Energy Efficient Microwave Systems
Lambert E. Feher
Energy Efficient Microwave Systems Materials Processing Technologies for Avionic, Mobility and Environmental Applications
123
Dr.-Ing. Lambert E. Feher Head Industrial Microwave Technology Institut f.Hochleistungsimpuls- u.Mikrowellentechnik Forschungszentrum Karlsruhe GmbH Hermann-von-Helmholtz-Platz 1 D 76344 Eggenstein-Leopoldshafen
[email protected]
ISBN 978-3-540-92121-9
e-ISBN 978-3-540-92122-6
DOI 10.1007/978-3-540-92122-6 Library of Congress Control Number: 2008942394 c Springer-Verlag Berlin Heidelberg 2009 This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: eStudio Calamar S.L. Printed on acid-free paper 9 8 7 6 5 4 3 2 1 springer.com
Acknowledgement
The author wants to acknowledge and thank for the fruitful collaborations to mutually investigate and realize new approaches for energy efficient industrial systems and technologies that contributed to this book. It turned out to move forward in technology by “getting back“ to quantum theory. I thank alphabetically to Dr. Arnold (FZK), Prof. Borie (FZK), Prof. D¨oring (FZK), Prof. Drechsler and colleagues (IFB, University of Stuttgart), Dr. Erb (Porsche), S. Filsinger and colleagues (EADS), Prof. G¨ornitz (University of Frankfurt) and wife, C. Heim (Airbus), Dr. Hennigsen (BASF), T. Herkner (GKN), Prof. Hermann (Airbus), Dr. Schattschneider (Hexion), Dr. Scherr (DLR), Dr. Szabo (FZK), Prof. Thumm (FZK), R. Tschunke (CPI), R. Wieseh¨ofer (V¨otsch) as well as numerous colleagues of DLR in Braunschweig, Stuttgart and former Fairchild Dornier and those not being stated here. Special thanks to my colleagues Dr. Akhtar, J. Dittrich, M. Huber, S. Layer, Dr. Link, V. Nuss, T. Seitz, Dr. Stanculovic, and C. Z¨oller. At last thanks to my family and friends S. Bubeck, J. Geier, A. Nestl and all others who are encouraging and accompanying this for years.
v
Contents
1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Decreasing Resources Consumption by Airplanes, Cars, and Wind Energy Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1 2
2
Industrial Microwave Sources at ISM Frequencies . . . . . . . . . . . . . . . . . Possible 24.15 GHz Sources and Their Properties . . . . . . . . . . . . . . . . . . . . .
5 6
3
Microwave Heating – Dielectric Properties and Energy Conversion . . Heating – A General Electromagnetic Problem . . . . . . . . . . . . . . . . . . . . . . . Electromagnetic Wave Propagation and Material Interaction . . . . . . . . . . . . Classical Debye Dissipation Model of Materials . . . . . . . . . . . . . . . . . . . . . . Detailed Consideration on Microwave Heating of Water . . . . . . . . . . . . . . . Ionic Contribution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A Non Classical Consideration on Microwave Heating of Water . . . . . . . Synthesis of Technical Macromolecular Plastics . . . . . . . . . . . . . . . . . . . . . . Quantum Representation of the Microwave Effective Electric Conductivity
9 9 10 12 13 13 14 17 19
4
Efficient Microwave Transmission Devices and Measurements . . . . . . . Standard Rectangular Waveguides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Waveguide Antenna Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dielectric Measurements and Material Data . . . . . . . . . . . . . . . . . . . . . . . . . . Dielectric Measurements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
23 23 25 30 31 33
5
Avionic Microwave Anti-/De-Icing Systems . . . . . . . . . . . . . . . . . . . . . . . . Basic Considerations of Icing in Aviation . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microwave Alternative – CAPRI Investigations . . . . . . . . . . . . . . . . . . . . . . Millimetre-Wave Investigations at FZK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Following the CAPRI-Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heating of CFRP Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion and Conclusions to the mm-Wave Experiments for Technological Design Preparations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrothermal CFRP-Airfoil In-Flight Model . . . . . . . . . . . . . . . . . . . . . . . .
35 35 38 39 40 41 43 44 vii
viii
6
7
Contents
Theoretical Basics of Microwave In-Flight Composite Heating . . . . . . . . . . Continuous Anti-Icing System Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Frequency Choice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microwave Source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Primary Waveguide System Components . . . . . . . . . . . . . . . . . . . . . . . . . . Secondary Waveguide System Components . . . . . . . . . . . . . . . . . . . . . . . . Power Requirements and Heating Performance . . . . . . . . . . . . . . . . . . . . . . . Common Aluminum Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Power Reduction with CFRP-Composites . . . . . . . . . . . . . . . . . . . . . . . . . Anti-Icing In-Flight Heating Performance . . . . . . . . . . . . . . . . . . . . . . . . . De-Icing Leading Edge System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
45 49 50 50 52 53 54 54 55 55 57
Processing Technology for Composite Materials . . . . . . . . . . . . . . . . . . . . The HEPHAISTOS-Systemline and Technology . . . . . . . . . . . . . . . . . . . . . . Microwave Process Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fundamental Investigations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Homogeneous Field Applicator Development . . . . . . . . . . . . . . . . . . . . . . Electrothermal Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Temperature Answer of Materials Exposed to a Homogeneous Field – Choice of Frequency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basic Investigations for the CFC Curing Process with Microwaves . . . . . . Anisotropic Electromagnetic Effects for CFRP Heating . . . . . . . . . . . . . . . . Design and Proof of the Modular 2.45 GHz HEPHAISTOS Conception . . Process Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Development of Thermo-Electric Foils . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Systematic Material Investigations for Microwave Processed CFC Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
59 62 66 69 69 73 76 78 80 82 86 89 90
Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
List of Figures
1.1 2.1 2.2 2.3 2.4 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 4.1 4.2 4.3 4.4 4.5 4.6 4.7
Energy efficient microwave heating/processing of large CFC structures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Costs/R&D-efforts overview on heating technologies . . . . . . . . . . . . . . . Design overview of 24 GHz vacuum electronic devices . . . . . . . . . . . . . . Left: power vs. size [W/cm3], right: power vs. weight [W/kg] . . . . . . . . Left: 24.15 GHz tubes and their efficiency, right: accelerating voltage over output power [kV/kW] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spectral density of electromagnetic radiation for matter at different temperatures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dipole structure of water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Real- and imaginary part of the dielectric constant in the vicinity of the eigenfrequency ω = fr . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Measured dielectric properties of liquid water . . . . . . . . . . . . . . . . . . . . . . Front view of a single water molecule, dark blue p-orbitals for oxygen, light blue s-orbitals of hydrogen . . . . . . . . . . . . . . . . . . . . . . . . . . Rotation of a single vaporized water molecule (left) and an example for a water droplet that is absorptive at 2.45 GHz (right) . . . . . . . . . . . . . Enthalpy diagram of a curing reaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . Water, alcohols (R = molecular rest), phenol (Ar = aromat), ether, and related dipole moments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxiran dipole and water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polytetrafluor-ethylene PTFE (Teflon) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rectangular waveguide with TE10 vector field visualization of E(x, y) and B(x, z) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Vector field visualization of E(x, y) and B(x, z) for the first rectangular waveguide modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electric and magnetic slot coupling – the equivalent electric and magnetic dipole is shown in the lower sketches . . . . . . . . . . . . . . . . . . . . . Effective slot coupling and S-matrix coefficients . . . . . . . . . . . . . . . . . . . Optimized microwave reflection of waveguide systems . . . . . . . . . . . . . . Measured reflected amplitude radiating into the water loaded applicator The MUT between two known dielectric materials in a waveguide section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 6 7 7 8 9 12 13 14 16 17 18 19 19 20 23 26 27 28 30 31 32 ix
x
4.8 5.1 5.2 5.3 5.4 5.5
5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 5.15 5.16 5.17 5.18 5.19 5.20 5.21 6.1 6.2 6.3 6.4 6.5 6.6 6.7
6.8 6.9 6.10
List of Figures
Setup for the measurement of dielectric properties for liquids . . . . . . . . . The catch probability depends on the droplet size. All droplets larger than 70 microns will be caught at the surface of the wing structure . . . . Wing underside of the DLR research aircraft DO-228. Note, that parts of the accreted ice could not be removed by the pneumatic boots . . . . . . CAPRI leakage wave applicator . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Actual large aircrafts composite replacements: nose radome, engine cowlings, flaps, wing to body fairing, spoilers, rudder etc. . . . . . . . . . . . . Comparison of the electromagnetic material properties of the GFRP and CFRP honeycomb samples in the frequency range of 22–40 GHz (here time domain) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental 30 GHz de-icing experiment set-up . . . . . . . . . . . . . . . . . . . Melting at the boundary layer GFRP-ice . . . . . . . . . . . . . . . . . . . . . . . . . . Setup for the a CFRP sample . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Temperature development of CFRP sample . . . . . . . . . . . . . . . . . . . . . . . . Principle and geometry for in-flight composite heating . . . . . . . . . . . . . . Electrical field penetration in CFRP-composite . . . . . . . . . . . . . . . . . . . . . Significant temperature reduction in microwave heated CFRP-composite Principle anti-Icing scheme for composite structures . . . . . . . . . . . . . . . . Avionic magnetron design (shown with 10 cm rulers) . . . . . . . . . . . . . . . Models of primary waveguide system components and sizes (top: WR 187 rectangular waveguide, left: tee, right: bend) . . . . . . . . . . . . . . . Shape and size of mode converter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Shape and size of taper . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heating performance (dry air) for Tair =−40. . .0◦ C at two different power levels for leading edge region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heating performance (dry air) for Tair =−40. . .0◦ C at two different power levels for runback ice region . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Secondary waveguide integration within CFRP-section . . . . . . . . . . . . . . De-icing leading edge section with parted strips . . . . . . . . . . . . . . . . . . . . Overview on composite parts for the A380 . . . . . . . . . . . . . . . . . . . . . . . . The “Orca”-autoclave at Airbus Stade, a 33 m long CFC production facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Process flow for composite fabrication . . . . . . . . . . . . . . . . . . . . . . . . . . . . Process cycle for resin infiltration technology . . . . . . . . . . . . . . . . . . . . . . A curved “Large Part” presentation CFRP-structure fabricated with the HEPHAISTOS-CA2 microwave processing system (1.6 × 1.8 m) . . The HEPHAISTOS-CA2 facility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hephaistos, painting ca. 525 BC. He is also known as Hephaestus and Vulcan (Roman). His attributes in iconography include the axe and tongs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HEPHAISTOS development milestones . . . . . . . . . . . . . . . . . . . . . . . . . . Anisotropy of layered CFC composites . . . . . . . . . . . . . . . . . . . . . . . . . . . In-principle set-up for VAP infiltration process (vacuum assisted process) developed by EADS, Germany . . . . . . . . . . . . . . . . . . . . . . . . . . .
33 36 37 38 39
40 40 41 42 42 44 47 48 50 51 52 53 54 56 56 57 57 59 60 61 62 63 64
65 67 68 68
List of Figures
6.11 6.12 6.13 6.14 6.15 6.16 6.17 6.18 6.19 6.20 6.21 6.22 6.23 6.24 6.25 6.26 6.27 6.28
6.29 6.30 6.31 6.32 6.33 6.34 6.35 6.36 6.37 6.38
6.39 6.40
Microwave assisted CFC fabrication flow (EADS Munich) . . . . . . . . . . . Basic polygonal orientation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Comparison of polygonal cross sections . . . . . . . . . . . . . . . . . . . . . . . . . . A conventional standard autoclave is changed by the HEPHAISTOSCA upgrade to a full electromagnetic “cold oven” system . . . . . . . . . . . . Schematic set-up of HEPHAISTOS-systems . . . . . . . . . . . . . . . . . . . . . . . Upgraded small standard autoclave at DLR Braunschweig with an HEPHAISTOS-SA system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HEPHAISTOS-CA1 system (1 m length, 1 m diameter) . . . . . . . . . . . . . Material model and boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . Ceramic like and high electric conducting material temperature profile [◦ C], x-axis: position within the sample [m] . . . . . . . . . . . . . . . . . Temperature homogeneity of CFC composite structure at 30 GHz . . . . . Stationary temperature homogeneity for materials at 2.45 and 24 GHz . THESIS3D-process simulation of a CFRP slab in a homogeneous microwave field at 30 GHz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Heating of a CFRP Slab in a homogeneous microwave field at 30 GHz . Right: Heating of CFRP slab at 2.45 GHz (HEPHAISTOS-CA2), Left: symmetrically curved panel at 2.45 GHz (HEPHAISTOS-BA) . . . Fibre layer and geometry of tensor directions . . . . . . . . . . . . . . . . . . . . . . Anisotropic orientations for composites . . . . . . . . . . . . . . . . . . . . . . . . . . . Experimental volumetric temperature answer of a unidirectional CFRP slab at 30 GHz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Energy density calculated for a CFRP slab (15 × 30 cm) with the DELFI-Code at 2.45 GHz (left) and initial temperature answer in the central plane of the slab (right) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Schematic HEPHAISTOS module (1 m length) with mounted WR340 waveguides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CFRP slab cured in CA2 with sketched locations of DMA test samples Dynamic Mechanical Analysis (DMA) of test samples cured in HEPHAISTOS-CA1 and CA2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microwave vacuum bag set-up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molecular structure of epoxy resin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dielectric properties of RTM6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Elastomer TE-foil assisted CFRP slab processing . . . . . . . . . . . . . . . . . . . Basic monomer of elastomeres . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . RTM6 resin injected test slab fabricated with HEPHAISTOS-CA1 . . . . Prepreg processing: Prepreg 913 system (left) and homogeneously cured sample (40 × 40 cm) after the process (HEPHAISTOS-CA1 System) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A T-profile fabricated with the HEPHAISTOS-CA1 according the VAP infiltration technology and cured at 180◦ C . . . . . . . . . . . . . . . . . . . . Large RTM6 injected CFRP Plate (60 × 20 cm) with integrated steps 2 / 6 / 20 mm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
xi
68 70 70 71 72 72 73 74 75 76 77 78 79 79 80 81 82
83 84 85 85 87 88 88 90 90 91
91 92 92
xii
6.41 6.42 6.43 6.44 6.45 6.46 6.47 6.48 6.49 7.1 7.2 7.3
List of Figures
Ultrasound investigations of the 3 sections of the step plate, thicknesses: left 2 mm, centre 6 mm, right 20 mm . . . . . . . . . . . . . . . . . . 93 Ultrasound C-scan of a charge of 3 standard test slabs . . . . . . . . . . . . . . . 94 Comparison of Celanese pressure tests with different oven systems using the VAP process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Comparison of interlaminary shears strength with different oven systems using the VAP process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Reaction mechanism for microwave cured CFC materials . . . . . . . . . . . . 95 A 40 kg aluminum tool for a VAP injected presentation structure (HEPHAISTOS-CA1) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 The curved VAP presentation CFRP-structure (HEPHAISTOS-CA1) . . 96 The prepreg-presentation structure of Fig. 6.5 ready for curing at FZK (HEPHAISTOS-CA2) on a curved metal tool provided by EADS . 97 The prepreg-presentation structure of Fig. 6.5 after curing – the excellent surface properties are obvious . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Comparison of the specific energy consumption for an identical certified standard process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Consortium of national funded microwave R&D project scheduled until 2010 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 R&D milestones for aerospace HEPHAISTOS-application . . . . . . . . . . . 101
List of Tables
3.1 4.1 4.2 4.3 4.4 5.1 6.1 6.2 6.3
Dielectric properties of water at 2.45 GHz . . . . . . . . . . . . . . . . . . . . . . . . . . Generic relations for TE and TM-waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . Field components for TE10 -mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameters for TE10 -mode in a WR-340 waveguide . . . . . . . . . . . . . . . . . . Measured material properties of different samples . . . . . . . . . . . . . . . . . . . Typical parameters of different magnetrons at different power levels (source: CPI, Palo Alto, U.S.A) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dielectric parameters for anisotropic CFRP components . . . . . . . . . . . . . . Technical data of HEPHAISTOS-CA1 (VHM100/100) and HEPHAISTOS-CA2 (VHM 180/200) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dielectric properties of bagging tool materials at 2.45 GHz . . . . . . . . . . . .
14 24 24 25 34 51 80 84 87
xiii
Biography
Lambert E. Feher was born in New York, NY, on May 6, 1966. He received the Dipl.Phys. degree and the Dr.-Ing. degree in Electrical Engineering from the University of Karlsruhe, Karlsruhe, Germany, in 1993 and 1997, respectively. His diploma thesis concerned theoretical investigations on current neutralization effects of light ion beams, and his doctoral thesis was devoted to computer simulations for the application of millimeter waves for industrial processing of materials. He received in 2008 his Habilitation for teaching at the University of Karlsruhe. From 1989 to 1993, he was a student employee at the Robert-Bosch GmbH, where he was involved in business administration. Since 1997, he has been with the Forschungszentrum Karlsruhe, Institute for Pulsed Power and Microwave Technology (IHM), where he has been involved in the field of microwave materials processing and industrial microwave system design. In 1998 and 1999, he visited the New Jersey Institute of Technology (NJIT), Newark, as a Post-Doctoral Researcher. Since 2001, he has been the Head of the Industrial Microwave Group, IHM. In 2002, he was selected for a scientific-technical excellence trainee program. He holds 20
xv
xvi
Biography
patents, and authored/coauthored 100 conference proceedings and reviewed papers. In 2003, he studied at the German University of Administrative Sciences in Speyer “Science Management”. Dr. Feher has been a member of the AMPERE-committee since 2005. He was the recipient of the 2004 innovation award by the Technology Region of Karlsruhe for the research on HEPHAISTOS technology. In 2006, he was selected to the 10 best for the Future Award by the Federal President of Germany. In 2008, he was awarded the Ricky Metaxas Pioneer Award for Microwave Quantum Interactions of Water.
Chapter 1
Introduction
High mobility and resources consumption today characterize a global, exportoriented economy. Increasing scarcity of resources and energy, however, will influence the cost structure of products, services, locations, and mobility. Provident, innovative developments give rise to new, resources-efficient production processes, systems, and materials. For some years now, innovative research approaches have been pursued to opening up energy-efficient applications with microwaves by technology transfer to industry [1–3]. Technical applications of microwaves have already become a natural part of modern life (e.g. kitchen microwave). Compared to conventional heating methods, microwaves have clear advantages. Microwave heating at 2.45 GHz and 915 MHz has been established as an important industrial technology about more than 50 years ago. The successful application of microwaves in industries, that always has to compete with conventional heating systems, has been reported e.g. by food processing systems, domestic ovens, rubber industry, vacuum drying etc. [4]. A lot of individual technological solutions in industries have been developed without giving a clear pattern for a key strategy to replace conventional established industrial technologies only by technical advantages until now. Proving the increased economical efficiency at the design stage of a novel microwave system can be a quite contradictive and time consuming procedure, because of the lack of direct comparability or knowledge of related parameters even for the working conventional systems. Also its acceptance and success on the global market, if the product is aimed e.g. at the consumer area, depends significantly on local availability and costs of electrical energy, central government policies and emission standards as well as the consumer’s behavior. As energy prices increase in the future, energy efficiency and related costs savings will influence significantly decisions for use of novel innovative microwave technologies. The generation of very homogeneous fields so far has been e.g. a key problem of microwave process technology, while it is an indispensable prerequisite for many industrial applications. In 1997 already, this problem was theoretically solved and the solution patented based on fundamental studies in a PH.D. Thesis [2,5] by the author. Under certain conditions, a hexagonal geometry of the microwave applicator was found to be ideal for the homogenization of the microwave fields. Another important step of technical development was the calculation of temperature L.E. Feher, Energy Efficient Microwave Systems, C Springer-Verlag Berlin Heidelberg 2009 DOI 10.1007/978-3-540-92122-6 1,
1
2
1
Introduction
responses of materials in electromagnetic fields. It was possible to predict, for instance, that process fields with disappearing thermal gradients can be generated in carbon fibre-reinforced composites [6]. For industrial application now, a modular microwave system line HEPHAISTOS with very homogeneous field distributions has been developed. HEPHAISTOS means High Electromagnetic Power Heating Autoclaveless Injected STructures Oven System and refers to the future time- and energy-efficient alternative to autoclave technology (thermal ovens under very high pressure). By the way, already in ancient times, the Greek god HEPHAISTOS was responsible for metal and oven processing. He is identical with Daedalus, famous for his “first mythological flight” and the “pasted wings”.
Decreasing Resources Consumption by Airplanes, Cars, and Wind Energy Plants The ancient problem faced by Daedalus and Icarus is of crucial importance to present aeronautical research, in particular since airplanes and wing components have been made of synthetic lightweight construction materials to an increasing extent. Use of the correct “paste”, or better, plastic material for wings, of course, is of primary significance. After about 80 years of expensive technical development, the potential of improved lightweight construction has been nearly exhausted for metals. In case of carbon fibre-reinforced composites (CFC), however, development is in full swing. Due to their low density of 1.55 g/cm3 compared to aluminum with 2.8 g/cm3 , these materials are particularly suited for lightweight construction. In synergy with the electrotechnical HEPHAISTOS technology, aeronautical research is developing new, lower-cost processes to inject polymer resins as “paste” between dry carbon fibres. Even the production step of resin injection will be accelerated and improved by novel, small, compact microwave injectors in the future. Methods used to save weight in the airplane are also used to reduce weight in vehicles. In the Formula 1, for instance, CFCs are applied in a way that is similar to that in aviation. Considering the wide application of composite materials in the automotive sector, where it is not only focused on synthetic carbon fibres, but also on natural fibres, such as flax and sisal, for car bodies, microwave technology opens up similar use and saving potentials. Apart from technical developments for industry up to large-scale industrial systems, fundamentals are investigated in this work to improve and accelerate the hardening of resins by specific microwave-sensitive materials or nanoparticles. Due to its modularity, the HEPHAISTOS technology is suited for very large systems and high-quality hardening of very large technical composite structures. This is not only relevant to aircraft construction and automotive, but also to sustainable energy generation with large wind energy plants and
Decreasing Resources Consumption by Airplanes, Cars, and Wind Energy Plants
3
Fig. 1.1 Energy efficient microwave heating/processing of large CFC structures
CFC rotor blades. Furthermore, a microwave based de-icing technology was developed for future CFC structures e.g. for wings used in-flight for aviation or blades used for off-shore wind energy blades in the northern hemisphere (Fig. 1.1). Since 2004, the HEPHAISTOS microwave technology has been successfully commercialized.
Chapter 2
Industrial Microwave Sources at ISM Frequencies
At Forschungszentrum Karlsruhe, Germany, the properties of using millimetre waves for a possible industrial use have been investigated as a spin-off project of the magnetic fusion program and gyrotron development since 1993 [8]. Primarily, the interests have been focused first on processing structural and functional ceramics (like low lossy alumina), where unique advantages of millimetre-wave heating and processing where shown as well as novel system design technologies and hybrid heating. To investigate other industrial applications, e.g. heating or processing organic based materials with microwaves a detailed feasibility study on microwave sources [9], their size and efficiency was performed first. In addition, to understand and predict the material dynamics, hot spot formation, thermal runaway phenomenon etc., several computational codes (MiRa, THESIS, DELFI) developed by the author have been used to clear the choice of internationally licensed ISM (Industrial, Scientific, Medical) frequencies for specific technological applications. Comparing e.g. computationally the properties for applicator design of mm-wave and 2.45 GHz microwave systems, significant differences are noticed as a result of their physical appearance. Figure 2.1 shows the competition for micro- and mm-wave technology depending on system costs (consisting of source, applicator, waveguide components, etc.) and R&D-efforts for developing an industrial solution. Mm-wave system components for industrial applications have to be individually designed and developed implying much higher investment and development costs as well efforts for electromagnetic compability compared to standard microwave and conventional technologies on the market. In general, the physical advantages of microwaves compared to conventional heating are very well known: • • • • • • • •
Volumetric heating of materials Selective heating High heating rates Reduction of processing time Increased product quality Processing of new materials Savings on energy consumption “Clean” technology etc.
L.E. Feher, Energy Efficient Microwave Systems, C Springer-Verlag Berlin Heidelberg 2009 DOI 10.1007/978-3-540-92122-6 2,
5
6
2 Industrial Microwave Sources at ISM Frequencies Technological Development High
Research
Mm-Wave Technology
Microwave Technology
Conventional Heating
Low
Industrial Low
High
System Costs
Fig. 2.1 Costs/R&D-efforts overview on heating technologies
From the point of view of standard microwave technology at ISM frequencies (as 915 MHz and 2.45 GHz), the need for using higher frequencies like 24.15 GHz (additional SM frequency) for industrial applications has to be carefully verified with respect to special physical/engineering advantages or to limits the standard microwave technology meets for the specific application problem [10]. Costs evaluation of industrial mm-wave systems have to be competitive, not only to conventional heating, but also to standard microwave solutions. An important point is in addition the availability of appropriate sources in power, size and efficiency. The following comparison of mm-wave system technology and standard microwave technology at 2.45 GHz has been performed for industrial processing and heating of materials.
