Machines & mechanisms

Multistage Gear Train Optimizer

Builds a compound spur gear train of 1 to 3 stages and gives its ratio, its error against an exact target ratio or output speed, the speed, turning sense and torque of every shaft with a mesh efficiency, the pitch-line speed, tooth force and centre distance of each stage, interference (undercut) and hunting-tooth checks, the gear layout and 61 phased animation frames. A bounded search of 324 candidate trains in exact integer arithmetic ranks every candidate, lists the top 20 and returns them all against their size with the trade-off front, and the continued-fraction convergents of the target give their smallest one-stage builds. Refuses invalid inputs plainly, warns of interference and of a search that misses the limits or the tolerance, and scores five challenges. For learning gear trains and integer ratio design.

Last updated

Calculator overview

Inputs and outputs

This summary comes from the calculator's published input and output contract.

Inputs

Number of stages (required)
About this input

N, the number of compound stages of your train and of the search: 1, 2 or 3.

Default 2 Range 1 to 3
Largest stage ratio (search) (required)
About this input

rmax, the largest ratio one searched stage may have.

Default 7 Range 2 to 10
Required output torque (required)
About this input

The torque in N m the output shaft must reach for the targets to be met.

Unit N m Default 250 Range 0 to 10000000
Pressure angle (required)
About this input

The pressure angle of the full-depth teeth: 14.5 degrees, 20 degrees or 25 degrees. It sets the interference limits.

Default 20 degrees Allowed 14.5 degrees, 20 degrees, 25 degrees
Ratio tolerance (required)
About this input

The ratio tolerance in %, in steps of 0.01%: a train is within it when its relative ratio error is at most this.

Unit % Default 1 Range 0.01 to 10
Gear train (row k is stage k; a row beyond the number of stages may be left blank) (required)
About this input

Your gear train: row k is stage k, with its driver teeth (on shaft k - 1), its driven teeth (on shaft k), each a whole number from 8 to 200, and its mesh angle from 0 to 360 degrees (the direction from the driver centre to the driven gear centre, counter-clockwise from +x). Rows 1 to N must be complete; a row beyond N may be blank or complete and is then kept but not used.

Default 3 rows
ColumnRange or allowed values
Driver teeth 8 to 200
Driven teeth 8 to 200
Mesh angle (deg) 0 to 360
Search window offset (required)
About this input

Shifts the drivers of the two- and three-stage search up by this many teeth from the smallest driver interference does not restrict.

Unit teeth Default 0 Range 0 to 30
Target ratio (required)
About this input

The target ratio, input speed over output speed, with at most three decimals. Used with the Ratio basis.

Default 30 Range 1 to 500
Target basis (required)
About this input

Ratio (the target ratio) or Output speed (the target output speed at the input speed).

Default Ratio Allowed Ratio, Output speed
Target output speed (required)
About this input

The target output speed in rpm. Used with the Output speed basis: the exact target is then 1000 x input speed over this, rounded half up to a whole number of thousandths.

Unit rpm Default 48 Range 0.01 to 10000
Input speed (required)
About this input

n in, the speed of the input shaft (shaft 0) in rpm.

Unit rpm Default 1450 Range 1 to 10000
Input torque (required)
About this input

T in, the torque on the input shaft in N m.

Unit N m Default 10 Range 0.01 to 100000
Hunting-tooth rule (search) (required)
About this input

Not required or Required: with Required every searched gear pair must share no common factor (a hunting-tooth pair).

Default Not required Allowed Not required, Required
Challenge number (required)
About this input

Selects a challenge: 0 for none, 1 to 5 for a predefined challenge whose settings replace the inputs it fixes.

Default 0 Range 0 to 5
Module (required)
About this input

m, the module in mm: a gear of z teeth has a pitch diameter of m z and an outside diameter of m (z + 2).

Unit mm Default 2 Range 0.5 to 10
Your predicted output speed (required)
About this input

Your predicted output speed for challenge 1, in rpm. Not used outside challenge 1 and never overridden.

Unit rpm Default 0 Range 0 to 10000
Your predicted turning sense (required)
About this input

Your predicted output turning sense for challenge 1: Not yet predicted, Same as the input or Opposite to the input. Not used outside challenge 1 and never overridden.

