Skip to content
AbilityBench

Free, untimed, 20 items with the working

Mechanical reasoning test with the working shown

Twenty diagram items that want a number — the output speed of a gear pair, the pull that holds a block and tackle, the force out of the wide piston — and print the derivation in your own figures the moment you answer, free and with nothing to sign. There is no clock anywhere on the page, because a test whose purpose is that you leave understanding gear ratios cannot also be measuring how quickly you read a drawing. The report scores gears, pulleys, levers, fluid and balance separately, restates the single rule each one turns on, and puts the guessing floor beside your count: four options an item means five right out of twenty is what knowing nothing produces.

  • 100% free
  • No signup
  • 20 items, 5 topics
  • No clock
  • Working after every answer

Five topics, a number for an answer, and the derivation printed the moment you commit to one. Nothing here is timed: the item stays on screen until you answer it, and the seconds are recorded only so the report can tell you which topic slowed you down.

What each topic reduces to

Gears and ratio
Teeth per minute is what two meshed gears share, so revolutions divide in the ratio of the tooth counts and turning effect multiplies in the same ratio the other way.
Pulleys and rope
Count the rope sections that lift with the load. Each one carries a share of it, and the free end has to travel that many times as far.
Levers and arms
Load times its distance from the pivot equals effort times its distance. Which of the three sits in the middle is what names the class.
Fluid and pressure
Pressure is the same everywhere on one level in a connected fluid, so force out over force in equals area out over area in — and area goes as the square of the diameter.
Turning and balance
Nothing turns when the turning effects cancel: mass times distance on one side of the pivot equals mass times distance on the other.

Answer with keys 1 to 4 or the buttons. Every drawing is generated from a set number and described in words for a screen reader, so no answer depends on being able to see it.

How to work through the five topics

Answer, read the derivation, and use the topic medians to find what you are deriving rather than knowing.

  1. Start with five items, one from each topic

    The short run deals exactly one item from gears, pulleys, levers, fluid and balance, which is enough to find out which of the five you have no rule for. It is a map rather than a measurement: a single item cannot separate a gap from a slip, and the panel at the end says so. The twenty-item run gives each topic four, which is where the per-topic numbers start meaning anything.

  2. Answer with 1 to 4, then read the derivation before moving on

    Every drawing is to scale — a gear's radius is set by its tooth count, so the pair you are looking at could physically mesh, and a lever's arms are drawn at the ratio the question states. After you commit, the page names the right option and works it out using the numbers you were just given rather than a textbook example. The next item does not arrive until you ask for it.

  3. Read the topic medians, not just the topic scores

    The report gives right-out-of-asked and a median time for each of the five topics. The time is the more diagnostic half: a topic you answered correctly in forty seconds is one you are reconstructing from first principles, and a topic you answered correctly in eight is one you know. Both look identical in a score out of twenty, which is the reason this page splits them.

Technical specifications

Items20 in a full run — four each from gears, pulleys, levers, fluid and balance — or 5 in a short run, one from each topic. Four options per item, one of them right
Guessing floor25% by construction, so 5 of 20 or 1.25 of 5 is the expected count from picking at random. The result panel states this next to your score rather than leaving you to work out whether 8 out of 20 means anything
PacingNone. No countdown, no per-item window and no time added or taken; an item stays on screen until you answer it. Seconds are recorded and reported as a median per topic, which is deliberation time rather than reaction time
Constants and idealizationsWater taken as 1000 kg/m³ and gravity as 9.81 m/s². Friction, rope stretch and the weight of the blocks, beams and pistons are ignored, and the item that depends on that says so in its own wording rather than in a footnote
How the drawings are builtGear radius is proportional to tooth count, so the pitch matches across a pair and the mesh drawn is one that could turn. A tackle's rope is a single path with its dead end on whichever block keeps the hauling part at the top, which is why the section count is countable rather than decorative. Lever arms are drawn at the ratio the numbers state
What voids an itemOnly two things: the window losing sight of you while an item is open, which makes the recorded seconds a measure of your email rather than your thinking, and stopping the run early with an item still on screen. Nothing is discarded for a fast answer, because an item with no time limit has no such thing
ReportedRight out of asked with the guessing floor beside it; right, asked and median seconds for each of the five topics; the rule each topic reduces to; and an item-by-item review with the full derivation for every item you answered
What leaves the pageNothing. The set number rebuilds the same items with the same numbers, the result token carries four figures, and neither one contains an option you chose

Frequently asked questions

Why does a compound pulley halve the force?

