Axial tension
Member pulled along its axis. Look for stretching, necking in ductile material, or tensile rupture, depending on material and loading.
Principles of Engineering • Unit 2 • Lesson 2.12
Structural Failure
Engineering question: What can failure evidence tell us about how a structure behaved—and how can we redesign it responsibly?
Your learning pathway
What you'll learn
Complete the resources in sequence. The on-page explanations, technical diagrams and practice are sufficient to complete the assigned case study independently.
01 · ClassifySeparate an applied loading mode from the way a component failed.
02 · InvestigateRecognize buckling, fracture, excessive deformation and connection failure using observations.
03 · JustifyRecommend redesign actions and explain how to verify them with evidence.
Suggested 45–50 minute learning sequence
Classify the forces8 minutes
Watch and take notes10 minutes
Explore column buckling12 minutes
Diagnose case studies15–20 minutes
The historical engineering investigation and extended reflection may be assigned as homework.
01 / First identify what is acting
Five loading modes engineers investigate
Loading mode describes the internal action or applied force effect. Failure mechanism describes how a structure stops meeting a requirement or loses its ability to carry load.
Axial compression
Member pushed along its axis. A short member may crush; a slender member may become unstable and buckle first.
Shear
Opposing parallel forces tend to slide nearby planes relative to one another; see local tearing near a connector or sheared fastener.
Bending
A transverse load creates curvature and bending stress. Excessive deflection can make a beam unusable even without rupture.
Torsion
A turning couple twists a shaft; it creates shear stress and can cause twisting deformation or fracture.
02 / What does failure look like?
Four common structural failure mechanisms
Observe the difference between loss of stability, material rupture, permanent/excessive deformation and failure at a connection.
Instability
Slender compression member bows sideways. A member can lose stability before its material crushes.
Material failure
Crack, split, rupture or permanent yielding appears in the loaded material. The failure mode depends on stress and material behavior.
Connection failure
Adhesive separates, a gusset tears, a fastener slips, or wood splits near a joint. The members themselves can be intact.
Serviceability is different from collapse
A shelf can sag too far to use, a walkway can vibrate excessively, or a beam can deform beyond its allowable limit without immediately fracturing. Structural design includes both strength and serviceability requirements.
For your bridge challenge, passing the 100 N load is not enough; the deflection must also remain at or below 10 mm under that load.
Observed sign ≠ proven root cause
A broken gusset could result from undersized adhesive area, poor surface preparation, an off-axis load, a brittle substrate or a combination. A photo of the break does not automatically identify the single cause.
Good failure analysis compares competing hypotheses to measurements and evidence.
03 / Video learning library
Watch and annotate real engineering explanations
Begin with column buckling and beam stresses. The short case-study video is optional for the historical investigation and should be viewed respectfully.
The Efficient Engineer
Understanding Buckling
Pause and explain: Sketch a straight and a buckled compression member, then explain why doubling unsupported length reduces the ideal Euler load.
Watch directly on YouTube ↗The Efficient Engineer
Understanding Stresses in Beams
Pause and explain: Label tension and compression in a bent beam, and explain why shear may also occur.
Watch directly on YouTube ↗Optional case-study video: Hyatt Regency walkway collapse and connection design
Tom Scott with Practical Engineering
The Disaster That Changed Engineering: The Hyatt Regency Collapse
Pause and explain: Compare the two hanger-rod arrangements and name the connection whose load path changed.
Watch directly on YouTube ↗Viewing context
This video discusses a real fatal engineering disaster without requiring graphic imagery. Study the design and review decisions—not the human loss as entertainment.
04 / Interactive compression investigation
What makes a slender member more likely to buckle?
In this illustrative classroom model, apply Euler's elastic buckling equation to a rectangular column. Vary length, thickness, modulus and idealized support condition. The model compares the applied force to the theoretical elastic threshold—not a tested failure load.
18.00 mm⁴Weak-axis I = bt³/12
49.35 NIdeal Euler critical load
0.81Applied force ÷ Euler load
180 mmEffective length KL
The strip width b is held at 8 mm. Effective E values are hypothetical—not certified values for classroom wood.
Pcr = π² EI / (KL)²
Applied load is below the model's theoretical elastic Euler threshold. That does not prove the real column safe.
Model limitations: Euler theory describes elastic instability of an ideal straight, slender column. It ignores crookedness, load eccentricity, joints, real boundary stiffness, material defects, inelastic behavior and local buckling. The curved drawing is an exaggerated illustration, not a calculated deformation or a failure animation. Even below Pcr, real members can fail for other reasons. No student should physically load a column to its predicted failure point without an instructor-managed safety setup.