Default Not yet predicted Allowed Not yet predicted, Same as the input, Opposite to the input
Smallest tooth count (search) (required)
About this input

zmin, the smallest tooth count the search may use.

Unit teeth Default 17 Range 8 to 60
Largest tooth count (search) (required)
About this input

zmax, the largest tooth count the search may use (not below zmin).

Unit teeth Default 100 Range 20 to 150
Mesh efficiency (required)
About this input

The efficiency of one mesh in %: each stage passes on the torque in its tooth ratio times this.

Unit % Default 98 Range 50 to 100

Outputs

Ratio tolerance used
About this output

The ratio tolerance used, % (the challenge's when a challenge fixes it).

Unit %
Required output torque used
About this output

The required output torque used, N m (the challenge's when a challenge fixes it).

Unit N m
Pressure angle used
About this output

The pressure angle used (the challenge's when a challenge fixes it).

No unit declared
Target ratio, exact
About this output

The exact target ratio, a whole number of thousandths.

No unit declared
Candidate trains searched
About this output

The number of distinct candidate trains the search formed.

No unit declared
Candidates within the tolerance
About this output

The number of valid candidate trains within the tolerance.

No unit declared
What the search covers
About this output

What the search covers with these settings: one stage is complete without the hunting rule; two and three stages report the best found, not a guaranteed best possible.

No unit declared
First driver of the search window
About this output

The first driver of the two- and three-stage search window (the smallest driver interference does not restrict, plus the window offset); blank for one stage.

Unit teeth
Output power
About this output

The output power, in W. Blank while challenge 1 awaits a prediction.

Unit W
Output turning sense
About this output

The output turning sense: Same as the input for an even number of stages, else Opposite to the input (every external mesh reverses it). Blank while challenge 1 awaits a prediction.

No unit declared
Model version
About this output

The model version.

No unit declared
Formula module versions
About this output

The formula modules the model is built from, with their versions.

No unit declared
Candidates on the trade-off front
About this output

The number of valid candidates on the trade-off front of ratio error against size.

No unit declared
Power lost in the meshes
About this output

The power lost in the meshes: the input power less the output power, in W. Blank while challenge 1 awaits a prediction.

Unit W
Output speed
About this output

The speed of the output shaft, in rpm. Blank while challenge 1 awaits a prediction.

Unit rpm
Output torque
About this output

The torque on the output shaft, in N m, with the mesh efficiency at every stage. Blank while challenge 1 awaits a prediction.

Unit N m
Candidates within the limits
About this output

The number of candidate trains within the tooth limits, the stage ratio limit, free of interference and, with the hunting rule, hunting-tooth.

No unit declared
Train efficiency
About this output

The efficiency of the whole train, the mesh efficiency to the power N, in %.

Unit %
Train ratio
About this output

The train ratio of your train: the product of the driven teeth over the product of the driver teeth (input over output speed). Blank while challenge 1 awaits a prediction.

No unit declared
The top 20 candidate trains by rank (valid first, then ratio error, then size)
About this output

The top 20 candidate trains by rank (valid first, then ratio error, then size): the rank, the driver and driven teeth of each stage (unused stages blank), the ratio, the ratio error in %, 1 when within the tolerance, the size and the class.

No unit declared
Required torque reached
About this output

Yes when the output torque reaches the required output torque, else No. Blank while challenge 1 awaits a prediction.

No unit declared
Train within the tolerance
About this output

Yes when your train ratio is within the tolerance of the exact target (an exact whole-number test), else No. Blank while challenge 1 awaits a prediction.

No unit declared
Search window offset used
About this output

The window offset used (the challenge's when a challenge fixes it).

Unit teeth
Train ratio error
About this output

The relative error of your train ratio against the exact target, in %. Blank while challenge 1 awaits a prediction.

Unit %
Teeth in the train
About this output

The total number of teeth in the stages of your train in use.

Unit teeth
The stages of your train: teeth, ratio, speeds, turning sense, torques, pitch-line speed, tooth force, centre distance, interference and hunting
About this output

The stages of your train: one row per stage in use with the driver and driven teeth, the stage ratio, the speeds in and out in rpm, the turning sense out (1 the input's sense, -1 opposite), the torques in and out in N m, the pitch-line speed in m/s, the tooth force in N, the centre distance in mm, 1 when it interferes and 1 when it is hunting-tooth. The ratio, speed, sense, torque, pitch-line speed and tooth force columns are blank while challenge 1 awaits a prediction.

