The mean free path is the average distance a gas molecule covers between one collision and the next. In room air it is about 65 nanometres — roughly two hundred times the diameter of the molecule itself, and small enough that a molecule crossing a one-millimetre gap collides some fifteen thousand times on the way. Pump most of the air out and that same molecule can cross a whole vacuum chamber without meeting anything. That transition is the single most useful thing this number tells you.
This calculator returns the mean free path from pressure, temperature and molecular diameter, and then reports the quantities that follow from it: how often a molecule collides, how long it survives between collisions, how many molecules occupy a cubic metre, and the Knudsen number for a length scale you choose. Arb Digital built it because the mean free path is usually quoted as a single memorised figure for air, when it is in fact a strong function of pressure and a weak one of temperature.
What This Mean Free Path Calculator Does
Pick a gas or type in your own molecular diameter, set the pressure and temperature in whatever units your gauge reads, and the headline result is the mean free path, displayed with an automatically chosen unit so that a nanometre-scale answer and a kilometre-scale answer are both readable.
The grid gives the collision frequency, which is the average speed divided by the mean free path; the mean free time, which is its reciprocal; the number density from the ideal gas law; and the Knudsen number, the ratio of mean free path to the characteristic length you entered. The subtitle names the flow regime that Knudsen number falls into, using the conventional boundaries.
Two adjacent tools cover neighbouring ground. The Knudsen number calculator is the dedicated page for the regime ratio and its consequences for slip and rarefaction; this page derives the mean free path that feeds it. The RMS velocity calculator handles molecular speed distributions in detail — root-mean-square, mean and most-probable speeds — where this page uses only the mean speed, and only as an intermediate step.
How to Use It
- Choose the gas. Seven common gases are pre-loaded with an approximate kinetic diameter and molar mass. Select the custom option to enter values from your own source.
- Enter the pressure and its unit. Pascals, kPa, millibar, bar, atmospheres or torr — use whatever your gauge is calibrated in rather than converting first.
- Enter the temperature. Celsius, kelvin or Fahrenheit; the calculator converts to absolute temperature internally, and refuses to run below absolute zero.
- Adjust the molecular diameter if you have a better figure. Mean free path depends on the square of this number, so it is by far the largest source of uncertainty in the answer.
- Set the characteristic length of your pipe, gap or chamber to get a Knudsen number and a flow regime alongside.
The Formula: How It's Calculated
Treat molecules as hard spheres of diameter d. A molecule sweeps out a collision cylinder of cross-section πd² as it moves, and it collides when another molecular centre falls inside that cylinder. Setting the expected number of centres in a cylinder of length λ equal to one, and correcting by √2 because the targets are moving too, gives:
λ = kBT ÷ (√2 × π × d² × P)
where kB is the Boltzmann constant, 1.380649 × 10−23 J/K by definition since the 2019 SI redefinition, T is absolute temperature and P is absolute pressure. The number density comes from the same ideal-gas substitution, n = P ÷ kBT, so the relation can equally be written λ = 1 ÷ (√2 π d² n). The derivation, including the reason for the √2, is set out in OpenStax University Physics Volume 2, section 2.2, and the collision-rate treatment is developed further in the HyperPhysics page on mean free path and molecular collisions.
To turn a distance into a rate you need a speed. The calculator uses the mean speed of the Maxwell–Boltzmann distribution, v̄ = √(8kBT ÷ πm), where m is the mass of one molecule. The collision frequency is then z = v̄ ÷ λ and the mean free time is its reciprocal.
A worked example matching the defaults on this page: air at 101,325 Pa and 20 °C (293.15 K) with d = 370 pm. The numerator is 1.380649 × 10−23 × 293.15 = 4.047 × 10−21. The denominator is √2 × π × (3.70 × 10−10)² × 101,325 = 6.163 × 10−14. Dividing gives λ = 6.57 × 10−8 m, or 65.7 nm. The mean speed at that temperature for M = 28.96 g/mol is 462.9 m/s, so the collision frequency is 7.05 × 109 per second — a molecule collides about seven billion times a second, and survives roughly 140 picoseconds between hits.
Why the Molecular Diameter Dominates the Uncertainty
Pressure and temperature you can measure. Molecular diameter you cannot, at least not directly — molecules do not have edges. What is quoted as a "diameter" is an effective collision diameter inferred from a measured transport property, and different properties give different answers for the same molecule. A diameter derived from viscosity, one derived from thermal conductivity and one derived from molecular-sieve exclusion can differ by ten per cent or more.
Because λ goes as 1 ÷ d², a ten per cent disagreement in diameter is a twenty per cent disagreement in mean free path. That is worth remembering before quoting the result to three significant figures. Values published for nitrogen sit between about 364 and 380 pm depending on the source and method, which alone moves the room-air answer between roughly 62 and 68 nm. The hard-sphere model is also an idealisation: real molecules attract at long range and repel softly at short range, so the effective diameter itself drifts with temperature.
Reading the Knudsen Number and the Flow Regimes
The mean free path only becomes actionable when compared with something. The Knudsen number, Kn = λ ÷ L, does that comparison, and the conventional bands are: below 0.01 the gas behaves as a continuum and ordinary fluid dynamics applies; from 0.01 to 0.1 is slip flow, where the no-slip wall condition starts to fail; from 0.1 to 10 is the transition regime, where neither continuum equations nor free-molecular ones work well; and above 10 is free-molecular flow, where molecules travel wall to wall without meeting each other.