Because two rope sections are holding the load up instead of one, and the rope cannot have different tensions along its length. Follow the rope from your hand: it passes over the sheaves and every section between the two blocks is the same piece of rope under the same pull. If two of those sections rise from the moving block, they share the weight between them, so each carries half and your hand only has to supply half. The price is distance — both sections have to shorten by the full lift, so you pull two meters of rope for every meter the load rises, and the work you do is exactly what it would have been lifting it by hand.

I took the ratio off the piston diameters and my answer was three times too small. Why?

Because force follows area, and area grows with the square of the diameter. A piston three times as wide has nine times the face for the pressure to push on, not three, so the same pressure produces nine times the force. This is the most reliably missed item in the set and it is missed in a specific direction: always short, always by the ratio itself. The same square catches people in reverse when the wide piston barely moves — it travels nine times less far, which is where the extra force is paid for.

Why are the pulley answers in kilograms rather than newtons?

Because that is how the question is asked on a shop floor and in every mechanical battery written for one. A kilogram is a mass and a newton is a force, so strictly a 60 kg load pulls down with about 590 N; but 'what pull holds this 60 kg crate' is answered in kilograms because gravity is the same on both sides of the division and cancels out. The item that does need real force units — the hydraulic press, where you are given a push rather than a weight — asks and answers in newtons, and the pressure item is in kilopascals.

Does ignoring friction make these answers wrong?

It makes them an upper bound on how well the machine can do, which is the number the questions are about. Every sheave a rope turns around costs some of the pull, so a real four-part tackle needs more than a quarter of the load and the shortfall grows with the number of sheaves — a rig with eight parts is nothing like eight times better than a single rope. The same applies to a lever on a stiff pivot and a press with tight seals. What survives the friction is the direction and the size of the trade, and that is what the drawings are testing.

Is mechanical reasoning the same thing as mechanical aptitude?

The two names get used for the same job family and describe two different tasks, which is why they are two pages here. A pictorial aptitude battery mostly asks you to compare — which arrangement needs less force, which way the second wheel turns — and puts a clock on it, because an employer wants to know how many you get through. Reasoning items ask you to produce the quantity: how fast, how much, how far. If a clock and a comparison format is what you are preparing for, the mechanical aptitude drill runs twenty of those at a published battery's tempo; this page is where the number underneath comes from.

Do I have to memorize formulas for this?

Five sentences cover all twenty items, and the page prints them before you start and again beside your score. Teeth per minute is shared between meshed gears; rope sections between blocks share the load and multiply the distance; load times its arm equals effort times its arm; pressure is equal on a level in a connected fluid, so force scales with area and area with the square of the diameter; and nothing turns when the turning effects cancel. Nobody needs a formula sheet for those, and an item that needed a sixth rule would be testing recall rather than reasoning.

Where is my percentile?

There is not one, and the reason is worth more than the number would be. A percentile requires a published mean and spread from a group that sat these items under these conditions, and twenty diagrams generated in your browser have no such group behind them; the commercial mechanical batteries that do have norm tables keep them for licensed customers. Since this page also keeps nothing you do, it could not build its own distribution even if that were sound. What you get instead is a count out of a stated number, the guessing floor beside it, and the working — which is the part that changes what you can do next time.

