Table of Contents

Page

Table of Contents............................................................................................................... 2

List of Figures..................................................................................................................... 4

List of Tables...................................................................................................................... 6

Forward/Notice................................................................................................................... 7

Acknowledgement............................................................................................................... 8

Executive Summary............................................................................................................ 9

1.0 Introduction................................................................................................................ 11

2.0 Task Description and Result.................................................................................... 13

2.1 Task 0 Project Management............................................................................ 13

2.2 Task I Materials and Product Forms Technology........................................... 14

2.2.1 Prepeg Formulations................................................................................ 14

2.2.2 Flame Retardants.................................................................................... 14

2.3 Task II Processing Technology........................................................................ 18

2.3.1 Test Panel Fabrication............................................................................. 18

2.3.2 Mechanical Testing Results for Plaque Mold Test Panels................... 20

2.3.3 Discussion of Mechanical Testing Results............................................ 21

2.3.4 Tensile Tests Performed on Flow Tool Specimens................................ 26

2.3.5 Processing and Testing of Unidirectional Tape Inserts......................... 32

2.3.6 Porosity Evaluation.................................................................................. 34

2.3.7 Studies of Fiber Distribution and Orientation........................................ 40

2.4 Task III Tooling Technology............................................................................. 42

2.4.1 Flow Tool.................................................................................................. 42

2.4.2 Prototype Tool for Bus Seat.................................................................. 47

2.5 Task IV Component Selection........................................................................... 59

Table of Contents

Page

2.6 Task V Component Fabrication......................................................................... 62

2.6.1 Seat Design Basis.................................................................................... 62

2.6.2 Flow Modeling and Analysis................................................................... 67

2.6.3 Manufacturing Concept........................................................................... 72

2.7 Task VI Component Evaluation........................................................................ 73

2.7.1 Design Check........................................................................................... 73

2.7.2 Comparison of Flow Simulation with Prototype...................................... 74

3.0 Future Work............................................................................................................... 76

4.0 Technology Transfer................................................................................................. 78

Appendix A Flow Simulations of Compression Molded Panels..................................... 78

Appendix B Cost Analysis Summary for Bus Seat Production..................................... 86

Glossary............................................................................................................................ 92

List of Abbreviations........................................................................................................ 93

Metric Chart..................................................................................................................... 94

List of Figures

Page

Figure 1.0-1 Project Schedule and Milestones............................................................... 12

Figure 2.2.2-1 Sample of LFT-PPGF40 without Flame Retardant after Dynatup Test 15

Figure 2.2.2-2 Sample of LFT-PPGF40 with Exolit AP751 after Dynatup Test........... 16

Figure 2.3.3-1 Instrumented Drop Weight Low Velocity Impact Test.......................... 23

Figure 2.3.3-2 Fiber Pullout of Nylon 6 Izod Specimen.................................................. 24

Figure 2.3.3-3 Izod Test Specimen – Polypropylene Matrix/Glass Fiber..................... 25

Figure 2.3.4-1 Compression Moldong Tool for Manufacturing Tensile Test Tabs from Fiber-reinforced Thermoplastic Composites..................................................................... 29

Figure 2.3.4-2 Finished Compression Molded Tabs Edge View

and Grip Surface..................................................................................... 30

Figure 2.3.4-3 Tensile Specimens taken Perpendicular and Parallel............................ 30

Figure 2.3.4-4 Typical Stress/Strain Curves from Glass Reinforced PP Composite Materials 31

Figure 2.3.5-1 Microscopic View through a Co-molded Insert...................................... 32

Figure 2.3.6-1 Schematic of Panel Fabricated using the Flow Tool Showing Position of Specimens taken for Porosity Characterization....................................................................... 35

Figure 2.3.6-2 Panel with 517 kPa Back Pressure and 91 Metric Tons Mold Force 38

Figure 2.3.6-3 Panel with 2067 kPa Back Pressure and 91 Metric Tons Mold Force 39

Figure 2.3.6-4 Panel with 2067 kPa Back Pressure and 363 Metric Tons Mold Force 40

Figure 2.3.7-1 Difference between Local Weight Fraction Glass Fiber and Mean for a Typical Panel 41

Figure 2.4.1-1 Basic Flow Tool Design........................................................................... 42

Figure 2.4.1-2 The Flow Tool Installed in a Press.......................................................... 43

Figure 2.4.1-3 Detailed views of the Flow Tool.............................................................. 43

Figure 2.4.1-4 Part Fabricated with the Flow Tool and Secondary Insert..................... 44

Figure 2.4.1-5 Insert for Test Specimens and Retainer for Testing

......................... Tooling Materials.................................................................................... 45

Figure 2.4.1-6 Secondary Insert for Testing Different Tooling Materials................... 45

Figure 2.4.1-7 Mold Insert for a Component Geometry with Different Rib Thickness 46

Figure 2.4.1-8 Mold Insert for a Component Geometry with Cross Ribs.................... 46

Figure 2.4.2-1 View of Lower Mold Half before Pouring of Epoxy over Copper Cooling Lines 48

Figure 2.4.2-2 Mixing Setup using a Large Drill Press................................................. 49

Figure 2.4.2-3 Partially Filled Lower Mold Half............................................................ 50

Figure 2.4.2-4 View of the Upper Mold Half Showing Cooling Lines and a Partially Filled Mold Cavity 51

Figure 2.4.2-5 Upper Mold Half during Filling Stage.................................................... 52

List of Figures

Page

Figure 2.4.2-6 End of Mold Pouring with Epoxy Molding Compound Flush with the Top of the Mold Box 53

Figure 2.4.2-7 View of the Upper Mold Half after Encasement in a Steel Case.......... 54

Figure 2.4.2-8 Photograph of Lower Half of Seat Mold after Molding........................ 55

Figure 2.4.2-9 Detail of Aluminum Mold Insert............................................................. 56

Figure 2.4.2-10 View of Lower Mold Showing Aluminum Insert and Pre-positioned Reinforcements 57

Figure 2.4.2.11 Lower Mold Detail with Aluminum Insert............................................. 58

Figure 2.6.1-1 Views of the Seat Design........................................................................ 64

Figure 2.6.1-2 Top Edge Subjected to Vertical Load..................................................... 65

Figure 2.6.1-3 Top Edge Subjected to Horizontal (lateral) Load................................... 65

Figure 2.6.1-4 Vertical Load Applied to Seat Base....................................................... 66

Figure 2.6.2-1.. Flow Sequence of Injection Mold Long Glass Fiber/propylene

......................... in Ribbed Seat Design............................................................................ 67

Figure 2.6.2-2 Parameters: a) fill time, b) pressure drop, c) confidence of fill, d) flow front temperature, e) quality prediction, and f) weld lines for glass fibers/polypropylene flow.......... 68

Figure 2.6.2-3 Meshed Model CADPress™ Simulation.............................................. 70

Figure 2.6.2-4 Simulation of Bus Seat Filling Sequence Near the End of the Compression Cycle – Initially Places-Charge Location is Blue

......................... on Seat..................................................................................................... 71

Figure 2.7.1-1 Seat Mounted on Test Frame for Line Loading.................................... 73

Figure 2.7.2-1 Actual Seat Molded with a Short Charge............................................... 75

Figure A-1 Pressure Diagrams from Flat Plaque – Charge on End......................... 82

Figure A-2a..... Flow front for flat plaque-charge at end................................................. 83

Figure A-2b..... Upper fiber orientation from flat plaque-charge at end......................... 83

Figure A-2c..... Lower fiber orientation detail flat plaque-charge at end....................... 83

Figure A-2d..... Maximum nodal pressure from flat plaque-charge at end.................... 83

Figure A-3a..... Flow front for flat tool simulation........................................................... 85

Figure A-3b..... Fiber orientation from flow tool simulation............................................ 85

Figure A-3c..... Fiber orientation from flow tool simulation............................................ 85

Figure A-3d..... Fiber orientation from flow tool simulation............................................ 85

Figure A-4a..... Maximum nodal pressure in flow tool simulation.................................. 85

Figure A-4b..... Deformation (mm) in flow tool simulation.............................................. 85

Figure B-1....... Summary of Prototype Bus Seat Fabrication......................................... 86

Figure B-2....... Distribution of Costs for Prototype Bus Seat........................................ 87

Figure B-3 ...... Contrast of Two Feasible Versions of the Bus Seat.............................. 87

Figure B-4....... Comparison of per Part Non-labor Costs with Three Designs............. 88

Figure B-5....... Contrast of Production Cost per Part of Three Molded Design with a Calculated Cost of the Current Bus Seat in Use......................................................................... 88

List of Tables

Page

Table 2.2.1-1 Material Combinations and Prepeg Forms Produced............................ 14

Table 2.2.2-1 Flame Spread and Smoke Testing Results.............................................. 17

Table 2.3.1-1 Table of Compression-molded Panels Produced with Plaque Tool......... 18

Table 2.3.1-2 Unidirectional Panels Fabricated............................................................. 19

Table 2.3.2-1 Mechanical Testing Summary of Selected Plaque Mold Panels............ 20

Table 2.3.3-1 Summary of Theoretical Tensile Properties of E-Glass

and Polypylene.......................................................................................... 21

Table 2.3.4-2 Ultimate Tensile Strength, Tensile Modulus, and Dynamic Modulus for Tensile and Panels Fabricated with 2067 kPa Back Pressure and 363 Metric Tons Mold Force................................ 28

Table 2.3.5-1 Unidirectional Composite Mechanical Properties................................... 33

Table 2.3.6-1 Panels for Porosity Studies Showing Process Variables......................... 37

Table 2.5-1 Candidate Bus Components Contrasted by Cost, Weight, And Fabrication.................................. 59

Table 2.5-2.... Candidate Bus Component Replacement Contrasted by Cost, Weight, and Fabrication............. 60

Table 2.5-3.... Expected Total Weight and Cost Savings per Bus................................. 61

Table 2.6.1-1. Properties Summary of Nylon/glass Ribbed Seat Design...................... 66

Table 2.6.2-1. Summary of Injection Molding and Material Parameters...................... 68

Table 2.6.2-2. Summary of Input Parameters for Flow Simulation............................... 69

Table 2.6.3-1. Production Parameters for the Seat......................................................... 72

Table A-1(a).. Sample Material Data Input in CADPress™ Program.......................... 78

Table A-1(b).. Sample Material Data Input in CADPress™ Program.......................... 79

Table A-2....... Process Variables for Flat Plaque and Flow Tool Simulation................. 80

Table B-1....... Cost Breakdown for Prototype Part........................................................ 86

Table B-2....... General Design Parameters for Version 1 Design................................. 89

Table B-3(a).. Breakdown of Design Version 1 Costs.................................................... 89

Table B-3(b).. Breakdown of Design Version 1 Costs.................................................... 90

Table B-3©.... Breakdown of Design Version 1 Costs.................................................... 90

Table B-4....... Contributing Costs and Estimated Sale Price Design Version 1............ 91

Foreword

This annual report describes an investigation of the use of long-fiber thermoplastic composites in transit bus applications for the Department of Transportation and the Federal Transit Administration. The goals of improved safety, reduced weight, and lower cost are very important to the transportation industry. Investigations into several candidate transit bus components were made to replace heavier conventional components with long-fiber thermoplastic components while simultaneously maintaining safety and reducing fabrication costs. A two-person bus seat was selected as the composite candidate and then was designed and compression-molded as a demonstration of the technology. The weight of the composite seat was 50 percent less than the conventional component and could be manufactured in commercial quantities with a 40 percent reduction in cost.