Possible 24.15 GHz Sources and Their Properties Several specific advantages have been proposed originally for using higher frequency microwaves than standard 2.45 GHz. These are • • • • • •
Enhanced coupling of the materials to the electromagnetic field Potential of increased field homogeneity within smaller volumes More compact and smaller components/applicators Overmoded waveguides for low loss transmission applicable Focusing and targeting of generated beams Optical transmission techniques and mirror systems applicable
A crucial point for new industrial systems is the availability of a set of appropriate sources in power, size and efficiency, as well as components. For research purposes, a 30 GHz gyrotron source has been used at FZK for detailed heating and processing considerations. For industrial use, different types of devices have to be compared first (Fig. 2.2). Anyway, the most common microwave source is the magnetron (e.g. kitchen microwave oven) and the klystron (e.g. radar applications). At very high
Possible 24.15 GHz Sources and Their Properties
7
Anode
27 cm
Wellenleiter Auskopplung
Kavität E e l
kt r o n e n -
Kathode wo
l e k
ca. 10 cm
30 cm
ca. 10 cm 20 cm 100 cm
Magnetron
EIO
Klystron
Permanent-Magnet Gyrotron (30 GHz) IAP
2 kg
6 kg
32 kg
90 kg
200 W
1–1.2 kW
3–4.5 kW
> 10 kW
24 GHz
24 GHz
24 GHz
30 GHz
Fig. 2.2 Design overview of 24 GHz vacuum electronic devices
frequencies, only the gyrotron and EIO (Extended Interaction Oscillator) attract due to their unique high power properties. The following overview compares the standard vacuumelectron tubes (single component) at the “near millimetre” ISM frequency of 24.15 GHz with typical power levels < 10 kW in terms of the tube’s size, weight and efficiency [11]. The design study for the magnetron, EIO and klystron has been performed together with the company of CPI in Palo Alto, U.S.A. on considerations for airborne use of higher frequency devices. Due to this, only a permanent magnet gyroton can be taken into account because of its compact size – these technological sources are built by GYCOM/IAP in Nizhny Novgorod, Russia. Due to their physical interaction principles, the EIO and the magnetron are most efficient for compact applications, where the component size is strictly limited. If one considers the component’s weight related to its microwave power, the EIO concept as well turns out to be the most lightweight source at 24.15 GHz (see Figs. 2.3 and 2.4) [12]. In terms of generated mm-wave power, the gyrotron is the most efficient source at 24 GHz. With a single stage depressed collector efficiencies of 50%-60% can be
200 180 160 140 120 100 80 60 40 20 0
1,60 1,40 1,20 1,00 0,80 0,60 0,40 0,20 0,00
Magnetron
EIO
Klystron Gyrotron
Magnetron
EIO
Fig. 2.3 Left: power vs. size [W/cm3 ], right: power vs. weight [W/kg]
Klystron Gyrotron
8
2 Industrial Microwave Sources at ISM Frequencies
16,0
60%
14,0
50%
12,0 40%
10,0
30%
8,0
20%
6,0
10%
4,0 2,0
0% Magnetron
EIO
Klystron Gyrotron
0,0 Magnetron
EIO
Klystron
Gyrotron
Fig. 2.4 Left: 24.15 GHz tubes and their efficiency, right: accelerating voltage over output power [kV/kW]
achieved. The EIO shows here the lowest performance. Another clear disadvantage of the EIO is obvious on the necessity of high accelerating voltage per power. The gain in compact dimensions and weight is reduced by the need of oversized power supplies and related costs [13]. As a result of this work, current technical parameters of the proposed components at higher frequencies unfortunately do not satisfy industrial requirements for commercial heating also with respect to energy efficiency or special avionic applications–furthermore, detailed comparisons on the electrothermal heating using mm-waves did not show any significant temperature homogeneity advantages for processed materials to microwaves (p. 79–81).The frequency choice is reduced in terms of costs, power availability and need, as well as the geometrical dimensions of the application. Applicable frequencies to be used are proposed for 915–927 MHz, 2.45 and 5.85 GHz.
Chapter 3
Microwave Heating – Dielectric Properties and Energy Conversion
Heating – A General Electromagnetic Problem It is well known, that microwaves can heat materials. An overview on typical sources has been given in the previous chapter. Nevertheless, the uniform processing of different materials raises substantial problems for microwave technology as physically speaking a monochromatic microwave frequency from a source is converted to a thermal radiation noise. This physical process is accompanied in general with a nonlinear thermal heat distribution within the exposed material as a result of the volumetric penetration of the electromagnetic wave. The phenomenon and interaction processes of materials with electromagnetic waves for industrial, technical
Spectral Density 1e–15 rho(f,30) rho(f,100) rho(f,200) rho(f,500)
1e–20
rho(f,1000) Magnetron(f) Licht(f) 1e–25 rho [Js/m3]
Infra(f)
1e–30
1e–35
1e–40 1e+08
1e+09
1e+10
1e+11
1e+12 f [Hz]
1e+13
1e+14
1e+15
1e+16
Fig. 3.1 Spectral density of electromagnetic radiation for matter at different temperatures
L. E Feher, Energy Efficient Microwave Systems, C Springer-Verlag Berlin Heidelberg 2009 DOI 10.1007/978-3-540-92122-6 3,
9
10
3 Microwave Heating
processing issues are complex, diversified and depend primarily on consequently tuned system and control properties, as well as the specific selection on material systems in their interaction to achieve a targeted reproducible temperature development. It has long been known, that an insulating dielectric material can be heated by applying electromagnetic energy. The common explanations are founded due to the ability of the electric field to polarize charges in the material and the obstacle of resulting polarization that cannot follow the high frequency reversals of the electromagnetic wave. For arbitrary materials, polarization effects of charge displacements coupled with direct conduction effects, space charge interfacial effects etc. are taken in appropriate combinations which rely on classical electrodynamics, chemistry and related dipole representations.
Electromagnetic Wave Propagation and Material Interaction To describe the electromagnetic field-material interactions, we consider first and Maxwell’s equations in the common notation for the electric field strength E the magnetic flux density B = μ0 μr j + ε0 εr ∂ E ×B ∇ ∂t ρ = =0 ·E ·B ∇ ∇ ε0 εr
= −∂ B ×E ∇ ∂t
(3.1) (3.2)
A material is described by a set of parameters: the electric permittivity εr , the magnetic permittivity μr and the charge density ρ. We consider physics, where no free charges or plasmas are involved (ρ = 0). Using the vector potential A = −∂ A E ∂t
=∇ ×A B
(3.3)
we describe the field-material interaction in a compact wave equation [15] x , t) − Δ A(
x , t) 1 ∂ 2 A( = −μ0 j 2 c ∂t 2
(3.4)
= 0 for the vector potential can be · A The Coulomb gauge condition ∇ imposed, as well as the case of zero charge densities (no discharges, sparks and plasmas with free charged particles, ρ = 0) [16]. Until now, no physical approximations for the propagating fields as far as the nature of εr and μr is valid have been used. To proceed, the current density j is now approached classically in first order assuming Ohmic-like losses represented by an artificially introduced material parameter, the effective electric conductivity σeff
Electromagnetic Wave Propagation and Material Interaction
j = σeff E = −σeff ∂ A ∂t
11
(3.5)
We remind that the electromagnetic vector potential of a propagating wave can be idealized as a separated time harmonic varying function of space x , t) = A( x )e− jωt A(
(3.6)
In current-free regions, a homogeneous Helmholtz-Equation results (x ) + k 2 (x ) A (x ) = 0 ∇2 A
(3.7)
with the wave number k 2 (x ) =
ω2 μr (x ) ε˜ c (x ) = k02 ε˜ (x ) c02
(3.8)
In the presence of an absorbing medium, the electric permittivity is complex, ε˜ c = εr − jε . To include easily the magnetic permeability, we introduce ε˜ (x ) = μr (x ) ε˜ c (x ) = εr − jεreff
(3.9)
The real part is given by εr = μr εr
(3.10)
and the imaginary part is given by εreff =
μr σeff μr σ + μr εr = ωε0 ωε0
(3.11)
which includes losses due to ionic conduction as well as dielectric relaxation. One defines the loss tangent by tan δ =
εreff ε Im(εr∗ ) = = Re(εr∗ ) εr εr
(3.12)
x , t)and B( x , t) are also time harmonic varying quantities given by For the fields E( = ∇ = B( x )e− jωt × A B
(3.13)
= jω A( x , t) = E( x )e− jωt E
(3.14)
Introducing the Poynting vector 1 S p = E ×B μ
(3.15)
12
3 Microwave Heating
we obtain the continuity equation for electromagnetic energy according to · S p + ∂ ∇ ∂t
2 1 2 1 2 ε E + = − σeff E B 2 2μ
(3.16)
The electric power absorbed in a given volume of a dielectric sample with a specific electric conductivity is then simply given by the expression Pabs = −
1 2
2 σeff E (x) d V
(3.17)
as is well-known from textbooks (for example [16, 17]). These derivations are complete standard and generally used for calculations of dielectric heating. This situation is simple for engineering, because the complex dielectric constant ε˜ (x ) can be easily measured for materials by sophisticated means and techniques (see p. 22, pp. 31–34). It may be mentioned that absorbed microwave fields in materials consist of evanescent properties.
Classical Debye Dissipation Model of Materials The interaction of an electric field with a dielectric has its origin in the response of a charged particle to the applied field. Molecules and molecular groups with dipole properties consist of a positive charge (of nuclei) and negative charge (of electrons) that is not distributed uniformly within their structure. A simple mean to describe the geometrical charge and structural properties is the term of electronegativity introduced by Pauling, which is illustrated in Fig. 3.2 with partial charges δ + and δ − as a result of distorted covalent bonds. The well known classical Debye theory describes the behavior of permanent dipoles in liquids and in solutions of polar molecules in non-polar solvents [14]. In this approach, the rotation of an ensemble of spherical dipoles is considered in a viscous medium with friction. The analysis results in the Debye equations for the real and imaginary part of the complex permittivity [17] ε = ε∞ +
εs − ε∞ 1 + ω2 τ 2
ε =
(εs − ε∞ )ωτ 1 + ω2 τ 2
(3.18)
δ−
ΔEN:1.2 Fig. 3.2 Dipole structure of water
δ+
H
δ−
O 104.5°
Hδ +
2δ +
Detailed Consideration on Microwave Heating of Water
13
Fig. 3.3 Real- and imaginary part of the dielectric constant in the vicinity of the eigenfrequency ω = fr
Transparent
Absorbing
Reflecting
εs represents the d.c. dielectric constant, ε∞ its value at very high frequencies and τ is related to the dipole’s relaxation time (Fig. 3.3). The classic model states, that the best dissipation results if the reorientation polarization fails to follow the applied field which raises the loss factor. For very high frequencies, the dipoles do not feel the impinging wave any more. The Debye mechanism is fairly inaccurate, as e.g. solids and polymers differ significantly in their physical consistence to the idea of dipoles in a non-polar solvent.
Detailed Consideration on Microwave Heating of Water Ionic Contribution Already water as a static dipole does not comply with the Debye approach. As an ampholyte, measurable electric charges are generated in a reaction with itself 2H2 O ⇔ H3 O + + O H −
(3.19)
This autoprotolysis results in a small conductivity of pure distilled water due to the presence of ions. Though, the conductivity of distilled water is about 10–4 S/m, which is not responsible for the microwave conductivity – the measured values for water are about a factor of 10,000 higher. As shown in Table 3.1 the conductivity can be increased by the allowance of ions like NaCl.
14
3 Microwave Heating Table 3.1 Dielectric properties of water at 2.45 GHz [17, 18] Dielectric properties at 2.45 GHz
Temperature [◦ C]
εr
ε˝
σ [S/m]
Ice Water (Dest.) Water (+NaCl) Water (Dest.) Steam, S.C. Phase
<0◦ C
3.18 76.7 67.0 52.0 <5
0.003 12.0 41.87 2.44 Unknown
0.0004 1.63 5.7 0.33 Unknown
25◦ C 25◦ C 95◦ C 374◦ C
A Non Classical Consideration on Microwave Heating of Water Dielectric measurements at microwave frequencies show that the dielectric properties decrease as water is heated and breaks down if the water undergoes the transition from liquid to a vaporized phase as shown in detail for water in the supercritical state (see Table 3.1 for steam), where both phases are available and cannot be distinguished [18]. This observation is contradictive to the simple dipole explanation for an individual water molecule, because the vaporization does not has any change on the polar electronegative character of water caused by the molecular structure in Fig. 3.2, and in conclusion, the temperature dependent decrease of the complex conductivity cannot be explained as well (Fig. 3.4).
Fig. 3.4 Measured dielectric properties of liquid water
Detailed Consideration on Microwave Heating of Water
15
In fact, a quantum mechanical description is proposed here for a more precise understanding. If a molecule absorbs electromagnetic energy E12 , its quantum state |Ψ = c1 |Ψ1 + c2 |Ψ2 + ... =
ci |Ψi
(3.20)
i
is changed from |Ψ1 to |Ψ2 . The change of state can be separated in general into three contributions for rotational energy Erot , oscillation energy Evib and electronic excitation Eel . The frequency f12 of the absorbed or emitted electromagnetic wave is given by E 1→2 = h f 1→2 = E el + E vib + Er ot = E 2 − E 1
(3.21)
The rotational transitions need the lowest energy, which is in the area of microwave and far-infrared, the energy states are quantized according 1 j( j + 1)h– 2 h– with the frequency fr ot = j and j = 0, 1, 2, 3... 2 J 2π J (3.22) to the next excitation state j and the molecules´ moment of inertia J [19]. Excitations at higher frequencies Evib occur in the infrared (10−1 eV), while direct electronic excitations Eel take place only in the Ultra violet regime (10 eV) and need not being considered for microwave interactions. The probability P1→2 for spontaneous emission or absorption for a rotational transition of state |Ψ1 to |Ψ2 is proportional to the matrix element Er ot =
P1→2 ∼ |Ψ1 | xˆ |Ψ2 |2
(3.23)
The radiated power can be shown to be [20] – 1→2 P1→2 = P = hω
4 ω1→2 |Ψ1 | 2e xˆ |Ψ2 |2 3c3
(3.24)
The classical calculations show for the radiated power [16] P=
2 ω4 pDipole 3 3c
(3.25)
which validates the idea of an assigned dipole moment for the field-material interaction according pDipole 2 |Ψ1 | 2e xˆ |Ψ2 |2
(3.26)
This matrix on the eigenfunctions, which represents the non-classical molecular orbital geometry, determines the dipole character in quantum terms. The phys-
16
3 Microwave Heating
ical consequences differ due to the Hamiltonian of the molecule considered in the Schr¨odinger equation ˆ r ot |Ψ = Er ot |Ψ H
(3.27)
= pDipole × E is generated by presence of Classically, a rotational moment M an electric field interacting with a dipole moment. In quantum terms, the energy eigenvalues (without knowing the precise|Ψ ) for a rotation depend primarily on the inertia moment J, that represents the geometrical mass distribution. Er ot =
1 j( j + 1)h– 2 2 J
(3.28)
In quantum terms, the rotation for a single tetragonal water molecule forces in good approximation a rotation around the unbounded oxygen’s p-orbital (Fig. 3.5). But if we estimate now J according the data of the well known orbital models, the resulting frequency in the ground state (j = 0) can be found in the range of about 500 GHz, which is about 200 times the frequency of technical decimeter waves (2.45 GHz). In consequence, a single water molecule does not absorb technical microwaves and should not be heated, which is of course contradictive to the common classical interpretations as pointed out for Fig. 3.2, but the result is consistent for obtained experimental observations for hot vapor mentioned in the beginning of this paragraph. The distinctive difference to understand the heating mechanism of water is to consider a combination of several single water molecules as a compound by dynamically varying smallest droplets. Recent investigations showed [21–23] that about more than 4 water molecules around a core molecule form a dynamic and structurally inhomogeneous networked water droplet by hydrogen bonds in the liquid
→
↑M →
→E →
pDipole
Fig. 3.5 Front view of a single water molecule, dark blue p-orbitals for oxygen, light blue s-orbitals of hydrogen
Synthesis of Technical Macromolecular Plastics
17
Fig. 3.6 Rotation of a single vaporized water molecule (left) and an example for a water droplet that is absorptive at 2.45 GHz (right)
1011 Hz 9
10 Hz phase. This network is rather labile; the water molecules continuously change their neighbors by their hydrogen bonds within a few picoseconds (this experimentally obtained value ν = 1/T = 1011 − 1012 H z corresponds to the quantum ground frequency of 500 GHz demonstrated here). Estimating J with the reduced mass of hydrogen bonded droplet of several single water molecules J=
m i ri2
(3.29)
i
shows a significant reduction in the quantum ground frequency including technical frequencies like 24 and 2.45 GHz (Fig. 3.6). The described absorption mechanism is still elastic – the dissipation for thermal energy is a result, where energy converts due to hydrogen bond interaction of the microwave induced rotating droplets and water molecular exchange to contribute to Brownian motion. We now can understand the behavior of water in Fig. 3.4: as ice is formed, a regular crystal pattern behaves rather transparent for microwaves with low absorption properties. If the ice is melting, the absorption rises significantly due to the formation of microwave strongly absorbing droplets, which convert the compound rotation by the exchange of single water molecules to thermal energy. It can be stated, that the immediate vaporizing of liquid water after melting is only prohibited by the effective action of the additional degrees of freedom of generating hydrogen bonds and interactions. As the temperature for water rises, the droplets reduce and weaken the hydrogen bonds with single water molecules leaving the droplet system and reducing the microwave absorption. As the water vaporizes, the structurally inhomogeneous networked compounds brake down to single water molecules that are fairly not absorptive for microwaves and the overall microwave absorption brakes down again.
Synthesis of Technical Macromolecular Plastics Due to reactions like polycondensation, polymerisation and polyaddition monomers are cross-linked to polymers. In each of these poly-reactions, macromolecules are
18
3 Microwave Heating Activated Complex
Enthalpy Activation Energy
Edukte Monomere Product
ReactionEnthalpy
Polymer
Reaction-Coordinate Fig. 3.7 Enthalpy diagram of a curing reaction
generated that differ in the number of their monomers and the polymerization degree. The polymerization reactions considered further in this work are related on trifunctional polymers, which form three-dimensional structured duroplastics. Thermoplastics, formed in linear chains out of bifunctional polymers are similar in the treatment on microwave processing. Another class of linearly, but weakly linked polymers are elastomers with rubber properties (Fig. 3.7). Due to molecular collisions, monomers and dimers etc. combine to higher order chains. Effective collisions lead via transition states and activated complexes by energy conversion to the formation of new products. In an activated complex, the old bonds are partially abandoned and the new bonds partially linked. The necessity of activation energy WA is a consequence of loosing bonds during a reaction. To begin a reaction, energy to overcome the barrier has to be fed. The thermodynamical relations are represented in the Arrhenius equation, WA
k = Ae− RT
(3.30)
The constant A represents the collision number as well as the orientation of the reactants, R = 1.380658 · 10−23 J/K represents the Boltzmann constant and the WA factor e− RT the fraction of particles within the Boltzmann distribution, which possesses energy at least the activation energy to contribute to the reaction. Based on experimental observations it is expected that microwaves enhance polymer reactions by lowering the activation energy. The collision number can be expected to be increased as well. This can be concluded in analogy to the previous considerations on water heating, if we consider the formal structure of substitution products of water, that are found in polymers. This means, that in the presence of a microwave field, the exposed materials which carry molecular rests of the categories of Fig. 3.8 underlie in addition to the thermal Brownian motion a further degree of freedom because of rotational resonances of a dipolar molecular rest, which can consist of a significant accumulation of massive molecular chains. The molecular rest additionally increases the iner-
Quantum Representation of the Microwave Effective Electric Conductivity
δ+
δ−
δ−
δ−
δ−
O
O
O
O
H δ+
H µ
19
δ+
= 1,8
H δ+
R µ
= 1,6
δ+
Hδ+
Ar µ
δ+
Rδ+
R
= 1,5
µ = 1,2
Fig. 3.8 Water, alcohols (R = molecular rest), phenol (Ar = aromat), ether, and related dipole moments δ−
ΔEN:1.0
δ−
O
CH δ+
ΔEN:1.2
CH2 δ+
δ+
H
O Hδ+
Fig. 3.9 Oxiran dipole and water
tia moment J, which causes the decrease for the quantum ground frequency to be absorptive in the microwave range. A microwave field presence and interaction will result in an enhanced mobility and collision frequency of the rotationally activated molecules lowering the energy barrier and changing properties as thermal conductivity, diffusivity, viscosity and cross-linking of the polymers, which will result as well in a better curing degree. A very important polymer, which will be considered for curing more in detail, is epoxy resin. The oxiran structure of Fig. 3.9 is a basic component of a complex epoxy polymer and a direct substitution product to water. It is obvious, that epoxy resins, phenolic resins and polyester resin have to be good microwave absorbers accompanied by a significant effective conductivity.
Quantum Representation of the Microwave Effective Electric Conductivity As previously pointed out, no direct classic ohmic-like electron transport happens in dielectric materials; the microwave effective electric conductivity of (3.5) and (3.16) is generated by contributions of the rotational deformation of molecular orbitals and rotational resonances of sufficiently massive molecular groups. The following example for Polytetrafluorethylene (PTFE, Teflon) shows that the polarity of a classic permanent dipole (see argumentations p. 12) is generally not responsible for microwave absorption. As the molecular structure of PTFE in Fig. 3.10 shows, a very strong electronegative polarization of the C-F bonds can be realized. But dielectric measurements for PTFE (see Table 4.4) give a very low value for the electric conductivity. This
20
3 Microwave Heating
Fig. 3.10 Polytetrafluorethylene PTFE (Teflon)
δ−
δ−
F
F
δ+
C F
δ−
δ+
C
ΔEN:1,6
F n δ−
is caused by the absence of accessible rotational states, which are hindered by the strongly linearly oriented polymer chains as a result of the strong polarization. If the purity and chain length of PTFE material is decreased, increasing values for the electric conductivity can be observed. We derive now a general representation for the electric conductivity under the – of a photon is absorbed by assumption, that the quantum energy W photon = hω W h 2 = ec f = an orbital electron at the equivalent field strength E¯ equivalent = photon eλ −23 V s 2 2 f within the wavelength λ = c/ f of the exciting wave (using (3.24) 1.37 · 10 m and (3.25)) according dP =
2 1 h2 4 ω4 σ 2 2 f d V = 3 d pDipole 2 e c 3c
(3.31)
Using (3.26), we solve for the electric conductivity and obtain the general relation
σi→ j =
2 16π 2 e4 d Ψi | xˆ Ψ j 2 – d V 3h c
(3.32)
which shows clearly that σi→ j is an intrinsic result of the geometrical orbital density for those quantum states, where a transition i → j is possible. The effective microwave electric conductivity is high, if the number of quantum transition states Ψi |Ψ j in the vicinity of the monochromatic microwave frequency is high: σeff =
all States
σi→ j
(3.33)
i→ j
In addition, a high microwave/thermal dissipation in the material is given if the variety of intra-material energy conversion paths is high which dissipate the absorbed power at continuously higher frequencies leading to a thermal equilibrium according Fig. 3.1. E.g. the microwave electric conductivity for ceramics is low, because it is a result of a higher order photon-phonon interaction with a very limited choice on transition and energy conversion paths in the lattice [2]. The microwave electric conductivity for polymers and water is significantly higher, because of the variety of rotational states and energy conversion choices as described
Quantum Representation of the Microwave Effective Electric Conductivity
21
in the previous paragraphs. For metals, the microwave electric conductivity is very high (∼107 S/m) because of the availability of a large number of collective quasifree electron states. Electrons from a conducting band can absorb the microwave energy. The quasi-free electrons will emit the absorbed microwave power again which causes the elastic reflection from metal surfaces and negligible power dissipation due to the lack of energy conversion paths in the electron gas. Detailed considerations for metals show that similar to ceramics photon-phonon interactions with the lattice result in very small dissipation. For metal powders, the regularity of the lattice (metal grain size small to microwave wavelength) is significantly changed such that the number of conversion paths increases which results in thermal heating, if the metal powder is exposed to a microwave field. In conclusion, the effective electric conductivity σeff can be used as a macroscopic parameter for comparison reasons representing a large number of diversified microscopic energy absorption and conversion paths p=
2 1 σeff E 2
(3.34)
Chapter 4
Efficient Microwave Transmission Devices and Measurements
Standard Rectangular Waveguides Efficient transmission of guided microwave power can be obtained inside hollow highly conducting geometries, where the operating wavelength has to be less the characteristic modal cut-off wavelength. The shape and dimensions of the waveguide determine these parameters. The Helmholtz-equation for the vector potential determines this idealized infinite geometry (x ) + k 2 (x ) A (x ) = 0 ∇2 A
(4.1)
We consider this structure using Cartesian coordinates and homogeneously filled with a dielectric material (usually air) (Fig. 4.1) such that the wave number is given by k 2 = k x2 + k 2y + k z2 = k02 ε˜ (x )
(4.2)
The z-axis points into the propagation direction of the continuing electromagnetic action in the waveguide. It is appropriate for the vector potential, to choose a solution of two independent components, which fulfill (4.1) for z.
