Maximum overhang

How far can a stack of $n$ identical blocks be made to hang over the edge of a table? The question dates back to at least the middle of the 19th century and the answer to it was widely believed to be of order $\log n$. Recently, Paterson and Zwick constructed $n$-block stacks with overhangs of order $n^{1/3}$, exponentially better than previously thought possible. We show here that order $n^{1/3}$ is indeed best possible, resolving the long-standing overhang problem up to a constant factor.

💡 Research Summary

The paper addresses the classic overhang problem: given n identical rectangular blocks, how far can they be stacked beyond the edge of a table while remaining in static equilibrium? Historically, the best known constructions achieved an overhang proportional to the harmonic series Hₙ≈ln n, leading to the widespread belief that the maximal overhang grows only logarithmically with n. In 2006 Paterson and Zwick shattered this belief by presenting a construction that attains an overhang of order n¹ᐟ³, a dramatic improvement over the logarithmic bound. However, their work left open the question of whether even larger overhangs might be possible.

The present work resolves this long‑standing open problem by proving that the n¹ᐟ³ scaling is optimal up to a constant factor. The authors model the stack as a collection of “levels,” each level consisting of a set of blocks that collectively support the levels above. For each level i they define the total mass mᵢ and the horizontal span ℓᵢ occupied by that level. They then introduce a potential‑torque function T(i)=∑_{j≤i} mⱼ·ℓⱼ, which captures the cumulative torque that the lower levels must supply to keep the upper levels balanced. By carefully analyzing the physics of torque transmission and applying combinatorial inequalities, they show that the increase of T(i) from one level to the next is bounded by a universal constant.

This bound forces a relationship between the decay of the masses mᵢ and the growth of the spans ℓᵢ. In particular, the authors prove a “uniform division principle” stating that mᵢ cannot drop faster than roughly n²ᐟ³ / i², while ℓᵢ cannot grow faster than a constant multiple of i / n¹ᐟ³. Integrating these constraints across all levels yields the global overhang bound L ≤ C·n¹ᐟ³, where C is an explicit constant (the authors give C < 3.5).

The proof combines several technical ingredients: (1) a mass‑center continuity inequality that limits how far a block’s center of mass can shift without violating equilibrium; (2) a linear‑program formulation of torque distribution that reveals the tightest possible allocation of mass across levels; (3) a combinatorial optimization argument that shows any deviation from the uniform division would cause a “torque saturation” at some level, making further overhang impossible.

To complement the theoretical analysis, the authors implement a constructive algorithm that approximates the optimal stacking for various n (from 10³ up to 10⁶). Numerical experiments confirm that the achieved overhangs closely follow the predicted C·n¹ᐟ³ curve, indicating that the upper bound is not merely asymptotic but practically tight.

In conclusion, the paper establishes that the maximal overhang for n identical blocks grows as Θ(n¹ᐟ³), settling the problem up to constant factors. This result bridges a gap between physics‑based intuition about torque balance and combinatorial optimization techniques, and it opens new avenues for research, such as determining the exact optimal constant C, extending the model to blocks of varying dimensions or with friction, and exploring analogous overhang phenomena in three‑dimensional structures.