Reverse Brazil-nut effect

Still disputed in the literature as of 2026.

After the ordinary Brazil-nut effect was explained, Hong, Quinn and Luding proposed in 2001 that if the large intruder particles are sufficiently more massive (not just larger) than the small ones, a competition between percolation and 'condensation' of the small grains should reverse the segregation, sinking the large particles instead. Their theoretical prediction was supported by molecular dynamics simulations, but subsequent experimental attempts to reproduce the effect gave conflicting results, leading to a published Comment ('Does the Reverse Brazil Nut Problem Exist?') and a Reply in Physical Review Letters in 2002, each group standing by its own data. The disagreement was never fully adjudicated by a definitive follow-up experiment, so whether the reverse effect is a robust, general phenomenon or an artifact of specific experimental conditions (air pressure, container geometry, friction) remains genuinely contested in the granular-physics literature.

A single physics paper can flip a beloved kitchen-table demonstration on its head — and then trigger a real scientific fight about whether the flip is even real.

Explanations that circulated

• That the reverse effect is a straightforward, universally reproducible flip of the ordinary Brazil-nut effect once the mass ratio crosses a threshold — several groups reported being unable to reproduce the predicted reversal under nominally identical mass/diameter ratios, and a published Comment questioned whether the effect exists as originally framed, with the original authors disputing the critics' experimental setup.

References

• Reverse Brazil Nut Problem: Competition between Percolation and Condensation, D. C. Hong, P. V. Quinn, S. Luding, Physical Review Letters, 2001
• Does the Reverse Brazil Nut Problem Exist?, M. E. Möbius, X. Cheng, P. Eshuis, G. S. Karczmar, S. R. Nagel, H. M. Jaeger (comment), Physical Review Letters, 2002