MANUFACTURING OF COMPOSITE PARTS VIA VARTM. D. Heider J. W. Gillespie, Jr. UD-CCM

Transcription

1 MANUFACTURING OF COMPOSITE PARTS VIA VARTM D. Heider J. W. Gillespie, Jr. UD-CCM UD-CCM 00000

2 Report Documentation Page Form Approved OMB No Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington VA Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it does not display a currently valid OMB control number. 1. REPORT DATE 26 AUG REPORT TYPE N/A 3. DATES COVERED - 4. TITLE AND SUBTITLE Manufacturing Of Composite Parts Via VARTM 5a. CONTRACT NUMBER 5b. GRANT NUMBER 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) 5d. PROJECT NUMBER 5e. TASK NUMBER 5f. WORK UNIT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) University of Delaware Center for Composite Materials Newark, DE PERFORMING ORGANIZATION REPORT NUMBER 9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR S ACRONYM(S) 12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited 11. SPONSOR/MONITOR S REPORT NUMBER(S) 13. SUPPLEMENTARY NOTES See also ADM001700, Advanced Materials Intelligent Processing Center: Phase IV., The original document contains color images. 14. ABSTRACT 15. SUBJECT TERMS 16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT UU a. REPORT unclassified b. ABSTRACT unclassified c. THIS PAGE unclassified 18. NUMBER OF PAGES 48 19a. NAME OF RESPONSIBLE PERSON Standard Form 298 (Rev. 8-98) Prescribed by ANSI Std Z39-18

3 Examples Of Current Composite Structures Fabricated Via VARTM Vacuum Assisted Resin Transfer Molding (VARTM) DD 21 Integrated Topside Solutions Heider ONR Workshop - 2

4 Vacuum-Assisted Resin Transfer Molding Schematic And Processing Sequence Vacuum Bag Peel Ply Injection Line Connected to Resin Bucket Distribution Media Breather Preform Vent Connected to Vacuum Port Processing Steps PREFORM MANUFACTURING Precision Laser Cutter Preform Binding Station IMPREGNATION STEP National VARTM Workcell Fully Automated Injection Station CURE STEP Induction Heating Localized Resistive Heating QA/QC Fiberoptic Health Monitoring Modal Analysis Heider ONR Workshop - 3

5 UD-CCM Intelligent VARTM Capabilities I Resin Characterization and New Resin Development Preforming Laser Cutter 3-D Preforms Heider ONR Workshop - 4 High-Performance (3TEX) Binder (Solectria) Complex Shape (Bally Ribbon) SMART-Preforms with integrated weaved sensors Permeability Station 2-D Fully automated 3-D Simulation LIMS Mold Filling Simulat 3-D Liquid Molding Simulation (LIMS 5.0) Analytical Tool for Design Optimization Sensor Flow and Cure Tool-Mounted (reusable) Embedded Bag-Mounted

6 UD-CCM Intelligent VARTM Capabilities II Control and Automation Fully Automated Sequential Injection Flow Rate Control Advanced VARTM Processing RTM-like Parts Surface Quality Dimensional Tolerances Co-Injection Resin Transfer Molding In-Plane Layer by Layer FASTRAC Elevated Temperature VARTM Heider ONR Workshop - 5 Signature Ballistic Fire Structural Fire Structural Tooling Rapid Prototyping Rapid Water Solvable Tooling Reusable Bagging Multifunctional Materials Structural Fire Ballistic Signal

7 Process Design Tool No strong VARTM experience needed to make basic decisions on material and injection scheme MATERIAL DATABASE Viscosity Permeability Infusion Temperature Gelation Time Down Select Database and Design capability can be increased over time P 0 : Injection Pressure y Layer 1:Distribution Layer x Layer 2:Structural Layer πp: Viscosity of Resin SIMULATION K h 1xx 1 π1 h 2 K 2xx π 2 K 2yy D d Heider ONR Workshop - 6 Flow front region h F (x) u F Vacuum Pressure=0 d: region with transverse flow=flow front length D: region without transverse flow=length behind the flow front region u F : flow front velocity PROCESS MODEL Analyze Flow Behavior - Infusion Time - Number of Sequential Injection Lines - Length of Non-Saturated Flow Region GUI AND OPTIMIZATION

