Two-Way Slab Design (NSCP 2015 / ACI 318) Using the Coefficient Method: Step-by-Step Worked Example

By Engr. Ruel H. Cepeda, Structural Engineer

A slab is two-way when its longer-to-shorter span ratio, L/S, is less than 2.0 — both directions carry significant moment. For L/S ≥ 2, see the companion article on one-way slab design (NSCP 2015 / ACI 318). This article covers the classical moment-coefficient method for slabs on beams on all four sides — minimum thickness, the Ca/Cb tables, middle-strip vs. column-strip steel, spacing, corner reinforcement, and shear — then a full numeric example on a 4.5 m × 6.0 m interior panel.

Confirm the Classification: Two-Way or One-Way?

Compute L/S using the longer span L and shorter span S. When L/S < 2, both directions need design and the panel is two-way; the method below, or the Direct Design Method (DDM), applies. When L/S ≥ 2, essentially all the load runs in the short direction and the one-way method governs instead. For a DDM treatment, see the sister site's two-way slab DDM guide on RHCES; the coefficient method below is the older, hand-calculation-friendly alternative, still widely used for beam-supported panels in Philippine practice.

Step 1 — Minimum Thickness for Slabs with Beams on All Sides

ACI 318 Table 8.3.1.2 (mirrored in NSCP 2015) sets minimum thickness by α_fm, the average flexural stiffness ratio of the four edge beams to the slab (α_f = E_cbI_b/E_csI_s for each beam, averaged). Deflection need not be computed if h meets these limits:

l_n is the clear span in the long direction (face-to-face of beams, mm); f_y in MPa; β is the ratio of clear spans, long/short. Note that the middle row's denominator grows with α_fm — stiffer beams permit a thinner slab — and reaches the (36 + 9β) form of the last row exactly at α_fm = 2.0. Run trial beam sizes through the free minimum beam depth calculator, which shares this same ACI 318 thickness-table family for beams and slabs.

Step 2 — Loads and Factored Design Load

Sum self-weight (h in m × 24 kN/m³), floor finish, and any partition allowance for dead load D; take live load L from NSCP 2015 Table 205-1 for the governing occupancy:

w_u = 1.2D + 1.6L

Keep the factored dead and live components separate too — w_u,D = 1.2D, w_u,L = 1.6L — since positive-moment coefficients are tabulated separately by load type (pattern-loading effects), while the negative coefficient at a continuous edge applies to the combined w_u.

Step 3 — The Moment Coefficient Method (ACI 318-63 Method 3)

This classical method is no longer printed in ACI 318 or NSCP 2015 themselves, but both codes permit any slab analysis that satisfies equilibrium and geometric compatibility, and the tables are reproduced in most Philippine RC textbooks and design handbooks. It assumes beams stiff enough to be unyielding edges, each edge either continuous or discontinuous. Design moments per metre width:

M_a = C_a·w·l_a²    M_b = C_b·w·l_b²

l_a = shorter clear span, l_b = longer clear span; M_a is for bars parallel to l_a, M_b for the long-direction bars. The table has nine edge-condition cases, indexed by m = l_a/l_b in 0.05 steps from 1.00 to 0.50, with separate coefficients for negative moment, positive dead-load moment, and positive live-load moment, in both directions:

• Case 1 — all four edges discontinuous (simply supported all around); no negative-moment coefficients.
• Case 2 — interior panel, all four edges continuous (used below).
• Cases 3–9 — one, two, or three edges discontinuous (e.g., Case 4 is the corner panel with two adjacent discontinuous edges); match the pictured edge pattern to your support condition.

Step 4 — Middle Strip vs. Column Strip

Each direction is divided into a middle strip and two column strips. The middle strip — the center half of the panel width — carries the full unit moment M_a or M_b above. In the two column strips (the quarter nearest each supporting beam, on each side) the method lets the moment taper linearly from the full middle-strip value at the middle-strip edge down to one-third of it at the beam face; in practice the column strip is designed for the average, two-thirds of the middle-strip unit moment.

Step 5 — Main Reinforcement (R_n/ρ Method)

Both orthogonal bar layers here are primary flexural steel — unlike a one-way slab, there is no separate temperature-and-shrinkage direction. For each design moment, on a 1000 mm strip:

R_n = M_u / (φ·b·d²)    ρ = (0.85f′_c/f_y)·[1 − √(1 − 2R_n/0.85f′_c)]    A_s = ρ·b·d

φ = 0.90, b = 1000 mm. Since the short-direction bars normally carry more moment (C_a > C_b when m < 1), place them as the outer layer for the larger effective depth. Compare the computed A_s against the two-way slab minimum, A_s,min = 0.0018·b·h for Grade 420 bars (0.0020·b·h for f_y < 420 MPa; 0.0018×420/f_y·b·h, but not less than 0.0014·b·h, above it) — note that this minimum is taken on the gross section b·h, not on b·d — and use whichever is larger; the free reinforcement calculator automates the A_s-to-spacing conversion.

