpm. Zone 1, zone 2, and nozzle temperatures were set for 410° F and mold temperature was set for 60° F for all materials. Injection time was 13 seconds and hold time was 25 seconds at a pressure of 650 psi. The eighteen materials molded in this fashion included HDPE and PP each with glass microbubble contamination. The levels of microbubble contamination were 0%, .1%, 1%, 2%, and 5% by weight. Also, mixtures of each polymer in the other at levels of 0%, .5%, 1%, 2%, and 5% by weight were molded into tensile bars. These tensile bars were used as molded for testing tensile properties and further modified for use in all other tests. 3.2.2 Tensile Properties - In order to determine tensile strength at yield, tensile modulus of elasticity, and percent elongation at yield, ASTM D 638-86, "Standard Test Methods for TENSILE PROPERTIES OF PLASTICS", was followed. A United Test System testing machine (Model SEM- 20) was used to test tensile bars which were injection molded by Himont to conform to Type I specimens under the ASTM standard. A 1000 pound load cell was used with a 20 cross-head speed of 0.2 in./min. The gage length used was 2 in. and the grip separation was 4.5 in. The stress- strain curves displayed on the computer screen were used to determine the tensile modulus of elasticity. Because of the manual nature of modulus determination the results may be slightly subjective (two points were manually chosen to define the initial linear portion of the stress-strain curve, the computer then generated the line and calculated the modulus). The tensile strength at yield and percent elongation at yield were calculated through use of the load cell and a laser extensometer. 3.2.3 Flexural Properties - The flexural properties tested were tangent modulus of elasticity and stress at 5 percent strain. ASTM D 790-86, "Standard Test Methods for FLEXURAL PROPERTIES OF UNREINFORCED AND REINFORCED PLASTICS AND ELECTRICAL INSULATING MATERIALS", was followed. The samples tested were bars cut from the tensile bars molded by Himont. These cut tensile bars conformed ix: the requirements established in ASTM D 790-86 (2.5" x 0.5" x 0.125"). The same UTS testing machine used for tensile testing was set up to perform flexural testing. The support span-to-depth ratio used was 16 to l to give a support span of 2 in. A cross-head speed of 0.05 in./min. was used with a 20 pound load cell. The test was automatically terminated at a strain of 5 percent and the strength was recorded. The tangent modulus of elasticity was determined in the same 21 manner as the tensile modulus and likewise may be somewhat subjective. 3.2.4 Impact Strength - Impact strength was determined by following .ASTMZ D 256-84, "Standard Test Methods for IMPACT RESISTANCE OF PLASTICS AND ELECTRICAL INSULATING MATERIALS". Again samples were cut from the tensile bars molded by Himont. The samples conformed to the standard (2.5" x 0.5" x 0.125"). These cut samples were then notched according to the standard to a depth of 0.1 in. with a TMI Notching Cutter (Model TMI 22-05). A. TMI Impact Tester (Model 43-02) was then used with a 5 pound pendulum to determine the impact strength of the samples. 3.2.5 Environmental Stress-Crack Resistance - In order to determine the susceptibility of the samples to failure under conditions of stress in the presence of detergents, ASTM D 1693-70 ,"Standard Test Methods for ENVIRONMENTAL STRESS-CRACKING (N? ETHYLENE PLASTICS", was followed. Samples were prepared as follows: Tensile bars injection molded by Himont were used to ensure good distribution of the contaminant in the polymer matrix. These tensile bars were then compression molded using a Carver Laboratory Press (Model M). The samples were placed in a steel frame of inside dimensions 6.6" x 6.6" x 0.075". Sheets of aluminum were placed on either side of the plastic tensile bars and then covered with thicker sheets of steel. This "sandwich" was placed into the press and processed as follows: 22 HDPE + contaminants: 10 minutes at: . 25,000 psi . 150° c cooled to 55° C without manually adjusting pressure PP + contaminants: 15 minutes at: . 30,000 psi o 1900 C cooled to 65° C without manually adjusting pressure The sheets of plastic molded from the process described above were left to cool and were cut into rectangular samples within 24 hours of their removal from the mold with a Jarmac circular saw. The samples conformed to condition B from. the .ASTM standard (1.5" x 0.5" x 0.075"). These samples were nicked to a depth between 0.012 and 0.015 inches using a nicking jig made by Custom Scientific Instruments, Inc. They were then bent, placed in sample holders and. immersed. in aa test tube filled. with full- strength Igepal C0-630, a non-ionic detergent. The bending clamp, transfer tool, and specimen holders were also made by C51, Inc. The stoppered test tubes were then placed in a constant temperature bath at 50° C and checked regularly for failure (every day for PP, every hour for HDPE up to 34 hours, then every four hours up to 50 hours, and finally at 80 hours). 3.2.6 Flow Rate - To help understand the effects that contamination may have on the reprocessing of the materials, 23 flow rates were determined following the standard, ASTM E) 1238-86, "Standard. Test. Methods for FLOW ‘RATES OF THERMOPLASTICS BY EXTRUSION PLASTOMETER". Tensile bars molded by Himont were cut into small pieces to be charged into the cylinder of the plastometer. Procedure A was used for the HDPE + contaminants because the method indicated that Procedure B should only be used for materials having flow rates greater than 0.5 grams per 10 minutes. Procedure B was used for the PP + contaminants so that the viscosity could be determined, this is not possible with Procedure A. The machine used in both procedures was a Ray Ran Melt Flow Indexer (Model 2 'A'). Procedure A — HDPE + contaminants: After heating the cylinder with the piston and die in place at 190° C for 15 minutes, 3 grams of material were charged into the cylinder and the piston was weighted with a 2.16 kg weight (Condition 190/2.16). At the appropriate time the extrudate was cut and a timer was started. After six minutes the extrudate was cut and, after cooling, it was weighed. The weight was multiplied by 1.67 (10 min/6 min) to determine the flow rate in g/10 min. Procedure B - PP + contaminants: After heating the cylinder, piston, and die at 230° C for 15 minutes, approximately 7 grams of material were charged into the cylinder and a 2.16 kg weight was placed on top of the piston (Condition 230/2.16). The extrudate was cut just as 24 the automatic timer was triggered at the appropriate time and position. When the piston traveled 6.35 nun the timer was automatically stopped and the time in seconds it took for the piston to travel 6.35 mm was recorded. The factor from Table 5 (ASTM D 1238-86) was then used to determine flow rate in g/10 min for the samples from the following equation: Flow rate, g/10 min = F/t where: F = factor from table 5 (ASTM D 1238—86) t time of piston travel for length L, 5 And the factor, F, is derived from: F = 427 x L x d where: length of calibrated piston travel, cm density of resin at test temperature, g/cm3 mean of areas of piston and cylinder x 600 (600 = 60 sec/min x 10 min) i:" II II 427 3.2.7 Differential Scanning Calorimetry - To determine the heat of fusion of the samples, ASTM D 3417-83, "Standard Test Methods for HEATS OF FUSION AND CRYSTALLIZATION OF POLYMERS BY THERMAL ANALYSIS", was followed. Approximately 12.5 milligrams of each sample was cut from the tensile bars molded by Himont. The samples were cut from the same area in each tensile bar to ensure uniform cooling rates. Also, 25 visual inspection helped to ensure homogeneity in such a small sample. The tests were performed. in a Du Pont Instruments 910 Differential Scanning Calorimeter (Model 910001-908). The cell was purged with nitrogen at 50 mL/min gas flow rate and the specimen (sealed inside an aluminum DSC pan) was placed in the cell along with an empty DSC pan as a known reference