Notice

This document is disseminated under the sponsorship of the United States Department of Transportation in the interest of information exchange. The United States Government assumes no liability for its contents or use thereof.

The United States Government does not endorse manufacturers or products. Trade names appear in the document only because they are essential to the content of the report.

Acknowledgment

The authors wish to express their appreciation to the Federal Transit Administration for the support of this work. We thank the C.A. Lawton Company for the use of their plastication equipment and North American Bus Industries for their guidance in component selection and design guidelines. We acknowledge the following for their contributions to the technical effort of this project: from Southern Research Institute, George Husman, Tim Hartness, Patric Moriarty, Mike Dyksterhouse, and Andy Grabany; from The University of Alabama at Birmingham, Uday Vaidya, Greg Janowski, Krish Chawla, Chad Ulven, Shane Bartus, Brian Pillas, and Jianguang Zhang; and from Polycomp, Joel Dyksterhouse. We acknowledge and thank Lisa White for her assistance with the preparation of this report.

Executive Summary

The objectives of this project were to develop thermoplastic composite materials and product forms and to demonstrate fabrication methods for molding these materials into components for use in buses and other mass transit applications. The primary goal was to demonstrate technologies that can provide lower cost, lighter weight, improved performance structures for mass transit applications.

The work described in this report was performed under a Cooperative Agreement with the U.S. Department of Transportation, Federal Transit Administration (FTA), FTA C-7, October 1, 2000, Project number: AL-26-7001. The work was performed during the period of March 15, 2001 to May 31, 2002.

The project brought together the expertise from several technical disciplines to perform this work. The team was headed by the Engineering Division of Southern Research Institute (SRI), Birmingham, Alabama, with technical support from The University of Alabama at Birmingham / School of Engineering (UAB); C.A. Lawton Company, Green Bay, Wisconsin; North American Bus Industries (NABI), Anniston, Alabama; and Joel Dyksterhouse, technical expert in thermoplastic composites.

This project was performed under a set of parallel and sequential tasks. Tasks I – III were technology base development efforts aimed at increasing knowledge and understanding of thermoplastic composite materials and product forms, in processing methodologies for molding these materials into structural components, and in tooling design and fabrication methodologies for producing cost-effective prototype and production tools. Tasks IV – VI were focused on selection, fabrication, and evaluation of a bus component that would demonstrate the technologies developed.

Several thermoplastic prepreg materials were formulated. Test panels were fabricated and tested from the prepreg materials produced. A compression molding flow tool was designed, manufactured and used to mold plaques for test specimens. Test panels were made with flame-retardants and then tested successfully for both flame spread and smoke density. Based upon cost and weight analysis, a bus seat was selected as the component for fabrication and testing. An all-composite 2-person bus seat was designed and analyzed. Prototype tooling was designed to compression mold the bus seat design in a production facility. The molding trial of the bus seat resulted in the successful production of 20 full-scale bus seats that were used for testing and validation studies. The molding operation was also simulated using unique long fiber thermoplastic analysis software, CADpressÔ for Thermoplastics. Comparison of the molded parts with the flow simulation was profitable and excellent press tonnage requirements were realized. The cost of producing a seat was at least 40% less than the current model with weight savings of approximately 50%. The mechanical testing of the seat back met specifications without damage or permanent deformation.

The intention of future work is to emphasize larger structures with additional processing technologies. The technology base will continue to expand with additional materials and fabrication methods.

1.0 Introduction

The work described in this report was performed under a Cooperative Agreement with the U.S. Department of Transportation, Federal Transit Administration (FTA), FTA C-7, October 1, 2000, Project number: AL-26-7001. The designated project was performed during the period of March 15, 2001 to May 31, 2002.

The objectives of this project were to develop thermoplastic composite materials and product forms and to develop and demonstrate fabrication methods for molding these materials into components for use in buses and other mass transit applications. A primary goal was to demonstrate technologies that can provide lower cost, lighter weight, and improved performance structures for mass transit applications.

Southern Research Institute (SRI), Engineering Division, located in Birmingham, Alabama, lead the team performing this project. SRI was responsible for materials and fabrication technology development, for fabricating demonstration components, and for overall project management.

The other team members were:

North American Bus Industries (NABI), a leading bus manufacturer located in Anniston, Alabama, was responsible for identifying and assisting the selection of demonstration components that could benefit from the use of thermoplastic composite materials and fabrication technologies. NABI provided component requirements and helped evaluate the demonstration components fabricated in this project.

C.A. Lawton Company, a manufacturer of plasticators and presses for compression molding, located in Green Bay, Wisconsin, was responsible for technical and equipment support in extrusion / compression molding and was also responsible for development of cost effective tooling fabrication technologies.

The University of Alabama at Birmingham (UAB), School of Engineering, located in Birmingham, Alabama, was responsible for materials and process modeling and analysis. This included process flow modeling, component design, performance prediction, and mechanical properties of composite materials.

Joel Dyksterhouse, consultant and technical expert in thermoplastic composite materials and processing and inventor of the DRIFT process, provided technical support to all tasks of the project.

This project was performed with parallel and sequential tasks, as can be seen in Figure 1.0-1, Program Schedule / Milestones. Tasks I – III were technology base development tasks aimed at increasing knowledge and understanding in thermoplastic composite materials, product forms, processing methodologies needed to mold these materials into structural components. Included in these tasks were tooling design and fabrication methodologies for producing cost-effective prototype and production tools. Tasks IV – VI were focused on selecting, fabricating, and evaluating a bus component that demonstrated the technologies. The objectives and progress reports for each of these tasks are presented in Section 2.0 of this report.

Figure 1.0-1 Project Schedule and Milestones

2.0 Task Description and Result

2.1 Task 0 – Project Management

During the first quarter, Cooperative Agreements were negotiated with each team member. The kick-off meeting with North American Bus Industries (NABI) was held May 10, 2001, at the NABI plant in Anniston, Alabama. At this meeting, potential demonstration components were discussed and actual parts were viewed. Initial results for component definition and selection are presented in Task IV. The kick-off meeting with C.A. Lawton Company was held May 17, 2001, at the Lawton plant in DePere, Wisconsin. Lawton’s participation in the project was directed at the development of tooling technology and fabricating tools for use in the program. Initial discussions of this area are presented in Task III.

The kick-off meeting with the University of Alabama at Birmingham (UAB) was held June 8, 2001. UAB support for the project focused on design, process modeling, and composite mechanics. UAB worked very closely with SRI on all tasks of the project.

During the second quarter, there were two design review meetings with C.A. Lawton, SRI and Joel Dyksterhouse at the C.A. Lawton Company. During the first meeting, the basic design for the flow tool was reviewed and necessary changes were made. The final design was discussed and approved at the second meeting.

Weekly meetings were established with UAB to discuss the ongoing work and to improve the design of the seat structure as well as the flow simulation.

The status of North American Bus Industries (NABI) changed during the second quarter. NABI supported all future work but did not ask for funding of their work.

The demonstration part selection was discussed with NABI.

During the third quarter, two design review meetings with C.A. Lawton and SRI took place at C.A. Lawton. At the first meeting in October, the final design for the flow tool, the ongoing fabrication work on the flow tool, and the seat tool design were the main objectives. At the second meeting in November, the seat design was reviewed and necessary changes were made. UAB was also a contributor during this review process. The final design was approved via phone conferences.

December 4^th, we had a meeting at SRI with the selected producer for the tooling material for the seat mold, to discuss the details and needs for tooling.

The bus seat surface models were fabricated by C.A. Lawton Co. during the fourth quarter. Mold plugs were cast over these models at their facilities in Green Bay, Wisconsin. SRI and UAB personnel were present to document and assist in this operation. Several additional meetings were made with SRI, UAB, and C.A. Lawton to finalize tooling issues before molding trials. Molding trials were scheduled and performed on two occasions with the supervision of SRI, UAB and C.A. Lawton.

2.2 Task I Materials and Product Forms Technology

2.2.1 Prepreg Formulations

During this project we produced several combinations of materials in various forms. From these raw materials, we fabricated panels for testing and evaluation in order to develop a database from which materials would be reasonably selected so that appropriate design decisions can be made. The following table summarizes the general material combinations and product forms produced. Refer to the Glossary or List of Abbreviations for terminology, abbreviations, or acronyms.

Resin

Fiber

Form of Prepreg Material

Polypropylene Homopolymer

E-glass

PP Compatible

12.7 mm pellets

25.4 mm pellets

6.35 mm tape

Polypropylene

Impact-Modified Copolymer

E-glass

PP Compatible

25.4 mm pellets

6.35 mm tape

Polyethylene

High Density

E-glass

Multi-purpose thermoplastic compatible

25.4 mm pellets

6.35 mm tape

Nylon 6

E-glass

Nylon compatible

25.4 mm pellets

6.35 mm tape

Nylon 6

E-glass

Multi-purpose

Thermoplastic compatible

6.35 mm tape

Polypropylene Homopolymer

Carbon

6.35 mm tape

Nylon 6

Carbon

6.35 mm tape

Table 2.2.1-1 Material Combinations and Prepreg Forms Produced

The formulation of additional resin and fiber combinations in following years will permit investigation of articles from materials possessing much greater strength and toughness as well as improved performance in high temperature applications. The choice of materials combinations was application driven. As new materials are developed they will be fabricated into panels and tested to establish mechanical properties critical to design and component fabrication.

2.2.2 Flame Retardants

The requirement for interior mass transit applications requires flame and smoke testing performed according to ASTM E-162 and ASTM E-662. Although high glass fiber content inherently promotes flame retardant characteristics, methods to augment flame and smoke performance were considered.

We investigated two different flame retardant concepts:

· Compounding a flame retardant directly into the polymer

· Applying the flame retardant as a top coat

Compounding of flame retardants:

A flame retardant can be directly compounded into the polymer during the impregnation process. Therefore, a non-halogenated flame retardant was selected to avoid toxicity associated with halogenated materials. There are a very limited number of flame-retardants available for polypropylene. We tested Exolit AP751, made by Clariant, in a 40% by weight glass reinforced polypropylene. The addition of the flame retardant dramatically changed the impact behavior of the test panel. Adding 5% by weight (with respect to the resin) of Exolit AP751 in the resin reduced the Dynatup impact capacity by approx. 10%. But more important, there was a change in the mode of failure. Figures 2.2.2-1 and 2.2.2-2 shows a sample with and without flame retardant after Dynatup impact testing. The sample with the flame retardant failed as a brittle material.

