Pergamon Journal of Terramechanics – Rapid Stabilization of Thawing Soils (TPD0307003)

1. Introduction

 

Thawing soils can reduce vehicle mobility on unsurfaced roads or trails and severely restrict, and sometimes even prohibit, off-road travel. In addition, traffick- ing may cause damage by rutting and tearing of surface vegetation, and subsequent erosion. In frost-susceptible soils, freezing temperatures draw soil water upward,

* Corresponding author. Tel.: +1-603-646-4100; fax: +1-603-646-4640.

E-mail address: shoop@crrel.usace.army.mil (S. Shoop).

 

0022-4898/02/$20.00 Published by Elsevier Ltd on behalf of ISTVS. doi:10.1016/S0022-4898(02)00019-8

 

182                                S. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194

 

forming ice lenses. Later, as surface temperatures rise, water from melting ice is trapped in the thawing layer by an impermeable frozen soil layer below. Additional water from snowmelt or precipitation can worsen conditions, as can low nightly temperatures that continue to draw soil moisture toward the surface. These condi- tions were evident during the deployment of US forces in Bosnia, where rapid sta- bilization of thawing soils was critical for the safe and timely movement of troops and supplies.

During March 1995, CRREL tested and evaluated a variety of expedient techni- ques suitable for stabilizing thawing ground. The field experiment, called Task Force Winter Thaw, was performed in cooperation with Fort McCoy, Wisconsin; the Wisconsin National Guard; the US Army Engineer School; the US Forest Service; and temporary-road product manufacturers Terramat and Uni-Mat International. Stabilization techniques were tested in three configurations: sloped sections with a 16–18% grade, a pentagonal-shaped road for cornering, and a level trail. We used our results to prepare a decision matrix for construction of expedient roads on thawing ground, which was referenced during winter deployment of US troops in Bosnia, 1996.

To supplement the data from Task Force Winter Thaw, the use of chemical sta- bilization methods for thawing soils was also evaluated. Two series of large scale tests were completed using CRREL’s Frost Effects Research Facility (FERF) during February and March of 1996, and laboratory tests were used to evaluate the pres- ence of organic materials in the soil on the effectiveness of the chemical stabilizers.

 

 

2. Task Force Winter Thaw

 

Task Force Winter Thaw evaluated a variety of mechanical methods to rapidly stabilize thawing soils. The techniques focused on methods that were new or uncommon in the military. Stabilized sections were constructed directly on the thawing surface. Stabilized surfaces consisted of chunkwood, tire chips, wood mats (Uni-Mats and pallets), tire mats (Terramats), fascines, tree slash, geosynthetics, and combinations thereof. Each test surface was trafficked with 100 passes using two vehicles: a HEMTT (Fig. 1), and an M60A3 tank (Fig. 2). The HEMTT (Heavy Expanded Mobility Tactical Truck) is a 30-ton 8×8, with tire pressures of 140 kPa (20 psi) on the front two axles and 690 kPa (100 psi) on the rear. The M60A3 is a main battle tank weighing 54 tons, with a ground pressure of 77 kPa (11.2 psi). To test the materials under the severe conditions caused by the turning action of a tracked vehicle, corner test sections were trafficked with the M60A3.

The test sections were evaluated during construction based on time and difficulty of construction, labor and equipment requirements, and material cost and avail- ability. During trafficking, surfaces were evaluated for rutting, lateral movement, material interference with vehicles, vehicle traction, handling, and ride quality. This information was gathered by assigning members of the Wisconsin National Guard to complete an evaluation survey during the construction and trafficking of each surface (approximately 30 individuals and over 100 surveys). The results were

 

1. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194 183

compiled by the CRREL project engineers as presented in Table 1. Additional details on the performance are given in Kestler et al. [1] and Shoop and Stark [2].

The use of chunkwood roads was initiated by the US Forest Service (USFS) to improve the utilization of unmerchantable timber [3]. Chunkwood consists of well graded, angular pieces of wood ranging from 1 to 20 cm length (and approximately 4 cm thick). The chuckwood test sections ranged from 30 to 45 cm fill thickness. The chunkwood sections exhibited good trafficability, excellent traction, and were a

 

Fig. 1. HEMTT crossing Unimats and Terramats on the level trail at Fort McCoy, Wisconsin.

 

Fig. 2. M60A3 tank on the chunkwood test section at Fort McCoy, Wisconsin.