sample. The cell was heated from ambient temperature to 180° C for HDPE + contaminants and to 200° C for PP + contaminants at a rate of 10° C/min. A base line was then constructed by connecting the two points at which the melting endotherm deviated from the straight base line. After manually choosing these two points (a slightly subjective task) a TA Instruments Thermal Analyst 2200 System displayed the DSC curve along with the heat of fusion in J/g. 3.2.8 Density - In order to determine the density of each of the samples, ASTM D 792-66, "Standard Test Methods for SPECIFIC GRAVITY AND DENSITY OF PLASTICS BY DISPLACEMENT", was followed. Rectangular samples were cut from the tensile bars molded by Himont. These samples were weighed in air. The samples were then weighed in deionized water along with a wire and sinker used to hold and sink the materials which were lighter than water. Next, the wire and sinker were weighed. while immersed in water. Specific gravity was computed as follows: Sp gr 23/23° C = a/(a + w — b) 26 where: m It apparent weight of specimen, without wire or sinker, in air b = apparent weight of specimen and sinker completely immersed and of the wire partially immersed in water apparent weight of totally immersed sinker and partially immersed wire 2 ll And the density of the samples was computed as follows: 23C D , g/cm° = sp gr 23/23° C x 0.9975 3.2.9 Scanning Electron Microscopy - In an attempt to characterize the structure of the polymer and contaminant blends and thus to help explain the results of the physical and mechanical testing, scanning electron micrographs were taken of the fracture surfaces of impact-tested samples of HDPE and PP, including pure samples, material with 5 percent glass microbubble contamination, and material with 5 percent polymeric contamination. .Almo, micrographs were taken of the glass ndcrobubbles alone t1) compare surface textures. Samples were coated with approximately 188 Angstroms of gold using a Polaron SEM Coating Machine. Then a Jeol JSM-T330 Scanning Microscope was used to view and photograph the samples. 3.2.10 Data Analysis - In all the graphs of this study the error bars represent the 95 percent confidence intervals for the population means of each treatment. If the lines between two treatments do not overlap, it can be said with 27 95 percent confidence that they are from populations with different means. Analysis of variance (ANOVA) at a confidence level of 95 percent was also run on each group of data to determine whether differences in means were attributed to variances within the same population or to treatments of individual groups (Appendix C). The literature value located in the bottom left corner of the graphs, gives typical properties of the uncontaminated polymer (either reported in the literature or given as product specifications). inns contamination level is given as weight percent and may be converted to approximate volume fraction through Table 1. Tables of raw data are contained in Appendix A for a more precise representation of the results shown in graphical form. CHAPTER IV RESULTS AND DISCUSSION 4.1 Density 4.1.1 Microbubble Contamination of the Polymer Matrices - A significant increase in density was found for all levels of :microbubble contamination. in. both. polymer matrices except for the 0.1 percent contamination levels in both polymers. This increase in density was quite different from. what was expected and led to the conclusion that significant breakage of the microbubbles had occurred during the injection molding process. This conclusion was later verified through the use of a scanning electron microscope. Figure 1 shows the calculated density of the resins at 100 percent breakage and at zero percent breakage along with the density actually measured for each resin. Calculations were performed to determine the amount of microbubble breakage in each of the polymer matrices. The glass was assumed to have a density of 2.5 g/cm3. With this assumption microbubble breakage was found to be an average of 88.7 percent and 92.3 percent for HDPE and PP, respectively. Unfortunately, this may not be typical of what would be seen in a closed loop (bottle to bottle) recycling application. Blow molding operations usually run at pressures up to 150 psi (17), whereas the injection molding processes used in this study had a holding pressure of 650 psi. The product information 28 29 HDPE/MB HDPE/PP 0 98 0.951 1 0.96 ria’fi 0.95 3 T t 1.0 94 . 1.0 949.: m m : L 5 ‘ \\ E \K 3, 0 92 V 3 0.948 1 \ d >. § §’ 0.9 K\ ‘3 0.947 3 - cn \ in 5 \ c ‘ c s 8 0.88 \\\\\ 8 0.946 f L 0.86 0 94s 5 0 ‘Lit Value: 0.952 g/cm3 \ 0.84 1 . i I 1 0.944 5 ' 0 1 2 3 4 5 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) PP/MB PP/HDPE 0 94 0.909 [ AII"1E===:F:fi—1 0.908 3 0.9 q d .1 "E \ T; 3 ‘ . 0.9075 ‘ 4’ £50.88 \\ E; L r,/’r :; ~ \\\k\. :; 0.907 j ’/,4’< ,zr’r ' H 0.86 7 *4 IE ‘ i‘\q ‘§ 0.9065 1”” 3 0.84 \\- 2 | ‘\\ 0 906- ‘ l 0-32 0.9055 j 0 ‘ Lit Value: 0.903 g/cm3 \+ 3 l 0.8 1 j, I 1 0.905 7 0 1 2 3 4 5 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) -I- Observed Density + Observed Density + Density at 100% Breakage _._ Rule of Mixures + Den31ty at 0% Breakage Figure 1: Effects of Contamination on Polymer Density. 30 pamphlet for the microbubbles indicated that at 500 psi a maximum of 20 percent breakage should occur. Using the averages of 88.7 and 92.3 percent microbubble breakage for HDPE and PP (see Appendix B), the volume fraction of broken and intact glass bubbles was calculated. The total volume fraction of glass (broken plus intact) was used in the model equations for spherical fillers (Table 21). Models for spherical fillers were used because the glass spheres were much easier to characterize than the broken pieces of glass. Also, the volume fraction of glass spheres was very similar to the volume fraction of the broken glass. Spheres comprised approximately 55.1 percent of the glass volume in HDPE and about 44.5 percent in PP (Table 1). 4.1.2 HDPE/PP Cross-Contamination - The only significant change in density from the uncontaminated resins seen for the various HDPE/PP blends was at the 5 percent contamination level of’ PP in ZHDPE where a decrease in density was observed (Figure 1). All other blends produced densities within the 95 percent confidence intervals for the uncontaminated resins. However, the trends seen were expected, and the observed densities are plotted against the expected densities based on a simple rule of mixtures in Figure 1. The greater than expected loss in density seen by HDPE may be indicative of a disruption of HDPE crystallinity caused by methyl groups present on the PP chains. The shift of observed density from the expected density in PP may be due to an error in the estimation of the population mean for 31 Table 1: Characterization of Glass in the Polymer Matrices. Average of 88.7% Microbubble Weight Percent Glass Breaks 8 for HDPE 0.1 1 2 5 Density of Contaminated Resin ll 0.949738 0.95186 0.954186 0.960961 [Wolurne Fraction Broken Bubbles || 0.000337 0.003344 0.006838 0.016236“ Volume Fratclion Intact Bubbles II 0.000412 0.004096 0.008131_ 0.019888 otal Volume Fraction Glass | 0.000749 0.00744 0.01477 0.036124 Intact Bubbles Comprise 55.1% of the Glass in HDPE. Average of 92.3% Microbubble WeighFPflement Glass Breaks efor PP 0.1 1 2 5 :“ Density of Contaminated Resin ll 0.90676 0.909976 0.91351 0.923862 IlVolume Fraction Broken Bubbles || 0.000334 0.003326 0.006813 0.016242“ “ET/£10m Fratction intact Bubbles || 0.000268 0.002668 0.005305 0.013029“ otal Volume Fraction Glass “0.000803 0.005995 0.011918 0.029271.“ Intact Bubbles Compn‘se 44.5% of the Glass in PP. 32 uncontaminated PP. It can be observed from Figure 1 that if the uncontaminated PP had a density of approximately 0.9056 the two lines would coincide very well. The large standard deviation for the pure PP helps to validate this theory. 