The dimensional analysis calculator above does the thing every physics and engineering student is told to do at the end of a derivation and almost nobody actually does by hand: it strips a unit expression down to the seven base dimensions and checks whether both sides of an equation carry the same ones. It parses products, quotients, brackets and powers, understands SI prefixes and named derived units, and tells you not just that two sides disagree but exactly which dimension is out and by how many powers.
Arb Digital publishes this alongside a large set of free physics and conversion tools because a dimensional check is the cheapest error trap in science. It costs seconds, it needs no data, and it catches the single most common class of algebra mistake — a term that has been divided when it should have been multiplied. Unlike the unit converter, which changes a quantity from one unit to another, this tool never converts anything. It asks a different question: what kind of quantity is this at all?
What This Dimensional Analysis Calculator Does
Every physical quantity in the SI can be written as a product of powers of seven base dimensions: mass (M), length (L), time (T), electric current (I), thermodynamic temperature (Θ), amount of substance (N) and luminous intensity (J). Speed is L T⁻¹. Force is M L T⁻². Electrical resistance is M L² T⁻³ I⁻². The NIST page on the SI units lists all seven base units and the quantities they measure.
The tool reduces whatever you type to that seven-number signature. It recognises the named derived units — newton, joule, watt, pascal, volt, ohm, farad, henry, tesla, weber, coulomb, siemens, hertz, lumen, lux, gray, sievert, becquerel and katal — and expands each into base units before comparing. It also accepts the litre, the minute, the hour, the day, the electronvolt, the bar and a handful of imperial units, since real problems arrive in mixed notation.
Prefixes are handled by trying an exact unit match first and only then peeling off a prefix. That ordering matters: m is the metre rather than milli-nothing, min is the minute rather than milli-inch, and T is the tesla while TW is the terawatt. Because a prefix is only a power of ten, it never changes the dimensional answer — a kilometre and a nanometre are both L.
The four supporting figures give the coherent SI unit for the left side, the reduction of the right side, a plain verdict on consistency, and the name of the physical quantity the signature corresponds to when it matches a standard one. That last box is the useful surprise: reduce an unfamiliar cluster of symbols and the tool often tells you it is a pressure, or an energy density, or a frequency.
How to Use It
- Type the left-hand side using unit symbols, not quantity symbols. Write kg*m/s^2, not m*a. The tool reads units; it has no way to know what your letters stand for.
- Group anything raised to a power with brackets. (m/s)^2 squares the whole speed; m/s^2 squares only the second. This is where most hand-checks go wrong, and the tool follows the same precedence rules algebra does.
- Put the other side of your equation in the second box. Leave it empty and the tool simply reduces the first expression, which is often all you want.
- Read the mismatch message, not just the verdict. When two sides disagree the tool names the base dimensions that differ and the power by which they differ, and that difference usually points straight at the offending term.
- Switch notation to suit the task. Dimensional notation is right for a consistency proof; coherent SI notation is right when you are about to label an axis or write down an answer.
How the Reduction Is Calculated
The parser walks the expression left to right, building a vector of seven exponents. Multiplication adds the two vectors, division subtracts them, and a power multiplies every exponent by the same number. Brackets recurse. Numbers contribute a vector of zeros, which is exactly right: 2, π and 1/2 are dimensionless, so they cannot rescue or ruin a consistency check.
Work the default through by hand. The kilogram contributes M¹. The metre contributes L¹. The s^2 in the denominator contributes T⁻². So kg*m/s^2 is M L T⁻². The newton is defined as the force that accelerates one kilogram at one metre per second squared, so it expands to exactly the same vector, and the verdict is consistent. Change the right side to J and the joule expands to M L² T⁻² — the mismatch is one extra power of length, which is precisely the metre of distance that turns a force into work.