 

Table 1

Guidelines for selecting rapid stabilization techniques for vehicle mobility on thawing, low bearing strength, ground; a gravel road is shown for comparison

 

strong, permeable, and lightweight alternative to gravel-fill roads. The US Forest Service reports chunkwood roads being used for over 10 years with no maintenance. Tire chips are produced by shredding old tires so that the pieces pass a 5-cm sieve. The tire chip fill was approximately 30 cm thick. Like chunkwood, the tire chips are a lightweight fill material having high permeability. They also have insulating prop- erties that reduce frost depth and can be used to recycle old tires. They have good trafficability and traction, but protruding bead and belt steel can puncture vehicle tires and the long-term environmental effects are unknown.

Six different geosynthetics were pretested by performing start, stop, and turning maneuvers with the M60A3 tank. Based on this initial evaluation, a geocomposite (double-sided geonet) and a high-strength geotextile (a nonwoven polypropylene reinforced with polyester fibers) were chosen for use on the test sections without a cover of fill material, as shown in Fig. 3. Geotextiles were also used as separators and reinforcement in combination with other stabilization techniques. Geosynthetics were unrolled, cut, and placed by hand. They were easy to install and improved trail conditions in nearly all cases, even when bare, as long as they covered the entire width of the trail. Fig. 4 shows the geosynthetics after trafficking by the HEMTT on the sloped sections. Covering the geosynthetic with a small amount of fill significantly decreased wear damage and tearing from the vehicle undercarriage or track, and enhance the geotextile performance.

Tree slash consists of branches and small trees (up to 8-cm diameter) laid directly on the travel way. The larger diameter branches or trunks were used to fill in ruts, and smaller branches were placed at 45⁰ to the roadway. More slash was added

 

Fig. 3. Geosynthetics and wood mat on sloped test sections at Fort McCoy, Wisconsin, before trafficking.

 

186                                S. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194

 

Fig. 4. Geosynthetic test sections after trafficking slope with HEMMTT.

 

during trafficking to replenish the surface as needed. Slash requires no special equipment or training but involves considerable manual labor for placement. It provided a good travel surface but was prone to catching in the vehicle track and undercarriage until it was packed down.

The tire mats used were a commercial product provided by Terramat, Corp., designed specifically for vehicle and equipment use on sensitive terrain (for forestry, oil and gas exploration, and construction). The mats consist of layers of tire tread and sidewalls, and were placed using chains and heavy machinery. The tire mats provided good traction in all conditions, and they were flexible and rugged.

Two types of wood mats were tested. One type was a relatively lightweight wood pallet constructed on-site and placed by hand or machine. Although many boards broke during trafficking, the pallets continued to provide traction and flotation. The second type of wood mat was a much larger and more rugged commercial product provided by Uni-Mat International. These were made of oak and require heavy equipment for proper placement. Uni-Mats were durable and reusable but were slippery on slopes. Uni-Mats were the only material tested known to withstand the extreme shear forces caused by the turning of tracked vehicles [4].

A fascine is a series of parallel pipes linked or chained together to create a flexible mat. It is usually used to fill a drainage ditch, allowing water to flow. The fascine was constructed with 10-cm-diameter PVC pipe, placed in low, wet areas, and cov- ered with mats or chunkwood to provide a good travel surface. Fascines were tested

 

1. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194 187

 

in two locations. One of the fascines performed as designed, but the other was pressed into the mud and became ineffective for drainage.

 

 

3. Chemical stabilization in cold

 

Although chemical stabilization methods are common for soils, their effectiveness is hampered by extremely wet soil and low temperatures. Therefore, the perfor- mance of techniques suitable for thawing conditions (wet, cold soil) needed to be evaluated. This was done through laboratory testing and large-scale experiments in CRREL’s Frost Effects Research Facility (FERF). The FERF test area was 42×7 m, where 30 cm of low plasticity silt (ML) was placed over existing materials. In the first set of tests (February 1996), the area was divided into 14 test sections, which were treated with combinations of portland cement, Rapid Set cement, hydrated lime, and quicklime. Sodium metasilicate, which chemically reacts with the soil creating secondary cementation [5], was also added to some sections. The chemicals were spread on the soil surface (Fig. 5), rototilled into the soil to a depth of 10–15 cm, and then compacted using a smooth drum roller. Air temperature was held at 5 ⁰C (soil at 7 ⁰C) and the test section strength was monitored with time using a Clegg impact tester. The water content of the test sections varied, and they were sorted into two groups: the wetter sections had an average water content of 26.7% (wet) and the drier sections had an average of 23.5% (moist). The sections were trafficked with 150 passes by a 5-ton truck after allowing 48 h for hardening (Fig. 6).