8 Automated Permeability Estimation (In-Plane Only) Online Capturing of Flow-Behavior Post-Processing of Image Files Online/Offline Noise Reduction Arrival Time Calculation for all Pixels (1024x1024) Permeability Estimation for Different Distribution Media 5.0E-04 Heider ONR Workshop - 7 Permeability Estimation (2D) for each preform/distribution media (Offline) Future work will incorporate 0.0E+00 online permeability estimation 1 DM1 DM2 Permeability [cm -2 ] 4.0E E E E-04 DM3 DM4 2 3 # of layers

9 Database: Preform Permeability Breather: Airtech Airweave N10 400g/m 2 Random mat: Vetrotex Unifilio g/m 2 Permeability [10-7 cm^2] Kx Ky Kz Breather Random mat Complex Non crimp Woven 24oz Non crimp : 320g/m 2 Complex: Vetrotex Stitchment 2400g/m 2 Woven: Boeing 300g/m 2 Woven: Vetrotex g/m 2 Heider ONR Workshop - 8

10 Added Permeability Data to Database of Commercially Available Distribution Media Distribution Media Comparison Data Courtesy of Gaetan Denis 1.00E-03 Lantor Soric XF Polybeam % Shading Material Colbond 7005 Colbond E E E E-05 Permeability [cm2] 4.76E E E E-06 Database includes now 5 Distribution Media (4 more in progress) Design tool chooses DM based on lead length and flow times Heider ONR Workshop - 9

11 Resin Arrival Times Measured By SMARTWeave Resin Arrival Time of 9 Layer 411-C50 Injection 85cP, 40inch by 6inch Resin Arrival Time of 15 Layer 411-C50 Injection 85cP, 40inch by 6inch Top-Layer [15] Top-Layer [9] Layer 12 Layer 6 Layer 9 Layer 3 Layer SW Node Number Bottom-Layer [0] Layer 3 Bottom-Layer [0] SW Node Number Resin Arrival Time of 30 Layer 411-C50 Injection 85cP, 40inch by 6inch Resin Arrival Time of 40 Layer 411-C50 Injection 85cP, 40inch by 6inch Top-Layer [30] Top-Layer [40] Layer 25 Layer 20 Layer 15 Layer 10 Bottom-Layer [0] SW Node Number Layer 5 Layer SW Node Number Layer 35 Layer 30 Layer 25 Layer 20 Layer 15 Increase of non-saturated flow region with number of layers Resin arrival times increase linearly with number of layers Important VARTM feature ==> Elimination of dry spots during sequential injection with correct opening of injection ports Optimization of injection length (sequential injection) is important to reduce cycle times, especially for thick-section and large-scale composite parts Heider ONR Workshop - 10

12 Influence of Sequential Injection Lines Injection Time Improvement [%] Injection Time Improvement with 20 Sequential Injection Gates Heider ONR Workshop - 11 Injection Time Improvement [%] Injection Time Improvement [s] Injection Length [cm] Injection Time Improvement [s] Injection Time 100cm Preform [s] Injection Time Versus Number of Injection Gates (24oz Woven Fabric, 50% Shading Material, SC-15) hr desired Injection Time (500cm) LOWER BOUND 100cm Preform Injection 500cm Preform Injection 15min desired Injection Time(100cm) LOWER BOUND # of Injection Lines Injection Time 500cm Preform [s]