No unit declared
Target basis used
About this output

The target basis used (the challenge's when a challenge fixes it).

No unit declared
Number of stages used
About this output

The number of stages used (the challenge's when a challenge fixes it).

No unit declared
Largest stage ratio used
About this output

The stage ratio limit used (the challenge's when a challenge fixes it).

No unit declared
Ratio and torque targets met
About this output

Yes when your train is within the tolerance and reaches the required torque, else No. Blank while challenge 1 awaits a prediction.

No unit declared
Output speed of the exact target
About this output

The output speed the exact target ratio gives at the input speed, in rpm.

Unit rpm
Target output speed used
About this output

The target output speed used, rpm (the challenge's when a challenge fixes it).

Unit rpm
Target ratio used
About this output

The target ratio used (the challenge's when a challenge fixes it).

No unit declared
Challenge result
About this output

No challenge selected, MET or NOT MET.

No unit declared
Continued fraction of the target ratio: each convergent and its smallest one-stage build within the limits
About this output

The continued fraction of the exact target: one row per step in use with the step, the term, the convergent numerator and denominator, its error in %, and its smallest one-stage build within the tooth limits and free of interference (driver and driven teeth, blank when none).

No unit declared
Challenge: the value scored
About this output

The value the selected challenge scores: the output speed in rpm (1), the train ratio error in % (2 and 3), the output torque in N m (4) or the total teeth (5); blank when not valid or awaiting a prediction.

No unit declared
Challenge score
About this output

The challenge score: 100 when met, else 0.

No unit declared
Housing envelope height
About this output

The height of the box round the outside circles of the gears in use, in mm.

Unit mm
Housing envelope width
About this output

The width of the box round the outside circles of the gears in use, in mm.

Unit mm
Continued-fraction convergents
About this output

The number of continued-fraction convergents of the exact target.

No unit declared
Housing envelope area
About this output

The area of that box, in mm2.

Unit mm2
Teeth in the best train
About this output

The total teeth of the best train found. Blank when no searched train is valid.

Unit teeth
Best train found
About this output

The best train the search found, each stage as driver to driven teeth, stages joined by a semicolon. Blank when no searched train is valid.

No unit declared
Best train error
About this output

The ratio error of the best train found, in %. Blank when no searched train is valid.

Unit %
Best train ratio
About this output

The ratio of the best train found. Blank when no searched train is valid.

No unit declared
Challenge feedback
About this output

Feedback on the selected challenge: the prediction prompt, the met text or the not-met hint (blank with none).

No unit declared
Challenge goal
About this output

The goal of the selected challenge (blank with none).

No unit declared
Best train within the tolerance
About this output

Yes when the best train found is within the tolerance, else No. Blank when no searched train is valid.

No unit declared
Candidate cloud: every active search row, its ratio error (at least 0.0001% for the log axis), its size and its class (0 the best found, 1 on the trade-off front, 2 valid, 3 no allowed driven count, 4 not hunting)
About this output

The candidate cloud: 324 rows, one per search row; for each distinct candidate its ratio error in % (at least 0.0001 for the log axis; exact trains sit on that floor), its size in teeth and its class (0 the best found, 1 on the trade-off front, 2 valid, 3 no allowed driven count, 4 not hunting-tooth); rows that repeat a train or are out of range are blank.

No unit declared
Animation frames: 61 frames over one input revolution, the time and each shaft's angle (degrees, not wrapped)
About this output

The animation frames: 61 rows, frame j = 0 to 60 over one input revolution, each with the time in s and the angle of shafts 0 to 3 in degrees (not wrapped; shafts beyond the number of stages blank). Frames 1 to 60 are blank while challenge 1 awaits a prediction.

No unit declared
Layout plot, largest y
About this output

The largest y of the square plot area, in mm.

Unit mm
Layout plot, smallest y
About this output

The smallest y of the square plot area, in mm.

Unit mm
Layout plot, largest x
About this output

The largest x of the square plot area, in mm.

