Introduction
Prokon continues to enhance SUMO‘s functionality, steadily turning it into a more powerful analysis platform. A notable recent feature is the ability to convert Plane Shells into Solid Meshes. I often use this capability when dealing with slab-to-column or slab-to-beam interactions. This raised a question: what if a slab on elastic foundation is simulated using Solid Meshes instead of the usual finite element plane elements? Would the results justify the additional modelling effort, or would it simply replicate what we already achieve with conventional methods?
Modelling FEA Solids Slab
The modelling process begins with a Plane Shell that defines the slab’s plan geometry. The next step is setting the mesh size. This value should not exceed the slab thickness; for example, if the slab is 200 mm thick, the maximum mesh size should also be 200 mm. Once the mesh size is specified, go to the Modify tab and select Mesh All. Switch to the Discretized Model view to assess mesh quality. This step is crucial because mesh refinement directly affects how easily supports and loads can be applied later.

With the mesh verified, select the slab and activate the Solid Mesh command from the Plane Shell contextual ribbon. This generates a 3D Solid element. Assign the appropriate material properties and adjust the color to distinguish it clearly from other model components.


Modelling Elastic Supports
Unlike plane elements, Plane Area Supports are not compatible with solid elements. Instead, supports must be assigned at each nodal point. The Array and Linear Copy tools significantly speed up this process. Start by selecting the Point Support command. Set all rotational fixities to Free, and all translational fixities to Free, except for Type Y, which should be set to Elastic. This activates an additional input field: Stiffness Y (kN/m).
The stiffness value is obtained by multiplying the nodal tributary area by the Subgrade Modulus. Values differ between corners, edges, and interior nodes, so careful calculation is required. To maintain clarity, consider applying color coding or naming conventions to distinguish between support types.


Next, provide horizontal restraint in both directions, otherwise the model will remain unstable. Corner supports are usually the most practical locations for this. With the restraints in place, loads can be defined. For 3D Solids, only Point Loads are currently supported, so loading must be assigned node by node. At this point, the model setup is complete.

Interpreting Analysis Results
The results stage reveals the main limitation of this approach. The slab deflection pattern is essentially uniform, with little variation across the surface. This produces unrealistic or unhelpful outputs. Even if SUMO simulated the elastic foundation perfectly, the Solid Results are not particularly useful for concrete slab design. They report stresses and shear forces, but these values do not translate directly into the bending moment information required for reinforced concrete checks.
By contrast, Shell Results provide bending moment data directly, making them far more suitable for practical design of concrete plane elements. For this reason, while solid modelling is an interesting exercise, the traditional shell-based approach remains the more efficient and design-relevant option.
Conclusion
Using Solid Meshes in SUMO to model a slab on elastic foundation is technically feasible, but it introduces significant inefficiencies. The requirement to define supports at every node, combined with the limitations of point-load application, results in a model that is time-consuming to set up and interpret. More importantly, the outputs generated by Solid elements—stresses and shear forces—do not directly support reinforced concrete design, whereas Shell elements provide the bending moment data that designers actually need.
In practice, Solid Meshes may offer value for research purposes or for highly localized investigations, but for routine slab-on-foundation analysis, Shell elements remain the more efficient and design-relevant approach.