Small-angle approximation: Difference between revisions

Revision as of 09:13, 30 November 2024

Latest revision as of 03:15, 18 August 2025

(23 intermediate revisions by 8 users not shown)

Line 18:

== Justifications ==

=== Graphic ===

The accuracy of the approximations can be seen below in Figure 1 and Figure 2. As the measure of the angle approaches zero, the difference between the approximation and the original function also approaches 0.

<gallery widths="500px" heights="400px">

File:Small_angle_compair_odd.svg|'''Figure 1.''' A comparison of the basic [[odd function|odd]] trigonometric functions to {{mvar|θ}}. It is seen that as the angle approaches 0 the approximations become better.

File:Small_angle_compare_even.svg|'''Figure 2.''' A comparison of {{math|cos ''θ''}} to {{math|1 − {{sfrac|''θ''²|2}}}}. It is seen that as the angle approaches 0 the approximation becomes better.

</gallery>

=== Geometric ===

[[File:Small angle triangle.svg|600pxframeless|upright=2]]

The red section on the right, {{math|d}}, is the difference between the lengths of the hypotenuse, {{mvar|H}}, and the adjacent side, {{mvar|A}}. As is shown, {{mvar|H}} and {{mvar|A}} are almost the same length, meaning {{math|cos ''θ''}} is close to 1 and {{math|{{sfrac|''θ''²|2}}}} helps trim the red away.

For a small angle, {{mvar|H}} and {{mvar|A}} are almost the same length, and therefore {{math|cos ''θ''}} is nearly 1. The segment {{mvar|d}} (in red to the right) is the difference between the lengths of the hypotenuse, {{mvar|H}}, and the adjacent side, {{mvar|A}}, and has length <math>\textstyle H - \sqrt{H^2 - O^2}</math>, which for small angles is approximately equal to <math>\textstyle O^2\!/2H \approx \tfrac12 \theta^2H</math>. As a second-order approximation,

<math display="block"> \cos{\theta} \approx 1 - \frac{\theta^2}{2}.</math>

The opposite leg, {{mvar|O}}, is approximately equal to the length of the blue arc, {{mvar|s}}. GatheringThe factsarc from{{mvar|s}} geometry,has length {{math|1=''s'' = ''AθθA''}}, fromand trigonometry,by definition {{math|1=sin ''θ'' = {{sfrac|''O''|''H''}}}} and {{math|1=tan ''θ'' = {{sfrac|''O''|''A''}}}}, and fromfor thea picturesmall angle, {{math|''O'' ≈ ''s''}} and {{math|''H'' ≈ ''A''}}, which leads to:

<math display="block">\sin \theta = \frac{O}{H}\approx\frac{O}{A} = \tan \theta = \frac{O}{A} \approx \frac{s}{A} = \frac{A\theta}{A} = \theta.</math>

Or, more concisely,

Simplifying leaves,

<math display="block">\sin \theta \approx \tan \theta \approx \theta.</math>

Line 61 ⟶ 55:

\end{align}</math>

where {{tmath|\theta}} is the angle in radians. For very small angles, higher powers of {{tmath|\theta}} become extremely small, for instance if {{tmath|1= \theta = 0.01}}, then {{tmath|1= \theta^3 = 0.000\,001}}, just one ten-thousandth of {{tmath|\theta}}. Thus for many purposes it is safesuffices to drop the cubic and higher terms and approximate the sine and tangent of a small angle using the radian measure of the angle, {{tmath|\sin\theta \approx \tan\theta \approx \theta}}, and drop the quadratic term and approximate the cosine as {{tmath|\cos\theta \approx 1}}.

If additional precision is needed the quadratic and cubic terms can also be included,

<math display=block>{{tmath|\sin\theta \approx \tan\theta - \approx tfrac16\theta.</math>^3}},

<math display=block>{{tmath|\cos\theta \approx 1 - \tfrac12\theta^2.</math>}}, and

The cosine of a sufficiently small angle is very nearly {{tmath|1}}, which sometimes suffices as an approximation. Where more precision is needed the quadratic term can also included, and cosine can be approximated as

{{tmath|\tan\theta \approx \theta + \tfrac13\theta^3}}.

