Ultimate Load of Beams
Using the ultimate moment $M_{\mathrm u}$ and plastic-hinge concept, find the ultimate load $F_{\mathrm{Pu}}$ of ideal elasto-plastic beams in determinate / indeterminate / continuous cases. Two tools: the static method (use equilibrium at the ultimate state) and the kinematic method (virtual work on the collapse mechanism) — both judge by "system becomes a mechanism" and give identical answers.
Click any item in the left Contents to jump · Click the green "Next" · Each animation has a ▶ button.
Determinate beam · Static & kinematic methods
Statically DeterminateAnimation 12.3.1: a prismatic simply supported rectangular-section beam under a midspan point load $F_{\mathrm P}$ growing from zero. The method that uses equilibrium at the ultimate state is the static method; alternatively, apply the virtual-work principle to the collapse mechanism — the kinematic method.
Elastic limit → plastic hinge → ultimate load
With increasing load:
- Outermost fibers at the midspan section yield first — the elastic-limit load is $F_{\mathrm{Ps}} = 4 M_{\mathrm s}/l$;
- A plastic zone forms at midspan (see Animation 12.3.1) and grows with load;
- Finally the midspan moment reaches $M_{\mathrm u}$ — a plastic hinge forms (see Animation 12.3.1) and the beam becomes a mechanism.
Static method Static
- Use moment equilibrium at the ultimate state;
- At each plastic hinge, $M = \pm M_{\mathrm u}$; draw the $M$ diagram and write equilibrium;
- Directly gives $F_{\mathrm{Pu}}$.
Kinematic method Kinematic / Virtual Work
- Impose a virtual displacement on the collapse mechanism;
- Use virtual work $W_{\mathrm{ext}} = W_{\mathrm{int}}$;
- For a simply supported beam with a single hinge: $F_{\mathrm{Pu}}\cdot(l/2)\cdot 2\theta = M_{\mathrm u}\cdot 2\theta$ gives $F_{\mathrm{Pu}} = 4M_{\mathrm u}/l$.
A determinate beam only needs a single plastic hinge to become a mechanism — $F_{\mathrm{Pu}}$ is tied to $M_{\mathrm u}$ by static equilibrium directly, no need to track hinge order or load history.
Single-span indeterminate beam · Animation 12.3.2
Indeterminate BeamAnimation 12.3.2: a fixed-pinned prismatic beam under uniform load $q$. Key property of single-span indeterminate beams: enough plastic hinges must form before the beam becomes a mechanism (collapses).
① Elastic stage · $M$ diagram (see Animation 12.3.2)
Elastic analysis gives $M_{A} = -q l^{2}/8$ at the fixed end, positive max near midspan $= q l^{2}/14.22$, and $M_{B} = 0$ at the pinned support. The fixed-end moment has the largest magnitude — it reaches $M_{\mathrm u}$ first.
② First plastic hinge at A
At $q \to q_{\mathrm s}^{A}$, the fixed end reaches $M_{A} = M_{\mathrm u}$ and the first plastic hinge forms. The beam now acts as a simply supported beam with moment $M_{\mathrm u}$ at $A$ — still can carry more load.
③ Second plastic hinge · mechanism forms
Further loading: some midspan section $C$ also reaches $M_{\mathrm u}$ — second plastic hinge; the system becomes a mechanism. From the ultimate-state $M$ diagram, $A$ carries $-M_{\mathrm u}$ and $C$ carries $+M_{\mathrm u}$.
Static-method solution
Kinematic-method solution
Using the collapse mechanism of the figure:
$W_{\mathrm{ext}} = W_{\mathrm{int}}$ gives $q_{\mathrm u} = 16 M_{\mathrm u}/l^{2}$ — identical result.
Two key properties of indeterminate structures' ultimate analysis
Key PropertiesAt first glance, ultimate analysis of an indeterminate beam seems to require tracking "how the plastic hinges develop". In fact, we can jump directly to the ultimate state — thanks to two key properties:
When computing the ultimate load of an indeterminate structure, the full elasto-plastic history is unnecessary, i.e. we do not need to know the hinge order or compatibility conditions — why?
Because by the time the last plastic hinge forms, the structure has already become statically determinate — temperature changes, support settlements, and other non-load effects produce no internal forces in a determinate structure, so they have no effect on the ultimate load either.
Property ① · Hinge order can be ignored
Just guess the collapse mechanism (which sections will yield), then apply the equilibrium conditions in the ultimate state to obtain the ultimate load. No need to track the elasto-plastic history, the order of hinge formation, or compatibility conditions.
Property ② · Temperature and settlement irrelevant
Temperature changes, support settlements, and similar "non-load" effects do not affect the ultimate load — because by the time the last hinge forms, the structure is already statically determinate, where these effects produce no internal forces.
These two properties mean that plastic analysis needs only to identify the final collapse mechanism — usually combinations of maximum-moment locations (support negative peaks + midspan positive peaks) — and then solve for $F_{\mathrm{Pu}}$ by equilibrium or virtual work. No elasto-plastic progression analysis is required. Plastic analysis is much simpler than it looks.
Example 12-1 · Fixed-pinned beam under uniform load
Worked ExampleFind the ultimate load $q_{\mathrm u}$ of the fixed-pinned prismatic beam of under uniform load. Section ultimate moment $M_{\mathrm u}$.
