A mate removes degrees of freedom. A joint removes degrees of freedom too — but deliberately leaves the ones that let the assembly actually move like the real mechanism will.
A common early mistake is mating every part in an assembly fully rigid — concentric here, coincident there, flush everywhere — because that's what makes the model "stay together" and stop drifting apart on screen. That approach works fine for a welded bracket, but it silently breaks the moment the assembly is supposed to represent a mechanism: a hinge that should rotate, a slider that should translate, a shaft that should spin in its bearing. Every standard mate (concentric, coincident, distance, angle, parallel) removes some number of the six possible degrees of freedom — three translations, three rotations — between two parts. The real distinction that matters isn't "mate vs. constraint," which are largely the same underlying math; it's whether the combination of mates you've applied leaves zero degrees of freedom (fully rigid, correct for permanent joints) or exactly the right number and type of degrees of freedom (correct for a mechanism that's meant to move).
Two rigid bodies in free space have six degrees of freedom relative to each other: translation along X, Y, Z and rotation about X, Y, Z. Every standard assembly mate removes a specific, predictable subset of those six. A concentric mate between two cylindrical faces removes four DOF (two translations perpendicular to the axis, two rotations about axes perpendicular to that axis) but leaves translation along the axis and rotation about the axis free. A coincident (flush) mate between two planar faces typically removes one translation and two rotations. Stacking mates is additive: apply enough of them and the total DOF removed reaches six, at which point the part is fully constrained and can't move at all — correct for anything meant to be permanently joined. But a mechanism's whole purpose is that some relative motion is supposed to exist, so the mate scheme has to stop short of removing all six deliberately: a concentric-plus-coincident combination on a pin and hinge bore, for instance, removes exactly five DOF and leaves the one rotation about the pin axis that makes it a functioning hinge rather than a welded joint. Most parametric CAD packages (SolidWorks, Inventor, Creo, NX) now offer named "mechanical mates" or joint types — hinge, slider, cam, gear, rack-and-pinion — that are really just pre-packaged combinations of the same basic geometric mates, chosen specifically to leave the right DOF open for that mechanism type.
Over-constraining an assembly "to be safe" is one of the most common sources of assembly rebuild errors and motion-study failures downstream. Fully mating a part that's meant to be a moving mechanism component doesn't just prevent that one part from moving — it can also produce mate conflicts (the solver reporting the assembly is over-defined, sometimes silently ignoring one of the redundant mates rather than flagging it clearly) when a later mate change tries to reintroduce motion the earlier mates already eliminated. It's also worth distinguishing over-constraint from redundant-but-consistent constraint: some CAD kernels tolerate certain redundant mates gracefully (for example, three planes all constrained parallel to each other, which is mathematically redundant but not contradictory) while flagging others as errors outright — the safer default for any part that could plausibly need to move later is to apply exactly the mates the design intent calls for, verify the remaining DOF count matches the intended motion (many CAD packages show this directly, e.g. SolidWorks's degrees-of-freedom indicator in the assembly tree), and only add more constraint than that once it's confirmed the part genuinely is meant to be rigid.
Explains the real distinguishing feature between a fully-mated (rigid) CAD assembly and one built as a working mechanism — not different underlying math, but how many of the six degrees of freedom the applied mates remove, and whether the ones left open match the motion the mechanism is meant to have.
Because the same set of mate types (concentric, coincident, parallel, tangent, distance, angle) is used for both permanently joined parts and mechanism joints, and CAD software doesn't force a distinction between the two use cases at the UI level, engineers new to assembly modeling often default to over-mating everything to make the model "behave" — stop parts from floating apart in the viewport — without considering whether that particular part is supposed to move relative to its neighbors.
Any two unconstrained rigid bodies in 3D space have six relative degrees of freedom: translation along X, Y, and Z, and rotation about X, Y, and Z. Each mate type removes a specific, well-defined subset: a coincident (flush) mate between planar faces typically removes one translation and two rotations (three DOF); a concentric mate between cylindrical faces removes two translations and two rotations (four DOF), leaving axial translation and rotation about that axis free; a distance or angle mate removes exactly one DOF each. Stacking mates removes DOF cumulatively until reaching zero (fully constrained, rigid) — going further than zero doesn't remove more motion, it creates an over-constrained (redundant or conflicting) condition the solver has to detect and resolve, sometimes by silently dropping one of the conflicting mates. Modern CAD packages' "mechanical mate" or "joint" types (hinge, slider, cylindrical, universal, cam, gear/rack-and-pinion) are pre-built combinations chosen to leave exactly the DOF a named mechanism type needs — a hinge/revolute joint leaves one rotational DOF, a slider/prismatic joint leaves one translational DOF, a cylindrical joint leaves both one translation and one rotation along the same axis.
Any assembly destined for a CAD-native motion study, a kinematic simulation, or eventual multibody dynamics analysis needs its mate scheme to reflect the actual mechanism kinematics, not just visual stability — an accidentally over-constrained joint won't animate at all when a motor or driver is applied to it, and the resulting error message ("cannot solve assembly" or "redundant mate detected") is one of the most common points of confusion for engineers newer to mechanism design. Getting this right early also matters for downstream interference/collision checking and for exporting the assembly to a dedicated multibody dynamics tool (like MSC Adams or Simulink Simscape Multibody), which typically expects joints defined with the correct DOF rather than a fully rigid assembly.
It means the applied mates collectively try to remove more degrees of freedom than physically exist between the two parts — for example, two separate mates each independently trying to fix the same rotation to different, conflicting values. The solver either flags this outright as an error, or in some cases silently satisfies one mate and treats the other as redundant, which can hide a modeling mistake rather than surface it clearly.
Yes, and this is the normal case — a gearbox housing's cover plate is typically fully mated (rigid) to the housing, while the shafts and gears inside are mated with exactly the rotational DOF each needs to spin. Treating the whole assembly with one uniform mating philosophy is itself often the underlying mistake.
Not usually — a living hinge is typically modeled as a single continuous part (thin flexible section connecting two thicker regions) rather than as two separate mated bodies, so its motion is captured through material flexibility (often verified with a nonlinear FEA bend simulation) rather than an assembly-level joint at all.
Similar in effect but different in mechanism — a rigid sub-assembly (sometimes called a "rigid group" or achieved by making a multi-body part) treats several components as one solid unit for simulation and analysis purposes, bypassing individual mate resolution between them entirely, whereas a fully mated assembly still resolves each mate individually but happens to leave zero net DOF as a result.
Try our Engineering Software Studio
More calculators, simulators, and guides for this discipline.
Engineering Software Handbook (Full Access)
✨ Premium ContentA zoomable interactive reader — free preview, then unlock the full set.