I am simulating the process of pipe burst induced groundwater flow in an unsaturated slope (see simplified sketch in Figure 1 below). According to the according to the k-function curve, a lower limit of soil suction will make the soil more permeable at high suction ranges (Figure 2), and should lead to faster groundwater flow and thus result in larger wetting extent. This is indeed validated in the SEEP/W software. However, I recently found that when the upper limit of soil suction is reduced from 100m to 1m, the size of wetting extent in the slope is reduced (see Figures 3 to 5). It looks a bit odd to me. Can anyone help with solving this probem? Thanks.
Sorry that I may made a mistake. I had a double check and found that smiliar results seemed to be obtained in SEEP/W: an increase of suction cap generally leads to greater wetting extent, while in some cases the opposite may emerge (not sure if it was due to numerical error). I was advised that suction cap is the pressure of cavitation for pore water, i.e., water becomes air bubbles, and thus the permeability to water will drop drastically. However, I am still not clear how the permeability of soils with suction higher than the capped value is determined in Plaxis. Seems that it does not follow the k function curve which is cut off at the suction cap.
Seems that I have omitted the impact of suction loss on the hydraulic gradient which is a key point of the Darcy's law: q = k*i. When the suction cap value is reduced, the hydraulic gradient is also decreased near pipe burst point, and therefore leads to slower groundwater flow. The change of flow rate due to reduction of suction cap is a trade-off evaluation between the increased soil permeability and reduced hydraulic gradient. It seems that the hydraulic gradient plays a more dominated role in the flow rate calculation with varied suction cap in my case. Any other point of view is welcome for discussion.