Neutron Tomography Measurement of Delayed Ettringite Formation in Concrete Richard A. Livingston Materials Science & Engineering Dept University of Maryland 14th ISNDCM Marina del Rey, CA, June 24, 2015
Co-Authors Amde M. Amde & Serge Feuze Civil Engineering Dept., U. of Maryland Daniel Hussey & David Jacobson Physical Measurement Laboratory, NIST Acknowledgements John Newman, Laser Technologies, Inc Stewart Sherman, National Ready Mix Concrete Association
Outline Concrete deterioration mechanisms Neutron tomography DEF case study Results Calibration methods Conclusions
Major Deterioration Processes in Concrete Mechanism DelayedEttringite Formation Abbreviation Expansive Phase Alkali Silica Reaction ASR ASR Gel Formula DEF Ettringite (CaO) 3 Al 2 O 3 (CaSO 4 ) 3 (H 2 O) 32 Freeze-thaw cycles F-T Ice H 2 O Na 2 O xsio 2 yh 2 O x = 4-22 y = 5-10 Rebar corrosion - Rust Fe 2 O 3 nh 2 O,FeO(OH)orFe(OH) 3
Single Spherical Aggregate Model E. Garboczi, CCR, 1997
Signatures of Expansion Types Based on the Garboczi Model Expansion Type Mechanism Crack Type Gap Thickness Uniform matrix DEF? F-T? Circumferential Aggregate radius Rim only DEF? Circumferential Aggregate radius Aggregate only ASR, F-T? Radial None
Conventional Analytical Methods X-ray diffraction Thermal analysis Scanning electron microscopy Fracture surface Polished section
Neutron vs X-ray Attenuation 8
Comparison of X-ray and Neutron Radiographs X-ray Neutrons 9
Phase Neutron Attenuation Coefficients of Common Phases in Concrete Formula Molecular weight g/mol H fraction Density g/cm 3 H density Atoms/cm 3 Attenuation coefficient* cm -1 Quartz SiO 2 60.09 0 2.65 0 0.284 Limestone CaCO3 100.09 0 2.72 0 0.258 CSH gel a (CaO) 1.7 (SiO 2 )(H 2 O) 1.8 187.83 0.019 2.61 0.0499 2.731 Calcium hydroxide Ca(OH) 2 74.08 0.027 2.23 0.0602 3.193 Ettringite (CaO) 3 Al 2 O 3 (CaSO 4 ) 3 (H 2 O) 32 1254.62 0.051 1.77 0.0903 4.675 ASR gel b Na 2 O xsio 2 y H 2 O 214.08-0.012-1.93 0.0314-1.833 - x = 4-22 y = 5-10 1563.78 0.032 2.46 0.0631 3.355 Water ice H 2 O 18.00 0.111 0.9 0.1000 5.081 a Allenet al. 2007 b Broeckmann, 2012 *Attenuation for bound H at 0.18 nm wavelength 10
NIST Neutron Imaging Facility 11
Neutron Camera Converter screen Mirror hν Target n Rotating stage CCD Camera
Steam Curing and DEF Primary ettringite forms during early hydration Normal concrete curing temperatures 30-40 C Steam curing at pre-cast plant 80 90 C Hypothesis Ettringite decomposes ~ 70 C In the field ambient moisture causes ekringite to reform DEF
Sample Preparation Two batches of concrete from same mix Control Potassium added, 1.2% as K 2 CO 3 Cast as prisms 3 x 3 x 11 Two curing conditions Room temperature Steam cured J. Newman, FHWA SBIR, 2011
Concrete Test Prism Measurement point
Simulated Steam Curing
Initial Thermal Cycling
Water Storage
Expansion of Concrete Prisms High Potassium Control 19
Drilling of 2 inch Cores 20
Raw Neutron Image Scan Neutron Beam L/D = 450 Fluence = 1.3 x 10 7 cm 2 /s 7 cm Image Capture Pixel Pitch = 25 μm Rotation step = 0.1 Range = 180 Image scan time ~15 sec. Replicate scans = 3 Total acquisition time = 26 hrs 6 cm
Tomographic Image of Core Bright rims around aggregates 22
Tomographic Image of Core Bright rims around aggregates O Ca AlS 23
Segmentation of Tomographic Slice Grayscale 2-D slice through tomographic volume Histogram of grayscale values segmented by concrete phase. 24
False Color Image Areal Fractions of Concrete Phases % Porosity 14.0 Aggregates 70.0 Paste 22.0 Ettringite 3.0 Sum 100 25
Internal H Standard Cement Phase Attenuation Coefficient cm -1 Plastic Attenuation Coefficient cm -1 CSH Gel 2.73 Polycarbonate 3.31 Calcium hydroxide 3.19 Polystyrene 3.96 Ettringite 4.67 Polyethylene 6.86
Major Deterioration Processes in Concrete Mechanism DelayedEttringite Formation Abbreviation Expansive Phase Alkali Silica Reaction ASR ASR Gel Formula DEF Ettringite (CaO) 3 Al 2 O 3 (CaSO 4 ) 3 (H 2 O) 32 Freeze-thaw cycles F-T Ice H 2 O Na 2 O xsio 2 yh 2 O x = 4-22 y = 5-10 Rebar corrosion - Rust Fe 2 O 3 nh 2 O,FeO(OH)orFe(OH) 3
Conclusions Neutron imaging can identify concrete hydrous phases Neutron tomography can capture 3-D spatial relationships among phases at 20 μm resolution Exhaustive volumetric sampling enables highly precise materials characterization Nondestructive nature makes it possible to observe reactions among phases over time Limitations include scarce beam time at neutron facilities
Thank you for your attention! Richard A. Livingston: rliving1@umd.edu