4.2 Differential Scanning Calorimetry 4.2.1 Microbubble Contamination of tflua Polymer Matrices - Figure 2 shows the observed heat of melting at various contamination levels plotted against the line for a simple rule of mixtures. As can be seen for both HDPE and PP the heat of melting is below that predicted by a rule of mixtures. This may signify a disruption of the crystalline regions for both the HDPE and the PP. 4.2.2 PP Contamination of HDPE - The trend seen for the HDPE/PP mixture are similar to the trends above (Figure 2). The sharp initial decrease in. heat of' melting is probably indicative of a disruption in the crystalline structure of EHHTL At higher levels of contamination the heat of melting of the blend begins to approach the weighted average of the two components. 4.2.3 HDPE Contamination of PP - A very different effect was seen with the PP/HDPE blends. Low levels of HDPE actually increased the heat of melting of the PP by more than the weighted average of the two components (Figure 2). This phenomenon may be explained by the high degree of nucleation caused by the presence of HDPE. Once nucleation is initiated by the HDPE the PP spherulites continue to grow at these sites. Without the presence of the HDPE, 33 HDPE/MB HDPE/PP 200 200 198 198 i 196 196% A A J § 194 . g" 194 if \ 192 { \\\\\ 192 \ “~_ g g \ 1 \ ‘ .H .H -i 1‘. 190 .1”. 190 0) Q) = = . l l \ u 188 .,_, 188 ° . \ ° . \I \ E 186 \ :3 186 x :1: ‘ :1: ‘ 184 \ 184 182 \ / l - 182 l . , l . 180 I 4 ; 180 E 0 l 2 3 4 5 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) PP/MB PP/HDPE 94 99 = 93 98 X 92 g / 3 91 3 97 / \ \ E 3 90 3 96 f s s / c; 89 : E .H -.-1 95 I:- 3 88 f. // a) 2) g / z z 94 _ H-t H E / ° 86 ° 93 3 u 1; : /\ / t8 : g 85 g 92% 84 ‘ 83 91 7 82 90E i 0 1 2 3 4 5 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) + Observed Heat of Melting + Rule of Mixtures Figure 2: Effect of Contamination on Heat of Melting. 34 nucleation in PP is much slower resulting in fewer (albeit, larger) spherulites (18). 4.3 Scanning Electron Microscopy 4.3.1 Microbubble Contamination of HDPE - Micrographs of the fracture surfaces of impact tested HDPE with 5 percent microbubble contamination showed good dispersion of the microbubbles in the polymer matrix (Figure 3). Figures 4 and 5 verify that many of the glass bubbles had indeed suffered. breakage during' processing. Figure 5 shows a typical microbubble/HDPE interface with poor wetting of the glass surface and 'very little adhesion. between the two components. The surface of this glass bubble can be compared to a glass bubble that was not in contact with the polymer (Figure 6). The broken bubbles resulted in many shapes and sizes which made them hard to characterize in model equations. 4.3.2 Microbubble Contamination of PP - Again micrographs of the fracture surfaces were taken and fairly good dispersion of the microbubbles was observed. A higher percentage of breakage seemed quite apparent from the micrographs (Figure 7), which was in accordance with what was calculated from changes in density. Polypropylene seemed to wet the glass surface much better than HDPE, but adhesion was still very minimal (Figure 8). The degree of polymer/glass adhesion and the percent microbubble breakage (through particle shape factors and percent volume loading) are important factors when considering the effects of Figure 3: Micrograph of HDPE with 5% Microbubble Contamination. .4. - ‘ 1BKU X1a009 Figure 4: Micrograph Showing Typical Amount of Microbubble Breakage. 36 a h...’ .h . 18KU 911.1386 000803 Figure 5: Typical HDPE/Microbubble Interface. IGKU XIaSBB Figure 6: Typical Surface of a Microbubble. 37 100P~m 000006 Figure 7: Micrograph of PP with 5% Microbubble Contamination. l'h )/ 10KU XI’BBB 10H": 000001 _ \ - 1.7 ' \ Figure 8: Typical PP/Microbubble Interface. 38 Figure 9: Micrograph of PP with 5% HDPE Contamination. 39 microbubble contamination (n1 mechanical auxi physical properties of these polymers. 4.3.3 HDPE/PP Cross-Contamination - SEM study of HDPE with small amounts of PP showed poor dispersion of the PP. Most areas looked similar to the uncontaminated HDPE while other areas contained large agglomerations of PP. Micrographs of PP with HDPE showed good dispersion, yet definite phase separation of the two polymers (Figure 9). The blend seemed to be anisotropic with elongation of the HDPE occurring in the direction of flow during the injection molding process. 4.4 Tensile Properties 4.4.1 Microbubble Contamination of HDPE - A general increase was seen in tensile strength at yield and in modulus of elasticity, with a sharp increase from the uncontaminated resin to the 0.1 percent level of contamination. After an initial increase at the 0.1 percent contamination level, a decrease was seen in elongation at yield. Figures 10, 11, and 12 show these trends quantitatively and the significance of various levels of contamination can be determined from these graphs. Models studied describing the effects of spherical inclusions with little adhesion to the polymer matrix predicted three trends. An increase in modulus with a decrease in both the tensile strength and elongation. Not represented in the equation by Katz and Milewski (3), however, was a qualitative description of why tensile strength may 26. 25 24 Tensile Strength at Yield (MPa) 21. Level of Contamination 23. 22. HDPE/MB 27 / 26 111 1111 1 / .5 ‘ 25 ' .5 24 5 23 \. l \ Lit. Value: 18-29 W8 1 l i l 1 2 3 4 5 (weight percent) 5 1111 1111 L111 1111 111L 22 111 S 0 PP/MB U H U! 1L —' ' Ezm. 04 E 30 ‘c H . .329J5‘ l - r ‘ | 3 29: F \ 32m.5‘ : 4 i”. \ 28 ' a , ‘\\ L 27.5 in .. 5 27‘ \~ 9‘ ‘ Lit. Value: 35 MPa 2 6 ' 5 l T l l 0 1 2 3 4 5 Level of Contamination (weight percent) + Observed Tensile Strength —0— Model Figure 10: 4O Tensile Strength at Yield (MPa) Level Tensile Strength at Yield (MPa) Level of Contamination 26. 26. 26. 25. 25. 25. 25. 25. 24. 24. 24. 33. 32. 31 30. 29. HDPE/PP 34 5 33 5 32 .5 31 5 30 S 29 of 1 2 Contamination 3 4 5 (weight percent) PP/HDPE 1114 1111 1111 1111 1111 1111 1 1111 1111 0 2 l 3 4 5 (weight percent) Effects of Contamination on Tensile Strength at yield. 41 initially increase for filler volumes up to 5 percent (recall, however, that the 55 weight percent filler corresponds to a 0.039 volume fraction of spheres and broken glass in HDPE). This phenomenon is explained by a work hardening effect on the polymer. Because there is little adhesion to the glass, the HDPE may draw around the particles and orient. This is induced by a viscous drag on the polymer by the glass. Work hardening is usually promoted. by' high aspect ratio fillers because of their greater surface area:volume ratio (3) and the strengthening in this case may be due largely to the broken glass in the matrix. A higher increase in modulus was observed than predicted and the loss of elongation was not as dramatic as the model had. predicted (Figures 12 and 11). Tensile strength experienced an increase which no model could quantitatively predict. All the tensile properties studied for the contaminated resins were higher than predicted by the models. Perhaps this is due to the presence of the broken bubbles (due to their shape), for which no model found could estimate the effects, along with microbubbles, on the polymer matrix. These significant differences, according to Analysis of variance at a confidence level of 95 percent, are attributed to the treatment of groups rather than to variations within the same group. 3.4.2 IMicrobubble Contamination of E¥>- Analysis of variance showed no significant changes in tensile strength HDPE/MB 15. mfll‘” 14. 13.5 12.5 11.5 10.5 \ Elongation at Yield (percent) 111111111 111111111 111111111 111111111 ’41 0 1 2 3 4 5 Level of Contamination (weight percent) PP/MB 10.54 T A 10f g i o 9.5 1 o 23 E! 9 U 8.5 H \ up (D 3:: 81a \\ 1., Z w 7.5: g : “\\L 44 7 7 u : g : 8 6.5 _ B 63 \\» E Lit Value: 12 percent 5'5 l l l I 0 1 2 3 4 5 Level of Contamination (weight percent) .gp. Observed Elongation + Model Figure 11: 42 Elongation at Yield (percent) H H i-' H N 00 ab Ln 0 H o H c H c U1 U U1 uh U1 U" U" %# 1111 1 l H N 11. U‘ H H 1111 1111 1111 1111-11 10.5 Level 10 Level of Contamination HDPE/PP 111 ‘ 7T\\1 I Elongation at Yield (percent) \0 \l of 1 2 3 4 5 Contamination (weight percent) PP/HDPE LL 0 1 2 3 4 5 (weight percent) Effect of Contamination on Elongation at yield. 