Johann J Cardenas

Gear3D — Sources

Every number in this app’s library traces to something. This document says what, and — just as importantly — how strongly each number was verified during the build.

Read §1 before using any of these values in a publication.


1. Verification status — read this first

Three levels are used throughout the data files:

Level Meaning
verified Retrieved from the cited source during this build and checked figure-for-figure.
medium Well-established published engineering practice, reproduced from the standard literature, but not opened against a primary document during this build. Check before publication use.
low Representative of the vehicle type; not a specific model, not a cited figure. Adequate for illustrating a configuration, not for quoting.

What was verified during this build:

What was NOT verified during this build:


2. Primary sources

FHWA Traffic Monitoring Guide — Scheme F vehicle classification

AASHTOWare Pavement ME Design — default traffic inputs

23 CFR 658 — Truck size and weight, route designations

Yoder & Witczak, Principles of Pavement Design; Huang, Pavement Analysis and Design

Tire and Rim Association Yearbook


3. Truck geometry — what is derived and what is typical

Derived, and therefore reproducible:

Typical practice, confidence medium — check before quoting:

Low confidence — representative only:


4. Tire dimensions

Computed exactly from the designation — no table, no uncertainty:

Table lookup, in src/data/tires.json:

Size Confidence Note
11R22.5 high 279 mm section / 1054 mm OD. The reference truck tire of the flexible-pavement literature; dimensions agree across the major medium-truck data books.
11R24.5, 12R22.5, 10.00R20, 11.00R20, 9.00R20 medium Reproduced from the standard truck-tire literature. Not checked against a TRA yearbook during this build.

A size that is not in the table is reported by the app as unknown and refuses to produce geometry. It is never guessed. Adding one requires a source and a confidence field, and the test suite fails without them.

Static loaded radius. Two models are implemented and the choice is exposed:


5. Aircraft — what is sourced, derived and assumed

The aircraft library shipped in v1.2 with four Boeing aircraft spanning gear codes D, 2D and 3D: 737-800, 757-200, 767-400ER and 777-300ER. v1.6 adds the three wing-plus-body aircraft — 747-400, 747-8 (both 2D/2D2) and A380-800 (2D/3D2) — which had been deferred since the first build. Seven aircraft, six gear codes.

Those three are sourced by a different and better method than the original four, described in 5.5. Read that section before comparing numbers between them: on the original four the outer width is an input and the track is derived; on the three new ones the track is published and the outer width is a cross-check.

An earlier attempt was abandoned because the sources could not be reached (the FAA database returned 403 to the fetch tool, and the ACAP PDFs exceeded its size limit). Both were tooling limits, not access limits: the FAA site serves the spreadsheet normally to a browser user-agent, and the PDFs download fine with curl. Every number below was retrieved and read directly.

5.1 Authoritative — taken verbatim

This table describes the original four aircraft. The three wing-plus-body aircraft take their geometry from the manufacturer footprint figures instead — see 5.5.

Quantity Source
Gear designation (Main_Gear_Config) FAA Aircraft Characteristics Database
Wheelbase, nose to main gear FAA Aircraft Characteristics Database
Main gear outer width FAA Aircraft Characteristics Database
MTOW FAA Aircraft Characteristics Database
Maximum design taxi weight Manufacturer ACAP, section 7.2
Tire size and tire pressure, nose and main Manufacturer ACAP, section 7.2
Percent gross weight on the whole main gear (95 %) FAA AC 150/5320-6G, G.1.3

ACAP editions used: 737 D6-58325-7 Rev C (Oct 2025), 757 D6-58327 Rev H (Dec 2024), 767 D6-58328 Rev K (Dec 2024), 777 D6-58329-2 Rev G (Dec 2024).

5.2 The outer-width trap

The FAA field is not the centerline tread. Its own data dictionary defines Main_Gear_Width_ft as “Distance between outer tires in the main landing gear.” Treating it as the track would push every main wheel outboard by half a dual spacing plus half a tire — for a 777 that is nearly a meter per side, and the figure would look entirely reasonable while being wrong.

So the track is derived, never assumed:

track = outerWidth − (wheelsAcross − 1) × dualSpacing − sectionWidth

Section width comes exactly from the three-part tire designation.

5.3 The cross-check that makes this trustworthy

Nothing in that derivation uses the manufacturers’ separately published tread figures, so agreement between them is real corroboration rather than circularity. With the dual spacings recorded in the data files:

Aircraft Derived track Manufacturer published tread Difference
737-800 5727 mm 5715 mm (18 ft 9 in) +12 mm
757-200 7302 mm 7315 mm (24 ft 0 in) −13 mm
767-400ER 9302 mm 9296 mm (30 ft 6 in) +6 mm
777-300ER 10 963 mm 10 973 mm (36 ft 0 in) −10 mm

All four agree to within about 13 mm, on quantities of 6 to 11 meters — which is roughly what quoting a tread to the nearest inch can account for, and no more. test/run.mjs asserts both the derivation and this cross-check, at a 15 mm tolerance.