Unit mm
Layout plot, smallest x
About this output

The smallest x of the square plot area round the layout (10% margin), in mm.

Unit mm
Smallest tooth count used
About this output

The smallest tooth count used (the challenge's when a challenge fixes it).

Unit teeth
Model Status
About this output

NOT VALID with the reason when an input cannot be used; CHECK when the results are shown but a stage of your train interferes, no searched train meets the tooth limits, or the best train found misses the tolerance; otherwise OK. A train that misses the target or the torque is a result, not a warning.

No unit declared
Largest tooth count used
About this output

The largest tooth count used (the challenge's when a challenge fixes it).

Unit teeth
Mesh efficiency used
About this output

The mesh efficiency used, % (the challenge's when a challenge fixes it).

Unit %
Module used
About this output

The module used, mm (the challenge's when a challenge fixes it).

Unit mm
Hunting-tooth stages
About this output

The number of stages of your train whose two tooth counts share no common factor.

No unit declared
Time between frames
About this output

The time between animation frames: 1/input speed, in s (60 frames make one input revolution).

Unit s
Gear layout: each gear's shaft, plane, teeth, centre, pitch and outside radius and start angle
About this output

The gear layout: one row per gear in use (gear 2k - 1 the driver and gear 2k the driven gear of stage k) with its shaft, plane (stage), teeth, centre x and y in mm, pitch and outside radius in mm and the start angle of tooth 0 in degrees, phased so the teeth interleave.

No unit declared
Input torque used
About this output

The input torque used, N m (the challenge's when a challenge fixes it).

Unit N m
Stages that interfere
About this output

The number of stages of your train whose pinion has too few teeth for its gear at the pressure angle (undercut).

No unit declared
Hunting-tooth rule used
About this output

The hunting-tooth rule used (the challenge's when a challenge fixes it).

No unit declared
Input speed used
About this output

The input speed used, rpm (the challenge's when a challenge fixes it).

Unit rpm
This page is provided by LogicCommons for informational purposes only. Results are analysis outputs computed from the inputs you supply and are not professional advice. Verify all inputs and results independently.

LogicCommons is in beta. If a result, label, or reference looks wrong, tell us here; we read every message.

Methodology

Purpose and model boundary

This lab builds a compound train of external spur gears with 1 to 3 stages and compares it with a target ratio. Stage k has a driver on shaft k - 1 and a driven gear on shaft k; the driven gear of one stage and the driver of the next turn together on the same shaft. Shaft 0 is the input and the last shaft the output. For your train the lab returns the ratio and its error against the target, the speed, turning sense and torque of every shaft with a mesh efficiency, the pitch-line speed, tooth force and centre distance of each stage, interference (undercut) and hunting-tooth checks, the gear layout and its housing envelope, and 61 animation frames of the shaft angles over one input revolution. Beside it, a bounded search of 324 candidate trains with the same number of stages, in exact whole number arithmetic, ranks every candidate, lists the top 20 and plots them all against their size with the trade-off front marked. In Advanced the continued fraction of the target shows the best whole-number approximations behind the search. Five scored challenges use the same model.

It is a lab for learning gear trains and whole-number ratio design: how stage ratios multiply, why each external mesh reverses the turning sense, why a ratio such as 30 needs more than one stage, what each extra mesh costs in torque, and how small a pinion can be before its teeth interfere. It is not a gear design tool. It checks no tooth strength, bending or contact stress, bearing loss, lubrication or clashes between gears that do not mesh, and the two- and three-stage search reports the best train it found, not a proven best possible train.

The model is the calculation authority. The page sends the values you set to the model and draws only what it returns: the gears at their centres and sizes, the phased teeth turning frame by frame, the candidate cloud and trade-off front, and the candidate table. The page computes no ratio, speed, torque, interference test or search itself.

