当浸水面积较小或饱和自重压力小于其湿陷起始压力时，黄土自重湿陷变形无法发生，其大孔隙结构得以保留，在长期的水力共同作用下，强度发生显著衰减，对地下工程的长期稳定性造成影响。为深入研究原位地层应力、实际水分渗流补给状态及不同浸水持续时间下原状饱和黄土试样的真实力学行为，开展了1组约415 d的现场试坑浸水试验，综合采用室内试验和原位旁压试验手段，讨论了长期浸水期间地层原位水平应力的改变，从剪切特性、压缩特征和原位强度和变形指标等角度研究了黄土的软化力学行为及其与浸水持续时间的对应关系。研究表明：（1）在长期浸水期间，黄土逐步由可塑状态向软塑甚至流塑状态演化，形成了具有大孔隙比、高含水率、中或高压缩性和低强度的欠压密饱和黄土。（2）在持续浸水期间，Qeol3黄土和Qel3古土壤的原位水平应力明显增加，深层Qeol2黄土的原位水平应力受到地层拱效应的影响，未发生明显改变。（3）黄土浸水软化存在时效特征，室内短期增湿试样无法反映土体结构强度缓慢弱化过程；当持续浸水约120 d后，各类强度和变形指标基本稳定。（4）随着浸水持续时间的延长，试样应力−应变曲线中弹性段逐步缩短，各类强度指标（无侧限抗压强度、三轴剪切峰值强度和原位旁压承载特征值）均发生明显降低；压缩系数随荷载增加呈现指数形态，即压缩系数峰值区间提前，数值大幅度增加；原位旁压模量和基床系数均发生大幅度降低。（5）在衰减稳定阶段，地层不均匀性始终存在，饱和Qeol3黄土和饱和Qeol2黄土的无侧限抗压强度为Qel3古土壤的21%～32%，地基承载力约为Qel3古土壤的47%；饱和Qeol2黄土和饱和Qeol3黄土的水平基床系数Kx分别约为古土壤的67%和28%。When theimmersion area is small or the self-weight is less than its collapsiblepressure

unsaturated loess does not collapse and its large pore structure canbe retained. However

long-term hydraulic action significantly reduces the strengthof saturated loess

potentially influencing the long-term stability of underground engineeringprojects. A field immersion test lasting 415days was conducted to collect undisturbedsaturated loess samples under various immersion duration

maintaining originalin-situ stress and actual water infiltration conditions. Indoor and in-situtests were conducted toexamine changes in in-situ lateralstress during long-term immersion. The relationship between the softeningmechanical behavior of loess and immersion duration was studied in terms of shear characteristics

compressive properties

and in-situ strength and deformation indicators. Resultsshow that: (1) During prolonged immersion

loess transitions from a plastic toa soft or even flow plastic state

forming under-compacted saturated loess witha high void ratio

high water content

medium or high compressibility

and lowstrength. (2) During prolonged immersion

the in-situ horizontal stress of Qeol3 loess and Qel3 paleosol significantly increase

while thedeep Qeol2 loess’s in-situ horizontal stress remainsstable due to the strata’s arch effect. (3) Loess softening is time-dependent

and short-term indoor humidification samples are insufficient to show thegradual weakening of soil structure. After 120 days of continuous immersion

various strength and deformation indicators stabilize. (4) As immersion durationincreases

the elastic segment of the stress-strain curve gradually shortens.Strength indicators

including unconfined compressive strength

peak shear strength

and in-situhorizontal bearing value

decrease significantly. The compression coefficientshows an exponential trend with increasing load

resulting in a significant increase in valueand advancement of the peak interval. Additionally

both in-situ lateralpressure modulus and subgrade coefficient undergo substantial reduction. (5) Inthe stable stage

formation heterogeneity always exists. The unconfinedcompressive strength of saturated Qeol3 loess and Qeol2 loess is approximately 21%−32% of Qel3 paleosol

and the bearing capacity is 47% of Qel3 paleosol. The horizontal bedding coefficient Kx of saturated Qeol2 loess is 67% of Qel3 paleosol

while that for saturated Qeol3 loess is 28%.