PAVEMENT PROBLEMS CAUSED

BY COLLAPSIBLE SUBGRADES

By Sandra L. Houston,1 Associate Member, ASCE

(Reviewed by the Highway Division)

ABSTRACT: Problem subgrade materials consisting of collapsible soils are com-

mon in arid environments, which have climatic conditions and depositional and

weathering processes favorable to their formation. Included herein is a discussion

of predictive techniques that use commonly available laboratory equipment and

testing methods for obtaining reliable estimates of the volume change for these

problem soils. A method for predicting relevant stresses and corresponding collapse

strains for typical pavement subgrades is presented. Relatively simple methods of

evaluating potential volume change, based on results of familiar laboratory tests,

are used.

INTRODUCTION

When a soil is given free access to water, it may decrease in volume,

increase in volume, or do nothing. A soil that increases in volume is called

a swelling or expansive soil, and a soil that decreases in volume is called a

collapsible soil. The amount of volume change that occurs depends on the

soil type and structure, the initial soil density, the imposed stress state, and

the degree and extent of wetting. Subgrade materials comprised of soils that

change volume upon wetting have caused distress to highways since the be-

ginning of the professional practice and have cost many millions of dollars

in roadway repairs. The prediction of the volume changes that may occur in

the field is the first step in making an economic decision for dealing with

these problem subgrade materials.

Each project will have different design considerations, economic con-

straints, and risk factors that will have to be taken into account. However,

with a reliable method for making volume change predictions, the best design

relative to the subgrade soils becomes a matter of economic comparison, and

a much more rational design approach may be made. For example, typical

techniques for dealing with expansive clays include: (1) In situ treatments

with substances such as lime, cement, or fly-ash; (2) seepage barriers and/

or drainage systems; or (3) a computing of the serviceability loss and a mod-

ification of the design to "accept" the anticipated expansion. In order to make

the most economical decision, the amount of volume change (especially non-

uniform volume change) must be accurately estimated, and the degree of road

roughness evaluated from these data. Similarly, alternative design techniques

are available for any roadway problem.

The emphasis here will be placed on presenting economical and simple

methods for: (1) Determining whether the subgrade materials are collapsible;

and (2) estimating the amount of volume change that is likely to occur in the

'Asst. Prof., Ctr. for Advanced Res. in Transp., Arizona State Univ., Tempe, AZ

85287.

Note. Discussion open until April 1, 1989. To extend the closing date one month,

a written request must be filed with the ASCE Manager of Journals. The manuscript

for this paper was submitted for review and possible publication on February 3, 1988.

This paper is part of the Journal of Transportation.Engineering, Vol. 114, No. 6,

November, 1988. ASCE, ISSN 0733-947X/88/0006-0673/$1.00 + $.15 per page.

Paper No. 22902.

673

field for the collapsible soils. Then this information will place the engineer

in a position to make a rational design decision. Collapsible soils are fre-

quently encountered in an arid climate. The depositional process and for-

mation of these soils, and methods for identification and evaluation of the

amount of volume change that may occur, will be discussed in the following

sections.

COLLAPSIBLE SOILS

Formation of Collapsible Soils

Collapsible soils have high void ratios and low densities and are typically

cohesionless or only slightly cohesive. In an arid climate, evaporation greatly

exceeds rainfall. Consequently, only the near-surface soils become wetted

from normal rainfall. It is the combination of the depositional process and

the climate conditions that leads to the formation of the collapsible soil.

Although collapsible soils exist in nondesert regions, the dry environment in

which evaporation exceeds precipitation is very favorable for the formation

of the collapsible structure.

As the soil dries by evaporation, capillary tension causes the remaining

water to withdraw into the soil grain interfaces, bringing with it soluble salts,

clay, and silt particles. As the soil continues to dry, these salts, clays, and

silts come out of solution, and "tack-weld" the larger grains together. This

leads to a soil structure that has high apparent strength at its low, natural

water content. However, collapse of the "cemented" structure may occur

upon wetting because the bonding material weakens and softens, and the soil

is unstable at any stress level that exceeds that at which the soil had been

previously wetted. Thus, if the amount of water made available to the soil

is increased above that which naturally exists, collapse can occur at fairly

low levels of stress, equivalent only to overburden soil pressure. Additional

loads, such as traffic loading or the presence of a bridge structure, add to

the collapse, especially of shallow collapsible soil. The triggering mechanism

for collapse, however, is the addition of water.

Highway Problems Resulting from Collapsible Soils

Nonuniform collapse can result from either a nonhomogeneous subgrade

deposit in which differing degrees of collapse potential exist and/or from

nonuniform wetting of subgrade materials. When differential collapse of

subgrade soils occurs, the result is a rough, wavy surface, and potentially

many miles of extensively damaged highway. There have been several re-

ported cases for which differential collapse has been cited as the cause of

roadway or highway bridge distress. A few of these in the Arizona and New

Mexico region include sections of 1-10 near Benson, Arizona, and sections

of 1-25 in the vicinity of Algadonas, New Mexico (Lovelace et al. 1982;

Russman 1987). In addition to the excessive waviness of the roadway sur-

face, bridge foundations failures, such as the Steins Pass Highway bridge,

1-10, in Arizona, have frequently been identified with collapse of foundation

soils.