Two sides are consistent when all seven exponents match. The comparison uses a small tolerance rather than exact equality so that fractional powers behave sensibly: the period of a pendulum goes as the square root of length over gravity, which involves exponents of one half, and floating-point arithmetic on halves does not always land on the same bits. The BIPM SI Brochure is the authority on how the base units and the derived units relate, and it is where every expansion used here comes from.
What a Dimensional Check Can and Cannot Catch
This is the section that matters most, and the one most quick explanations skip. A dimensional check is a filter, not a proof. It catches a large and important class of error and is completely blind to another.
It catches structural mistakes: a term multiplied instead of divided, a squared quantity left unsquared, a radius used where an area was meant, a density mistaken for a mass. All of these change the exponent vector, and the check fires immediately. It also catches unit mixing at the point where it bites — an expression that is fine in SI but has a stray hour or foot in it will still reduce correctly dimensionally, which is a reminder that dimensional consistency and numerical correctness are separate tests.
It cannot catch anything dimensionless. The one half in the kinetic energy formula, the 2π in a circumference, the factor of four in an inverse-square law, a sign error, a swapped pair of variables that happen to share dimensions — every one of these survives. Two lengths are dimensionally identical whether one is a radius and the other a wavelength, so substituting one for the other passes the check and gives a wrong number.
The honest way to use the tool, then, is as a first gate. If an expression fails, it is definitely wrong and you have been handed a specific clue. If it passes, you have learned only that it is not wrong in this particular way, and the numerical check still has to happen.
Arguments of Functions Must Be Dimensionless
A rule that catches a great many real errors: anything inside a sine, cosine, exponential, logarithm or power series must be dimensionless. There is no such thing as the sine of three metres. The series expansion makes the reason obvious — it would add a length to a cubed length to a fifth-power length, and those cannot be added.
Every exponential decay in physics respects this quietly. In radioactive decay the exponent is time divided by a lifetime, and the two times cancel. In an RC circuit the exponent is time divided by the product of resistance and capacitance, which is why that product is called the time constant and why it must have the dimension T. In the Boltzmann factor the exponent is an energy divided by another energy, and the constant that makes temperature into an energy is what Boltzmann's constant is for.
Use this as a design check. If you have written an exponent and it does not reduce to a plain number, you have missed a constant, and the constant you have missed usually announces itself by the dimensions it needs to supply. That trick — reading the dimensions of a missing constant off the equation — is one of the few genuinely creative uses of dimensional analysis, and it is how several famous constants were first characterised before they were measured.
Angles, Radians and the Quantities That Look Dimensionless
The radian and the steradian are dimensionless by construction: an angle is a length divided by a length. That makes them awkward, because they vanish from a dimensional check even when the physics badly wants them kept. Angular velocity in radians per second reduces to T⁻¹, which is the same signature as frequency in hertz, yet the two differ by a factor of 2π and confusing them is a classic error that no dimensional tool can flag.
Several other quantities collapse to nothing for legitimate reasons. Strain is a length over a length. Refractive index is a speed over a speed. Efficiency, the coefficient of friction, Mach number, Reynolds number and the fine-structure constant all reduce to a bare number. Seeing a blank dimensional formula is therefore not an error message — it is the correct answer for a whole family of quantities, and dimensionless groups are among the most useful numbers in fluid mechanics and heat transfer precisely because they carry across scales.
Temperature deserves its own warning. Kelvin differences and kelvin absolutes share the dimension Θ, but degrees Celsius and Fahrenheit involve an offset, and an offset scale cannot be handled by exponent arithmetic at all. This tool treats temperature dimensionally and never converts it; use the dedicated converter for that, and never put a Celsius reading into a formula that expects an absolute temperature.
Where Dimensional Analysis Earns Its Keep
Beyond checking algebra, dimensional reasoning can construct a formula you do not know. If you believe the period of a pendulum depends only on its length, its mass and the acceleration of gravity, there is exactly one combination of those three that has the dimension of time: the square root of length divided by gravity. Mass cannot appear, because there is nothing to cancel its M. That result — that a pendulum's period does not depend on the bob's mass — falls out of dimensions alone, with no differential equation and no experiment.
