When you create a probe by selecting a specific x, y, z location on the model (using the Hit Point Coordinate option) or using a coordinate system, Mechanical chooses a singular element of a singular body that contains the x, y, z location. In the event the probe lies between two bodies, the application still only picks one of the two bodies (and one element) to use for the result calculation. Result values are interpolated from the element's nodes to the x, y, z location.

When picking a specific x, y, z location, you can obtain the probe result directly at the closest corner node, without extra interpolation, by right-clicking on the probe object in the tree and choosing Snap to mesh nodes from the context menu. The identification number of the closest corner node is displayed as the Node ID in the Details pane of the probe in the Results category. See the Interpolation section for additional information.

As a result of an element selection limitation, the application can select unscoped adjacent elements that will then contribute to a probe's result.

When you set the Location Method property to Geometry Selection, the scoping algorithm initially selects a group of nodes on the highlighted geometry (face, edge, etc.). The algorithm then selects any elements attached to those nodes and can include unscoped adjacent elements, that is, elements not contained in the scoping (green highlight).

As a result, probe types that have element-nodal results, such as Structural Force Reaction/Moment Reaction and Magnetostatic Force Summation/Torques, the scoping algorithm selects these unscoped adjacent elements and these additional elements can contribute to the probe results.

Although a limitation, this behavior enables you to compare element-nodal probes to nodal probes scoped to boundary conditions. For example, an element-nodal Force Reaction probe scoped to a face can be compared to a nodal probe scoped to a boundary condition that occupies the face.

When the result that is associated with a probe is purely elemental (one value per element regardless of the number of nodes), then the probe does not include elements from unscoped bodies. For example, a Volume probe scoped to a body provides the exact volume of the body.

Shell element node-based results (like stress and strain) exist at the top, bottom, and middle of the shell element (or the layer). Therefore, a shell node can have three values for a given layer.

For result probes on shell models that are scoped by Geometry Selection, the probe normally considers the top value and bottom values at the scoped nodes and picks either the maximum or minimum value. Based on the probe type, the Spatial Resolution property enables you to select whether the application uses the maximum or minimum value.

If you scope your probe to a Coordinate System, the application performs an interpolation using the values at the top and bottom of the shell.

With these situations in mind, your scoping may present results with unexpected or non-intuitive values.

For example, consider a probe scoped to a coordinate system that is situated near a vertex at the mid-plane. For this situation, the interpolation is a simple arithmetic average. However, what if the Top value at the node is -1000 and the Bottom value at the same node is 1000, a very real scenario for shell models. The coordinate system probe would report (-1000 + 1000)/2 = 0.

Now consider the probe scoped by geometry to the same Vertex. It would report max(-1000, 1000) = 1000 if the Spatial Resolution property was set to Use Maximum.