Fig. 4.1 Rectangular waveguide with TE10 vector field visualization of E(x, y) and B(x, z)
L.E. Feher, Energy Efficient Microwave Systems, C Springer-Verlag Berlin Heidelberg 2009 DOI 10.1007/978-3-540-92122-6 4,
23
24
4
Efficient Microwave Transmission Devices and Measurements
H H = ∇ ×A E
H = A H (x , t)z A
(4.3)
E =∇ E ×A H
E = A E (x , t)z A
(4.4)
H and H E represented in (4.1) are of scalar type, The resulting equations for E where the index E stands for E-waves (Transverse Magnetic TM) and H for Hwaves (Transverse Electric TE). For simplicity, the waveguide consists of an ideally conducting metal, such that the tangential electric field component can be neglected. The properties and formulas for TEmn and TEmn waves (the indices m, n denote the mode number for the x and y direction as a consequence of the tangential E-field to vanish harmonically at the boundaries, k˜ m,n and λ˜ m,n represent the cut-off wave numbers and wavelengths) are well known from text books (Table 4.1). The TE10 -mode, having the lowest frequency in a particular waveguide, called the dominant mode is technically the most commonly used transmission pattern. We denote the well known components for the electric and magnetic field of this mode (Table 4.2).
k˜ m,n βm.n
Table 4.1 Generic relations for TE and TM-waves
2 mπ 2 = + nπ a b 2 = k 2 − k˜ m.n 0
2ab λ˜ m,n = √ m 2 b2 + n 2 a 2 λ0 λm,n = ˜ 2 f 1− f ω c vp = = ˜ 2 βm.n f 1− f ˜ 2 dω f vg = =c 1− dβm.n f
Table 4.2 Field components for TE10 -mode Ex = 0 E y = Ae± jβ10 z sin Ez = 0
πx a
Hx = ∓ Hy = 0 Hz =
πx β10 π e± jβ10 z A sin ωμ0 a a
πx 1 π e± jβ10 z A cos jωμ0 a a
Waveguide Antenna Design
25
Table 4.3 Parameters for TE10 -mode in a WR-340 waveguide a b λ0 k0 k˜ 1,0 β1.0 vg vp ˜f λ˜ λ1,0 Z T E10
8.6 cm 4.3 cm 12.25 cm 51.3 m−1 36.51 m−1 36.05 m−1 m 2.1 · 108 = 0.7c s m 4.28 · 108 = 1.43 c s 1.74 GHz 17.2 cm 17.42 cm 536.61Ω
An overview on relevant parameters of the TE10 mode in a standard WR340 waveguide for use at 2.45 GHz Microwave technology is given in Table 4.3. Higher order modes cannot propagate in this waveguide due to their evanescent damping.
Waveguide Antenna Design For technical use, a rectangular TE10 mode is appropriate to transmit microwave power from a source to a power dissipating load. The load is situated within an applicator where a waveguide feeds electromagnetic power by radiation. The waveguide walls can be perphorated with radiating elements (RE) for generating appropriate antenna field patterns (Fig. 4.2). The feeding waveguide underlies several engineering criteria • • • •
High efficiency (minimized reflections inside the waveguides), Appropriate choice on radiating elements (RE) Low efforts and costs for the waveguide fabrication System modularity
Slotted waveguide radiating elements fulfill above requirements, because they offer many parameters for the optimization of a microwave heating system. The shape of RE-slots in a waveguide wall has been investigated in detailed numerical studies [29, 30]. The RE-slots have to be designed according the following requirements • • • •
Comparable and normalized slot resistances/admittances Equivalent linear polarization for the RE-slots Easy manufacturing and costs Power handling capability
26
4
Efficient Microwave Transmission Devices and Measurements
E
B
B TE10
E TE 20
E
TE11
B
B
TM11
E
Fig. 4.2 Vector field visualization of E(x, y) and B(x, z) for the first rectangular waveguide modes
The optimum for the RE-slots on physical considerations within a specific parameter range has been a curved “banana like” profile that follows the current density lines in the waveguide wall. For engineering and manufacturing purposes, a rectangular slot shape with rounded edges turned out to be comparable for the required performance but versatile to predict and design its properties by analytical formulas for the waveguide. By proper calculation of these RE-slots location, the efficiency of the microwave system can be significantly optimized for a desired EM field distribution and optimized tuning to a load (Fig. 4.3). The waveguide walls are considered to consist of a very high electric conductivity (ideally perfect conductor). The RE-slots are a discontinuity in these waveguide walls that transmit a power signal to the environment. Due to the skin effect of the guided electromagnetic wave, current densities are induced in the waveguide walls = ρ inserted in the by the tangential magnetic field. Maxwells´ equation ∇ · E ε0 εr ∂ continuity equation ∇ · j + ∂t ρ = 0 yields ˙ = −∇ · j ∇·D
(4.5)
that shows easily the generation of a time dependent electromagnetic field signal. The wall current densities can be immediately calculated by ji = n i × H i
(4.6)
Waveguide Antenna Design
27
p(+) m(+) m(–)
p(–) E
H
Fig. 4.3 Electric and magnetic slot coupling – the equivalent electric and magnetic dipole is shown in the lower sketches
for each wall i = 1..4 with n i the outside normal vector of the waveguide wall and i the tangential field at the interface. H For the inner narrow waveguide walls b, the expressions for the surface current densities are trivial j y = H0 e− jkz z x=0,a
(4.7)
For the broad waveguide walls a the current densities yield ⎛
− cos 0
j = ⎜ ⎝
−ik z πa
πx
sin
⎞
a
πx
⎟ − jk z ⎠ H0 e z
(4.8)
a
Figure 4.4 on next page shows, where significant fields are generated by the cutting radiator elements: The properties of a waveguide can be described in technical terms of impedance, admittance and reflection coefficients. The signals a1 and b1 at Port 1 represent the complex amplitudes of the transversal fields, whereas the signal states a2 and b2 of these fields are detected at Port 2. The transformation of the amplitudes from Port 1 to Port 2 due to inconsistencies and radiation elements is characterized by the elements of the scattering matrix S according
b1 b2
=
S11 S12 S21 S22
a1 a2
(4.9)
28
4
Efficient Microwave Transmission Devices and Measurements
Fig. 4.4 Effective slot coupling and S-matrix coefficients
a1 b1 Port 2 b2
Port 1
a2
It is easily evaluated that S11 = ab11 determines the reflection coefficient at a2 =0 Port 1, and S22 = ab22 the reflection at Port 2. The parameters S12 = ab12 a =0 a1 =0 1 b2 and S21 = a1 represent the transmission coefficients for each section of the a2 =0
waveguide. The reflection coefficient S11 is very appropriate to be accessed by measurements with probes within a waveguide section. The measured impedance is represented by z in =
1 + S11 1 − S11
(4.10)
and the transmitted power from Port 1 to Port 2 yields P2 = P1 (1 − |S11 |2 )
(4.11)
If we think of subsequent ports with constant reflection properties, the equation can be rewritten in infinitesimal form d P = P2 − P1 = −P |S11 |2 dl
(4.12)
P(l) = P0 e−|S11 |
(4.13)
with the solution 2
l
This is valid for an infinite long waveguide with constant damping properties. The real waveguides are in contrast finite and consist of changing properties to fulfill the global reflection optimization, which results for the power transport to be described realistically in the waveguide according P(l) = P0 e−2α(l)l
(4.14)
Waveguide Antenna Design
29
with α(l) representing the attenuation function of the limited waveguide. For an efficient power transmission, the normalized impedances zn =
1 + S11 |n 1 − S11 |n
(4.15)
have to be optimized in each section n for an entity of N sections or RE-slots to avoid overall reflecting properties as pointed out in the case of exponential decay. This is possible due to the condition z in =
N
z n + z shor t ≈ e j2π = 1
(4.16)
i=1
All sectional impedances interfere in the complex plane to a unit circle ofe j2π . The impedance zin can be expressed as well by the measured normalized impedance of the first section or RE-slot that contains the full properties of the subsequent following discontinuities z in =
z 1.in + i tan(β L) ≈ e j2π = 1 1 + i z 1.in tan(β L)
(4.17)
The admittances for the RE-slots are straightforward represented 1 =1= gi z in i=1 N
yin =
(4.18)
and expressed by the individual conductances of each RE-slot. To calculate analytically the offsets of each RE-slot from the broad wall waveguide centreline, an implicit formula for the conductances of longitudinal slots is evaluated for each xn [32] gn =
2.09aλ1,0 cos2 bλ0
π λ0 2λ1,0
sin2
πx n
a
(4.19)
For transversal RE-slots, the resistivity determines the individual slot properties 1 =1= ri yin i=1 N
z in =
(4.20)
which is analytically represented for RE-slots in the broad wall by rn =
480 λ31,0 cos2 73 4π 2 λ0 ab
π λ0 4a
sin2
πx n
a
(4.21)
30
4
Efficient Microwave Transmission Devices and Measurements
Inclined RE-slots are as well possible, but not considered for these feeding waveguides, because the polarization properties of the inclined RE-slots interfere and a pure linear polarized radiation is preferred.
Experimental Verification S11 parameters (reflections inside the waveguides) and S12 parameters (mutual coupling between the waveguides) of the RE-slotted waveguides were measured. The measurements were performed under the idealization that the waveguides radiate into free space and for the realistic case to radiate into a large loaded cavity (HEPHAISTOS-CA2). A vector network analyzer (VNA) has been used for these low power tests. The gain to a commonly optimized waveguide system (lin2) is depicted in Fig. 4.5, where the reflectivity represented by the S11 -Parameter could be minimized to less than 3%, –30.5 dB (lin1) with a pure vertically polarized radiation into free space. To check the capability of these high efficient devices for technical use, the waveguides have been mounted on a large cavity chamber (HEPHAISTOS-CA2) with water load. The measurements on the loaded HEPHAISTOS system showed identical levels for the reflected amplitude inside the waveguide. The good agreement for the experimentally obtained results is shown in Fig. 4.6.
Fig. 4.5 Optimized microwave reflection of waveguide systems
Dielectric Measurements and Material Data
31
Fig. 4.6 Measured reflected amplitude radiating into the water loaded applicator
Dielectric Measurements and Material Data The measuring of dielectric parameters is essential to consider and evaluate properties on coupling and transmission of microwave exposed materials [38]. In industrial processes, the complex behavior of materials combinations can only be understood in their response taking all different contributions of the complex dielectric constants into account. Various kinds of resonant and non-resonant methods are available to measure the dielectric properties of materials [36, 37]. To measure material parameters of solid materials and liquids at the technical ISM frequency of 2.45 GHz, a WR-340 waveguide for obtaining the reflection and transmission coefficients was chosen. For measuring high conducting solid materials like carbon reinforced composites (CFC), a novel technique has been applied [39, 40, 42]. The solid materials under test (MUT) are placed into a centre section of the waveguide setup, which can be represented as a three-layer medium, where the regions 1 and 3 are filled with a medium of known properties, and the centre region 2 represents the MUT of a known thickness (Fig. 4.7). The overall reflection and transmission coefficients of this three-layer media can then be computed by representing each layer (junction) with a wave-amplitude transmission [T] matrix. The advantage of using the transmission matrix representation is that the [T] matrix of a number of cascaded sections can simply be determined by multiplying the matrices of the individual sections together [Toverall ] =
T11
T12
T21
T22
= [T0 ] T0,1 [T1 ] T1,2 [T2 ] T2,3 [T3 ] T3,0
where the subscript 0 corresponds to the air medium.
(4.22)
32
4
Efficient Microwave Transmission Devices and Measurements
Fig. 4.7 The MUT between two known dielectric materials in a waveguide section [41]
The propagation constant for each layer i is represented by βi =
k02 εi∗ − kc2 = k0 ε˜ i , i= 0, 1, 2, 3
(4.23)
with ε˜ as the effective complex permittivity (see pp. 10–12) according ε˜ i = εi∗ −
c 2a f
2 (4.24)
The impedance of each layer is easily calculated (Z0 represents the impedance of free-space) Z0 Zi = √ ε˜ i
(4.25)
The individual T-matrices for the layers can be expressed in terms of the effective complex permittivity and the propagations constant [39]. For linking measured values obtained by the VNA measurements with the computational approach, the scattering matrix S (pp. 27–31) is essential and can be expressed in terms of the overall T matrix. T12 T11 T22 − T12 T21 S11 S12 1 = (4.26) [Soverall ] = T22 S21 S22 1 −T21 The overall reflection and transmission coefficient of this three-layer medium is then determined from the overall transmission matrix using an equivalent circuit theory approach by iterative adaption of the computed values to the measured values minimizing the errors [39]. A first computational step employs an analytical approach to obtain approximate values for the material properties. The second step
Dielectric Measurements
33
uses a least square optimization algorithm to determine accurate values of the permittivity and conductive values. If both the amplitude and phase of the reflection and transmission coefficients of the MUT are measured accurately, then the frequency dependence of the material properties can be determined as well [40].
Dielectric Measurements The VNA is first calibrated at the coaxial reference plane using the AutoCal kit supplied by the manufacturer. After calibrating at the coaxial plane, the coaxialwaveguide adapters are connected at the two ports of the Network Analyzer, and the MUT is placed in the core section of the waveguide connected between the two ports to measure the reflection and transmission coefficients. The relative permittivity and conductivity of a number of dielectric samples have been determined using these methods. The reconstructed dielectric properties have been compared with values for samples of known dielectric properties, and a good agreement was found. For natural materials (e.g. wood), the values can largely differ due to their water content and composition of basic materials. Measurements on the temperature dependence of liquid materials have been realized by preheating the material in a glycol bath that was adjusted to the targeted temperature. The procedure for determining liquid material data is similar to the procedure described in the previous section – geometrical corrections have to be considered due to the
Fig. 4.8 Setup for the measurement of dielectric properties for liquids
34
4
Efficient Microwave Transmission Devices and Measurements
Table 4.4 Measured material properties of different samples Material
Permittivity εr
Conductivity σe [S/m]
Teflon PVC PE Alumina Wood GFRP CFRP
2.04 2.91 2.51 9.55 3.40 3.80 2530
0.00090 0.00400 0.00004 0.02000 0.09950 0.00510 28.3000
use of a cylindrical glass equipment. Temperature dependent data of the MUT up to 200◦ C could be obtained by this method (Fig. 4.8). An overview on dielectric parameters for several different solid materials is given in Table 4.4. Polyethylen and Teflon turned out to have a very low electric conductivity, PVC as another polymer is quite high in comparison. A glass-fibre reinforced composite (GFRP) shown in Table 4.4 is rather similar to PVC. The difference to carbon fibre reinforced plastics (CFRP) is outstanding, as the permittivity and the conductivity are extremely increased. Anisotropic properties of CFRP are measured and discussed on pp. 80–82.
Chapter 5
Avionic Microwave Anti-/De-Icing Systems
Basic Considerations of Icing in Aviation For aviation, a suitable alternative for currently used in-flight anti-/de-icing technologies for today’s aircrafts with metal structures and future aircrafts with replaced composite structures is necessary. During the two years Brite/Euram CAPRI-Project (Civial Aircraft Protection against Rain and Ice, 1990–1992), some fundamental aspects of ice protection were investigated. The program included theoretical and experimental work, in which the development of a 2.45 GHz microwave system was carried out at a fundamental technological level [44]. The following investigations and developments performed at FZK have been together in collaboration with the aircraft manufacturers of Airbus and Fairchild Dornier to develop a new avionic technology. A detailed description of the components for prototype development and testing of related systems according to JAR (Joint Aviation Requirements) is included. The full system integration has been tested and will be visualized in this work. In-flight ice formation on airplanes is one of the most critical and actual problems for civil aviation [45]. Ice accretion on surfaces depends physically on the water droplet temperature (Outside Air Temperature, OAT), the content on liquid water in the cloud, the droplet size, aircraft speed and the horizontal extent of the icing cloud. The droplets tend to remain in the liquid phase at temperatures as low as −40◦ C (Super Cooled Large Droplet, SLD). Then, the ice is formed due to the impingement of activated ice nuclei contained in droplets on the aerodynamic surface (see also pp. 14–17 on the quantum physics of liquid water and droplets). The number of activated nuclei reduces continuously from 0 to −40◦ C to zero, where no more icing takes place. The highest potential for icing is therefore at temperatures from 0 to −10◦ C, which is frequently met at altitudes lower 22,000 ft. (approx. 7000 m). Most commuter and regional aircraft are operated below this altitude [46]. The ice formed on aerodynamic surfaces adversely affects the flight behavior. The following effects may occur: • Reduced laminar air flow/increased air resistance • Reduced lift L.E. Feher, Energy Efficient Microwave Systems, C Springer-Verlag Berlin Heidelberg 2009 DOI 10.1007/978-3-540-92122-6 5,
35
36
5 Avionic Microwave Anti-/De-Icing Systems
• Increased weight • Increased fuel consumption • Disruption of the air flow (stalling) at low speed (take-off, runway approach). Two categories of methods against icing exist • Anti-Icing: Prevent ice formation due to a preheated surface (continuously working system during ice encounter) • De-Icing: Removal of ice that had been allowed to build up to a specific extent, from the protected surface. The typical profile of accreted ice formed on a leading edge can be seen in Fig. 5.1. Several trajectories of impinging droplets are shown. Due to elastic properties of freezing water, “run back ice” can emerge at more rear positions of the leading edge [46]. As this problem is essential for flight safety, various conventional, standard anti-/de-icing methods are in use: 1. Anti-/De-icing with hot air: The hot air generated by engines and taken from the compressor is led to the endangered positions via a pipeline and valve system. 2. De-icing with liquids: De-icing liquid is taken from a reservoir and led to the endangered positions via pipelines, pumps and valves made of porous sheet metal. 3. Electric anti-/de-icing: Ohmic heating of anti-/de-icing mat directly at the endangered position. 4. De-icing with pneumatic systems: Accreted ice is removed by inflatable boots.
Fig. 5.1 The catch probability depends on the droplet size. All droplets larger than 70 microns will be caught at the surface of the wing structure
Basic Considerations of Icing in Aviation
37
In addition, better approaches for in-situ ice detection (ice evidence probe), meteorological prediction and avoidance of air space with high risk on ice encounter, better training of pilots as well as the development of advanced exiting procedures have to be carefully considered to reduce serious in-flight incidents. For example, conventional hot air de-icing methods for aluminum structures, which are primarily in use for larger aircraft, are characterized by very high energy consumption (low efficiency) during the flight. In addition to advanced turbines, novel materials and composites have to be used in order to reduce the weight and, hence, the fuel consumption considerably. Moreover, these composite materials have a far worse thermal conductivity than metals. This is associated with the risk of strong local overheating and possible delamination. Consequently, an increased technical expenditure of small pipeline systems with a smaller efficiency and increased weight would be required for homogeneous heating of the composite structures. Using instead electric heating equipment for composite materials leads to a similar situation. As an alternative, mechanical systems are currently used (Fig. 5.2). For turboprop powered aircraft, the available hot air is not sufficient and mechanic systems such as pneumatic boots are used for protecting the structure. Unfortunately, the de-icing pneumatic boots are regarded as unreliable (potentially small contents of icing water entering the boots by malfunctioning the system) and inefficient for ice removal (danger of ice bridging if operated too early) by many pilots, who would prefer a continuously working anti-icing system.
Fig. 5.2 Wing underside of the DLR research aircraft DO-228. Note, that parts of the accreted ice could not be removed by the pneumatic boots
38
5 Avionic Microwave Anti-/De-Icing Systems
Microwave Alternative – CAPRI Investigations Preliminary studies to the CAPRI investigations indicated that ice absorbs much less microwave power than water, making a radiating 2.45 GHz microwave de-icing system impractical due to high power levels and radiation into the environment. As a result, it was possible to demonstrate the physical basic feasibility, but not the technological realization of such a concept for aircraft ice protection. Figure 5.3 shows the system set-up of the 2.45 GHz microwave system. The system was built as a leakage wave applicator with a dielectric composite panel at the leading edge section, where a magnetron source (2.45 GHz) feeds a rectangular waveguide for coupling the microwave power through the composite to the ice/water layer outside. The results of this approach are not satisfactory. It was shown, that most of the microwave power passes through the water and is radiated into free space. Further work (changed illumination, improved horn, field stirring) did not result in any improved effectivity for ice protection. For a pure ice layer, it was measured with a different experimental set-up, that a thickness up to 10 cm absorbs negligible amounts of microwave power. The project committee drew the following conclusion concerning a direct water/ice heating by radiation: “Microwave Anti-Icing does not appear feasible at a frequency of 2.45 GHz. A system operating at higher frequencies (up to millimetre range) could result in much more efficient heating of the water layer.”
Fig. 5.3 CAPRI leakage wave applicator
Millimetre-Wave Investigations at FZK
39
Millimetre-Wave Investigations at FZK The increasing requirements to be met by large aircraft manufacturers, such as • • • • •
Increasing number of passengers Larger ranges covered Reduced fuel consumption and emissions Increased flight safety Reduced operating costs
make it necessary to develop novel high efficient avionic technologies. In collaboration with Airbus, considerations on de-/anti-icing systems using millimetrewaves as a consequence to the CAPRI-investigations have been performed. For basic experiments at FZK, a 30 GHz gyrotron system was used first (Fig. 5.4). In addition to advanced turbines, novel materials and composites have to be used in order to reduce the weight and, hence, the fuel consumption considerably. The use of extremely lightweight materials in the area of the fuselage, tailplane and the outer wing etc. is considered to be the key for the substitution of heavier conventional structural materials. Materials that may replace the previously applied aluminum structures are CFRP/GFRP (Carbon Fibre Reinforced Plastics/Glas Fibre Reinforced Plastics). By an increased stiffness of these composite materials, the aerodynamically optimum form of large wings can be achieved. But, composite leading edges cannot be not used until now in aviation, because conventional heating systems for de-/anti-icing overheat and delaminate these structures. Several CFRP/GFRP material samples have been investigated which were made available by Airbus. These samples consisted of two pairs of identical, hexagonally laminated honeycomb material samples of variable size and absorption properties.
Fig. 5.4 Actual large aircrafts composite replacements: nose radome, engine cowlings, flaps, wing to body fairing, spoilers, rudder etc.
40
5 Avionic Microwave Anti-/De-Icing Systems
Following the CAPRI-Approach In an absorption chamber, the CFRP/GFRP composites were subjected to low-power measurements in the frequency range of 22–40 GHz in order to compare the dielectric properties of these material structures. The different behaviors of the GFRP material and the CFRP material are obvious from Fig. 5.5. Here, the average signal (of 22–40 GHz) returning from the sample is plotted as a function of time. As far as the GFRP sample is concerned, practically the complete signal is returned (0 dB attenuation). A small reflection maximum only indicates the thickness of the material. In contrast to this, the CFRP material is a strong attenuator. Attenuation of the measured signal is about 30 dB. As a result, especially the GFRP material is suited for a direct de-icing by a leakage wave applicator through the airfoil structure in a similar approach according to the CAPRI-project. Further millimetre-wave de-icing experiments showed the feasibility of this approach. For this purpose, a hexagonal metal shielding plate with a hole in the centre was inserted into the hexagonal 30 GHz millimetre-wave applicator (see Fig. 5.6). The composite samples were located above the hole (HF-outlet window).
Absorbing Material, Time Domain
Permeable Material, Time Domain
Shh
Shh 0
0
dB –20
dB–20
–40
–40
–60
–60
–80
–80
–100
–100 –120
–120 –2
–1
0 t [ns]
1
–2
2
–1
0 t [ns]
1
2
Fig. 5.5 Comparison of the electromagnetic material properties of the GFRP and CFRP honeycomb samples in the frequency range of 22–40 GHz (here time domain)
Metalic screening plate with circular HF-outlet window
Gaussian Beam 30 GHz
Fig. 5.6 Experimental 30 GHz de-icing experiment set-up
Following the CAPRI-Approach
41
Temperature Development at the Ice Rim 20
Ice
Ice melted
Heating of water 10 0 Ice water
T [°C] –10 –20 –30 –40
0
100
200 t [s]
Fig. 5.7 Melting at the boundary layer GFRP-ice
The samples were covered with water ice that was cooled by liquid nitrogen and separated from the samples by a thin glass/plastic skin. Figure 5.7shows the temperature development of such a de-icing experiment, where the radiation directly couples to an ice load placed on the surface of the GFRP composite according Fig. 5.6. The experiments were performed to study in detail the melting in the area of the ice/composite surface layer. For this reason, a thermocouple was frozen onto the ice in the composite transition region. The holding temperature of the compound material again was specified to be 80◦ C. The mean heating rate of the composite material was about 50◦ C/min. After about 120–130 s, the absorption behavior changes abruptly (jump of loss tan δ) due to the onset of ice melting and increasing heating of the ice water. The heating process is shown in Fig. 5.7. The ice layer breaks and is lifted off from the composite material on a warm, absorbing water layer. It could be proven by this experiment, that the millimetre-wave radiating de-icing mechanism is much more effective than at 2.45 GHz (note the CAPRI-project conclusions p. 38).
