Yield-stress materials sit at the center of a quiet theoretical war. One model says they flow at all stress levels — slowly under small loads, faster under large ones. The other says they don't flow at all below a critical stress. Both are widely used. Both can fit the same data. They disagree about what the material is actually doing.
The resolution is surprisingly clean. When you apply a stress below yield to a microgel or an emulsion, the strain response is bounded and periodic. The material deforms, oscillates, and returns. It does not flow.
The confusion came from slip. In standard rheometers, thin layers of material at the wall can slide rather than deform, creating apparent flow that isn't bulk flow at all. When parallel superposition rheometry properly accounts for this artifact, the bounded response emerges clearly. The “sub-yield flow” was the instrument's misinterpretation.
But the bounded response isn't simple elasticity either. It's nonlinear viscoelasticity — the material's stiffness and damping both depend on how far it's been pushed. The response is recoverable, but the path depends on the amplitude. The sub-yield regime is doing something, just not flowing.
This breaks a convenient assumption embedded in many constitutive models: that nonlinearity and yielding are the same transition. If materials are nonlinear below yield but don't flow, then yielding isn't the onset of nonlinearity. It's the onset of irreversibility. These are different thresholds, and conflating them has shaped decades of model development.
The difference matters because it's a category error about what a threshold is. Below yield, the material is stubborn — it bends but won't break. The models that predicted flow were measuring the instrument, not the material.