Figure 2.2.2-1 Sample of LFT-PPGF40 without Flame Retardant after Dynatup Test

Figure 2.2.2-2 Sample of LFT-PPGF40 with Exolit AP751 after Dynatup Test

We also produced flame retardant samples for Nylon using DSM MelaPur 200/70 during impregnation. When the prepreg fed into the plasticator, the retardant material was not thermally stable during plastication and decomposed resulting in extreme lofting due to the gas production. Therefore this compounded version was not considered for any further testing in compression molding applications.

Flame retardant as a topcoat:

Two products were tested: Flame Seal FX-PL from Flame Seal Products, Inc. and Thermaflex II-C from Avtec Industries. Both flame retardant products are rated as UL-V0 rating. The adhesion to the substrate made from long fiber thermoplastic composite (LFT) appeared to be adequate. These materials were applied by brush or airless spray to the polypropylene panels so that the surface finish and flame retardancy characteristics could be evaluated. Informal tests indicated that both products appeared to provide good flame retardancy. The Flame Seal product was selected for formal testing on the basis of surface finish after drying.

Polypropylene homopolymer panels containing 40% glass by weight were produced and coated with the Flame Seal FX-PL product. After curing at 60°C, the samples were submitted to an independent testing lab, SGS U.S. Testing Company of Tulsa, Oklahoma, for evaluations with respect to flame-spread and smoke density generation according to ASTM standards E162-95 and E662-95.

Results of Flame Retardant Tests

The results from the independent flammability test laboratory were favorable. The data in Table 2.2.2-1 summarizes the flame spread and smoke generation data with the standards for bus seat applications.

Test Performed	Test Result	Test Requirement
Flame-Spread, Is	3.01	Is<35
Smoke Density –Flaming mode, Ds(1.5), Ds( 4)	0.19, 0.96	Ds(1.5)<100. Ds(4)<200
Smoke Density –Non-flaming mode, Ds(1.5), Ds( 4)	0.29, 1.26	Ds(1.5)<100. Ds(4)<200

Table 2.2.2-1 Flame Spread and Smoke Testing Results

As can be readily noted from the results, the surface coating of the flame retardant was sufficient to pass both flame-spread and smoke density tests with the polypropylene/glass composite formulation. The addition of another coating to the surface of the bus seat or the addition of pigments to the flame retardant could change the result to some degree; however, the result of the flame and smoke test was positive. For commercial application the tests would be performed with particular color coatings and full-scale flame spread tests.

2.3 Task II Processing Technology

2.3.1 Test Panel Fabrication

During the fourth quarter of this effort, test panels were initially compression molded with a 0.61m x 0.46m meter plaque tool and later panels with a 0.6m x 0.24m flow tool. Two molds were used for fabricating samples because the smaller flow tool had to be constructed. The smaller flow tool was preferred since higher compression pressures could be realized during processing. Flow simulations of both the plaque tool and flow-tool are found in Appendix A. The record of panel types is listed in Table 2.3.1-1 denoting the basic material combinations and product forms made with the plaque tool. Each of the panel identification numbers represent 10 to 15 individual panels made for mechanical testing and fiber content analysis.

The purpose of producing different panels was to investigate the effects of different resins, fiber length, and fiber content on the panel mechanical properties. Selected panels were mechanically tested as noted in the next section.

Panel ID	Resin Type	Fiber Type	Fiber Length mm	Fiber Content Wgt%	Color
010626-1	High melt flow PP homopolymer	E glass	25.4	40	Natural
010626-2	High melt flow PP homopolymer	E glass	25.4	40	Black
010723-1	Medium melt flow PP high impact copolymer	E glass	25.4	40	Black
010723-2	Medium melt flow PP high impact copolymer	E glass	25.4	40	Natural
010919-1	High melt flow PP homopolymer	E glass	25.4	40	Natural
010919-2	High melt flow PP homopolymer	E glass	12.7	40	Natural
010920-1	High melt flow PP homopolymer	E glass	25.4	50	Natural
010920-2	High melt flow PP homopolymer	E glass	25.4	60	Natural
011601-1	Medium melt flow polyethylene (PE)	E glass	25.4	40	Natural
011601-2	High melt flow PP homopolymer	E glass	12.7	60	Natural
011601-3	High melt flow Nylon 6	E glass	25.4	40	Natural

Table 2.3.1-1 Table of Compression-molded Panels Produced with Plaque Tool

In addition to the compression-molded panels, several unidirectional fiber test panels were made. The unidirectional panels were fabricated by wrapping tape prepreg over a frame. The frame-wrapped material was placed in a heated press to consolidate the tapes into a panel that provided sufficient area for mechanical testing purposes. A list of these unidirectional panels is found in Table 2.3.1-2 below. The unidirectional panels provided opportunity to minimize the effects of random fiber orientation that normally occur during compression molding.

Panel #	Description	Fiber	Fiber Content Wgt%	Number Plys
105	High melt flow PP with thermoplastic size on E glass	E glass	60	2
107	High melt flow PP with thermoplastic size on E glass	E glass	60	2
108	High melt flow PP with thermoplastic size on E glass	E glass	60	2
109	High melt flow PP with thermoplastic size on E glass	E glass	60	2
110	Medium melt flow PE with thermoplastic size on E glass	E glass	60	2
111	Medium melt flow PE with thermoplastic size on E glass	E glass	60	2
114	High melt flow Nylon 6 with Nylon size on E glass	E glass	60	6
115	High melt flow Nylon 6 with Nylon size on E glass	E glass	60	2
116	High melt flow Nylon 6 with Urethane size partially compatible for Nylon on E glass	E glass	60	2
117	High melt flow Nylon 6 with Urethane size partially compatible for Nylon on E glass	E glass	60	2
118	High impact PP copolymer with PP compatible size on fiber	E glass	60	2
119	High impact PP copolymer with PP compatible size on fiber	E glass	60	2
120	Nylon 6 with Nylon and PP compatible size on fiber	E glass	60	2
121	Nylon 6 with Nylon and PP compatible size on fiber	E glass	60	2
122	High melt flow PE with thermoplastic size on fiber	E glass	60	2
130	High melt flow PP with thermoplastic size on E-glass	E glass	60	4
132	High melt flow PE with thermoplastic size on E-glass	E glass	60	4
134	High melt flow PP homopolymer	carbon	60	4
137	High melt flow Nylon 6	carbon	60	4

Table 2.3.1-2 Unidirectional Panels Fabricated

2.3.2 Mechanical Testing Results for Plaque Mold Test Panels

Randomly oriented panels were selected for mechanical testing. The tests performed were chosen to enhance investigation of material combinations and to assist in resolving test issues associated with long fiber thermoplastic composite materials. A summary of the mechanical property tests is found in Table 2.3.2-1.

The data in Table 2.3.2-1 were characterized by panel type, fiber content, and fiber length. The mechanical tests focused on impact and tensile properties. The impact data include the standard notched Izod test from samples oriented longitudinally (direction of material flow) and transversely (normal to material flow), as well as an instrumented Falling Dart test. The Falling Dart test and the Izod test showed total energy absorbed and were normalized with respect to the thickness of the sample. It should be noted that the two impact tests are very different and that direct comparison of the two methods is difficult. The last three columns of the table display the tensile properties of ultimate strength and modulus of specimens 25.4 mm wide and 229 mm long oriented longitudinally to material flow in the mold. The last column is a nondestructive ultrasonic modulus. The randomly oriented specimens were taken from plaque mold panels that were compression molded at 363 metric tons of force on a 0.28 square meter area that yielded a 12.8 Mpa compression pressure. The plastication back pressure of the extruder was about 500 kPa during the panel production.

Material #/ Fiber Orientation	Material/ Fiber Content	Fiber Length	Long Izod	Trans. Izod	Falling Dart	Long. Ult. Strength (see note 1)	Long Mod.	Uson Mod
		mm	J/m	J/m	J/m	MPa	GPa	GPa
1 Random	PP Copoly, blk E glass/40%	25.4	411	215	3393	73.1	7.45	-
2 Random	PP Copoly nat . E glass/40%	25.4	365	193	3016	73.9	6.62	-
3 Random	PP Homo nat E glass/40	25.4	380	177	2845	77.3	6.9	-
4 Random	PP Homo nat E glass/40	12.7	464	169	3117	84.4	7.6	10.4
5 Random	PP Homo nat. E glass/50	25.4	449	204	3005	92.4	10.3	-
6 Random	PP Homo nat. E glass/60	25.4	550	178	3244	112.7	12.1	16.3
7 Random	High Density Polyethylene (HDPE) nat., E glass/40%	25.4	752	283	2786	108.8	-	13.0
8 Random	PP Homo nat. E glass/60	12.7	531	327	3631	119.6	-	16.8
9 Random	PA6 nat. E glass/40%	25.4	716	303	3052	35.4	-	12.4
10 Unidirectional	PP Homo nat. E glass/60	-	-	-	-	405.2	18.1	-

Table 2.3.2-1 Mechanical Testing Summary of Selected Plaque Mold Panels

Note 1: These initial panels continued porosity due to low process back pressure. See section 2.3.4 for reduced porosity data.

2.3.3 Discussion of Mechanical Testing Results

The consideration of the mechanical properties for the list of materials in Table 2.3.2-1 showed several important characteristics related to fiber length, fiber concentration, resin type, and test specimen orientation. Most of the panels tested were fabricated from 25.4 mm long pellets. Two were fabricated with 12.7 mm pellets and 1 panel tested was made with unidirectional tape.

As a background comparison, a theoretical calculation of composite strength and modulus was made based on the strength and modulus of E glass fiber and the weight percentage of that fiber in a polypropylene matrix. The theoretical tensile data calculated in Table 2.3.3-1 assumes that there is no effect of fiber length on the ultimate translation of mechanical properties. E glass fiber has a tensile strength of about 3445 MPa and a modulus of about 76 Gpa. The unidirectional properties were based on glass volume fraction and the in plane random properties were assumed to be one half of the corresponding unidirectional properties

Weight Percent E-glass	Orientation	Tensile Strength MPa	Tensile Modulus GPa
40	Random	330	7.28
40	Unidirectional	660	14.6
50	Random	453	9.96
50	Unidirectional	900	19.9
60	Random	600	13.2
60	Unidirectional	1200	26.4

Table 2.3.3-1 Summary of Theoretical Tensile Properties of E-Glass and Polypropylene

The modulus of the test specimens is generally observed to be roughly equal to the theoretical modulus and the tensile strength less than the theoretical strength. The measured modulus therefore was roughly the same as the random values for a given fiber content. The measured tensile strength was on the order of 22 to 25 percent of the theoretical tensile strength expected from a random glass composite. Several factors contribute to the lower tensile strength numbers. The first reason is that the method of tensile testing used a straight-sided specimen that is predisposed to break in the grip. The second reason is that there was some porosity in the specimens due to the lower processing pressure of the large plaque mold and in the back pressure in the plasticator. These two conditions were addressed when specimens were formed with the flow tool constructed later in the program and are noted in Section 2.3.4. A third reason for less than optimal tensile performance may be due to weakness in the fiber matrix bond.