Heating of CFRP Composites The same set-up has been used for heating of carbon fibre reinforced structures. The 10 kW gyrotron at 30 GHz served as well as the radiation source (Fig. 5.8). It was demonstrated in these experiments that 1. Homogeneous heating of honeycomb samples (CFRP, GFRP) to 60◦ C without local overheating or material damage.
42
5 Avionic Microwave Anti-/De-Icing Systems
Fig. 5.8 Setup for the a CFRP sample
2. The use of high initial heating rates (instantaneous heating of the material). 3. The holding of the operating temperature specified for the sample. 4. A direct millimetre-wave heating (according the CAPRI and FZK-GFRP experiments) of the ice by transmitted radiation is not possible due the high absorptive properties of CFRP-materials. In Fig. 5.9, a typical temperature profile for heating is shown. An initial heating rate (60◦ C/min) and a holding temperature (60◦ C) over 5 min were given. A slight transient effect which is due to the control after having reached the holding temperature and to cooling following the switching off of the field after 5 min can be
Heating Experiment Absorbing Sample#2 70
0
100
200
300
400 500 Thermo1 70
60
60
50
50
40
40
30
30
T [°C]
20
0
100
200
300 t [s]
Fig. 5.9 Temperature development of CFRP sample
400
20 500
Discussion and Conclusions to the mm-Wave Experiments
43
noticed. After having been subjected to the mm waves, the samples themselves were found to be neither modified nor damaged. The ice melting here is only indirectly caused by heat transfer of the millimetre-wave heated CFRP-slab [47].
Discussion and Conclusions to the mm-Wave Experiments for Technological Design Preparations The CAPRI set-up used GFRP-like composites for the leading edge section. The experiments aimed to heat compact ice formed on the surface. This approach is questionable on several issues. The coupling to ice and GFRP composite was proven to be very poor at 2.45 GHz. Most of the power is radiated by transmission into the environment. According their proposal in the project conclusions, the electromagnetic coupling to the formed ice could be found to be enhanced with millimetrewaves. The CAPRI leakage wave applicator at 2.45 GHz was not designed in an appropriate manner. The dimensions have been chosen just for radiating through the waveguide and transmitting through the GFRP composite into the environment. Power adjustment for small, thin layers of water or thick ice on the composite is therefore not possible and, as the experiments showed, therefore not achievable by changing or adding any other microwave components (enhanced horn, mode stirrer etc.). In addition for real composite structures, a metal mesh for lightning protection on the surface is prescribed that makes the leakage wave concept obsolete. In the CAPRI project, no CFRP composite investigations have been performed and considering the leakage applicator approach, no radiation through CFRP structures is applicable. CFRP composites have been investigated first at FZK for anti-/de-icing with microwaves/millimetre-waves. As the experiments with CFRP showed at FZK, the electromagnetic behavior and heating properties/mechanisms are totally different to GFRP composites, as well as the later described designs of applicable technological systems. Graphite fibres are among the strongest and stiffest composite materials being combined with matrix systems for high performance structures like slats, leading edges, stabilizers, tailplanes etc. Their outstanding design properties are their high strength-to-weight and stiffness-to-weight ratios. Using their specific structuredynamical advantages, substantial anti-/de-icing problems occur, if those relevant in-flight parts are intended by a manufacturer to be used for novel aircraft leading edges or turbine in-let designs. Today, conventional in-use anti-/de-icing methods for aluminum airfoils are characterized by very high energy consumption. The inefficiency is primarily the result of losing most of the power along the tubular pipeline system until the remaining hot air reaches the endangered structures. The situation is more severe for CFRP/GFRP composites heated using hot air. CFRP/GFRP composites consist of an extremely bad thermal conductivity compared to aluminum. This is associated with the consequence of significant thermal
44
5 Avionic Microwave Anti-/De-Icing Systems
gradients leading to strong local overheating and possible delamination. Therefore, conventional thermal heating is not applicable for aerodynamically relevant structures. Advanced anti-/de-icing methods will have to meet the following requirements: 1. Small energy consumption, preventive effect. 2. Small system weight. 3. High effectivity, reliability and no interference with the electronic aircraft equipment (fly-by wire) and the radio/navigation systems by outside radiation. For flight-safety purposes, the price of microwave devices for a working de-icing system is not of primary importance. Millimetre-wave devices at the ISM frequency of 24.15 GHz have been severely considered on their applicability – the design proposals considering relevant issues such as waveguides and mirrors and related manufacturer tolerances, power supplies, structural system integration, weight, efficiency etc. showed the frequency choice had to be significantly reduced. As it turned out in brief, an ISM frequency of 5.85 GHz is suitable in all manners for the currently considered in-flight application.
Electrothermal CFRP-Airfoil In-Flight Model For quantitative calculations, we build up a model containing the electrodynamical and thermal interactions to determine the heating and power contributions for a metallic screened airfoil. Figure 5.10 shows a small one dimensional segment of a dielectric composite structure of thickness xa covered by a lightning protection (metallic mesh) on the “outside” environmental surface. The metal mesh is essential due to aviation requirements to cover the composite fuselage against lightning strikes as a Faraday cage. Microwaves are generated within this shell containing the CFRP composite that is endangered for outside ice-accretion and has to be heated.
Interior
Fig. 5.10 Principle and geometry for in-flight composite heating
Composite
Metal Sheet
Microwave
Airflow 0
xa
Theoretical Basics of Microwave In-Flight Composite Heating
45
The microwaves penetrate the composite from the interior for a certain extent. A specific electromagnetic field pattern penetrates volumetrically the composite and generates a thermal heat flux to the outside surface. Determined by thermal diffusion in the laminate, a thermal signature of the heating field distributes on the surface keeping it free from accreting droplets or removing formed ice, if the temperature is high enough at the interface. Due to the high electrical conductivity of CFRP and the presence of a metallic lightning protection (copper mesh), no radiation is enabled to be transmitted to the environment.
Theoretical Basics of Microwave In-Flight Composite Heating CFRP/GFRP composite materials siders their mechanical, thermal be described by tensors. As an ductivity σˆ T in the orientation layer
are in general anisotropic materials if one conand electromagnetic properties. They have to example we write the diagonal thermal conof the fibre direction of a single laminated ⎛
⎞ σT ⊥¯ 0 0 σˆ T = ⎝ 0 σT ¯ 0 ⎠ 0 0 σT ⊥
(5.1)
The tensor notation for the electric conductivity σˆ etc. is similar. The determining equation for the essential anti-/de-icing power needs as well as for the structural composite temperature distribution depending on all relevant parameters in the in-flight situation is denoted in the stationary state ∂t∂ T (x , t) = 0 by Poissons´ equation − ∇(σˆ T · ∇T (x )) = pe f f (x , T )
(5.2)
The source term of the power density pe f f (x, T ) is not trivial and contains all source and loss terms according (microwave heating, thermal radiation loss, loss by convection, loss by ice load/meteorological condition)
d A 4 x )σˆ E( x ) − εk B T 4 (x) − Tair pe f f (x, T ) = 12 E( d V Slat d A −h air f low (T (x ) − Tair ) + pmeteo d V Slat
4 dA 1 4 2 = σT ⊥¯ E x (x) − εk B T (x) − Tair 2 d V Slat d A −h air f low (T (x) − Tair ) + pmeteo d V Slat
(5.3)
46
5 Avionic Microwave Anti-/De-Icing Systems Relevant Parameters σT Therman Heat Conductivity σe f f Electrical Conductivity ε Emissivity h air f low Heat Coefficient pmeteo Meteorological In-Flight Situation
Tambient
Temperature of Airflow
x¯ )|2 | E(
Electrical Field
In the case of microwave anti-/de-icing, an electromagnetic field solution of the heating electric field within the composite structure and its environment has to be known. The related solution is found by solving Helmholtzs´ equation for the electric field ˆ2 E Δ E (x) + K (x) = 0
(5.4)
ˆ 2 = kˆ 2 represents the frequency as well as the electromagnetic The tensor K properties of the anisotropic laminated materials. We use the orthogonalized tensor directing parallel and perpendicular to the orientation of the fibre ⎛
ω2 ∗ ε¯ c2 ⊥
ˆ2 = ⎜ K ⎝ 0 0
0 ω2 ∗ ε c2 ¯ 0
⎞ 0 ⎟ 0 ⎠ ω2 ∗ ε c2 ⊥
(5.5)
by the complex permittivity ∗ . rij denotes the electrical permittivity of the material and σij its corresponding electrical conductivity determined by the orientation of the fibres as above. 2 σ 2 j arctan ωεσikε ω σ ik 2 ∗ 2 2 ˆ ij = 0 r ik δ δk j = k0 εr2 ik + 2ik 2 e K εik δk j = k0 εr ik − j kj 2 c ωε0 εr ik ω ε0 (5.6) Solving the full coupled electromagnetic and thermal equations is a rather complex procedure. For this study, we just refer to the obtained solutions for the electromagnetic heating pattern and the temperature answer according the boundary conditions in Fig. 3.10. E(0) = E 0 e jϕ0
E(xa ) = 0
(5.7)
The solution for the electromagnetic heating field within the dielectric composite structure can be calculated ⎤ cos2 (mβ (2xa − x) + mϕ) e2mδxa + cos2 (mβ (2xa − x) − mϕ) e−2mδxa − ⎥ ⎢ 2 cos(2mβ(xa − x)) cosh(2mδxa ) − cos (2m (βxa + ϕ)) e2mδxa − ⎥ ⎢ ∞ 2mδxa ⎥ ⎢ 2 − cos (2m (βxa − ϕ)) e−2mδxa + 2 ⎢ cos (2m (βx a + ϕ)) e ⎥ E (x) = αm ⎢ 2 2mδ(2xa −x) 2 −2mδ(2xa −x) ⎥ cos + cos − + mϕ) e − mϕ) e (mβx (mβx ⎥ ⎢ m=−∞ ⎦ ⎣ cos (2mβ (xa − x) + 2mϕ) e−2mδ(xa −x) − cos (2mβ (xa − x) − 2mϕ) e2mδ(xa −x) − 2 cos (2mβxa ) cosh (2mδ (xa − x)) + 2 cos(2mϕ) + cos (2mβ(2xa − x)) + cos(2mβx) (5.8) ⎡
Theoretical Basics of Microwave In-Flight Composite Heating
47
With αm =
Em 2 cos(2mβxa ) − 2 cosh(2mδxa )
(5.9)
The obtained solution is an exact solution for Helmholtz’s equation, taking the full complex electromagnetic properties (5.5) into account. The penetration of the microwaves into the laminate for volumetric heating can be calculated and the patterns of field profiles for different phase states can be examined and compared. In the same way, the maximum essential field strengths for heating are estimated. The full field solution is used in the next step to solve the stationary heat equation (5.2). The boundary conditions include convection (air flow), thermal radiation from the surface and a potential ice load by taking into account the thermal material parameters of CFRP composites. Depending on the anti-/de-icing power density (power per area), the one dimensional temperature signature (profile) within the composite can be calculated, as well the temperature that is achieved within the interior of the screened structure. Figure 5.11 depicts the electrical field patterns at a microwave frequency of 5.85 GHz for different phase states φ0 within a laminate of typical xa =3 mm diameter. The corresponding geometrical representation of the microwave, airflow, lightning protection etc. is visualized in Fig. 5.10 (x = 0 relates to the laminate surface within the shell, xa =0.003 m relates to the environmental surface, the microwaves penetrate from x < 0, the air streams enters from x > x). 5.85 GHz Heating Pattern in the Airfoil 800 700
E2 [V2/m2]
600
Averaged Field 0 1/2 π π 3/2 π
500 400 300 200 100 0 0,0000
0,0005
0,0010
0,0015 x [m]
Fig. 5.11 Electrical field penetration in CFRP-composite
0,0020
0,0025
0,0030
48
5 Avionic Microwave Anti-/De-Icing Systems
The red line shows the penetration and profile of the averaged field
1 E (x) = 2π 2
2π
2 E (x, ϕ0 )dϕ0
(5.10)
0
About one half of the airfoils diameter (0.15 mm) is basically volumetrically heated, the remaining volume is dominated by thermal heat transfer to the right surface exposed to the environment, where a constant thermal gradient is built up. Considering the solution, the penetration pattern related to the phase state corresponds to the reflectivity of the CFRP depending on the impinging wave from the waveguide. To get an optimal power deposition in the composite structure, the radiated and excited modes within the shell have to be well tuned to the CFRP load. This can be achieved by a very efficient design of the coupling waveguide in a multimode applicator approach. The electromagnetic heating pattern deals as a nonlinear, volumetric power source located within the composite. As the model showed, the temperature maximum is definitely situated within the structure, but the potential peak temperature effect is negligible small. Despite, the reaction and variation on different phase states for the temperature answer is very small as well. As a very important result, conventional heating by conductive thermal transfer from the inside and heating with a volumetric contribution differ consequently for the maximum temperature peak to be achieved, when an equivalent power flux through the structure and boundary conditions are considered. The model shows
Temperature T(x) for Πmeteo = 46 kW/m2, Ta = 35°C 130
Delamination of Carbon Reinforced Material
120 110 100
Conventionally heated
T [°C]
90 80 70
Significant temperature reduction with microwaves
60
Microwave heated
50 40
Surface Temperature of Slat 35°C
30 0,0000
0,0005
0,0010
0,0015 x [m]
0,0020
0,0025
Fig. 5.12 Significant temperature reduction in microwave heated CFRP-composite
0,0030
Continuous Anti-Icing System Design
49
a significant temperature reduction of 40◦ C (in example in Fig. 5.12, geometry according Fig. 5.10, heating field according Fig. 5.12) along the cross section of the composite structure. If it is conventionally heated, the slat is likely to delaminate as pointed out in the introduction of this chapter. We can conclude that internal temperature gradients are significantly and sufficiently reduced using microwaves due to the volumetric penetration of the power into the structure reducing the dimension in x, where the low thermal conductivity of the composite usually builds up a high thermal temperature gradient [48].
Continuous Anti-Icing System Design The technological design presented, as an example for the DO-328, comprises highly increased efficiency for advanced aircraft. The microwave anti/-de-icing system can become a fully integrated avionic system situated within the airfoil structure. The benefits of lightweight carbon reinforced materials can be combined with the unique heating properties of higher frequency microwaves. The objective of the following design is to demonstrate detailed properties and advantages of for higher frequency microwave technology • • • • • • • •
System design for a novel microwave anti-/de-icing technology Power requirements (Icing conditions according JAR 25 App. C) Weight requirements System integration System control Reliability of the components, maintainability Electronic Interference Development program for prototype
The system has to be suitable for future demands and developments in aviation and can be adapted as well for off-shore windmill rotor blades. The microwave system in principle is depicted in Fig. 5.13. The system is closed, fully electromagnetically screened with all compact components internally mounted and spatially truncated to the leading edge structure (see “System Integration” for technical realization). A thin metallic lightning protection layer covers the complete laminated composite structure reflecting internally the microwaves as a resonant microwave applicator. Due to the high electrical conductivity, the carbon reinforced structure is heated volumetrically from the inner side by the absorbed microwaves. No radiation leaves the stabilizer such that no interference with avionic systems is possible. The system consists of a compact switch mode power supply, a lightweight 5.85 GHz microwave source (magnetron), a waveguide system and distributed temperature/ice sensors. The magnetron launches the microwaves into the primary waveguide system. This waveguide splits within the vertical tailplane to feed the secondary waveguide system for the horizontal tailplane (left and right). Each coupling waveguide with a length of 3 m is oriented parallel to the leading edge of the
50 Fig. 5.13 Principle anti-Icing scheme for composite structures
5 Avionic Microwave Anti-/De-Icing Systems Anti-/De-Iced Structure Secondary Waveguide System
TE11 - Mode
Primary Waveguide System
TE10 - Mode Magnetron
stabilizer containing a radiating slot pattern for contact less heating the potentially ice accreted nose. Several design approaches depending on the choice of the frequency, waveguides, excited mode, mode competition, power efficiency, high frequency components, system weight, etc. have been investigated. As a result, the following design implies the most feasible and effective approach for a straight avionic integration [49].
Frequency Choice The microwave anti-icing system is designed to work at an international free ISM (Industrial, Scientific, Medical) frequency of 5.85 GHz (freespace λ = 5.13 cm). On the one hand, a very high power transmission from the magnetron to the ice accreted surface can be achieved by this frequency; on the other hand, the geometrical sizes of 5.85 GHz standard components suit directly into the small aircraft structure According JAR 25, the avionic anti-/de-icing system has to be protected to effects of HIRF (high intensity radiated fields). At this frequency, the exposure field strengths for certification to be met for the system are 3200 V/m (peak) and 280 V/m in average. A certification HIRF environment produces the conditions for testing, that no interference by outside emitters affects the microwave avionics.
Microwave Source Figure 5.14 shows the shape and size of the avionic magnetron with a monochromatic frequency to be used as the microwave source. Basically, the maximum microwave power and efficiency of a magnetron is a function of frequency: the higher the frequency, the smaller gets the available peak power and efficiency. The table of Table 6.1 summarizes the relevant parameters of three different magnetron devices at different power levels and frequencies (source: Communication
Continuous Anti-Icing System Design
51
Fig. 5.14 Avionic magnetron design (shown with 10 cm rulers)
and Power Industries CPI, Palo Alto, CA, U.S.A). All of these magnetrons can be cooled just by forced air (fan). The geometrical sizes (see Table 5.1)and the weight (less 2 kg) of the components do not depend in this regime on the frequency or the overall power. According CPI, at 5.85 GHz the magnetron output power can be designed to be significantly higher than 3 kW (continuous wave, cw). The efficiency for the 3 kW magnetron (6 GHz) is about 63%, for the 2 kW design at 8 GHz about 59%, and for Table 5.1 Typical parameters of different magnetrons at different power levels (source: CPI, Palo Alto, U.S.A) Parameter
Units
3 KW Design
2 KW Design
1 KW Design
Frequency High voltage Beam current Heater voltage Heater current Anode dissipation Anode cooling Coolant flow rate Tube size Tube weight
GHz V mAdc V A W
6.0 5700 835 9.5 20 2000 Forced air 120 3×3×6 4
8.0 4800 700 8 17 1500 Forced air 80 3×3×6 4
8.0 3400 500 5.5 12 750 Forced air 40 3×3×6 4
cfm Inches lbs
52
5 Avionic Microwave Anti-/De-Icing Systems
the 1 KW design at 8 GHz about 58%. Due to very high transmission efficiency in the waveguide systems (discussed in the next sections), the overall system efficiency is primarily determined by the efficiency of the magnetron.
Primary Waveguide System Components The primary waveguide system (system scheme see Fig. 5.13) consists of exclusively rectangular waveguide components. The microwave source (magnetron) feeds a rectangular launcher waveguide to excite the fundamental TE10 mode. The waveguide used is a standard rectangular waveguide WR187 with inner sizes of 47.55 × 22.149 mm. For splitting the power equally and guiding it into two opposite directions, a rectangular waveguide tee is introduced. To change the microwave power direction horizontally by a certain angle, a rectangular waveguide bend is used, which consists of two partial standard WR 187 with a centrally curved segment. The waveguide components in their original sizes together with 10 cm rulers are shown in Fig. 5.15. To assemble and fix the components, the waveguides have to be connected by flanges. With respect to light components, the waveguide, tee and bend, being standard components, will be chosen to be made of aluminum. If further weight reduction is intended, these components may be manufactured out of CFRP, covered by a aluminum mesh. The WR 187 waveguide transmission over 1 m is about 96% of the input power.
Fig. 5.15 Models of primary waveguide system components and sizes (top: WR 187 rectangular waveguide, left: tee, right: bend)
Continuous Anti-Icing System Design
53
Secondary Waveguide System Components The secondary waveguide system is basically cylindrical. A well designed transition of the rectangular waveguide to the cylindrical waveguide is essential for a low reflective transmission of the microwave power. The working mode for transmission in the secondary waveguide system was chosen to be a TE11 -mode. The TE11 -mode has a special directed field structure, which is very suitable for concentrating the highest field intensity along the long coupling waveguide to the leading edge (Fig. 5.16). To convert the rectangular TE10 -mode to the cylindrical TE11 -mode, the field profile has to be adiabatically distorted to fit without significant reflections and false modes in length and diameter (dconverter = 4.437 cm) to the cylindrical output of the converter. The length of this component can be chosen between Lconverter – 10..20 cm. The standard cylindrical waveguide C36.4 (transmission 99.7%/m) situated along the leading edge has a inner diameter of dwaveguide = 5.86 cm. To fit the appropriate radii of this waveguide with the smaller mode converter, a taper has to be introduced. The length of the taper is calculated to be LTaper = 12.3 cm. Possible parasite modes to be excited are only the TE21 -mode and TM01 -mode. The calculation showed that the presented design comprises less than 0.1% of the power lost to parasite modes. The shape and size of the taper is visualized in Fig. 5.17. The coupling waveguide (each 2 m length) is flanged to the taper and integrated into the CFRP structure (see Fig. 5.20). This waveguide consists of a special slot
Fig. 5.16 Shape and size of mode converter
54
5 Avionic Microwave Anti-/De-Icing Systems
Fig. 5.17 Shape and size of taper
pattern, to create a specific field profile within and along the composite structure to be heated in a desired and suitable way. The slot pattern is depending on the dimensions of the specific aircraft dimensions and structure and can to be worked out according the procedures and analysis presented on pp. 23–30.
Power Requirements and Heating Performance Common Aluminum Approach We estimate roughly the essential microwave power the magnetron has to provide for the total anti-/de-icing system of the DO-328 tailplane. JAR 25 shows the requirements for maximum continuous icing conditions. According the literature, a power flux of about Πaverage = 5 − 6 kW/m2 for aluminum structures in average is sufficient to fulfil the anti-/de-icing in-flight requirements. The contour, where ice is accreted at the stabilizers nose was taken to be about 10 cm. The length of one stabilizer is about 3 m, so that we obtain the necessary ice accreted surface to be Astabilizer = 0.3 m2 . Both stabilizers are fed by one source. We calculate the necessary power Psour ce = 2A Stabili zer Πaverage = 3 − 3.6 kW . As a result, one 3–4 kW magnetron source is sufficient for the system calculated for a conventional aluminum situation.
Power Requirements and Heating Performance
55
Power Reduction with CFRP-Composites We consider the thermal heat flux balance for the CFRP-structure. T˙ C FC =
4
! 1 4 ΠC FC − εk B T CFC − Tair − h air f low T CFC − Tair cCFC ρCFC d
The averaged initial heating rate (if T¯ Str uctur e = Tair ) of the structure depends directly on the material parameters determined by the heat capacity and the density. If we introduce the values, we obtain cCFC ρCFC ≈ 0.65 c Alu ρ Alu
(5.12)
We see that the average initial heating rate for the CFRP is T˙ CFC - Structure = 1.54T˙ Alu−Str uctur e
(5.13)
higher than compared to the conventional aluminum case. As a result, if we require the same heating performance (heating rate) for CFRP heated structures as for former used aluminum structures such that ΠCFC = 0.65Π Alu
(5.14)
the necessary power is reduced about 35% for CFRP composite structures. It is shown by these estimations that a 3–4 kW magnetron should be sufficient to fully supply the whole DO-328 tailplane with power for anti-/de-icing. At this step, the heating properties have to be demonstrated in an experimental prototype set-up.
Anti-Icing In-Flight Heating Performance Figure 5.18 demonstrates for a leading edge CFRP-structure the heating perfor(green mance (dry air) for two different area power densities Πaverage = 5.8 kW m2 lines) and Πaverage = 8.0 kW (brown lines). Depending on the outside air temm2 perature Tair = −40◦ C, −30◦ C, −20◦ C, −10◦ C, 0◦ C, the corresponding averaged dynamical heating rate for the composite structure is depicted. , we If we consider the heating rates of the trajectories for Πaverage = 5.8 kW m2 obtain a value of 1.2◦ C/s. For Tair = −40◦ C air temperature, about Tstructure = is available, the composite temperature +5◦ C is achievable. If Πaverage = 8.0 kW m2 is increased to about Tstructure = 20◦ C. The highest potential on icing is found between temperatures of Tair = −10 to 0◦ C. The corresponding structural temperatures to be achieved are Tstructure = 32◦ C and Tstructure = 44◦ C, which is sufficient for successful anti-icing. If we take a look at the heating rates, we see that they are high enough for an immediate response, if ice encounter is likely (ATC, icing detector etc.).
56
5 Avionic Microwave Anti-/De-Icing Systems
Fig. 5.18 Heating performance (dry air) for Tair =−40. . .0◦ C at two different power levels for leading edge region
Figure 5.19 shows the same relations for the runback ice region, where an increased heat transfer coefficient due to the stronger attacking air stream is relevant. with Tair > −30◦ C, the strucWe see clearly, that for Πaverage = 5.8 kW m2 ture can be kept on temperature values above 0◦ C. In the ice endangered temperature regime, the runback ice area can be kept sufficiently warm for anti-icing. The full system integration is visualized in Fig. 5.20 and could be performed at a real DO-328 tailplane, which is provided to FZK by Fairchild Dornier.
