A large area VHF plasma source for atmospheric air plasma treatment of coated surfaces

Transcription

1 A large area VHF plasma source for atmospheric air plasma treatment of coated surfaces Brandon Byrns, Daniel Wooten, and Steve Shannon North Carolina State University Department of Nuclear Engineering IEEE International Conference on Plasma Science 29 June 2011, Chicago IL 1

2 Introduction Plasmas for coating removal Current methods including grit blast and water jet have problems Large amounts of hazardous waste Flash rusting, substrate damage Hazardous work enviroment Plasma assisted removal of coatings is one of several technologies being evaluated as an environmentally friendly replacement technology One challenge has been design of a high power density, large area source for efficient removal of large surfaces (ships, planes, etc.) ghts.com/product/deadman/100 2.htm 2

3 VHF Coaxial Air Plasma Source Plasma generated using a 162 MHz VHF Coax Source generator Transmission line match used both as source assembly structure and impedance match Up to 600W power delivery forms a 2 inch diameter discharge with no Plasma Torch observable electrode erosion or damage Overview of talk preliminary build design and characterization Source design Electrical Characterization Optical Emission Spectroscopy Pictures roughly scale in size for comparison 3

4 Circuit Model Two coaxial transmission 162MHz 3.5kW lines in parallel One terminated by a short Other terminated with a plasma impedance Plasma impedance represented by capacitance in parallel with a resistance and inductance Ground Termination AE Ovation OD ~ 3cm Length ~ 5cm 50 Coaxial feed to source Inner Conductor OD ~2cm ~ 4cm ~30cm Z 0 Z 0 Z p C p Variable parameters: l p, l g OD, ID, ODE, n e Where Z p = R p +i R p/ me 4

5 Circuit Model: Adding the plasma Relation between plasma density, collision i frequency, and expected plasma impedance Zp: insert equation here Terminating match with this load as a function of number density gives a first order estimate of match size as well as an estimate of trajectory as we move to higher h power densities. Initial target density for coax design ~10 11 cm -3 Increasing electron density Total device impedance for a range of electron density for a given set of dimensions 5

6 Experiment Source/match characterization: Took data for short, middle, and long distance for series and shunt (9 settings total) No plasma initially; vacuum impedance was measured to characterize matching network component Used Helium to ignite plasma due to low breakdown voltage: 2-6kV for He compared to 30-60kV for air Results in drop of power needed for breakdown by a factor of 100 6

7 Electrical Measurements Electrical measurements performed on device without plasma Dt Determine the resonant frequency based device dimensions Want resonance with Q~10 at 162MHz db f res = 162 MHz FWHM Frequency (MHz) 7

8 Video Initial helium strike at 300W wall interaction observed at ignition, then discharge reverts to center glow Transition to air also has a brief interaction with the wall, followed by stable center glow at 300W Power then increased to 500Wto produce final air plasma glow 8

9 Reflected Power data Change in impedance with changing electron density shown in model and experiment Impedance trajectories differ between model and experiment Possible explanations: Experimental data obtained from generator phase data, which can have significant errors in phase angle Model is not complete and lacks the ability to accurately describe the plasma 595W 490W 420W 350W 280W 9

10 Impedance Matching From impedance data taken from generator the plasma impedance can be calculated Assume the transmission lines are terminated to the plasma impedance These can then be compared to plasma model to determine plasma a parameters a Power Plasma Impedance Electron Density (cm -3 ) 280 ( i)Ω 5.5 x ( i)Ω3i)Ω x ( i)Ω 1.35 x ( i)Ω 1.5 x ( i)Ω4i)Ω 14x

11 Non-Normalized Normalized Spectroscopy Intensity W 420 W 350 W Wavelength (nm) 11

12 0.01 Area-Normalized Spectroscopy Similar Ch i Chemistry over power density d i N2 (C B) and N2+ (B X) OH (A X) N2+ (B X) N2 (C B) No ormalized Counts NO γ (A X) O 595 W W 350 W Wavelength (nm) 12

13 25000 No free lunch! Electrode Damage at higher powers (>600W) Al AlO Wavelength (nm) At high powers plasma sparks blue, resulting in damage to electrode Once significant ifi tdamage occurs, plasma only burns blue Recorded spectrum matches the spectrum of Aluminum in air Plan to evaluate new materials to extend power capability 13

14 Future Work Repeat spectroscopy data with 0.5m spectrometer isolate lines of interest, measure rotational temperature, re etc. Determine paint removal efficiency Model plasma to use impedance measurements to determine plasma parameters Laser absorption spectroscopy Electrode materials for higher power densities Scale up both in power density and plasma volume current power capability should be able to support ~5cm diameter discharge with comparable power density 14

15 Acknowledgements This project is supported by the State of North Carolina and University of North Carolina General Assembly as well as generous gift donations from both Applied Materials Inc. and Advanced Energy Inc. The authors wish to thank Mohamed Bourham, Jerry Cuomo, o, Richard Guanari, Stephen Hudak, Anthony McWilliams, Anatoli Melechko, and the staff at the NCSU Precision Machine Shop for valuable input and support during this project. 15