2015 2nd International Conference on Material Engineering and Application (ICMEA 2015)
ISBN: 978-1-60595-323-6
Early-Stage Microstructures of Cemented Paste Backfill with
Different Binders
Juanrong Zheng*1, Xiaoyu Lu
College of Civil Engineering, Zhengzhou University, Zhengzhou City in Henan Province 450002, China
ABSTRACT: The early-stage (within 7 days of hydration) microstructures of cemented paste
backfill (CPB) with different binders were compared using SEM and EDXS. The two types of CPB,
denoted CPB1 (mixed with 5% Portland cement, PC) and CPB2 (mixed with 5% complex binder,
CB), were prepared. In CPB, white particle accumulations (WPA) smaller than 10 µm filled the
spaces surrounding the big tailings, small amounts of the WPA also adhering to the surfaces of
the big tailings. The PC products self-aggregated within the CPB1 paste, and the CB hydration
products were relatively evenly dispersed throughout the CPB2 paste. The initial
microstructure of the CPB2 was more compact than that of the CPB1.
1. INTRODUCTION
Cemented paste backfill (CPB) is typically comprised of 70-85% solids by mass with amounts
of fine particles sufficient for preventing particle precipitation and segregation and, thereby,
water bleeding during pipeline transportation (Kesimal, et al, 2003; Fall et al, 2005). Binders are
added to the tailings of CPB in order to increase their mechanical strength. Binders have a 28day curing strength of approximately 2 MPa. Binders typically account for 3-7% of the total
solid mass of CPB. However, a binder can account for 75% of the cost of a cement backfill.
Portland cement is primarily binder in CPB (Ouellet, et al, 2007; Benzaazoua, et al, 2004;
Yilmaz, et al. 2011). Currently, a binder composed of slag powder (75%), Portland cement
(15%), lime (5%), and anhydrite (5%) is used in most of the mines in China (Du, et al. 2012;
Han 2001; Yang, et al. 2014) .
Although Ouellet et al. (2007) studied CPB microstructures using mercury intrusion
porosimetry (MIP), no research concerning the application of SEM and EDXS to study the
early-stage microstructures of CPB has been published.
1
Corresponding author: College of Civil Engineering, Zhengzhou University, No.100 Science Road, Zhengzhou
City, Henan Province 450000, People's Republic of China.
E-mail: [email protected]. Tel.: 13938257703.
733
2. MATERIALS AND METHODS
2.1 Raw materials
The raw materials tested in this study included binders, tailings, and deionized water. The mine
tailings were obtained from an iron mine in Anhui province, China. The major, moderate, and
minor mineralogical contents of the tailings included dolomite [CaMg(CO3)2], quartz (SiO2),
and hematite (Fe2O3), respectively. The particle size distributions of the tailings with particles
larger than 0.075 mm and smaller than 0.075 mm were determined using a sieve analysis and
laser imaging, respectively. The particle size distributions of the tailings are shown in Table 1.
The chemical compositions are shown in Table 2.
Table 1. Particle size distribution of the tailings.
Sieve hole sizes /mm
1.18-0.60
0.60-0.30
0.30-0.15
0.15-0.075
0.075-0.020
<0.02
Cumulative /%
1.67
14.1
20.0
35.0
90.0
100
Table 2. Chemical compositions of the tailings and binders/wt%.
Type
SiO2
Tailings 41.4
Al2O3
Fe2O3
CaO
MgO
SO3
Na2O
K2O
f-CaO
Loss
0.3
5.5
18.0
11.7
-
1.5
1.2
-
18.5
PC
25.26
6.38
4.05
54.67
2.68
2.66
0.075
0.60
0.2
2.59
CB
29.37
11.99
1.10
42.54
8.08
3.74
0.46
0.47
0.10
3.17
Table 3. Physical properties of the binders.
