International Summit on Cement Hydration Kinetics and Modeling * a Summary and outcomes

INTERNATIONAL SUMMIT ON CEMENT HYDRATION KINETICS AND MODELING*

A SUMMARY AND OUTCOMES

Dr. Joseph J. Biernacki, PETennessee Technological University

Department of Chemical Engineering

August 19, 2010Federal Highway Administration

AcknowledgementsThe 51 Participants of the International Summit on Cement Hydration Kinetics and

ModelingNational Science Foundation (NIST)

Federal Highway Administration (FHWA)W. R. Grace

BASFMapei

Canadian Research Center on Concrete InfrastructureNatural Science and Engineering Research Council of Canada

Center for Manufacturing Research, Tennessee Technological University

* The International Summit on Cement Hydration Kinetics and Modeling was held on July 27, 28 and 29, 2009 in at Laval University, Quebec, Quebec City, Canada.

Special AcknowledgementsT. Xie, The Origins and evolution of cement hydration models, Comp. Concr., submitted (2010).

J. J. Thomas, Digital and Mathematical Modeling of the Nucleation and Growth of Cement Based Materials , International Summit on Cement Hydration Kinetics and Modeling (2009).

S. Bishnoi, A. Kumar and K. L. Scrivener, Nucleation and Growth Kinetics and the Density of C-S-h, International Summit on Cement Hydration Kinetics and Modeling (2009).

J. Bullard, How can C-S-H growth behavior be predicted? Questions from a modeling perspective , International Summit on Cement Hydration Kinetics and Modeling (2009).

A. Luttge, Vertical scanning interferometry… , Kinetics Summit (2010).

J. Chen, et al., “A coupled grid-indentation/SEM-EDX study on low w/c PC pastes… ,” Kinetics Summit (2010).

A. A. Jeknavorian, Impact of Water Reducers and Superplasticizers on the Hydration of Portland Cement , International Summit on Cement Hydration Kinetics and Modeling (2009).

D. Silva, Impact of Accelerators and Retarders on the Hydration of Portland Cement , International Summit on Cement Hydration Kinetics and Modeling (2009).

L. Roberts, Admixtures of C-Ash – A Challenge to Model, International Summit on Cement Hydration Kinetics and Modeling (2009).

V. Kocaba and K. L. Scrivener, Effect of SCMs on Hydration Kinetics of Cementitious Systems, International Summit on Cement Hydration Kinetics and Modeling (2009).

Outline Background Demographics of Participants Introduction Topical Report (based on 35 Summit presentations*)

Mechanisms (7) Modeling (5) Experimental Methods (9) Admixtures (5) Supplemental Cementitious Systems (3) Alternative Cements (4) Thermodynamics (2)

Roadmap

*see for full presentation slides http://blogs.cae.tntech.edu/hydration-kinetics/

http://blogs.cae.tntech.edu/hydration-kinetics/

BackgroundSummer 2007 – Biernacki, Hansen (UM) and Bullard (NIST) meet as part of ongoing

NSF grant activities; we discussed and agreed to try to organize a workshop of some form on hydration kinetics

Fall of 2007 – The workshop concept took on the form of a US-Canadian joint effort with invitation of researchers from the European Community; Jacques Marchand (Laval University) agreed to host the event

September 2007 – Biernacki and Hansen submit proposal to NSF for US-Canadian workshop entitled, “International Summit on Cement Hydration Kinetics and Modeling”

June 2008 – NSF grant award received for ~$25k

Fall 2008 and Winter of 2009 – Biernacki, Constantiner (BASF) and Bullard raised an additional ~$29k including support from the FHWA and a number of industrial sponsors

July 2009 – Workshop held at Laval University, Quebec, Canada

Participant DemographicsCountry Faculty Industry/Lab StudentCanada 2 5 4United States 13 10 7Europe 3 5 2Asia/Other 1Totals 18 21 13

52 Participants6 Countries

Program Early Age Hydration (Stages I, II and III)

Mechanisms Modeling Instrumentation and Experimentation

Effect of Supplemental Cementitious Materials

Post Peak Hydration (Stage IV) Thermochemicstry and Geochemical

Viewpoint Alternative Cement Systems

Introduction Why is hydration important? What is this presentation about?

