1. The effect of compressive pre-stress on the thermal expansion behaviour of anisotropic nuclear grade graphite M. Haverty, W. Bodel, B.J. Marsden Nuclear Graphite Research Group The University of Manchester maureen.haverty@postgrad.manchester.ac.uk

2. Contents 1.General Methodology used 2.Preliminary Study 3.Main Study 4.Comparison between two studies 5.High Resolution images

3. Motivation Previous studies have shown a change in CTE with stress – Gilsocarbon: Applied uniaxial compressive and tensile stress 1 – Steels: Applied uniaxial tensile stress and pre- stress beyond elastic limit 2 Are these changes observed in PGA? What causes these changes? 1.Preston, S.D. & Marsden, B.J., 2006. Changes in the coefficient of thermal expansion in stressed Gilsocarbon graphite. Carbon, 44(7), pp.1250–1257 2.Rosenfield, A.R. & Averbach, B.L., 1956. Effect of Stress on the Expansion Coefficient. Journal of Applied Physics, 27(2), pp.154–156. 3..

4. CTE MEASUREMENT

5. Standards and Methodologies Used CN821-1 “Advanced technical ceramics- Monolithic ceramics-Thermo-physical properties-Part 1: Determination of thermal expansion” – Provides test method including, sample size; reference standards and heating rates Accuracy if followed: – 5K/min  0.5 x 10 -6 K -1 – 2K/min  0.1 x 10 -6 K -1

6. CTE Measurement equipment Netzsch Proteus DIL – Pushrod dilatometer – Nitrogen atmosphere Al 2 O 3 Reference standard – Glow runs – Filler piece for short samples (Al 2 O 3 )

7. Thermal Expansion Measurements Al 2 O 3 Sample Holder Al 2 O 3 Pushrod Al 2 O 3 Filler Piece Graphite sample

8. PRELIMINARY TEST

9. Method Samples were cut in the AG and WG directions Uniaxial compressive stress applied – Compressive strength: 27 MPa Max temp: 250 °C Sample No.Stress (%) AG1/WG10 AG2/WG220 AG3/WG340 AG4/WG460 AG5/WG580 AG6/WG690

10. Applying Stress Universal Load Testing Machine Compressive stress Sample subdivided into two sister samples Excess used to face off to correct tolerance Φ =12 mm H =18 mm Φ =12 mm H =6 mm

11. Comparison of unstressed samples with the literature Comparing unstressed sample values with Sutton and Howard 3 Two blocks of PGA, measured in WG and AG directions Average CTE for 50°C increments, e.g. 100-150°C CTE plotted at midpoint of increment e.g. 125 °C 3. Sutton, A.L. & Howard, V.C., 1962. The role of porosity in the accommodation of thermal expansion in graphite. Journal of Nuclear Materials, 7(1), pp.58–71.

13. CTE Preliminary Results Average CTE plotted Reference temperature of 50 °C used, e.g. 50-250 °C CTE plotted at midpoint of temperature increment

26. MAIN STUDY

27. Samples Φ =6 mm H =6 mm CTE measured on 10 samples to ascertain sample variability Two samples in each direction selected randomly for pre- stressing

28. Method CTE measured Pre-stress applied (fraction of compressive strength) Properties re-measured Max temp: 450 °C AG: 0%; 20%; 40%; 90% WG: 0%; 20%; 40%

29. Comparison of unstressed samples with the literature Comparing unstressed sample values with Sutton and Howard 3 Two blocks of PGA, measured in WG and AG directions Average CTE for 50°C increments, e.g. 100-150°C CTE plotted at midpoint of increment e.g. 125 °C 3. Sutton, A.L. & Howard, V.C., 1962. The role of porosity in the accommodation of thermal expansion in graphite. Journal of Nuclear Materials, 7(1), pp.58–71.

32. Sample Variability Temp Range (°C)AG CTE (x 10 -6 K -1 ) St. Dev (x 10 -6 K -1 ) WG CTE (x 10 -6 K -1 ) St. Dev (x 10 -6 K -1 ) 50-100 0.9 ± 0.10.1 50-150 3.4 ± 0.10.11.1± 0.10.2 50-200 3.5 ± 0.10.11.2 ± 0.10.1 50-250 3.6 ± 0.10.11.3 ± 0.10.1 50-300 3.7 ± 0.10.11.5 ± 0.10.1 50-350 3.8 ± 0.10.11.6 ± 0.10.1 50-400 3.9 ± 0.10.11.7 ± 0.10.1 50-450 4.0 ± 0.10.11.9 ± 0.10.1

50. COMPARISON BETWEEN PRELIMINARY AND MAIN STUDY

56. Observations Both preliminary and main studies show good agreement in CTE measurements Pre-stress causes increase in CTE – Approximately 20% increase in CTE at 90% Pre-stress in AG direction – Approximately 10% increase in CTE at 40% Pre-stress in AG direction Minimum pre-stress required to cause change has yet to be determined Residual strains not directly proportional to pre- stress

57. What causes change in CTE? Crystal re-orientation may be occurring Other authors have attributed the increase in CTE in steels to lattice distortion 2 Elastic limit must be reached in steels before permanent changes to CTE occur – Residual strains in graphite – Is there a minimum residual strain for permanent changes in CTE to occur?

67. Acknowledgements This project is funded by EPSRC through the FunGraphite consortium The author would like to thank the following for their advice and help with the project: Dr. Marc Schmidt; Gary Harrison; the staff at FEI; David Mortimer; Judith Shackleton The SEM images were taken at the Dalton Cumbrian Facility and the author would like to thank DCF for access to the equipment