 

Fig. 5. Chemicals being placed on the chemical stabilization test sections in the Frost Effects Research Facility, prior to mixing and compaction.

 

188                                S. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194

 

The combinations containing cement were the most effective and are shown in Fig. 7. Complete details of the test program are reported in Stark and Afflcek [6], and results are summarized below:

 

• In the moist test sections, the Rapid Set cement hardened, or, ‘‘set-up,’’ in approximately 1 h and reached its ultimate strength in less than 1 day. The strength decreased after trafficking, possibly because of the breaking of the cement bonds. In the ‘‘wet’’ test sections, the soil treated with Rapid Set cement gained strength much more slowly. The strength of the Rapid Set cement section for both the wet and moist conditions were approximately the same after 6 days.
• For the moist soil conditions, the portland cement section with sodium metasilicate (PC+Meta) reached the same strength as the Rapid Set sections in 2 days and continued to gain strength throughout the test. The portland cement section with no additives took 4 days to reach the same strength as the Rapid Set sections ( 7).
• Lime (both quicklime and hydrated lime) modifies fine-grained soils through ion exchange on the clay This effectively increases the soil optimum moisture content and lowers the plasticity index, thus having an effect similar to drying the soil. However, the lime test sections in the FERF did not gain any strength, probably owing to the low temperatures, short curing time, small percentage of clay fines in the soil.

 

 

Fig. 6. Measuring rut depths after trafficking.

 

1. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194 189

Fig. 7. Strength gain of chemically treated soils as measured using the Clegg impact tester (meta=sodium metasilicate, PC=Portland Cement, CBR=California Bearing Ratio).

 

In March 1996, an additional test was conducted in the FERF to simulate the application of a chemical treatment on a thawing trail in the field. Before applying the treatment, the soil was cooled, watered, and trafficked. The chemical treatment was spread and mixed using a grader. The surface was then compacted. The entire area was treated as a single test section using a mixture of portland cement, Rapid Set cement, and polyacrylate. Air temperature was held at 3 ⁰C (soil at 6 ⁰C), strength was monitored with the Clegg impact hammer, and the surface was traf- ficked with 50 passes after 24 h. The following observations are based on this test:

 

• The surface was rutted by the tires behind the blade on the grader and by the forklift carrying the cement to the test section. It proved very difficult to get good mixing on the rutted surface, and the grader had a difficult time blading the wet soil. The grader’s scarifying teeth were eventually used to mix the cement and soil. In areas where the cement concentration was high, the soil gained considerable strength in 24 h, while areas with low cement concen- tration gained minimal strength.
• Polyacrylate absorbs soil moisture and was used to make the soil more friable and easier to mix. The absorbent will chemically break down with time and does not affect compaction.
• Even with the mixing difficulties, the test section was able to withstand five times more traffic, based on rut development, 24 h after

 

190                                S. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194

 

The final test on chemical stabilization was a laboratory study to assess the effec- tiveness of Rapid Set cement and portland cement stabilization in soils containing organic material. Organic soils contain acids that chemically bind the calcium ions, lowering their concentration in the pore water to levels below that required for cement to harden. To simulate soil mixes with organic material, mortar mixes were made with cement, sand, water, and various amounts of tartaric acid (0–10% tar- taric acid to cement, by weight) [7]. The strength was tested after 7 days. The sam- ples made with Rapid Set gained approximately two-thirds of their (no organics) strength as long as the organic content was 4% or less. The mixtures of portland cement mortar did not gain any strength when 1% or more organic material was present.

 

 

4. Operation Joint Endeavor, Bosnia, winter 1996

 

• Winter conditions

 

The area of central and northern Bosnia is primarily forest and agricultural land, where the weather is comfortably warm in summer, and cold and wet in winter. Much of the soils of lowland Croatia and the valleys in northern Bosnia are specified as pseudogleys [8], which are a sandy silty loam with a clay layer at 38- to 41-cm depth (near Tuzla). Surface water stagnates above the claylayer during the winter and frequently seeps to the surface in low areas. Because of the saturated conditions, the pseudogleys tend to remain cold longer than well-drained soils that may be nearby. The pseudogleys dry out in May, and they subsequently wet and dry very quickly after summer rains because of evaporation; thus, there may be brief muddy periods alternating with dust problems in the summer. Meadows and farm fields have a 15- to 25-cm humus layer containing organics that may affect chemical stabilization.