Specific surface
Type
Area
/ (m2/kg)
Setting time
Flexural strength
/min
/MPa
Standard
consistency
/%
Initial
Final
setting
setting
Compressive
strength
/MPa
3d
28d
3d
28d
PC
353
26.2
150
210
5.1
8.0
26.6
51.3
CB
385
31.6
219
243
2.2
7.8
6.1
30.2
The two binders used in this study included Portland cement (PC) and a complex binder (CB)
comprised of blast-furnace slag powder (BFSP) (75%), PC (15%), lime (5%), and anhydrite
(5%). The chemical compositions and physical properties of the binders are shown in Tables 2
and 3, respectively.
734
2.2 Methods
The samples were prepared according to the mixture proportions shown in Table 4. The samples
were placed in sealed plastic barrels and maintained at 20°C. The samples were removed after
the predetermined amounts of time had passed. A small piece of each sample was removed and
immersed in CH3COCH3 for 24 hours in order to terminate hydration. Then, the samples were
dried at 60°C for 24 hours. The samples were subjected to the SEM and EDXS analysis seven
days after hydration. A JSM-7500F SEM (Japan), with an operating voltage of 15 kV and a
current intensity of 10 μA, was used for the SEM analysis. The EDXS, with an operating
voltage of 15 kV, was used to examine the chemical compositions of the samples.
Table 4. The mix proportions of the paste samples and compressive strength after curing.
Paste name
Mix proportion
Compressive strength
/(kg)
/KPa
Tailings
PC
BC
Water
7d
28d
CPB1
95
5
-
28.2
530
680
CPB2
95
-
5
28.2
899
1400
3. RESULTS AND DISCUSSION
3.1 Microstructures of the CPB1 and CPB2 samples after seven days of hydration
Figure 1 displays the microstructures of CPB1 after seven days of hydration. After seven days
of hydration, the CPB1 sample exhibited a large amount of white particle accumulations (WPA)
around the big tailings. Although the contours of the tailings were still recognizable and the
paste still contained noticeable gaps, the CPB1 sample had an open structure in that the tailings
were not entirely interconnected. In the CPB1 sample, 30-μm aggregates of cement hydration
products were observed, indicated by arrows in Figure 1. The hydration products primarily
consisted of fiber-like CSH, as shown in Figure 2. These results indicated that the PC hydration
products self-aggregated in the CPB1 paste.
The EDXS analysis indicated that the chemical compositions of the WPA in the CPB1 sample,
indicated by point C in Figure 3, consisted of O/Ca/Fe/Mg/S/Al/Si with a Ca/Si molar ratio was
15.3. These results suggested that the Ca(OH)2 in the PC hydration products could have been
deposited on the surface of the WPA, leading to a substantial increase in the calcium contents of
the WPA. The EDXS analysis also indicated that the big tailings in the CPB1 sample, indicated
735
by point D in Figure 3, consisted of O/Ca/Mg. In addition, identical to the Ca/Mg molar ratio of
the tailings, the Ca/Mg molar ratio was equal to 1.09, as shown in Table 2. These results
indicated that no hydration products were deposited on the surfaces of the big tailing particles.
Figure 1. Micromorphology of the CPB1
after seven days of hydration.
Figure 2. Fibriform CSH in the CPB1
after seven days of hydration.
Figure 3. Surfaces of the big tailings in the CPB1 after seven days of hydration.
Compared to the CPB1 sample, the CPB2 sample exhibited more WPA surrounding the big
tailings and, thereby, a more compact microstructure (Figure 4). However, the CPB2 sample
was still characterized by an open structure at this point in time. In the CPB2 sample, needlelike ettringite crystals approximately 5 μm in length were deposited around the WPA, on the
surface of the big tailings, and in the gaps in the paste, as shown in Figure 5. The EDXS
analysis indicated that the WPA in the CPB2 sample, indicated by point E in Figure 6, was
comprised of O/Mg/Ca/Si/Na/Al, with a Ca/Si molar ratio of 1.7. In addition to the ettringite
crystals and WPA, a layer of mesh-like substances also covered the surfaces of the big tailings,
as indicated by point F in Figure 6. The EDXS results indicated that this layer was comprised of
O/Mg/Ca/Si/Na/Al/S/K, with a Ca/Si molar ratio of 3.1. Thus, the chemical composition of the
mesh-like substances was similar to that of the WPA, suggesting that the CB hydration products
736
were relatively evenly dispersed throughout the CPB2 paste.