Why is hydration important?meters

centimeters

millimeters

micrometers

nanometers

http://books.google.com/books?id=o4J2FX43s-8C&pg=PA27&lpg=PA27&dq=clinker+petrography&source=bl&ots=ZX_YYKLRw9&sig=XInuOa0eq992BgAXR5shszrcfO4&hl=en&ei=YBydSvyOL-KMtgebioTDBA&sa=X&oi=book_result&ct=result&resnum=4#v=onepage&q=clinker%20petrography&f=false

Hydration and Common Field Problems

Shrinkage and Related Cracking – the result of volume change autogenous drying due to hydration

Corrosion – the consequence of undesirable transport to and from the environment due to inadequately developed and/or controlled microstructure

Alkali Silica Reaction – a secondary hydration process

Adverse Admixture Interactions – unexpected or unknown effects on hydration

Curing – adequate hydration

Kinetics and Materials Metals – understanding the kinetics of ore processing,

solidification/crystallization and solid-state phase transformation has given us super-alloys, stainless steel, lightweight alloys, corrosion resistant metals and high temperature refractory metallurgy

Polymers – by controlling the chemical reaction kinetics of organic synthesis it is now possible to reliably produce mega-tons of polyethylene, produce designer co-polymers, liquid crystals and strong, lightweight matrix materials for composites

Semi-conductors – without detailed knowledge of the kinetics of doping modern computes would be impossible

Petroleum Refining – kinetic models make it possible for refineries to be agile machines that can move between product distributions in response to rapidly changing market demands and crude prices and availability

Typical Calorimetric Response

Stage I – DissolutionStage II – DormancyStage III –AccelerationStage IV - Deceleration

What is this presentation about?

1. This presentation will summarize some of the most recent findings related to cement hydration, both experimental and computational.

2. The “roadmap” presented here is not a proposal, but rather a summary of research needs, it is not the answer to many question, but instead identifies what the questions are.

3. Finally, the roadmap suggests directions in which to move, not the details of every turn.

Mechanisms





1. Formation of thick barrier layer (prevalent hypothesis ~1980)2. Formation of hydroxylated surface (suggested hypothesis ~2001)3. Etch pit to stop flow dissolution transition (suggested hypothesis ~ 200x)




1. Thermodynamic destabilization of thick barrier layer (suggested hypothesis ~2008)2. Mechanical destabilization of thick barrier layer (suggested hypothesis ~20xx)3. Nucleation driven (suggested hypothesis ~2009)4. Surface area driven (suggested hypothesis 2010)



1. Onset of diffusion control (prevalent hypothesis ~1980)2. Space filling (prevalent hypothesis ~2007)





1. Onset of diffusion control (prevalent hypothesis ~1980)2. Space filling (prevalent hypothesis ~2007)

1. Onset of diffusion control (prevalent hypothesis 200x)2. Onset of topochemical reaction/diffusion control (2010)3. Slow secondary densification (suggested hypothesis 2010)



Barrier Layer Hypothesis1. Rapid dissolution 2. Formation of meta-stable layer 3. Nucleation of stable C-S-H and destabilization of meta-stable layer 4. Growth

Modeling

Origins and Evolution of Hydration Models

Mass Continuity-based Models Nucleation Models

Microstructure Simulation Tools

Jander (1927)

Brown (1985)

Ginstling (1950)Taplin (1968)

Pommersheim (1979)

Avrami (1939)

Bentz 2004)(volume filling)

Cahn (1956)

Thomas (2007)

Jennings (1986)

Bentz (1999) van Breugel (1995)

Pignat (1999)

Bishnoi (2009)

Bullard (2008)

Knudsen (1984)

Particle Size Distribution Models

Taplin (1972)

Brown (1989)

Pommersheim (1987)

2T. Xie and J. J. Biernacki, The Origins and evolution of cement hydration models, Comp. Concr., submitted (2010).

