The Circular Girder in an Intze Tank — Why It's the Trickiest Member to Design

Every reinforced concrete member you meet early in a structures course is, sooner or later, a straight line: a beam spans between two supports, a column runs vertically between floors. The circular girder at the base of an Intze tank's cone breaks that pattern completely — it's a closed ring, supported not at two points but continuously (or at a handful of columns around its circumference), carrying a symmetric load from every direction at once. That geometry alone is enough to introduce a force most students never have to design for in an ordinary beam: torsion, showing up not because anything is applied off-centre, but because the girder is curved.

Why a straight beam analogy breaks down

Cut a small segment out of the circular girder and look at it in isolation. Load comes down onto it from the cone above and the columns below, same as any beam — but because the segment is curved, the load path along its own length isn't straight either. A curved member carrying load along its own curved axis develops torsion simply from that geometry, even under a perfectly symmetric, evenly distributed load with no eccentricity anywhere. This is fundamentally different from why an ordinary beam might see torsion (an off-centre slab, or a load applied away from the shear centre) — here, torsion is baked into the shape itself.

What lands on the girder, and from which direction

The circular girder sits exactly where two members meet: the conical slab above it, and the columns below it. The cone, narrowing as it goes down, pushes inward and downward onto the girder at an angle set by the cone's slope. The columns push back up, of course, but not necessarily perfectly aligned with where the cone's thrust lands — and it's this combination of a sloped incoming thrust, a vertical reaction, and the ring's own curvature that produces bending, direct compression, and torsion simultaneously in the same member.

Why the classic design breaks the load into three rings

Rather than trying to solve one impossibly complicated member, the standard Intze design (and this is really the elegant part) distributes the outward and inward thrusts across three separate rings, each doing one clear job:

• B1, at the top of the cylindrical wall, resists the outward hoop thrust from the top dome.
• B2, at the wall-to-cone junction, resists the combined outward pull of the wall's own hoop tension and the top of the cone's outward-leaning thrust.
• B3 — the circular girder itself, at the bottom of the cone — is where the cone's inward thrust and the bottom dome's outward thrust meet, and where the column reactions come in from below. This is the one carrying torsion, because it's the one doing double duty: closing the ring and transferring load down to discrete column points instead of a continuous support.

B1 and B2 are comparatively simple — pure hoop tension rings, well understood, sized directly from the thrust and reinforced accordingly. B3 is the one that needs the fuller torsion-plus-bending-plus-shear treatment, precisely because it's also the load-transfer point down to a finite number of columns rather than a continuous ring support.

Why this specific ring gets the "balance" benefit

This is also exactly the ring where the Intze balance shows up, discussed in more detail elsewhere on this blog: the cone's inward push and the bottom dome's outward push at this same level substantially cancel each other when the bottom dome's proportions are chosen well. That cancellation reduces the net hoop force the girder has to carry — but it does nothing to reduce the torsion, which comes from the ring's curvature and its discrete column supports, not from the balance of radial thrusts. Even a perfectly balanced Intze tank still needs its circular girder designed properly for torsion; the balance only simplifies one part of the problem, not all of it.

Why this matters for anyone actually designing one

It's tempting, working through an Intze tank design, to treat the circular girder the way you'd treat any other ring beam in the structure — check the hoop force, provide steel, move on. The presence of genuine torsion (from curvature) combined with bending and shear (from the discrete column supports) means this member needs a proper combined-stress check, not a simplified one. It's the one part of the whole tank where skipping a step actually shows up as a real design gap, not just a conservative simplification.