13 Decisions Required for an Optimal Sequential Injection of Large Parts 1. A minimum number of inlets (lower bound for the number of inlets) are required to assure fill times are less the gel time to ensure complete fill. 2. Increasing number of inlets will reduce cycle time but add cost (additional bagging setup (labor) and hardware requirement, resin waste, etc.). The minimum spacing and thus upper bound for the number of injection lines should be related to the flow front lead length. Analytical studies have shown that the lead length is strongly dependent on the permeability of the distribution media and the preform permeability and thickness. 3. The optimum timing for opening of the sequential injection gate is when no dry-spot can develop under the injection gate. Opening of the injection when the tool surface under the gate is wetted ensures complete wet-out and a minimum penalty on cycle time (optimum opening would be shortly before the tool surface under the gate is wetted out). Heider ONR Workshop - 12

14 Design Tool Demonstration Heider ONR Workshop - 13

15 Design Example: Hull Section BASELINE: Fabric: 42 layers of 24oz. Woven Fabric Resin: DOW Derakane Momentum Resin Distribution Media: 50% Shading Material Part Dimension Hull Section 12 feet by 8 feet by 1 inch Infusion time approximately 30 minutes Center injection scheme will reduce one dimension by a factor of 2 We assume injection along the width of the part Problem reduces to an infusion of a 6 feet part Heider ONR Workshop - 14

16 Design Summary Minimum spacing given by non-saturated lead length ~50cm A maximum of six injection should be used Cycle Time [min] Infusion Time vs. Number of Injection Lines # of Injection Lines If total infusion time should be below 30 minutes then the optimum number of inlets equals TWO (Total of three)!!! Heider ONR Workshop - 15

17 Design Example II: Hull Section BASELINE: Fabric:40 layers of 24oz. Woven Fabric Resin: Applied Poleramics SC-15 Resin Distribution Media: 50% Shading Material Part Dimension Hull Section 12 feet by 8 feet by 1 inch Infusion time approximately 90 minutes (3x Derakane ) Change in resin type results in increase in viscosity but allows longer gelation and infusion time Heider ONR Workshop - 16

18 Design II Change in resin type results in increase in viscosity but allows longer gelation and infusion time Cycle Time [min] Infusion Time vs. Number of Injection Lines # of Injection Lines If total infusion time should be below 90 minutes then the optimum number of inlets equals THREE!!! Heider ONR Workshop - 17

19 Design Example III Hull Section BASELINE: Fabric: New fabric with Twice the Permeability Resin: Applied Poleramics SC-15 Resin Distribution Media: 50% Shading Material Part Dimension Hull Section 21 feet by 8 feet by 1 inch Infusion time approximately 90 minutes Change in fabric type results in increase in permeability decreasing infusion time Heider ONR Workshop - 18

20 Design III Infusion Time vs. Number of Injection Lines Change in fabric type results in increase in permeability allowing faster infusion Cycle Time [min] # of Injection Lines If total infusion time should be below 90 minutes then the optimum number of inlets equals TWO!!! Heider ONR Workshop - 19

21 SMARTWEAVE and SMARTMolding Sensors Low cost sensors measures conductivity of resin SMARTweave (patented by ARL) uses embedded wires, creating nodal measurement points SMARTMolding sensors are tool-mounted Resin arrival Gelation behavior SW Signal SW vs T g log(condt/condtg) = C 1(T-Tg)/((C2+C3Tg)+(T-Tg)) η vs T g log(η T/η Tg) = C1(T-Tg)/((C2+C 3Tg)+(T-Tg)) 6x10-7 Viscosity, η Conductivity (1/ohm-m) 4x10-7 2x10-7 Viscosity (cp) Tg Tg vs α α(tg) =ln(tg/tgo)/(ln(tg/tgo)-cprln(tg/tginfinity)) Degree of Free Conversion, Volume Model α30c SW Data 30C Viscometer Data 30C Free Volume Model 30C Time (Min) 300 Heider ONR Workshop Time (Min)