Double the unsupported length
Hold E, b, t and K constant. What happens to the theoretical elastic buckling force if length doubles?
Reveal the explanation
Because Pcr ∝ 1/L², doubling unsupported length reduces the theoretical Euler critical load to one-quarter.
Double the thickness
For the same bending axis and fixed width b, what happens to I and the theoretical buckling force if thickness doubles?
Reveal the explanation
I = bt³/12. Doubling thickness multiplies I by 8, so the ideal elastic buckling load also multiplies by eight if E, L and K remain unchanged.
05 / Interactive failure detective
Diagnose the evidence—not just the shape
Select a case, read the observations, and form a hypothesis before revealing the engineering explanation. These are fictional instructional scenarios, not proven reconstructions.
Compression member
Case A · Bowed chair leg
Observed: A long narrow wooden chair leg develops sideways bowing when vertically loaded. No visible splitting is present near the middle at first.
What is known?
The member is slender and loaded in compression; it moves sideways.
What is uncertain?
End restraint, wood defects, off-center load, and whether crushing also developed.
Evidence to collect
Length, section dimensions, load location, video of lateral displacement and end connections.
Reveal possible mechanism and redesign
Likely mechanism: Compression buckling. Increase the weak-axis section stiffness, shorten unbraced length or brace the leg. Verify by comparing otherwise identical tests and measuring the load at onset of lateral deflection.
06 / Fully worked forensic example
From a symptom to a defensible redesign
Scenario: a 200 mm-long, 8 mm-wide, 3 mm-thick ideal wooden compression strip bows sideways at a load where no obvious crack has formed. Use illustrative E = 9,000 N/mm² and pinned–pinned ends.
1 · State evidence and assumptions
Observation: lateral bowing during axial compression.
Hypothesis: elastic instability could contribute.
Assumptions: ideal straight elastic pinned member; 8 × 3 mm uniform rectangle, buckling about the weak axis, E = 9,000 N/mm².
I = 8(3³)/12 = 18 mm⁴
2 · Calculate a theoretical load
Pcr = π²(9,000)(18)/(200²)
Pcr ≈ 39.97 N for this simplified model.
If unsupported length becomes 100 mm with all other inputs unchanged, Pcr ≈ 159.89 N, four times the original theoretical value.
3 · Interpret and propose a testable change
A shorter free column length has a much larger theoretical Euler load. Recommend a physical lateral brace or shorter unsupported segment and measure onset of lateral movement under instructor-controlled testing.
Do not claim a fourfold real-world safety margin. Brace effectiveness, connection behavior and wood imperfections are not checked by this equation. If a real chair has become unstable or damaged, remove it from service; do not experiment with people sitting on it.
07 / Historical engineering investigation
A change in connections can change the entire load path
On July 17, 1981, the second- and fourth-floor suspended walkways at the Hyatt Regency hotel in Kansas City collapsed. The National Bureau of Standards investigation identified insufficient box-beam/hanger-rod connection capacity. An approved change from continuous rods to a two-rod arrangement increased the force carried by the upper connection, approximately doubling that connection demand for the two walkways.
Conceptual load-path comparison
Not-to-scale teaching schematic. This represents the principle of the altered support path, not fabrication detail, geometry or the complete structural investigation. Actual connection force changes varied by location and load case.
What the investigation showed
The modified rod arrangement placed the lower walkway's load onto the upper walkway box beam connections. Investigators found that those connections had inadequate capacity; the original connection design had weaknesses as well.
Engineering responsibility
A seemingly small manufacturing change can alter the load path and required connection capacity. Revised drawings must be reviewed and verified by appropriately responsible professionals, rather than assumed equivalent.
08 / Independent engineering task
Three failure cases and three redesign decisions
Analyze the chair leg, bending shelf and separated glued corner described in the Failure Detective. Keep these examples distinct from the later full bridge manufacturing project.
1 · RecordWrite only observable facts
2 · DiagnoseState forces and possible mechanism
3 · RedesignPropose one controlled improvement
4 · ValidateSpecify a safe comparative test
| Scenario | Observation to begin with | What you must decide |
|---|---|---|
| A · Chair leg | Long slender member bows sideways under axial compression. | What evidence distinguishes buckling from crushing? Which geometric change targets instability? |
| B · Shelf | A supported shelf deflects excessively under books but shows no visible crack. | Is this a serviceability problem, a strength failure, or both? How can you test the redesign? |
| C · Glued corner | Two wooden members separate at a glued corner; neither long member visibly snaps. | What connection hypotheses remain? How will you distinguish adhesive failure from wood splitting? |
Practical safety: Do not deliberately break furniture, climb on bridges, or apply unknown loads to classroom structures. Physical tests in Lesson 2.13 require teacher-selected specimens, safety glasses, shields and approved loading apparatus. Keep hands and faces away from loaded members.