Effects of fiber orientation provided an interesting contrast of Material 3 and 4 in which the shorter fiber material displayed a slightly higher ultimate strength, modulus, longitudinal notched Izod energy and falling dart (Dynatup) impact energy absorption. One would generally expect the longer fiber material to exhibit higher numbers on all counts; however, just the opposite occurred. The most likely explanation of this phenomenon is that the shorter fibers were more highly oriented than the 25.4 mm fibers. The materials were otherwise the same. Therefore it is important not to hastily conclude that shorter fibers are superior in strength and modulus in composite applications. It does; however, clearly illustrate the importance of fiber orientation in compression molding processes and the effect of fiber length upon the degree of orientation.

The issue of orientation of test samples when forming compression-molded panels is real. Noting the difference between the longitudinal and transverse Izod measurements in all of the test specimens illustrates this point. There was a two or three fold difference in the Izod measurements when the orientations are contrasted. In addition to the predictable variations in fiber orientation, there are random variations in fiber concentration that also occurs. Therefore design decisions related to mechanical performance of compression-molded components must make with a statistical perspective.

Comparison of Materials 4, 5, and 6 yields a trend of increasing strength, modulus, and impact resistance with increasing fiber concentration in the same material that one might reasonably expect. In this case all of the fibers were originally the same length and demonstrated increasing performance with increased fiber loading. It should be noted that the modulus was greater than expected for a random three-dimensional fiber distribution indicating orientation effects during processing. Tensile strength was one fourth to one third of the theoretical under the same conditions. The lower than expected tensile strength can be attributed to testing and fabrication factors noted previously.

Consideration of the unidirectional panel, Material 10, illustrates the fact that the true unidirectional specimen can exhibit about 3 times the random orientation strength; however, is still only about 48 percent of the theoretical tensile strength. The issue of predisposition of sample breakage in the grips and fiber alignment give the most credible explanation to the less than optimal strength.

Impact data in Table 2.3.2-1 were performed by longitudinal and transverse Izod and low velocity falling dart (Dynatup) measurements. Notched Izod measurements were made from samples taken from both the direction of fiber orientation and normal to that direction. Dynatup measurements were performed on panels 76.2 mm or 101.6 mm square.

Low velocity impact tests were conducted to evaluate damage initiation, progression and fracture of the thermoplastic composite samples. The equipment used to conduct the tests is an instrumented Instron 8250-drop tower, as shown in Figure 2.3.2-1. The basic principle of operation is to drop a tup of known weight from a set height onto the test sample. The maximum load and maximum energy absorbed by the test sample and the damage to the sample is assessed.

Figure 2.3.3-1 Instrumented Drop Weight Low Velocity Impact Test

Configurations that were tested include:

· Varying fiber length; 25.4 mm and 12.7 mm

· Varying matrix materials; High impact PP copolymer, high melt flow PP, medium melt flow PE, Nylon 6

Table 2.3.2-1 summarized the impact test results normalized to the thickness of the samples. The typical failure is through localized indentation of the top surface, and an accumulation of damage to the tensile side of the samples. The effects of additives on impact performance need to be considered for any design. The addition of fire retardants reduced the peak load and energy absorbed impact properties of the material as noted in Section 2.2.2.

Correlation of the Falling Dart and Izod tests is not easily accomplished by merely comparing numbers. An excellent example of this is found in Material 6, which exhibited a high longitudinal Izod value, but also demonstrated a lower Falling Dart energy absorption compared to the other tests. This difference may be more related to the matrix properties

Another interesting example was a Nylon 6/glass specimen that did not bond well with glass fiber. The fiber pullout of this sample, shown in Figure 2.3.3-2, dissipated considerable energy in the Izod test, but the lack of substantive bonding did not permit good tensile performance. The tensile strength of that material was also one third of comparable materials. Figure 2.3.3-2 can be contrasted with a polypropylene Izod sample shown in Figure 2.3.3-3, which exhibited shorter fibers, indicative of better coupling of the polymer to the fiber surface that translate to better tensile performance, but less impact capacity.

Figure 2.3.3-2 Fiber Pullout of Nylon 6 Izod Specimen

Figure 2.3.3-3 Izod Test Specimen – Polypropylene Matrix/Glass Fiber

It is important that Izod and Falling Dart measurements must be very carefully analyzed for a particular application. The highest Falling Dart measurements were found in Materials 1, 5, and 7. Materials 5 and 7 both had 60-weight percent fiber content and Material 1 was composed of high impact polypropylene copolymer with 40 weight percent glass. This observation is not unexpected since improved impact performance is usually associated with higher fiber loading and impact resistant matrices.

2.3.4 Tensile Tests Performed on Flow Tool Specimens

Background

As noted in previous sections, it was expected that there were negative influences of porosity on the performance for our test specimens related to mold pressure and also to the back pressure seen in the plasticator during charge formation. In addition to the fabrication issues related to the specimens, the test method was not conducive to consistent breakage in the gauge section of the specimen. Therefore, when the flow tool was available, panels were made under conditions that would reduce porosity and the test specimens could be tabbed to assist in a proper introduction of load into the test piece. Panels were fabricated with the processing parameters of 1000 kPa plasticator back pressure and 363 metric tons press force (25 MPa pressure) to form tensile specimens. The matrix was a polypropylene copolymer with E glass reinforcement at 40% by weight. These panels had the lowest porosity level (see Section 2.3.6 on porosity) and, therefore, should have higher, more reproducible tensile properties. Samples were taken both parallel (designated as “L”) and perpendicular (designated as “T”) to the long side of the panels to investigate the anisotropy of material properties.

Procedure

It was necessary to cut and establish an undamaged gage length to perform a tensile test. In polymer-matrix composites, this is typically done by cutting rectangular pieces from the panels and bonding tabs with a 10° taper on the ends. The present study developed a compression molding technique for making tabs and found a method to bond the tabs onto the test piece.

A compression-molding tool was machined from aluminum plate (see Figure 2.3.4-1a). A charge from the same material as that to be tested was heated to 200°C for 10 minutes and placed onto the mold. The mold (preheated to 100°C) was placed in a hydraulic press and a force of 19 metric tons applied for 60 seconds. A photograph of the tool, charge, and a finished tab is also shown in Figure 2.3.4-1b. The process resulted in dimensionally acceptable tabs, shown in Figure 2.3.4-2.

The following process was used to bond the tabs to the test piece:

· Clean both the test piece and the tabs using acetone.

· Roughen the surface of the tabs and the test piece using 120 grit silicon carbide papers.

· Clean again with acetone.

· Apply Threebond PP/PE primer (TB 1797) to the test piece and tab surface. Allow drying for 5 minutes.

· Apply Threebond adhesive (TB1743) to the tab and bond to test piece. Clamp the tabs onto the test piece and allow curing overnight (6-8 hours). Ensure that the tapered portion of tabs are properly aligned and clamped.

· Remove clamps and clean of excess adhesive (do not use acetone).

Four samples that were taken from each direction (parallel and perpendicular to the long side of the panel) were tested. Samples were tested in a servo‑hydraulic tensile test machine at a displacement rate of 2 mm/s with a gage length of 241 mm, which resulted in a strain rate of 8×10^-3/s. Modulus measurements were made with a clip-on extensometer, which was detached prior to failure.

A non-destructive, dynamic modulus measurement (Grindosonic) was made to independently verify the tensile modulus data and to give a correlation between the two methods.

Results

The tab fabrication and bonding were generally successful, with specimen failure occurring either in the tensile gage section or the tapered gage/tab transition. The tabs remained bonded during testing in all cases. Typical specimen break mode is shown in Figure 2.3.4-3. Further development of the tab design and specimen preparation is needed, with the goal of developing a standard testing protocol for this class of engineering materials.

Examples of stress/strain curves from representative samples are shown in Figure 2.3.4-4, with both the low strain portion (non-linear elastic) used to determine modulus and the entire test to fracture. These curves are typical of glass fiber reinforced thermoplastic composite materials.

The ultimate tensile strength (UTS) and Young’s Modulus (E) from the tensile tests are shown in Table 2.3.4-2, along with the dynamic modulus data. The averages for all tensile tests were 75.5 MPa and 6.4 GPa for UTS and modulus, respectively. The dynamic modulus measurements averaged 5.6 GPa, slightly lower than measured in tension at larger strains. The panels exhibited a higher UTS and lower modulus perpendicular to the axis; the dynamic modulus was more anisotropic that the tensile modulus. The values obtained for both the UTS and modulus fall within the expected and acceptable parameters for glass fiber reinforced thermoplastics. The highest tensile strength numbers were improved from 73.9 MPa to 82.4 MPa when compared to the corresponding polypropylene copolymer/40% E glass specimens tested from the plaque tool. Modulus measurements were not significantly affected by the tabbing and process changes with both plaque mold and flow tool samples measuring in the vicinity of 7 GPa, which was close to the 2 dimensional in-plane theoretical modulus of 7.28 GPa. The theoretical UTS was not replaced and remained close to 25% of the theoretical valve.

UTS and tensile modulus was not identified for the two orientations but were quite similar indicating a lack of strong fiber orientation.

	UTS (MPa)	Tensile E (GPa)	Dynamic E (GPa)
Perpendicular to Panel Axis
T1	86.7	7.15	5.50
T2	90.3	5.53	4.70
T3	79.5	5.49	4.72
T4	73.1	6.00	4.80
Mean	82.4	6.04	4.93
Parallel to Panel Axis
L1	63.1	6.50	6.20
L2	66.9	6.63	6.60
L3	64.6	5.65	6.00
L4	71.6	8.35	6.63
Mean	66.6	6.78	6.36

Table 2.3.4-2 Ultimate Tensile Strength, Tensile Modulus, and Dynamic Modulus for Tensile and Panels Fabricated with 2067 kPa Back Pressure and 363 Metric Tons Mold Force

(a)

(b)

Figure 2.3.4-1 (a) Compression Molding Tool for Manufacturing Tensile Test Tabs from Fiber-reinforced Thermoplastic Composites, and (b) Bottom Tool with Charge (left) and Finished Tab (right)

Run #	Material	Back Pressure (kPa)	Lower Mold Temperature (°C)	Upper Mold Temperature (°C)	Mold Pressure (metric ton)
020306-1	High impact PP copolymer- 25.4 mm flake	517	40	35	91
020306-2	High impact PP copolymer- 25.4 mm flake	517	40	35	181
020306-3	High impact PP copolymer- 25.4 mm flake	517	40	35	272
020306-4	High impact PP copolymer- 25.4 mm flake	517	40	35	363
020306-5	High impact PP copolymer- 25.4 mm flake	1034	40	35	91
020306-6	High impact PP copolymer- 25.4 mm flake	1034	40	35	181
020306-7	High impact PP copolymer- 25.4 mm flake	1034	40	35	272
020306-8	High impact PP copolymer- 25.4 mm flake	1034	40	35	363
020306-9	High impact PP copolymer- 25.4 mm flake	2067	40	35	91
020306-10	High impact PP copolymer- 25.4 mm flake	2067	40	35	181
020306-11	High impact PP copolymer- 25.4 mm flake	2067	40	35	272
020306-12	High impact PP copolymer- 25.4 mm flake	2067	40	35	363
020306-13	High impact PP copolymer- 25.4 mm flake	517	60	55	91
020306-14	High impact PP copolymer- 25.4 mm flake	517	60	55	181
020306-15	High impact PP copolymer- 25.4 mm flake	517	60	55	272
020306-16	High impact PP copolymer- 25.4 mm flake	517	60	55	363

Table 2.3.6-1 Panels for Porosity Studies Showing Process Variables

Figure 2.3.6-2 Panel with 517 kPa Back Pressure and 91 Metric Tons Mold Force

High levels of porosity (dark gray) are visible. The light phase in these figures is the glass fiber. Elongated glass fibers are in the plane of polish. The compression direction is vertical.