Fig. 5.19 Heating performance (dry air) for Tair =−40. . .0◦ C at two different power levels for runback ice region
De-Icing Leading Edge System
57
Fig. 5.20 Secondary waveguide integration within CFRP-section
De-Icing Leading Edge System A very efficient, modular and lightweight design has been developed for de-icing of leading edge structures. This approach is especially for small and medium sized aircraft applicable, because they lack sufficient electric energy for anti-icing systems. The leading edge structure (see Fig. 5.21) consists of an inner GFRP layer and a surface CFRP layer, which is coated by a lightning protection (metal mesh). For heating, a 5.85 GHz slotted rectangular waveguide WR187 is introduced at the rear of the leading edge box for each section (design according procedures pp. 23–30). A tailplane or wing structure is divided into several sections, which take the same amount on electric power for the surface to be heated. Along the leading edge, con-
Fig. 5.21 De-icing leading edge section with parted strips
58
5 Avionic Microwave Anti-/De-Icing Systems
tinuously is heat exchanged to the environment generated by the magnetrons. The magnetrons are cooled by a special liquid and joined together in one circulation. The sections are switched on one after another, such that only for each wing one magnetron is consuming power. For the other sections, the forming of ice is allowed, until this section is heated and the ice is melted on the interface and taken away by the aerodynamic forces. The truncation at the leading edge and sections by continuously anti-icing (parted strips) guarantees, that the ice removal is complete and no increasing ice agglomeration is possible any more (runback icing).
Chapter 6
Processing Technology for Composite Materials
Current airplanes still consist mainly of metal which are mainly aluminum alloys. In the Airbus A320, the weight fraction of carbon fibre-composite (CFC) materials already amounts to 15%. These composite materials consist of fibre materials and a hardened polymeric resin. The amount CFC materials is continuously increasing as it is shown for the new megaliner A380 which will be delivered in September 2007 (Fig. 6.1). While each A380 still contains around 60% of aluminum but 22% of carbon fibre-reinforced composites, the share of the latter material shall rise to 52% in the new planned long-range aircraft A350 XWB that is comparable to the new 787 of Boeing. The situation for the increasing CFRP applications and technology development has to be carefully considered in view of future market requirements [52]. A maximum of 30% of weight reduction is estimated by the direct substitution of metallic structures with CFRP composites. Considering operational costs, the final total
Fig. 6.1 Overview on composite parts for the A380
L.E. Feher, Energy Efficient Microwave Systems, C Springer-Verlag Berlin Heidelberg 2009 DOI 10.1007/978-3-540-92122-6 6,
59
60
6 Processing Technology for Composite Materials
saving for MTOW (Maximum Take Off Weight), taking fuel and full boarded aircrafts into account, is about 7.5% giving a significant saving for the operating airliner company as well as reduced fuel consumption for the long hault flight operation. As investigations on the overall ecological consumption in terms of energy and related tool materials for conventional CFRP production in comparison to aluminum fabrication showed, no significant reduction on prices and resource consumption is achievable with the existing process technologies for CFRP [55]. Beside this, a considerably decreasing price of carbon fibres in the near future is expected which is caused by new production facilities due to an immanent market demand in the transportation industry. This shows clearly the need to focus on new and costs saving innovative processing technologies to industrialize a widespread application of lightweight CFRP composite structures. The specific bottleneck in the aerospace and transportation field is the price/kg in comparison to aluminum, which is not competitive due to the stated high fabrication costs (material costs just about 30% of the actual end price, 70% for processing issues). The highest potential for a significant cost reduction is therefore to be found on the manufacturing process which implies substantial long time and high energy consumption, as well as a very low degree of automation. Today, CFC materials are still hardened with heat in heavy industry ovens at high pressure, the so-called autoclaves (Fig. 6.2). To achieve further technological progress and a cost reduction for CFC processing, the following issues have to be considered more details: • • • • • •
Reduced cycle times Lower cost tool materials Reduced production of volatiles Process automation Large part capability Reduction of energy consumption
Fig. 6.2 The “Orca”-autoclave at Airbus Stade, a 33 m long CFC production facility
6
Processing Technology for Composite Materials
Fig. 6.3 Process flow for composite fabrication
61 heat + pressure
fibres mold
composite part
resin
A typical manufacturing process in a conventionally used production autoclave can last up to 10 hours for a cycle which is an inherently long time. The CFC composite components are prepared usually on a metal tool (mold) that consists of the exact contour for the technical part (Fig. 6.3). The CFC structure preparation steps are a rather expensive procedure due to the selection of specific high quality tool materials, mainly a variety of polymeric foils which have been qualified for aerospace application. The prepared CFC structures are sealed for applying a vacuum bag technology. The quality of the vacuum and its preservation is essential for the final quality of the structure, as well as a very close and temperature homogeneous process control. For industrial aerospace use, all process steps are certified and underlie restrictive requirements for temperature homogeneity. For currently used autoclave processes, considerations to substitute tool materials by less expensive choices are continuing. The quality of sealing, temperature homogeneity as well as the application of pressure determines the volatile content and quality for conventionally cured CFC structures. As mentioned before, automation is one of the key issues for CFC production. The integration of more and more preparation steps into one optimized production line increases versatility, speed and quality of the workflow. To satisfy the need to substitute more and more aircraft parts by CFC structures, their sizes for wings and fuselages have to increase continuously which is limited by autoclave capacities and temperature homogeneity capabilities. This “large part problem” has to be overcome as well as the tremendous need on electrical power for heating the full autoclave and cooling it down after the curing process has ended. The general manufacturing process for CFC structures can be subdivided into two different approaches [53]: • The most commonly used materials for aerospace composite fabrication are prepreg materials, due to their easy handling and storing issues. The fibres are pre-impregnated with specific resin compositions and are commercially available as very thin plane foils. Currently used and certified prepreg systems need the application of pressure in an autoclave to achieve the required quality during the curing. The fabrication procedure is highly optimized by tape laying machines, but basically limited by these machines for increasing sizes and complexity of the parts. At this time, new pressure-less prepreg systems have been adapted and are available on the markets which comply with future pressure-less fabrication and
62 Fig. 6.4 Process cycle for resin infiltration technology
6 Processing Technology for Composite Materials T curing infiltration t
curing technologies. Prepreg materials for aviation are usually cured depending on the resin system at temperatures of 120–180◦ C. • Another preferred approach is the wet or injection technology, where resin is infiltrated into a dry lay up of carbon fibre weaves. By this technique produced parts show excellent surface detail and accuracy. The need for shipping and storing refrigerated prepreg is removed as well because fibres and resin can be joined on demand. This technique can be used very efficiently for fabricating large and complicate 3D structures. This approach does not need a pressurized environment, as very high fibre contents and quality for the structures can be achieved with a simple vacuum bag technology. The degree of automation for this approach has to be increased in the future. The curing temperatures depend on the only certified resin RTM6 in aviation, where a temperature of 180◦ C has to be applied. For prepreg-materials, a very complex chemical procedure prepares the impregnated fibres for use, where the fabrication process has to be immediately performed when the lay-up of the structure is initiated. For injection technology, an intermediate preform step can be performed, where the dry fibre weaves (which have been impregnated with a binder) are heated on a tool to about 80◦ C. The polymerization of the binder keeps the contour for the fibre weaves, which can be stored for a later curing cycle (Fig. 6.4). A typical temperature cycle for an injection process consists of two holding times, the first to ensure the full infiltration of the lay-up, usually at temperatures between 80 and 120◦ C and in addition the curing cycle at 180◦ C.
The HEPHAISTOS-Systemline and Technology A novel industrial microwave systemline HEPHAISTOS (High Electromagnetic Power Heating Automated Injected STructures Oven System) for curing of carbon fibre reinforced plastics (CFRP) has being developed. These systems integrate advantageously the basic processing steps as tooling, tempering of the resin and lay up, the impregnation of the fibres, pre-forming techniques as well as finally the process curing of the composite structures. The advantages obtained by microwave technology for composite processing are • Selective heating of composite materials • Volumetric heating of the laminates High heating rates – reduction of processing time
The HEPHAISTOS-Systemline and Technology
63
Fig. 6.5 A curved “Large Part” presentation CFRP-structure fabricated with the HEPHAISTOS-CA2 microwave processing system (1.6 × 1.8 m)
• High temperature homogeneity for curing • Savings on energy consumption “Clean” heating technology • High degree on automation As a result, a cured “large part” aerospace CFC prepreg structure is shown in Fig. 6.5 that was produced with an HEPHAISTOS-CA2 system. The presented structure takes about the size that is available in the CA2 applicator chamber. The technology complies perfectly with novel autoclave-free fabrication methods which are under special consideration of industrial non-autoclave heating system manufacturers and composite end users. The internationally patented HEPHAISTOS systems work in principle without additional pressure and heat the CFC component selectively in an inertia-free manner – the oven environment is no longer heated actively. By the choice of novel, highly efficient modular antenna and waveguide systems designed for the HEPHAISTOS-systems, nearly the complete generated microwave power can be transmitted into the applicator chamber without any remarkable losses. Moreover,
64
6 Processing Technology for Composite Materials
Fig. 6.6 The HEPHAISTOS-CA2 facility
the microwaves act volumetrically, i.e. without heat conduction, the microwave enters the material to be heated and immediately leaves a heat input inside. In this way, a high heating rate is generated directly in the component caused by the microwaves at low energy consumption (Fig. 6.6). The autoclave is today the most popular system in the high performance composite field, as prepregs used until now usually need a high pressure environment during the curing process for forming. The concerns on this state of the art approach are • • • • • •
Long process times Inherently energy inefficient system Major difficulties caused by large thermal gradients and slow heat up times Bag failures Complex expensive tooling High acquisition costs
A unique alternative for heating are microwaves, since microwave heating at 2.45 GHz and 915 MHz has been established as an important industrial technology about more than 50 years ago. The successful application of microwaves in industries that is always competing with conventional heating systems, has been reported e.g. by food processing systems, domestic ovens, rubber industry, vacuum drying etc. The use of microwaves in the composite field shows new approaches on the process interaction, e.g. for the fibre materials, precursors, resin systems, lay-up
The HEPHAISTOS-Systemline and Technology
65
preparation with direct consequences on the adjustable material properties. CFRP materials are heated volumetrically with microwaves, offering the opportunity of very high heating rates and improved mechanical performance. Efforts on microwave and millimetre-wave composite curing as well microwave effects have been reported in [54–57] and proposed for technological application. In comparison to conventional heating where the heat transfer is diffusive and depends on the thermal conductivity of the material, the microwave field penetrates the material and acts as an instantaneous heat source at each volumetric point of the structure. A lot of individual technological solutions in industries, where the physical benefits of using microwaves result finally in commercial gain, have been developed. Investigations on heating CFRP with microwaves have been undertaken for decades but failed due to the lack of a sufficient homogeneous electromagnetic heating field or reasonable costs for suitable sources and components providing industrial scale up. The HEPHAISTOS concept originally developed and patented at FZK solves these long existing problems. These ideas have been tested initially during collaboration with DLR, Braunschweig at an early fundamental stage [88, 89] (Fig. 6.7). The system line is named after Hephaistos, who is the builder and craftsman for the Greek gods being also responsible since the past for oven and transportation technologies. The name HEPHAISTOS (High Electromagnetic Power Heating Automated Injected STructures Oven System) stands as well for the technological concept of this innovative approach.
Fig. 6.7 Hephaistos, painting ca. 525 BC. He is also known as Hephaestus and Vulcan (Roman). His attributes in iconography include the axe and tongs
66
6 Processing Technology for Composite Materials
Distinctive physical and technological advantages of microwave heating compared to conventional heating are: • Volumetric heating of composite materials: The microwaves penetrate instantaneously the materials and generate locally a specific heating content. • Reduction of cycle times: Due to the volumetric heating, the heating rates with a structure can be strongly increased gaining for a higher throughput of products. • Selective heating: Microwaves only heat the composite structure, the oven and components “keeps cool”. • Energy savings: Only the lightweight structure is directly heated. The oven will not be cooled actively, which results as well in reduction of cycle times [58]. • Rapid control of the process: Due to the instantaneous and volumetric heating within the composite structure processes can be applied, which are impossible in conventional ovens due to their inertia on temperature changes especially if local overheating occurs for exothermal reactions. • Upscale: The HEPHAISTOS-technology can be applied for large structures. • Automation: The fabrication process as well as control and sensoring can be adjusted very advantageously with the HEPHAISTOS-technology, such that overall optimizations of the process chain can be gained. • Reduced hardware-costs of the heating system: Due to the use of standard industrial 2.45 GHz components, an industrial HEPHAISTOS-system is much more convenient than autoclave systems. The most notable effect processing CFRP materials with microwaves is their volumetric heating, offering the opportunity of very high heating rates. In comparison to conventional heating where the heat transfer is diffusive and depends on the thermal conductivity of the material, the microwave field penetrates the material and acts as an instantaneous heat source at each point of the sample [85, 86]. The CFRP can be selectively heated, keeping the oven environment cool [90]. Spatial temperature homogeneity is crucial for qualified material properties. The samples must be exposed therefore to excellent homogeneous field distributions [92]. This is essential, as the carbon fibres imply a very high microwave reflectivity and a tendency for arcing and breakdowns at loose ends in areas of inhomogeneous microwave patterns. This problem is very severe and one of the major phenomena, that made microwaves not applicable yet for CFRP processing. A specific multimode applicator development, which is described in the following chapters in detail aims to tackle these problems.
Microwave Process Development Process developments for CFC structures are continuously proceeding. For the HEPHAISTOS technology, the generation of an industrialized and certified process is essential for the commercial application of the microwave processing technology in the aerospace field. At first, fundamental research had to be performed to
Microwave Process Development
67
understand the heating mechanism and electrothermal interaction of CFC structures with electromagnetic waves. The original early idea of the HEPHAISTOS applicator concept was generated in a Ph.D. Thesis of the author considering higher frequency processing of materials [2]. The continuing work has been accompanied first by the DLR in Braunschweig [57] and later by the EADS Composite Research Centre in Ottobrunn for a detailed fundamental test and evaluation program of a large number of processed CFC materials and structures. The results of this fundamental program launched the industrial development of the HEPHAISTOS system line due to a technological transfer cooperation. In addition detailed process steps to combine with microwave advantages were worked out and have been proofed by experimental tests (Fig. 6.8). The challenge for microwave curing turned out to be [99]: • • • • • • •
Generation of highly homogeneous electromagnetic fields Handling of highly electrical conducting carbon fibres Generation of highly homogeneous temperature fields Control of exothermic polymerization reaction of epoxy polymers Energy efficient power transfer Adaption of certified tool materials for CFC lay-ups Temperature measurement and control
To understand the full properties of CFC materials, related 3D structures [71] and their various tool materials with microwaves, dielectric measurements of the materials and components have been taken, e.g. to consider and evaluate the profound differences of the anisotropic fibre directions for microwave heating and processing as well to adapt tool materials without a tendency of overheating or hot-spot formation (Fig. 6.9). The CFC composite lay-up is a series of subsequent foils starting on a highly cleaned metal tool surface, which contains the contour of the technical part. The dry carbon fibre weaves are contained in a release ply and for injection, a breather made of a microporous membrane is stacked under the bagging film which is hermetically closed by a sealant. With a valve system, a vacuum hose is connected to the lay-up as well as an opposite directed injection hose, where the pre-heated resin is infiltrated into the lay-up. A typical set-up for an infiltration process is depicted in Fig. 6.10. The use of microwaves is advantageous to all specific process steps in the production chain.
Fig. 6.8 HEPHAISTOS development milestones
68
6 Processing Technology for Composite Materials
• fibres (e.g. carbon) + matrix (e.g. epoxy) • orthotropic mechanical properties • stacking of layers with individual orientation
The mechanical properties of the laminate can be designed. This determines the electromagnetic material properties. Fig. 6.9 Anisotropy of layered CFC composites microporous membrane
resin distributor
vacuum
resin
Fig. 6.10 In-principle set-up for VAP infiltration process (vacuum assisted process) developed by EADS, Germany
Fig. 6.11 Microwave assisted CFC fabrication flow (EADS Munich) [108]
Figure 6.11 shows a completed microwave assisted flowchart for production. These are the steps for CFC preforming, the preparation of frozen RTM6 resin and the precise pre-heating by a microwave flow heater (“Microwave Injector”), the preheating of the dry weave in a microwave applicator chamber and the curing of the
Fundamental Investigations
69
injected CFC laminate. Even the steps to join CFC parts can be done much more advantageously using the microwave applicator chamber, because only the adhesive can be heated selectively by the microwaves which gives a rapid and highly energy efficient assembling procedure.
Fundamental Investigations Homogeneous Field Applicator Development At Forschungszentrum Karlsruhe the possibilities of employing high frequency micro- and millimetre waves for a possible industrial use have been investigated as a spin-off project of the magnetic fusion program since 1993. Primarily, interest has been focused first on processing structural and functional ceramics such as low lossy alumina [77], where unique advantages of millimetre-wave heating and processing have been demonstrated as well as novel system design technologies and hybrid heating. Experiments and simulations have shown in general that achieving homogeneous field distributions over most of a large applicator volume is not trivial, even at high frequencies in the millimetre wave regime (30 GHz). A large cylindrical cavity (1 cm wavelength, cavity height 50 cm, diameter 50 cm) with the use of mode-stirring devices turned out to be inadequate under industrial quality requirements due to focusing effects, standing wave structures and field concentrations in the centre. A synthesis of field measurements and 3D virtual computer experiments (MiRa-Code) [65] led to a special hexagonal applicator design showing significantly reduced field fluctuations and increased homogeneity over most of the large applicator volume [66]. As a measure for the quality of the field within a considered volume (the processing area), the design parameter Δd was defined σ| E 2 | Δd = 2 E
(6.1)
σ|E 2 | represents the standard deviation of the electric energy density in the given volume, E 2 is the spatial averaged energy density in the considered area. The increasing quality of the field homogeneity at a high field level can be realized by a decreasing value of Δd changing the applicator configuration. For avoiding disadvantages such as focusing effects, caustics, structured standing waves (resonator modes) and field concentrations in the coupling plane topological plane structures as the screening geometry of the applicator have been investigated and compared to each other. Two different basic configurations (Fig. 6.12) with polygonal applicator cross sections are possible. (i) Beam targeting at a segment of the plane surface for a direct reflection and (ii) Beam splitting at a polygonal corner forming two outgoing beamlets moving into different directions.
70
6 Processing Technology for Composite Materials
Fig. 6.12 Basic polygonal orientation
Beam directly reflecting
Beamsplitting
Hexagonal Inset
For both orientations, calculations have been performed by varying the number of polygonal edges from a quadratic shape to an octagonal shape. The results on the calculated field quality can be seen in Fig. 6.13. A significant minimum for Δd can be observed. Experiments verified the calculations, showing a significant intrinsic gain in field homogeneity for microwave exposed and processed materials. These considerations originally obtained at higher frequency microwaves (30 GHz) have been patented for technological applications [5]. For the industrial development of CFRP processing systems, a sophisticated concept for lower frequencies (2.45 GHz, 915 MHz) has been developed by intense use of DELFI-simulations [77, 88, 89, 96].
Results of theΔd -Parameter for a polygonal series
1 0,9 0,8 0,7 0,6 0,5 0,4 0,3
1 0,9 0,8 0,7 0,6 0,5 0,4 0,3
Quadratic Pentagonal Hexagonal Heptagonal Octagonal Cylindrical Cavity Cylindrical cavity, Mode Stirrer
Cross section geometry Fig. 6.13 Comparison of polygonal cross sections [2]
Beam reflecting
Beam splitting
Fundamental Investigations
71
Monomode applicators involve only controllable field properties in small specific applicator regions, whereas multimode ovens promise the possibility for low field fluctuations in larger regions. An industrial microwave oven for production needs is successfully designed, if the available processing volume containing uniform field properties is about the size of the whole cavity and reflections are minimized. Conceptions of the millimetre-wave technology at FZK have been transferred by DELFI simulations to 2.45 GHz technology to realize these demands to provide a versatile designable large part capability for CFRP processing [89] (see p. 82). The principal hardware contribution of the system development is realized for a specific modular hexagonal applicator containment providing an excellent homogeneous electromagnetic field distribution [95, 99, 105]. The fabrication process can be performed in pressurized environments up to 5 bar or at standard conditions. The processes are measured remotely by optical thermocouples, infrared sensors or shielded low-cost thermocouples. The HEPHAISTOS-concept follows a two paths strategy: 1. as a stand alone processing system 2. as an upgrade system for existing autoclaves The hexagonal HEPHAISTOS microwave system containment is very compact and can be used as well as an inset to upgrade at low costs existing conventional autoclave systems as depicted in Fig. 6.14. The upgrade system consists of a compact inset module that can be inserted into an autoclave, changing it to a cool microwave system. The autoclave can still be applied as a pressure shell. But because on the very cost-intensive nature for autoclave based composite processing technologies as pointed out before, autoclave-free approaches without pressurizing are most promising for future broadband industrial application. The commercial applicator development targets preferred at modular stand alone systems for CFC pressure-less processes in combination with specific infiltration techniques such as VAP (Fig. 6.15).
Fig. 6.14 A conventional standard autoclave is changed by the HEPHAISTOS-CA upgrade to a full electromagnetic “cold oven” system
72 Fig. 6.15 Schematic set-up of HEPHAISTOS-systems
6 Processing Technology for Composite Materials
Infrared Temperature Sensoring
Power Supply
Microwave Processing Applicator
Process Control
Cooling
Magnetrons + Waveguide System
Since the beginning of 2003, a first HEPHAISTOS-CA1 with a full hexagon has been used as a pilot system at FZK to follow the early fundamental systems HEPHAISTOS-SA (used in Braunschweig at 2001–2002) and HEPHAISTOS-BA (used in Karlsruhe at 2001–2003), which have been shaped as semi-hexagons. Former basic investigations for CFRP heating have been performed with a 30 GHz gyrotron installation (Fig. 6.16). The HEPHAISTOS-SA semi-hexagonal system has been used as an inset within a pressurized chamber at DLR Braunschweig to process prepreg-materials. The maximum pressure that can be applied to the magnetrons is about 5–6 bar. The length of a commercial hexagonal HEPHAISTOS applicator module has been assigned for each system to 1 m [106, 107, 110, 112]. Arbitrary system lengths can be achieved by serially assembling the 1 m modules. For microwave feeding, 12 optimized nonresonant slotted waveguides WR340 are arranged around the 6 edges of the hexagonal applicator. Due to the high transmission efficiency of the waveguide systems, no related circulators and additional cooling devices are necessary.
Fig. 6.16 Upgraded small standard autoclave at DLR Braunschweig with an HEPHAISTOS-SA system
Electrothermal Considerations
73
Fig. 6.17 HEPHAISTOS-CA1 system (1 m length, 1 m diameter)
The operating frequency has been chosen to the ISM frequency of 2.45 GHz. The air cooled magnetrons are directly mounted on the WR340 waveguides and consist each of a power of 700–800 W. The power supplies are designed using standard components and have been arranged using an IR-camera to keep the heat losses on the circuits and transformator of the power supply as low as possible (Fig. 6.17). The commercial HEPHAISTOS-CA-1 system is based on the first HEPHAISTOS-CA1 prototype system at FZK which has been evaluated for the application in composite manufacturing together with EADS Corporate Research Centre and the Institute of Aircraft Design (University of Stuttgart). The HEPHAISTOS-CA2 and CA3 have been changed by diameter and number of subsequent modules [110].
Electrothermal Considerations Unfortunately, considering materials in general, the physical occurrence of inhomogeneous temperature profiles limits the immediate technological access for microwave heating/processing systems even when a homogeneous electromagnetic heating pattern of very high quality is available. Solving the exact thermal and electromagnetic equations, it turned out that an idealized homogeneous field does not correspond in general to a homogeneous spatial temperature distribution [79, 81, 98]. Heat and temperature within a material are determined by the non-linear inhomogeneous heat equation [2]
74
6 Processing Technology for Composite Materials
∂ T (x, t) − ∇ (σT (x, T )∇T (x, t) = Pe f f (x, t, T ) ∂t
cv (x, T )ρ(x, T )
(6.2)
σT (x, T ) is the thermal conductivity, ρ(x, T ) the local density and cv (x, t) is the heat capacity of the material. The effective power density pe f f (x, t, T ) includes
contributions from electromagnetic microwave heating pelec (x, T, t), heat losses by thermal radiation prad (x Sur f ace , T ), conduction and convection.
