Diaphragm walls are rectangular shaped cast in place deep foundations. There are two critical phenomena occurring, according to which the final quality can be affected: bentonite suspension exfiltration and concrete placement. Some imperfections seem to appear recurrently on the surface of the final wall. The defects are known as shadowing pathologies. The main reasons can be attributed to the dual effect of exfiltration mechanisms and kinematics of concrete flow. The objective of this study is developing a numerical tool to prevent the appearance of shadowing pathologies by visualizing the concrete flow in the presence of a bentonite suspension. This paper presents the results obtained from 2D and 3D models of diaphragm wall construction using COMSOL Multiphysics. The CFD model helped in solving a multifluid and particularly a two-phase flow. The 2D modeling has considered a fresh slurry and an exfiltrated (or polluted) suspension neighboring soil and followed concrete flow with two rheological behaviors in two reinforcement configurations. Then, 3D simulations were compared to actual experimentation results, which were undertaken to construct diaphragm walls in the laboratory. By comparing the results of the simulations to the experimental outcomes, it has been possible to validate the model. The resulting simulations could clearly explain the occurrence of the pathology where the flow pattern and volume fraction of the fluid flow were determined. From the results obtained, it can be conducted that a compliant concrete mix but at the lower limit for the consistency recommendations, leads to pathologies, just like a polluted slurry.

Due to high reactivity and relatively low cost, nano zero-valent iron (nZVI) has become an alternative material for in-situ remediation of contaminated sites. However, factors such as short transport distance and easy deposition in porous media also seriously restrict its injection remediation effect. The optimum ratio of bentonite and kaolin supported nano zero-valent iron (K-nZVI) in the remediation agent was determined by sedimentation and rheological tests. The transport characteristics of deionized water and bentonite suspensions carrying K-nZVI in porous media under different injection pressures were investigated using simulating column tests. The results show that bentonite suspensions could significantly improve the stability and dispersibility of K-nZVI. The proportion of bentonite and K-nZVI are 5% and 0.4%, respectively, which is the best ratio of the remediation agent. The transport capability of K-nZVI carried by deionized water increases with the increase of injection pressure, while there is a critical injection pressure for bentonite suspensions carrying K-nZVI remediation agent. The numerical simulation results show that the diffusion radius of K-nZVI is positively correlated with the injection pressure and negatively correlated with the viscosity of the remediation agent. The results provide theoretical guidance for the remediation project of heavy metal pollution in non-ferrous smelting sites.