Figure 2.3.6-3 Panel with 2067 kPa Back Pressure and 91 Metric Tons Mold Force

Limited of porosity (dark gray) is visible. The compression direction is vertical.

Figure 2.3.6- 4 Panel with 2067 kPa Back Pressure and 363 Metric Tons Mold Force

Virtually no porosity (dark gray) is visible. The compression direction is vertical.

2.3.7 Studies of Fiber Distribution and Orientation

Studies of the fiber distribution, orientation, and size molding were performed after compression. It was necessary to experimentally determine these parameters in order to (1) assess the composite processing, (2) analyze mechanical property data, and (3) confirm the CADpress_™ modeling results.

The weight fraction of glass fiber was measured using the burnout method. In this technique, each sample was weighed, placed in a weighed aluminum pan, and heated with the following cycle:

· Heat to 150^oC, hold for 30 minutes

· Step 50^oC per hour up to 400^oC

· Hold for six hours

· Cool to room temperature

The pan and sample were reweighed, with the weight loss being attributed to the polymer matrix. It has been confirmed that this procedure completely removes the polypropylene matrix.

The weight fraction of glass fiber in a typical 0.61m x 0.46m plaque panel was determined in 108 locations. The 38mm diameter specimens were cut in a grid pattern that was defined by a row and column number. The mean weight fraction of fiber of this panel was 38.6% with a standard deviation of 1.3%. The difference between the panel mean and each row/column location is noted in Figure 2.3.7-1. The center of the starting charge was positioned at approximately row 3, column 5. The local variation in weight fractions was 36.5 to 43.0 wt.%, which was an expected variation for a composite product of this size.

		Column
		1	2	3	4	5	6	7	8	9
Row	1	1.8	1.7	1.8	1.5	1.8	1.7	1.1	2.2	4.4
	2	0.3	0.2	-0.2	0.1	0.7	0.5	0.5	0.5	4.4
	3	0	-0.9	-0.2	-0.1	0.3	-0.3	-1	-0.9	1.1
	4	-0.1	-0.1	-1.8	-1.2	-0.6	-1.2	-1.1	-0.5	0.5
	5	-1.4	-1.1	-1.1	-0.5	-1.2	-1.7	-1.7	-1.7	-2.4
	6	-1.2	-1.1	-1.2	-1.9	-1.4	-0.8	-1.2	-1.1	0.1
	7	0.2	0.3	-0.5	-0.7	-0.5	-0.4	0.3	1.7	3.2
	8	2.3	0.6	-0.3	-0.9	-0.9	-0.4	-0.3	0.8	1.6
	9	1.2	0.8	-0.5	-0.7	-0.8	-0.2	-0.1	0.6	1.2
	10	0.5	0.3	-0.6	-0.6	-0.7	-1.2	0	0	1
	11	0.7	-0.4	-0.5	-0.3	-0.4	-1.2	-1.1	0.5	2.2
	12	0.1	-0.5	-1.2	-2.2	-1.1	-1.3	-0.2	2.5	3.7

	Below 1 standard deviation of mean (<37.3%)
	Within 1 standard deviation of mean (between 37.3 and 39.9%)
	Between 1 and 2 standard deviations above mean (between 39.9 to 41.2%
	More than 2 standard deviation above mean (>41.2%)

Figure 2.3.7-1 Difference between Local Weight Fraction Glass Fiber and Mean for a Typical Panel

It should be noted that the greatest deviation in fiber content occurred at the edges of part. These areas were the terminal flow points for material movement.

Plaque and Flow Tool Flow Simulation

During the course of this project, the flat plaque tool and the flow tool geometries were simulated using the CADpress_™software. These simulations were performed to demonstrate the capabilities of the software and determine the degree of orientation that could be expected in the larger bus seat mold. Several simulations are recorded in Appendix A for reference.

2.4 Task III – Tooling Technology

2.4.1 Flow Tool

In order to effectively determine material combination properties a generally flat compression-molding tool was needed. Figure 2.4.1-1 shows the basic design of the tool that was designed and constructed. The lower mold was built upon a base tool that included all necessary cooling and heating features as well as precision alignment guidance for the shear edge, a basic insert, an ejector plate and a base plate. The base tool was made from aluminum. The ejector plate was prepared to accommodate the maximum number of ejector pins needed for some of the inserts. The ejector pins served two functions: ejection of the part and gas venting during the compression molding process.

The base tool was mechanically stabilized against deflection during compression molding with large support posts between base plate and tool base. Stops of different thickness permit adjustment of the plaque thickness. Precision dowel pin alignment guides insure protection of the shear edge against damage from lateral relative movement of the upper and lower mold halves.

Figure 2.4.1-1 Basic Flow Tool Design

The flow tool permitted production of test panels for test specimens and gave an opportunity to examine material flow properties, fiber orientation and fiber distribution on vertical walls.

Figure 2.4.1-2 shows the flow tool. Figure 2.4.1-3 shows some detail views on the tool including the guidance and the secondary insert that can be made from different materials to test thermo mechanical behavior under production conditions. The secondary insert was made from epoxy tooling resin in this case. The resulting part can be seen in Figure 2.4.1-4. The highly filled epoxy casting system, RP 4037 from Vantico, was used for the secondary tooling insert.

Figure 2.4.1-2 The Flow Tool Installed in a Press

Figure 2.4.1-3 Detailed Views of the Flow Tool

Figure 2.4.1-4 Part Fabricated with the Flow Tool and Secondary Insert

Additional inserts are designed, but will be built during the second year of the project. Renderings of the four inserts designs are shown in Figures 2.4.1-5, 6, 7, and 8. To prevent material sticking to the first epoxy secondary insert, it was sealed with a special coating, RenShape-Express Sealer No.1, before using a conventional mold release.

Figure 2.4.1-5 Insert for Test Specimens and Retainer for Testing Tooling Materials

Figure 2.4.1-6 Secondary Insert for Testing Different Tooling Materials

Figure 2.4.1-7 Mold Insert for a Component Geometry with Different Rib Thickness

Figure 2.4.1-8 Mold Insert for a Component Geometry with Cross Ribs

2.4.2 Prototype Tool for Bus Seat

The prototype tool demonstrated the feasibility of using a filled epoxy system as a tooling material for compression molds allowing production of a limited number of demonstration parts. The primary requirements for this system were mechanical and thermal stability. Therefore, a high temperature resin system was specified and the casting system RP4037 from Vantico was selected for this tool.

Machining a master part from a CAD/CAE-data set and then casting the mold, including the pipes for heating/cooling, around the master part, created the prototype tool. The master was dimensional corrected for shrinkage of the molding material. The surface texture was added to the master part before the tool casting operation. Mold alignment pins were included in the metal frame that encased the cast epoxy portion of the tool.

The bus seat was situated in the tool in a V orientation. This orientation reduced the side load and the tension transferred to the tooling material. This shape is also necessary to fit the tool into the press. The necessary open press dimension (daylight) would have been too great for most of the 2000 ton-presses that were available.

Lawton Pattern Company used the model of the seat supplied by Southern and made the final design changes. Once approved by Southern, Lawton revised the model to include the anticipated shrinkage of both the epoxy tooling material and the molded thermoplastic composite material. The master model was used to create the mold models. After approval, Ren BoardÔ material was used to create the master patterns. Surf CamÔ software created the tooling paths for the CNC router from the master model data to machine the master patterns. The patterns were completed and a painted texture was applied to the front surface of the seat. Strong wooden boxes were built around the two master patterns to mold the epoxy tooling to the size recommended for the steel tooling encasement. The master patterns were located and mounted within the open boxes. Appropriately sized copper cooling lines were placed in each mold. Two pass-through devices were cast into each mold for handling purposes. Several coats of wax release were applied to the Ren BoardÔ masters and copper cooling lines. The epoxy was mixed and poured into the mold boxes and left to cure. Approximately 12 hours later, the master patterns were removed from the cast epoxy and the wood boxes were stripped off the molds. Final polishing was performed on the epoxy surface before curing the epoxy during a prescribed time sequence and temperature process. After the curing operation, the molds were sent to Snyder Mold Company and assembled into a heavy structural steel framework. The remaining mold features were added, that is, the ejector pins, machined shear edges, and ejector box. Once completed, the mold was placed in a press for final confirmation of clearances and alignment. A liquid sealer coating was applied to both of the mold surfaces prior to shipping.

The following section describes and illustrates the steps involved in casting the epoxy tooling. In Figure 2.4.2-1 the mold for the seat bottom was being fabricated. The copper cooling channels that were cast into the mold are visible. At this stage two coats of special epoxy coating was brushed on the mold surface and tubing. The purpose of the coating was to insure a smooth, bubble-free surface and prevent the aluminum filler from penetrating the mold surface and firmly bonding to the bulk of the epoxy casting.

Figure 2.4.2–1 View of Lower Mold Half before Pouring of Epoxy Over Copper Cooling Lines

The mold box was progressively filled with the casting compound that contained aluminum powder and aluminum aggregate. The aluminum filler provided improved thermal conductivity, increased the compressive strength and reduced the potential exotherm that occurs during the cure of the epoxy. The materials were pre-measured and mixed with a large drill press as illustrated in Figure 2.4.2-2. The both mold halves were simultaneously poured with time between pourings to permit air bubbles to rise out of the epoxy. The cure time was approximately 2 to 4 hours. At no time was the top surface layer allowed to cure during the pour in order to avoid lines in the surface of the mold.

Figure 2.4.2-2 Mixing Setup Using a Large Drill Press

The Figure 2.4.2-3 illustrates the distribution of poured epoxy in the deeper sections of the mold while the upper portions of the mold were washed with fresh molding compound to keep the surface layer uncured.

Figure 2.4.2‑3 Partially Filled Lower Mold Half

The Figures 2.4.2-4 and -5 show the upper half of the mold during casting. The white plastic pipes were included to provide a space to insert lifting rods through the cured mold halves to place the rough cast tools in the heat-treating furnace and also to position them for machining. The copper tubing was for cooling purposes as in the other mold half. All copper tubing was silver-soldered and leak tested under pressure before any epoxy casting started.

