EN

Calculate Focal Length of a Thick Lens

The thin lens equation on the thin lens page assumes a negligible thickness. Once a lens has a real thickness e at its centre, you need the lensmaker's equation, which adds a correction term that depends on e, the refractive index n, and the two radii of curvature R₁ and R₂.

Standard convention (light travels left to right): R > 0 if the centre of curvature of the surface lies to the right of it, R < 0 if it lies to the left. As a data-entry shortcut, R = 0 stands for a flat surface (R → ∞).

eR₁R₂axis

Inputs

Shape and unit

Fills in n, R₁, R₂ and e (mm) — it does not affect the calculation itself.

Lens geometry

≈ 1.5 for ordinary crown glass.

0 = flat surface.

0 = flat surface.

Result

  • Focal length f 101.7 mm
Refractive index n 1.500
Radius R₁ 100.0 mm
Radius R₂ -100.0 mm
Centre thickness e 10.00 mm
Focal length f (thick lens) 101.7 mm
Focal length under the thin-lens approximation (e = 0) 100.0 mm (a 1.69% difference)
Optical power 1/f 9.833 δ
Lens type Converging

1/f = (n−1)·[1/R₁ − 1/R₂ + (n−1)·e / (n·R₁·R₂)]

From the thin lens to the thick lens: where the correction term in e comes from

The name gives it away. The lensmaker's equation is what a glass shop actually needs: two curves you ground, the glass you poured, and how thick the blank ended up at the centre. A textbook thin lens skips that last number. This page does not.

Each surface is a refracting power of its own. The front face contributes (n − 1)/R₁, the back face (1 − n)/R₂. Stick the two together with no glass between them and you just add. That is the thin-lens limit, and it is the first thing to recover as a sanity check:

e = 0 → 1/f = (n − 1)·(1/R₁ − 1/R₂)

Give the blank a real centre thickness e and a ray that has already bent at the first surface travels a short way through glass before it meets the second. It hits at a slightly different height. Gullstrand's extra product — written here as (n − 1)e / (n R₁ R₂) inside the brackets — is that height change, folded into one term. In air, the whole thing reads:

1/f = (n − 1)·[1/R₁ − 1/R₂ + (n − 1)·e / (n·R₁·R₂)]

A flat face kills the extra term as well (R → ∞ in the product). So a plano-convex lens has an equivalent power that does not depend on e at all. Thickness still slides the principal planes around; 1/f stays put. That is worth remembering when you are tempted to “correct” a condenser for thickness it does not need in this formula.

Reading the signs of R₁ and R₂

Light runs left to right. R is positive when that surface's centre of curvature sits to the right of the surface. Left of it: negative. That is the whole rule. The traps sit around it.

A “symmetric” biconvex magnifier is not +R and +R. Front surface bulges left, centre to the right, R₁ > 0. Back surface bulges right, centre to the left, R₂ < 0. Type +100 and +100 and you have drawn a meniscus, not a magnifier. The diagram above the form follows the signs you type — if it does not look like the piece on the bench, fix the signs before you trust f.

  • R = 0 in this form means a plane (R → ∞, so 1/R = 0). It is not an infinitely sharp curve.
  • This sign rule is not the “real-is-positive” convention used for object and image distances. Mixing the two is the classic lab-book accident.
  • Flipping the lens swaps which radius is first. For equal |R| biconvex glass the thin-lens power is unchanged; the thick-lens term is not, once the two radii differ.

When thickness actually matters

Ask what e is compared with the radii, not what e is in millimetres. A +2 D reading magnifier, 4 mm thick, radii around half a metre: the thick-lens shift is a few thousandths of a dioptre. School bench, ignore e.

Spectacles stop being that kind. A plus lens is fat in the middle; at +5 D and 6 mm centre thickness you already pick up 0.20 D — see the worked example. Camera glass is worse. A 50 mm focal-length element that is 8–12 mm of glass has e/f around 0.2. Nobody designs a photographic objective with the thin-lens formula; each surface is traced on its own.

Rule of thumb, then. Weak magnifier, teaching lens, mild +1.50 D: e = 0 is fine. Strong Rx, condenser, any camera group: use the lensmaker form with e in it.

The six common shapes

Pick a preset and the signs fill themselves. Power follows the thickness at the centre: fat in the middle converges, thin in the middle diverges.

  1. Biconvex — R₁ > 0, R₂ < 0. Plus power. Hand magnifier.
  2. Plano-convex — one radius entered as 0. Plus. And 1/f = (n − 1)/R, no e term.
  3. Plus meniscus — both radii positive, |R₁| < |R₂|. Plus-power spectacle lens.
  4. Biconcave — R₁ < 0, R₂ > 0. Minus. Eyepiece field lens, strong myopia.
  5. Plano-concave — one flat, one hollow. Minus.
  6. Minus meniscus — both radii the same sign, |R₁| > |R₂|. Minus-power spectacle lens.