Since ∇σT (x, T ) → 0 is assumed for the heated samples, the temperature field in the stationary state ∂T / ∂t = 0 is finally described by Poisson’s equation
ΔT (x, t) = −
pe f f (x, T ) σT (T )
(6.3)
which is of great interest for the determination of phenomenon in the process holding cycle (Fig. 6.18). An analytic solution of (6.3) with respect to boundary conditions of an idealized homogeneous electric field can be obtained for the temperature T(x) T (x) = S + κ[P(x) + A(x)]
(6.4)
With the parameters κ=
Pelec χ, 4AσT S=
4
χ=
βδ β sinh (2δxa ) − δ sin (2βxa )
(6.5)
Pelec εambient 4 + T 2Ak B ε(0) ε(0) ambient
(6.6)
# 1" cosh(2δ(x − xa )) + cosh(2δx) − cosh(2δxa ) − 1 2 δ # 1" P(x) = 2 cos(2β(x − xa )) + cos(2βx) − cos(2βxa ) − 1 β
A(x) =
Homogeneous Field
Fig. 6.18 Material model and boundary conditions
(6.8)
Homogeneous Field
Material
0
(6.7)
X
Electrothermal Considerations
75
κ represents an electrothermal coupling coefficient, S a surface term; in (6.8) A(x) refers to the absorption properties and P(x) to the dielectric permeability properties of the material. ε determines the emissivity and kB the Boltzmann constant. The solution is fully consistent to Maxwell’s equations, where β and δare the real and imaginary part of the wave number k x = β − jδ
n¯ 2 =
εr2 +
σ2 ω2 ε02
(6.9)
The refraction index n¯ includes additional contributions of the electrical conductivityσ. β and δ are finally given by
σ 1 arctan β = k0 n¯ cos 2 ωε0 εr
σ 1 δ = k0 n¯ sin arctan 2 ωε0 εr
(6.10)
As a result, we compare the temperature answer of a ceramic type (Al2 O3 ) low loss material (low electric, but high thermal conductivity) with a CFRP like material (high electric, but low thermal conductivity). The samples are exposed to an idealized homogeneous ambient millimetre-wave field of 30 GHz [82]. Figure 6.19 shows the resulting stationary inverse temperature predictions within the slab dimension of xa = 10 cm. For the low lossy material (left), a stationary hot spot [79, 81] emerges in the centre. The right graph shows the temperature distribution under the same circumstances and power level for a carbon reinforced composite like material leading to stable overall temperature homogeneity for the slab. The basic physical difference of the temperature answers for the samples is obvious [83]. A basic experiment performed with a honeycomb CFRP sample at 30 GHz verified the analytic predictions. Figure 6.20 shows the measured temperature differences of two arbitrarily placed thermocouples in a CFRP composite slab heated at 30 GHz in a hexagonal applicator
Fig. 6.19 Ceramic like and high electric conducting material temperature profile [◦ C], x-axis: position within the sample [m]
76
6 Processing Technology for Composite Materials
Fig. 6.20 Temperature homogeneity of CFC composite structure at 30 GHz
that provides a homogeneous field. The stationary temperature state was achieved after 50 s. The obtained temperature differences are about ± 1◦ C and represent fluctuations within the accuracy of the used metal thermocouples.
The Temperature Answer of Materials Exposed to a Homogeneous Field – Choice of Frequency Another important step for the technical development and frequency choice for the HEPHAISTOS-Technology was the generation of a general classification for the electrothermal behaviour of materials in the holding cycle (stationary state). This resulted in an extensive use of the analytical methods presented in this work and numerical codes (DELFI, THESIS3D) developed by the author for predicting the temperature answer of materials while exposed to homogeneous electromagnetic fields [69, 74]. To obtain a quantitative comparable measure which is basically independent of the specified environment, the normalized statistical variance of the temperature distribution within a considered volume of an arbitrary material has been investigated for different process conditions. The detailed calculations have been presented e.g. in [82, 94, 98], where more than 70 materials with their electromagnetic and thermal material parameters have
Temperature Answer of Materials
77
been taken into account. The calculations for the statistical variance of the temperature in the holding cycle were performed at the ISM frequencies of 24.15 and 2.45 GHz. The calculated results showed that the materials obey for the resulting temperature homogeneity (temperature answer) in a natural order: Organic materials can answer nearly completely homogeneous for the resulting temperature field, followed by polar materials and finally inorganic materials. The obtained order is consistent to the quantum mechanical considerations of available transition states for heating and energy conversion (pp. 24–21) corresponding to the behavior and composition of the microwave effective conductivity for ceramics, water polymers etc. Further, it could be predicted and calculated precisely by the analytical and computational tools at FZK, that under specific circumstances, process fields with vanishing thermal gradients can be applied for CFRP composite materials[10, 56]. Figure 6.21 shows the universal potential of microwave processing with homogeneous fields for industrial applications on organic polymer and reinforced materials [94, 98, 104]. But the calculations even showed, that the resulting temperature homogeneity achieved by the 2.45 GHz HEPHAISTOS-technology can not be improved by increasing the frequency (e.g. 24 GHz), which was previously understood to give an opportunity for finer resolved microwave field distributions. This result can be physically understood that in the stationary state all patterns on phase distributions have been thermally diffused within the volume of the heated materials, such that no artifact on a frequency signal is remaining in the resulting temperature distribution. Curing experiments at 30 GHz of CFC slabs with resin showed in addition a very high untunable reflectivity of the samples and a questionable quality of the cured CFRP [87, 93], because the polymerization showed unacceptable brittle properties during mechanical tests. At 2.45 GHz, the penetration of the microwaves (volumetric heating) is much higher at a very high degree of homogeneity in comparison to 30 GHz with proofed very good material properties. Therefore, the HEPHAISTOS operating frequency could be decided to use the commercial frequency of 2.45 GHz.
0.0000
Organic
Marbledry
GlassCeramic
BariumTitanate
NaturalRubber
Water
Milk
0.0005
Fig. 6.21 Stationary temperature homogeneity for materials at 2.45 and 24 GHz
Polymers
0.0010
Butter
Variance
0.0015
RubyMica
2.45 GHz 24 GHz
0.0020
Inorganic
Polar Materials
FusedQuartz
Effect of Frequency on Temperature Homogeneity
MagnesiumSilicate
0.0025
78
6 Processing Technology for Composite Materials
Basic Investigations for the CFC Curing Process with Microwaves The considerations on the temperature answers and homogeneity for CFC materials until now have been focused on the holding cycle. To simulate the temperature development for a CFC composite heating, a numerical time dependent code THESIS3D [2, 69, 93] has been used. As in a real experimental process, a heating rate and a holding temperature have been defined. The virtual CFC slab was assigned for the process such that the averaged temperature in the volume follows the prescriptions. Figure 6.22 depicts the simulated temperature answer of the CFRP slab. The minimum and maximum temperature of the slab, as well as the temperature averaged over the sample volume can be observed. As predicted in the analytical approaches, the temperature differences in the holding state vanish within the whole volume of the slab. During the heating rate cycle, a temperature gradient represented by the observed difference of the maximum and minimum temperature of the slab is occurring. The volumetric temperature homogeneity of the slab is shown in Fig. 6.22 as well, which is initially at a maximum peak due to the volumetric penetration of the surrounding homogeneous microwave field. The peak of the temperature variance is consequently decreasing and reduced to zero at the holding temperature. The initial inhomogeneity peak is of course proportional to the exposed electromagnetic power of the homogeneous microwave field. The simulated heating behavior of a CFC slab was proofed experimentally at 30 GHz by exposing a cured slab to the homogeneous microwave field of the hexagonal applicator.
140 0,014 0,012
100
0,010
80
0,008 average Temp maximal Temp minimal Temp Variance
60 40
0,006
Variance
Temperature in °C
120
0,004 0,002
20
0,000
0
–0,002 0
100
200
300 Time in s
400
500
600
Fig. 6.22 THESIS3D-process simulation of a CFRP slab in a homogeneous microwave field at 30 GHz
Basic Investigations for the CFC Curing Process with Microwaves
79
Fig. 6.23 Heating of a CFRP Slab in a homogeneous microwave field at 30 GHz
The heating was interrupted during the heating ramp to open the oven doors and an outside installed infrared camera on the top of the oven was used to compare the predicted temperature distribution of the THESIS3D simulation with the experiment [93]. The results are depicted in Fig. 6.23 and an excellent agreement has been achieved. Because of a time delay due to the opening of the oven the experimental observed distribution is smoothed slightly by thermal diffusion. Figure 6.24 shows two important results taken at 2.45 GHz in different HEPHAISTOS-systems: • The structure of the heating pattern does not change significantly from 30 to 2.45 GHz (left picture, Fig. 6.23 and right picture Fig. 6.24) • The influence of a curved geometry of the CFC part does not have a significant change on the temperature distribution (therefore in consequence the homogeneity of the surrounding microwave field in the hexagonal HEPHAISTOS cannot be changed by the differently shaped load as well).
Fig. 6.24 Right: Heating of CFRP slab at 2.45 GHz (HEPHAISTOS-CA2), Left: symmetrically curved panel at 2.45 GHz (HEPHAISTOS-BA)
80
6 Processing Technology for Composite Materials
Anisotropic Electromagnetic Effects for CFRP Heating As discussed for microwave de-icing, fibre reinforced composites are a generally anisotropic material which determines the mechanical, thermal and electromagnetic properties of the laminate. The fibres are embedded in a homogeneous polymer matrix of cured resin. The dielectric properties are represented by the complex dielectric tensors, when no magnetic properties are involved (μr = 1) ⎞ ∗ ε⊥ ¯ 0 0 σˆ e εˆ ∗ = ⎝ 0 ε∗¯ 0 ⎠ = εˆ r − j ωε 0 ∗ 0 0 ε⊥ ⎛
(6.11)
∗ ε⊥ ¯ denotes the complex permittivity in a laminate layer perpendicular to the fibre orientation, ε∗¯ denotes the permittivity properties parallel to the fibre orientation ∗ and ε⊥ denotes the complex permittivity perpendicular to a laminate layer. Due to the plane geometry of a fibre layer as depicted in Fig. 6.25, the elements ∗ ∗ ∗ ∗ ε⊥ ¯ and ε⊥ coincideε⊥ = ε⊥ ¯ . To obtain the dissipated power density in a fibre layer, the representing equation gets more complicate
pelec (x , t) =
1 x , t) = 1 (σ E 2 (x, t)+σ⊥ (E ⊥2 (y, t)+ E 2¯ (z, t))) (6.12) E(x , t)σˆ e E( ⊥ 2 2
To estimate the major contribution for heating the CFRP layer, dielectric measurements for the RTM6 resin matrix and the different fibre directions of CFRP at 2.45 GHz have been taken that shows profound differences (Table 6.1). If we consider an ideal homogeneous at the thin layer for each polarization field to be equal in the amplitude E (x, t) = E ⊥ (y, t) = E ⊥¯ (z, t), the heating is caused mainly by the contribution parallel to the conducting carbon fibre at the surface
Fig. 6.25 Fibre layer and geometry of tensor directions
Table 6.1 Dielectric parameters for anisotropic CFRP components RTM6 CFRP⊥ CFRP
εr
σe [S/m]
3.11 48.10 4521.40
0.02 12.50 88.30
Anisotropic Electromagnetic Effects for CFRP Heating
pelec (x , t) ≈
1 σ E 2 (x, t) 2
81
(6.13)
The components E ⊥ (y, t) and E ⊥¯ (z, t) transmit through the anisotropic fibre layer. A CFC laminate is a series of individual fibre layers with in general changing fibre directions as depicted in Fig. 6.26. If we consider a unidirectional laminate (same Figure, left) consisting of several layers, the E (x, t) component is immediately absorbed at the surface of the sample, while the perpendicular components both transmit through the sample and result in a significant volumetric heating of the subsequent layers
2 1 Pelec (x , t) = σ⊥ E ⊥ (y, t) + E ⊥2¯ (z, t) d V (6.14) 2 v
The temperature answer of the volumetric microwave heating can be expected to be similar to the profile discussed in Fig. 6.19 (left) (Fig. 6.27). This could be proofed experimentally by heating a thick unidirectional slab at 30 GHz showing that the heating pattern changed completely in comparison to the common heating profiles for multidirectional materials like shown in Figs. 6.23 and 6.24. A significant hot spot [79, 81] in the centre of the slab is realized due to the volumetric heating of the perpendicular field components [103]. For a multidirectional composite, which is usually used for material stability two perpendicular components of the electric field always transmit a layer to the next one while one parallel component is the dominant heating source of the individual considered layer. The microwave signal has therefore a higher penetration depth in anisotropic fibre materials than in materials containing the same amount on carbon powder, which would damp the impinging microwave equally for each polarization component.
Fig. 6.26 Anisotropic orientations for composites
82
6 Processing Technology for Composite Materials
Fig. 6.27 Experimental volumetric temperature answer of a unidirectional CFRP slab at 30 GHz
Design and Proof of the Modular 2.45 GHz HEPHAISTOS Conception For the concept realization of the commercial CFRP processing system line, the described analytical modeling and numerical simulation capabilities have been intensively used with respect to • • • • • • • •
Applicator design Waveguide systems Fibre/matrix composition of the laminate Material dynamics Hot spot formation Process/temperature control Exothermal heat wave propagation Tooling arrangements
A detailed description of the numerical methods and theoretical approaches for the MiRa-Code (Microwave Raytracer) and THESIS3D-Code is given in [2]. To model microwave components (e.g. waveguides) and field distributions within applicator chambers or exposed materials at lower frequencies (like the commercial frequency of 2.45 GHz) the DELFI-Code (Distribution of ELectromagnetic FIelds) was developed by the author [59]. The method is based on a staggered-grid Yee algorithm in a Finite Difference Time Domain (FDTD) scheme [60, 61] to reduce the allocated memory and to speed-up the computation at a high accuracy for the electromagnetic fields. Wave propagation up to 24 GHz has been successfully simulated with the DELFI-package. The code implies the following features • • • • •
Explicit 3D FDTD-code solving the full time dependent Maxwells´ equation. Arbitrary geometries (applicator, load, waveguides). Arbitrary materials. Reduced memory needs, fast and stable algorithm. Code for complex microwave system design and engineering.
Design and Proof of the Modular 2.45 GHz HEPHAISTOS Conception
83
By use of the DELFI-Code, the results and field homogeneity properties obtained at 30 GHz were successfully transferred to the commercial modular HEPHAISTOS applicator concept. As an example, the microwave energy density at the surface of a rectangular CFC slab processed in an HEPHAISTOS-BA system at 2.45 GHz calculated with the DELFI-Code is visualized in Fig. 6.28. The calculated fluctuations are less than 3% to the average value showing the high homogeneity of the applicator [74]. On the right the initial temperature answer of the slab in the central middle plane of the slab is calculated with the THESIS-3D code. As a result, the very low phase fluctuations do not interfere the temperature signal in the central middle plane of CFC slab. The entity of the numerically obtained results has been the prerequisite for a successful technological development. The experimental work with the HEPHAISTOS processing technology proofed consequently the procedures and showed supreme material quality properties of processed CFC materials. A program of intense material tests has been performed at Institute for Aircraft Design (IFB, University of Stuttgart) and EADS (Composite Research Centre, Munich) which will be pointed out in detail at p. 90. The modularization for the HEPHAISTOS systems has been assigned to the axial dimension to realize the capability of long tunnels. Independent to the diameter of the chamber, the length of a module comes to 1 m, such that a standardized WR340 waveguide feed can be used for all of the HEPHAISTOS systems. 12 optimized nonresonant slotted waveguides are arranged around the 6 edges of the hexagonal metal shell. Due to the high transmission efficiency of the waveguide systems, no related circulators and water cooling circuits are needed. The air cooled magnetrons are directly mounted on the WR340 waveguides. A conceptual module with 12 waveguides is depicted in Fig. 6.29. The power supplies can be mounted on plates thermally isolated on the applicator shell keeping the efforts for the high voltage wiring as low as possible. The commercial HEPHAISTOS processing systems are realized by choosing the diameter and the number of modules defining the length of the systems. It is
Fig. 6.28 Energy density calculated for a CFRP slab (15 × 30 cm) with the DELFI-Code at 2.45 GHz (left) and initial temperature answer in the central plane of the slab (right)
84
6 Processing Technology for Composite Materials
Fig. 6.29 Schematic HEPHAISTOS module (1 m length) with mounted WR340 waveguides
a prediction of the theoretical concept that the field homogeneity does not change or suffer by the change of these parameters. Table 6.2 shows e.g. the technical data for the HEPHAISTOS-CA1 (1 m length, 1 m diameter) and CA2 (2 m length, 1.8 m diameter). The CA1 consists of a volume that is about 5–6 times smaller than the CA2 which reduces significantly the number of available eigenmodes. To proof the Table 6.2 Technical data of HEPHAISTOS-CA1 (VHM100/100) and HEPHAISTOS-CA2 (VHM 180/200) Technical Data of HEPHAISTOS VHM Model
VHM 100/100
VHM 180/200
Rated temperature degree C Working chamber volume, ltr. approx. Inside dimensions Hexagon diameter mm width × height × depth mm Outside dimensions width × height × depth mm Weight, approx. kg Nominal voltage V, Hz Mikrowave rating kW
400 750
400 4200
1050 1050 × 910 × 1050
1800 1800 × 1560 × 2000
1500 × 800 × 1470 850 400 V, 50/60 Hz approx. 9,6 (12 × 0,8 kW)
Connected load Magnetron Heating system Working chamber material Control system
22,4 2,45 microwave stainless steel 1.4301 S!MCON/32-System
4200 × 2900 × 2200 1800 400 V, 50/60 Hz approx. 30 (24\break × 1,25 kW) 65 2,45 microwave stainless steel 1.4301 S!MCON/32-System
kW GHz
Design and Proof of the Modular 2.45 GHz HEPHAISTOS Conception
85
Fig. 6.30 CFRP slab cured in CA2 with sketched locations of DMA test samples
conservation of the predicted high degree of field homogeneity for the hexagonal concept in a process, two identical quadratic multidirectional prepreg slabs (0/90) have been cured under the same process conditions but each separated in the CA1 and CA2. The size of the slabs has been 400 × 400 × 2 mm. The process was defined for both with a heating rate of 5◦ C/min and a holding temperature of 135◦ C for 2 hours (Fig. 6.30). Five different locations for taking test samples out of the cured slab have been chosen. The test samples have been investigated with the Dynamic Mechanical 1000 149.87°C 148.51°C 149.25°C 147.26°C 150.45°C 149.93°C
Shear Loss Modulus (MPa)
800
146.81°C
–– –––– – –––– ––––– · ––– – – ––– ––– ––––– – –– –– – ––––––– –––– ––––– ·
CFK, CA 1, ProbeNr. 1 CFK, CA 1, Probe Nr. 2 CFK, CA 1, Probe Nr. 3 CFK, CA 1, Probe Nr. 4 CFK, CA 1, Probe Nr. 5 CFK, CA 2, Probe Nr. 5 CFK, CA 2, Probe Nr. 4 CFK, CA 2, Probe Nr. 3 CFK, CA 2, Probe Nr. 2 CFK, CA 2, Probe Nr. 1
146.91°C 146.63°C
600
400
200
0 0
50
100
150
Temperature (°C)
200
250
Universal V3.2B TA Instruments
Fig. 6.31 Dynamic Mechanical Analysis (DMA) of test samples cured in HEPHAISTOS-CA1 and CA2
86
6 Processing Technology for Composite Materials
Analysis (DMA) method that determines the mechanical properties of polymers. To compare the samples, the glass transition temperature Tg of each sample has been measured. The evaluation for these samples performed at the University of Stuttgart showed the following immanent results: • Tg for the 10 samples (CA1 and CA2) has been measured between 146 and 150◦ C • No significant differences in the polymerization link of the polymers for the CA1 und CA-2 samples could be observed The detailed course of the shear loss modulus is depicted in Fig. 6.31, showing the close peaks for the glass transition temperature representing the high processing versatility of the modular HEPHAISTOS concept that documents respectively the affecting field homogeneity.
Process Development The load-bearing mechanism in fibrous composites is greatly dependent on the connection between the fibres and the matrix. The matrix in the composites of topic is an epoxy resin, i.e. a polymer thermoset and a curing agent. For Prepreg materials, several certified systems are used in aerospace industries, while for injection technology the only certified resin used is RTM6 produced by Hexcel. The here described investigations on the process development had to show finally, that the material properties and load-bearing mechanisms of CFC structures fabricated in the HEPHAISTOS-systems are suited for aerospace application. A joined test program of mechanical material tests and measurements was carried out in collaboration with EADS Munich and University of Stuttgart (Institute for Aircraft Design, IFB). Figure 6.32 shows the in-principle lay up for microwave curing. As pointed out in the last chapter on the fundamentals of microwave heating for CFRP (e.g. Fig. 6.24), edges, corners and boundaries of CFC structures underlie a stronger heating. This effect can be easily reduced by shielding these boundary areas with an adhesive metal tape. Without the exception of the bagging film and silicon, the choice of aerospace tool materials is not changed due to their compability to the microwave curing process. Detailed dielectric measurements have been successfully taken for thin foils and bagging materials in a TE10 -waveguide. The most endangered material necessary to keep the vacuum during the process is the vacuum bag at the interface lay-up/environment that is usually a very thin polyamide polymer. The investigations showed the common materials to be suitable for microwave processes with temperatures up to 120◦ C, beyond this limit a local overheating due to melting and loss of the vacuum is likely. This problem is more severe, the more folds and wrinkles a vacuum bag has to consist of, which often cannot be avoided for asymmetric CFC lay-ups. A very universal choice for a vacuum bag material are Polyimide based materials like Upilex that is a novel low
Process Development Fig. 6.32 Microwave vacuum bag set-up
87 Thermo-Electrical Foil Bagging Film
Sealant Laminate Release Film Vacuum Hose
Aluminum Foil
Tool
cost Capton polymer which can resist temperatures up to 350◦ C. The versatility of capton foils for microwave processing is about 100%. Table 6.3 shows the differences in thickness (Upilex nearly twice as thick to the standard vacuum foil) and in terms of electric conductivity, which is about a third for Upilex (room temperature) and shows an in-principle lower amount on microwave dissipation than standard polyamide. In addition, the ratio will increase to differ at higher temperatures, as the polyamide gets more viscous (phase transition to liquid phase, see p. 14 for water) at temperatures higher than 120◦ C causing a significant higher microwave absorption. The membrane material used for the VAP process consists of a quite high electric conductivity, but consists of an extremely thin thickness. The foil is directly positioned on the CFC weave lay-up and soaked at higher processing temperatures by injected resin. Therefore, the temperature development of this tool material is governed by the lay-up environment; no volumetric run-away due to the extreme low thickness of this foil is possible. A material/overheating fail for microwave processes using these materials, as well for the standard release materials and fleeces has not been reported during all experiments. For the aerospace certified RTM6 resin, temperature dependent measurements of the dielectric properties have been performed at FZK-IHM. Table 6.3 Dielectric properties of bagging tool materials at 2.45 GHz Material
Thickness [mm]
Permittivity εr
Conductivity σ e [S/m]
Vacuum foil VAP membrane Release film Vacuum fleece Upilex
0.080 0.025 0.180 3.600 0.140
4.4 5.2 3.4 1.2 4.3
0.100 0.500 0.030 0.004 0.030
88
6 Processing Technology for Composite Materials ΔEN:1.2
δ− Δ EN:1.0
CH2 δ+
δ−
O
δ−
O CH δ+
CH2
δ+
δ+
C
O
C
δ+
CH δ+
CH2 δ+
n
O
O δ+
H
δ−
δ−
CH3 δ+
CH2
δ+
CH3
δ−
O
O
CH3
CH2 δ+
CH δ+
CH2
CH3
Fig. 6.33 Molecular structure of epoxy resin
The molecular composition of epoxy resin is shown in Fig. 6.33. The graph highlights the numerous “water-like” oxiran functional groups that form microwave absorbing molecular rests. In addition to the thermal Brownian motion, quantum rotational resonances of corresponding molecular rests (see p. 14) result in an enhanced mobility and collision frequency of the epoxy molecules. It can be concluded by the results depicted in Fig. 6.34 that the lowest viscosity and best properties (lowest void content) for injecting RTM6 for a microwave process is at temperatures between 110 and 145◦ C, as the dielectric properties decrease
Fig. 6.34 Dielectric properties of RTM6
Development of Thermo-Electric Foils
89
rapidly at higher temperatures. Depending on the degree of saturation into the tow sheet of the resin (VAP up to 70% of fibre content in the volume) and the quality of the adhesion for the fibre with the matrix the loading mechanics and strength of the composite are determined. To investigate and evaluate these material parameters, mechanical and thermal failure methods are used.
Development of Thermo-Electric Foils Further decisive optimization, power savings and reliability for the microwave process can be achieved using special Thermo-Electric (TE) foils. This additional tool component for the microwave process has been introduced to • thermally isolate the CFC structure from the environment • heat microwave reflecting materials. The unique selective heating mechanism of microwaves is significantly enhanced if direct thermal heat-loss from the CFC structure to the environment is reduced. This is important for very thin CFRP composites, as the volumetrically deposited power in very thin structures is immediately conducted by thermal conductivity to the cold metal tool causing a drop in the resulting heating rate and a related heating of the metal tool. But if the thickness of a CFC slab exceeds the electromagnetic coupling depth (1.5–2 mm at 2.45 GHz, see Fig. 5.12 for microwave penetration at 5.85 GHz for the interface CFC/metal), the instantaneous volumetric power deposition in the CFC structure causes controllable resulting heating rates to decouple thermally from the cooler massive metal tool, where a temperature gradient from the iso-thermal interface composite/tool to the bottom of the metal tool is generated. This means, more power is left volumetrically in the CFC structure as there can be lost by the bad perpendicular thermal conductivity of the fibres to the metal interface. For VAP processes, the TE-foil is essential to heat the dry, microwave reflecting carbon weave at low power to the injection temperature at 110–145◦ C. Without a resin matrix system, the components of the dielectric tensor for the raw carbon fibres increase extremely such that the microwave power absorption is due to the reflectivity reduced to a negligible amount. The TE-foils takes the microwave absorbing role and heats the lay-up in a fully controllable procedure to the injection temperature. Figure 6.35 shows the IR-sensored thermal distribution of a CFC-slab, covered by an elastomer TE-Foil. The overall temperature homogeneity sensored is very high (the fibre thermocouple wire can be seen lifting the TE foil slightly from the slab gathering a small temperature increase monitored by a white color), both for the TE-foil and the enclosed slab – for the heating ramp as well as in the holding stage (Fig. 6.35 taken at 180◦ C of the CFRP slab). Very promising base materials for the TE-foil development are elastomeres as they are used for vacuum sealants in the CFRP production.