Figure 2.4.2-4 View of the Upper Mold Half Showing Cooling Lines and a Partially Filled Mold Cavity

Figure 2.4.2-5 Upper Mold Half during Filling Stage

Both mold halves were poured to the brim of the mold boxes as shown in Figure 2.4.2-6. It was permissible to pour the final mold layers on a second day as long as none of the critical mold surfaces were exposed. Both mold halves were initially cured at room temperature for 24 hours then carefully transported to a curing oven in which the mold castings were heated up to 180°C over 24 hour period. The long thermal ramp time was needed because of the thickness of the tool and the need to avoid generating internal stresses during the curing process itself. After the curing cycle was over, the mold castings were shipped to the tool and die facility where the edges were machined in preparation for encasement in a steel chase.

Figure 2.4.2-6 End of Mold Pouring with Epoxy Molding Compound Flush with the Top of the Mold Box

After the sides of the mold were machined, the halves were welded in a steel chase. All ejector mechanisms were steel as were all components that would be bolted or clamped in the press. The mold halves were aligned and the final shear edge was machined after the guide pins were located and installed to assure proper registration. The upper portion of the mold is illustrated in Figure 2.4.2-7 mounted in the molding press.

Figure 2.4.2-7 View of the Upper Mold Half after Encasement in a Steel Chase … note the red cooling hoses and welded ribs on mold case

During the first molding trial, the lower portion of the mold fractured under load. The break occurred at the rear-most rib where the cast epoxy was thinnest. Residual polypropylene/glass material was forced into the crack. It was later learned that a small gap was present around the edge of the tool that effectively left the epoxy core of the tool unsupported. Repair of the tool was achieved by infusing epoxy on all fours sides of both mold halves though holes drilled in the steel chase. After the epoxy cured, the mold core was completely supported by the steel frame. The crack in the lower mold half was machined for an aluminum insert without cutting the cooling lines that lay beneath the surface. The aluminum insert shaped like the original part was fastened into the mold with screws.

Figure 2.4.2-8 Photograph of Lower Half of Seat Mold after Molding …note residual material in left-most rib

Figures 2.4.2-8, 2.4.2-9, 2.4.2-10 and 2.4.2-11 show some details of the aluminum insert. The fit was very good and the mold functioned well enough to mold 20 complete parts for testing.

The ejector system was a standard 4 cylinder hydraulic system that effectively ejected the part. The entire tool weight was approximately 8500 kg.

Figure 2.4.2-9 Detail of Aluminum Mold Insert

Figure 2.4.2-10 View of Lower Mold Showing Aluminum Insert and Pre-positioned Reinforcements

Figure 2.4.2-11 Lower Mold Detail with Aluminum Insert

2.5 Task IV – Component Selection

The bus component choice for prototype design and production was based upon several criteria. Cost and weight savings were the two chief considerations. A tabular analysis of the candidate bus components with respect to weight, cost, lifetime and function is summarized in Table 2.5-1.

	Bumper beam	Seat structure (2 passenger)	Flooring	Side impact panel
Existing weight / part	40.7 – 56.8 kg.	21.4 – 27.3 kg.	approx. 8.4 lbs.	n. a.
Existing costs / part	approx. $ 590	approx. $ 400 - $ 625	$ 68.00 / sheet	n. a.
Existing lifetime	n.a. [1]	approx. 12 years (without cushion)	approx. 6 years	n.a. [2]
Parts / bus (average)	2	14	8	12

Table 2.5-1 Candidate Bus Components Contrasted by Cost, Weight, and Function

The following manufacturing criteria were estimated based upon bus production of 500 to 700 units per year:

· Weight saving.

· Cost saving.

· Manufacturing feasibility using LFT materials.

· Low cost prototype tooling feasibility.

A tabulation manufacturing criteria for each component is listed in Table 2.5-2.

The flooring did not show any weight or cost savings. If assembly and replacement costs had been included, there could be a small cost savings in cost. Nevertheless, there was no predicted weight savings as long as the sound damping requirements were needed to maintain the level of comfort.

The side impact panel required a high level of surface finish, similar to a class-A finish expected on passenger cars. There would be some micro-waviness at the surface if standard LFT-materials were used. There was no economical LFT-system available to solve this problem. Therefore, material and process development needed to be done to make this part economically. The impact panel could be addressed in future efforts related to this project.

The bumper beam and seat structures showed the most promising combination weight and cost savings. The calculated cost and weight estimates are shown in Table 2.5-3.

	Bumper beam	Seat structure (2 passenger)	Flooring	Side impact panel
Fabrication approach	Vacuum bag forming	Compression back molding, 2 parts (seat and carrier)	Sheet lamination	Compression molding
Expected material	PP/fiber glass fabric 4:1	LFT-pellets + local reinforcement [3]	PP/GF fabrics and foam	LFT-XX/PP + cover sheet or painting [4]
Expected cycle time	12 min. / part	70 sec./ part	170 sec. / part	45 sec.
Expected weight	approx. 32.7 kg.	approx. 13.6 kg.–19.1 kg.	10.8 kg/sheet	n.a.
Expected costs / part [5]	$ 115	$ 110 (without cushion)	$ 71.50	n.a. [6]
Expected mold costs (production)	$ 30,000	$ 250,000 to $ 350,000 for both molds	$ 25,000	$150,000
Expected final costs / part [7]	$ 308	$ 250 (without cushion)	$ 121	n.a.
Remarks	Low volume	High requirements on flammability	High replacement costs, number of replacements vary with different customers, today use of plywood sheets (3/4” x 5’ x 7.5’), edges and cutouts has to be protected against water, sound damping required

Table 2.5-2 Candidate Bus Component Replacement Contrasted by Cost, Weight, and Function

	Bumper beam	Seat structure
Expected weight savings / bus	35 lbs.	238 lbs.
Expected cost savings / bus	$ 564	$ 2100

Table 2.5-3 Expected Total Weight and Cost Savings per Bus

The seat structure had the highest potential for both cost and weight savings; however, the seat structure required an additional constraint on material properties with respect to flammability, smoke and toxicity. One would also expect more sophisticated tooling in the case of the seat. With these considerations in mind, the seat structure was selected for the prototype component.

2.6 Task V – Component Fabrication

2.6.1 Seat Design Basis

The basic seat dimensions were based upon 'Standard Bus Procurement Guideline', Reference SAE J 826. This guidelines required that a passenger seat frame and its supporting structure be constructed and mounted so that space under the seat is maximized to increase wheelchair maneuvering room, and is also free of under-seat obstructions to facilitate cleaning.

The following seat design description is a paraphrased version of SAE J826 with metric units:

“The structure shall be fully cantilevered from the sidewall with sufficient strength for the intended service, and supported by a pedestal attached to the floor. The lowest part of the seat assembly that is within 305 mm of the aisle shall be at least 254 mm above the floor. All transverse objects including seat backs in front of forward facing seats shall not impart a compressive load in the excess of 454 kg onto the femur of passengers ranging in size from a 5th percentile female to a 95th percentile male during a 10-g deceleration of the bus. Permanent deformation of the seat resulting from two 95th percentile males striking the seat back during this 10-g deceleration should not exceed 50.8 mm, measured at the aisle side of the seat frame at height H. Seat back should not deflect more than 356 mm, measured at the top of the seat back, in a controlled manner to minimize passenger injury.

The seat assembly shall withstand static vertical forces of 227 kg applied to the top of the seat cushion in each seating position with less that 6.35 mm-permanent deformation in the seat or its mountings. The seat assembly shall withstand static horizontal forces of 227 kg evenly distributed along the top of the seat back with less that 6.35 mm permanent deformation in the seat or its mountings.”

Seat Design and Analysis

The seat geometry was created with Pro/Engineer (Pro/E) software. The seat was modeled by the Swept Blend option that generated a single swept profile including the side ribs while maintaining a wall thickness of 5 mm.

Molding Issues / Discussion about Orientation of Seat

Discussions on prototype molding issues were conducted at the C.A.Lawton Company, Green Bay Wisconsin. The original seat design incorporated ribs that were perpendicular to the seat bottom that was consistent with the orientation of the seat in the mold, as it would be installed in the bus. However, because of potential cracking of the metal-filled epoxy mold, the decision was made to tilt the seat back position by 20° to reduce stresses on the mold. Figure 2.6.1-1 shows several views of the seat.

The design was made to permit inclusion of optional hybrid polypropylene-carbon preform reinforcements. The primary structural elements were as follows:

a) Rectangular ribs running the full length of the back to accommodate carbon fiber tape

b) Rounded ribs running to partial profile to accommodate carbon tows.

c) Rounded edges to accommodate carbon tows

d) Ribs at an angle to the bottom to strengthen and stiffen the bottom

The rib thickness was specified as two thirds of the of the 5 mm nominal wall thickness. The depth of the ribs was set as four times the nominal wall thickness. The number and position of ribs was optimized.

Representative analysis for extreme cases (without any support ribs or top and front edge reinforcements) is shown in Figures 2.6.1-2, 2.6.1-3 and 2.6.1-4. In the final design, several ribs were added as indicated and the top and front edge was reinforced. The seat model was modeled as fully constrained at the four bottom support locations.

The following loading conditions were applied to the model:

Case 1: A load of 2500 N (~500lbs) was applied vertically along the top edge of each seat. This was modeled with a distributed load of 2500 N among 33 nodes on each side. This was equivalent to a load of 5000 N is applied on the two seats together. The maximum stress, 108 Mpa, was localized near the junction of the two seats at the top and the maximum displacement was 152 mm (Figure 2.6.1-2a and b).

Figure 2.6.1-1 Views of the Seat Design

Case 2: A load of 2500 N (~500lbs) was applied laterally along the top edge of each seat. A load of 2500 N was distributed among 33 nodes on each side. Therefore, a total load of 10000 N was applied to the two seats. The maximum stress, 216 Mpa, was localized near the junction of the two seats at the top and the maximum displacement was 234 mm (Figure 2.6.1-3a and b).

Case 3: A load of 2500 N (~500lbs) was applied to the seat vertically. A load of 2500 N was distributed among the 225 nodes on each side. A total load of 5000 N was applied to the two seats. The maximum stress for the two seats was 11 MPa and the maximum displacement was 2 mm (Figure 2.6.1-4a and b).

Figure 2.6.1-2 Top Edge Subjected to Vertical Load

Figure 2.6.1-3 Top Edge Subject to Horizontal (lateral) Load

Figure 2.6.1-4 Vertical Load Applied to Seat Base

The physical properties of the ribbed seat for nylon with 40% glass by weight are summarized in Table 2.6.1-1 below.