43 HDPE/MB HDPE/PP 1300 q 1050 7; 3 4- 7; . 1 94 2 Q4 .1 5 3 E 1000 3‘ 1200 j 3 j g : , ,II’ 3 950‘ 3 1100 ‘ ,I, . /l H : up H .. a: . m j I 900 B : 3 ’/,//’ S 1000 j " 3 1 L 3 ‘ I”’J 3 esoii‘ 'c 'c g / g 2 ' v/ 1 1 3 900 3 ~ .,., .,., 800 . in U) .4 c: 3 c: . o , w e-o ~ . [-1 1 : Lit Value: 620-895 We . 800 l I I 750 0 1 2 3 4 5 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) PP/MB PP/HDPE 1800 2200 I; 3 In? '1 o‘ E on 0‘ .1 5 1700 3 a I ._ 3. 5. 2000 H 1600 f H . ‘\\\\ U : U H I H . :3 1500 7* ‘ \- g g. _ H g 1800 0.4 1400 : I“ . O / O . in , U1 :3 1300 1 / 3 1600 r-i . r—i .- 3 5 Al :1 'U a “U 9 1200 ; g a, a, 1400 H : r-i 1'1 1100 j H 2 : 2 ,2: Lit. Value: 1138-1551 MP8 .2 1000 ‘ I I I 1200 I 0 1 2 3 4 S 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) + Observed Modulus —.— Model Figure 12: Effects of Contamination on Tensile Modulus of Elasticity. 44 due to the level of microbubble contamination up to 5 percent. However, variations in elongation and modulus are attributed to the presence of the microbubbles. Elongation is decreased and modulus is increased, which is expected (Figures 11 and 12). The same models were used here as for the effects of microbubble contamination on HDPE. 4.4.3 PP Contamination of HDPE — ANOVA showed no significant change for elongation at yield and tensile modulus. Tensile strength at yield had a quite erratic curve (Figure 10), suggesting, perhaps, that like elongation and modulus, tensile strength of HDPE is also unaffected by low levels of PP contamination. 4.4.4 HDPE Contamination of F¥>- Tensile Strength at yield and modulus of PP both increased up to 2 percent HDPE contamination and then began to fall (Figures 10 and 12). This initial increase is in accordance with prior research done by Iovinger and Williams (14), Noel and Carley (15), and Deanin and Sansone (16), and is attributed to an increase in the crystallinity of PP as well as an increase in links between the separate spherulites of the PP. However, in these prior works, maxima were observed at 20 percent, 25 percent, and 10 percent, respectively. Elongation at yield increased initially (Figure 11), which was quite unexpected because of the incompatibility of the two polymers. 45 4.5 Flexural Properties 4.5.1 Microbubble Contamination of HDPE - Both strength at 5 percent strain and the flexural modulus showed a sharp increase from the uncontaminated resin to the resin with 0.001 weight fraction microbubbles. 11 rather linear increase of lesser slope ‘was seen. for* higher levels of contamination (Figures 13 and 14). The model predicted well the slope of the line after the initial increase, but could not account for the dramatic initial increase, quantitatively or qualitatively. No quantitative model was found for flexural strength, however, the results can be explained. by the viscous drag applied to the molecular chains by the microbubbles on the outer surface where strain occurs (3) and perhaps by an increase in compressive strength at the inner surface where compression occurs. The variations were attributed to the treatment of different groups according to ANOVA, at a 95 percent confidence level. 4.5.2 Microbubble Contamination of PP - Strength at 5 percent strain showed no relation to ndcrobubble inclusion according to ANOVA. Flexural modulus showed an initial decrease followed by a fairly steady increase (Figure 14). The model predicts an increase of flexural modulus, however, the observed modulus increased with a steeper slope. 4.5.3 PP Contamination of HDPE - The flexural properties of HDPE were enhanced by the addition of PP (Figure 12). Because it is well documented that PP disrupts HDPE crystallinity this effect was not expected. 46 HDPE/MB HDPE/PP 27 - 25 5 2&5 ‘ ' - .. : / - i" to . I; 25 8 26‘ / 9 i c: 1 / c. . .,., . / ,,_, - g 25.5 . :3 24.5- 33 . u . U} 1 U) i- - 25 ‘ :3 : g d / u 2 u 1 l “ 2L5 m 24‘ . . 5 5 4 U“ at .. 5 24 E . 3 3.3 23.5 . in ‘n 5 § 23.5 t : ‘I 23 f l 23 4' i 0 1 2 3 4 S 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) PP/MB PP/HDPE 46.5 _ 47.5 g l 46 i a E j E 46.5 3 2 45.5 z E s- V 1 a 45: £45.53 H z I. ‘I- K H 3 1 a: 44.5 g . / a: g :0 " are 44.5 2 m 44 m ‘ U 0 U (U m a. 4: 43.5 g 43.5 5 'fi 14 u . 9 l 9 2 a, 43 0) 3L 3 , :4n.5; l ”in.5: “ 3 I 42 41.5 ‘ I . T 0 1 2 3 4 S 0 1 2 3 4 5 Level of Contamination (weight percent) Level of Contamination (weight percent) Figure 13: Effects of Contamination on Flexural Strength at 5% Strain. 47 HDPE/MB 1010 4)- 990 ’6 ‘P 970 A” g //’ ._ g . 950 H 5 _ / § / 93o . 7‘0 / H 52 m 910 r-‘I h - 890 1 (Lit. Value: 1240 MPa 870 I I I I 0 1 2 3 4 5 Level of Contamination (weight percent) PP/MB 1750 1700 ‘ g I :31: 1650 . /) m E 1/// 3 1600 ,, g a: '0 -4 g I T ‘1, 1550 :3 * / g ‘T 4,2’V’ll1 § 1500 . .r' H In 1450 j I Lit. Value: 1650 MP8 1400 7 I I 0 1 2 3 4 5 Level of Contamination (weight percent) + Observed Modulus + Model Figure 14: Elasticity. HDPE/PP 1030 1010 990 970 950 I I *””‘ 930 1 910 l 890 870¥ I 0 1 2 3 4 5 Level of Contamination (weight percent) Flexural Modulus (MPa) PP/HDPE 1700 1111 1650 1111 1600 1550 111L111J_1 1500 Flexural Modulus (MPa) 1450 1111A 1400 I I O 1 2 3 4 5 Level of Contamination (weight percent) Effects of Contamination on Flexural Modulus of 48 4.5.4 HDPE Contamination of PP - The flexural properties of PP decreased with the addition of HDPE (Figures 13 and 14). Apparently, the inferior properties of HDPE combined with the incompatibility of HDPE and PP dominated over the increased crystallinity of the PP. 4.6 Impact Strength 4.6.1 Contaminated HDPE - The HDPE samples used for impact testing all resulted in non-breaks with the heaviest weight available (30 pounds). Although the results cannot be quantified, it (uni be concluded that sufficient embrittlement to cause failure in HDPE did not occur with microbubble or PP contamination. 4.6.2 Microbubble Contamination of PP - A.very sharp decrease in impact strength was seen at a microbubble contamination level of 0.1 weight percent and then a slight increase was seen (Figure 15). This sharp initial drop seems to be consistent with prior research done by Bigg (6). In the study' by' Bigg, PP impact strength. was shown to decrease sharply with the addition of 0.01 volume fraction filler and then level off toward a flatter slope. Whereas fibers (as used by Bigg) usually tend to distribute the impact stress over a larger area perpendicular to the force of impact (3), this is not the case for spheres. Thus, greater reduction of strength was observed for the glass spheres than was reported by Bigg for fibers. 49 PP/MB Izod Impact Strength (J/m) Lit. Value: 42.7 Ma 0 1 2 3 4 5 Level of Contamination (weight percent) PP/HDPE 15 14 _a_u—+— 13 12 11 Izod Impact Strength (J/m) \0 / ~11 ‘ Lit. Value: 42.7 J/m I I 0 l 2 3 4 Level of Contamination (weight percent) (1‘ Figure 15: Effects of Contamination on Izod Impact Strength. 50 4.6.3 HDPE Contamination of PP - Although HDPE is often added to PP to improve PP impact resistance, very different results were observed in this study. A sharp decline in strength was seen at a contamination level of only .5 percent HDPE (Figure 15). At higher levels impact resistance did start to improve but at 5 percent contamination impact strength was still below that of the uncontaminated polymer. 4.7 Environmental Stress-Crack Resistance 4.7.1 Microbubble Contamination of HDPE - In order to find the 50 percent failure point, F50, for each level of microbubble contamination, the failure time of each sample within the group was plotted against the percentage of samples that had failed up to that point (see Appendix). Logarithmic probability graph paper was used and the intersection of the best line fit with the 50 percent failure line gave the F50. A general decrease in ESCR was seen for the HDPE/MB samples (Figure 16) which was expected because of molded in stresses and a higher area of contact between the HDPE and the detergent. 