Models…


And More Models…


And…


Ginstling, Brown and Jander

0.00

2.00

4.00

6.00

8.00

10.00

12.00

14.00

0.00 5.00 10.00 15.00 20.00 25.00

Heat

Flow

Rat

e (m

W/g

)

Time (h)

Experimental

Brown

Ginstling-Brounshtein

Jander


Pommersheim and Clifton

Associated ~Citation Name

Diffusivity DH2SiO4 (×10-9 m2/s)

Zero-order Growth Rate k (×10-10 m/s)

Bullard

0.007CSH(I) 0.525 CSH(II)

0.007 meta-stable CSH 0.7 H

NA

Bishnoi and Scrivener NA ~3 NA=Not Applicable.

Associate Model Name Diffusivity DH2SiO4 (×10-9 m2/s)

Zero-order Growth Rate k (×10-10 m/s)

Jander 0.00000003 to 0.000003* NA

Ginstling and Brounshtein 0.00000003 to 0.000003* NA

Brown NA 0.1

Pommersheim$ 0.1 inner product

0.00002 to 0.002 outer product 0.000033 meta-stable layer

0.12 to 10 #

Knudsen 0.000000097^ 0.15&

NA=Not Applicable. * Diffusivities were computed assuming a range of concentrations between 2.5 and 200 M for H2SiO4

-2, see Bullard (in preparation). $ Diffusivities were computed assuming the pore solution has a concentration of 25 M for H2SiO4

-2, see Nonat (2001). # Pommersheim assumed a 1st order reaction model. ^Knudsen’s diffusion parameter is not to be confused with a diffusivity, e.g. Knudsen’s parameter is not a transport property, thought a parabolic -like rate law is assumed. & Knudsen assumes an integrated form of rate law that does not produce to a zero-order differential form when differentiated, e.g. Knudsen’s rate parameter is not a zero-order kinetic rate constant.


Knudsen


What really is the problem?

original core location

unreacted core3 Ca+2 + 4 OH- + H2SiO4-2

(CaO)3SiO23H

2O

p(b i-

a i) H

2O

(1-p)

(CaO) ao

SiO2(H

2O) bo

p (CaO)aiSiO2(H2O)bi

(1-p)(bo-ao)H2O

2-p(ai+ao) Ca(OH)2

3-pa

i Ca+

2 4+

2p(1

-ai)

OH-

(1-p

) H2S

iO4-2

C-S-

H pr

oduc

t lay

er

2J. J. Biernacki and T. Xie, An Advanced Single Particle Model for C3S and Alite Hydration, J. Am. Ceram. Soc., submitted (2010).

Modelingrecent advances

J. Thomas (2007), A new approach to modeling the nucleation and growth kinetics of tricalcium silicate hydration, J. Am. Ceram. Soc., 90(10), 3282-3288.

J. Bullard (2008), A determination of hydration mechanisms for tricalicum silicate using a kinetic cellular automaton model, J. Am. Ceram. Soc., 91(7), 2008-2097.

S. Bishnoi and K. Scrivener (2009), Studying nucleation and growth kinetics of alite hydration using ic, Cem. Concr. Res, 39, 849-860.

J. J. Biernacki and T. Xie (2010, submitted), An advanced single particle model for C3S and alite hydration, J. Am. Cer. Soc.

Nucleation Models Bulk Nucleation Conceptual Description

cement grain

cement grain

Avrami’s Equation

nuclei

nAtke1

Nucleation ModelsBoundary Nucleation Conceptual Description

cement grain

cement grainCahn’s Equation

nucleinuclei

J.J. Thomas, Digital and Mathematical Modeling of the Nucleation and Growth of Cement Based Materials, International Summit on Cement Hydration Kinetics and Modeling (2009).

J.J. Thomas, Digital and Mathematical Modeling of the Nucleation and Growth of Cement Based Materials, International Summit on Cement Hydration Kinetics and Modeling (2009).




Experimental data from: S. Garrault and A. Nonat, Langmuir, 17 8131-8138 (2001).

J. Bullard, How can C-S-H growth behavior be predicted? Questions from a modeling perspective, International Summit on Cement Hydration Kinetics and Modeling (2009).

Advanced Single Particle Model - TTU

Model with minimal optimization compared to experimental hydration calorimetry dataset.

J. J. Biernacki and T. Xie, An Advanced Single Particle Model for C3S and Alite Hydration, J. Am. Ceram. Soc., submitted (2010).