Inputs and units

Input Meaning and unit
Number of stages N, a whole number from 1 to 3 (default 2): the stages of your train and of the search.
Gear train Row k is stage k: driver teeth and driven teeth, each a whole number from 8 to 200, and the mesh angle from 0 to 360 degrees, the direction from the driver's centre to the driven gear's centre, counterclockwise from the +x axis. Rows 1 to N must be complete; a row beyond N may be blank, or complete and kept but not used. Default: 17 to 93 at 0 degrees, 17 to 93 at 60 degrees, and 17 to 17 at 120 degrees (used only with 3 stages).
Input speed From 1 to 10,000 rpm (default 1,450 rpm).
Input torque From 0.01 to 100,000 N m (default 10 N m).
Target basis Ratio or Output speed (default Ratio).
Target ratio Input speed over output speed, from 1 to 500 with at most three decimals (default 30). Used with the Ratio basis.
Target output speed From 0.01 to 10,000 rpm (default 48 rpm). Used with the Output speed basis.
Ratio tolerance From 0.01% to 10% in steps of 0.01% (default 1%).
Required output torque From 0 to 10,000,000 N m (default 250 N m).
Mesh efficiency The efficiency of one mesh, from 50% to 100% (default 98%).
Module m, from 0.5 to 10 mm (default 2 mm): a gear of z teeth has a pitch diameter of m·z and an outside diameter of m·(z + 2).
Pressure angle 14.5 degrees, 20 degrees or 25 degrees (default 20 degrees), for full-depth involute teeth.
Smallest tooth count (search) zmin, a whole number from 8 to 60 (default 17).
Largest tooth count (search) zmax, a whole number from 20 to 150 and not below zmin (default 100).
Largest stage ratio (search) rmax, a whole number from 2 to 10 (default 7).
Hunting-tooth rule (search) Not required or Required (default Not required). With Required every searched pair must share no common factor.
Search window offset A whole number from 0 to 30 teeth (default 0): shifts the drivers of the two- and three-stage search up.
Challenge number 0 for free exploration, or 1 to 5 to load a challenge.
Your predicted output speed From 0 to 10,000 rpm (default 0). Used only in challenge 1.
Your predicted turning sense Not yet predicted, Same as the input or Opposite to the input. Used only in challenge 1.

Your own train is not limited by zmin, zmax or rmax, and a stage may step up (driver larger than driven); the challenges check your teeth against their tooth limits.

Where each input is set:

  • On the canvas only: the gear train and the number of stages (tap a gear to select it, drag its rim to change its teeth in whole teeth, drag a driven gear round its driver to change the mesh angle in whole degrees, or press + and - on a focused gear; Enter opens exact entry), and your predicted turning sense for challenge 1.
  • In the input panel: every other input.
  • A candidate in the search table can be applied, which puts its teeth on the canvas.

Governing relationships

Ratio and error. With D the product of the driver teeth and Nn the product of the driven teeth over the stages in use, the train ratio is Nn / D, input speed over output speed. The target is kept as a whole number of thousandths, I: with the Ratio basis I = 1000 times the target ratio, and with the Output speed basis I = 1000 times the input speed over the target output speed, rounded half up. The ratio error is

error (%) = 100·|1000·Nn - I·D| / (I·D),

and a train is within the tolerance when |1000·Nn - I·D|·10000 ≤ (tolerance in hundredths of a percent)·I·D. Both sides are whole numbers, so the test is exact. The output speed of the exact target is 1000·(input speed) / I.

Speeds, turning sense and torque. The speed of shaft k is the input speed times the product of the driver teeth over the product of the driven teeth up to stage k. Every external mesh reverses the turning sense, so shaft k turns as (-1)^k times the input: the output turns the same way as the input for an even number of stages. With η the mesh efficiency as a fraction, the torque on shaft k is

T_k = T_in·(driven product / driver product up to stage k)·η^k,

the train efficiency is 100·η^N %, the output power is T_N·n_N·2π/60 (W), and the power lost in the meshes is the input power less the output power.

Per stage. The stage ratio is driven over driver teeth; the pitch-line speed is π·m·z_driver·n_in,stage / 60000 (m/s); the tooth force is 2000·T_in,stage / (m·z_driver) (N); the centre distance is m·(z_driver + z_driven) / 2 (mm). A stage is hunting-tooth when its two tooth counts share no common factor, so every tooth meets every tooth of its partner in turn.

Interference. For full-depth teeth and pressure angle φ, a pinion of z teeth meshes without interference with a gear of at most

N_G = (z²·sin²φ - 4) / (4 - 2·z·sin²φ)

teeth while z·sin²φ < 2; when z·sin²φ ≥ 2 the pinion meshes even with a rack (any gear). A stage of your train interferes when its smaller gear has too few teeth for its larger one by this relation. At 20 degrees this gives, for pinions of 13, 14, 15, 16 and 17 teeth, largest gears of 16, 26, 45, 101 and 1,309 teeth.