<math display=block>\cos\theta \approx 1 - \tfrac12\theta^2.</math>

==== Dual numbers ====

One may also use [[dual numbers]], defined as numbers in the form <math>a + b\varepsilon</math>, with <math>a,b\in\mathbb R</math> and <math>\varepsilon</math> satisfying by definition <math>\varepsilon^2 = 0</math> and <math>\varepsilon \ne 0</math>. By using the MacLaurin series of cosine and sine, one can show that <math>\cos(\theta\varepsilon) = 1</math> and <math>\sin(\theta\varepsilon) = \theta\varepsilon</math>. Furthermore, it is not hard to prove that the [[Pythagorean identity]] holds:<math display="block">\sin^2(\theta\varepsilon) + \cos^2(\theta\varepsilon) = (\theta\varepsilon)^2 + 1^2 = \theta^2\varepsilon^2 + 1 = \theta^2 \cdot 0 + 1 = 1</math>

==Error of the approximations==

[[File:Small angle compare error.svg|thumb|upright=2300px|'''Figure 3.''' A graph of the [[relative error]]s for the small angle approximations. ({{tmath|\tan \theta \approx \theta}}, {{tmath|\sin \theta \approx \theta}}, {{tmath|\textstyle \cos \theta \approx 1 - \tfrac12\theta^2 }}) ]]

Near zero, the [[relative error]] of the approximations {{tmath|\cos \theta \approx 1}}, {{tmath|\sin \theta \approx \theta}}, and {{tmath|\tan \theta \approx \theta}} is quadratic in {{tmath|\theta}}: for each order of magnitude smaller the angle is, the relative error of these approximations shrinks by two orders of magnitude. The approximation {{tmath|\textstyle \cos \theta \approx 1 - \tfrac12\theta^2 }} has relative error which is quartic in {{tmath|\theta}}: for each order of magnitude smaller the angle is, the relative error shrinks by four orders of magnitude.

Line 79 ⟶ 69:

Figure 3 shows the relative errors of the small angle approximations. The angles at which the relative error exceeds 1% are as follows:

* {{tmath|\cos \theta \approx 1 }} at about 0.14 radians (8.1°)

* {{tmath|\tan \theta \approx \theta}} at about 0.17 radians (9.9°)

* {{tmath|\sin \theta \approx \theta}} at about 0.24 radians (14.0°)

* {{tmath|\textstyle \cos \theta \approx 1 - \tfrac12\theta^2 }} at about 0.66 radians (37.9°)

==Slide-rule approximations==

[[File:K&E Decilon slide rule left end with SRT scale.jpg | thumb | 300px | The left end of a [[Keuffel & Esser]] Deci-Lon slide rule, with a thin blue line added to show the values on the S, T, and SRT scales corresponding to sine and tangent values of 0.1 and 0.01. The S scale shows arcsine(0.1) = 5.74 degrees; the T scale shows arctangent(0.1) = 5.71 degrees; the SRT scale shows arcsine(0.01) = arctangent(0.01) = 0.01*180/pi = 0.573 degrees (to within "slide-rule accuracy").]]

[[File:K&E Decilon slide rule right end with line.jpg | thumb | 300px | The right end of a K&E Decilon slide rule with a line to show the calibration of the SRT scale at 5.73 degrees.]]

Many [[slide rule]]s – especially "trig" and higher models – include an "ST" (sines and tangents) or "SRT" (sines, radians, and tangents) scale on the front or back of the slide, for computing with sines and tangents of angles smaller than about 0.1 radian.<ref>{{cite book |title=Communications Technician M 3 & 2 |date=1965 |publisher=Bureau of Naval Personnel |page=481 |url=https://books.google.com/books?id=FYB3o7iGvb8C&pg=PA481 |access-date=7 March 2025}}</ref>

The right-hand end of the ST or SRT scale cannot be accurate to three decimal places for both arcsine(0.1) = 5.74 degrees and arctangent(0.1) = 5.71 degrees, so sines and tangents of angles near 5 degrees are given with somewhat worse than the usual expected "slide-rule accuracy". Some slide rules, such as the K&E Deci-Lon in the photo, calibrate 0.1 to be accurate for radian conversion, at 5.73 degrees (off by nearly 0.4% for the tangent and 0.2% for the sine for angles around 5 degrees). Others are calibrated to 5.725 degrees, to balance the sine and tangent errors at below 0.3%.

