Geant4 study · cylindrical tracking detectors

Fold the Detector

Can sensors folded like origami catch more particles than flat tiles? In simulation, the best fold raises acceptance by 2.36 percentage points over a matched flat barrel, for 5.69% more material.

What the fold costs, what it buys

Each fold against the flat barrel. Up is more particles caught; right is more material. Hover or focus a point.

01230246810extra material, % of X₀acceptance gain, ppBarrelMiura-oriKreslingYoshimura
Control… facets

Barrel

846 flat tiles · control

83.18%acceptance
Baseline for every gain shown
⟨X/X₀⟩ = 3.977×10⁻³
See the results Tap for details

The control

846 flat tiles around a cylinder. This is what a conventional detector looks like.

  • Tile columns141
  • GapsTrue gaps by design
  • Material ⟨X/X₀⟩3.977×10⁻³

Reference. Every gain on this page is measured against it.

Lightest… facets

Miura-ori

Minimal material cost

83.67%acceptance
Acceptance+0.49 pp
Material+1.88% X₀
See the results Tap for details

How it folds

A herringbone of parallelograms, mapped onto a cylinder.

  • Gain at the ends94.4%
  • Thinly covered bins0.04%
  • Extra material+1.88%

Best for designs limited by multiple scattering. A small gain that costs almost nothing.

Most acceptance… facets

Kresling

Maximum acceptance

85.54%acceptance
Acceptance+2.36 pp
Material+5.69% X₀
See the results Tap for details

How it folds

A twisted, triangulated cylinder that shortens and rotates when compressed.

  • Gain at the ends97.5%
  • Thinly covered bins1.45%
  • Gain vs facet count+2.31 pp at 6 → ~0 at 32

Best for the most coverage, if tracking can handle material near the fold vertices.

… facets

Yoshimura

Dominated by Kresling

84.48%acceptance
Acceptance+1.30 pp
Material+8.01% X₀
See the results Tap for details

How it folds

A diamond pattern, familiar from how thin cylindrical shells buckle.

  • Gain at the ends98.9%
  • Thinly covered bins1.86%
  • Fold 15° → 75°84.5% → 94.2%, material triples

Set aside. Most material, most thin spots, and fewer particles caught than Kresling.

F
v²
μ
ΔE
G
ρ