22 Electric Time Domain Reflectometry (E-TDR) Approach E-TDR is a method of sending a high-speed electrical pulse along a transmission line, and detecting reflections returning from impedance discontinuities within the line. In other words acquisition speed (50 GHz) is so fast that is possible to analyze transition even in short (10mm) electromagnetic circuits. High Speed Oscilloscope Sampling Circuit (20 GHz) Ste p Generator (30ps) Incident E i Sensor Line Z 0, ε eff 2.1 Air Reflection E r Air Epoxy resin Discontinuity Z d, ε eff 3.3 (a) Schematic of the E-TDR technique (a) and equivalent circuit diagram of the transmission line (b). I x (t) L I x+ x (t) L R E x (t) C G R E x+ x (t ) C G (b) Heider ONR Workshop - 21 x x+ x

23 Example: TDR Flow Measurement System Results SC TDR-F sensor response Surface coplanar transmission line gives very high sensitivity and high signal to noise ratio. Voltage [V] Time [ns] Arrival time of E-TDR sensor during level change in the U-shaped tube. Heider ONR Workshop - 22 Level measurements with TDR-F sensor [mm] Comparison of E-TDR and Visual measurements 3mm accuracy 2003 University has of Delaware been All rights demonstrated reserved Visual measurement of level [mm]

24 TDR Sensor Detection of Multiple Flow Fronts TDR Flow Sensor Response During Transversal Flow of the Resin 1st position 1st position Sensor Variation of the Reflected Voltage [V] Last position Last position Time [ns] Time dependant locations of resin in the preform Experimental data showing the movement of multiple flow fronts Heider ONR Workshop - 23

25 TDR Cure Monitoring Dielectric Constant (TDR) On-line Cure Sensor Comparison TDR DC AC-1Hz Time [min] LogLoss Factor at 1 Hz SMARTweave DC-Voltage [V] Change of Dielectric Constatnt (TDR) Correlation of the TDR sensor data (90 F and 180 F) versus DSC data α ( ) t TDR ( ) ε = ε t TDR CURE MONITORING ADVANTAGES 180 F 90 F Degree of Cure (DSC) Accurate on-line cure monitoring comparable to laboratory (FTIR and DSC) test equipment Low cost (flexible circuits can be mass-produced) Multiple sensor configurations for embedded or tool-mounted (reusable) versions are possible Sensing capability through intermediate layers: release agent, gel coat and others Heider ONR Workshop - 24

26 Motivation for an Intelligent VARTM Workcell Current Industrial Practice Prototypes, not production Trial and error High variability No automation, sensing, or control Manufacturing base limited to a few companies with know-how Costs not competitive with traditional approaches No two parts the same Design/Modeling of Infusion Fundamentals of mixing of reacting systems Controlled infusion Sensors Actuators Software Preform consolidation mechanics QA/QC Implementation and Validation of an Intelligent VARTM Workcell Technology Transfer Heider ONR Workshop - 25

27 Intelligent Process Control Intelligent Process Control requires Real-Time Process Evaluation Real-Time Process Simulation Integrated Sensors Maximize Automation Learning Capability Network Capable Modeling and Simulation Sensing and Control Advancement of Liquid Composite Mold Filling Processes by Intelligent Processing Increase production repeatability Reduce trial and error in process design Improve quality Introduce control Experimental Validation and Automation IPC system allows Repeatability Dimensional Control Scrap Reduction Eliminate Post-Inspection Increase Production Rates Process Traceability QA/QC of Process SPC Reduce Touch-Labor Reduce Cost IPC Industrial Hardened Software and Hardware Heider ONR Workshop - 26

28 Schematic of a Fully Automated VARTM Production Cell OBJECTIVES Automated Resin Delivery System Smart Tooling Tool-Mounted Sensors Resistive Heaters for Cure Control Automated Material Handling of Preforms Automated Vacuum Stations Reusable Bagging A Material Technologies Process Technologies Resins B Tooling Vacuum Material Handling Intelligent Processing & Control Intelligent Processing and Control Automation of all Processing Steps Resin Delivery Tile integration Preforms Sensors Heider ONR Workshop - 27