Layout and phase. Shaft 0 sits at the origin, and each stage's driven gear sits one centre distance from its driver in the direction of the mesh angle. Pitch radius m·z/2, outside radius m·(z + 2)/2. Each gear's start angle is phased so that a tooth space of the driven gear lies opposite a driver tooth on the line of centres, and the drawn teeth interleave. The housing envelope is the box round the outside circles of the gears in use.

Animation. Frame j = 0 to 60 covers one input revolution, 6 degrees of the input per frame, and the time between frames is 1 / (input speed) seconds. Shaft k turns by (-1)^k·6j times the driver product over the driven product up to stage k, from its start angle. Rolling without slip holds exactly at every mesh.

The bounded search. For each tooth count, the search first finds the largest allowed driven count: the smallest of zmax, rmax times the driver, and the interference bound above. Each searched stage is a reduction or 1:1, with teeth from zmin to zmax.

  • One stage: every driver from zmin to zmax with the two nearest allowed driven counts to I·z/1000. Without the hunting-tooth rule this is complete: it finds the most accurate allowed one-stage train.
  • Two stages: drivers from the smallest driver that interference does not restrict (plus the window offset) over six values on each stage; the first driven count near z₁·√(target), with nine offsets from -4 to +4; the closing driven count the exact nearest whole number that completes the target.
  • Three stages: drivers over six, six and nine values with an even split near the cube root of the target, and the exact closing count.
  • With the hunting-tooth rule, a pick that shares a factor with its driver is replaced by the second nearest count when that one is hunting-tooth.

A candidate is valid when every stage has an allowed driven count and, with the hunting rule, every pair is hunting-tooth. A row that repeats an earlier row's exact train is dropped. The candidates are ranked by valid first, then ratio error, then total teeth, then row; the top 20 are listed. A valid candidate is on the trade-off front when no other valid candidate has an error and a size both no larger, with one strictly smaller.

Continued fraction (Advanced). The Euclidean algorithm on (I, 1000) gives the terms a_k of the continued fraction of the exact target, and the convergents p_k / q_k from p_k = a_k·p_(k-1) + p_(k-2) and q_k = a_k·q_(k-1) + q_(k-2): the best whole-number approximations of the target. For each, the smallest one-stage build is the smallest multiple of q_k that reaches zmin, with the same multiple of p_k, shown when it lies within the tooth limits and is free of interference.

Calculation sequence

  1. Check the inputs in the order of the list under Errors and warnings.
  2. Take the values used: in a challenge, its own settings, except the inputs it leaves to you. The predictions are never replaced.
  3. Form the exact target I and the tolerance in hundredths of a percent.
  4. Build the tooth-limit table: the largest allowed driven count for every tooth count from 8 to 150, and the smallest driver interference does not restrict.
  5. Form the 324 search rows, their driven counts, ratios, errors, sizes and codes; drop repeats; rank them, mark the front and pick the best.
  6. Compute the continued fraction and its builds.
  7. Compute your train: ratio, error, shaft speeds, senses and torques, stage values, interference, hunting, layout, envelope and frames.
  8. Set the status and, in a challenge, score it.

Outputs and interpretation

Your train:

  • Train ratio, train ratio error (%) and train within the tolerance (Yes or No).
  • Output speed (rpm), output turning sense (Same as the input or Opposite to the input), output torque (N m), output power and power lost in the meshes (W), and train efficiency (%).
  • Required torque reached and ratio and torque targets met (Yes or No). A train that misses the target or the torque is a result, not a warning.
  • Teeth in the train, stages that interfere and hunting-tooth stages.
  • Housing envelope width, height (mm) and area (mm²), and the time between frames.
  • The stage table: teeth, stage ratio, speeds in and out, turning sense out (1 the input's sense, -1 opposite), torques in and out, pitch-line speed, tooth force, centre distance, interference and hunting.

The target: the exact target ratio (whole thousandths) and the output speed of the exact target.