The 767-400ER was wrong until v1.6.1, and this table is how it showed

The 767 row previously read 9322 mm, 26 mm — twice any other residual, and the only one this rounding argument could not explain. It was recorded as “agreeing to within a few centimeters” and left alone.

It was not rounding. Its dual spacing was 1143 mm, taken as a round 45 in because no consulted document stated it. FAARFIELD 2.1.1 stores the 767-400 ER main gear wheel coordinates explicitly as X ±22.900 in, i.e. 45.800 in (1163 mm). Because the track is derived from the dual spacing, that error propagated straight into the geometry. Correcting it drops the residual from 26 mm to 6 mm and removes the outlier.

What actually moved is worth stating precisely, because it shows the derivation behaving as designed: the outboard tire of each dual pair did not move at all. The FAA outer width is the authoritative datum and is held, so a corrected dual spacing is absorbed inside it — the inboard tire moves 20 mm inboard and the strut centerline 10 mm, while the outer tire edge stays put. This is the same property 5.4 relies on when it says the outer width is preserved whatever dual spacing you enter.

Two things are worth taking from this beyond the number:

5.4 Assumed — declared, and shown in the app

On the original four aircraft, two quantities are not constrained by any source consulted:

Every aircraft unit lists these in assumedFields, the schema fails validation if that array is missing, and the app shows an amber notice naming them whenever an aircraft is loaded. Set them from FAARFIELD before using the output for pavement work — and note that changing a dual spacing re-derives the track, so the authoritative outer width is preserved whatever you enter.

Neither assumption applies to the 747-8 or the A380-800: their footprint figures publish every spacing, and both declare assumedFields: []. The 747-400 declares exactly one, NLG.tire.rimDiameter, because Boeing states its nose tire as 49X17 — a two-part Type VII designation that omits the rim. The overall diameter and section width come from the two published numbers and do not depend on that assumption.

5.5 Wing-plus-body gear — the aircraft that were deferred (v1.6)

The 747-400, 747-8 and A380-800 were left out of v1.2 through v1.5 with this reason recorded: their wing-plus-body layouts need the longitudinal and transverse offsets of the body gear relative to the wing gear, “a single outer width closes a two-strut layout; it cannot close a four-bogie one.”

That was correct, and it was a data problem rather than a modeling one. The offsets are not in the FAA database, and they are not in FAARFIELD either — FAARFIELD analyzes one gear at a time and stores the wing gear and the body gear as separate entries (B747-400 and B747-400 Belly), so it carries the bogie geometry but not the distance between the two.

They are stated plainly in the manufacturers’ own airport planning documents, which is where they came from:

Aircraft Document Figure
747-400 Boeing ACAP D6-58326-1 Rev E (Sep 2023) §7.2.1 landing gear footprint
747-8 Boeing ACAP D6-58326-3 Rev C (Aug 2023) §7.2 landing gear footprint
A380-800 Airbus AC A380, issue Nov 01/24 §7-2-0 footprint, sheets 1 and 2

The sourcing is inverted relative to 5.2. These figures publish the track, both gear positions and every spacing directly, so nothing is derived from an outer width. The outer width instead becomes an independent check:

Aircraft track + dual + section Published outer width Difference
747-400 10 998 + 1118 + 482.6 = 12 598.6 mm 41 ft 4 in = 12 598.4 mm (Boeing) 0.2 mm
747-8 10 998 + 1189 + 533.4 = 12 720.4 mm 41 ft 9 in = 12 725.4 mm (Boeing) 5 mm
A380-800 12 456 + 1350 + 530 = 14 336 mm 47.05 ft = 14 341 mm (FAA) 5 mm

The residuals are the figures’ own rounding — Boeing draws to the nearest inch, the FAA tabulates to 0.01 ft.

Independent corroboration from FAARFIELD. The FAA’s FAARFIELD 2.1.1 aircraft library stores explicit per-wheel coordinates. It agrees with the manufacturer figures on every bogie dimension:

Aircraft FAARFIELD wheel coordinates Manufacturer figure
747-400 X ±22 in, Y 58 in 44 in dual, 58 in tandem
747-8 X ±23.4 in, Y 56.5 in 46.8 in dual, 56.5 in tandem
A380 wing X ±674.37 mm, Y 1699.26 mm 1350 mm dual, 1700 mm tandem
A380 body X ±764.54 / ±774.70 mm, Y 0 / 1699.26 / 3398.52 mm 1530 / 1550 / 1530 mm dual, 1700 mm tandem

Two wholly independent sources agreeing to the millimeter, including the A380 body bogie’s 20 mm wider middle axle — which is why dualSpacingByRow exists rather than the spacing being averaged. FAARFIELD’s per-strut MgPercent also reproduces the load split from the other direction: 0.2375 × 4 struts = 95 % for both 747s, and 0.19 × 2 wing + 0.285 × 2 body = 95 % for the A380, with both giving an equal 4.75 % per tire.