4.7.2 PP Contamination of HDPE — PP contamination reduced the ESCR of HDPE (Figure 16). This is in contrast to the study by Christensen, et al. (13), in which PP improved HDPE's resiStance to stress—cracking. Figure 16: F50 (hours) ESCR, F50 (hours) ESCR, 51 HDPE/MB 125 1_. 1 1A 100 1 1 L 1 75 1 1 L 1 50 25 Mi 1- 0 l I I 0 1 2 3 4 5 Level of Contamination (weight percent) HDPE/PP 125 100 1 1 1 1 75 U1 C 25 1 1 1 n1 l I I I I I 0 1 2 3 4 5 Level of Contamination (weight percent) Effects of Contamination on Environmental Stress-Crack Resistance. 52 4.7.3 Contaminated PP — All of the PP samples tested either broke upon bending or did not break at all, up to 800 hours. No F50 can be reported for these groups, but it can be said that sufficient deterioration of PP was not imparted by the contamination to cause failure below 800 hours. Unrelated to the ESCR but worth mentioning is that failure upon bending increased with increased filler concentration (see Appendix B). Flow Rate 4.8.1 Microbubble Contamination CA? the Polymer Matrices - Because the index length was not measured when determining HDPE flow rate, the flow rate cannot be converted to viscosity (see Appendix B), and no model can quantitatively predict the effect of a filler on HDPE flow rate. After an initial increase, the flow rate of HDPE declined as microbubble contamination increased (Figure 17). This decrease in flow rate is expected because of a disruption of the flow pattern and perhaps a hindrance to molecular orientation. The method of flow rate measurement used for PP allows the conversion to Viscosity and a model for viscosity can thus be used to predict the effects of fillers on PP flow rate. PP flow rate was reduced with an increase of ndcrobubble contamination, however, the effect is not as drastic as predicted by the Einstein model Figure 17). 53 4.8.2 HDPE/PP Cross-Contamination - The general effects seen on flow rates for the cross-contaminated polymers follows what would be expected intuitively. As the less viscous PP is added to HDPE more material flows through the capillary in a given time period (Figure 17). Increasing amounts of the viscous HDPE tend to decrease the flow rate of the polyblend melt. HDPE/MB 0.2971fl 0. 28 .27 .26 .25 .24 ?low Rate (g/10 min) * \ ‘1' 4. Ln. Value: 0.35 gm min I I I 0 1 2 3 4 5 Level of Contamination (weight percent) .23 .22 0.21 PP/MB 5. 5. E -r-4 E 2 5 53 ‘\\\ I‘\\\ 3 \ \I. m 5 m 3 . \\\\\ n o E‘. 4.8 <\ Lit. Value: 4 g/ 10 min. 4.6 I I I 0 1 2 3 4 5 Level of Contamination (weight percent) + Observed Flow Rate —.— Model Figure 17: 54 Flow Rate (g/10 min) .295 .285 .275 .265 .255 .245 .235 0.225 Level Flow Rate (g/lO mun) HDPE/PP U" U‘ U1 U‘ U1 U" U" U" tn 0 o e o a e o 0 .5 H CDKDU‘ 0 1 2 3 4 5 of Contamination (weight percent) PP/HDPE 9 8‘4 7 6‘/\ 5=/\ 31 2 I I\\fi#“‘- O 1 2 3 4 5 Level of Contamination (weight percent) Effects of Contamination on Flow Rate. CHAPTER V SUMMARY AND CONCLUSIONS In the processes used in this study, sample molding resulted in a high degree of microbubble breakage. This led to an increased density and may have decreased crystallinity in both polymer matrices. Polypropylene present in HDPE seemed to decrease the percent crystallinity of HDPE, while HDPE seemed to increase the percent crystallinity of PP. Many properties were slightly enhanced by various contamination levels (however, no added value is thought to accompany these increases), while most other properties tested stayed within typical ranges for the uncontaminated polymers. The main problem encountered for HDPE was a decrease in ESCR with the addition of either microbubbles or PP. This could result in bottle failure if closed-loop recycling is sought with products that accelerate environmental stress-cracking (such as oils and detergents). For PP the effect of contamination on impact strength is of greatest concern. The addition of 0.1 percent microbubbles or 0.5 percent HDPE (by weight) both result in a loss of impact strength of over 50 percent. This sharp initial effect at low levels of contamination is very characteristic of most of the properties studied. The changes in flow rates caused by contamination were not drastic enough to render the polymer unprocessable, yet 55 56 there are several possible concerns. When changes in processing' parameters are necessary, finding the optimum conditions can prove to be quite costly and time consuming. Also, the presence of the abrasive microbubbles could shorten the life of some of the processing equipment. In an attempt to resolve the few problems encountered in this study, further experimentation is recommended in several areas. In order to improve the rheology of the polymer blends and thus the ultimate polymer properties, processing parameters along with appropriate compatibilizer use may be investigated. In an attempt to improve the ESCR of HDPE and the impact strength of PP with microbubble contamination, surface treatment of the glass may be considered. Also, in View of implementing a cost effective and comprehensive curbside recycling program (concentrating still on the light fraction) several more separation processes must be investigated. This ambitious goal includes the separation of: -Expanded polystyrene from the other light constituents present in common household waste (PP, HDPE, LDPE). —Low density polyethylene from HDPE and PP. -Blow molding grade PP from injection molding grade PP -Blow molding grade HDPE from injection molding grade HDPE. -Blow molding grade HDPE homopolymer from blow molding grade HDPE copolymer. LIST 05' REFERENCES 10. 11. LIST OF REFERENCES . Vane, L.M., and Rodriguez, F., "Selective Dissolution: Multi-Solvent, Low Pressure Solution Process for Resource Recovery from Mixed Post-Consumer Platics," SPE RETEC, Recy. Tech. of 1990's, 100-106, Nov. 29-30, Chicago, IL. "Key Markets Post Solid Gains in 1993," Modern Plastics, 73-81, January 1994. Katz and Milewski, Handbook of Fillers for Plastics, Van Nostrand Reinhold Company, New York 1987. . Manson, J.A. and Sperling, L.H., Polymer Blends and Composites, Plenum Press, New York, 1976. Nicolais, L. and Nicodemo, L., "Strength of Particulate Composites," Polymer Engineering and Science, Vol.13, No. 6, 469, November 1973. Bigg, D.M., "Mechanical Properties of Particulate Filled Polymers," Polymer Composites, Vol. 8, No.2, 115—122, April 1987. Levy, S. and Dubois, J.H., Plastics Product Design Engineering Handbook, Chapman and Hall, New York 1984. Fraser, R., "Environmental Stress-Cracking of Plastics," Plastics and Polymers, 102-103, June 1975. Sheldon, R.P., Composite Polymeric.Materials, Applied Science Publishers, New York 1982. Folks, M.J. and Hope, P.S., Polymer Blends and Alloys, Blackie Academic & Professional, New York 1993. Paul, D.R. and Newman, 8., Polymer Blends, Academic Press, New York 1978. 57 12. 13. 14. 15. 16. 17. 18. 