Beruflich Dokumente
Kultur Dokumente
James Pommersheim
Jemei Chang
Bucknell University
Lewisburg, PA 17837
ABSTRACT
Introduction
911
912 Vol. 18, No. 6
J. Pommersheim and J. Chang
Conceptual Model
STAGE I
SO 4 diffuses in
dissolves
~ , , forms ©
STAGE 2
---~Stege 1 ~ 2.
C3A species diffuse out
Stage
STAGE 3 TIME, t
Hexagonal ~heees form
H20 ~ ~'~
FIG. 2
FIG. i
Schematic Description of
Hydration Stages
First Sta~e
Second Stage
Third Stage
Mathematical Models
The k i n e t i c models developed here are based on the conceptual model and
rate laws and conservation equations w r i t t e n for a dispersion of s i n g l e - s i z e d
p a r t i c l e s in s o l u t i o n . The s o l u t i o n is considered to be uniformly mixed and to
remain isothermal.
First S~age
dmA dme I
= - N (5)
dt dt 3 s
where
me is the moles of e t t r i n g i t e
dr A i aCs re
(6)
dt = ~ LADsc dA rA(re _ rA)
Vol. 18, No. 6 915
C3A, HYDRATION, KINETICS, GYPSUM
dr e dr A
- (1 - d) (7)
dt re dt
dA
with d =
( i - c ) de
where
r, 2 i (rA) 3 2/3 t
1 - (~) + { [d + (1 - d) ] - I ~ = 2- (9)
rAO 1-d rAO ts
Equation (10) shows that the time constant varies with the square of the
i n i t i a l particle size. Thus, larger particles w i l l take s i g n i f i c a n t l y longer to
hydrate during t h i s period than smaller.
916 Vol. 18, No. 6
J. Pommersheim and J. Chang
rAi i i/3
- [i - ~ (g/a)] (12)
rAO
This occurs (at t l ) when a l l the gypsum has been consumed S u b s t i t u t i o n of
equation (12) i n t o equations (8) and (9) g i v e s , r e s p e c t i v e l y , the p a r t i c l e s i z e ,
r e l , and reduced time, t L / t s , at the end of the f i r s t stage.
Second Stage
dmA dme
= - 2 = NA (13)
dt dt
where
i
r e = [ r e l 3 + (I + -- d) (r~ 3 - rAl3)] l/3 (14)
2
rA 2 I r 2 re I z 2(t - tl)
I - (-) + [ ( e ) ( ) ] - (15)
rAl I + i/2 d rAl rAl tA
Vol. 18, No. 6 917
C3A, HYDRATION, KINETICS, GYPSUM
During this stage r e remains greater than r A, and both the aluminate and
e t t r i n g i t e layers w i l l be shrinking. The time constant for diffusion of
dissolved C3A species through the e t t r i n g i t e layer is given by:
where
aCA is the concentration difference of C3A species~between the inside and
outside of the e t t r i n g i t e barrier layer, mol/m~
The model predicts that the second stage w i l l end abruptly when:
re2 _ rA2
Ll - ( g / a ) ] Z / 3 (17)
rA 0 rA 0
This occurs when all the e t t r i n g i t e has been consumed. Substitution of equation
(17) into equation (15) predicts the time t 2.
Third Stage
There are three time constants for the processes which occur during the
t h i r d stage, tw, for diffusion of water through the pores of a hexagonal phase
barrier layer formed adjacent to the C3A p a r t i c l e ; t x , for dissolution of t h i s
layer; and tc, for c r y s t a l l i z a t i o n of cubic phase from solution. These l a t t e r
steps take place through mass transfer films adjacent to p a r t i c l e surfaces.
Since the steps occur in series, the one with the longest characteristic time
w i l l control the hydration process. For the C3A - H2O system this has been
found to be t w [9].