One idea behind all five topics, and why knowing it does not stop you getting the drawing wrong

Every item in this set is the same trade seen from a different angle: a simple machine cannot give you work you did not put in, so anything it makes easier it makes longer. Four rope sections under a load quarter the pull and quadruple the rope you haul. A lever arm three times the load’s needs a third of the force and moves three times as far to shift the load the same distance. A gear pair that turns four times slower turns four times harder, which is the whole content of a reduction gearbox. A piston three times the diameter gives nine times the force and moves a ninth as far, because its area is what the pressure acts on and area carries the square. Once you can see which quantity is being multiplied and which is paying for it, the arithmetic is division, and the five topics stop being five things to remember.

The reason a page like this is worth taking anyway is that knowing the rule and reading the drawing are separately learnable. That is one of the better-documented findings in physics education: Hestenes, Wells and Swackhamer’s Force Concept Inventory (1992, The Physics Teacher) showed that students who could pass a mechanics exam kept their pre-instruction intuitions about forces intact, and McCloskey’s work on intuitive physics (1983, Scientific American) catalogued how systematic those intuitions are — people predict a ball leaving a curved tube will keep curving. The oldest example is still the most startling: Piaget and Inhelder’s water-level task, in which adults are asked to draw the surface of the water in a tilted bottle and a substantial share draw it tilted with the bottle rather than horizontal. Vasta and Liben went back over three decades of that literature in 1996 (Current Directions in Psychological Science) and the puzzle they described is that the failure survives being told the answer. None of these figures is printed here as a score, because none of them was collected on these items — they are the reason the page shows working instead of just marking you.

Two things go wrong on practice material for this topic, and both are visible rather than arguable. The first is a diagram that could not physically work: gears drawn with mismatched tooth pitch, or a tackle whose rope path cannot be followed, teach a reading habit that fails the moment a real battery draws it properly — which is why a gear’s radius here is set by its tooth count and the rope is one continuous path. The second is a score with no derivation, which tells you that seven of twenty were wrong and leaves you no way to find out whether it was one missing rule or seven careless readings. The format itself has a caveat worth stating: the well-known pictorial batteries in this family, the Bennett Mechanical Comprehension Test among them, mostly ask for a comparison rather than a quantity, so asking for the number is a deliberate departure from that format in the direction of teaching. For the comparison format under a clock, take the mechanical aptitude drill instead. For the figural and rule-finding items that sit beside mechanics in most batteries, there is pattern recognition, the matrix format and deductive reasoning; for the numbers-in-a-table half of an industrial paper, numerical reasoning and the Watson-Glaser practice page cover what usually comes next.

A reaction time here is the interval between the frame that painted the stimulus and the timestamp the browser attached to your key, both read from the same monotonic clock. What neither can see is the display pipeline behind it, so on a 60 Hz screen roughly 16 ms of every figure below is the machine rather than you. That is the timing floor: two numbers closer together than that are the same number, and this page reports no precision it cannot support.

The only figures this page reports in time are per-topic medians of tens of seconds, and the floor above is four orders of magnitude below them. It is stated anyway, because the seconds recorded here come off the same clock the timed pages use, and because an item interrupted by the window losing focus is discarded for exactly that reason.

This is a measurement exercise, not a clinical assessment. It reports what you did on this page against a stated reference and nothing more — it cannot establish a mechanical or spatial difficulty. Only a qualified professional, working with more than a browser, can make that judgment.

Not affiliated with, endorsed by, or connected to Pearson Education, Inc.. Bennett Mechanical Comprehension Test is a trademark of its owner and is used here only to name the assessment this page prepares for. The questions on this page are our own: no part of the published test is reproduced, and a score here is not comparable to a score from the real instrument.

Where the derivation is computed

Every number on this page is worked out by JavaScript running in the tab you are reading it in. Your answers, your reaction times and your score are never uploaded, logged or kept — which is also why the test carries on working after you disconnect from the network, and why nothing here can be held back behind an email address.

The derivations are built by the same code that built the numbers, in this tab, at the moment the item was generated — there is no answer key on a server to fetch and no request made when you commit to an option. The set number in the footer rebuilds the identical twenty items; it holds no record of what you chose.