Volume	8.7 E-03 m³
Surface Area	2.38 m²
Density	1485 kg/m³
Mass	12.92 kg
Center of Gravity with respect to coordinate frame; X Y Z	2.16e-01 m, -1.96e-02m, -6.24e-04 m

Table 2.6.1-1 Properties Summary of Nylon/glass Ribbed Seat Design

2.6.2 Flow Modeling and Analysis

Flow modeling studies were pursued along two areas. First, the component was simulated in a mold flow analysis to evaluate 40% and 50% glass-filled polypropylene under injection-molding conditions. The degree of fill, fill time, temperature of flow fronts, and confidence of fill were modeled. Second, the bus seat was modeled as a long fiber compression molded part with CADpress_™ for Thermoplastics software, and performed studies for fiber orientation, flow analysis, shrinkage, and warpage.

Figures 2.6.2-1 and 2.6.2-2 illustrate the flow patterns for the seat design with oriented ribs using traditional injection molding. The injection location was established at the highest point (the bottom side curvature) of the seat. Table 2.6.2-1 summarizes the assumed molding and material parameters.

Figure 2.6.2-1 Flow Sequence of Injection Mold Long Glass Fiber/polypropylene in Ribbed Seat Design

Polymer, Glass, Weight Fraction	PP, Long Glass Fiber, Wf = 40%	PP, Short Glass Fiber, Wf = 40%	PP, Short Glass Fiber, Wf = 50%
Mold Temperature	40 °C	40 °C	40 °C
Melt Temperature	220 °C	260 °C	260 °C
Max Injection Pressure	100 MPa	100 MPa	100 MPa
Injection Time	7.68 sec	9.55 sec	9.58 sec
Injection Pressure	27.92 MPa	11.18 MPa	12.78 MPa
Viscosity @ 240 °C	165 poise	101 poise	107 poise

Table 2.6.2-1 Summary of Injection Molding and Material Parameters

a) b) c)

d) e) f)

Figure 2.6.2-2 Parameters: a) fill time, b) pressure drop, c) confidence of fill, d) flow front temperature, e) quality prediction, and f) weld lines for glass fibers/polypropylene flow

It can be seen from the Figure 2.6.2-1 that the ribs could be filled adequately and uniformly. The 40% weight fraction long glass fiber of higher viscosity (165 poise) required much higher injection pressures (27.92 MPa) than the 40% and 50% short fiber PP, where the viscosity was 101 and 107 poise respectively. The confidence of fill on the part was adequate for the injection location chosen. The weld lines seen in Figure 2.6.2-2 f were typical for injection molding, but were not expected repeated in the compression molding process.

Compression molding simulation studies were performed with the CADpress_™ for Thermoplastics software. The CADpress_™ software performed two integral processes in the development and implementation fiber reinforced polymers; flow simulation and the determination of mechanical and thermomechanical properties. The first stage of the simulation determined flow behavior during mold filling of the compression molding process in which the flow-induced fiber orientation was developed. The induced fiber orientation determined the mechanical and thermomechanical properties of the part. Based upon the filling and fiber orientation information, process-induced shrinkage and warpage were determined. Figure 2.6.2-3 represents the meshed model of the seat ready for CADpress_™ simulation. Knowledge of flow characteristics, material properties, and mechanical behavior assisted estimates to optimize the seat mechanically while reducing processing and material costs.

The required parameters for input to the CADpress_™ simulation are listed in Table 2.6.2-2.

2.6.3 Manufacturing Concept

The manufacturing concept was based on compression co-molding unidirectional fiber reinforced inserts with 1” long LFT-pellets. The pellets were manufactured with a 40% glass load. The insert was fabricated using the same resin as for the LFT-pellets and

carbon fiber. The insertwas preheated in an oven and placed in the tool. Immediately after insert placement, the LFT charge was placed in the tool and the part was compression molded. The geometry of the plasticated material (charge) was determined by running the model of the flow simulation to get the best calculated mold filling. The bus seat required a plasticated charge formed in three parts. Two of the parts were each half of the full capacity of the extruder. The third charge was manually extruded within 60 seconds of the placement of the other two charges. The charges were approximately 6 inches wide and 1 inch thick as extruded, but grew slightly in dimension during residence time in the mold before compression.

Table 2.6.3-1 lists the primary processing parameters. The 425 second cycle time associated with the epoxy tooling system was longer than expected on a production tool made from steel or aluminum. The longer cycle time needed is the result of the lower heat transfer capability of epoxy versus metal and the resistance to high temperatures over a long time.

Table 2.6.3-1 Production Parameters for the Seat

2.7 Task VI – Component Evaluation

2.7.1 Design Check

Based upon the SAE J826 design guidelines, the seat assembly should withstand static vertical forces of 227 kg (500 pounds) applied to the top of the seat cushion in each seating position with less that 6.35 mm permanent deformation in the seat or its mountings. The seat assembly also should withstand static horizontal forces of 227 kg (500 pounds) evenly distributed along the top of the seat back with less that 6.35 mm permanent deformation in the seat or its mountings. Figure 2.7.1-1 illustrates the seat mounted on a test frame used to perform the latter line load test on the seat back.

Figure 2.7.1-1 Seat Mounted on Test Frame for Line Loading

The line-loading test was performed by applying a 454 kg load (twice that of the prescribed 227 kg) on the back of the seat. The experiment was conducted by placing weights on the seat that was mounted in a load frame by four inserts. It was observed that there was no notable permanent deformation on the seat back thus meeting that design requirement. No visible indication of damage to any portion on the seat was observed. Final failure was realized by insert pullout at the attachment points. The suggested improvement was to use of larger diameter inserts (about 50 mm to 75 mm).

2.7.2 Comparison of Flow Simulation with Prototype

Comparison of the theoretical flow simulation with the actual molding operation was helpful to obtain confidence in the predictive abilities of the CADpress_™ software and also to envision its application in other composite formulations and other parts. More importantly, the choice of charge shape and placement was successfully determined in a simulation environment without the cost of a mold fabrication or the cost of press trials. The bus seat design was simulated with various charge placements and it was found that the part could be filled with a 340 metric ton press force assuming that the charge could be placed evenly over the sitting area of the bus seat.

Comparison of the previously mentioned simulation in Section 2.6, Figure 2.6.2-4, with the actual filling sequence, experimentally determined in Figure 2.7.2-1, showed very similar flow behavior of the trial to the simulation. The areas of particular interest were the upper corners of the seat where the delivery of plasticated material was carried preferentially in larger channels that formed potential voids or would eventually determine weldment areas. Figure 2.7.2-1 also illustrated the potential sensitivity of charge placement and platen leveling effects on the symmetry of the flow front. This particular part needed about 1200 metric tons of closing force to obtain the best material consolidation; however, the actual filling force needed was predicted to be one third of the consolidation force, about 340 metric tons. The actual press force required for filling during the molding trial for the 20 bus seats was 360 metric tons.

The predictive power of the flow simulation model was only as good as the input parameters used. The use of different resins or fiber content completely changes the results of the simulation. It is therefore critical to have accurate knowledge of viscosity versus temperature and shear properties of the plasticated composite as well as the pressure-volume temperature relationships of the resin.

Minor issues did arise regarding the meshing of the model. Meshing choices can contribute to anomalous simulation behavior. Meshing technique is beyond the scope of this work, but should be noted as an important simulation factor.

Figure 2.7.2-1 Actual Seat Molded with a Short Charge …view from the back side of the part

The ribs in the bottom of the seat did exhibit one characteristic that the simulation did not reveal. There were small pockets of gas trapped in the ribs by the charge placement and subsequent compression. These pockets were evidenced by small notches seen in the ribbed structure of the part and are slightly visible in Figure 2.7.2-1. The placement of ejector pins and gas vents could solve these issues in a production tool.

Over all it was determined that the filling predictions and the flow front calculations were helpful and reasonably accurate. Additional experience with this software will be useful in future efforts.

3.0 Future Work

The next year’s efforts will be focused on new material development, processing methods and material characterization. Investigation is planned in the area of testing protocol and sample geometry since it is very critical to the behavior of LFT-materials. The second year efforts will be directed at structural sub-elements of transit buses such as flooring and frame units.

4.0 Technology Transfer

At least two seats will be assembled in a bus for the Altoona-test. This test is planned to be coordinated with NABI during testing their new composite bus and will be at Penn State University facility in Altoona. The results of this test will benefit later production designs.

NABI recommended two bus seat manufacturers for technology transfer. Discussion will start soon to evaluate possible manufacturing of an all-composite bus seat and use it in mass transit applications.

The technology, shown in this project, could be transferred to other projects, e.g. Maglev or Light Rail. It would allow weight reduction as well as minimizing problems resulting from magnetic effects on metal parts.

Appendix A Flow Simulations of Compression Molded Panels

During the course of this study test panels were molded to test processing parameters, material formulation and molding parameters. In addition to direct testing, analytical simulations of the molding process were undertaken to observe the characteristics of molded panels with regard to fiber orientation, material flow and mold pressure. The following simulations illustrate the pressure and flow characteristics predicted with CADpress^™ (The Madison Group, Inc.) software that was specifically designed to model long fiber thermoplastic compression molding. Table A-1(a) and Table A-1(b) are lists of required parameters needed for modeling of the polymer matrix and the fiber with the CADpress^™ software. No discussion of methods to determine parameters will be given in this report. The parameter list is included to underscore the model detail.

MATERIAL DATA:
Material:	Long Fiber Polypropylene
Thermal Conductivity:	0.1700 W/(m K)
Thermal Diffusivity:	0.0800 mm²/s
No Flow Temperature:	150.00 ^oC
Carreau P1:	4149.00 Pa s
Carreau P2:	1.00 s
Carreau P3:	5.99E-01
Reference Temperature:	200.00 ^oC
Standard Temperature:	36.44 ^oC
Volumetric Fiber Content:	0.19
Fiber Interaction Coefficient:	0.14
Length / Fiber Diameter:	1400
Heat Transfer Coefficient (air):	5.00e-6 W/(mm²K)
Heat Transfer Coefficient (mold):	2.00e-3 W/(mm²K)
Poisson's Ratio of Fiber:	0.22
Fiber Modulus:	72000 N/mm²
Coefficient of Linear Expansion (fiber):	5.400e-6 1/K
Poisson's Ratio of Matrix:	0.25

Table A-1(a) Sample Material Data Input in CADpress^™ Program

Matrix Modulus:
P1	1030 N/mm²
P2	0.8
P3	3.00e-5 1/K
Modulus of Elasticity of Matrix:
E 1:	1030 N/mm²
Poisson's Ratio:
Q 1:	3.46E-01
Q 2:	1.890e-3 1/K
Q 3:	-1.542e-6 1/K²
Q 4:	-3.050e-8 1/K³
Coefficient of Linear Expansion:
KAL 1:	34100 1/K
KAL 2:	54800 1/K
KAL 3:	238000000 bar
KAL 4:	30700 bar
KAL 5:	1.17E-07
KAL 6:	1.060e-1 1/K
KAL 7:	2.980e-3 bar
Crystallization Temperature:
KGMT 1:	133.781 ^oC
KGMT 2:	0.023 ^oC/bar
Temperature Dependent Data:
Thermal Conductivity of Fiber:	0.00085 W/(mm K)
Specific Heat Capacity of Fiber:	0.810 J/(g ^oC)
Density of Fiber:	2.580 g/cm³

Table A-1(b) Sample Material Data Input in CADpress_™ Program

Execution of the simulation model required that critical details of press dynamics and mold temperatures be provided. Two simulations, given as illustrations later in this appendix, used a similar set of process conditions although the models represented different geometries. The process variables used in the flat plaque panel and the flow tool are listed in Table A-2.