Figure 4. Micromorphology of the CPB2 after
seven days of hydration.
Figure 5. Rod-shaped ettringite crystals in the
CPB2 after seven days of hydration.
Figure 6. Surfaces of the big tailings in the CPB2 after seven days of hydration.
In summation, the CPB was characterized by a low binder content. The number of hydration
products, porosity, and solid connectivity were dependent on the type of binder. After seven
days of hydration, the CPB was still characterized by an open structure in that the tailings were
still not entirely interconnected. After seven days of hydration, the amount of uniform WPA and
hydration products in the CPB2 sample was greater than that in the CPB1 sample. Thus, the
early-stage microstructure of the CPB2 sample was more compact than that of the CPB1 sample.
This was the primary reason the compressive strength of the CPB2 sample after seven and
twenty-eight days were 1.7 and 2.1 times greater than that of the CPB1 sample after seven and
twenty-eight days, respectively (Table 4). However, further studies concerning the effects of
each component of the composite binders on their compressive strengths should be conducted.
737
4. CONCLUSIONS
(1) In the CPB1 sample with 5% PC, WPA smaller than 10 µm in length primarily filled the
spaces surrounding the surfaces of the big tailings, with small amounts adhering to the
surfaces of the big tailings. The PC hydration products self-aggregated within the CPB1
paste.
(2) In the CPB2 with 5% CB, WPA primarily filled the spaces surrounding the big tailings, with
small amounts adhering to the surfaces of the big tailings. Needle-like ettringite crystals
approximately 5 µm in length surrounded the WPA after seven days of hydration. The CB
hydration products were relatively evenly dispersed throughout the CPB2 paste.
(3) Both the CPB1 and CPB2 samples exhibited open structures after seven days of hydration in
that the tailings were not entirely interconnected. However, after seven days of hydration,
the amount of uniform WPA and hydration products in the CPB2 sample was greater than
that in the CPB1 sample. Thus, the early-stage microstructure of the CPB2 sample was
more compact than that of the CPB1 sample.
ACKNOWLEDGMENTS
This study was funded by the National Natural Science Foundation of China (Project No.
51274174). The authors would like to thank Professor Zhang Wenyan of Zhengzhou University
for his assistance with the SEM analysis.
REFERENCES
Kesimal, A. Ercikdi, B. Yilmaz, E. (2003) The effect of desliming by sedimentation on paste backfill
performance. Min Eng, 16 pp. 1009-1011.
Fall, M. Benzaazoua, M. Ouellet, S. (2005) Experimental characterization of the influence of tailings
fineness and density on the quality of cemented paste backfill. Min Eng, 18(1) pp.41-44.
Ouellet, S. Bussiere, B. Aubertin, M. et al. (2007) Microstructural evolution of cemented paste backfill:
Mercury intrusion porosimetry test results. Cem Conc Res, 37 pp. 1654-1665.
Benzaazoua, M. Fall M., Belem, T. (2004) A contribution to understanding the hardening process of
cemented pastefill. Min Eng, 17(2) pp. 141-152.
738
Yilmaz, E. Belem, T. Bussiere, B. et al. (2011) Relationships between microstructural properties and
compressive strength of consolidated and unconsolidated cemented paste backfills. Cem Conc
Compos, 33(6) pp.
Du Ju-qing, Gao Qian, Nan Shi-qing, et al. (2012) Develop on a new cementing Material of full tailings
backfilling, Metal Mine,431(5) pp.152-155. (in Chinese)
Han Ke-feng. (2001) Application of Fine Grounded Granulated Blast-furnace Slag in Whole Tailings
Cementation Filling. Metal Mine, (7) pp.45-47. (in Chinese)
Yang ,Zhi-qian, Xiao Bo-lin, Gao Qian, et al. (2014) Experimental study on development of a new
cementitious material based on Rod Grinding in Jinchuan Nickel Mine. Ferrous Metal (mine section),
66(5) pp.65-68. (in Chinese)
739