The winter soil water and temperature regime in central Bosnia, Croatia, and Herzegovina have diurnal cycles, with low temperatures drawing water up to the surface frost at night, and higher temperatures thawing the ground frost during the day. This water migration, in combination with winter precipitation, produces muddy soils in the winter and difficult travel conditions on unsurfaced roads, trails, and on agricultural land, as shown in Fig. 8.

 

• Frost predictions

 

To estimate the extent of the frost or thaw problem, daily soil frost thickness was simulated using the MIDwest FROst CALculator (MIDFROCAL) model [9]. The model increments daily frost thickness by computing freeze or melt using daily maximum and minimum air temperature, precipitation (liquid equivalent), snow depth, and computed estimates of soil moisture. Soil characteristics used in the model were for alluvial or podzolic soils, depending upon location.

Frost was simulated using climate data for six locations nearby, using as many years as available for each location (24 years or fewer for each), for two different

 

1. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194 191

Fig. 8. US Navy Seabees use a forklift to place crates of tents in a muddy field in Croatia, January 1996.

DoD photo by Sgt. Larry D. Aaron.

 

Fig. 9. Results from the soil frost model for Slavonskiy Brod. Each line represents the frost depth predicted based on 1 year of historical climate data.

 

192                                S. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194

 

surfaces: bare soil and snow-covered soil. Bare soil simulations represent roads and parking areas regularly cleared of snow. Snow-covered soil represents fields that are not plowed of snow. In snow-cover cases, actual snow measurements were used, if available.

Results indicated that frost is generally not deep in Bosnia because temperatures do not stay extremely low for long periods. An example of frost predictions for Slavonsky Brod is shown in Fig. 9. Frost thickness of 5–10 cm are common, with rare events reaching 20–30 cm. Frost is often sporadic, with freezing lasting for days to weeks, followed by a thaw that melts all frost from the soil. Then frost may occur again during the next cold spell. By 4 March there is about a 20% chance of frost still being in the soil, or recurring, at snow-free locations in Bosnia and southern Hungary. For locations with transitory seasonal snow cover, there is about a 30% chance of frost after 4 March.

 

• Recommendations

 

Based on experimental programs in Wisconsin and the FERF, predictions of conditions in Bosnia, and engineering judgment, suggested methods for stabilization of thawing ground are given in Table 2. These suggestions were geared toward the

 

Table 2

Suggested methods to stabilize thawing ground for different traffic requirements

Parking and work areas

Mats (Unimats, plastic mats, landing mats, etc.) on a nonwoven geotextile separator (to aid drainage, prevent sinking of mat, prevent mud from coming up through cracks and aid in subsequent removal).

Foot traffic and living areas

Lightweight mats over geotextile for separation. Elevate tents on mats, gravel, and stilts.

Localized trouble spots

Mats or reinforcement geosynthetic with gravel. Geosynthetics can also be used to wrap granular or in-situ material to provide lateral support.

Bridge site egress/ingress

Tire mats (tire sidewalls fastened together and layered with treads) provide vehicle flotation and traction even when wet or on slopes. Can be used alone or with geosynthetics.

Main supply routes and base camp roads [10]

—using standard geotextile design [11] and construction survivability [12] guidelines

For soils of CBR <1 deeper than 25 cm, use geosynthetic reinforcement covered with 45 cm of gravel or crushed rock.

For soils of CBR between 1 and 3; or for soils with CBR <1 with a shallow bearing layer, a geotextile separator with 45 cm of fill is required. An alternative is to use reinforcement geotextile or geogrid covered with 25 cm of gravel or crushed rock

Chemical methods

Chemicals can be mixed with and strengthen the soil for any of these situations. Rapid Set cement provides the fastest strength gain at low temperatures (just above freezing). The rate of strength gain for portland cement is reduced at low temperatures, but additives like sodium metasilicate can reduce the curing time and increase ultimate strength. When organics soils are present, Rapid Set has reduced benefits (2/3 the normal strength) and portland cement will not harden

 

1. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194 193

 

deployment of troops in Bosnia and, therefore, are biased toward the materials available and the requirements of the situation. Methods used at Fort McCoy are most suitable to immediate needs or localized areas. The use of the geotextiles with gravel was considered primarily a method for road building, rather than rapid sta- bilization, and is considered standard practice [10–12].