“Simple” Challengesthe effect of w/c ratio

J. J. Biernacki and D. Kirby, The Effect of Water to Cement Ratio on Early Age Hydration Behavior, unpublished (2010).

Alite

Type I Portland Cement

Mechanisms and Modeling Roadmap

Does a semipermeable layer actually form at early ages? An affirmative answer might lead to pathways for either preventing its formation or prolonging its existence. If there is such a layer, knowledge of the trigger for its disappearance (e.g. thermodynamic or mechanical instability) could lead to the design of admixtures for targeting that trigger.

What role is played by surface defects, such as stacking faults and dislocations, in governing the dissolution rates of clinker phases in cement? Answering this question could lead to approaches such as annealing or chemical pretreatments that could eliminate or deactivate such defects.

To what extent do species in solution adsorb on cement phases or hydration product phases and modify the dissolution or growth rates of those phases? There is persuasive evidence that adsorption of calcium sulfate onto active dissolution sites of aluminate phases is responsible for the set-controlling properties of gypsum in cement, but understanding in this area is still in its infancy.

What are the transport properties of the bulk C-S-H products formed and how do they evolve with time? Does C-S-H form by a two-stage growth process and what is the bulk density of C-S-H as a function of

time? When and where do C-S-H nuclei form and what is the formation rate? What factors are responsible for the strong interactions between silicates and aluminates in cement

clinker hydration? There is recent experimental evidence that incorporation of aluminate ions in C-S-H is highly dependent on aluminate concentration in solution but also poisons its growth rate.

What are the form of the rate laws, e.g. what is the reaction order, and what are the elementary reactions that control the reaction rates, not only at early age, but at any age?

What is the actual morphology of C-S-H growth and what controls the morphology since many forms have been observed?

Are there signatures in early-age calorimetry measurement that indicate long-term kinetics and performance?

Mechanisms and Modeling Roadmap cont’d

Continue to develop various solution-phase-driven models that incorporate kinetics, thermochemistry and transport phenomena.

Develop multiple modeling paths and strategies that corroborate findings and lead towards useful engineering tools as well as model-based research instruments including fast algorithms for PC and similar platforms.

Continue to extend and exploit computational resources as necessary and needed to accommodate changing needs, i.e. utilize massively parallel processing and super computer facilities as needed.

Consider alternative computational strategies to accelerate the development of rigorous models, i.e. fast single particle models, representative volume approaches, etc.

Exploit the body of knowledge on true multi-scale modeling. Improve the dissemination of modeling tools to promote their use and development. Incorporate more molecular-level modeling strategies, i.e. kinetic Monte Carlo, etc. Develop suitable structural analogs for various anhydrous and hydrated cement phases for

use in molecular modeling. Develop focused experimental program driven in part by model development and designed

to provide information for parameter estimation and to answer mechanistic questions. Specific questions that must be addressed experimentally and within the construct of existing and new models to be developed are included under Mechanisms above.

Experimental Methods

Methods, “new” and “old” “Old”

Isothermal calorimetry IR spectroscopy Electron microscopy and associated chemical micro-analysis

(EDS/WDS) Nuclear magnetic resonance (NMR) Coherent and incoherent X-ray and neutron scattering

“New” Nuclear reaction resonance analysis (NRRA) Time-domain reflectometry dielectric spectroscopy (TDR-DS) Vertical scanning interferometry (VSI) Nano-X-ray tomography Atomic force microscopy imaging and associated micro- and

nano-probe techniques

Methods, “new” and “old” “Old”

Isothermal calorimetry IR spectroscopy Electron microscopy and associated chemical micro-analysis

(EDS/WDS) Nuclear magnetic resonance (NMR) Coherent and incoherent X-ray and neutron scattering

“New” Nuclear reaction resonance analysis (NRRA) Time-domain reflectometry dielectric spectroscopy (TDR-DS) Vertical scanning interferometry (VSI) Nano-X-ray tomography Atomic force microscopy (AFM) imaging and associated

micro- and nano-probe techniques

Vertical Scanning Interferometry

*A. Luttge, “Vertical scanning interferometry… ,” Kinetics Summit (2010).

Nano-indentation

*J. Chen, et al., “A coupled grid-indentation/SEM-EDX study on low w/c PC pastes… ,” Kinetics Summit (2010).