The meniscus still looks wrong the first time you meet it: both faces curve the same way, yet one prescription is plus and the other minus. Whichever face is steeper wins. One more sign quirk: R₁R₂ is negative for a biconvex, so the thickness term cuts the power a little; it is positive for a plus meniscus, so thickness adds a bit of plus. Same millimetres of glass, opposite nudge.

Principal planes H and H'

A thin lens has one parking spot for every distance: “the lens”. Give it thickness and that spot splits. Incoming rays behave as if they were heading for the object principal plane H. Outgoing rays behave as if they had left from the image principal plane H'.

The focal length this calculator reports is HF on the way in and H'F' on the way out. It is not the gap from the last glass surface to the focal point. That gap is the back focal length, and it equals f only when H' sits on the exit vertex — almost never, once e is real.

On a wrapped spectacle meniscus, H can sit in the air in front of the first surface. Measure from the glass the way you would with a thin lens, and every conjugate is off. That is why an optician quotes vertex distance from the back surface, not from a fictional mid-plane.

Link to optical power and lens combinations

P = 1/f, in dioptres when f is in metres. Written with the two surface powers it is already a doublet: P = P₁ + P₂ − (e/n) P₁ P₂. Same skeleton as Gullstrand for two thin lenses in air, except the reduced gap is e/n rather than e.

Optical power covers lenses in contact (add the dioptres). Lens combination covers two thin lenses with air between them. A thick lens is the in-between case: one piece of glass, two surfaces that refuse to share a plane.

Worked example: an ordinary spectacle lens

A plus meniscus in everyday plastic, n = 1.50. Front R₁ = 50 mm, back R₂ = 100 mm, centre thickness e = 6 mm. Roughly a strong +5.00.

Ignore e first:

1/f₀ = (1.50 − 1)·(1/0.050 − 1/0.100) = 0.50 × 10 = 5.00 m⁻¹

So f₀ = 200 mm and P₀ = 5.00 D. Put the 6 mm back in:

(n − 1)·e / (n·R₁·R₂) = 0.50 × 0.006 / (1.50 × 0.050 × 0.100) = 0.40 m⁻¹

Then 1/f = 0.50 × (10 + 0.40) = 5.20 m⁻¹: focal length 192.3 mm, power 5.20 D. Thickness added 0.20 D — close to a quarter-dioptre workshop step. A mild +1.50 would barely move. An aphakic +8 would shift by more than half a dioptre. That is the point at which “treat it as thin” stops being honest.

Limits of the model

Air on both sides. noutside = 1 is baked into the (n − 1) factors. Drop the same blank in water (n ≈ 1.33) and both surfaces change: you replace (n − 1) with (nglass − nwater), and f stretches a long way. A contact lens under a tear film, a mask port, an eye in aqueous humour — different formula.

No aberrations. Spherical, chromatic, coma, oblique astigmatism: all of them show up as soon as the beam is more than a paraxial thread. The lensmaker equation returns the paraxial focal length, not whether the picture is sharp. A camera lens is judged on the rest.

Surfaces are spherical, the index is one number (no cemented achromat), and e is the centre thickness. A minus lens is thinner there than at the edge; you still type the centre value.

The meniscus spectacle lens

Walk into a high-street optician and you will not be sold a biconvex +4.00. That shape is a fat lentil. It will not sit in a frame, and the rear bulge fights the eyelashes. A meniscus wraps around the eye: convex front, concave back, the whole shell following the orbit.

Off-axis vision is the other reason. Look sideways through a flat plus lens and you pick up oblique astigmatism you can feel. Best-form curves (Tscherning's ellipse, the Ostwalt branch) are menisci. The dispenser picks a base curve on that branch, then compromises with whatever the chosen frame will take.

So a plus Rx (hyperopia, a near add) is a converging meniscus, thick in the middle. A minus Rx (myopia) is a diverging meniscus, thin in the middle, sometimes chunky at the rim. Both presets are in the drop-down.

1.5 vs 1.67 index plastic

n = 1.50 is everyday CR-39, or near enough crown. n = 1.67 is the high-index plastic offered once a prescription starts to make the lens look like a bottle-bottom. Same target power, higher n lets you grind shallower curves — longer radii — and the blank comes out thinner.

That +5.00 thin-lens target needed 1/R₁ − 1/R₂ = 10 m⁻¹ in 1.50 plastic. In 1.67, (n − 1) = 0.67, so you only need about 7.5 m⁻¹. Flatter faces, smaller e, a smaller thickness correction. That is the whole point of going up in index. The bill: a lower Abbe number (around 32 against ~58 for CR-39), so more chromatic fringing at the edge of a large frame, and more surface reflection, which is why an AR coat is more or less mandatory at 1.67.

Changing n in the calculator and leaving the radii alone does not tell that story. It just makes the lens stronger. To compare two plastics you have to re-cut R₁, R₂ and e, the way a surfacing lab would.

Related calculators