09 / Independent checks
Try first; reveal answers second
Show the observation, force mode, possible mechanism and remaining uncertainty in your response.
Load mode vs. failure
A slender vertical member bends sideways when compressed. Name the loading mode and likely failure mechanism.
Reveal the explanation
Loading mode: axial compression. Possible mechanism: buckling instability. Sideways bowing is evidence, but the precise cause must be tested.
What failed on the shelf?
A shelf bends 14 mm under books, above a stated 6 mm limit, yet returns to its original shape when unloaded. Is this automatically fracture?
Reveal the explanation
No. It fails the serviceability deflection limit but does not show fracture or permanent yielding from that description.
Joint failure vs. material failure
A glued corner opens under load but both long wooden members remain intact. Does this prove the wood itself was weak?
Reveal the explanation
No. The evidence points first to the connection/load transfer. Surface preparation, glue bond, grain direction, geometry and loading must be investigated.
Buckling length
With unchanged E, I and K, a pinned compression member's unsupported length doubles. How does theoretical Euler load change?
Reveal the explanation
It becomes one-quarter of the original value because Pcr ∝ 1/L².
Thickness scaling
With fixed 8 mm width and weak-axis rectangular bending, the 3 mm thickness doubles to 6 mm. How does I change?
Reveal the explanation
I = bt³/12, so I becomes eight times greater. This predicts eight times the elastic Euler load only if other model assumptions stay valid.
Loading direction
An M3 bolted plywood gusset has a hole very close to the edge. What potential failure mechanisms should you check?
Reveal the explanation
Potential risks include wood splitting, edge tear-out, bearing damage, net-section rupture and bolt/connection shear. A real connection needs geometry, material and load checks.
Evidence or inference?
You see a crack in a beam and write: “The material was too brittle.” Which part is observation and which is inference?
Reveal the explanation
Observation: a crack exists. Inference: the material was too brittle. Test competing causes including overload, defect, fabrication damage and load direction.
Design review
Why should a structural connection be rechecked when a shop drawing changes the way two members attach?
Reveal the explanation
Because the load path, connection forces and possible failure modes may change even if the overall member layout looks similar.
Reference library
Reliable engineering resources
These references support the videos and include deeper explanations, equations and documented investigation findings.
Understanding Buckling
Visual reference explaining why slender compression members become unstable and what effective length represents.
Read the Euler buckling explanation ↗Strength, Ductility and Toughness
Review the difference between yield, fracture, ductility and energy absorption from Lesson 2.11.
Review stress–strain behavior ↗NIST historical investigation
Primary engineering investigation of the 1981 Hyatt Regency suspended-walkway connection failures.
Read the NIST investigation summary ↗Materials Testing Lab — Lesson 2.13
Next lesson: observe real controlled load, displacement and specimen behavior, comparing failure evidence to the predicted mechanisms.
Open the next lab ↗Student submission
Exactly what to submit
Submit a POE 2.12 Structural Failure case-study report or a teacher-assigned engineering notebook equivalent. The central deliverable is a reasoned analysis of three separate failure scenarios supported by observations and proposed tests.
Required evidence
- Three labeled case analyses: chair leg, shelf and glued corner.
- For each case, a small annotated sketch with loads, supports, observed damage/deformation and likely load path.
- For each case, the loading mode, hypothesized failure mechanism and at least one plausible alternative explanation.
- For each case, one targeted redesign recommendation and an instructor-safe test or measurement that would evaluate it.
- At least one numerical Euler buckling comparison with E, I, L and K listed, showing units and why the assumption matters.
- A short historical case response explaining how altered hanger-rod connections changed load transfer and what review should have occurred.
- A conclusion separating what was observed, what was calculated and what remains uncertain.
Check before submitting
- Loading modes are not confused with failure mechanisms.
- The deflection limit is distinguished from material rupture or collapse.
- The buckling calculation is labeled theoretical and not treated as tested safe capacity.
- Proposed fixes address the diagnosed mechanism instead of automatically adding more material everywhere.
- Connection analyses include joining details and a possible load path.
- All three cases, the historical case and any additional assigned questions are included.
Due date: Check the Daily Agenda and follow your teacher's submission procedure.
Review lesson ↑