Identification of Collapsible Soils

There have been many techniques proposed for identifying a collapsible

soil problem. These methods range from qualitative index tests conducted on

674

disturbed samples, to response to wetting tests conducted on relatively un-

disturbed samples, to in situ meausrement techniques. In all cases, the en-

gineer must first know if the soils may become wetted to a water content

above their natural moisture state, and if so, what the extent of the potential

wetted zone will be. Most methods for identifying collapsible soils are only

qualitative in nature, providing no information on the magnitude of the col-

lapse strain potential. These qualitative methods are based on various func-

tions of dry density, moisture content, void ratio, specific gravity, and At-

terberg limits.

In situ measurement methods appear promising in some cases, in that many

researchers feel that sample disturbance is greatly reduced, and that a more

nearly quantitative measure of collapse potential is obtainable. However,

in situ test methods for collapsible soils typically suffer from the deficien-

cy of an unknown extent and degree of wetting during the field test. This

makes a quantitative measurement difficult because the zone of material

being influenced is not well-known, and, therefore, the actual strains, in-

duced by the addition of stress and water, are not well-known. In addition,

the degree of saturation achieved in the field test is variable and usually

unknown.

Based on recently conducted research, it appears that the most reliable

method for identifying a collapsible soil problem is to obtain the best quality

undisturbed sample possible and to subject this sample to a response to wet-

ting test in the laboratory. The results of a simple oedometer test will indicate

whether the soil is collapsible and, at the same time, give a direct measure

of the amount of collapse strain potential that may occur in the field. Potential

problems associated with the direct sampling method include sample distur-

bance and the possibility that the degree of saturation achieved in the field

will be less than that achieved in the laboratory test.

The quality of an undisturbed sample is related most strongly to the area

ratio of the tube that is used for sample collection. The area ratio is a measure

of the ratio of the cross-sectional area of the sample collected to the cross-

sectional area of the sample tube. A thin-walled tube sampler by definition

has an area ratio of about 10-15%. Although undisturbed samples are best

obtained through the use of thin-walled tube samplers, it frequently occurs

that these stiff, cemented collapsible soils, especially those containing gravel,

cannot be sampled unless a tube with a much thicker wall is used. Samplers

having an area ratio as great as 56% are commonly used for Arizona col-

lapsible soils. Further, it may take considerable hammering of the tube to

drive the sample. The result is, of course, some degree of sample distur-

bance, broken.bonds, densification, and a correspondingly reduced collapse

measured upon laboratory testing. However, for collapsible soils, which are

compressive by definition, the insertion of the sample tube leads to local

shear failure at the base of the cutting edge, and, therefore, there is less

sample disturbance than would be expected for soils that exhibit general shear

failure (i.e., saturated clays or dilative soils). Results of an ongoing study

of sample disturbance for collapsible soils indicate that block samples some-

times exhibit somewhat higher collapse strains compared to thick-walled tube

samples. Block samples are usually assumed to be the very best obtainable

undisturbed samples, although they are frequently difficult-to-impossible to

obtain, especially at substantial depths. The overall effect of sample distur-

bance is a slight underestimate of the collapse potential for the soil.

675

译文：

湿陷性地基引起的路面问题

作者：...

摘要：在干旱环境中，湿陷性土壤组成的路基材料是很常见的，干旱环境中的气候条件、沉积以及风化作用都有利于湿陷性土的形成。在这方面包括了一种使用常用的实验室设备和测试方法获得这些问题的土壤的体积变化的可靠估计的预测技性讨论。对典型的路面路基提供了一种方法去预测相关的应力和相应的湿陷张力。基于熟悉的实验室测试结果，使用相对简单的方法评估潜在体积的变化。

引言：当土壤接触到水的时候，可能体积会减小或扩大，也可能不变化。遇到水体积增大的土叫做膨胀土，而体积减小的称为湿陷性土。土壤的类型结构、最初的土壤密度、施加应力状态以及土壤浸湿的程度范围决定了体积变化量的大小。自从专业实践开始由这些遇水体积变化的土组成的路基材料已经导致了许多公路病害，并且在维修方面已经花费了数百万美元。处理这种路基材料做出经济决策的第一步是做出可能发生的体积变化的预测。

每个工程项目都有不同的设计考虑、经济限制和风险因素，所有这些情况都必须考虑到。然而，最好的和最合理的设计可能会具有更大的经济优势相比于可靠的体积变化预测。例如，典型的处理膨胀黏土的技术有：(1)在现场用例如石灰、粉煤灰或者水泥等处置处理；(2)设置渗流屏障或者排水设施；(3)进行适用性散失的计算来变更设计来接受预期膨胀。为了做出最经济的决定，体积变化（特别是不均匀的体积变化）的量必须要精确计算，并且要从计算出的数据上估测出路面的平整度。同样，不寻常的设计技术可利用到任何道路问题中。 这里将重点对以下两点提供简单和经济的方法：(1)决定路基材料是否是湿陷性膨胀性或者其他;(2)估算湿陷性土在路基中极有可能发生的体积变化量。这些信息将会是工程师做出合理的决定。湿陷性土在干旱地区是非常常见的。这种土的形成过程以及计算可能发生的体积变化量将在下文中介绍。

