Bending stress and extreme fibers

From moment to safety factor

The bending moment alone does not say how strongly a cross-section is stressed: for the same moment, a smaller section modulus WW leads to a larger extreme fiber stress. SpanDraft therefore calculates in two steps. It first computes the maximum elastic bending stress from the bending moment of the solver and the section modulus of the cross-section. It then relates this stress to the strength value RrefR_\mathrm{ref} of the material and yields the safety factor SS.

  1. ∣M∣max|M|_\mathrm{max}/ Wmin/\,W_\mathrm{min}⋅Wmin\cdot W_\mathrm{min}
  2. σb,max\sigma_{\mathrm{b,max}}Rref/R_\mathrm{ref}/Rref/R_\mathrm{ref}/
  3. SS

Here Wmin=min⁡(W+,W−)W_\mathrm{min}=\min(W^+,W^-) is the smaller section modulus. This chapter covers the first step. The second is described in the chapter Safety factor and assessment limits.

Why the stress is linearly distributed

The derivation follows the Euler–Bernoulli model. The coordinate dd measures the signed distance from the stress-free neutral axis in the direction of the normal n\boldsymbol{n}; the case shown is that n\boldsymbol{n} points in the positive ww direction. Plane cross-sections remain plane. The longitudinal strain therefore varies linearly with dd and with the curvature κ\kappa.1

ε(d)=−κ d\varepsilon(d)=-\kappa\,d
(1)

In the linear-elastic material, the stress is proportional to the strain. It is therefore also linearly distributed.1

σ(d)=E ε(d)=−E κ d\sigma(d)=E\,\varepsilon(d)=-E\,\kappa\,d
(2)

Without an external axial force, the resultant axial force is zero: tensile and compressive stresses must cancel over the cross-section. For a homogeneous cross-section, it follows that the neutral axis coincides with the centroidal axis.1 At the same time, the moment of the stresses must correspond to the bending moment. This yields the well-known relation M=EIκM=EI\kappa, and with it κ\kappa can be eliminated from equation 2.1

σ(d)=−M dI\sigma(d)=-\frac{M\,d}{I}
(3)

For a given bending moment, equation 3 does not contain Young’s modulus: the elastic bending stress then depends only on the moment and the cross-section geometry.1 The material enters elsewhere: EE governs deformation, while the strength value provides the reference for the assessment. The sign follows the SpanDraft convention: for a positive, sagging moment, the side with positive dd is in compression.

T-section with centroidal axis y and coordinate d. Next to it, the strain and the stress over the height: both are zero in the centroidal axis, vary linearly and reach the larger magnitude at the lower extreme fiber, which is farther away.
Figure 1: Strain and stress over the height of the T-section for a positive moment.

Extreme fibers and section moduli

Because the distribution is linear, the stress magnitude grows with the distance from the centroidal axis. It is largest at the extreme fibers, that is, where the cross-section reaches its greatest distance on each side of the axis. This can be a single point or an entire edge. The distances c+c^+ and c−c^- and the two section moduli of the selected bending axis are determined in the chapter Cross-section geometry and properties.

W+=Ic+,W−=Ic−W^+=\frac{I}{c^+},\qquad W^-=\frac{I}{c^-}
(4)

Inserting the edge distances into equation 3 yields the two extreme fiber stresses with their sign:

σ+=σ(c+)=−MW+,σ−=σ(−c−)=+MW−\sigma^+=\sigma(c^+)=-\frac{M}{W^+},\qquad \sigma^-=\sigma(-c^-)=+\frac{M}{W^-}
(5)

W+W^+ and W−W^- thus belong to the geometric sides of the cross-section. They do not stand for tension and compression.

The governing extreme fiber

If the edge distances are equal on both sides of the bending axis, for example for a cross-section symmetric about this axis, W+=W−W^+=W^- holds. If they differ, as for the T-section about the axis y-yy\text{-}y, the extreme fiber that is farther away carries the larger stress magnitude. Since both section moduli contain the same II, the smaller WW belongs to it. This holds independently of the sign of the moment: a change of sign swaps tension and compression but not the magnitudes (Figure 2).

The same T-section twice with stress distribution. For a positive moment, the top side is in compression and the bottom side in tension, for a negative moment the reverse. The magnitudes remain the same, the larger magnitude always lies at the marked lower extreme fiber.
Figure 2: Stress distribution in the T-section for positive and negative moment. The extreme fiber that is farther away is marked.

SpanDraft therefore determines the maximum elastic bending stress, the largest stress magnitude in the beam, directly with the smaller of the two section moduli:

σb,max=∣M∣maxmin⁡(W+,W−)\sigma_{\mathrm{b,max}}=\frac{|M|_\mathrm{max}}{\min\left(W^+,W^-\right)}
(6)

Analytical

Beam theory provides a stress with sign for every location and every extreme fiber. Tension and compression can thus be stated separately.

SpanDraft

SpanDraft reports exactly one non-negative value: the largest stress magnitude in the beam. The result contains no assignment to tension or compression or to a specific side of the section. The orientation of the section in the bending plane is not part of the model.

For manually defined cross-sections, only one section modulus is entered per axis. SpanDraft then sets W+=W−=WW^+=W^-=W, and equation 6 becomes ∣M∣max/W|M|_\mathrm{max}/W. This is a model assumption and not a property derived from the geometry. The one bending axis that is selected for the beam model is always used.