== Angle sum and difference ==

Line 100 ⟶ 99:

=== Astronomy ===

In [[astronomy]], the [[angular size]] or angle subtended by the image of a distant object is often only a few [[arcsecond]]s (denoted by the symbol ″), so it is well suited to the small angle approximation.<ref name=Green1985/> The linear size ({{mvar|D}}) is related to the angular size ({{mvar|X}}) and the distance from the observer ({{mvar|d}}) by the simple formula:

:<math>D = X \frac{d}{206\,265{''}}</math>

where {{mvar|X}} is measured in arcseconds.

The quantity {{val|206265|u="}} is approximately equal to the number of arcseconds in 1 radian, which is the number of arcseconds in a [[circle]] ({{val|1296000|u="}}), divided by {{math|2π}}, or, the number of arcseconds in 1 radian.

The exact formula is

:<math>D = d \tan \left( X \frac{2\pi}{1\,296\,000{''}} \right)</math>

and the above approximation follows when {{math|tan ''X''}} is replaced by {{mvar|X}}.

For example, the [[parsec]] is defined by the value of d when {{mvar|D}}=1 AU, {{mvar|X}}=1 arcsecond, but the definition used is the small-angle approximation (the first equation above).

=== Motion of a pendulum ===

{{main|Pendulum (mechanics)#Small-angle approximation}}

The second-order cosine approximation is especially useful in calculating the [[potential energy]] of a [[pendulum]], which can then be applied with a [[Lagrangian mechanics|Lagrangian]] to find the indirect (energy) equation of motion. When calculating the [[frequency|period]] of a simple pendulum, the small-angle approximation for sine is used to allow the resulting differential equation to be solved easily by comparison with the differential equation describing [[simple harmonic motion]].<ref>{{cite book |last1=Baker |first1=Gregory L. |last2=Blackburn |first2=James A. |chapter=Pendulums somewhat simple |title=The Pendulum: A Case Study in Physics |publisher=Oxford |year=2005 |doi=10.1093/oso/9780198567547.003.0002 |at=Ch. 2, {{pgs|8–26}} |isbn=0-19-856754-5 |chapter-url=https://archive.org/details/pendulumcasestud0000bake/page/8/mode/2up |chapter-url-access=limited }} {{pb}} {{cite journal |last=Bissell |first=John J. |year=2025 |title=Proof of the small angle approximation {{tmath|\sin \theta \approx \theta}} using the geometry and motion of a simple pendulum |journal=International Journal of Mathematical Education in Science and Technology |volume=56 |number=3 |pages=548–554 |doi=10.1080/0020739X.2023.2258885 |doi-access=free }}</ref>

The second-order cosine approximation is especially useful in calculating the [[potential energy]] of a [[pendulum]], which can then be applied with a [[Lagrangian mechanics|Lagrangian]] to find the indirect (energy) equation of motion.

When calculating the [[frequency|period]] of a simple pendulum, the small-angle approximation for sine is used to allow the resulting differential equation to be solved easily by comparison with the differential equation describing [[simple harmonic motion]].

===Optics===

In optics, the small-angle approximations form the basis of the [[paraxial approximation]].

=== Wave Interferenceinterference ===

The sine and tangent small-angle approximations are used in relation to the [[double-slit experiment]] or a [[diffraction grating]] to develop simplified equations like the following, where {{mvar|y}} is the distance of a fringe from the center of maximum light intensity, {{mvar|m}} is the order of the fringe, {{mvar|D}} is the distance between the slits and projection screen, and {{mvar|d}} is the distance between the slits: <ref>{{Cite web|url=http://hyperphysics.phy-astr.gsu.edu/hbase/phyopt/slits.html|title=Slit Interference}}</ref><math display="block">y \approx \frac{m\lambda D}{d}

</math>