Nano-indentation

*J. Chen, et al., “A coupled grid-indentation/SEM-EDX study on low w/c PC pastes… ,” Kinetics Summit (2010).

Experimentation Roadmap Extend as necessary and apply the vertical scanning interferometry (VSI) technique in

an attempt to answer at least a portion of the questions regarding mechanisms. Further develop x-ray nano tomography into a quantitative technique and apply it to

study the rate of cement phase reaction in both model systems and portland cements and for blended systems containing silica fume, blast furnace slag and fly ash.

Further explore the use of nuclear resonance reaction analysis (NRRA) as a tool for elucidating the barrier layer hypothesis.

At this time, broadband time-domain-reflectrometry (BTDR) dielectric spectroscopy (DS) is a rather elusive technique that needs to be made generally available to the community at large. The datasets coming out of the single laboratory where the experiments are being conducted also need to be made generally available to the modeling community or need to be integrated with a modeling effort in an attempt to extract mechanistic kinetic information.

More datasets that combine techniques need to be developed so that modelers can impose multiple constraints in an effort to produce unique parameter sets with physical meaning.

Establish an open network with researchers in the broader community, both those doing modeling and experimentation, so that they have access to datasets and instrument time on unique tools such as VSI, nano x-ray tomography, NRRA and BTDR-DS.

Admixtures

A. A. Jeknavorian, Impact of Water Reducers and Superplasticizers on the Hydration of Portland Cement, International Summit on Cement Hydration Kinetics and Modeling (2009).

A. A. Jeknavorian, Impact of Water Reducers and Superplasticizers on the Hydration of Portland Cement, International Summit on Cement Hydration Kinetics and Modeling (2009).

D. Silva, Impact of Accelerators and Retarders on the Hydration of Portland Cement, International Summit on Cement Hydration Kinetics and Modeling (2009).

D. Silva, Impact of Accelerators and Retarders on the Hydration of Portland Cement, International Summit on Cement Hydration Kinetics and Modeling (2009).

Admixtures Roadmap Identify the elementary steps (reactions) for classic hydration

acceleration and retardation, e.g. for CaCl2 and sucrose activity respectively.

Isolate and identify physiochemical interactions with various cement surfaces and quantify the rate controlling processes, i.e. rate of adsorption of neat admixture chemical “A” on alite.

Isolate and identify chemical interactions with various cementitious ionic species, i.e. rate of solution phase chelation of Ca+2 by sucrose.

Design experiments explicitly to be used for kinetic model development with the objective of having quantitative outcomes.

Utilize molecular modeling as a way to discover mechanistic steps in admixture-influenced hydration kinetics.

Supplemental Cementitious Materials






Supplemental Cementitious Materials Roadmap

Isolating the rate of reaction of the SCM or SCM phases is generally very difficult. Most reactive SCM’s are amorphous rather than crystalline and some, i.e. fly ash, contain more than one reactive constituent, that is, may contain more than one reactive glassy phase each having its own reactivity.

Characterization of the microstructure of even the parent SCM can be difficult , for materials such as fly ash which may contain more than one reactive glassy phase.

So called “filler effects” are difficult to separate from nucleation related effects since both may have similar apparent outcomes, i.e. slightly alter the size and location of the primary calorimetry peak (Stage 3 and 4).

Generalized solubility models for the range of glassy phases are not readily available.

Alternative Cements

Alternative Cements Roadmap While there are kinetic datasets for the various classes of cements, there is a general lack

of a cohesive unified theory for the common cement forms, e.g. those that are indirectly derived from the portland family of anhydrous crystalline cements and those derived predominantly from glassy raw materials and requiring high alkali content activator solutions.

Although the kinetic processes share features in common with those of C3S-based cements, it seems that side-by-side studies of these features have not been conducted.

There is a general lack of information regarding long-term durability for many classes of alternative cements. While somewhat outside of the scope of this roadmap is should be acknowledge that such is an obstacle in path for more widespread development and use of such materials.

There are a number of economic hurdles at this time, including the use of bauxite as a raw materials and the high cost of alkali agents, which if relieved by further development or materials engineering might reduce production cost and enable the introduction of alternative cements into various markets where they are presently not economically viable.

