Research interests and past experience VARIABILITY IN LIQUID COMPOSITE MOULDING TECHNIQUES: PROCESS ANALYSIS AND CONTROL.

Transcription

1 Research interests and past experience VARIABILITY IN LIQUID COMPOSITE MOULDING TECHNIQUES: PROCESS ANALYSIS AND CONTROL Nuno Correia FEUP 17 November 2005 INEGI Composite Materials and Structures Research Unit

2 OUTLINE MY PhD. The RoadLite project Liquid composite moulding (LCM) Variability in LCM Manufacturing repeatability Control techniques MY Current interests Vehicle applications - expertise Control of liquid composite moulding (LCM)

3 ROADLITE FIRST PROTOTYPE

4 SECOND PROTOTYPE Table 1. Trailer cost analysis Chassis type description value k Steel Total 7 to 12 Composite Materials Labour Running gear Total ~ 3 ~ 5 ~ 4 ~12 Table 2. Weight savings on composite trailer Chassis type in total in chassis Steel 4130 kg 1930 kg Composite 3740 kg 1540 kg Weight saving Saving ratio ~ 390 kg ~ 9.5 % ~ 20 %

5 VACUUM INFUSION - VI Mould Preform Distribution medium Vacuum bag Atmospheric pressure

6 COMPACTION IN VI Mould Gate Inlet Vent Fluid hvi hrtm = ( ) f t ( ) f t Lead-lag Outlet Front

7 ADDING VI TO LIMS Permeability (10 11 m 2 ) Random mats 1 ( 1 v ) 3 f K = kc S0 vf Aligned fabrics Fibre Volume Fraction Pressure (MPa) Solid mechanics and flow property models 0.3 Random mats v = v P Fibre Volume Fraction B f f 0 comp Aligned fabrics Error in pressure field < tolerance Pressure solving algorithm LIMS VI

8 VI SIMULATIONS RoadLite DEMONSTRATOR Fig. a. Nodal fill times Fig. b. Fibre volume fraction 1124 compliant shell elements Compliance of 12 layers of Formax FGE 117 μ = 120 mpa.s Fig. c. Thickness distribution

9 VARIABILITY Inlet system Quantity of resin Viscosity Height of container Piping layout Outlet system Pump pressure Leaks Inside the mould Reinforcement permeability Reinforcement density HPM placement Flow brake position Mould geometry Vac. bag folds Inlet positioning and cross section

10 MATERIAL PROPERTIES: VARIABILITY 0.45 Examples of variability in fibre angle, permeability and compliance of textiles HOES, K., DINESCU, D., VANHEULE, M., SOL, H., PARNAS, R., BELOV, E. AND LOMOV, S. (2002) Statistical distribution of permeability values of different porous materials. Proc. of the Tenth European Conference on Composite Materials (ECCM10) June 3-7, 2002, Brugges, Belgium Fibre angle (deg) normal curve experimental data Fig. a. Fibre orientation on a ±45º stitched textile. BTI EBXHD 936 Probability μ Permeability x10-12 m ( K ) = m σ = m 12 2 Fig. b. Permeability distribution for Syncoglas twill woven RE 144/255 (Hoes et al., 2002) Fibre volume fraction v f 60% 50% 40% 30% Compaction pressure kpa Max value Avg value Real data Min value Fig. c. Distribution of observed fibre volume fractions for three layers of FORMAX FGE 117

11 RELIABILITY RTM Monte-Carlo method 4 gates Syncoglas RE 144/255 1xSTD simulations 800 elements However, The Monte-Carlo due to the method large allows number the of elements simulation it is of not a large possible number to use of Monte-Carlo experiments. to estimate These can the then number of analysed simulations in order required to understand for statistical the statistical convergence nature of the results process.

12 TRAILER SIMULATIONS RTM σ Κ = 0.10 K Flow The ideal case is replaced by any number of possible variable scenarios allowing numerical experimentation and reliability characterization Flow fronts Fig. a. Distribution of mould filling time

13 VIRTUAL ON-LINE CONTROL RTM EXAMPLE Controlled injection Table 3. Material data used in simulations Permeability data Fibre v. fraction K σ / K v f RE 144/ m % 50% No control Computer selects between two injection pressure levels (0.3 and 1 bar) at each of the four corner gates in five steps. Results show a high degree of flow front steering when compared with non controlled simulation

14 Conclusions: Expertise Development of LCM models Stochastic modelling: Monte-Carlo Monte-Carlo based control strategies (online control)

15 Conclusions: Expertise Applied research

16 Applied research - Vehicles

17 Applied research - Smart materials ESF S3T: Fibre optic monitored wind turbine blades with PZT actuators

18 Work in progress APPLIED FUNDAMENTAL Vehicles: UUV, UAV, Bus bodies, chassis, trailers. Control (stochastic results and neural networks) Wind energy smart materials Models for the compaction of fibre Magnetic nano-fluids (viscosity control)

19 Possible collaboration INEGI CCM U. Delaware University of Nottingham Europrojects Southfields VT Group Timisoara - Romania