29 SMARTMolding Prototype Cell 1. Manual and AutomaticControl and Monitoring 1. Pinch Valves 2. Vacuum 3. Precision Scales 4. CCD-Camera 5. Tool-Mounted SMARTMolding Sensors 2. Automation 1. Sequential Injection Control with feedback from tool-mounted sensors Heider ONR Workshop - 28

30 SMARTMolding Full Production Cell Heider ONR Workshop - 29

31 Automation, Sensing, & Incoming Material Control Allow Repeat Manufacturing Resin Preform Tiles Periodic Checks Vendor certifies incoming material Other Consumables Barcode Material Handling is semi-automated Material is traced (Witness Photos) Actuators, Sensor REPEAT MANUFAC- TURING Cure Cycle is fully automated Heated Tooling TC for Mold, Surface and Resin Infusion is fully automated Sensor feedback traces repeatability Heider ONR Workshop - 30

32 SMARTMolding Features Automation Features Operator Login Shell Selection via Barcode Automated Tool Selection Automated Vacuum Control (Infusion, Dwell) and Leak Check Controlled On-Line and Off-Line Mixing Supervised Infusion Sensor Feedback from tool-mounted and/or SMARTPad sensors Fully Automated opening/closing of Valves Script Files for Sequential Injection Process Variation Easy to Implement Allows Dwell of Last Infusion Lines Timed Dwell (Reduction in Pressure to 7Hg) is Automated Process and Sensor Information are stored for QA/QC Heider ONR Workshop - 31

33 SMARTPad Sensor Info Injection Line 1 Injection Line 2 Injection Line 3 Shell Demo 5/29/02 SMARTPad Sensor Response 15:00 14:55 Shell Demo 5/29/02 Arrival Time on Sensor Max Diference. Between Sensor Activation = 395s Voltage [V] Injection Line 4 Time 14:50 14: : :35 14:40 14:45 14:50 14:55 15:00 Time 14:35 Sensor Line 1 Sensor Line 2 Sensor Line 3 Sensor Line 4 Sensor Line 6 Sensor Line 7 Sensor Line 8 Sensor Line 9 Two SMARTPad sensors under each injection line was attached to tool New PC board was developed for simple connectorization Sensor response indicates Arrival time at each sensor SPC Optimization of Infusion Scheme Uniformity of fabric permeability Heider ONR Workshop - 32

34 SMARTMolding Cure Monitoring New Tool-Mounted Torlon Sensor New Embedded or Surface Mounted Flexible Circuit Sensor Dwell Starts INFUSION Temperature Sensor Is Wetted Out Sensor Output DWELL Maximum Sensor Response Degree of Cure ~50% Heider ONR Workshop :00:00 AM 9:12:00 AM 10:24:00 AM 11:36:00 AM Monitors resin arrival The degree of cure can be observed up to 50%

35 Automated On-Line Mixing CAV-IHS Shell Infusion Resin Weight During CAV-IHS Infusion 06/21/ Strokes 8000 Resin Weight [gr] Injection Scale Vent Scale 2 Strokes :00 11:15 11:30 11:45 12:00 12:15 12:30 12:45 Heider ONR Workshop - 34 TOTAL Automated Infused Weight: 128lb

36 Sensor Data Review Timing Learning Curve Cycle Time Versus # of Cycles Automation enables repeatable processing times from cycle 1 Individual Time For Each Processing Step Is Recorded Witness Photo Of Tool Fill Is Displayed Vacuum Loss During Vacuum Check Is Shown Request for Operator Comment when Actual Time Step is Larger Than Nominal Heider ONR Workshop - 35 Actual Cycle Time (% of average) Part # 300 Average Cycle Time until Cure = 6 hours Conventional Cycle Time [% of Baseline]

