---

This paper describes a series of laboratory model tests on strip footings supported on unreinforced and geogrid-reinforced sand
with an inside void. The footing is subjected to a combination of static and cyclic loading. The influence of various parameters
including the embedment depth of the void, the number of reinforcement layers, and the amplitude of cyclic load were studied.
The results show that the footing settlement due to repeated loading increased when the void existed in the failure zone of the
footing and decreased with increasing the void vertical distance from the footing bottom and with increasing the reinforcement
layers beneath the footing. For a specified amplitude of repeated load, the footing settlement is comparable for reinforced sand,
thicker soil layer over the void and much improved the settlement of unreinforced sand without void. In general, the results
indicate that, the reinforced soil-footing system with sufficient geogride-reinforcement and void embedment depth behaves much
stiffer and thus carries greater loading with lower settlement compared with unreinforced soil in the absent of void and can
eliminate the adverse effect of the void on the footing behavior. The final footing settlement under repeated cyclic loading becomes
about 4 times with respect to the footing settlement under static loading at the same magnitude of load applied.

---

When geogrid reinforcement is used as a treatment method for improving soft subgrade as a roadway foundation, a top layer of subgrade is usually excavated and backfilled with geogrid-reinforced aggregates. This treatment method produces an adequate platform for the planned roadway construction site, where heavy traffic loading is constantly moving. This paper presents a quantitative assessment of subgrade improvement by geogrid reinforcement based on numerical modelling and parametric studies. First of all, the preliminary numerical models were verified by comparing the analysis results with previous studies. Secondly, the major numerical models in this study were assumed to be a simplified simulation of a geogrid-reinforced two-layer system with an aggregate layer above a subgrade layer. The numerical models were applied a quasi-static loading and unloading cycle, in order to monitor the permanent deformation at the surface of the models. Afterwards, thickness of aggregate layer, and subgrade CBR values were varied in order to summarize the outcomes of each case. This approach makes it possible to quantify the effects of geogrid reinforcement and aggregate material in terms of an enhanced California Bearing Ratio (CBR) of a single subgrade clay layer. Results have shown that when the aggregate thickness is up to 450mm, the contribution of enhanced CBR is mostly from aggregate material. However, when the aggregate thickness is about 150mm with a relatively weak subgrade material, the inclusion of geogrid material can contribute about 50% of the enhanced value.