The Bridge That Held 18 kg
In a high school engineering design class, teams build bridges from a fixed kit: 200 g maximum mass, 30 cm span, 5 cm width, white glue only. The test fixture loads the center of the deck until failure. The rubric awards full marks at 25 kg. One team’s bridge held 18 kg, then a glued gusset peeled away; the team wants to double every member and add glue the morning of testing. Another team reached 26 kg with a lighter truss. The teacher has one 50-minute period before the scheduled test and must keep the test fair for all teams. The school’s design cycle model requires students to define criteria, prototype, test, analyze, and iterate, but the district pacing guide allots only two prototypes. The team’s log shows they tested once at 10 kg, changed glue, and did not record drying time. The teacher suspects the failure is a joint detail, not overall member strength.
1. Using the design process, diagnose the most likely cause of the 18 kg failure and explain why doubling every member may not improve performance.
Read the worked answer
The log and failure mode point to a joint problem, not a member-strength problem. A gusset that peels away indicates the glue bond failed in shear or peel before the main members reached capacity. Doubling every member adds mass, which counts against the 200 g limit and increases dead load; it also may not change the force at the joint if the connection is still too small or poorly clamped. The team should inspect the failed surface: if the glue pulled cleanly from one surface, surface preparation or clamping pressure is suspect; if the wood itself tore, the bond may have been stronger than the wood. They should also check drying time, because the log shows they changed glue and did not record it. A better iteration is targeted: increase the bond area with larger gussets or lap joints, sand and clean surfaces, control glue amount, clamp for the full cure time, and reweigh. They should test one variable at a time. The 10 kg first test and 18 kg second test suggest progress, but without drying records the second result is not a clean comparison. Diagnosis should precede reinforcement.
2. What should the teacher do about the request to modify the bridge on the morning of testing, given fairness, time, and the two-prototype pacing limit?
Read the worked answer
The teacher should not allow one team to rebuild overnight while others do not, because that makes the test a contest of extra time rather than design learning. The fairest move is to separate testing from iteration. If the schedule allows, announce a common 25-minute iteration block for any team that wants it, with the same materials, scale, and mass check; then retest all revised bridges. If the schedule does not allow, test the bridge as built and treat the failure as the team's data for a required failure analysis and redesign proposal. The teacher can also allow a documented repair only if every team receives the same option and the repair is disclosed in the test record. The two-prototype pacing limit is a curriculum constraint, not a reason to give one team an advantage. The teacher should ask the team to predict the failure load of the doubled design and compare that prediction with the mass budget. This keeps the design cycle honest: iteration is assessed by the quality of the change and the evidence, not by whether the teacher rescues the team before the test.
3. How should the rubric be adjusted, if at all, to reward evidence-based iteration rather than only the highest load?
Read the worked answer
The rubric should not award full credit for load alone. Split it into: problem definition and criteria, prototype craftsmanship, test protocol and record keeping, failure analysis, iteration quality, and final performance. Final performance can be scored on load-to-mass ratio as well as absolute load, because the 200 g limit makes efficiency part of the design problem. A bridge that holds 18 kg at 120 g may be a better design than one that holds 20 kg at 199 g. Add a documentation criterion for drying time, clamp time, and test conditions, since the team's missing record is a real weakness. Add an iteration criterion that rewards a specific, evidence-based change rather than 'add glue everywhere.' Keep the 25 kg target as a benchmark, but do not make it the only path to full marks; otherwise students learn to overbuild and hide failures. If possible, test at least two specimens per team or report a range, because a single test is noisy. The teacher should state these criteria before the next test so the change is transparent.