Model Output
dQ da
- n (-AH)- (18)
dt dt
Model Predictions
100 i i
2000
1000 3.05 2 2
S00 : 0.1S
2
*
t ' x tO 3
',
4
0.2S
6
25
o~ 10
i•AA
20
t(h)
•
3O
A ,
40
*
50
FIG. 3 FIG. 4
mes thicker i t impedes the diffusion of ions through i t , and the reaction rate
turns lower u n t i l the point at which all the gyosum in the system has been con-
sumed (vertical dashed line I ) . An apparent induction period is entered as
indicated by the low f l a t region. This is common for all the curves u n t i l the
f i r s t stage ends. In the second stage, the conversion and heat release rates
rise again. As the e t t r i n g i t e layer becomes thinner, the diffusion of C3A spe-
cies outward through the e t t r i n g i t e to the solution also becomes easier, and the
reaction accelerates. The second stage ends when all the e t t r i n g i t e has reacted
( l i n e 2). Discontinuous break points on each curve show the switching from one
stage to another. The jumps are higher at lower values of (g/a). As shown,
lower gypsum to C3A ratios give shorter induction periods and higher heat
release peaks. Very large conversion rates are predicted to occur at the end of
the second stage, when the e t t r i n g i t e barrier layer is completely converted to
monosulfate, and the unreacted C3A again comes into direct contact with the
solution.
For the f i r s t two staqes of hydration Table 1 oresents the results of this
analysis.
Table I. %del Time Constants (h) for First Two Stages, from
Regression Analysis of ~1ori and ~linegishi's Data [28].
P1ethod ts a
t t_~.l __tz ts/t a
+constrained
Figure 4 shows the model f i t for the X-ray diffraction method, where both
t l and t~ were set at their measured values before the regression analyses were
performea. This is the most constrained of the three methods, but even with
these constraints there is reasonable agreement. The sharp break at the end of
the second stage, corresponding to the dissolution of all the e t t r i n g i t e , is
steeper and occurs later than the experimental data shows. The model also pre-
dicts a jump at the end of the f i r s t stage ( t i = 12 hours, marked by the dashed
vertical line) which is not matched by the data. For the other methods (not
shown) this jump occurs earlier and is less pronounced. This could be a result
of the presence of an intermediate or transition stage. Chatterji [13] has
suggested that monosulfate may be formed in the presence of gypsum near the end
of the f i r s t stage of hydration. T h i s could also give the lower heat liberation
rates observed when e t t r i n g i t e disappears.
Discussion
Conclusions
The differential equations of the model predicted the size and amounts of
reaction products, and the rate and degree of hydration and heat release. Model
input included d i f f u s i v i t i e s , phase densities, water to solid and ~ypsum to alu-
minate ratios, and saturation concentrations. The remaining (unknown) model
parameters were time constants, for diffusion of ions through the pores of the
e t t r i n g i t e barrier layer (t s and ta) , and for hydration of C3A after the
e t t r i n g i t e had been consumed (tw, t x and t c ) .
Comparing the model to experimental heat release data using the time
constants as regression parameters a reasonable f i t was obtained. Results
Vol. 18, No. 6 921
C3A, HYDRATION, KINETICS, GYPSUM
Acknowledgements
Nomenclature
C& C3A conversion
CA C3A species concentration, mol/m~
Cs sulfate ion c,mcen~ration, mol/m~
d dimensionless density ratio, dA/(1-¢)d e
DA d i f f u s i v i t y of C3~, m2/s
dA molar density of C3A, mol/m3
de molar density of e t t r i n g i t e , mol/m3
£ porosity of e t t r i n g i t e layer
g/a gypsum-tricalcium aluminate molar ratio
~A shape factor of C3A particle
mA moles of C3A
me moles of e t t r i n g i t e
~ A, NS diffusion rate of C3A species and sulfate ions, mol/s
heat evolution, cal/mol
cA radius of C3A, m
re radius of C3A particle covered by e t t r i n g i t e phase, m
rAO i n i t i a l radius of C3A particle, m
rAi, rA2 radii of C3A particle at end of f i r s t and second stage, m
rel, re2 radii of C3A particle covered by e t t r i n g i t e phase at end of
f i r s t and second phases, m
t l , t2 duration of f i r s t stage, s, end of second "stage, s
tA, t c, t s, time constants, for diffusion of C3A species through e t t r i n g i t e ,
tw, t x for crystallization of cubic phase from solution, for diffusion of
sulfate ion through e t t r i n g i t e , for water diffusion through hexagon-
al phase, and for dissolution of hexagonal phase, s
References