Schematic of Initial Solidification

Transcription

1 ANNUAL REPORT 2008 Meeting date: August 6, 2008 Design & Installation of Novel Sensors into the Continuous Casting Mold Michael K. Okelman & Brian G. Thomas Department of Mechanical Science and Engineering University of Illinois at Urbana-Champaign Schematic of Initial Solidification What s the temperature, heat flux, stress behavior in this region? Liquid flux film (low friction) Meniscus Solidified Flux Flux Rim Flux Powder Liquid Flux Steel surface level Oscillation Mark Solidifying Steel Shell Solid flux film (high friction) Copper Mold (images courtesy Brian Thomas, personal communication) Molten Steel Pool Contact Resistances Air Gap University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 2

2 Installation of Sensor Strip into Continuous Casting Mold k=109 W/m- C (temperature in Celsius) adhesive & coating applications to +600 C University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 3 Heat flux can be measured by a thermopile with two layers of ceramic TFTC Designed for Use in a Continuous Casting Mold q'' = k T x 1 1 T 2 x 2 University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 4

3 Air Gap Effect on Heat Transfer Air gaps increase hot face temperature by up to 20%, depending on geometry of gap University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 5 More Specifically temperature (K) temperature (K) Original Geometry (2mm x 0.1mm) distance from hotface (mm) gap centerline gap edge far from gap max temp at hot face 314 C 700 Thicker Gap (2mm x 0.2mm) gap centerline 680 gap edge 660 far from gap max temp at hot face 319 C distance from hotface (mm) 700 gap centerline 680 gap edge 660 far from gap distance from hotface (mm) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 6 temperature (K) Wider Gap (4mm x 0.1mm) max temp at hot face 379 C Doubling the thickness of the gap increases the hot face temperature by 5 C, doubling the width of the gap increases the hot face temperature by 65 C A wider gap makes it more difficult for heat to conduct around the gap, increasing the temperature at the hot face For a 2mm wide sensor even a 0.1mm gap produces a hot face temperature variation of only 28 C

4 Air Gap Effect on Stress The copper goes into compression (σ max = 67.8 MPa) while the nickel coating layer goes into tension (σ max = 276 MPa) The nickel coating layer might yield, debond, or spall off in patches An air gap 2 mm wide and 0.1 mm thick causes an ~10-50% increase in the stress in the nickel coating layer, compared to a nickel coating layer without an air gap University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 7 Plating FBG Sensors FBG sensors can be plated near the mold face, contacting the mold face, or in clusters to achieve high spatial resolution University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 8

5 FBG Sensor Performance Embedded FBG (Experimental) Free-floating FBG (Experimental) Embedded FBG (Theoretical - CTE) Embedded FBG (Theoretical - Beam) λ = T T = λ + m ( λ 0 ) T0 λ, wavelength (nm) Free-floating FBG (Theoretical) (T 0, λ 0) embedded λ = T λ = T where m=sensitivity/slope (numbers on plot are theoretically calculated), λ=bragg wavelength, and λ 0 and T 0 are the Bragg wavelength and temperature measured at calibration (T 0, λ 0) free-floating T, temperature (Celsius) It is possible to predict the temperature measured by the embedded FBG sensor, but the best temperature reading can be obtained by inserting the FBG sensor into a hollow stainless steel tube (to decouple temperature and strain effects) University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 9 Theoretical Sensitivity/Slope, m Given as change in wavelength divided by the change in temperature (Δλ/ΔT) Depends on how fiber is used: Free-floating (open air) depends only on temperature effects easily calculated given optical properties of FBG Embedded depends on temperature and strain effects can predict mechanical strain using "CTE" or (bimetallic) "beam" method University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 10

6 Residual Stress in Coating Layer Compressive residual stress in the nickel coating layer due to electroplating process produces mechanical strain on FBG sensor This stress can be determined using data from calibrated FBG ( T ' ) σ = Eα s T 0 ~100 MPa (compression) Variable E α s T ' T 0 Description elastic modulus of coating layer CTE of coating layer "reference" temperature at center wavelength of free-floating FBG "reference" temperature at center wavelength of embedded FBG Value 207 GPa 13.1 x 10-6 / C 48 C 10 C University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 13 Conclusions Conventional thermocouples cannot accurately quantify temperature (& more complex behavior) at meniscus TFTC can be silver pasted on CC mold (and nickel plated), or FBG sensor can be inserted via a plated SS tube (preferable) The signal output by FBG sensors embedded in a nickel coating layer has been investigated and can be predicted with simple equations University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 14

7 Acknowledgements Continuous Casting Consortium Members National Science Foundation DMI (Sensor) Prof. Li at University of Wisconsin - Madison M. Powers & Sumitec Siemens VAI Services C. Gulyash, K. Parrish, & MEL Machine Shop J. Soares & Center for Microanalysis of Materials at Frederick Seitz Materials Research Laboratory C. Graham, S. Fiegle, & ArcelorMittal - Riverdale Other Graduate students University of Illinois at Urbana-Champaign Metals Processing Simulation Lab Michael K. Okelman 15