Setup Lab
Crossbind
Check whether front and rear axle references share a plane, then separate the geometry from diagonal-load clues and measured rolling effects.
ILLUSTRATED PLANE CHECK
Geometry and scale load are related but not identical.
load diagnostic ΔX = (LF + RR) − (RF + LR)
- h
- vertical separation in inches or millimeters
- ΔX
- difference between diagonal scale sums in lb-force or newtons
- state
- driver, steering, cable, kingpin, load simulator, and supports
There is no general conversion from ΔX to h without stiffness, support, and mass-distribution data. A load difference is not a measured geometric lift, and it does not prove increased total rolling resistance.
SEPARATE SYMPTOM FROM CAUSE
Use independent geometry and load checks.
- Verify the fixture plane and equal support heights.
- Define four matched references permitted by the plan.
- Measure lift without side force; remove and reset the car.
- Measure corner load separately with the exact race state.
- Change one permitted variable and repeat both methods.
Illustrative diagnostic, not a target
Diagonal sums of 103 lb and 97 lb give ΔX = 6 lb. If that survives resets, report a repeatable load imbalance. Test floor, pads, driver, ballast, and geometry before naming a cause.
GARAGE WORKSHEET
The minimum record needed to act on this guide.
QUESTION THIS STATION ANSWERS
Do four matched axle references occupy one plane, and is any separate diagonal-load clue repeatable?
BRING
- Rigid coplanar rail fixture or a spindle-mounted string system used to its instructions
- Matched support blocks, machinist level, feeler gauges or indicator appropriate to the fixture
- Four scales only as a separate load diagnostic
LOCK BEFORE READING
- Three seated reference points with no hand or string side force
- Fourth reference named in advance; same driver, cable, steering, and clamp state
- Safe supports that cannot roll, tip, or suspend the car unpredictably
WRITE THE RAW RECORD
- Geometric fourth-point gap h for at least three complete resets
- LF+RR and RF+LR from an independent scale process
- Fixture plane checks, support locations, contact direction, and whether the symptom survives a fixture reversal
DECISION GATE
Report geometric lift and diagonal load as different measurands. There is no general conversion between them. Correct only a repeatable car-following result and only through a method the current plan permits.
VERIFY NEXT
Restore the legal stack, recheck triangulation, then measure all four loaded spindle directions.
GEOMETRY FIRST
Diagonal load is one possible outcome, not proof.
Crossbind describes axle contact references that are not coplanar in the defined static setup. Depending on chassis stiffness, support height, driver and ballast placement, cable tension, and clamp state, a possible outcome is a repeatable difference between diagonal scale sums. That load clue does not by itself prove a twisted axle system or identify its cause.
For an idealized linear rolling model, Frr = Σ(Crr,i·Ni). If every contact has the same constant coefficient Crr, redistributing normal load N while total weight stays fixed leaves Σ(Crr·N) = Crr·W unchanged. Real Derby wheels, bores, bearings or bushings, axles, and steering can respond differently to load, clearance, runout, flex, and contact state, so cross-plane geometry may accompany additional loss. Measure the geometry and load independently, then use controlled roll or track tests before making a loss claim.
MEASUREMENT METHODS
Choose repeatability over cleverness.
RAIL FIXTURE
Parallel supports
Level two rigid rails side-to-side, front-to-back, and diagonally. Support matching axle points, lift one end slightly, and observe whether both sides touch together.
FOUR SCALES
Diagonal-load clue
On a truly level scale plane, compare LF+RR with RF+LR. Repeat after settling the car, and treat a surviving difference as a diagnostic that still needs independent geometry checks.
Sighting across axles and testing on a concrete floor can find large errors, but they are less repeatable. Wheels add diameter variation, so a fixture that references axle square stock or a legal loaded stand is preferable.
CORRECTION ORDER
Fix crossbind without losing the rest of the alignment.
- Confirm the fixture and floor are level.
- Confirm weight, driver posture, kingpin stack, and fastener state are repeatable.
- Measure crossbind before changing toe or cable tension.
- Use only the shims, washer arrangements, plates, or adjustments expressly allowed by the current build plan.
- Apply final torque slowly and remeasure; clamp load can change the result.
- Recheck spindle alignment and steering neutral after every correction.
The 2005 forum describes a particular feeler-gauge shimming method. Treat that as historical technique, not blanket permission. Today's plan and inspector determine what is legal.
SOURCES & FURTHER READING
Check the primary material.
- Zero Error forum archive: Crossbind Historical 2005 team discussion. Useful shop theory, not a current rule or official instruction.
- New Hampshire Soap Box Derby: Tips from the locals Licensed local-program notes on alignment, weights, tuning, and driver position.
- ISBD Rule Book - linked for the 2026 World Championship Official PDF currently linked from Race Week resources; the file itself says revised June 2025.
- OpenStax University Physics: Conditions for static equilibrium University-level reference for force and moment balance used in axle-load and center-of-gravity calculations.
- NIST Engineering Statistics Handbook: Uncertainties of calibrated values NIST guidance on uncertainty in future results corrected by a calibration curve, including calibration-curve and future-measurement contributions. It does not characterize the whole Derby measurement process by itself.