The same logic underlies scale modelling. A wind-tunnel test of a small aircraft predicts the behaviour of a full-size one when the dimensionless groups match, because the dimensionless groups are the only things the physics can actually depend on. This is the Buckingham pi theorem in practice: a relationship among n quantities involving k base dimensions can always be rewritten as a relationship among n minus k dimensionless groups, which is a large reduction in the number of experiments needed.
What dimensional analysis cannot supply is the dimensionless constant out front. It tells you the pendulum period is proportional to the square root of length over gravity; it cannot tell you the factor is 2π. That has to come from the actual derivation or from one measurement. Knowing the shape of an answer without its coefficient is still an enormous head start, and it is why the technique survives in fields where full derivations are impossible.
Arb Digital builds free tools like this one because useful pages earn attention. If you want tools, calculators or content built for your own audience, we can help.
Browse All Free Tools Talk to Arb DigitalCommon Mistakes to Avoid
- Writing quantity symbols instead of unit symbols — the tool reads units. Enter kg*m/s^2, not m*a, or it will read your mass as a metre.
- Forgetting brackets around a squared group — (m/s)^2 and m/s^2 are different quantities, and dropping the brackets is the single most common input error here.
- Treating a passed check as a correct formula — dimensionless factors, signs and swapped same-dimension variables all survive a dimensional test untouched.
- Assuming radians behave like a unit — angular velocity and frequency share a dimensional signature but differ by 2π, and no dimensional check will ever tell you which one you meant.
- Putting Celsius or Fahrenheit into a formula — those are offset scales. Dimensional consistency says nothing about the offset, and the arithmetic will be wrong even when the check passes.
Related Free Tools From Arb Digital
Once the dimensions check out, convert the actual numbers with the unit converter, or use the specialised energy converter, pressure converter and speed converter. For presenting a result properly the significant figures calculator and the scientific notation converter handle precision and magnitude. Applied checks are easier with the force calculator, the kinetic energy calculator and the density calculator, and every tool Arb Digital publishes is listed on the free online tools hub.
Frequently Asked Questions
Mass, length, time, electric current, thermodynamic temperature, amount of substance and luminous intensity, written M, L, T, I, theta, N and J. Every SI quantity is a product of powers of these seven, and the tool reports that set of powers.
No. It can only prove one wrong. Dimensionless factors such as one half or two pi, sign errors, and swapped variables that share dimensions all pass the check, so a passing result means the equation is not wrong in this particular way and nothing more.
Because it is dimensionless, which is a legitimate answer. Strain, refractive index, efficiency, friction coefficients, Mach number and Reynolds number are all ratios of like quantities, so every exponent cancels to zero.
Use a caret and bracket anything that should be squared as a group. Writing m/s^2 squares only the second, giving acceleration, while (m/s)^2 squares the whole speed and gives the dimensions of energy per unit mass.
No. A prefix is only a power of ten, so a kilometre, a millimetre and a metre all reduce to length. The tool strips prefixes for the dimensional answer, but the numerical factor they represent still matters when you actually compute a value.
Because the series expansion adds the argument to its own square and cube, and quantities with different dimensions cannot be added. If an exponent in your working has dimensions, a constant is missing from it.
Not dimensionally. Radians are a length over a length and so vanish, leaving angular velocity and frequency with the identical signature of one over time even though they differ by a factor of two pi. That distinction has to be tracked by hand.
No, and deliberately so. It reports what kind of quantity an expression is, not what a value becomes in another unit. Use the unit converter for numerical conversion once the dimensions are confirmed.
This tool is provided for educational and study use. It performs an ideal dimensional reduction and does not convert values, handle offset temperature scales, or validate the numerical constants in a formula, so a passing check is a filter rather than a guarantee of correctness.
