Why an Air Fryer Browns the Crust Before the Middle Cooks

This page contains affiliate links. If you buy through one we may earn a commission, at no cost to you.

Because the machine is better at the outside of the bread than at the inside, and the gap is measurable. An air fryer is a small convection oven with an aggressive fan, and what the fan raises is the rate at which heat crosses from the air into the surface. That shortens the time before the crust browns almost exactly in proportion. It shortens the time for the middle to reach temperature by much less, because once the heat is inside the dough the fan has no further say.

For a 30 mm focaccia the numbers come out at 4.66 times faster at the surface and 2.55 times faster at the centre. The ratio between those, 1.83, is the whole complaint. And there is a floor underneath all of it that no machine can pass: 9.2 minutes for that thickness, set by conduction through dough alone.

Two of our own pages already say this in words, correctly, and this page is the arithmetic they were owed.

Clock one: the crust cannot brown until it is dry

Start with the surface, because it fails first and it is the reason food comes out of the basket. A wet dough surface cannot get hot. While liquid water is evaporating from it, every joule arriving goes into latent heat at 2,257 kilojoules per kilogram, and the surface sits pinned near 100 degrees Celsius. Browning needs 140 to 170. So the crust cannot begin to colour until the outer layer has dried out, and the time that takes is simply the water in that layer divided by the flux arriving.

Take the outer millimetre of a dough at roughly 1,100 kilograms per cubic metre and 40 percent water. That is 0.44 kilograms of water for every square metre of surface, and 993 kilojoules to evaporate it.

Now the flux. With air at 200 degrees and the surface pinned at 100, the driving difference is 100 kelvin, and the flux is the heat transfer coefficient times that difference. A conventional oven, where most of the heat arrives by radiation and gentle natural convection, runs somewhere around 15 watts per square metre kelvin, so 1,500 watts per square metre, and the millimetre takes 662 seconds, or 11 minutes. An air fryer runs around 70, so 7,000 watts per square metre, and it takes 142 seconds. Two minutes and twenty four seconds.

Our page on why air fryer food is not crispy is looking at the same physics from the opposite side and gets it right: "A dry surface browns; a wet one just sizzles off steam until the moisture is gone." That is this clock, exactly, and the figure attached to it is 142 seconds on a dough surface.

Clock two: the centre does not care about the fan

The middle is a conduction problem and it runs on different machinery. What governs it is the thermal diffusivity of the dough, about 1.3 times ten to the minus seven square metres per second, and the square of the distance the heat has to travel. For a 30 mm focaccia with both faces exposed, that distance is 15 mm, and the natural time scale is the distance squared over the diffusivity: 1,731 seconds.

How much of that you actually need depends on the Biot number, which is the coefficient times the half thickness over the dough's conductivity. It is the ratio of the resistance inside the dough to the resistance in the air film. In a conventional oven it is 0.45, so most of the resistance is still in the air and improving the air helps. In an air fryer it is 2.1, so most of the resistance has moved inside the dough, where the fan cannot reach.

Run the slab solution for the centre reaching 96 degrees and the answers are 45.3 minutes in the oven, 17.7 minutes in the air fryer, and 9.2 minutes in a machine with unlimited airflow. That last figure is the one worth memorising. It is the conduction floor. No wattage, no temperature and no fan gets a 30 mm focaccia's middle to temperature faster than 9.2 minutes, and the air fryer has already captured 76 percent of the distance from the oven to that floor.

So the air fryer is faster at both jobs. It is just disproportionately faster at the one you can see. In an oven the crust sits dry and browning for 75.6 percent of the bake. In an air fryer that share is 86.7 percent. The crust is committed earlier, it looks finished earlier, and the thing people do when food looks finished is take it out.

What the wattage on the box is actually for

A Ninja Air Fryer Pro AF141 publishes a 5 quart basket, 1750 watts and a 400 degree Fahrenheit ceiling, and 1750 watts sounds like it ought to dominate everything above. Check it against the transfer path.

A 180 mm round focaccia presents about 0.0254 square metres per face. At the 7,000 watts per square metre the convective coefficient delivers, that is 178 watts into the top and the same into the bottom, so 356 watts total is crossing into the food. The other 1,400 watts is heating the chamber, replacing what leaves through the vent, and recovering after you open the drawer.

Which tells you what wattage buys. It is a recovery figure and a ceiling figure. It decides how fast the chamber comes back after you pull the basket out to look, and whether the machine can hold 450 degrees against its own losses. It does not decide the rate at the food, because the rate at the food is set by the coefficient, and the coefficient is the one number no air fryer listing we carry publishes. Not an airflow in cubic feet per minute, not an air velocity, nothing. Across the whole category we publish watts and degrees, and neither of them is the variable in the equation.