58 Harris, M.G., "The Physical Properties and Effects of Polymeric Contamination on Post-Consumer Recycled High Density Polyethylene," Polyolefins VII RETEC, 671-678, Houston, TX February 27,1991. Christensen, R.E., Austin R.G., and Clayton D.M., "Polyolefin Blend Recycle Studies," SPE ANTEC, 794-798, Detroit, MI, May 3-7, 1992. Lovinger, A.J. and Williams, M.L., "Tensile Properties and Morphology of Blends of Polyethylene and Polypropylene", JOurnal of Applied Polymer Science, Vol. 25, 1703-1713, 1980. Kausch, H.H., Polymer Fracture, Sperling-Verlag, New York, 1978. Kinney, Enginnering Properties and Applications of Plastics, John Wiley and Sons, Inc. New York 1957. Lee, N.C., Plastic Blow.MOlding Handbook, Van Nostrand Reinhold, New York 1990. Yang, D., Zhang, B., Yang, Y., Fang, 2., Sun, G., and Feng, 2., "Morphology and Properties of Blends of Polypropylene with Ethylene-Propylene Rubber", Polymer Engineering and Science, Vol. 24, No. 8, 612-617, June 1984. APPENDIX A 59 Table 2: Effects of Contamination on Polymer Density. (g/cm3) HDPE Level of MB Contamination (wt °/o) Sample No. 0 0.1 1 2 5 1 0.9493 0.9497 0.9526 0.9580 0.9643 2 0.9498 0.9493 0.9520 0.9577 0.9645 3 0.9495 0.9494 0.951 2 0.9579 0.9609%| Average 0.9495 0.9495 0.9519 0.9579 0.9632 H St. Dev. 0.0003 0.0002 0.0007 0.0001 HDPE Levl Emmi %) Sample No. 0 0.5 1 =_====== _ __ - 1 0.9493 0.9494 0.9491 0.9488 0.9453 2 0.9498 0.9479 0.9498 0.9473 0.9467 3 0.9495 0.9497 0.9492 0.9468 0.9462 Average 0.9496 0.9488 0.9495 0.9471 0.9464 St. Dev. 0.0002 0.0013 0.0004 0.0003 0.0003 PP Level of MB Contamination (wt °/e) Sample No. 0 0.1 1 2 5 1 0.9070 0.9069 0.9102 0.9154 0.9214 2 0.9063 0.9068 0.9108 0.9139 0.9219 3 0.9061 0.9065 0.9104 0.9145 0.9216 Average 0.9064 0.9067 0.9105 0.9146 0.9216 oomxs 0£m07 cuxmz I I LevelWH—oPE Contamination (wt °/e) 05 1 2 5 cum57 osmxs mem (MKW4 2 II 0.9063 0.9057 0.9061 0.9062 0.9072 3 0.9061 0.9058 0.9058 0.9064 0.9077 Average II 0.9064 0.9057 0.9061 0.9064 0.9074 StDmL fl QGXB cuxm1 0£m04 00mm: 0mxm 60 Table 3: Effects of Contamination on Heat of Melting. (J/g) Level of MB Contamination (wt %) 0.1 1 2 5 1 * WWT2 180.0 2 192.7 193.4 183.8 193.0 185.0 Average 195.8 189.7 181.3 186.1 182.5 St. Dev. 4.3 5.3 3.6 9.8 3.5 1 HDPE Level of PP Contaminatiol'i-(Tvt: °/.) 1 Sample No. 0 0.5 1 2 5 1 198.8 188.4 _I=88.4 182?? 2 192.7 181.3 199.6 185.9 185.5 Average 195.8 184.9 194.0 184.4 183.9 81. Dev. 4.3 5.0 7.9 2.2 2.2 PP Level of MB Contamination (wt °/e) Sample No. 0 0.1 1 2 5 1 .7673 86.79 86.26 91 .42 79.28 2 96.1 1 94.64 ----- 0.91 90.71 Average 91.42 90.72 86.26 46.17 85.00 II St. Dev.__ 6.63 5.55 ----- 64.00 8.08 _ PP Level of HDPE Contamination (wt %) Sample No. 0 0.5 1 2 5 1 86.73 85.46 95.30 85.45 96.47 2 96.11 98.80 94.93 96.49 101.1 Average 91.42 92.13 95.12 90.97 98.79 St. Dev. 6.63 9.43 0.26 7.81 3.27 beau—5::— l 61 Table 4: Effects of Contamination on Tensile Strength at Yield. (PSH HDPE ll Level of MB Contamination (wt %) Sample No. 0 0.1 1 2 5 II 1 3620 3626 3624 3661 3864% 2 3642 3691 3654 3714 3872 3 3615 3709 3702 3651 3920 4 3652 3757 3679 3709 3897 5 3645 3742 3693 3686 3872 6 3650 um --- ---- ---- Average 3637 3705 3670 3684 3885 St. Dev. 16 51 32 28 23 HDPE Level of PP Contamination (wt °/o) Sample No. 0 0.5 1 2 5 1 3620 3581 3719 3601 3590 2 3642 3772 3661 3626 3677 3 3615 3714 3722 3621 3680 4 3652 3767 3689 3634 3642 5 3645 3797 3757 3631 3645 6 3650 ---- --- --- ---- Average 3637 3726 3710 3623 3647 Lit. Dev. 16 87 36 13 36 PP ll Level of MB Contamination (wt %) Sample No. 0 0.1 1 2 5 1 4402 4269 4204 4292 4327 2 4424 4382 4387 4334 4314 3 4392 4497 4322 4372 4445 4 4397 4412 4422 4357 4537 5 4384 4424 4349 4357 4314 Average 4400 4397 4337 4342 4387 St. Dev. 15 83 83 31 100 PP Level of HDPE Contamination (wt %) Sample No. 0 0.5 1 2 5 1 ‘ 4402 4550 4665 4617 4457 2 4424 4710 4673 4678 4547 3 4392 4610 4567 4630 4495 4 4397 4638 4567 4703 4600 5 4384 4648 4735 4622 4592 Average 4400 4631 4641 4650 4538 St. Dev. 15 58 73 38 62 62 Table 5: Effects of Contamination on Elongation at Yield. (Percent) ’ I 13.942 16.620 18.940 . II 2 12.852 13.512 12.990 12.399 10.641 3 13.273 13.353 13.272 12.675 10.770 4 12.713 12.517 12.802 12.526 11.322 5 12.595 11.385 13.296 12.890 11.294 6 12.939 ---- --- --.« -.... II Average 13.052 13.477 13.260 12.724 10.974 0.493 1.949 “v1 of Cotamnanionu—( * 0.5 1 2 5 13.942 ' 2102 16.281 13.953 IL 2 12.852 13.936 12.737 14.778 12.379 II 3 13.273 14.551 11.935 12.960 14.148 4 12.713 12.821 12.805 13.221 12.730 5 12.595 13.505 13.289 12.762 13.797 II 6 Average St. Dev. PP Sample No. 1 2 3 4 5 Average 81. Dev. PP Level of HDPE Contamination (wt °/e) Sample No. 0 0.5 1 2 5 1 10.131 10.767 10.774 9.757 10.071 II 2 8.915 10.036 10.295 9.217 9.553 I 3 8.692 10.016 10.227 9.400 10.347 II 4 9.255 10.382 10.693 9.255 9.961 5 I 9.693 9.942 10.479 9.390 9.256 II Average II 9.337 10.229 10.494 9.404 9.838 I II 51. Dev. II 0.583 0.346= 0.239 0.213 0.433 I 63 Table 6: Effects of Contamination on Tensile Modulus of Elasticity. (PSD HDPE Level of MB Contamination (wt %) Sample No. 0 0.1 1 2 5 1 7' 126300 128805 128377 154254 190610“ 2 128741 132262 155853 156749 160186 3 129658 165780 124209 156983 167270 4 129548 140788 130736 168730 170438 5 121795 137287 132302 161061 160127 Average 127208 140984 134295 159555 169726 St. Dev. 3315 14603 12431 5680 12508 «E HDPE Level of PP Contamination (wt %) Sample No. 0 0.5 1 2 3| — 126300 119218 126723 113750 135571 128741 137211 136015 124909 135921 129658 133022 143846 116326 149369 129548 153692 135666 130526 128406 121795 149216 156164 129592 115690 127208 138472 139683 123021 132991 3315 13685 11028 7646 12286 _ Level of MB Contamination (wt %) I 0 OJ 1 2 5 _ 222751 192243 185311 141531 246689 2 192692 198597 220084 189219 220783 3 190267 178576 201602 175083 223781 4 200688 206632 181524 187756 225101 5 196106 184160 196920 171622 254556 Average 200501 192042 197088 173042 234182 13040 11173 15249 19218 15344 Level of HDPE Contamination (wt %) 0 05 1 2 5 H 1 222751 203469 211847 297700 261116 2 192692 207063 218253 304502 330880 3 190267 187999 199899 310483 241640 4 200688 192989 202255 314810 254727 5 196106 218357 198121 298641 234191 Average 200501 201975 206075 305227 26451 1 St. Dev. 13040 11964 8626 7416 38583 Jul-sf * 64 Table 7: Effects of Contamination on Flexural Strength at 5% Strain. (PSI) Leveoi MB ntaminati (wt I°) 0.1 1 3352 ' 3636 I 3567 I I I652 3849 2 3402 3522 3638 3641 3863 3 3359 3518 3586 3667 3856 4 3348 3494 3558 3638 3874 5 3329 3525 3570 3626 3691 Average 3358 3539 3584 3645 3867 St. Dev. __, 27 __ __ -- 32 15 16 Level of PP Contamination (wt %) 0.5 1 1 3325 I 3530 3584 ' 3480 I 2 3402 3522 3501 3461 3548 3 3359 3603 3622 3473 3534 4 3348 3558 3650 3445 3518 5 3329 3556 3449 3452 3546 Average 3358 3554 3561 3462 3537 St. Dev. 27 32 84 14 12 l PP Level of MB Contamination (wt %) I Sample No. 0 0.1 1 2 5 I — 1 6601 6193 6417 6443 6722 2 6270 6181 6308 6534 6373 3 6313 6174 6534 6342 6782 4 6176 6306 6376 6407 5 6549 6246 6299 6448 Average 6395 6255 6362 6399 6546 St. Dev. 180 165 1 14 92 191 PP Level of HDPE Contamination (wt %) Sample No. 0 0.5 1 2 5 6601 6450 6277 II 6157 6061 6270 7079 6123 6210 6368 6313 6507 6270 6325 6251 --- 6496 6248 6140 6094 ---- 6279 6267 6159 6304 Average 6395 6562 6237 6198 6216 51. Dev. _ 180 303 65 76 133 UIhOJN-P ' .21“ '- 11} 65 Table 8: Effects of Contamination on Flexural Modulus of Elasticity. (PSI) Level of MB Contamination (wt %) 0 0A 1 2 5 127900 ' 138I700 " "142000 2 128000 134500 140500 138100 141600 3 126300 136700 128700 140100 141200 4 127700 134500 138400 141000 141800 5 126400 134000 135600 136000 148200 Average 127260 135680 136760 137720 142960 St. Dev. ._838 1985 4943 3090 2944 _ Level of PP Contamination (wt °/e) jI 0 05 1 2 5 1 127900 146700 137200 139300 130800 2 128000 138800 146200 139900 146500 3 126300 136800 137400 137700 145700 4 127700 136100 139200 132200 144500 5 126400 132600 137100 135800 133500 Average 127260 138200 139420 136980 140200 St. Dev. 838 5252 3886 3109 7444 Level of MB Contamination (wt °/e) SmnmeNo. 0 01 1 2 5 1 214200 208100 214%. 