Soil stabilization

This Technical Summary pertains to project TPF-5(215), “Best Practices for Soil Stabilization Design and Construction,” completed September 2015.

This Research Summary accompanies Report 2024-15, Understanding the Performance of Road Base Stabilization Additives, published in May 2024.

Base stabilization additives are used to increase the strength and stiffness of road foundations on weak and susceptible soils. The Minnesota Department of Transportation (MnDOT) quantifies the structural contribution of pavement layers by introducing granular equivalency (GE) factors. While numerous additives exist for improving the performance of aggregate base layers, this study focuses on proprietary additives including Base One, Claycrete, EMC SQUARED, PennzSuppress and Roadbond EN1. The laboratory study revealed that EMC SQUARED was the superior stabilizer, with an optimum dosage set 15% higher than the manufacturer recommended dosage (MRD). The long-term performance of proprietary additives was monitored by considering full-scale field implementation with optimum additive dosages obtained from laboratory investigation. Controlled sections without stabilization exhibited higher values in the California Bearing Ratio (CBR) and composite elastic modulus right after construction, while the impact of stabilizers on the increasing strength of the full depth reclaimed (FDR) base was revealed after two years of construction. Falling-Weight Deflectometer (FWD) tests demonstrated a progressive increase in the stiffness of stabilized sections over time, surpassing the control section's stiffness after two years. The economic analysis utilizing Life Cycle Cost Analysis (LCCA) indicated that stabilized sections, particularly those treated with EMC SQUARED, offered lower Equivalent Uniform Annual Cost (EUAC) values across various maintenance scenarios. These findings suggested potential cost savings over a pavement's life cycle with higher GE factors of recycled asphalt pavement base aggregate treated with proprietary additives. The findings will contribute to a comprehensive understanding of the benefits, feasibility, and design considerations associated with using commercial stabilizers in FDR base layers.

A series of laboratory tests were made by compacting 18 different soils at a variety of density and moisture conditions in boxes. Density determinations were made with the nuclear apparatus and with the sand cone and by calculation from the volume of the box. Comparisons of the sand cone and nuclear results to the box-volume results gave root mean squared errors (RMSE) which indicated their probable errors. Accuracy of the sand cone and the direct transmission device was about the same, and slightly better than the backscatter device. Improvements on the calibration curves furnished by the equipment supplier were made with the box tests. The field tests on soils permitted the comparison of a large number of nuclear tests by both direct transmission and backscatter taken in a limited area on several projects with sand cone results. The direct transmission results checked the sand cone densities much better than the backscatter results. The nuclear densities were calculated both from. the laboratory box-volume correlation curves and also from a correlation with the field sand cone densities. Root mean squared errors are used to depict the comparisons. Comparison of moisture contents measured with the backscatter device to oven dry values as well as other methods and the effect of these determinations on calculated dry densities are presented. Nuclear density measurements of bituminous pavements check displacement densities made on cores quite closely. The time requirements for various types of density tests and notes on the maintenance of the equipment are briefly discussed. The summary and conclusions list the probable errors in the nuclear methods using the box-volume or sand cone results as standards, and also list the recommended equations for determining densities from count ratios. The direct transmission device gives better results than the backscatter, especially in the field tests. The manner in which the equipment could be used for additional measurements is indicated.

An experimental field project was constructed in southeastern Minnesota 1965 to evaluate the effectiveness of stabilizing the upper portion of a silty subgrade with two different bituminous materials, RT-6 and MC-2. The project contained six sections in which the upper 6 inches of the subgrade was stabilized with bituminous material and capped with 3 or 5 inches of gravel. Two unstabilized control sections were designed with a 7-in. gravel base. The best performing stabilized section was on a par with the control sections, but was much more costly. The relative performance of the stabilized sections could be related to results from laboratory tests conducted on similar soil-bituminous mixtures.

This report covers two phases of the study. Chapter 2 presents data from tests on five silty soils from various locations in southeastern Minnesota. These data demonstrate the similarity and also the differences between these soils. These soils were tested in stabilized mixtures as well as in an unstabilized condition. Three other small samples were also obtained and the limited data on these soils are also presented for information. The second phase of the study was concerned with setting up a laboratory procedure for conditioning and for testing stabilized mixtures. For this part of the investigation only one soil was tested. A large supply of Soil No. 1 from Houston County was obtained and used for all tests relative to this approach

Southeastern Minnesota, as well as other large portions of the Midwestern United States, is largely covered with a blanket of loess. This blanket is so extensive that entire counties have almost no available soils other than silt loams. These silty soils are poor highway materials and create a problem for the engineers. This problem is compounded by the lack of granular material which may require long hauls or similar expensive procedures. One way to alleviate the problem is to improve or upgrade the native soil sufficiently to permit its use as a subbase and reduce the thickness of granular material required. This improvement requires some form of stabilization and the purpose of this study is to investigate the possibilities of bituminous stabilization.

The project was constructed by Wadena County in 1962 to evaluate the effectiveness of stabilizing the upper portion of a sandy subgrade with bituminous materials. It was also hoped to correlate the performance of the test section's with results of laboratory tests conducted in accordance with the recommendations of Investigation No. 608, "Bituminous Stabilization -Laboratory Study", recently published by the Minnesota Highway Department. The stabilized subgrade performed satisfactorily as base material on this project. However, similar results should not necessarily be expected on projects which have a different type of embankment soil or different traffic conditions.

This study compares the performance of a conventional flexible pavement (bituminous surface with granular base and subbase over untreated soil) with the performance of pavement designs incorporating lime stabilized soils. Included in the project were seven experimental sections each 1/4 mile long having various thicknesses of base and subbase and percentages of lime incorporated into the embankment soil. The report describes the soils, testing, design, construction and performance of the project. The lime stabilized test sections cost more and had more cracking and lower present serviceability indexes than the control sections.

The project was initiated in an attempt to develop a procedure using standard laboratory tests for designing soil-bituminous mixtures for use on highway construction projects. The Hveem stabilometer and cohesiometer and the unconfined compression tests were evaluated for this purpose using several types of soils and bituminous materials. Procedures were developed for molding the specimens, determining their desired moisture content and conditioning them prior to testing. The design method developed in this study appears to be more applicable to plastic soils, but probably has merit in designing some types of granular soil-bituminous mixtures also. The recommended design procedures given in this report were used in designing an experimental field stabilization project in Wadena County in north central Minnesota. The soils stabilized were fine sand, loamy sand and loamy fine sand. Results of this study will be described in the final report on Investigation No. 612 "Experimental Bituminous Stabilization Project - Wadena County" soon to be published.