Two claims in our catalogue that survive the arithmetic

Worth doing, because not every published figure is wrong and these two are defensible. A CHEFMAN 2 quart mini air fryer states that it "cooks food up to three times faster than a conventional oven". The arithmetic above gives 4.66 times at a surface and 2.55 times at a centre, so three sits inside the range. For the kind of thin, high surface area item a 2 quart basket actually holds, three is conservative. For a thick one it is optimistic. "Up to" is doing honest work.

A Breville Smart Oven Air Fryer states that its "super convection can cut cooking time by up to 30%". The physical ceiling available to convection on a 30 mm slab is the whole distance from 45.3 minutes to 9.2, which is a 79.7 percent cut, so a claimed 30 is well inside what the mechanism can deliver. Its five quartz elements and the stated aim of reducing cold spots are addressing a different problem again, which is uniformity across the cavity rather than depth into the food.

The sheet of paper that triples the problem

Here is the practical finding, and it is larger than anything else on this page. A pack of 500 disposable air fryer paper liners is a reasonable thing to own, rated to 450 degrees Fahrenheit, silicone coated, sized for 5 to 8 quart baskets. Put one flat under a focaccia and you have removed the underside of the bread as a heat path.

That is not a modest change. A slab heated on one face with the other insulated behaves exactly like a slab of twice the thickness heated on both faces. So a 30 mm focaccia on a solid liner is a 60 mm problem: the distance doubles, the natural time scale goes up fourfold to 6,923 seconds, and the centre reaches 96 degrees at 53.7 minutes instead of 17.7.

Three times the centre clock. Meanwhile the top surface is unaffected, so its clock is still 142 seconds and the crust is still committed in under two and a half minutes. The liner has taken a bake that was 86.7 percent crust time and made it 95.6 percent.

Our guide on using air fryer liners without ruining airflow already gives the right instruction, saying to "avoid liners that seal the whole basket floor" and to "leave gaps around the edges so air moves under and around the food". It is right, and the number attached to it is three times. A perforated liner, or one cut smaller than the food so the basket holes stay open under it, keeps the second face working. A flat sheet across the whole floor is the single most effective way to undercook the middle of anything thick.

Why thickness is the only lever that matters

Everything in the first panel is a factor of two or three. Thickness is a square law, so it is the only variable with real authority.

At 20 mm the centre reaches temperature in 9.4 minutes. At 30 mm, 17.7. At 40 mm, 27.1. The pure square law would predict a factor of four between the thinnest and the thickest, and the real factor is 2.9, because the Biot number rises with thickness too and slightly offsets it. Either way, rolling a focaccia to two thirds of its thickness takes 8 minutes off the centre and nothing off the crust, which is the only move on this page that improves one clock without hurting the other.

And the second lever, which our own page already recommends. Our air fryer cooking times guide says: "Baked goods such as cakes and muffins can brown on top before the center sets, so lower the temperature and cover loosely with foil if needed." Both halves of that are right and the arithmetic says why. Lowering the air temperature reduces the surface flux proportionally, so the surface clock stretches directly, while the centre clock stretches much less because it is governed by diffusion rather than by the driving difference. Covering loosely with foil does the same thing more sharply: it blocks the jet, dropping the effective coefficient at the top surface to something like 25, which pushes the surface clock from 142 seconds to 397 and the centre clock from 17.7 minutes to 30.8. You give up 13 minutes and buy back a 2.8 times lower surface flux, which is the right trade whenever the crust is the thing failing.

The same page is also right that "Dense or thick foods gain less. A whole chicken or a thick roast still needs time for heat to travel to the center, and rushing the temperature only burns the outside." That sentence is the conduction floor, stated in English, three years before anyone put 9.2 minutes next to it.

The honest limit

Every figure here rests on two heat transfer coefficients, 15 and 70, which are representative values for the two machine types rather than numbers read off a listing. If an air fryer is really at 50, its centre clock becomes about 20 minutes rather than 17.7. If it reaches 100, about 15.5. The individual times move.

What does not move is the shape, and the shape is the answer. Whatever the two coefficients are, the surface clock is inversely proportional to the coefficient and the centre clock is not, because the centre clock is a diffusion time multiplied by a weak function of the Biot number. So any machine that improves the air will always improve the crust more than the crumb, and the better the airflow the wider that gap gets. There is no air fryer setting, and no air fryer, for which this is not true.

Which means the fix was never going to be on the machine. It is a thinner dough, a lower dial, a loose cover, a liner with the holes left open, and a thermometer in the middle instead of a judgement from the top. None of those is a purchase, and together they are worth more than the difference between any two air fryers in this catalogue.