Use of Tube Suction Test for Evaluation of Aggregate Base Materials in Frost/Moisture Susceptible Environment By Imran Syed Texas Transportation Institute Texas A&M University College Station, Texas 77843-3135 (409) 862-3647, FAX (409) 845-0278 i-syed@tamu.edu Tom Scullion, P.E. Texas Transportation Institute Texas A&M University College Station, Texas 77843-3135 (409) 845-9913, FAX (409) 845-0278 t-scullion@tamu.edu And Robert B. Randolph Soil Stabilization Products Company, Inc. Merced, CA 95344 (209) 383-3296, FAX (209) 383-7849 rbrandolph@sspco.org Paper Submitted for Presentation and Publication in the 2000 Annual Meeting of Transportation Research Board Washington, D.C. August 1999 Syed, Scullion and Randolph ABSTRACT A simple tube suction test has been developed by the Finnish National Road Administration for evaluating unstabilized granular base material. It consists of monitoring the capillary rise of moisture within a 150-mm diameter by 200-mm high cylinder of compacted aggregate. A probe is used to measure the dielectric constant at the surface of the sample. The dielectric constant is a measure of the "free" or unbound water within the aggregate sample. It is this unbound water which is thought to be directly related to strength of the material and to its ability to withstand repeated freeze thaw cycling. Measurements of the dielectric constants are made over a period of 10 days. The poorest performing materials are those that rapidly reach saturation and exhibit high surface dielectric values. This paper reports a study that was conducted on four marginal aggregates from Alaska, Nevada, New Mexico, and Texas. These aggregates have been reported to be poor performers in their environments. Each aggregate failed the tube suction test. A mineralogical evaluation was performed to identify the mineral composition of each aggregate and to explain their high affinity for moisture. Subsequently, the aggregates were treated with a concentrated liquid stabilizer (CLS). Upon treatment the aggregates showed improved performance in the tube suction test. The treated Alaska, Nevada and New Mexico aggregates were used in construction projects, the performance reports of which have been positive. The Tube Suction test described in this paper has potential to identify moisture susceptible aggregates and to guide in selecting the optimal stabilizer type and amount. Key Words: moisture susceptibility, suction/dielectric measurements, base stabilization, X-ray diffraction, pavements, concentrated liquid stabilizer. Syed, Scullion and Randolph ACKNOWLEDGMENT This work was funded by the Texas Department of Transportation (TxDOT). The support of Dr. Dar-Hao Chen, P.E., of the TxMLS Data Analysis Program in the Pavement Design Section of TxDOT is greatly appreciated. Help extended by Dr. J. B. Dixon, Dr. G. N. White, and Dr. J. P. Harris in carrying out the mineralogical evaluations is appreciated. Syed, Scullion and Randolph INTRODUCTION Free moisture is frequently the root cause of damage to flexible pavements. Although the sources of water and the mechanics of how moisture damages a pavement are understood, these principles are not widely incorporated into design. Primary source of moisture in pavement structures is rainwater, which infiltrates through surface of the pavement. Moisture can also enter a pavement from subsurface sources such as lateral seepage from a drainage ditch or subsurface flow from a high water table, a spring, wet subgrades, etc. It is well known that moisture infiltration reduces the strength and stiffness of most aggregate base materials, contributing to many pavement failures, and is particularly damaging in severe climates. The freeze-thaw process can have a dramatic influence on the permanent and resilient deformation of these pavement structures(1). Recent documented failures (2) of both flexible and stabilized bases indicate that the standardized index tests traditionally used to evaluate untreated aggregates and chemically bound aggregate materials can carry significant risk when relied upon for pavement designs in the absence of test methods to evaluate moisture susceptibility. STATEMENT OF PROBLEM This study focuses on four reported problem aggregates from the states of Alaska, Nevada, New Mexico, and Texas. Performance problems were compounded due to extreme freeze-thaw conditions in Alaska, Nevada, and New Mexico. The base course in Texas was in a region of wet subgrades and high rainfall. Each aggregate was tested using the AASHTO Classification System (AASHTO M145-87 or AASHTO M283-83) that included particle size distribution and Atterberg Limits. It is important to note that barring the aggregate from New Mexico, each aggregate passed current AASHTO (M145-87) and individual state specifications. Goal of Study i. ii iii To demonstrate that the Tube Suction Test can be used to identify moisture susceptible granular materials. To identify mineral components within the materials that may be causing the problem. To evaluate if the moisture susceptibility of these poorly performing materials as measured by the Tube Suction Test is reduced by treatment with a CLS which has a field performance history of successful stabilization of these materials. To verify laboratory results by actual field performance studies. iv Syed, Scullion and Randolph AGGREGATES USED IN THIS STUDY Alaska Aggregate These are from the stockpile used to supply a 1991 Federal Highway Administration Experimental Feature Project on Elliott Highway, beginning north of Fairbanks at the transition point from asphalt to gravel road. This unbound granular material had performed poorly, it was subject to constant potholing, washboarding, and gravel loss and required regular