Step 6 — Bar Spacing

ACI 318 caps main-bar spacing at critical sections to the lesser of 2h or 450 mm for two-way slabs — tighter than the 3h/450 mm limit for one-way slabs, since both directions are primary steel here. Convert required A_s to spacing with s = 1000·A_b/A_s, then round down to a practical value not exceeding the 2h cap.

Step 7 — Corner Reinforcement

At an exterior corner where two discontinuous edges meet (Cases 1, 4, 6 and 7 — not the fully continuous interior panel, Case 2), the slab tends to lift and crack diagonally. Where the edge beams are stiff (ACI 318 / NSCP 2015 trigger this where an edge beam has α_f > 1.0, or the slab sits on walls), provide special reinforcement, top and bottom, sized for the panel's largest positive moment per unit width, extending from the corner one-fifth of the longer span in each direction (the ACI 318 family wording; some older texts and other codes use the shorter span). Run it in two layers parallel to the sides, or place the top steel parallel to the diagonal and the bottom steel perpendicular to it.

Step 8 — Shear Check

A beam-supported two-way slab transfers load to its beams by tributary area — triangular for the short-span beams, trapezoidal for the long ones, split by 45° lines from each corner. The slab's own one-way (beam-type) shear peaks along the long edges, where the short-direction strips land (and at mid-length of the short edges), at approximately V_u ≈ w_u·l_a/2 per metre width — slightly conservative, since Method 3's load-distribution table assigns only a fraction W_a of the load to the short direction. Check it against:

φV_c = φ·0.17λ√f′_c·b·d

φ = 0.75, λ = 1.0 (ACI 318-14/NSCP 2015 form); ACI 318-19 gives V_c = 0.66λ_sλ(ρ_w)^1/3√f′_c·b·d, worth checking too for lightly reinforced slabs. This is one-way (beam) shear, not punching shear — punching applies to slabs on columns without beams (flat plates/flat slabs), not to a beam-supported panel like this one.

Worked Example — 4.5 m × 6.0 m Interior Panel

Given: Interior panel, all four edges continuous (Case 2), beam centerlines 4.5 m (short) × 6.0 m (long), 300 mm-wide beams, trial h = 150 mm. f′_c = 21 MPa; Grade 415 bars, 12 mm dia., 20 mm cover. Floor finish (assumed) 1.0 kPa; ceiling and services (assumed) 0.5 kPa; live load L = 2.4 kPa (office).

1. Thickness check

Clear spans: l_n,short = 4.5−0.3 = 4.2 m; l_n,long = 6.0−0.3 = 5.7 m; β = 5.7/4.2 = 1.357. Assume α_fm ≈ 1.2 (typical moderately stiff beams; compute the actual value from your section properties for final design) — 0.2–2.0 family, 125 mm floor.
h_min = 5700×(0.8+415/1400) ÷ [36+5×1.357×(1.2−0.2)] = 5700×1.0964 ÷ (36+6.79) = 6249.6÷42.79 = 146.1 mm < 150 mm provided — OK (and above the 125 mm floor). Had the beams been stiffer, α_fm > 2.0, the (36+9β) row would have allowed 6249.6÷48.21 = 129.6 mm.

2–3. Loads and factored load

Self-weight = 0.150×24 = 3.60 kPa; D = 3.60+1.00+0.50 = 5.10 kPa; L = 2.40 kPa.
w_u = 1.2(5.10)+1.6(2.40) = 6.12+3.84 = 9.96 kPa; w_u,D = 6.12 kPa; w_u,L = 3.84 kPa.
m = l_n,short/l_n,long = 4.2/5.7 = 0.74 ≈ 0.75 (nearest tabulated value).

4. Coefficients and moments (Case 2, m = 0.75)

M_a,neg = 0.069×9.96×4.2² = 0.069×9.96×17.64 = 12.12 kN·m/m.
M_b,neg = 0.022×9.96×5.7² = 0.022×9.96×32.49 = 7.12 kN·m/m.
M_a,pos = (0.028×6.12×17.64)+(0.045×3.84×17.64) = 3.02+3.05 = 6.07 kN·m/m.
M_b,pos = (0.009×6.12×32.49)+(0.014×3.84×32.49) = 1.79+1.75 = 3.54 kN·m/m.

5–6. Main steel (middle strip) and spacing

d_short = 150−20−6 = 124 mm (outer layer); d_long = 150−20−12−6 = 112 mm (inner layer).
A_s,min = 0.0018×b×h = 0.0018×1000×150 = 270 mm²/m, both directions (gross section, so it is the same for the outer and inner layers).
Short, negative: R_n = 12.12×10&sup6;/(0.9×1000×124²) = 0.876 MPa; ρ = 0.0430×[1−√(1−2(0.876)/17.85)] = 0.0430×0.0503 ≈ 0.00217 → A_s = 0.00217×1000×124 = 269 mm²/m, a hair under A_s,min → 270 mm²/m.
Short, positive: R_n = 0.439 MPa; ρ = 0.00107 → A_s = 133 mm²/m < A_s,min → 270 mm²/m.
Long, negative: R_n = 7.12×10&sup6;/(0.9×1000×112²) = 0.631 MPa; ρ = 0.00155 → A_s = 173 mm²/m < A_s,min → 270 mm²/m.
Long, positive: R_n = 0.313 MPa; ρ = 0.00076 → A_s = 85 mm²/m < A_s,min → 270 mm²/m.
Cap: 2h = 300 mm. 12 mm bars, A_b = 113.1 mm²: s = 1000×113.1/270 = 419 mm > 300 mm in all four cases, so the cap governs — use 12 mm@300 mm o.c. both directions, top and bottom, middle strips (377 mm²/m ≥ all four required values). If your reviewer reads Grade 415 strictly as f_y < 420 MPa, A_s,min = 0.0020×1000×150 = 300 mm²/m — still satisfied by 377 mm²/m.