The search: the best train found (each stage as driver to driven teeth), its ratio, error, size and whether it is within the tolerance; the number of candidates searched, within the limits and within the tolerance; the number on the trade-off front; the first driver of the search window; and a sentence on what the search covers. The top 20 candidates list each train's teeth, ratio, error, tolerance flag, size and class.

The continued fraction (Advanced): each step's term, convergent, its error and its smallest one-stage build.

Reading the canvas: the gears are drawn at the returned centres with the returned radii and turn through the returned frames, slowed down, with the real time labelled. The candidate cloud plots every searched train's error, on a log axis, against its total teeth: exact trains sit on the bottom line, labelled exact, and the trade-off front holds the trains no other train beats on both error and size. A train far to the left and low down is small and accurate.

Challenges

Every challenge fixes the speeds, torques, efficiency, module, pressure angle and tooth limits, and leaves the window offset free; challenges 2 to 5 leave you the number of stages and the gear train. The model scores each challenge MET (100 points) or NOT MET (0), shows the value it scored, and gives a one-line hint while it is not met. Challenges 2 to 5 also require every tooth count of your train to lie within the challenge's tooth limits and no stage to interfere.

  1. Predict the output. Predict the output speed (within 2%) and the turning sense of this two-stage train. The value scored is the output speed; it is met when your sense is right and your speed is within 2% of it.
  2. Factor the ratio. Build a train within 0.5% of 12:1 with at most two stages and teeth from 17 to 80. The value scored is the train ratio error.
  3. Hunting teeth at a motor speed. Reach 48 rpm from 1,450 rpm within 1% with every gear pair hunting-tooth (no common factor). The value scored is the train ratio error.
  4. Torque with the fewest stages. Deliver at least 270 N m within 1% of 30:1 from 10 N m, losing 4% at every mesh. The value scored is the output torque; it is met when both the ratio and the torque targets are met.
  5. The smallest train. Build a train within 0.5% of 30:1 with no more than 215 teeth in all. The value scored is the total teeth.

Challenge 1 holds back the answer until you predict the turning sense. While it reads Not yet predicted, the ratio, error, speeds, sense, torques, power and the target checks are blank, the stage table shows only the teeth, centre distances, interference and hunting columns, the frames after frame 0 are blank, and the challenge reads NOT MET with a prompt to predict first. The gears are drawn at rest so you can count their teeth. Your predictions change nothing outside challenge 1.

What to take from the lab: stage ratios multiply; a large ratio is easier to reach with several stages, but every mesh costs torque and reverses the turning sense; and whole-number teeth can only approximate most ratios, which is why the search reports an error.

Validation and status logic

The status reads the first match in the list under Errors and warnings. The NOT VALID checks run on the values you entered, also while a challenge is selected; the CHECK items run on the values used, also inside challenges. A challenge number that is not a whole number from 0 to 5 counts as no challenge selected.

Assumptions and limitations

  • External spur gears, 1 to 3 compound stages, full-depth involute teeth. No planetary trains, idlers or internal gears.
  • One efficiency per mesh; no bearing or churning loss. No tooth strength, bending or contact-stress check.
  • Interference by the standard minimum-pinion relation for full-depth teeth; no profile shift.
  • Gears in different stages may overlap in plan view; clashes of a gear with a shaft it is not on are not checked.
  • The one-stage search without the hunting rule is complete; with the hunting rule it tries the two nearest allowed counts per driver and can miss a hunting pair further from the target. The two- and three-stage search is a bounded pattern of 324 rows: the best train it finds, not guaranteed to be the best possible.
  • The ranking reads ratio errors exactly; equal errors fall to size, then to row. A value within about 1×10⁻⁹ (relative) of a threshold may flip a result.
  • The animation covers one input revolution in 60 frames and is played slowed down.

Restrictions and non-computing states

Under NOT VALID every result, table, layout and frame is blank; a selected challenge reads NOT MET with a score of 0, and its goal and hint stay shown. When no searched train is valid, the best-train outputs are blank and the status reads the CHECK for no train within the limits. In challenge 1, until you predict, the results are held back as described under Challenges.