The governing location

Material and cross-section are constant over the beam length. The extreme fiber stresses are therefore everywhere proportional to the moment, and the largest stress magnitude lies where the largest moment magnitude lies. SpanDraft needs no search of its own for this. The solver already knows the largest and the smallest bending moment as analytical extreme values, and the evaluation selects the one with the larger magnitude from them.

∣M∣max=max⁡(∣Mmin∣, ∣Mmax∣)|M|_\mathrm{max}=\max\left(|M_\mathrm{min}|,\,|M_\mathrm{max}|\right)
(7)

There is thus neither a second calculation nor additional evaluation locations or values from the diagram. Value and location come unchanged from the selected extreme value of the solver, including the side at a jump location. “Left” and “right” denote the side along the beam axis and not the upper or lower extreme fiber.

Figure 3 shows this for a beam with an overhang: the support moment with the larger magnitude determines the maximum bending stress, the smaller span moment remains below it.

Beam with two supports, overhang and two point forces. Below, the moment curve with a positive span moment and a support moment of larger magnitude that is negative. At the bottom, the curve of the moment magnitude divided by the smaller section modulus: its maximum lies exactly below the support moment.
Figure 3: Moment curve and stress magnitude of a beam with an overhang from a SpanDraft calculation run. Hollow dot: span moment.
Deep dive: equal moment magnitudes

In symmetric systems, the magnitudes of the largest and the smallest moment can coincide numerically. To keep rounding residues from deciding the selection by chance, two magnitudes count as equal if they differ relatively by at most 64 ε\varepsilon, with the machine epsilon ε\varepsilon of about 2.2⋅10−162.2\cdot10^{-16}. This is no technical tolerance and no absolute threshold in SI units.

In such a tie, the smaller position counts and then the left side before the right. If these are also equal, the smallest moment remains selected. The stress practically does not depend on this choice, but the reported location does. In the unloaded case, moment and stress are exactly zero.

Example: T-section

The example is the single-span beam from the chapter Stations and equation system: L=4 mL=4\,\mathrm{m}, point force F=1300 NF=1300\,\mathrm{N} at 1.2 m. The span moment is F⋅a⋅b/L=1092 NmF\cdot a\cdot b/L=1092\,\mathrm{Nm}. Because the system is statically determinate, this value holds independently of the cross-section. New here is a T-section without fillet, with the flange on top, and the material S235JR.

T-section with the flange on top, 100 mm high and 100 mm wide, flange and web 10 mm thick, with centroidal axis y and the edge distances c plus and c minus.
Figure 4: T-section of the example, dimensions in mm.

The table compares the hand calculation with the result of a SpanDraft calculation run. The centroid lies 71.3158 mm above the bottom edge, thus close to the flange. The lower extreme fiber is therefore governing.

QuantityAnalyticalSpanDraft
AA in mm²19001900
II in 106 mm410^6\,\mathrm{mm^4}1.800041.80004
W+W^+ in 103 mm310^3\,\mathrm{mm^3}62.753862.7538
W−W^- in 103 mm310^3\,\mathrm{mm^3}25.240525.2405
∣M∣max|M|_\mathrm{max} in Nm10921092
σb,max\sigma_{\mathrm{b,max}} in MPa43.263943.2639

W−W^- is the smaller section modulus here and thus governing. If the direction of the force is reversed, the moment changes its sign, but σb,max\sigma_{\mathrm{b,max}} remains the same at 43.2639 MPa. The value would also remain the same for a section rotated by 180°: W+W^+ and W−W^- swap roles, the smaller one remains the same.

When tension and compression are to be assessed separately

The largest stress magnitude and the largest tensile stress can lie at different locations. This is shown by the beam from Figure 3 with the same T-section. It is 6 m long and rests at 0 and 4.5 m. At 2 m, 1000 N act downward, and at the end of the overhang 600 N. A hand calculation with equation 5 yields the following extreme fiber stresses at the two moment extrema; tension is positive.

LocationMM in NmTop in MPaBottom in MPa
x=2 mx=2\,\mathrm{m}711.11−11.33+28.17
x=4.5 mx=4.5\,\mathrm{m}−900+14.34−35.66
T-section at two locations. At x equal to 2 m: top minus 11.3, bottom plus 28.2. At x equal to 4.5 m: top plus 14.3, bottom minus 35.7.
Figure 5: Extreme fiber stresses in MPa at both moment extrema, tension is positive.

For this system, SpanDraft reports σb,max=35.66 MPa\sigma_{\mathrm{b,max}}=35.66\,\mathrm{MPa} at x=4.5 mx=4.5\,\mathrm{m}. This is the largest stress magnitude; according to the hand calculation, this fiber is in compression there. The largest tensile stress is 28.17 MPa and lies 2.5 m away.

For the comparison with a single strength value that applies equally to tension and compression, the magnitude is the appropriate quantity. If tension and compression are to be assessed separately, the elastic extreme fiber stresses can be determined as above from the moment curve and from W+W^+ and W−W^-, provided the orientation of the section in the bending plane is known. For brittle materials, adhesive bonds or the buckling of compressed thin sheets, this alone is not sufficient: separate failure criteria and verifications are required there.

What this value covers

σb,max\sigma_{\mathrm{b,max}} is the elastic normal stress from the bending moment about the selected axis. Shear stresses, local stress peaks and superpositions with other stresses are not contained in it. The classification and the specific limits are described in the chapter Safety factor and assessment limits. How the values appear in the application is shown in the help under Bending stress and safety factor.

Sources

  1. David Roylance. Stresses in Beams. Mechanics of Materials, Massachusetts Institute of Technology / MIT OpenCourseWare. 2000. Accessed .