Room air in a one-millimetre channel gives Kn ≈ 6.6 × 10−5, firmly continuum, which is why ordinary pipe flow calculations and the Reynolds number work. Take the same channel down to 1 Pa and Kn rises to about 6.7, deep in the transition regime, where pumping speed no longer scales the way a continuum conductance formula predicts. The same shift explains why microfluidic gas channels can show measurable slip at atmospheric pressure: shrink L instead of dropping P and you cross the same boundary.
Where the Mean Free Path Actually Gets Used
Vacuum engineering is the obvious case. The distinction between viscous and molecular pumping regimes is a mean-free-path distinction, and it determines which pump technology works, how conductance is calculated, and how long a chamber takes to reach base pressure. Thin-film deposition depends on it directly: in sputtering and evaporation the useful question is whether an atom leaving the source reaches the substrate without scattering, which is a comparison of source-to-substrate distance against λ.
It matters in atmospheric science too. Aerosol particles smaller than the mean free path of air are not dragged by a continuum fluid; they experience individual molecular impacts, which is why the Stokes drag law needs a slip correction below roughly a micrometre. It sets the scale for heat conduction in gases — the reason a vacuum flask works is that below a certain pressure the gas can no longer relay heat across the gap by successive collisions. And it explains why gas thermal conductivity is almost independent of pressure over a wide range: fewer carriers, but each travelling proportionally further, and the two effects cancel until the mean free path reaches the size of the container.
Temperature: A Much Weaker Lever Than Pressure
At fixed pressure, λ is proportional to absolute temperature, because heating a gas at constant pressure thins it out. Going from 20 °C to 200 °C, a rise most people would call large, takes the absolute temperature from 293 K to 473 K — a factor of 1.61, so the mean free path grows by 61%. Compare that with pressure, where a single decade of pumping changes λ by a factor of ten. If you are trying to move a system between flow regimes, pressure is the lever; temperature is a correction.
At fixed density rather than fixed pressure, the mean free path does not change with temperature at all, since λ = 1 ÷ (√2πd²n) contains no T. Only the speed changes, so the molecules cover the same distance faster and the collision rate rises as √T. Being clear about which quantity is held constant is the difference between the two statements, and it is the source of a great deal of confusion. The underlying P–V–T relation is handled by the ideal gas law calculator, and the density side by the air density calculator.
Arb Digital builds accurate, fast technical calculators that rank for the terms your customers search and earn links from people who use them. See what we have already published, or tell us what your readers need.
Browse All Free Tools Talk To Arb DigitalCommon Mistakes to Avoid
- Using gauge pressure instead of absolute pressure — the relation needs absolute pressure, and near vacuum the difference between the two is the entire answer.
- Forgetting the √2 — the simplified derivation that ignores target motion overstates the mean free path by about 41%.
- Quoting a diameter-derived result to three significant figures — the effective collision diameter is itself uncertain, and λ depends on its square.
- Applying continuum fluid formulas at high Knudsen number — above about 0.01 the no-slip wall condition begins to fail, and above 0.1 continuum results stop being trustworthy.
- Mixing up constant-pressure and constant-density statements — heating at constant pressure lengthens the mean free path, but heating at constant density does not change it at all.
Related Free Tools From Arb Digital
Feed this result into the Knudsen number calculator for the rarefaction regime, or the RMS velocity calculator for the full molecular speed distribution. The number density calculator covers molecules per unit volume in its own right, the ideal gas law calculator handles the P–V–T relation, and the air density calculator gives mass density for humid air. For continuum flow, use the Reynolds number calculator and the pipe flow calculator, and convert units with the pressure converter. The rest are in the free online tools hub.
Frequently Asked Questions
About 65 to 68 nanometres at one atmosphere and 20 degrees Celsius, depending on which effective molecular diameter you use. With a diameter of 370 picometres the calculation gives 65.7 nm, which is roughly 180 times the size of the molecule itself.
No. The mean free path depends only on temperature, pressure and molecular diameter. Molar mass enters the collision frequency and the mean free time, because heavier molecules move more slowly and therefore take longer to cover the same distance between collisions.
Because the other molecules are moving as well. A naive derivation treats the targets as stationary, which understates the rate at which encounters happen. Averaging over the relative velocities of two molecules drawn from the same Maxwell-Boltzmann distribution introduces a factor of the square root of two.
It is inversely proportional to absolute pressure at constant temperature. Halving the pressure doubles the distance between collisions, and each decade of pumping multiplies it by ten, which is why the number spans nanometres at atmospheric pressure and metres in high vacuum.
By the conventional bands, above about 10. Below 0.01 the gas is a continuum, 0.01 to 0.1 is slip flow, and 0.1 to 10 is the transition regime where neither continuum nor free-molecular treatments are reliable. The boundaries are conventions rather than sharp physical thresholds.
Because a molecule has no hard edge. The quoted diameter is an effective collision diameter inferred from a measured property such as viscosity, thermal conductivity or sieve exclusion, and different properties give different values. Since the mean free path depends on the square of the diameter, a ten per cent disagreement becomes twenty per cent in the result.
At constant pressure, yes, in direct proportion to absolute temperature, because a hotter gas at the same pressure is less dense. At constant density it does not change at all; only the molecular speed rises, so collisions happen more often over the same distance.
This tool is provided for educational and engineering-estimate use. It applies the hard-sphere kinetic model with an ideal-gas number density and does not account for intermolecular attraction, real-gas behaviour near condensation, gas mixtures with dissimilar species or surface effects. Vacuum system design and any process depending on rarefied gas behaviour should be verified against measured data.