Errors and warnings

The status messages, first match wins:

  • NOT VALID: the number of stages must be a whole number from 1 to 3
  • NOT VALID: stage k needs whole-number driver and driven teeth from 8 to 200 and a mesh angle from 0 to 360 degrees; a row beyond the number of stages may be left blank: k is the lowest row at fault.
  • NOT VALID: the input speed must be from 1 to 10,000 rpm and the input torque from 0.01 to 100,000 N m
  • NOT VALID: choose the target basis from the list
  • NOT VALID: the target ratio must be from 1 to 500 with at most three decimals, and the target output speed from 0.01 to 10,000 rpm
  • NOT VALID: the target output speed must give a reduction from 1:1 to 500:1 at this input speed: Output speed basis only.
  • NOT VALID: the ratio tolerance must be from 0.01% to 10% in steps of 0.01%, and the required output torque from 0 to 10,000,000 N m
  • NOT VALID: the mesh efficiency must be from 50% to 100%, the module from 0.5 to 10 mm, and the pressure angle one of the listed choices
  • NOT VALID: the tooth limits must be whole numbers, the smallest from 8 to 60 and the largest from 20 to 150 and not below the smallest, and the stage ratio limit a whole number from 2 to 10
  • NOT VALID: choose the hunting-tooth rule from the list, and give a window offset that is a whole number from 0 to 30
  • NOT VALID: the challenge number must be a whole number from 0 to 5, the predicted output speed from 0 to 10,000 rpm, and the predicted turning sense one of the listed choices
  • CHECK: the pinion of stage k has too few teeth for its gear at this pressure angle, so the teeth would interfere (undercut); use more pinion teeth, a smaller stage ratio or a larger pressure angle: k is the lowest stage of your train that interferes.
  • CHECK: no train in the search meets the tooth limits; widen the tooth range, raise the stage ratio limit or change the number of stages
  • CHECK: the best train found misses the ratio tolerance; try another number of stages, a wider tooth range or the window offset
  • OK: the inputs are valid and none of the above applies.

References

This laboratory is an educational model of a compound train of external spur gears. Its ratio is the product of the driven tooth counts over the product of the driver tooth counts, each external mesh reverses the turning sense, and each mesh passes on the torque times its tooth ratio and a mesh efficiency. A pinion's interference with its gear follows the standard minimum-pinion relation for full-depth involute teeth. A bounded search over whole-number tooth counts approximates a target ratio in exact integer arithmetic, and the continued fraction of the target, from the Euclidean algorithm, gives its best whole-number approximations.

Sources and further reading

  • Spur gear terms (pitch circle, module, pressure angle), ordinary and compound gear trains, the velocity ratio as driven over driver teeth, and the change of direction at each external mesh: Y. Zhang, S. Finger and S. Behrens, Introduction to Mechanisms, chapter 7, Gears, part of Rapid Design through Virtual and Physical Prototyping, Carnegie Mellon University. Carnegie Mellon University
  • Interference and undercut of involute teeth, and the smallest pinion that avoids them for full-depth teeth: University of Almería, Machine and Mechanism Theory, lesson 7, Gear Theory, The Interference Phenomenon. University of Almería
  • Gear types and geometry, and the power losses of gear meshes: J. J. Coy, D. P. Townsend and E. V. Zaretsky, Gearing, NASA Reference Publication 1152, NASA, 1985. NASA Technical Reports Server
  • The interference relation between pinion and gear tooth counts, and the design of gear trains to a ratio: R. G. Budynas and J. K. Nisbett, Shigley's Mechanical Engineering Design, McGraw-Hill, chapter 13, Gears: General.
  • The Euclidean algorithm, the Stern-Brocot tree of rationals and Brocot's method of approximating a ratio with whole numbers: R. Backhouse and J. F. Ferreira, "On Euclid's algorithm and elementary number theory", Science of Computer Programming 76(3), 2011, pages 160-180. arXiv:1506.05981

No table or text from any of these publications is reproduced, and none of their authors or publishers has reviewed or endorses this laboratory.

This laboratory is for learning. Its results hold only within its stated ranges; it checks no tooth strength, wear or lubrication, so a train from this lab is not a gearbox design.

Links were current at build time. If one moves, search the document's title rather than reporting a broken link as a defect.

Results are informational and not professional advice. See our Terms of Use. Powered by SpreadsheetWeb · About · Privacy · Terms