calcnote ACI 318-19
Methodology Feedback Run a calc

T-beam vs rectangular beam: ACI 318-19 comparison

A beam cast monolithically with a slab uses the slab as a compression flange. ACI 318-19 Table 6.3.2.1 limits the effective flange width. When the compression zone (stress block depth a) stays within the flange, the section behaves as a wider rectangular beam. When a exceeds the flange thickness, a two-part calculation accounts for the overhang and web separately. This page compares both cases side by side using the same engine that powers the live T-beam calculator. For effective flange width rules, see the flange width spoke.

Effective flange width (brief)

ACI Table 6.3.2.1 limits how much slab acts as flange:

Common inputs

Both cases share these properties:

Web widthbw = 12 in.
Total depthh = 24 in.
Clear cover1.5 in. (Table 20.6.1.3.1)
Concretef′c = 4,000 psi, normal-weight (λ = 1.0)
SteelGrade 60 (fy = 60 ksi)
Stirrups#3 @ 8 in., 2 legs (Av = 0.22 in2)
Factored loadsMu = 250 k-ft, Vu = 40 kip

Case A: neutral axis in the flange (a ≤ hf)

4-#8 bars (1 layer), bf = 48 in., hf = 6 in.

$$d = 24 - 1.5 - 0.375 - 0.500 = 21.625\ \text{in}$$
$$A_s = 4 \times 0.79 = 3.16\ \text{in}^2$$

Whitney stress block using the full flange width:

$$a = \frac{A_s\,f_y}{0.85\,f'_c\,b_f} = \frac{3.16 \times 60}{0.85 \times 4 \times 48} = 1.162\ \text{in}$$

Since a = 1.162 in. < hf = 6.0 in., the neutral axis stays in the flange. The section behaves as rectangular with width bf = 48 in.

$$c = \frac{a}{\beta_1} = \frac{1.162}{0.85} = 1.367\ \text{in}$$
$$\varepsilon_t = 0.003 \times \frac{d - c}{c} = 0.003 \times \frac{21.625 - 1.367}{1.367} = 0.04447$$

εt = 0.04447 far exceeds 0.005, so the section is tension-controlled and φ = 0.9 (Table 21.2.2).

$$\phi M_n = 0.9 \times 3.16 \times 60 \times (21.625 - 0.581) = 3{,}591.0\ \text{k-in} = 299.2\ \text{k-ft}$$
$$\text{DCR} = \frac{M_u}{\phi M_n} = \frac{250}{299.2} = 0.835$$

When a < hf, ignore the web entirely for flexure. The wide flange provides so much compression area that the stress block is shallow. This is the common case for typical slab-beam construction.

Case B: neutral axis in the web (a > hf)

6-#8 bars (2 layers), bf = 24 in., hf = 3 in.

$$d = 20.625\ \text{in (2 layers: centroid shifts down)}$$
$$A_s = 6 \times 0.79 = 4.74\ \text{in}^2$$

First try the rectangular assumption with the full flange width:

$$a = \frac{A_s\,f_y}{0.85\,f'_c\,b_f} = \frac{4.74 \times 60}{0.85 \times 4 \times 24} = 3.485\ \text{in}$$

a = 3.485 in. > hf = 3.0 in., so the neutral axis drops into the web. Split the compression zone into flange overhangs and web:

Two-part decomposition

$$A_{sf} = \frac{0.85\,f'_c\,(b_f - b_w)\,h_f}{f_y} = \frac{0.85 \times 4 \times (24 - 12) \times 3}{60} = 2.04\ \text{in}^2$$
$$A_{sw} = A_s - A_{sf} = 4.74 - 2.04 = 2.70\ \text{in}^2$$
$$a_{\text{web}} = \frac{A_{sw}\,f_y}{0.85\,f'_c\,b_w} = \frac{2.70 \times 60}{0.85 \times 4 \times 12} = 3.971\ \text{in}$$

Moment contributions

$$M_{n,f} = A_{sf}\,f_y\!\left(d - \frac{h_f}{2}\right) = 2.04 \times 60 \times (20.625 - 1.5) = 2{,}340.9\ \text{k-in}$$
$$M_{n,w} = A_{sw}\,f_y\!\left(d - \frac{a_{\text{web}}}{2}\right) = 2.70 \times 60 \times (20.625 - 1.986) = 3{,}019.6\ \text{k-in}$$
$$M_n = M_{n,f} + M_{n,w} = 5{,}360.5\ \text{k-in}$$

Strain check:

$$c = \frac{a_{\text{web}}}{\beta_1} = \frac{3.971}{0.85} = 4.671\ \text{in}$$
$$\varepsilon_t = 0.003 \times \frac{20.625 - 4.671}{4.671} = 0.01025 > 0.005$$

Tension-controlled, φ = 0.9.

$$\phi M_n = 0.9 \times 5{,}360.5 = 4{,}824.5\ \text{k-in} = 402.0\ \text{k-ft}$$
$$\text{DCR} = \frac{250}{402.0} = 0.622$$

When a > hf, the overhang compression and web compression have different moment arms. Asf acts at (d − hf/2) and Asw acts at (d − aweb/2). The engine iterates because the initial a (assuming full bf width) must be rechecked against hf.

Shear: same formula, but ρw matters

For shear, ACI Table 22.5.5.1 uses bw (web width) regardless of section type. The T-beam flange does not contribute to Vc. However, Table 22.5.5.1 provides two equations, and the one that governs depends on the longitudinal steel ratio ρw = As / (bw d):

When minimum stirrups are present, the engine uses max(Vc,a, Vc,b).

Case A (4-#8, d = 21.625)

$$\rho_w = \frac{3.16}{12 \times 21.625} = 0.01218$$

Vc,a = 32.82 kip, Vc,b = 30.21 kip. Eq (a) governs: Vc = 32.82 kip.

Case B (6-#8, d = 20.625)

$$\rho_w = \frac{4.74}{12 \times 20.625} = 0.01915$$

Vc,a = 31.31 kip, Vc,b = 33.50 kip. Eq (b) governs: Vc = 33.50 kip.

With more longitudinal steel (higher ρw), eq (b) can exceed eq (a). T-beams that need more tension steel for the two-part flexure check may also get a higher Vc.

Crack control and deflection

Summary comparison

Property Case A (NA in flange) Case B (NA in web)
Flange width bf48 in.24 in.
Flange thickness hf6 in.3 in.
Bars4-#8, 1 layer (As = 3.16)6-#8, 2 layers (As = 4.74)
Effective depth d21.625 in.20.625 in.
Stress block a1.162 in. (< hf)3.971 in. (> hf)
Flexure methodRectangular with bfTwo-part: Asf + Asw
φMn299.2 k-ft402.0 k-ft
Flexure DCR0.8350.622
Vc controlling eqTable 22.5.5.1(a)Table 22.5.5.1(b)
Vc32.82 kip33.50 kip
φVn51.38 kip50.65 kip
Shear DCR0.7790.790