Two disagreements worth knowing about

The FAA’s tabulated wheelbase is not defined consistently for these aircraft. For the 747-8 it equals the centroid of the four bogies (97.3 ft against a computed 29 655 mm, agreeing to 2 mm). For the A380 it equals the nose-to-body-gear dimension (104.6 ft = 31 881 mm exactly). For the 747-400 it matches neither, nor their midpoint — it gives 87.9 ft where Boeing’s figure gives 84.0 ft, the commonly published 747-400 wheelbase. This library uses the manufacturer figures and treats the FAA field as a cross-check only.

The FAA’s MTOW for the 747-400 is the -400ER figure (910 000 lb), which exceeds the maximum taxi weight in Boeing’s own table for the -400. The manufacturer table is used instead: MTOW 875 000 lb, MDTW 877 000 lb, from the same column of ACAP §2.1.1.

What is still not included

Nothing from the original build spec. Additional weight variants (the A380 alone has fifteen) are not separate entries because the geometry is identical across them — only the weights differ, so a variant changes tire loads and nothing else. A380-800 WV000 is the one carried, and its variant is stated on every weight.

5.6 Tandem spacings, retired as assumptions (v1.9)

The 757-200, 767-400ER and 777-300ER shipped from v1.2 with MLG.tandemSpacing declared in assumedFields: no consulted source constrained it, so a plausible round number in inches was chosen and said to be chosen. The FAARFIELD 2.1.1 aircraft library does constrain it.

Aircraft Assumed FAARFIELD Error
757-200 1143 mm (45 in) 1143 mm (Y ±571.5) none
767-400ER 1422 mm (56 in) 1371.6 mm (Y ±685.8) 50.4 mm too long
777-300ER 1448 mm (57 in) 1463.04 mm (Y −1463.04 / 0 / +1463.04) 15.0 mm too short

All three are now sourced and out of assumedFields. Only NLG.dualSpacing remains assumed on the two-strut aircraft; FAARFIELD models the main gear only, because the nose gear carries too little load to matter to thickness design.

Why a wrong value survived seven releases. A tandem spread is symmetric about the bogie center, so changing it moves both axle lines equally and leaves the wheelbase (a centroid), the track and the outer width completely untouched. Every derivation check in the suite passed with the wrong number in place. This is the argument for declaring assumptions in the data rather than trusting tests to find them: the tests could not have.

5.7 FAA Order 5300.7 gear configurations (v1.9)

Sixteen configurations in aircraft/faa-5300-7.json. Every one is schematickind: "schematic", flagged in the app — and the split between what is measured and what is nominal is not the same as it is for the real aircraft above.

Verified, and cited per gear. The wheel geometry — track, dual spacing, tandem spacing, uneven bogie offsets, wing-to-body offset — of the ten configurations whose representative aircraft appears in the FAARFIELD 2.1.1 library: S (F-15C), 2S (C-130), 2T (C-17A), 2D/D1 (DC10-30/40), 2D/2D1 (A340-600 WV000), 5D (An-124), 7D (An-225), C5 (C-5), D2 (B-52), Q2 (IL-76T). Gross weights and tire pressures come from the same entries. This is the library already used in §5.3 to corroborate the 747 and A380, read directly during this build.

Nominal, and declared in assumedFields on every unit:

No outer width is stated on any of them. On the real aircraft the FAA’s published outer width is the datum and the track derives from it (§5.2). Here the track is measured and the outer width would depend on the nominal tire, so stating one would present a placeholder as a datum. mainGearOuterWidth is null throughout and a test asserts it.

Low — the six pure patterns. T, Q, 2Q, 3S, 3T, 3Q have no aircraft behind them, or none whose geometry any consulted source publishes (Q’s representative, the HS-121 Trident, left service in 1985). They are drawn to one nominal scale — 49x19.0-22 tires, 1400 mm lateral pitch, 1450 mm longitudinal, strut centers 1800 mm outboard of the bogie half-width, wheelbase 20 000 mm — so that Figure 2’s twelve cells stay comparable, which is the point of that figure. Not one of those numbers describes an aircraft.

Two idealizations, both declared in the data. The C-17’s real bogie has its two rows offset laterally from each other by 38.1 mm, and one wheel in each row sits 292.1 mm out of line with the other two; both are squared up here and both are in the FAARFIELD coordinates if needed. Loads on all sixteen use the equal per-tire split that 95 % on the main gear implies, which is not how FAARFIELD apportions the wing/body aircraft (DC-10 78/17, A340-600 72/23); the geometry is sourced, the per-tire loads are nominal, and the unit notes say so.

Order 5300.7 itself — the naming convention, Table 1’s pressure codes, Table 3’s eighteen rows with their wheel counts, nose gear types, typical aircraft and the historic FAA/USAF/Navy concordance, and Figures 2–20 — was read in full during this build and is transcribed in src/core/gearcode.js. The test suite reproduces all eighteen published wheel counts from the names alone.


6. How to check any number in this app

Every axle carries a source string and every load a basis string, both shown in the properties panel when you select an axle. Export unit.json alongside any figure and the citations travel with it.

The test suite (npm test, 175 checks as of v1.9) fails the build if any axle, gear, load, GVW, MTOW, tire pressure or multi-axle group spacing lacks provenance, and includes a negative control that confirms the validator actually rejects a missing source rather than passing vacuously.