Home / Articles / Shelters SIG-08
High-Strength and Layered Concrete Under Contact Explosion
Two new C60 tests showed our conservative boundary is too conservative for high-strength concrete; with the fix, seven of seven tests agree. Sandwich composite slabs lie outside the scope of empirical engines.
1. A new paper, two new tests
Yin and Li (2026) subjected four full-scale slabs (2.0 × 2.0 m) to contact explosions: two monolithic high-strength C60 slabs (f′c = 62.2 MPa; 3 kg TNT on the 300 mm slab, 5 kg on the 400 mm slab) and two UHTCC–UHPC–UHTCC sandwich slabs of the same thicknesses. The monolithic slabs give two new independent tests for the site’s contact-charge assessment.
2. What the engine said, what the test showed
The scaled thickness T/W1/3 is 0.208 and 0.234 m/kg1/3; both slabs showed extensive scabbing, no breach. UFC §4-55 and Morishita both predict spall: correct. But the site’s conservative boundary (0.24), derived earlier from the C40 tests of Zhang et al. (2024), predicts “breach” for both, and because the joint assessment selects it, the joint result is too conservative here.
That boundary came from a single data set of normal-strength concrete. The fix: the conservative boundary now applies only for f′c ≤ 50 MPa; for high-strength concrete the decision is the less favourable of UFC and Morishita. Result: Zhang’s five tests and Yin and Li’s two, i.e. seven of seven, are classified correctly. This adjustment also rests on limited data and is documented as such.
3. Sandwich slabs: where the engines stop
The paper’s main subject is sandwich slabs: high-toughness composite (UHTCC) faces and an ultra-high-performance concrete (UHPC) core. At the same charge and thickness, rear-face damage changed from extensive scabbing in the monolithic slabs to confined scabbing or only bulging and cracking. According to the simulations, local resistance is governed by the UHPC core thickness, while the allocation of the UHTCC layers shifts damage between the front and rear faces.
These slabs cannot be computed with the site’s engines: UFC §4-55 and Morishita are for monolithic normal- and high-strength concrete; layered and fibre composites are outside their scope. Such a shelter wall must be designed with tests and validated numerical models.
4. Rules
- A conservative adjustment is limited to its data. A boundary derived from one data set should not be carried to another material class.
- New tests check old adjustments. Every new independent data point is added to the engine’s regression set.
- An engine’s scope is stated. Empirical thresholds are not used for composite and layered walls.
References
- Yin, X. and Li, Q. (2026). Rear-face failure control of large-scale UHTCC-UHPC-UHTCC sandwich slabs under contact explosion: Experimental and numerical investigations. International Journal of Impact Engineering 218:105851.
- Zhang, W. et al. (2024). Sci. Rep. 14:25700.
- Morishita, M. et al. (2004). Concrete Research and Technology 15(2):89–98.
- U.S. DoD (2008). UFC 3-340-02, §4-55.
- archi-civil.com engine/contact-spall.js (v637) — tests/v637-high-strength-contact.test.js.
The numbers in this article were produced by the site's engines; code values are given with their clause numbers.