It appears that some alternative cement systems exhibit kinetic features that are like those of the C3S-based portland cement system, including behaviors such as an early dissolution peak, dormancy and a main hydration peak, i.e. supesulfated cements and calcium sulfoaluminate cements. While this is well known among the community of researchers, it might be beneficial to study such systems side-by-side in an effort to resolve common or dissimilar features that could lead to a more refilled and clarified mechanistic theory of cement hydration.

Roadmap cement phase solubility (AAS)transport properties (NRRA/NMR) barrier layer (NRRA/BTDR-DS) global rates (calorimetry) nucleation (VSI)

morphology of C-S-H (SEM/TEM/AFM/X-ray) density (SEM/EDS/nano-X-ray tomography) product composition and microstructure (AFM/SEM/EDS)

admixture isotherms (?)admixture nucleation effects (VSI) admixture barrier layer effects (NRRA) admixture-pozzolan interactions (?)

Continuum

HydratiCA icMolecular

Kinetics of (?) :aluminate phase hydration aluminates-silicates co-hydration sulfates-aluminates co-hydration sulfates-aluminates-silicates co-hydration

Kinetics of (?) :slag, ash and silica fume co-hydration…

Text in () indicates suggested analytical methods.

unifying kinetic-thermodynamic theory (kMCS-VSI)

Research Simulation Environments Engineering Tools

New Materials “Highway”

Concluding Remarks There is a general lack of resource organization and dissemination of

tools for modeling cement hydration. A National resource for hydration data should be considered wherein a database of computer models, thermophysical properties (thermodynamic datasets and thermodynamic models), crystallographic information files (CIF), kinetic datasets, models and modeling tools and their associated source codes, etc., can be easily accessed by the research community at large. Presently, huge amounts of time are spent by research teams searching for, reviewing and assembling such information independently.

There is presently no focal point for hydration research in the US, but there should be. Concrete is the primary building materials for the world’s infrastructure and the US must continue to remain competitive and be a global leader in concrete materials technology. The lack of a generalized, universal theory governing chemical transformations, microstructure development and properties of complex hydrate synthetic mineral-based materials impedes the pace of development.

Publications and Dissemination J. W. Bullard, H. M. Jennings, R. A. Livingston, A. Nonat, G. W. Scherer, J. S. Schweitzer, K. L.

Scrivener, and J. J. Thomas, Mechanisms of Cement Hydration at Early Ages, Cem. Concr. Res., (submitted, 2010).

T. Xie and J. J. Biernacki, The Origin and Evolution of Cement Hydration Models, Comp. Concr., (submitted, 2010)..

J. J. Thomas, J. J. Biernacki, J. W. Bullard, S. Bishnoi, J. S. Dolado, G. W. Scherer and A. Luttge, Modeling and Simulation of Cement Hydration and Microstructure Development, Cem. Concr. Res., (submitted, 2010).

A. Luttge, Experimental Methodologies for Investigating Cement Hydration Kinetics, Cem. Concr. Res. Cem. Concr. Res., (in preparation, 2010).

J. Cheung, L. Roberts, A. Jeknavorian and D. Silva, The Impact of Admixtures on Hydration Kinetics, Cem. Concr. Res., (submitted, 2010).

K. Scrivener, B. Lothenbach, and D. Hooton, The Effect of Supplemental Cementitious Materials on Hydration Kinetics, Cem. Concr. Res., (in preparation, 2010).

M. Juenger, F. Winnefeld, J. Provis and J. Ideker, Advances in Alternative Cementitious Binders, Cem. Concr. Res., (submitted, 2010).

J. J. Biernacki, et al., Paving the Way for a More Sustainable Concrete Infrastructure – A Roadmap for the Development of a Comprehensive Description of Cement Hydration Dynamics, TBD (in preparation, 2010).

International Summit on Cement Hydration Kinetics, http://blogs.cae.tntech.edu/hydration-kinetics/

Cement Hydration Kinetics and Modeling, http://blogs.cae.tntech.edu/jbiernacki/welcome-to-cement-hydration-kinetics-and-modeling/

http://blogs.cae.tntech.edu/hydration-kinetics/

http://blogs.cae.tntech.edu/jbiernacki/welcome-to-cement-hydration-kinetics-and-modeling/

Documents

International Summit on Cement Hydration Kinetics and Modeling * a Summary and outcomes