90
6 Processing Technology for Composite Materials
Fig. 6.35 Elastomer TE-foil assisted CFRP slab processing
Fig. 6.36 Basic monomer of elastomeres
CH2
H
H
C
C
CH2 n
Figure 6.36 shows their molecular structure, which basically consists of no polar bonds. Electric conducting additives can originate an optimal microwave coupling to the microwave permeable material for industrial fabrication. The elastomer foil can be chemically adapted to resist temperatures up to 200–220◦ C. Further developments target to substitute the bagging/vacuum foil by a TE-bagged set-up. Alternatives on doped thin silicon materials for thermoelectric heating are under future consideration within the aerospace field. It could be demonstrated with the HEPHAISTOS systems that with the use of TE-foils even GLARE structures (metal composites with glass fibre inlays) have been cured and in consequence successfully evaluated at EADS providing a universal processing procedure for all kind of aerospace composites.
Systematic Material Investigations for Microwave Processed CFC Materials Extensive work for processing and testing CFRP materials has been performed very fruitfully together in a joined collaboration with Institute for Aircraft Design IFB, Stuttgart and EADS Corporate Research Centre, Ottobrunn. Especially the following parameters have been evaluated: Heating rates, resin systems, prepregs materials, textile semifinishes, infiltration techniques, difficult and complexly shaped geometries as well as tooling materials [104, 109]. For all these parameters, intense
Systematic Material Investigations for Microwave Processed CFC Materials
91
Fig. 6.37 RTM6 resin injected test slab fabricated with HEPHAISTOS-CA1
mechanical test to investigate the material quality have been performed where only a small summary/overview will be given in this chapter. A RTM 6 resin injected standard test slab cured at 180◦ C is visualized in Fig. 6.37 (30 × 15 cm). The complete processes were run completely automated with the HEPHAISTOS technology [96, 97, 100]. A typical process in an HEPHAISTOS-CA system contains the stages: initial heating and tempering phase up to 110◦ C, resin injection phase according the VAP process, final heating for curing to 180◦ C and at last the cooling down phase of the sample. A more detailed description of the VAP procedures is given in pp. 68–71 (Fig. 6.38). To demonstrate to feasibility of 3D structures, a T-profile has been successfully cured (Fig. 6.39). A specific difficulty for the CFRP fabrication is the immediate change of the material thickness. To consider this obstacle, different tests in prepreg- and infiltration technique have been applied successfully at significant high heating rates (8◦ C/min).
Fig. 6.38 Prepreg processing: Prepreg 913 system (left) and homogeneously cured sample (40 × 40 cm) after the process (HEPHAISTOS-CA1 System)
92
6 Processing Technology for Composite Materials
Fig. 6.39 A T-profile fabricated with the HEPHAISTOS-CA1 according the VAP infiltration technology and cured at 180◦ C
As a result, Fig. 6.40 shows a large sized CFRP plate (60 × 20 cm) fabricated in the HEPHAISTOS-CA1 microwave system. The curing was performed at 180◦ C using a standard RTM6 resin system [101]. Identical structures have been processed as well in prepreg [102, 104]. The utilized resin systems (e.g. 913) show a high exothermal reactivity. Due to the inertialess control of the volumetric microwave heating high heating rates can be applied without the danger of reactive overheating. An identical curing process in a conventional autoclave failed due to the high inertia for the temperature control that caused a total loss of the step sample due to strong overheating in the thicker sections and delamination [109] (Fig. 6.41). The successfully HEPHAISTOS cured step plate was further investigated at EADS, Ottobrunn. It turned out by ultrasound scans, that the three different sections of the plate of Fig. 6.40 showed a high homogeneity in each section appropriate for aerospace application [105, 106].
Fig. 6.40 Large RTM6 injected CFRP Plate (60 × 20 cm) with integrated steps 2 / 6 / 20 mm
Systematic Material Investigations for Microwave Processed CFC Materials
93
Fig. 6.41 Ultrasound investigations of the 3 sections of the step plate, thicknesses: left 2 mm, centre 6 mm, right 20 mm
For the mechanical material classification program, a precisely defined standard process for comparison reasons has been performed in the HEPHAISTOS system as well as in conventional systems: autoclave (autoclave used without pressure) and convection heat oven to obtain comparable data of the material properties. The materials, weave, resin etc. have been taken from the identical source. As a result, the investigations of the Dynamic Mechancial Analysis (DMA), Differential Scanning Calorimetry (DSC), ultra sound and Interlaminary Shear (ILS)-tests showed in general enhancements of the material qualities for the microwave processed materials [108–110, 112]. Each test method was applied on a significant number of standard CFC slabs as shown in Fig. 6.37. Figure 6.42 shows a charge of three RTM6 injected test slabs documenting the high material homogeneity (the difference in amplitude reported by the measurements is approximately 4–5 dB). On the C-scan image the pleats of sample RTM6-I9 (right), which can be seen by naked eye on the surface, are visible. More quantitative results summarized by the complete test slab charges are given in the following charts. As the results in Figs. 6.43 and 6.44 show, enhancements of the mechanical material properties in comparison to conventional curing have been proofed. The theoretical considerations on the anisotropic dielectric properties of CFC materials showed that the microwave heating of the fibres is caused mainly by the contribution parallel to the conducting carbon fibre pelec (x , t) ≈
1 σ E 2 (x, t) 2
(6.15)
In conventional systems, the heat is transferred by the thermal conductivity of the resin matrix system (the thermal conductivity of the carbon fibres is poor) while
94
6 Processing Technology for Composite Materials
Fig. 6.42 Ultrasound C-scan of a charge of 3 standard test slabs
Fig. 6.43 process
Comparison of Celanese pressure tests with different oven systems using the VAP
Systematic Material Investigations for Microwave Processed CFC Materials
95
Fig. 6.44 Comparison of interlaminary shears strength with different oven systems using the VAP process
the power is driven into the structure by the thermal gradient of the environment temperature and the local temperature of the composite (Fig. 6.45). With microwaves, the direct heating of the parallel microwave field component at the fibre improves the interfacial bonding and enhances the mechanical interlocking between the resin and the fibres. In addition, an improvement for the resin flow in a microwave field has a direct effect on the void content obtained for the microwave cured composites. The mobility and fluidity of the resin is enhanced due to rotational
O CH2
O
CH3
CH CH2
O
O CH2
C
CH3
n React io l e Chann
n React io l e Chann
O CH2
O CH CH 2
Electromagnetic Field CH3 C
O
O CH 2
CH
CH2
Epoxy Matrix
CH3
Fig. 6.45 Reaction mechanism for microwave cured CFC materials
CH
CH2
96
6 Processing Technology for Composite Materials
Fig. 6.46 A 40 kg aluminum tool for a VAP injected presentation structure (HEPHAISTOS-CA1)
quantum states as an additional degree of freedom compared to conventional thermal heating. The diffusivity for the resin in the microwave increases forcing trapped air to leave micro-holes in the fibre. A better contact between the resin and the fibre surface contour is the result which causes better mechanical material properties as reported in these investigations [116].
Fig. 6.47 The curved VAP presentation CFRP-structure (HEPHAISTOS-CA1)
Systematic Material Investigations for Microwave Processed CFC Materials
97
Fig. 6.48 The prepreg-presentation structure of Fig. 6.5 ready for curing at FZK (HEPHAISTOSCA2) on a curved metal tool provided by EADS
At last, the next pictures give vivid impressions for complexly shaped structures processed in large metal tools. The structures have been fabricated in the HEPHAISTOS-CA1 and CA2 systems (next page) (Figs. 6.46, 6.47, 6.48 and 6.49).
Fig. 6.49 The prepreg-presentation structure of Fig. 6.5 after curing – the excellent surface properties are obvious
Chapter 7
Summary and Outlook
In this work, energy efficient microwave system and materials processing technologies for avionic, mobility and environmental applications have been developed and presented for the first time. The developments have been in close interaction of theoretical and computational methods with experimental work in very close cooperation to industry. The microwave technologies showed that significant energy savings for in-flight applications and future lightweight composite aircrafts could have been demonstrated. Novel industrial microwave systems with enormous innovative potentials for energy and resource efficient materials processing are now available [114]. Large high frequency heating systems based on these developments will get more and more important for the materials processing industry as the kWh-prices for energy consumption will continuously increase for the future. In Fig. 7.1, a comparison of the total net energy consumption for an identical standard RTM6 curing process performed in a microwave system (HEPHAISTOSCA1), a thermally heated system VTU 200/200/300 and an autoclave at EADS is visualized. As the results show, the microwave technology takes the lowest consumption. The autoclave needed in addition electric power after the holding cycle (t = 4 h) to cool the complete shell and environment of the autoclave chamber down; the electric consumption of the microwave and the thermal heating system could be switched off at this point. The gain for the microwave energy efficiency can still be optimized by developing a specific microwave process. The heating rate using microwaves can be easily increased with appropriate microwave sources to reduce the processing time. Estimations show, that power savings of the factor 8–10 are realistic to be achieved for large installations at reduced overall cycle times and high material quality. Due to the volumetric heating of the microwaves, the curing of the resin starts earlier. By optimizing the resins´ polymerization with microwaves adding specific chemical additives another reduction for the processing time and electric power level will be possible. These additives are worked out together with the companies of BASF and Hexion Specialties in a national funded project (BMBFproject, see Fig. 7.2) [115]. The development on new and fast high quality resin systems is very important, as the demands in curing for automotive industry for large numbers of parts is within a timescale of minutes in contrast to aerospace, where the current certified autoclave L. E Feher, Energy Efficient Microwave Systems, C Springer-Verlag Berlin Heidelberg 2009 DOI 10.1007/978-3-540-92122-6 7,
99
100
7 Summary and Outlook
Fig. 7.1 Comparison of the specific energy consumption for an identical certified standard process
processes demand several hours for rather few parts. For the future, a convergence of these demands can be foreseen – microwave technology offers here a unique contribution to activate the polymerization on the molecular and quantum level combined with a volumetric penetration of the structure. The stage of development is now continuously focusing on “tailored microwave processes” after a reliable system-line
Fig. 7.2 Consortium of national funded microwave R&D project scheduled until 2010
7
Summary and Outlook
101
Fig. 7.3 R&D milestones for aerospace HEPHAISTOS-application
of high frequency systems is industrially available. For aerospace composites, a qualification procedure for the mature systems and processes will be the final step towards the industrial application which is targeted in 5–7 years (see Fig. 7.3) by the aerospace industry for the A350 and A31X program [116]. The challenge of energy supply security and the volatility of fossil fuel prices have become pressing on solutions for the increasing electricity demand. In addition to the processing of lightweight composites for aerospace and mobility applications, large wind energy structures suffer similar demands for fast and reliable processing. As an outlook, the HEPHAISTOS system technology will be adapted in future work to the needs for large and thick rotor blades. As well the development for continuous HEPHAISTOS systems is worked out as the HEPHAISTOS-CA3 system, being currently the largest microwave chamber system in the world (7000 l volume) is equipped first time with a throughput capability and will be taken into service in September 2007. The anti-/de-icing microwave technology discussed in this work will be available for aircrafts in the future that change leading edge structures of wings and tailplanes from metal to composites. Turbine inlets of large passenger aircrafts are on current consideration with aerospace manufacturers for future application of novel microwave anti-/de-icing methods. Even for wind energy rotor blades, the icing problem is similar especially for remotely working off-shore wind parks in the northern hemisphere. These composite rotor blades could be equipped with an automated microwave de-icing system which takes the needed energy directly from the wind energy turbine such that generator halts for security reasons on icing conditions are not necessary any more.
102
7 Summary and Outlook
The significant advantages of these technologies as well as a comparison of the microwave material properties with conventionally processed samples have been shown evidently. The observed and independently proofed material enhancements could be explained by theoretical considerations of the electromagnetic interactions within the laminate. As a conclusion, modeling and simulation have been playing an important role to achieve homogeneous field and temperature conditions for materials processing. Concepts from 30 GHz mm-wave technology have been successfully transferred in this work to the lower ISM frequency of 2.45 GHz and continuously optimized. Unfortunately, the use of higher frequencies in the mm-wave regime does not show evidence for industrial application at present as pointed out in pp. 5–8, 43, 50, 76–86. The reasons are • Poor net energy transmission efficiency to the load • Temperature answer for materials at higher frequencies is not enhanced • Availability of appropriate mm-wave sources, equipment and related costs Nevertheless, mm-waves and gyrotron technology are excellent for scale up considerations and fundamental research topics. A powerful computational and analytical set of tools (MiRa-, DELFI-, THESIS-Code) has been used in this work to clarify and predict material interactions of microwaves at 30 and 2.45 GHz. Detailed investigations on dielectric measurements of anisotropic CFRP composites and related heating mechanisms have been intensively discussed and explained. The outstanding high values of the measured complex dielectric constant of CFRP materials have to be explained theoretically in the future. As the treatment on the dielectric constant of water showed, quantum mechanical considerations are very essential and will be intensified in the future for profound explanations of microwave heating with details on the composition of the microwave effective electric conductivity and related energy conversion paths in materials.
References
Chapter 1 1. Yu.V. Bykov, A.G. Eremeev, V.A. Flyagin, V.V. Kaurov, A.N. Kuftin, A.G. Luchinin, O.V.Malygin, I. Plotnikov, V.E. Zapelov, L.Feher, M. Kuntze, G. Link, M. Thumm, “Gyrotron Installation For Millimetre-Wave Processing Of Materials”, Proc. Vakuumelektronik und Displays, ITG Fachtagung 2.-3. Mai, Garmisch-Partenkirchen, 1995, pp. 103–108 2. L. Feher, “Computer Simulations for Process-Technical Application of Millimetre-Waves for Industrial Processing of Materials” (Ph.D. Thesis, FZKA 5885, Forschungszentrum Karlsruhe GmbH, 1997) 3. G. Link, L. Feher, M. Thumm, Sintering of Advanced Ceramics Using a 30-GHz, 10-kW, CW Industrial Gyrotron, IEEE Transactions on Plasma Science, Vol. 27, No. 2, April 1999, pp. 547–554 4. E.C. Okress (Editor),“Microwave Power Engineering”, Vol 2., Academic Press, New York, 1968 5. L. Feher, G. Link, M. Thumm, Hochmodiger Mikrowellenresonator f¨ur die Hochtemperaturbehandlung von Werkstoffen, DE196 33 245 C1, Erteilt 17.7.1997, PCT, EP, US, JP, RU, KR, IN 6. L. Feher, M. Thumm, Numerische Modellierung f¨ur die industrielle Materialverarbeitung mit Millimeterwellen, Nachrichten – Forschungszentrum Karlsruhe, 28(1996), pp. 215–223 7. L. Feher, M. Thumm, Simulationsunterst¨utzte Entwicklung industrieller Millimeterwellentechnologie f¨ur die Materialprozeßtechnik, Mikrowelleneinsatz in den Materialwissenschaften, der chemischen Verfahrenstechnik und in der Festk¨orperchemie, M. WillertPorada (Hrsg.), Shaker Verlag, Berlin, 182–197 (1998)
Chapter 2 8. R.A.Cairns, A.D.R. Phelps, “Generation and Application of Microwaves”, SUSSP Publications, 1997, ISBN 0-7503-0-474-0 9. J. Bretting, “Technische R¨ohren – Grundlagen, Funktionen, Anwendung”, H¨uthig Buch Verlag Heidelberg, 1991, ISBN 3-7785-1645-0 10. L. Feher, G. Link, M. Thumm, Innovative challenge of commercializing industrial millimetre-wave processing technology at 24 GHz, 2nd World Congress on Microwave & Radio Frequency Processing, Orlando, USA, April 2–6, 2000
103
104
References
11. L. Feher, G. Link, M. Thumm, Comparison of Microwave and Millimetre-Wave Materials Processing, 25 th International Conference on Infrared and Millimetre Waves, September 12–15 th, 2000, Bejing, China, pp. 475–476 12. M. Thumm, L. Feher, Millimetre-Wave-Sources Development, Present and Future, 8th Conference on Microwave and High Frequency Heating, Bayreuth, Germany, Sept. 3–7, 2001, Springer, pp. 15–23, ISBN 3-540-43252-3 13. M. Thumm, L. Feher, K-band vacuum electron tubes for materials processing, present and future, 3rd IEEE Internat. Vacuum electronics, Conf. (IVEC 2002), Monterey, Calif., April 23–25, 2002, pp. 371–372
Chapter 3 14. K. Kopitzki, “Einf¨uhrung in die Festk¨orperphysik”, Teubner Studienb¨ucher, 1986 15. G.Lehner, “Elektromagnetische Feldtheorie”, Springer Verlag 1990, ISBN 3-540-56873-5, p. 450 16. J. D. Jackson, “Klassische Elektrodynamik”, Verlag W. de Gruyter, 1982 17. A.C. Metaxas , R.J. Meredith, “Industrial Microwave Heating”, IEE Power Engineering Series 4, Peregrinus Ltd., 1988, ISBN 0909048893 18. K. Okada, Y. Imashuku, and M. Yao, “Microwave spectroscopy of supercritical water ”, The Journal of Chemical Physics, December 8, 1997, Volume 107, Issue 22, pp. 9302–9311 19. C.Cohen-Tannoudji, B.Diu, F.Laloe, “Quantenmechanik Teil 2”, De Gruyter, Berlin 1997, ISBN 3-11-01 5859-0, p. 68 20. T. Fließbach, “Quantenmechanik”, Spektrum Akademischer Verlag 2000, ISBN 3-82740996-9, p. 20, p. 301 21. H. E. Stanley, S. V. Budyrev, M. Canpolat, S. Havlin, O. Mishima, M. R. Sadr-Lahijany, A. Scala, F. W. Starr, The puzzle of liquid water: a very complex fluid, Physica D, 133 (1999) 453–462. 22. P. G. Kusalik and I. M. Svishchev, The spatial structure in liquid water, Science 265 (1994) 1219–1221. 23. G. Malenkov, Structure and dynamics of liquid water, Journal of Structural Chemistry, Volume 47, Supplement 1, September 2006 , pp. S1–S31(1) 24. L. Feher, E. Borie, Electromagnetic Wave Propagation – A methodical Discussion on analytical and numerical Field Prediction Techniques, 29th Conference on Infrared and MillimetreWaves, September 24th–October 1st, 2004, Karlsruhe, pp. 835–836 25. L. Feher, M. Thumm, System design development for microwave and millimetre-wave materials processing, Intense Microwave Pulses IX, Orlando, Fl., April 2–3, SPIE 2002, pp. 75–79
Chapter 4 26. W.J.R. Hoefer, The Transmission-Line Matrix (TLM) Method, Numerical Techniques for Microwave and Millimetre-Wave Passive Structures, T. Itoh Ed., John Wiley & Sons, New York 1989 27. H. Henke, “Elektromagnetische Felder”, Springer Verlag 2004, ISBN 3-540-40417-1, p. 272 28. K. K¨upfm¨uller. W. Mathis, A. Reibiger, “Theoretische Elektrotechnik”, Springer Verlag 2005, ISBN 3-540-20792-9, p. 598 29. S. Stanculovic, L. Feher, M. Thumm, Optimization of slotted waveguides for 2.45 GHz applicators, 9th Int. Conf. on Microwave and High Frequency Heating, 2nd–5th September 2003, Loughborough, U.K., pp. 313–316
References
105
30. S. Stanculovic, L. Feher, M. Thumm, Optimization of slotted Waveguides for 2.45 GHz Applicators using novel slot Types, Novel Materials Processing by Advanced Electromagnetic Energy Sources, March 19–22, 2004, Osaka, Elsevier, pp. 135–138, ISBN 0-080-44504-7 31. S. Stanculovic, L. Feher, M. Thumm, Optimization of slotted waveguide feeding for 2.45 GHz applicators using new slot shape, 13rd Int. Conf. on Antennas, Nice, France, November 8–10, 2004, pp. 392–393 32. S. Silver, Microwave Antenna Theory and Design, New York, Mc Graw-Hill, 1949 33. S. Stanculovic, L. Feher, M. Thumm, Design of travelling wave slotted waveguide feeds for 2.45 GHz industrial microwave heating system, 10th International Conference on Microwave and Radiofrequency heating, Modena, Italy, Sept. 12–15, 2005, pp. 454–457 34. S. Stanculovic, L. Feher, M. Thumm, Design of travelling wave slotted waveguide feeds for 2.45 GHz industrial microwave heating system, 10th International Conference on Microwave and Radiofrequency heating, Modena, Italy, Sept. 12–15, 2005, pp. 454–457 35. S. Stanculovic, L. Feher, M. Thumm, Slotted waveguides for 2.45 GHz industrial applicators, Noordwijk, ESA, 2005, pp. 21–24 36. M.Afsar et al., The measurement of the properties of materials, Proceedings IEEE, Vol. 74, pp. 183–199 37. J.B. Jarvis et al., Improved technique for determining complex permittivity with the transmission/reflection method, IEEE Trans. Microwave Theory Tech, Vol. MTT-38, pp. 1096–1103, August 1990 38. W.B. Westphal (Editor), “Dielectric Constant and Loss Data”, Report Number AFML-TR740-250, Part III, June 80 39. J. Akhtar, L. Feher, M.Thumm, A Robust Optimization Algorithm for the Reconstruction of Dielectric Properties of Lossy Composite Materials, 29th Conference on Infrared and Millimetre-Waves, September 24th–October 1st, 2004, Karlsruhe, pp. 387–388 40. J. Akhtar, L. Feher, M. Thumm, A Generalized Approach For Measuring The Dielectric Properties Of Lossy Composite Materials, 4th World Congress on Microwave and RF Applications, November 7th–12th, 2004, Austin, USA, pp. 146–147 41. J. Akhtar, L. Feher, M. Thumm, A Multi-Layered Waveguide Technique for Determining Permittivity and Conductivity of Composite Materials, German Microwave Conference, Karlsruhe, 2005 42. J. Akhtar, L. Feher, M. Thumm, Measurement of Dielectric and Conductive Properties of Avionic Materials at 2.45 GHz Using Two-Step Approach, IMPI s 39th Annual Microwave Symposium, July 13–15, 2005, pp. 58–61 43. J. Akhtar, L. Feher, M. Thumm, The Measurement of Complex Permittivity Tensor of Uniaxial Anisotropic Composite Materials Using a Waveguide Method, 10th International Conference on Microwave and Radiofrequency Heating, Modena, Italy, Sept. 12–15, 2005, pp. 24–27
Chapter 5 44. Del Core (Editor), Advances in Onboard Systems Technology, EC Aeronautical Research Series, John Wiley & Sons, New York, 1994, pp. 1–29. 45. M. P. Simpson, P.M. Render, Investigation of the certification and operational procedures for turboprop aircraft in icing, The Aeronautical Journal, October 1999, pp. 449–454. 46. M.T. Brahimi, P. Tran, D. Chocron, F. Tezok, I. Paraschivoiu, Effect of Supercooled Large Droplets on Ice Accretion Characteristics, 35th Aerospace Sciences Meeting & Exhibit, Reno, NV, January, 6–10, 1997.