227600 247000 2 209100 209200 217700 232200 233800 3 216800 210000 231600 223500 253300 4 «an 211100 220000 224500 232600 5 «u- 226000 216600 221500 237500 Average 213367 212880 221460 225860 240840 St. Dev. 391 7 741 6 5973 41 73 8970 PP Level of HDPE Contamination (wt %) Sample No. 0 0.5 1 2 5 1 214200 225300 207200 205600 208100 2 209100 249800 204700 209600 219500 3 216800 229000 210600 215100 216100 4 ----- 221300 209800 208100 208300 5 ----- 212000 211400 207400 218200 Average 213367 227480 208740 209160 214040 St. Dev. 3917 13992 2755 3618 5468 E u an: 66 Table 9: Effects of Contamination on Izod Impact Strength. Impact Strength (FT.LBJ|N.) and Type of Break (6=Complete Break. h=Hinge Break) I PP Level 61 MB_CII—Ientm'nination (wt °/o) Sample No. 0 0.1 1 2 5 1 0.2436 0.0936 0.117h 0.1186 0.151 6 2 0.2446 0.1046 0.1066 0.1046 0.1226 3 0.288 c 0.093 c 0.187 6 0.185 6 0.166 6 4 0.2026 0.1056 0.1636 0.1166 0.1196 5 0.236 0.1056 0.1286 0.1276 0.1176 6 0.256 c 0.092 c 0.117 6 0.104 h 0.107 6 7 0.231 c 0.092 6 0.115 h 0.128 6 0.129 c 8 0.117h 0.1046 0.1286 0.1276 0.1196 9 0.2246 0.1156 0.1626 0.1056 0.1196 10 0.2366 0.0936 0.1636 0.1166 0.1086 11 0.216 0.1056 0.1526 0.1166 0.1676 12 0.253 6 --- --- 0.127 6 - Total Average 0.229 0.100 0.140 0.123 0.129 St. Dev. 0.042 0.008 0.026 0.022 0.022 Hinge Average 0.117 --- 0.116 0.104 «- St. Dev. «- --- 0.001 --- --- Complete Average 0.239 0.100 0.145 0.124 0.129 St. Dev. 0.024 0.008 0.027 0.022 0.022 PP Level of HDPE Contamination (wt %) Sample No. 0 0.5 1 2 5 1 0.243 6 0.127 6 0.173 6 0.162 c 0.197 6 2 0.2446 0.1396 0.1396 0.1846 0.1636 3 0.2886 0.1156 0.1526 0.1276 0.1286 4 0.2026 0.1156 0.1496 0.1516 0.146 5 0.236 0.1036 0.103h 0.156 0.1536 6 0.2656 0.1076 0.1266 0.1156 0.1636 7 0.231 c 0.105 c 0.093 6 0.104 h 0.162 6 8 0.117h 0.1166 0.1056 0.1266 0.1386 9 0.224 6 0.094 c 0.139 6 0.058 h 0.221 6 10 0.2366 0.1846 0.09311 0.11611 0.1886 11 0.216 0.1156 0.116h 0.1276 0.1166 12 0.253 c --- - --- --- Total Average 0.229 0.118 0.126 0.129 0.161 St. Dev. 0.042 0.019 0.026 0.033 0.031 Hinge Average 0.117 --- 0.104 0.093 --- St. Dev. --- --- 0.012 0.031 --- Average 0.239 0.118 0.135 0.143 0.161 St. Dev. 0.024 0.019 0.026 0.023 0.031 67 Table 10: Effects of Contamination on Environmental Stress-Crack Resistance. Hours to Failure 0 0A 1 2 8 6 11 11 17 12 13 11 18 13 15 13 20 13 23 14 22 13 25 17 24 14 26 18 50 16 32 19 80 17 33 26 80 21 80 27 80 80 80 30 F50 30.0 15.3 26.3 17.6 HDPE Level of PP Contamination (wt %) No. 0 05 1 2 8 4 5 17 10 14 18 10 15 20 10 16 22 10 16 24 12 13 19 50 17 13 25 80 20 15 26 80 38 16 29 80 42 26 30 30.0 13.3 1 1.6 17.2 tomxlmmrxwro—s .5 o romxloamewm-e .4. o 68 Table 11: Effects of Contamination on Flow Rate. (g/10 min) 7 I IPE I Level of MB Contamon (wt %) 1 2 5 I I I I I. I .0755 0.281 0.224 0.229 2 0.249 0.271 0.267 0.232 0.220 3 0.244 0.301 0.261 0.220 0.20 Average 0.250 0.284 0.270 0.225 0.223 0.016 0.010 0.006 Level of PP Contamination (wt °/e) 0.5 1 2 5 0.279 0.235 0.269 0.274 2 0.249 0.286 0.237 0.264 0.281 3 0.244 0.296 0.271 0.257 0.296 Average 0.250 0.287 0.248 0.263 0.284 51. Dev. . 0.007 0.009 0.020 £11006 0.011 I Level of MB Contamination (wt %) Sample No. I 0 0.1 1 2 5 ‘ 1 ' 5.31 5.38 5.32 5.05 5.00 2 5.57 5.43 5.46 5.56 5.12 3 5.38 5.36 5.41 5.42 4.91 Average 5.42 5.39 5.40 5.34 5.01 St. Dev. 0.13 0.04 0.07 0.26 0.11 PP Level of HDPE Conthination (wt °/.) Sample No. 0 0.5 1 2 5 ' 1 5.31 5.85 5.39 4.93 4.80 ' 2 5.57 5.71 5.35 5.35 5.10 3 5.38 5.81 5.39 5.28 5.28 Average 5.42 5.79 5.38 5.19 5.06 St. Deva 0.13 0.07 0.02 0.23 0.24 APPENDIX B 69 Table 12: Calculated Microbubble Breakage in HDPE and PP. I” HDPE/MB I30 Level of Glass MicrIoquDtllés Aided (wt 4.); 0.1 Density of Resin 0. 9495 0. 9495 0. 9519 0. 9579 0. 95632 olume Fraction Intact Bubbles II 0.0000 0.000691 0.004049 0.003819 0.017333 IIVolume Fraction Broken Bubbles II 0.0000 0.000308 0.003349 0.007116 0.016544 Percenta e of Bubbles Broken 81.05 88.83 94.71 90.17 otal Volume Load Glass 0.0000 0.000999 0.007398 0.010934 0.033877 Average Percentage of Microbubble Breakage in HDPE 88.69 PPIMB Level of Glass Microbubble Added (wt %) 0 OJ 1 2 5 Density of Resin 0.9064 0.9067 0.9105 0.9146 0.9216 IIVolume Fraction Intact Bubbles 0.0000 0.000341 0.002027 0.003979 0.015729 IIVolume Fraction Broken Bubbles 0.0000 0.000327 0.003395 0.00676 0.015918 Percenta e of Bubbles Broken 90.21 94.15 94.23 90.68 otal Volume Load Glass 0.0000 0.000668 0.005422 0.010738 0.031647 Average Percentage of Microbubble Breakage in PP 92.32 7O Hmummflndhhmflndummn+n 8/ Icahn o-o ILII—-:-o b Figure 19: Graphical Method Used to Determine Environmental Stress-Crack Resistance 71 Figure 13: Number of PP Samples to Break Upon Bending When Testing for Environmental Stress-Crack Resistance. Level of MB Contamination 0 0.1 1 2 5 No. of Failures U n Bendi 1 6 6 1 8 Level of PP Contamination O 0.5 1 2 5 No. of Failures U on Bendin 1 1 1 3 3 where, HF‘H'JUH'U 72 . . Shear Stress V18 cosrty = smeRMB Shear Stress = , 21!!! L 2 SNMIRM£=12§%21 nr'r = Load on ram, kg Orifice radius, mm Barrel radius, mm Orifice length, mm Index length, mm Index extrusion time, sec. Figure 19: Equations Used to Convert Flow Rate to Viscosity. APPENDIX C 73 APPENDIX C Analysis of Variance Anova: Densi’g - HDPE/MB Smnmew' ” Groups Count Sum Average Variance Pure HDPE 3 2.848552 0.949517 6.29E-08 HDPE/.1%MB 3 2.848409 0.94947‘ 5.56E-08 HDPE/1%MB 3 2.855837 0.951946i 4.99E-07 HDPE/2%MB 3 2.873616 0.957872 1.65E-08 HDPE/5%MB 3 2.889679 0.963226 4.05E-06 ANOVA Source of Variation I 35 df MS F P-value Fcrit Between Groups 0.000432 4 0.000108 115.3218 2.53E-08 3.47805 Within Groups 9.37E-06 10 9.37E-07 Total 0.000442”' 14 74 Anova: Density - HDPE/PP I Summary I Groups Count Sum Average Variance Pure HDPE 3 2.848552 0.949517 6.29E-08 HDPE/.5%PP 3 2.846998 0.948999 9.51 E-07 HDPE/1%PP 3 2.848124 0.949375 1.41E-07 HDPE/2%PP 3 2.842985 0.947662 1.11E-06 HDPE/5%PP 3 2.838158 0.946053 4.96E-07 ANOVA Source of Variation I 53 df MS F P-value F crit Between Groups 2.57E-05 4 6.44E-06 11.65937 0.000878 3.47805 Within Groups 5.52E-06 10 5525-07 Total 3.13E-05 14 75 Anova: Density - PPIMB Summary Groups Count Sum AveEge Variance Pure PP 3 2.719345 0.906448 2.485-07 PP/.1%MB 3 2.720183 0.906728 3.63E-08 PP/%MB 3 2.731359 0.910453 8.46E-08 PP/2%MB 3 2.743789 0.914596 5E-07 PP/5%MB 3 2.764944 0.921648 5.29E-08 ANOVA Source at Variation 7 I 55 df MS F P-value F crit Between Groups 0.000483 4 0.000121 654.7633 4.76E-12 3.47805 Within Grou 3 1845-06 10 1.84E-07 Total 0.000484 14 76 Anova: Density tor PP/HDPE Summary I . Groups Count Sum Average Variance Pure PP 3 2.719345 0.906448 2.48E-07 PP/.5%HDPE 3 2.717153 0.905718 3.5E-09 I PP/%HDPE 3 2.718342 0.906114 1.41 E-07 I PP/2%HDPE 3 2.719199 0.9064 3.26E-08 I PP/5%l-IDPE 3 2.722335 0.907445 4.16E-08 I I ANOVA Source of Variation I 33 df MS F P-value Fcrit Between Groups 4.925-06 4 1235-06 13.17115 0.000538 3.47805 Within Groups 9.33E-07 10 9.33E-08 Total 5.85E-06 14 I 7 7 Anova: Tensile Strength at Yield - HDPE/MB Summary Groups Count Sum Average Variance Pure HDPE 6 21824 3637.333 251.0667 HDPE/.1%MB 5 18525 3705 2631.5 HDPE/1%MB 5 18352 3670.4 1001.3 HDPE/2%MB 5 18421 3684.2 786.7 HDPE/5%MB 5 19425 3885 537 ANOVA Source of Variation I 35 df Ms F P-value Fcn't Between Groups 1958106 4 48952.64 48.76377 2.46E-10 2.840096 Within Groups 21081.33 21 1003.873 Total 2168919 25 7 8 Anova: Tensile Strength at Yield - HDPE/PP I I I l Summary I I I l Groups Count Sum Average Variance 7 Pure HDPE 6 21824 3637.333 251.0667 HDPE/.5%PP 5 18631 3726.2 7501.7 HDPE/1%PP 5 18548 3709.6 1318.8 l HDPE/2%PP 5 181 13 3622.6 170.3 a, HDPE/5%PP 5 18234 3646.8 1316.7 ‘ ANOVA Source 01 Variation I 33 df MS F P-value Fcn't Between Groups 43772.21 4 10943.05 5.409021 0.003732 2.840096 Within Groups 42485.33 21 2023.1 1 1 Total 86257.54 25 79 Anova: Tensile Strength at Yield - PPIMB Summary I I Groups Count Sum Average Variance Pure PP 5 21999 4399.8 227.2 PP/.1%MB 5 21984 4396.8 6890.7 ; PP/%MB 5 21684 4336.8 6945.7 : PP/2%MB 5 21712 4342.4 978.3 I PP/5%MB 5 21937 4387.4 10030.3 I I l ANOVA ‘ Source of Variation SS df MS F P-value F crit Between Groups 18690.96 4 4672.74 0.931857 0.465463 2.866081 Within Groups 1002888 20 5014.44 Total 1 18979.8 24 8O Anova: Tensile Strength at Yield - PP/HDPE I . I Summary I I Groups Count Sum Average Variance I i Pure PP 5 21999 4399.8 227.2 I PP/.5%HDPE 5 23156 4631.2 3395.2 PP/%HDPE 5 23207 4641.4 5346.8 PP/2%HDPE 5 23250 4650 1466.5 PP/5%HDPE 5 2691 4538.2 3812.7 g I ANOVA I g I s : Source of Variation I 33 df MS F P-value Fcrit Between Grou s 226001 4 56500.26 19.82688 9.86E-07 2.866081 Within Groups 56993.6 20 2849.68 Total 2829946 24 I“! 