maintenance grading(3). Nevada Aggregate This material had been placed on the road in late Fall and then there was a delay of six to eight months until temperature conditions warmed to be fully suitable for hot mix asphalt placement operations. The unsurfaced access road had switchbacks and grades as steep as 13% and provided year round access for 200 or more daily trips by triple trailer trucks weighing approximately 60 tons, fully loaded. The project owners recognized that they had a moisture susceptible aggregate that required constant grading maintenance and daily watering in dry weather in order to service the truck traffic prior to paving. New Mexico Aggregate City of Gallup used this material for paving projects and observed poor performance over four and a half miles of distressed chip seal surfaces and aggregate surfaced streets that had deteriorated to the point that winter access had become problematic even for four wheel drive vehicles and trucks. Gallup experiences severe winter conditions and frequent freeze-thaw cycling, compounded by low bearing strength clay subgrade soils. Texas Aggregate This material was a mixture of siliceous river gravel mixed with locally available "caliche" (low quality limestone) fines. To reduce the plasticity index, this material was treated with 1½ percent of lime. It was used as base course material in a pavement section that was subjected to accelerated pavement testing, using the Texas Mobile Load Simulator (TxMLS). Test was conducted on a frontage road of US59 in Victoria, Texas, in the Texas Department of Transportation (TxDOT) Yoakum District. Detailed results of accelerated pavement test were reported by Hugo et. al. (4). Rutting and cracking developed quickly during testing, resulting in generally poor performance. Poor performance of this pavement test section was attributed to water ingress into the base. CURRENT CLASSIFICATION METHODS AASHTO and State DOT's AASHTO Designation: M145-87, Recommended Practice for The Classification of Soils and Soil-Aggregate Mixtures for Highway Construction Purposes, and AASHTO Designation: M 283-83, Standard Specification for Coarse Aggregate for Highway & Airport Construction was used to classify the four aggregate materials based on the AASHTO Classification System. Also, classification system used by each state highway departments (DOT) in Alaska, Nevada, New Mexico, and Texas was used to classify the aggregate materials. Results are tabulated in Tables 1 and 2. Syed, Scullion and Randolph PROPOSED CLASSIFICATION METHOD Tube Suction Test According to research conducted by Finnish National Road Administration and Texas Transportation Institute (1), electrical properties can be used to classify the strength properties of base course aggregates. Studies showed that the value of dielectric constant and electrical conductivity related to both strength and deformation properties, and the frost susceptibility of base course aggregates. Moisture affects the mechanical properties of all pavement materials and soils. However, magnitude of these effects depends on the material physical and chemical properties, moisture content, and even on the saturation history. Both positive and negative porewater pressures in a soil have a major effect on shear strength and volume change(6). Negative pore pressure is often described by the term "suction" which is used to describe the Gibbs free energy, a thermodynamic quantity which generates tension in the pore water between soil particles(7). Total suction is composed of two components: matric and osmotic suction(8). In base materials with low moisture content, suction can increase the resilient modulus of aggregates(9, 10), but when moisture content increases, the effect of suction decreases(8). Based on ultramicroscopic observations, it has been concluded that due to the very small size of clay particles, powdered clay particles dissolve in water and form a colloidal solution(11). Water molecule is bipolar and is attracted to the negatively charged clay particles. Height of capillary water, h in cm, is given by following equation: h ' 2 cos r wg is surface tension, dynes/cm is contact angle in degrees between the liquid and surface 3 w is density of the liquid in g/cm g is acceleration due to gravity in cm/s 2 r is capillary radius in cm Contact angle for water against most soils is very small. So, cos . 1 3 = 73 dynes/cm, and g = 981 cm/s 2 Also, if we substitute w for water = 1 g/cm , we arrive at the simplified equation: where: h ' .15 r It is thus evident that height of capillary rise is inversely related to the radius of capillary. In current laboratory testing procedures we lack the ability of quantifying the volumetric water that is held by capillary forces within pavement base materials. Tube Suction Test is a promising laboratory testing technique that needs further evaluation. Syed, Scullion and Randolph Test set up and typical results are shown in figure 1. This test measures the suction/dielectric properties of materials by monitoring the capillary rise of moisture within a 200-mm high cylinder of compacted aggregate. Moisture conditions of the aggregate surface are monitored with a probe that measures the dielectric constant. The value of dielectric constant is a measure of "free" or unbound water near the surface of aggregate sample. This test is not a simple measure of moisture content of the material but an assessment of the state of bonding of water within fine aggregates. MINERALOGICAL EVALUATION Pretreatment and Fractionation To identify component minerals, the base sample was separated into different fractions based on particle size, i.e. sand (2 mm-50 µm), silt (2 µm-50 µm), coarse clay (0.2 µm-2 µm), and fine clay (