37 Implementation Hardware Software Pinch Valves Vacuum Control and Sensing Leak Check Sensors Temperature Flow Cure Scales for Flow rate Recipes QA/QC Database Graphical User Interfaces Material Lay-up Infusion Cure Control Heider ONR Workshop - 36

38 SMARTMolding Software Suite Design Tool IPC System Data Review Statistical Package Simple Interface, Limited to simple geometries Predicts Flow Times, Lead Length Optimizes # of Seq. Injection Lines Database with Material Properties Automates the VARTM Process Records the processing steps Reporting of collected data Enables statistical analysis Guidance Software to define process recipe Recipe GUI Heider ONR Workshop - 37

39 Database Overview Configuration Tables Recipe Tables Setup (directories, DAQ settings) Material Info (Resin, Fabric, Core) Operator Info Material Sequence Bag and Tool Selection Infusion Information Max. Leak Resin Info (Amount, Type, Ratios) Sensor Setup Seq. Infusion Script (Valves Sensors) Dwell Info (Temperature, Time) QA/QC Tables Heider ONR Workshop - 38 For each part Sensor Feedback Cycle Time

40 Manufacturing Queue Heider ONR Workshop - 39 Allows central administration of VARTM production Enables monitoring of production status

41 Material Lay-up Station Login feature Automatic part selection Recording of cycle time On-line help via work instruction Heider ONR Workshop - 40

42 Infusion Station (b) (c) (a) Login feature Automatic part selection Recording of cycle time On-line vacuum integrity check (Figure b) Allows integration of industrial mixer hardware Sequential Injection automation (Figure c) Timed room-temperature dwell (Figure d) Records sensor feedback (d) Heider ONR Workshop - 41

43 Automated Infusion End A) Minimum Resin Amount Infused B) Net gain into Part below 10gr/min C) All sensors are wetted out Infusion stops when A ˆ B ˆ C = TRUE Heider ONR Workshop - 42

44 Help Through Work Instructions Heider ONR Workshop - 43 Lay-up Includes instructions about lay-up, infusion and staging using HTML MSDS Pictures AutoCAD drawings Video Infusion

45 Report Basic Information Operator Material Sequence during Lay-up with Cycle Time Info Opening/Closing of Valves Weight and Fiber Volume Info Charts Cycle Time for each processing step Arrival Time of Flow Sensors Heider ONR Workshop - 44

46 Infusion History Report a. Start time b. End time c. Process time Heider ONR Workshop - 45

47 IPC Demonstration PEGASUS / CAV IHS High-strength Aluminum Alloys; FSW and VARTM Externally Stiffened Top Plate Fabricated with FSW Composite Nonstructural Armor Wheeled Platform Offer Breakthrough Technologies Resin Weight During CAV-IHS Infusion 06/21/02 Aluminum Alloy Spaceframe Nodes Low-cost Composite Structural Armor Side Plates Strokes Resin Weight [gr] Injection Scale Vent Scale 2 Strokes :00 11:15 11:30 11:45 12:00 12:15 12:30 12:45 SOFTWARE Resin Preform Tiles Periodic Checks Vendor certifies incoming material REPEAT MANUFAC- TURING Other Consumables Barcode Cure Cycle is fully automated Heider ONR Workshop - 46 TOTAL Automated Infused Weight: 128lb Material Handling is semi-automated Material is traced (Witness Photos) Actuators, Sensor Infusion is fully automated Sensor feedback traces repeatability Vehicle Helps Army Meet 2008 FCS Timeline Wheeled Vehicle Designed and Built In Less Than 8 Months

48 Fun to Watch!!! Heider ONR Workshop - 47

49 BETA-Site Technology Transition Sensor Development Surface Warfare Center Division Sequential Injection SMART Molding Automation SMARTMolding Design Tool Heider ONR Workshop - 48 SMARTMolding