7. Column strip and corner steel

Column-strip demand = 2/3 of the middle-strip moments above (worst case 2/3×12.12 = 8.08 kN·m/m → A_s ≈ 177 mm²/m), but A_s,min = 270 mm²/m and the 2h cap still apply there, so the same 12 mm@300 (377 mm²/m) simply continues through the column strips; 10 mm@275 (286 mm²/m) is the leanest alternative that still clears A_s,min and the 300 mm cap. This is an interior, all-edges-continuous panel (Case 2), so Step 7's corner-reinforcement provision does not apply — it applies only at the building's actual exterior corner panels, using their own case.

8. Shear check

V_u ≈ w_u·l_a/2 = 9.96×4.2/2 = 20.92 kN/m.
φV_c = 0.75×0.17×√21×1000×124 = 72.5 kN >> 20.92 kN — OK.
ACI 318-19 cross-check: λ_s = √(2/(1+124/250)) = 1.16 → 1.0; ρ_w = 377/(1000×124) = 0.00304; V_c = 0.66×(0.00304)^1/3×√21×1000×124 = 0.66×0.145×4.583×124,000 = 54.3 kN; φV_c = 40.8 kN > 20.92 kN — still OK, no stirrups needed.

Assumptions & Limitations

• Applies to nonprestressed, solid two-way slabs on beams on all four sides (L/S < 2); flat plates/flat slabs on columns need the Direct Design Method or Equivalent Frame Method and a punching-shear check instead.
• The Case 2 (interior panel), m = 0.75 coefficients shown are quoted from the standard ACI 318-63 Method 3 tables; the actual m here is 0.737, and interpolating to it would raise C_a,neg only slightly (to about 0.070) without changing the bar selection. Read the exact case and m ratio for your own panel from the published table before finalizing bar sizes.
• α_fm = 1.2 in the worked example is assumed; compute it from actual beam and slab section properties, using the effective T-beam section (web plus a slab flange each side, as defined in the ACI 318 / NSCP 2015 two-way slab provisions).
• Floor finish and ceiling/services allowances are assumed values; confirm against the actual finish and MEP schedule.
• A_s,min pairs Grade 415 with 0.0018·b·h (treating it as Grade 420), common in Philippine practice; the stricter 0.0020·b·h reading is also satisfied here. Verify the fy-based formula for other grades, and confirm against the code edition adopted by the building official of record.
• The shear check uses the conservative w_u·l_a/2 shorthand and the ACI 318-14 / NSCP 2015 V_c; the ACI 318-19 size-and-ρ-dependent V_c is shown as a cross-check only.

Frequently Asked Questions

How is the coefficient method different from the Direct Design Method (DDM)?

The coefficient method (ACI 318-63 Method 3) works panel-by-panel from a fixed Ca/Cb table keyed to the edge-continuity case and span ratio m, valid only when beams on all four sides are stiff enough to act as unyielding supports. The DDM instead distributes a total static moment along a column line among column and middle strips by stiffness-ratio rules, and covers slabs with or without beams. See the sister site's two-way slab DDM guide for that method.

Why does the short direction (Ca) carry more moment than the long direction (Cb)?

Load splits between the two spanning directions in proportion to their stiffness — roughly as the inverse fourth power of the span — and the shorter, stiffer strip deflects less under the same load, so it attracts more of it. As the panel elongates (m farther below 1.0), Ca keeps rising and Cb falls toward zero — an infinitely long panel behaves as one-way, spanning only the short direction.

Why did the code's maximum spacing (2h), not computed A_s, govern bar spacing throughout the worked example?

The 150 mm slab is lightly to moderately loaded, so three of the four flexural steel areas fall below A_s,min = 0.0018bh = 270 mm²/m and the fourth only just reaches it — and even 270 mm²/m of 12 mm bars works out to a 419 mm spacing, well above the 2h = 300 mm cap; the maximum-spacing rule steps in before flexural demand does. This is common for typical loads and slab thicknesses in the 125–175 mm range — worth checking on every panel, not assumed.

Getting the case, the coefficients, and the strip split right the first time avoids rebar rework and shear surprises on site. Try the minimum beam depth calculator and the reinforcement calculator, or browse all free web tools. Offline spreadsheet versions are on the download page for use without internet.