PLAXIS is a geotechnical finite element program, in 2D and 3D versions, that models soil as a continuous material with realistic stress-dependent behavior, together with the structures in and on it, the groundwater, and the sequence in which everything gets built. It is a powerful tool and I use it regularly. It is also frequently the wrong tool, and an honest engineer will tell you when you do not need it, because a full finite element analysis (FEA) on a problem that does not call for one is a way to spend your money making a simple answer look complicated.
When you do not need it
Most routine geotechnical questions have been solved with closed-form methods, design charts, and limit-equilibrium analysis for decades, and those methods remain entirely appropriate. Bearing capacity of a spread footing, settlement of a simple foundation on reasonably uniform soil, the stability of a straightforward slope: these are well served by established hand methods and dedicated slope-stability tools. Using finite element analysis on any of them adds cost and the appearance of rigor without adding real understanding. If your problem fits in a chart, use the chart.
When you do need it
PLAXIS is worth its cost when the problem has interactions that simpler methods cannot represent honestly.
The clearest case is genuine soil-structure interaction: a deep excavation with bracing or tieback anchors, a tunnel, a complex retaining system, an integral abutment, anything where the soil and the structure push on each other and the answer depends on their combined stiffness. The second case is deformation. Limit-equilibrium analysis gives you a factor of safety against failure, but it cannot tell you how much something will move. When the design is controlled by tolerable settlement or wall deflection rather than by collapse, you need a method that computes displacements, and that requires finite element analysis. Staged and time-dependent problems are the third: an embankment built in lifts over soft ground, consolidation settlement developing over months or years, loads applied in a sequence that changes the result. And slope problems with complicated geometry, layered strata, pore pressures, or reinforcement are often beyond what a limit-equilibrium model can represent, and need a continuum analysis.
The part that separates good analysis from confident nonsense
The danger of a tool like this is that it will always produce a result, beautifully rendered, whether or not your inputs deserve to be believed. The skill is in the soil constitutive model and its parameters. A simple Mohr-Coulomb model is fine for some problems and badly misleading for others, where a Hardening Soil model or another advanced formulation is needed to capture how stiffness changes with stress and strain. Choosing the model and justifying the parameters, ideally against the actual soil data, is the engineering. The software is just the calculator.
How this fits with the other analysis tools
PLAXIS is general-purpose geotechnical finite element analysis, which is a different thing from CANDE, the specialized finite element tool for buried structures. The two rest on the same physics and almost never apply to the same problem, and part of scoping a job correctly is knowing which one the problem actually requires. It also pairs naturally with the geotechnical work I have written about elsewhere: you might model improved ground directly to predict its settlement, or run a continuum analysis to corroborate a laterally loaded shaft result in unusual soil.
If someone has told you your project needs finite element geotechnical analysis, it is worth confirming that it really does. Sometimes the answer is yes and it is exactly the right investment. Sometimes the answer is that a chart and an afternoon will do, and I will tell you so. Either way, this falls under my geotechnical finite element services.