8 1 Anova: Elongation at Yield - HDPE/MB I I I I 1 Summary I Groups Count Sum Avera a Variance Pure HDPE 6 78.314 13.05233 0.243496 HDP El. 1 %MB 5 67.387 1 3.4774 3.798278 I HDPE/1%MB 5 66.3 13.26 0.186626 HDPE/2%MB 5 63.62 12.724 0.084905 I HDPE/5%MB 5 54.868 10.9736 0.098424 ‘ ANOVA Source of Variation I as df MS F P-value Fcrit Between Groups 20.23654 4 5.059135 5.938477 0.00233 2.840096 Within G roups 17.89042 21 0.851925 Total 38.12695 25 82 Anova: Elmation at Yield - HDPE/PP Summary Groups Count Sum Averagg Variance Pure HDPE 6 78.314 13.05233 0.243496 HDPE/.5%PP 5 70.703 14.1406 1.354005 HDPE/1%PP 5 62.868 12.5736 0.305565 HDPE/2%PP 5 70.002 14.0004 2.257332 HDPE/5%PP 5 67.007 13.4014 0.628563 ANOVA Source of Variation I as df MS F P-value F crit Between Groups 8.675384 4 2.168846 2.347799 0.087556 2.840096 Within Groups 19.39934 21 0.923778 Total 28.07473 25 83 Anova: Elongation at Yield - PPIMB Summary Groups Count Sum Average Variance Pure PP 5 46.686 9.3372 0.339501 PP/.1%MB 5 47.088 9.4176 0.168605 PP/%MB 5 47.927 9.5854 0.714608 PP/2%MB 5 43.434 8.6868 0.112675 PP/5%MB 5 39.031 7.8062 0.290086 ANOVA Source of Variation SS df MS F P-value F crit Between Groups 10.74237 4 2.685594 8.26095 0.00042 2.866081 Within Grou 3 6.501901 20 0.325095 Total 17.24428 24 ‘F—l 84 Anova: Elongation at Yield - PP/HDPE Summary Groups Count Sum Average Variance Pure PP 5 46.686 9.3372 0.339501 PP/.5%HDPE 5 51.143 10.2286 0.11946 PP/%HDPE 5 52.468 10.4936 0.057279 PP/2%HDPE 5 47.019 9.4038 0.045498 PP/5%HDPE 5 49.188 9.8376 0.187112 ANOVA Source of Variation I 83 df MS F P-value Fcrit Between Groups 5.096274 4 1.274069 8.506842 0.000353 2.866081 Within Groups 2.995397 20 0.14977 Total 8.091671 24 85 Anova: Tensile Modulus of Elasticity - HDPE/MB Summary Groups Count Sum Average Variance Pure HDPE 5 636042 1272084 10988305 HDPE/.1%MB 5 704922 140984.4 2.13E+08 HDPE/1%MB 5 671477 1342954 1.55E+08 HDPE/2%MB 5 797777 1595554 32259520 HDPE/5%MB 5 848631 1697262 1.56E+08 ANOVA Source of Variation I 88 df MS F P-value F crit Between Groups 6.31 E+09 4 1.58E+09 13.89189 1.41E-05 2.866081 Within Groups 2.27E+09 20 1.13E+08 Total 8.58E+09 24 86 Anova: Tensile Modulus of Elasticity - HDPE/PP Summary Groups Count Sum Average Variance Pure HDPE 5 636042 1272084 10988305 HDPE/.5%PP 5 692359 1384718 1.87E+08 HDPE/1%PP 5 698414 1396828 1.22E+08 HDPE/2%PP 5 615103 1230206 58460519 HDPE/5%PP 5 664957 132991.4 1.51 E+08 ANOVA Source of Variation I as df Ms F P-value Fcrit Between Groups 1.03E+09 4 2.56E+08 2.421849 0.082093 2.866081 Within Groups 2.12E+09 20 1.06E+08 Total 3.14E+09 24 87 Anova: Tensile Modulus of Elasticity - PPIMB l Summary I . Groups Count Sum Average Variance I Pure PP 5 1002504 2005008 1.7E+08 PP/.1%MB 5 960208 1920416 125E408 PP/%MB 5 985441 1970882 2.33E+08 PP/2%MB 5 865211 1730422 3.69E+08 PP/5%MB 5 1170910 234182 2.35E+08 . ANOVA I Source of Variation I SS df MS F P-value F crit Between Groups 9.83E+09 4 2.46E+09 10.84882 7.66E-05 2.866081 Within Grou s 4.53E+09 20 226E+08 Total 1.44E+10 24 88 Anova: Tensile Modulus of Elasticity - PP/HDPE Summary Grougs Count Sum Average Variance Pure PP 5 1002504 2005008 1.7E+08 PP/.5%HDPE 5 1009877 2019754 1.43E+08 PP/%HDPE 5 1030375 206075 74405290 PP/2%HDPE 5 1526136 3052272 55004044 PP/5%HDPE 5 1322554 2645108 1.49E+09 ANOVA Source of Variation I 88 df MS F P-value F crit Between Groups 4.46E+10 4 1.11E+10 28.86172 4.69E—08 2.866081 Within Groups 7.72E-1-09 20 3.86E+08 Total 5.23E+10 24 89 Anova: Flexural Strengh at 5% Strain - HDPE/MB Summary Groups Count Sum Average Variance Pure HDPE 5 16790 3358 728.5 HDPE/.1%MB 5 17695 3539 3090 HDPE/1%MB 5 17919 3583.8 1020.2 HDPE/2%MB 5 18224 3644.8 239.7 HDPE/576MB 5 19333 3866.6 271.3 ANOVA Source of Variation SS df MS F P-value F crit Between Groups 6780894 4 1695228 158.441 7.68E-15 2.866081 Within Groups 21398.8 20 1069.94 Total 6994882 24 9O Anova: Flexural Strength at 5% Strain - HDPE/PP Summary 4 Groups Count Sum Average Variance Pure HDPE 5 16790 3358 728.5 HDPE/.5%PP 5 17769 3553.8 1005.2 HDPE/1%PP 5 17806 3561 .2 7078.7 HDPE/2%PP 5 17311 3462.2 208.7 HDPE/5%PP 5 17683 3536.6 142.8 ANOVA Source of Variation I as df MS F P-value F crit Between Groups 1470662 4 36766.54 20.06053 9E-07 2.866081 Within Groups 36655.6 20 1832.78 Total 1837218 24 91 Anova: Flexural Stren h at 5% Strain - PPIMB Summary Groups Count Sum Average Variance Pure PP 3 19184 6394.667 32392.33 PP/.1%MB 5 31273 6254.6 27139.3 PP/%MB 5 31811 6362.2 13029.2 PP/2%MB 5 31994 6398.8 8484.7 PP/5%MB 5 32732 6546.4 36381.3 ANOVA Source of Variation I SS df MS F P-value F crit Between Groups 2182628 4 54565.57 2.4256 0.085717 2.927749 Within Grou 5 4049227 18 22495.7 Total 623185 22 W 11.5%» ran" a all-w 92 Anova: Flexural Strflh at 5% Strain - PP/HDPE Summary Groups Count Sum Average Variance Pure PP 3 19184 6394.667 32392.33 PP/.5%HDPE 5 3281 1 6562.2 91825.7 PP/%HDPE 5 31 185 6237 4176.5 PP/2%HDPE 5 30991 6198.2 5709.7 PP/5%HDPE 5 31078 6215.6 17745.3 ANOVA Source of Variation I 69 df MS F P-value Fcrit Between Groups 47261 1.8 4 1 18153 3.919463 0.018471 2.927749 Within Groups 5426135 18 30145.19 Total 1015225 22 93 Anova: Flexural Modulus of Elasticity - HDPE/MB I Summary . Groups Count Sum Average Variance Pure HDPE 5 636300 127260 703000 HDPE/.1%MB 5 678400 135680 3942000 HDPE/1%MB 5 683800 136760 24433000 HDPE/2%MB 5 688600 137720 9547000 HDPE/5%MB 5 714800 142960 8668000 ANOVA i I Source of Variation I 88 df MS F P-value F crit Between Groups 6.42E+08 4 1.61 E+08 16.9738 3.25E-06 2.866081 Within Groups 1.89E+08 20 9458600 Total 8.31 E+08 24 94 Anova: Flexural Modulus of Elasticity - HDPE/PP Summary 7 Groups Count Sum Average Variance Pure HDPE 5 636300 127260 703000 HDPE/.5%PP 5 691000 138200 27585000 HDPE/1%PP 5 697100 139420 15102000 HDPE/2%PP 5 684900 136980 9667000 HDPE/5%PP 5 701000 140200 55420000 ANOVA Source of Variation I 35 df MS F P-value Fcrit Between Groups 5.53E+08 4 1.38E+08 6.376679 0.001776 2.866081 Within Groups 4.34E+08 20 21695400 Total 9.87E+08 24 95 Anova: Flexural Modulus of Elasticity - PPIMB Summary Groups Count Sum Average Variance Pure PP 3 640100 213366] 15343333 PP/.1%MB 5 1 137400 227480 1 .96E+08 PP/%MB 5 1043700 208740 7588000 PP/2%MB 5 1045800 209160 1 3093000 PP/5%MB 5 1070200 214040 29898000 ANOVA I I Source of Variation I I ' ss df MS F P-value F crit Between Groups 1.16E+09 4 2.89E+08 5.116109 0.00623 2.927749 Within Grou s 1.02E+09 18 56448370 Total 2.17E+09 22 96 Anova: Flexural Modulus of Elasticity - PP/HDPE Summary Groups Count Sum Average Variance Pure PP 3 640100 2133667 15343333 PPI5%HDPE 5 1 137400 227480 1 .96E+08 PPP/oHDPE 5 1043700 208740 7588000 PP/2%HDPE 5 1045800 209160 1 3093000 PPI5%HDPE 5 1070200 214040 29898000 ANOVA Source of Variation I as df MS F P-value Fcrit Between Groups 1.16E+09 4 2.89E+08 5.116109 0.00623 2.927749 Within Groups 1.02E+09 18 56448370 Total 2.17E+09 22 97 Anova: Flow Rate - HDPE/MB Summary Grows Count Sum Avera a Variance Pure HDPE 3' 0.75 0.25 4.3E-05 HDPE/.1%MB 3 0.851 0.283667 0.000241 HDPE/1%MB 3 0.809 0.269667 0.000105 HDPE/2%MB 3: 0.676 0.25333 3.73E-05 HDPE/5%MB 3% 0.669 0.223 2.7E-05 4 ANOVA Source of Variation I 38 df Ms F P-value Fcrit Between Groups 0.008571 4 0.002143 23.59949 4.45E-05 3.47805 Within Groups 0.000908 10 9.08E-05 Total 0.009479 14 98 Anova: Flow Rate - HDPE/PP Summary Groups Count Sum Average Variance Pure HDPE 3 0.75 0.25 4.3E-05 HDPE/.5%PP 3 0.861 0.287 7.3E-05 HDPE/1%PP 3 0.743 0.247667 0.000409 HDPE/2%PP 3 0.79 0.263333 3.63E-05 HDPE/5%PP 3 0.851 0.283667 0.000126 ANOVA Source of Variation J 33 df Ms F P-value Fcrit Between Groups 0.004055 4 0.001014 7.367975 0.00494 3.47805 Within Groups 0.001376 10 0.000138 Total 0.005431 14 99 Anova: Flow Rate - PPIMB Summary Groups Count Sum Average Variance Pure PP 3 16.26 5.42 0.0181 PP/.1%MB 3 16.17 5.39 0.0013 PP/%MB 3 16.19 5.396667 0.005033 PP/2%MB 3 16.03 5.343333 0.069433 PP/5%MB 3 15.03 5.01 0.01 1 1 ANOVA . Source of Variation 1 ss df MS F P-value Fcrit Between Groups 0.351307 4 0.087827 4.18355 0.03024 3.47805 Within Grou 3 0.209933 10 0.020993 Total 0.56124 14 I 100 Anova: Flow Rate - PP/HDPE I Summary Groups Count Sum Average Van'ance I I Pure PP 3 16.26 5.42 0.0181 I PPI5%HDPE 3 17.37 5.79 0.0052 PPP/oHDPE 3 16.13 5.376667 0.000533 . PP/2%HDPE 3 15.56 5.186667 0.050633 PP/5%HDPE 3: 15.18 5.06 0.0588 I ANOVA I Source of Variation 1 SS df MS F P-value F cn't Between Groups 0.9258 4 0.23145 8.683717 0.002725 3.47805 Within Groups 0.266533 10 0.026653 Total 1.192333 14