In addition to these suggestions, modifications to standard operating procedures were also suggested, such as lowering tire pressures (to off-road specifications) to reduce rutting and promote healing of damaged surfaces [13], and traveling at night when ground is frozen. Special consideration should also be given to improving the site drainage. Suggestions and experiences from personnel in Bosnia were subsequently assembled in two handbooks: Beat the Mud and Keep the Convoys Rolling, prepared by the US Army Engineer Waterways Experiment Station and CRREL [14].

 

 

5. Summary

 

Several experiments on soil stabilization techniques suitable for thawing soils, and the results of these tests, are summarized here. Mechanical techniques for rapid stabilization were tested at Fort McCoy in March 1995. Chemical stabilization techniques suitable for thawing ground were evaluated in the laboratory and in large-scale tests during the winter of 1996. The experience from these test programs was useful for US troop deployment in Bosnia during the winter of 1996 but may also be applied to forestry, mining, oil, construction, and agriculture operations, where it is necessary to travel on ground that is thawing, contains excess water, or has low bearing capacity.

 

 

Acknowledgements

 

The field experiment, called Task Force Winter Thaw, was performed in coop- eration with Fort McCoy, Wisconsin; the Wisconsin National Guard; the US Army Engineer School; the US Forest Service; and temporary-road product manufacturers Terramat and Uni-Mat International. Expertise provided by Karen Henry of CRREL was critical to the success of the McCoy test program and subsequent results presented in this paper. This paper was presented at the 7th European ISTVS Conf., Ferrara, Italy, Oct. 1997.

 

References

 

• Kestler MA, Shoop SA, Henry KS, Stark JA, Affleck RT. Rapid stabilization of thawing soils for enhanced vehicle mobility: a field demonstration project USA Cold Regions Research and Engi- neering Laboratory, CRREL report 99-3, 1999.
• Shoop SA, Stark JA, Stabilizing thawing soils for improved vehicle mobility, T96011. USA Cold Regions Research and Engineering Laboratory,1996.
• Arola R, Hodek RJ, Bowman JK, Schulze Forest roads built with chunkwood. In: Chunkwood:

 

194                                S. Shoop et al. / Journal of Terramechanics 39 (2003) 181–194

 

production, characterization and utilization. North Central Experiment Station, USDA Forest Service General Technical report HC-145, St. Paul, MN, June 1991.

• Poyer UNI-MAT video. Houston, (TX, USA): UNI-MAT International 1995.
• Moh Z. Soil Stabilization with cement and sodium additives. Journal of the Soil Mechanics and Foundations Division, Proceedings of the American Society of Civil Engineers 1962;88(SM6):81–
• Stark J, Affleck R. Chemical stabilizers for rapid stabilization of thawing soils. USA cold regions research and engineering laboratory. CRREL Internal report 1996.
• Clare KE, Sherwood PT. Further studies on the effect of organic matter on the setting of soil-cement App Chem 1956;6:317–24.
• Jelenic Soils of Yugoslavia. Beograd: Jugoslovensko Drustvo Za Proucavanje Zemljista; 1969.
• Ryerson C. A moisture-based model for calculating daily changes in seasonal soil frost depth. Publications in Climatology 1979;32(2):1–115.
• Henry K. Personal communication. Hanover (NH): USA Cold Regions Research and Engineering Laboratory; 1996.
• Giroud JP, Noiray L. Geotextile reinforded unpaved roads. Journal of the Geotechnical Engineering Division, American Society of Civil Engineers 1981;107(GT 9):1233–54.
• Guide specifications and test procedures for geotextiles. Task Force 25 report. Washing- ton (DC): Subcommittee on New Highway Materials, American Association of State Transportation and Highway Officials, 1990.
• Kestler MA, Bert RL, Nam SI, Smith C. Reduced tire pressure: review of an alternate road usage technique for reducing springtime damage on low volume roads. ASCE Cold Regions Engineering Conference, Anchorage, AK, May 2002.
• Beat the mud and keep the convoys rolling. Vicksburg (MS): US Army Engineer Waterways Experiment Station; 1996.