PROCESS DATA:

Maximum Pressure:	3000.00 kN
Startup velocity:	25.00 mm/s
Closing profile:	[mm]	[mm/s]
	100	25
Connection type:	linear
Melt temperature:	200.00 ^oC
Mold temperature (bottom):	60.00 ^oC
Mold temperature (top):	60.00 ^oC
Idle time:	15.00 s
Ambient temperature:	22.00 ^oC
Limiting Temperature:	75.00 ^oC
Cooling / Curing time:	150.00 s

MATERIAL:
Polypropylene
E-glass fiber
Volumetric fiber content:	0.19
Length / diameter fiber:	1400
PROCESS DATA:

Maximum Pressure:	3000.00 kN
Startup velocity:	25.00 mm/s
Closing profile:	[mm]	[mm/s]
	100	25
Connection type:	linear
Melt temperature:	200.00 ^oC
Mold temperature (bottom):	60.00 ^oC
Mold temperature (top):	60.00 ^oC
Idle time:	15.00 s
Ambient temperature:	22.00 ^oC
Limiting Temperature:	75.00 ^oC
Cooling / Curing time:	150.00 s

MATERIAL:
Polypropylene
E-glass fiber
Volumetric fiber content:	0.19
Length / diameter fiber:	1400

Table A-2 Process Variables for Flat Plaque and Flow Tool Simulation – charge at end

Simulations of a flat plaque and a flow tool panel illustrate the software capabilities and provide general understanding to molding dynamics that operate during compression molding processes. Figure A-1 graphically summarizes a set of pressure relationships for the flat plaque tool as a function of mold closing position, mold opening position, mold velocity, and time. It instructive to note that the pressure buildup time is less than one second since mold closure and compression needs to be rapid in order to avoid freezing of the material before the mold is filled and fully consolidated.

Figure A-1 Pressure Diagrams from Flat Plaque – Charge on End

Figure A-2(a) shows the flow front sequence of the molten composite charge as it moved from the initial charge placement location in dark blue. The time course of the flow front proceeded from blue to red with time. The charge was not perfectly centered as can be seen by slight flow asymmetry; however, maximum nodal pressure seen in Figure A-2(d) is much more sensitive to charge placement and yielded a more asymmetrical pattern with the highest pressure shown in red and lowest in blue. This illustrates the potential relationship of charge placement and the highest levels of consolidation pressure. The fibers on the top front, Figure A-2(b), were more oriented with the rapid flow of the front along the length and the edge of the panel due to race-tracking effects. The lower side, Figure A-2(c), fiber orientation was influenced by the compaction of the material. It is difficult to appreciate the modeling of 5 layers for fiber orientation; however, the variation in coloring provides a sense of potential local variation in mechanical properties due to flow orientation.

The following two figures present compression molding simulation of the flow tool used to test material formulations. The flow tool is a flat panel with two deep ribs and a thick centrally located rib. The flow tool permitted molding of deep ribs and a thick rib to test material flow and orientation. The simulated charge was placed (shown in blue in the flow front diagram) in front of the two deep ribs. The flow front dynamics are illustrated in Figure A-3(a) and progress from the initial charge location to the edge of the panel. There was simultaneous flow of polymer in the ribs and the front of the panel at about 40% flow front. There was a high degree of orientation around the tabs and the region surrounding the deep ribs as can be seen by comparing Figures A-3(b), A-3(c), and A-3(d). In the deep ribs fibers are very oriented along the edges. Local peak pressure is strongly correlated to the initial charge placement as seen in Figure A-4(a). Also substantial local fiber orientation contributed to calculated post-molding deformation. In Figure A-4(b) the

deformation revealed a deflection of 7 mm at the edge of the panel (away from the deep rib region). The deformation calculations clearly illustrate the important of flow orientation of fibers on the magnitude and location of warpage. Initial charge placement has profound effects upon final part tolerance.

The compression molding simulation for long fiber composites demonstrated that the modeling software was capable of assisting in determining press tonnage needed to form components as well as yielding information on part warpage and fiber orientation. The value of simulation is that multiple processing experiments can be made without constructing a physical tool or using valuable press time. The bus seat design was analyzed this way and yielded results consistent with the calculations. Particularly important was the minimum press tonnage required to fill the mold.

Appendix B Cost Analysis Summary for Bus Seat Production

Several cost analyses were performed on the bus seat production starting with the prototype seat fabricated in this effort and including two proposed designs based on experience with the prototype seat. The prototype bus seat was compression mold using 25.4 mm glass fiber with polypropylene matrix. The composite material was 40 percent glass by weight. The basic press requirement was a 1000-ton fast down-acting type. The prototype bus seat material cost and weight data are summarized in Figure B-1.

Figure B-1 Summary of Prototype Bus Seat Fabrication

The materials costs for the prototype parts are summarized in Table B-1.

Material Cost per Part	$27.65
Metal Inserts Fastener	$ 4.00
Material Costs Overhead 10%	$ 2.77

Total Materials Cost/Part	$30.42

Table B-1 Cost Breakdown for Prototype Part

The fractional distribution of costs on the prototype is shown in Figure B-2 include the labor, tooling, and overhead expenses.

Figure B-2 Distribution of Costs for Prototype Bus Seat

There are two feasible alternative designs based on our experience with the prototype seat. The costs for theses two variations are outlined in Figure B-3 with a sale price assuming a 40 percent profit. A comparison of non-labor cost of the three seat designs is displayed graphically in Figure B-4.

Figure B-3 Contrast of Two Feasible Production Versions of the Bus Seat

Figure B-4 Comparison of per Part Non-labor Costs with Three Designs

The cost of the three designs was compared to estimate production cost of the seat currently in use at NABI for commercial application. The relative comparison of cost of the three composite designs with current technology is found in Figure B-5. In all cases the compression molded seat was less expensive to produce and about one half the weight of the current model.

Figure B-5 Contrast of Production Cost per Part of Three Molded Designs with a Calculated Cost of the Current Bus Seat in Use

Tables B-2, B-3(a), B-3(b), B-3(c), and B-4 provide a complete breakdown and summary of the Version 1 design costs noted above. This analysis gives the detail used in the comparison of the seat designs.

Economic Analysis for Bus Seat

Version 1:	Optimized production with local reinforcement, 30% less weight

Material:		LFT PPGF40, local reinforcement
Material costs LFT ($/kg):		$1.76
Weight LFT / part (kg)		7.28
Material costs local reinforcement ($/kg):		$18.70
Weight Insert / part (kg)		0.5
Cycle time (sec)		90

Process:	Compression molding
Equipment:	1000 t press, plasticator, material feeding, manual handling

Table B-2 General Design Parameters for Version 1 Design

Material costs
Material costs/part:			$22.16
Metal Inserts			$4.00
Material costs overhead:		10%	$2.22
Total material costs / part			$24.38

Production Costs
Availability of equipment
Availability			90%
Production days / year			250
Shifts / day			3
Production hours / year			5,400

Depreciation	Calculated at a production time of (in years)		5
Equipment Costs			$1,100,000.00
Depreciation/year			$220,000.00
Depreciation / hour			$40.74
Depreciation / part			$1.02

Table B-3(a) Breakdown of Design Version 1 Costs

Machine costs:
Energy costs/h
Max. power rating (kW)	500
Average:	70%
Cost ($/kW):	$0.07
Energy costs/h:		$24.50
Area costs:
Area (m²):	800
Costs per m²/month	$5.00
Area costs/h:		$8.89
Maintenance (% of equipment costs)	1%
Maintenance costs/h:		$2.04
Machine costs/h		$35.43
Machine costs/part		$0.89

Table B-3(b) Breakdown of Design Version 1 Costs

Labor costs:
Labor costs:	$30.00
Number of workers:	2
Total labor costs/h:		$60.00
Total labor costs/part:		$1.50

Costs of money:
% of ½ equipment costs:	7%
Costs of money/part		$0.18

Total production costs/part:		$3.58

Flame coat and painting/part		$30.00

Mounting bracket		$50.00

Tooling costs
Tool costs		$300,000.00
Average parts per bus	14
Bus/year	600
Live time of tool (in years)	5
Total parts	42000
Tooling costs/part		$7.14

Table B-3(c) Breakdown of Design Version 1 Costs

Cost summary:
Material costs/part:		$24.38
Production costs/part:		$3.58
Flame coat and painting/part		$30.00
Mounting Bracket		$50.00
Total:		$107.96
Overhead costs production:	11%	$11.88
Tooling costs		$7.14
Total production costs/part:		$126.98

Margin:	40%	$50.79
Sales price:		$177.77

Table B-4 Contributing Costs and Estimated Sale Price Design Version 1

Glossary

CADpress_™ Software for mold-filling of long fiber thermoplastic composites, The Madision Group, Inc.

Copolymer Combination of two or more polymers into one molecular chain

Dynatup Trade name for an instrumented falling dart impact test

E glass Common grade of fiberglass used in commodity composites

Falling dart test Impact test defined in ASTM standard D-3029

High melt flow Polymer with melt flow index greater than 50 g/10 minutes

Homopolymer Polymer consisting of similar subunits in the molecular structure

Izod test Impact test defined in ASTM standard D-256

Longitudinal Sample orientation in the direction of predominant fiber orientation

Prepreg Fiber reinforcement that has been intimately combined with specific quantity of polymer resin to be used in molding parts

Random Material with fiber orientation equal in all directions

Size Coating applied to fiber bundles to assist with respect to handling and surface compatibility

Transverse Sample orientation orthogonal to predominant fiber orientation

List of Abbreviations

Blk Black (color)

E Young’s modulus

GPa Gigapascal

HDPE High density polyethylene

LFT Long fiber thermoplastic (composite)

MPa Megapascal

Nat. Natural color

PA6 Nylon 6

PE Polyethylene

PP Polypropylene

PVC Polyvinylchloride

PVT Pressure-volume-temperature (relationship)

Uni Unidirectional (composite)

Unidir Unidirectional (composite)

Uson Ultrasonic

UTS Ultimate tensile strength

[2] Depends on driving behavior, not enough experience (new part)

[3] LFT based on PP or Nylon and glass fiber, local reinforcement might be made from the same resin and carbon fiber

[4] An LFT material for a class-A finish was not developed at this time.

[5] Based on material, labor costs and depreciation for production (5 years). It does not include costs for part assembly, tooling costs or profit margin.

[6] Costs cannot be calculated because the material cannot be selected before development is done.

[7] Includes assembly parts, tooling costs and 40% margin. All calculations based on 600 buses / year and 5 years amortization.

Bumper beam

Bumper beam

2.6.1 Seat Design Basis

Seat Design and Analysis

Molding Issues / Discussion about Orientation of Seat