Talk:Fresnel number

Fresnel diffraction at F>1??
On the page on diffraction, it says that fresnel diffraction is the more general case, which leads me to think that there is no F at which fresnel diffraction is more or less accurate. In other words, I'm saying fresnel diffraction can be used when F is anything, not just greater than one. Is this right? Fresheneesz 05:44, 20 March 2006 (UTC)


 * Nevermind.. Fresnel diffraction should be used when F is largeish. But this leaves out a bit of fresnel numbers. What if your F is .98 ? what then? Is there more accurate and general ways of determining diffraction? Fresheneesz 05:50, 20 March 2006 (UTC)

reply: For large L then the case will be Fraunhofer. Therefore Fraunhofer is valid for values of F less than 1.

Fresnel number definition
As far as I know, the given definition of Fresnel number is valid for the radius and not the diameter of the diffracting aperture. For the diameter you will have : F=(D/2)^2/(lambda*R)

reply: the defintion in the page is identical to the above one where D/2 := a and R := L

Fresnel number derivation
Does anyone have a source on how this equation was derived? My Fourier Optics (Goodman) text pg 85 simply defines Fresnel number as it is defined here without explanation.Mollynet (talk) 01:43, 4 December 2009 (UTC)


 * Many (most?) dimensionless numbers are not *derived*, they are *defined*. They are simply a collection of relevant parameters that are arranged so the units cancel out.  For them to be useful, the range of values should differentiate among some phenomena of interest.  Often, a physical interpretation can be attached to these useful dimensionless numbers.  But in some sense they are quite arbitrary.  Convention is often required in their definition.  For instance, one could have defined Fresnel number using diameter instead of radius.  One could have put in a factor of 2 to make the number of cycles be based on the full wavelength instead of the half-wavelength.  I realize that this reply comes almost five years after the question, but the discussion may be helpful to other readers.  Kimaaron (talk) 20:01, 20 April 2014 (UTC)

Fresnel number convention
The transtion from Fresnel to Fraunhofer regime is gradual and it is part a matter of convention at what F-value it should be drawn. Such conventions may be different in different subdisciplines e.g. optics and microwaves. In Antenna Theory - Analysis and Design by C.A. Balanis (2:nd ed, John Wiley & Sons 1997, ISBN 0-471-59268-4) which primarily deals with microwaves the limit is drawn at L=2D^2/\lambda, i.e. F=1/8. This means that the curvature of the wavefront is less than \lambda/16 over an area with diameter D at distance L. —Preceding unsigned comment added by 150.227.15.253 (talk) 11:09, 24 February 2010 (UTC)

"estimed"?
Is this some kind of optics jargon? I've never heard this word before except as a typo of "estimated". — Preceding unsigned comment added by 68.198.133.72 (talk) 15:54, 2 June 2014 (UTC)

reply: thanks for the English correction!

Definitions incoherent
''According to the user manual for the Zemax optical design software, the correct approximation for propagation in the near field follows the angular spectrum method. This approximation works well when at the propagation position the distance to the aperture is of the same order as the aperture size. This propagation regime satisfies $$\ F \gg 1$$.

''The correct approximation for the propagation in the far field is Fresnel diffraction. This approximation works well when at the observation point the distance to the aperture is bigger than the aperture size. This propagation regime verifies $$\ F \sim 1$$.

''Finally, once at the observation point the distance to the aperture is much bigger than the aperture size, propagation becomes well described by Fraunhofer diffraction. This propagation regime verifies $$\ F \ll 1$$.

These characterizations are unclear and confused.

(1) The Fresnel number F depends on all three of a, λ, and L, yet in the verbal geometric characterizations, only the ration of a to L is mentioned. This makes no sense. Most likely there are some unstated assumptions about the size of the optical system and the wavelength of the light.

(2) In the first paragraph, the verbal geometric criterion is a rough equality, yet the F criterion is a strong inequality. This makes no sense.

(3) In the second paragraph, the regimes is characterized in turn by "far field" (inequality), "bigger" (inequality), and F ≈ 1 (equality).

(4) The second and third paragraph, taken together, tend to suggest to the uninitiated that the region where Fraunhofer diffraction becomes effective is distinct from and beyond the far field. This really should be phrased differently.

178.38.43.206 (talk) 10:09, 8 June 2015 (UTC)

reply: the definition of Fresnel number can be rewritten as

FN = a^2/(lambda*L) = a^2/(lambda*L) * lambda/lambda = (a/lambda)^2*lambda/L.

Hence:

1) The sense is given considering a and L such as normalized to lambda.

2) & 3) FN >> 1 and FN << 1 are clear optical regimes, while FN ∈ O(1) is unclear, this is why the Gaussian Beam is definitely better (cfr. text).

4) The text declares only that Fraunhofer propagation applies just for FN << 1: the far field. Where to fix this threshold is not unique.