美国亚利桑那州皇家经济学会高级助理教授Tempe

注：讨论开放至1989年4月1日。增加截止日期一个月，必须要有ASCE期刊经理批准的书面请求。这篇文章是提交复审的初稿，可能出版的时间在1988年2月3日。本文是运输杂志收录的的一篇文章。114工程卷，6号，1988年11月。ASCE，ISSN 0733-947x \ / \ / \ / 88 0006-0673 1美元+每页15美元。22902号文件

湿陷性土

湿陷性土的结构

湿陷性土有高孔隙率、低密度和较弱的黏性等特点。在干旱地区，有很高的蒸发量，而降水量较低。因此，当有降水时只有地面土壤湿润。沉积作用和气候条件共同造成了湿陷性土的形成。尽管湿陷性土存在于非沙漠地区，但干旱环境中蒸发量远超降水量这一特点非常有利于湿陷性土结构的形成。

当土壤在蒸发过程中变干后，毛细张力使其余的水进入土壤颗粒的界面，同时带出可溶性盐、粘土和粉砂颗粒。随着土壤继续变干，可溶性盐、黏土和粉砂颗粒逐渐从溶解状态脱离出来，大量的颗粒物聚集在一起。这就导致这种土壤在低含水量时具有较高的表面强度。然而，当遇到水时，由于结合材料的弱化和软化，土壤承受应力超过浸湿之前，会使土结构发生崩塌。这样，如果提供给土壤水量高于自然状态水量，可能在较低水平的压力时就发生崩溃，或许

就在上覆土压力作用下。额外的负荷，如交通荷载或桥梁结构的存在，增加了湿陷性，特别是对于浅层土。无论怎样，触发湿陷性的原因就是加入水。

湿陷性土引起的公路问题

不均匀的沉陷可能是因为地基矿床存在不同程度的不均匀性或者是地基材料湿度不一样。当路基土发生微分崩溃时，结果是一个粗糙的、波浪状的表面，并潜在存在许多英里路基的广泛灾害。已经有一些报道，微分崩溃已被引用作为道路或公路桥梁病害的原因。其中一些在亚利桑那州和新墨西哥州地区包括靠近本森,亚利桑那州1 - 10,部分和新墨西哥附近Algadonas的1 - 25部分。除了道路表面的过度波动，桥梁基础的问题，比如在亚利桑那州斯坦通公路桥梁，其他的经常被确定地基土的崩溃。

鉴别湿陷性土 已经有许多技术，提出了鉴别湿陷性土的问题。这些方法的范围从干扰样品进行质量指标的测试到比较浸湿前后土的性状再到现场观测技术。在所有的情况下，工程师首先必须要知道是否被浸湿的土壤含水量在天然含水量之上，如果是，那么就要确定潜在的浸湿范围。大部分鉴定湿陷性土的方法在本质上都是定性的，没有提供潜在崩塌规模的大小。这些定性的方法是基干密度、水分含量、空隙率、比重和阿太堡界限之上的。

原位检测出现在某些较有前途的研究中，因为许多专家认为样品干扰大大减少，而定量检测更能得到潜在的崩塌结论。然而，湿陷性土原位测试方法在现场测试时通常遭受润湿分布范围和程度未知这一问题的困扰。由于该区域材料材料以及水和应力的影响是未知的，使得定量检测难以进行。此外，在现场试验取得的饱和度是变化的，通常也不能确定。

根据最近的研究，表明鉴别湿陷性土的最可靠地方法是在试验室中观测最优质的原状样品在接触到水时的反应。简单的土压缩试验结果将表明土壤是否是湿陷性的，与此同时，还能得到这些区域潜在湿陷应力的直接测量值。直接测量的方法存在的现在的问题包括样品干扰以及测到的饱和度可能低于实验室测得的。

未受干扰的样品质量是与收集样品的管的面积比有很大关系的。面积比是收集到的样品横截面积与样品管横截面积的比。根据定义，A型薄壁样品管具有10%-15%的面积比率。虽然最好通过薄壁管来获得原状样品，但实际情况下样品很容易发生僵硬、胶结，尤其是封闭的砂砾，所以通常情况用较厚的管来采样。亚利桑那州的土壤通常有56%的面积比。另外，可能需要相当大的锤击该管一驱动样品。当然，其结果是，样品一定程度的干扰、断键、致密化，并相应的减小实验室测量时的崩溃。然而，符合定义的湿陷性土，样品管插入时导致局部剪切破坏，比一般剪切的土壤样品量更少干扰失败。