① Mechanism · expression for $M_{C}$
From the elastic $M$ diagram, the fixed end $A$ forms a plastic hinge first. Assume the second hinge lies at section $C$ a distance $x$ from the pinned support $B$. From the equilibrium we get the reaction at $B$:
So the moment at $C$ (distance $x$ from $B$):
② Maximum positive moment (stationary condition)
$C$ is the maximum positive-moment location — $\mathrm d M_{C}/\mathrm d x = 0$ gives
Rearranging: $q_{\mathrm u} = 2 M_{\mathrm u}\,/\,[\, l(l - 2x)\,]$.
③ Exact hinge location
Substituting back into the $M_{C}$ expression and simplifying: $x^{2} + 2lx - l^{2} = 0$, giving
Therefore
Example 12-2 · Ultimate load of a stepped beam
Stepped BeamFind the ultimate load $F_{\mathrm{Pu}}$ of the stepped beam of Animation 12.3.5. Segment $AB$ has ultimate moment $M_{\mathrm{u1}}$, segment $BC$ has $M_{\mathrm{u2}}$ with $M_{\mathrm{u1}} > M_{\mathrm{u2}}$.
① Possible plastic-hinge locations
For a stepped beam, since $AB$ and $BC$ have different ultimate moments, plastic hinges may form not only at the maximum-moment sections $A$ and $D$, but also at the step $B$ — because the $BC$ side has smaller $M_{\mathrm{u2}}$, so $B$'s $BC$ side is more easily capped.
② Mechanism with hinges at $B$ and $D$
If hinges form at $B, D$, both reach $M_{\mathrm{u2}}$ in the figure. The moment at $A$ is then
For this mechanism to be valid, $A$ must be able to carry this moment: $M_{\mathrm{u1}} \ge M_{A}$. Hence
③ Kinematic method
For the mechanism of the figure, the virtual-work equation is
The ultimate load:
For indeterminate stepped beams, do not assume hinges only at elastic maximum-moment locations — step sections (on the side with smaller $M_{\mathrm u}$) can reach the ultimate moment first. One must enumerate all possible mechanisms and pick the one yielding the smallest $F_{\mathrm{Pu}}$.
Collapse mechanisms of continuous beams
Continuous Beam MechanismsA multi-span continuous beam may collapse in two mechanism types: single-span mechanism (one span collapses alone) or combined mechanism (two or more spans collapse together).
Single-span mechanism Single-Span
A single span forms its own mechanism; adjacent spans remain rigid in place.
Applies when:
- the span carries larger load;
- the span has smaller $M_{\mathrm u}$;
- or adjacent spans are unloaded.
Combined mechanism Combined
Two (or more) spans form a mechanism together — plastic-hinge locations in adjacent spans are compatible.
Applies when:
- adjacent spans are loaded in the same direction;
- support negative moments of adjacent spans counteract;
- variable stiffness / load distribution favor it.
If all spans have uniform section and are loaded in the same direction, only the single-span mechanism is possible — no combined mechanism occurs.
Reason: under same-direction loads the support negative moment reduces the in-span moment on both sides, so if adjacent spans collapse jointly, one of them would exceed $M_{\mathrm u}$, violating the yield condition.
Unequal positive & negative ultimate moments
In practice (e.g. reinforced-concrete beams) the positive / negative ultimate moments may differ — $M_{\mathrm u}^{+}$ vs $M_{\mathrm u}^{-}$. Each plastic-hinge moment must be substituted individually in the mechanism equation:
Example 12-3 · Continuous beam
Continuous Beam ExampleFind the ultimate load of the equi-section continuous beam (spans $AB$ and $BC$, both carrying uniform load $q$). Section ultimate moment $M_{\mathrm u}$.
① AB span mechanism alone
By §12.3.6's rule (equi-section + same-direction loads), only the single-span mechanism is possible. $AB$-span mechanism: two plastic hinges — negative at $A$ ($M_{A} = -M_{\mathrm u}$) and positive at some interior point in $AB$.
By kinematic method with hinges at $A$, midspan, and $B$, let midspan deflection be $\delta_{1}$ and rotation $\theta_{1} = 2\delta_{1}/l_{1}$:
Giving $q_{\mathrm u}^{AB} \cdot l_{1}^{2}/4 = 3 M_{\mathrm u}\cdot (2\delta_{1}/l_{1})$, i.e.
② BC span mechanism alone
Similarly:
③ Take the smaller as the true ultimate load
Take the smaller of the two — i.e. the longer span fails first:
Once that span collapses, the whole continuous beam loses capacity — so the real ultimate load is the smaller of the two — a direct application of the minimum theorem (rigorously proved in §12-4).
- Determinate beams' ultimate load = $M_{\mathrm u}$ divided by a geometric factor — differs from the elastic limit by the shape factor $\alpha$;
- Indeterminate beams need enough plastic hinges to become a mechanism;
- Plastic-hinge location obtained from $\mathrm d M/\mathrm d x = 0$ — exact value for a fixed-pinned beam under uniform load is $x = 0.414\, l$;
- Two key properties: hinge-order irrelevance and temperature/settlement irrelevance;
- Stepped beams' plastic hinges can occur at step sections; enumerate possible mechanisms;
- Equi-section continuous beams with same-direction loads collapse only in the single-span mode; the longest span collapses first;
- Next, §12-4 gives the upper / lower / uniqueness theorems for proportional loading.