106
References
47. L. Feher, M. Thumm, High Frequency Microwave Anti-/De-Icing System for Carbon Reinforced Airfoil Structures, SPIE 15th Annual International Symposium on Aerospace, Simulation and Controls, Orlando, U.S.A., 16–20 April 2001, pp. 99–110 48. L. Feher, M. Thumm, Design of Avionic Microwave De-/Anti-Icing Systems, 8th Conference on Microwave and High Frequency Heating, Bayreuth, Germany, Sept. 3–7, 2001, Springer, pp. 695–702, ISBN 3-540-43252-3 49. L. Feher, Enteisung von Flugzeugen mit Mikrowellen, DE 197 50 198 gemein-schaftlich mit DaimlerChrysler Aerospace Airbus GmbH, Erteilt 31.5.99, PCT, US, JP, RU 50. L. Feher, Kompaktes mikrowellentechnisches System zum Enteisen und/oder Vorbeugen einer Vereisung der a¨ ußeren Oberfl¨ache von meteorologischen Einfl¨ussen ausgesetzten Hohlraum- oder Schalenstrukturen, DE 10016261, Erteilung 18.10.2001 51. L.Feher, Confidential Report “Feasibility Study DO-328 Anti-/De-Icing System” for Fairchild Dornier, 2002
Chapter 6 52. C.Y. Niu, Composite Airframe Structures, Hong Kong Conmilit Press Ltd., 1992 53. J. Brandt, K. Drechsler, J. Filsinger, New Approaches in Textile and Impregnation Technologies for the Cost-Effective Manufacturing of CFRP Aerospace Components, ICAS 2002, Toronto, (2002) 54. S. Zhou, J. West, M. C. Hawley and J. Wei, “Fast and Uniform Microwave Adhesive Bonding of Polymers”, Proceedings of the Midwest Advanced Materials and Processing Conference, Dearborn, Michigan, September 12–14, 2000. 55. Y. Qiu and M.C. Hawley, ”Uniform Processing of Complex Geometry Composite Parts Using Microwaves”, International Mechanical Engineering Congress & Exposition 1999, ASME, Nashville, TN, November 1999 56. L. Feher, M. Thumm, Millimetre-Wave Processing of Composite Materials, 2nd IEEE International Vacuum Electronics Conference, Noordwijk, Netherlands, 2–4 April 2001, pp. 83–84 57. L. Feher, A.S. Herrmann, J. Kleffmann, M. Kleineberg, A. Pabsch, Verfahren und Mikrowellensystem zur thermischen Prozessierung von aus Ausgangsmaterialien zusammengesetzten Formk¨orpern zu formbest¨andigen dreidimensionalen Kompositen, DE-PS 19 929 666 58. B. Reßler, M. Achternbosch, K.-R. Br¨autigam, C. Kupsch, G. Sardemann, Technikfolgenabsch¨atzung – Theorie und Praxis Nr. 1, 11. Jg. (2002) 59. A. Taflove, “Computational Electrodynamics – The Finite-Difference Time-Domain”, Artech House, 1995, ISBN 0-89006-792-9 60. T. Barts, “Maxwell s Grid Equations”, Frequenz 44,1990, 9 61. K.S. Yee, Numerical solution of initial boundary value problems involving Maxwell s equations in isotropic media , IEEE Trans. Antennas Propagation, Vol- AP-14, No. 5, May 1966, 302–307 62. A.Sommerfeld, “Optik”, Verlag Harri Deutsch, ISBN 3-87144-377-8, 180 63. V.A. Borovikov, B.Ye. Kinber, ‘‘Geometrical theory of diffraction”, IEEE Waves Series 37, ISBN 0-85296-830-2 64. Yu A. Kravtsov, Yu I. Orlov, “Caustics, Catastrophes and Wave Fields”, Springer Verlag, ISBN 3-540-56587-6, 1993 65. L. Feher, G. Link, M. Thumm, “Microwave Raytracing in Large Overmoded Industrial Resonators”, Proc. Meeting American Ceramic Society 95, Microwaves, Theory and Application in Materials Processing III Symposium, Cincinnati, April 30–May 3, 1995, 159
References
107
66. L. Feher, M. Thumm, “Numerische Bestimmung von Feldverteilungen f¨ur die Materialprozesstechnik mit Millimeterwellen”, Proc. Klausurtagung Mikrowelleneinsatz in den Materialwissenschaften, der chemischen Verfahrenstechnik und in der Festk¨orperchemie, Dortmund, 2.-3. November, 1995, 3 67. L. Feher, G. Link, M. Thumm, Microwave raytracing in large overmoded industrial resonators, Symposium on Microwaves: Theory and Application in Materials Processing III, Annual Meeting of the American Ceramic Society in Cincinnati, 1–4 May 1995, pp. 159–165. 68. L. Feher, G. Link, M. Thumm, Theoretical aspects for microwave ray tracing calculations in screened structures, Proc. Latsis Symposium 1995 on Computational Electromagnetics, ETH Z¨urich, Switzerland, 1995, pp. 236–241. 69. L. Feher, G. Link, M. Thumm, “The MiRa/THESIS3D-Code Package for Resonator Design and Modeling of Millimetre-Wave Material Processing”, Proc. MRS Spring Meeting, Symposium Microwave Processing of Materials V, San Franciso April 8–12, 1996, 363–368 70. G. Barton, “Elements of Green s Functions and Propagation”, Oxford Science Publications, ISBN 0-19-851998-2, 345 71. K. Drechsler, [Miravete, A. (Ed.)], “3-D textile reinforced composites for the transportation industry”, In: 3-D textile reinforcements in composite materials Cambridge: Woodhead, Publishing Ltd, 1999, 43–66 72. K. Drechsler, “Beitrag zur Gestaltung und Berechnung von Faserverbundwerkstoffen mit dreidimensionaler Textilverst¨arkung”, Dissertation an der Fakult¨at f¨ur Luft- und Raumfahrttechnik der Universit¨at Stuttgart, Stuttgart 1992. 73. L. Feher, M. Thumm, HEPHAISTOS – ein neuartiges Mikrowellensystem f¨ur die Produktion von kohlefaserverst¨arkten Verbundwerkstoffen, Nachrichten – Forschungszentrum Karlsruhe, 35(2003), pp. 123–127 74. L. Feher, M. Thumm, Microwave Innovation for Industrial Composite Fabrication. The HEPHAISTOS Technology, IEEE Transactions on Plasma Science, Vol. 32, No. 1, February 2004, pp. 73–79 75. L. Feher, G. Link, M. Thumm, The MiRa/THESIS-Code package for resonator design and modelling of millimetre-wave material processing, 1996 Spring Meeting of the Materials Research Society, Microwave Processing of Materials V, San Francisco, 1996, Symposium Proc., Vol. 430, pp. 363–368. 76. L. Feher, M. Thumm, Modeling of millimetrewave materials processing, Proc. 21st Int. Conf. on Infrared and Millimetre Waves, Berlin, 1996, Contributed Paper AW2. 77. L. Feher, M. Thumm, Advanced modelling and simulation of millimetre-waves ceramics sintering with the THESIS3D-Code, Proc. 2nd European Workshop on Microwave Processing of Materials, Karlsruhe, 1997, pp. 19–20. 78. L. Feher, G. Link, M. Thumm, Optimized design of an industrial millimetre wave applicator for homogeneous processing of ceramic charges, 6th Int. Conf. on Microwave and High Frequency Heating, Fermo, Italy, Sept. 9–13, 1997, pp. 443–446 79. L. Feher, G. Kriegsmann, G. Link, M. Thumm, Investigations on self-consistent Field Calculations for Hot Spot Development in Millimetre/Microwave Processed Ceramics, Klausurtagung Mikrowelleneinsatz in den Materialwissenschaften, der chemischen Verfahrenstechnik und in der Festk¨orperchemie, Dortmund, M¨arz 1998, p. 19 80. L. Feher, G. Link, M. Thumm, Optimized applicator design for industrial millimetre wave processing, CIMTEC 98, June 14–19, 1998, Florence, Italy 81. L. Feher, G. Link, M. Thumm, Hot Spot Formation and Development in Millimetre/ Microwave Processed Ceramics, International Conference on Microwave Chemistry, Prague, Sept. 6–11, 1998, Book of Abstracts OR15 82. L. Feher, G. Link, M. Thumm, Electrothermal Heating Model for Microwave/HybridProcessed Materials, 24th International Conference on Infrared and Millimetre-Waves, Monterey, California, Sept. 6–10, 1999, F-B5
108
References
83. L. Feher, G. Link, M. Thumm: Electrothermal Heating Effects and Temperature Gradients in Microwave Processed Materials, 7th International Conference on Microwave and High Frequency Heating, Valencia, Spain, Sept. 13–17, 1999, pp. 435–438 84. L. Feher, M. Thumm, Commercial potentials of microwave systems for processing composite materials, 13th Joint Russian-German Workshop on ECRH and Gyrotrons (STCMeeting), IPP Greifswald, July 16–21, 2001, pp. 455–470 85. L. Feher, M. Thumm, Application of Higher Frequency Microwaves for Composite Materials Heating/Processing, Displays and Vacuum Electronics, ITG, Garmisch-Partenkirchen, May 2–3, 2001, pp. 363–368 86. L. Feher, M. Thumm, Industrial Higher Frequency Microwave Processing of Composite Materials, 8th Conference on Microwave and High Frequency Heating, Bayreuth, Germany, Sept. 3–7, 2001, Springer, pp. 682–686, ISBN 3-540-43252-3 87. Hunyar, C., L. Feher, M. Thumm, Investigations in millimetre-wave processing of an unidirectional carbon-fibre reinforced composite (CFRP) material, Intense Microwave Pulses IX, Orlando, Fl., April 2–3, 2002, pp. 81–89 88. L. Feher, C. Hunyar, G. Link, P. Pozzo, M.Thumm, Development of Novel System Technology for Microwave Processing of CFRP, ICOPS 2002, Banff, Canada, May 26–30, 2002 89. L. Feher, M. Thumm, HEPHAISTOS – Development of a Novel Automated Microwave Processing System for Carbon Reinforced Fibre Plastics (CFRP), 3rd World Congress on Microwave & RF Applications, Sydney, Australia, 22–26 September 2002, pp. 291–299 90. L. Feher, M. Thumm, Microwave Innovation for Industrial Composite Fabrication -The HEPHAISTOS Technology, 2003 IEEE International Conference on Plasma Science, June 2–5, 2003, Jeju, Korea 91. C. Hunyar, L. Feher, M. Thumm, Investigation of heating effects of CFRP samples in a 30 GHz gyrotron furnace by experiment and simulation, 15th Joint Russian-German STC Workshop on ECRH and Gyrotrons, Karlsruhe/Stuttgart/Garching, June 25–July 1, 2003 92. L. Feher, A. Flach, V. Nuss, P. Pozzo, T. Seitz, HEPHAISTOS – A novel 2.45 GHz Microwave System for Aerospace Composite Fabrication, 9th Int. Conf. on Microwave and High Frequency Heating, 2nd–5th September 2003, Loughborough, U.K., pp. 441–444 93. C. Hunyar, L. Feher, M. Thumm, Experimental and simulation approach for the description of heating effects in millimetre-wave processed CFRP composites, 9th Int. Conf. on Microwave and High Frequency Heating, 2nd–5th September 2003, Loughborough, U.K., pp. 437–440 94. M. Paulson, L. Feher, M. Thumm, Improving Heating uniformity by parameter optimization of a stationary electro-thermal model, 9th Int. Conf. on Microwave and High Frequency Heating, 2nd–5th September 2003, Loughborough, U.K., pp. 209–212 95. L. Feher, A. Flach, V. Nuss, P. Pozzo, T. Seitz, M. Thumm, Innovation for Aerospace CFRP Fabrication with High Electromagnetic Power Heating – The HEPHAISTOS System, 28 th Conference on Infrared and Millimetre-Waves, September 29th–October 3rd, 2003, Otsu, Japan, pp. 317–318 96. L. Feher K. Drechsler, F. Filsinger, Neueste Entwicklungen bei der Mikrowellenh¨artung von Faserverbundwerkstoffen, 10. Nationales Forum “Impulse f¨ur die Faserverbundtechnologie, SAMPE Deutschland, 12.–13. Februar 2004, Dresden 97. L. Feher, M. Thumm, Aerospace CFRP Structure Fabrication with the 2.45 GHZ HEPHAISTOS System, Novel Materials Processing by Advanced Electromagnetic Energy Sources, March 19–22, 2004, Osaka, Osaka, Elsevier, pp. 129–133, ISBN 0-080-44504-7 98. M. Paulson, L.Feher and M. Thumm, Quasi-stationary Electro-thermal Heating Model for Microwave/Hybrid-processed Materials Using Greens Function Techniques, Novel Materials Processing by Advanced Electromagnetic Energy Sources, March 19–22, 2004, Osaka, Osaka, Elsevier, pp. 125-128, ISBN 0-080-44504-7 99. L. Feher, K. Drechsler, A. Flach, J. Filsinger, T. Herkner, V. Nuss, T. Seitz, HEPHAISTOS -A novel modular 2.45 GHz Microwave Processing System for Pressure less Aerospace
References
100.
101.
102.
103.
104.
105.
106.
107.
108.
109.
110.
111.
112. 113.
114.
115.
116.
109
Composite Fabrication, 25th SAMPE Europe International Conference, March 30-April 1st, 2004, Paris L. Feher, V. Nuss, T. Seitz, M. Thumm, Aerospace Structural Composite Fabrication with the 2.45 GHz HEPHAISTOS System, 2004 International Symposium on Microwave Science and Its Application to Related Fields July 28th-30th, 2004, Takamatsu, Japan, pp. 86–90 L. Feher, A. Flach, V. Nuss, T. Seitz, M. Thumm, Aerospace Composites Microwave Processing with the 2.45 GHz HEPHAISTOS System, 29th Conference on Infrared and Millimetre-Waves, September 24th–October 1st, 2004, Karlsruhe, pp. 833–834 L. Feher, A. Flach, V. Nuss, T. Seitz, M. Thumm, Industrial Composite Curing with the 2.45 GHz HEPHAISTOS System, 4th World Congress on Microwave and RF Applications, November 7th-12th, Austin, USA C. Hunyar, L. Feher, M. Thumm, Simulations and Experiments on the Effects of MillimetreWave Heating of Orthotropic and Anisotropic CFRP Composites, 4th World Congress on Microwave and RF Applications, November 7th-12th, 2004, Austin, USA, 170–171 L. Feher, K. Drechsler, J. Filsinger, Composite Manufacturing by using a novel modular 2.45 GHz Microwave Processing System, 36th International SAMPE Technical Conference, San Diego, November 15–18, 2004 L. Feher, K. Drechsler, J. Filsinger, F. Karl, Development of the modular 2.45 GHz HEPHAISTOS-CA2 Microwave Processing System for Automated Composite Fabrication, SAMPE EUROPE International Conference 2005, Paris, April 4th–7th, 2005 L. Feher, K. Drechsler, M. Thumm, Development of an Industrial 2.45 GHz HEPHAISTOSCA2 Microwave Processing System for Avionic Composite Fabrication, IMPI s 39th Annual Microwave Symposium, July 13–15, 2005, pp. 53–57 L. Feher, T.Seitz, V. Nuss, Mikrowellenresonator, eine aus einem solchen Mikrowellenresonator modular aufgebaute Prozessstrasse, ein Verfahren zum Betreiben und nach diesem thermisch prozessierte Gegenst¨ande/Werkst¨ucke”, DE 103 29 411, 22.9.2005 J. Filsinger, L. Feher, New approaches for the application of microwave heating for processing composite materials in the aeronautic industry, IMPI s 39th Annual Microwave Symposium, July 13–15, 2005, pp. 50–52 R. Gr¨aber, L. Feher, K. Drechsler, J. Filsinger, Microwave curing of high performance composite materials for aerospace applications with Hephaistos, IMPI s 39th Annual Microwave Symposium, July 13–15, 2005, pp. 62–65 L. Feher, K. Drechsler, J. Filsinger, Development of the Industrial 2.45 GHz HEPHAISTOSCA2 Microwave Processing System for Composite Fabrication, 10th International Conference on Microwave and Radiofrequency heating, Modena, Italy, Sept. 12–15, 2005, pp. 56–59 L. Feher, E. Borie, Electromagnetic Wave Propagation–A methodical discussion on numerical field prediction techniques, 10th International Conference on Microwave and Radiofrequency heating, Modena, Italy, Sept. 12–15, 2005, pp. 412–414 L. Feher, K. Drechsler, Development of Industrial 2.45 GHz Microwave Processing Technology for Composite Applications, Composites Europe 2005, October 6–7, 2005, Barcelona L. Feher, et el., Gigahertz and Nanotubes – Perspectives for Innovations with Novel Industrial Microwave Technology, Advanced Engineering Materials, Vol. 8, No.1–2, February 2006, pp. 26–32 L. Feher, K.Drechsler, R.Wieseh¨ofer, J.Filsinger, SAMPE Journal, The Industrial HEPHAISTOS System Line for Microwave Processing of High Performance Composites, SAMPE Journal Vol. March/April, 2007 L. Feher, K. Drechsler, R. Wieseh¨ofer, Advancements in CFRP Microwave Processing with the HEPHAISTOS-System, SAMPE EUROPE International Conference 2007, Paris, April 3rd–5th, 2007 L. Feher, New Design and Manufacturing Perspectives for Composite Processing with the HEPHAISTOS Microwave Technology, IIR Deutschland Conference “Composites and Lightweight Structures in Aircraft”, Hamburg, July, 4th 2007
Index
A A380, composite parts, 59 Absorption mechanism, 17 Aircraft manufacturers, 39 Anisotropic electromagnetic effects for CFRP heating, 80–82 Anti-/de-icing methods, 36 advanced requirements, 36 conventional hot air de-icing methods, 37 experimental 30 GHz de-icing experiment set-up, 40 high energy consumption, conventional, 43 microwave alternative – CAPRI investigations, 38 temperature development of such, 41 Anti-icing system design, 49–50 advantages of, 49 frequency choice, 50 microwave source, 50–52 avionic magnetron design, 51 typical parameters of different magnetrons, 51 primary waveguide system components, 52 components and sizes, models, 52 principle for composite structures, 50 secondary waveguide system components, 53–54 shape and size of mode converter, 53 shape and size of taper, 54 Autoclaves, 60, 64, 99 Avionic microwave anti-/de-icing systems basic considerations of icing in aviation, 35–37 continuous anti-icing system design, 49–50 frequency choice, 50 microwave source, 50–52 primary waveguide system components, 52
secondary waveguide system components, 53–54 de-icing leading edge system, 57–58 electrothermal CFRP-airfoil in-flight model, 44–45 following CAPRI-approach, 40–41 heating of CFRP composites, 41–43 microwave alternative – CAPRI investigations, 38 millimetre-wave investigations at FZK, 39 mm-wave experiments for technological design preparations, 43–44 power requirements and heating performance anti-icing in-flight heating performance, 55–56 common aluminum approach, 54 power reduction with CFRPcomposites, 55 theoretical basics of microwave in-flight composite heating, 45–49 C CAPRI-approach, 40–41 electromagnetic material properties of GFRP and CFRP, 40 heating of CFRP composites, 41–43 setup for, 42 temperature development, 42 CAPRI leakage wave applicator, 38 CAPRI-project, 35 Carbon fibre-reinforced composites (CFC), 2 curing process with microwaves, 78–79 laminate, 81 Carbon fibre reinforced plastics (CFRP), 62–63 anisotropic properties of, 34 electrical field penetration in, 47 significant temperature reduction in microwave heated, 48
111
112 thermal heat flux balance, 55 See also CAPRI-approach Carbon Fibre Reinforced Plastics/Glas Fibre Reinforced Plastics, see CFRP/GFRP CFC, see Carbon fibre-reinforced composites (CFC) CFRP/GFRP, 39, 40, 43, 45 Classic model, 13 Composite fabrication, process flow for, 61 Composite materials, processing technology for, 59–62 anisotropic electromagnetic effects for CFRP heating, 80–82 CFC curing process with microwaves, 78–79 electrothermal considerations, 73–76 frequency choice, 76–77 fundamental investigations homogeneous field applicator development, 69–73 HEPHAISTOS-systemline and technology, 62–66 microwave process development, 66–69 microwave processed CFC materials, material investigations for, 90–97 modular 2.45 GHz HEPHAISTOS conception, 82–86 process development, 86–89 thermo-electric foils, 89–90 Composites, anisotropic orientations for, 81 Coulomb gauge condition, 10 D Debye dissipation model of materials, 12–13 Debye theory, 12, 13 De-icing, 36 leading edge system, 57–58 DELFI-Code, 5, 70, 71, 76, 82, 83, 102 Diagonal thermal conductivity, 45 Dielectric measurements, 33–34 measured material properties of different samples, 34 setup for measurement of dielectric properties for liquids, 33 Dielectric measurements and material data, 31–33 effective complex permittivity, 32 Dielectric parameters, 31 Dipole moments, 15, 16 water, alcohols, phenol and ether, 19 Dipole structure of water, 12 Distribution of ELectromagnetic FIelds code, see DELFI-Code
Index DMA, see Dynamic mechanical analysis (DMA) Dominant mode, 24 Dynamic mechanical analysis (DMA), 85–86 E Efficient microwave transmission devices and measurements dielectric measurements, 33–34 and material data, 31–33 standard rectangular waveguides, 23–25 waveguide antenna design, 25–30 experimental verification, 30 EIO, see Extended Interaction Oscillator (EIO) Electric and magnetic slot coupling, 27 Electric conductivity, 21 tensor notation for, 45 Electric permittivity, 10 Electric power absorption, dielectric sample, 12 Electromagnetic energy, 12 Electromagnetic field-material interactions, 10 Electromagnetic radiation, 9 Electromagnetic vector potential, 11 Electronegativity, 12 Electrothermal CFRP-airfoil in-flight model, 44–45 Epoxy resin, 19, 88 Extended Interaction Oscillator (EIO), 7 and magnetron, 7 F Frequency choice, 76–77 G Glass-fibre reinforced composite (GFRP), 34 Gyrotron, 7 H Helmholtz-equation, 23 Helmholtz-Equation, homogeneous, 11, 23 HEPHAISTOS-CA system, 91, 92, 96–97 HEPHAISTOS-systemline and technology, 2, 62–67, 71–73, 83–84, 86, 90–92, 99 High electromagnetic power heating automated injected structures oven system (HEPHAISTOS), see HEPHAISTOS-systemline and technology Homogeneous field applicator development, 69–73 I Inclined RE-slots, 30 Industrial microwave sources at ISM frequencies, 5–6
Index possible 24.15 GHz sources and their properties, 6–8 In-flight ice formation, 35 catch probability depends on droplet size, 36 effects, 35–36 methods against icing, 35 microwave alternative – CAPRI investigations, 38 Injection technology, 62 J Joint Aviation Requirements (JAR), 35 K Klystron, 6 L Liquids, material parameters, 31 Loss tangent, 11 M Macromolecular plastics, synthesis of technical, 17–19 enthalpy diagram of curing reaction, 18 Magnetic flux density, 10 Magnetic permittivity, 10 Magnetron, 6 Materials under test (MUT), 31 between two dielectric materials in waveguide section, 32 Maximum take off weight (MTOW), 59–60 Maxwell’s equations, 75 Microwave anti-/de-icing methods as alternative, 38 system set-up of 2.45 GHz, 38 vs. conventional heating, physical advantages of, 5 electric conductivity, 20–21 using frequency higher than standard 2.45 GHz , advantages, 6 heating, advantages of, 66 injector, 68–69 process development, 66–69 processed CFC materials, material investigations for, 90–97 theoretical basics of, in-flight composite heating, 45–49 Microwave heating – dielectric properties and energy conversion classical Debye dissipation model of materials, 12–13
113 detailed consideration on microwave heating of water ionic contribution, 13–14 non classical consideration on microwave heating of water, 14–17 electromagnetic wave propagation and material interaction, 10–12 heating – general electromagnetic problem, 9–10 quantum representation of the microwave effective electric conductivity, 19–21 synthesis of technical macromolecular plastics, 17–19 MiRa-Code, 5, 69, 82, 102 Mm-wave system components, 5 costs/R&D-efforts, 6 system costs, compared to microwave, 5 See also Gyrotron Modular 2.45 GHz HEPHAISTOS conception, 82–86 Monomer of elastomeres, 90 Monomode applicators, 71 O Oxiran dipole and water, 19 P Pneumatic boots, 37 Poisson’s equation, 45, 74 Polygonal cross sections, comparison of, 70 Polygonal orientation, 69–70 Polytetrafluorethylene (PTFE, Teflon), 19, 20 Power density, 45 Power requirements and heating performance anti-icing in-flight heating performance, 55–56 heating performance, 56 common aluminum approach, 54 power reduction with CFRP-composites, 55 Poynting vector, 11 Prepreg-materials, 62, 86, 91, 97 R Radiating elements (RE), 25, 26 Resin infiltration technology, process cycle for, 62 Resin´s polymerization, 99 RE-slots, 26, 29 inclined, 30 shape in waveguide wall, 25 Rotational moment, 16 RTM6, 86, 99 resin system, 92
114 S Solids, material parameters, 31 T TE and TM-waves, generic relations for, 24 TE10 -mode, 24 field components for, 24 parameters in WR-340 waveguide, 25 rectangular, to transmit microwave power from source to power dissipating load, 25 Thermo-electric foils, 89–90 Thermoplastics, 18 THESIS-Code, 5, 76, 78, 79, 82, 83, 102 THESIS3D, 76, 78, 79, 82 U Upilex, 86–87 V Vacuum assisted process (VAP) infiltration process, 68, 89 24 GHz vacuum electronic devices, design overview of, 7 Vector network analyzer (VNA), 30
Index W Water dielectric properties of, 14 dipole structure of, 12 front view of water molecule, 16 oxiran dipole and, 19 rotation of single vaporized water molecule, 17 Waveguide antenna design, 25–30 effective slot coupling and s-matrix coefficients, 27 experimental verification, 30 feeding waveguide, 25 RE-slots designing requirements, 25 vector field visualization for first rectangular waveguide modes, 26 wall current densities, 26 walls, 26 current densities, inner narrow/broad, 27 gain to commonly optimized, 30 optimized microwave reflection of, 31 properties of, 27–28 real waveguides, 28 Wet technology, 62