TL;DR
This study investigates the strength characteristics of Deep Cement Mixing Integrated Drilling, Mixing and Jetting (DMJ) piles in the Yellow River Floodplain. Researchers found that DMJ pile strength follows a normal distribution and is strongly correlated with depth, especially at the pile core. Crucially, they established a linear relationship between electrical resistivity and unconfined compressive strength, offering a promising non-destructive method for evaluating DMJ pile quality.
What Problem Does This Paper Address?
Highway construction in China frequently encounters soft soil conditions, leading to inadequate bearing capacity and issues like excessive settlement and differential settlement. Deep cement mixing (DCM) technology has been used to reinforce foundations, but in areas like the Yellow River Floodplain, the high consolidation of clay presents challenges, sometimes resulting in suboptimal pile quality. The DMJ method, which combines drilling, mixing, and high-pressure jetting, aims to improve pile quality. However, the spatial variability of strength within these DMJ piles, particularly the interplay between the jetting-influenced outer layer and the mixed inner core, is not fully understood. This lack of comprehensive insight into spatial variability poses challenges for accurate strength assessment and can lead to increased engineering costs due to conservative design practices.
Methods
The study employed a multi-pronged approach to assess DMJ pile strength:
- Unconfined Compressive Strength (UCS) Testing: Core samples from field-tested DMJ piles were subjected to UCS tests to determine their strength.
- Statistical Analysis:
- Kolmogorov-Smirnov (K-S) Test: Used to evaluate whether the distribution of pile strength followed a normal distribution, a key assumption for many statistical analyses.
- Autocorrelation Function (ACF) Method: Applied to analyze the spatial variability of pile strength along the depth of the pile, helping to understand how strength fluctuates and correlates with depth. This method helps quantify the "fluctuation range" or scale of fluctuation (SOF).
- Resistivity Non-Destructive Testing (NDT): Electrical resistivity measurements were taken on core samples. This technique was chosen because it's non-destructive and has shown potential correlations with soil strength in previous research. The study used an AC current to avoid electrochemical effects that DC current can induce.
Datasets
The study utilized core samples obtained from field-tested DMJ piles constructed for a highway foundation reinforcement project in Dezhou, Shandong Province, China. Two types of DMJ piles (Pile 1 and Pile 2) were tested, differing in grouting pressure (25 MPa for Pile 1 and 15 MPa for Pile 2). The soil profile at the site was characterized as typical Yellow River Floodplain clay and silty clay interbedded structure. A total of 127 samples were analyzed (66 from Pile 1, 61 from Pile 2).
Key Insights and Ideas
- Normal Distribution of Strength: The K-S test confirmed that the strength of DMJ pile bodies generally follows a normal distribution at a 5% significance level. This is a crucial finding for statistical modeling and reliability analysis.
- Depth-Dependent Strength: A clear correlation was observed between pile core strength and depth, suggesting that deeper sections might exhibit different strength characteristics. This is particularly pronounced in the pile's inner core.
- Outer vs. Inner Pile Strength: The strength of the pile's outer side (influenced by jetting) showed less pronounced correlation with depth compared to the inner core. This highlights the distinct formation mechanisms and potential variability within the pile body.
- Resistivity as a Strength Indicator: A significant positive linear correlation was found between electrical resistivity and unconfined compressive strength. This is a major practical insight, as resistivity is a readily measurable, non-destructive parameter.
- Cement Dosage Matters: The study implies that cement dosage significantly impacts pile formation quality and strength uniformity. Lower cement dosage (as in Pile 2) led to lower average strength at the center compared to the sides.
Key Results
- The K-S test indicated that DMJ pile strength distributions are approximately normal (P > 0.05).
- The autocorrelation function analysis revealed that the Scale of Fluctuation (SOF) for the pile core was generally higher than for the outer side, indicating better correlation with depth for the core. For instance, Pile 1's center SOF was 10.7, while its outer side SOF was 5.5. Pile 2's center SOF was exceptionally high at 74, suggesting strong depth correlation.
- A linear relationship was established between unconfined compressive strength (y) and resistivity (x) for core samples: $y = 8.4 + 19x$, with an R² value of 0.87. This indicates that resistivity can effectively predict strength.
- The study noted that the outer side of the pile exhibited more pronounced strength variations and higher maximum strength values compared to the center, likely due to the high-pressure slurry injection and its interaction with the soil.
Limitations and Future Work
- Spatial Variability Complexity: While the study analyzed depth-wise variability, the paper acknowledges that factors like soil quality variations, cement mixing uniformity, and inherent engineering construction uncertainties can influence strength and resistivity. The spatial variability of DMJ piles is complex and requires further investigation.
- Indoor vs. Field: The study acknowledges that design strengths for field construction are often lower than indoor test results, suggesting a need for further research to bridge this gap.
- Generalizability: The findings are based on tests in the Yellow River Floodplain. Further research is needed to confirm the applicability of these findings to other soil types and geological conditions.
- Future Work Suggestions: The paper implicitly suggests that the established resistivity-strength correlation can be further utilized for non-destructive evaluation of DMJ piles. Future work could focus on developing more refined empirical models and validating them across a wider range of conditions. Investigating the influence of other factors like curing time and environmental conditions on this correlation would also be valuable.
Summary updated