4

I would like to find the

$$\color{red}{\huge\sin 12°}$$ without a scientific calculator and Maclaurin formula because my students of an high school don't know this approach. I have thought starting from $\sin 18°=(\sqrt 5-1)/4$ and the cosine $\cos 18°=(\sqrt{10}+2\sqrt 5)/4$ (from the geometry we know that the side of the regular decagon is the golden section of the radius).

Being $\sin (2\alpha)=2\sin(\alpha)\cos(\alpha)$ I know $\sin 36°$, but for the $\sin 12°$ I have thought $$\cos3α=\cos(2α+α)=\cos2α\cosα−\sin2α\sinα=$$$$\cosα⋅(2\cos2α−1)−2\sin2α\cosα$$

$$=2\cos3α−\cosα−2\cosα(1−\cos2α)=$$$$2\cos3α−\cosα−2\cosα+2\cos3α=4\cos3α−3\cosα$$ Considering the formulas of trisection of an angle I have: $$\boxed{\cosα=\frac 43\cos3α −3\cos \frac\alpha3}$$ We could put $\cos(α/3)=z$ and thus I will have the equation

$$4z^3-3z-\cosα=0$$

It is then a matter of solving that third degree equation, of which there is a solution method (which reminds me of the famous querelle between Tartaglia and Cardano...).

My problem it is this:

You want to build a platform like the one in the figure to overcome a height difference $h$ of 15 cm. The inclination must be 12°. What is the length $\ell$ of the platform? (the link is https://invalsi.zanichelli.it/taoDelivery/DeliveryServer/runDeliveryExecution?deliveryExecution=kve_de_https%3A%2F%2Finvalsi.zanichelli.it%2Ffirst.rdf%23i164517281905829660045946)

enter image description here

My students have not a scientific calculator during the test.

enter image description here

How can I solve the problem simply?

coudy
  • 5,843
Sebastiano
  • 7,649

4 Answers4

7

A $\large 1400^+$ years old approximation is $$\sin (12 {}^{\circ})\sim\frac{224}{1069}$$ which is a relative error of $0.78$% and absolute error of $1.6\times 10^{-3}$.

This will make your students aware that there were very good mathematicians long time ago.

3

The degree 3 equation has an explicit solution but it is a bit complicated. $$\sin(12^\circ) = {1\over 4} \sqrt{{5+\sqrt{5} \over 2}} -{\sqrt{3}(\sqrt{5}-1)\over 8}$$

As far as I understand, you want an approximation to the value $\sin(12^\circ)$ that may replace the one given by a calculator.

Before the advent of calculators, engineers used the following approximation for the sine, assuming $x$ small, given in degrees. $$ \sin(x) \simeq {x\over 60}$$ In your case, this gives $$\sin(12^\circ) \simeq {12 \over 60} = 0.2$$ This is close to the true value which is $\sin(12^\circ) = 0.20791169081775931$...

coudy
  • 5,843
  • Where can I know that $\sin(x)\simeq x/60$? Thank you very much for your response. – Sebastiano Feb 18 '22 at 08:49
  • 1
    @Sebastiano $\sin x\approx x$ when $x$ is small and in radians. The factor $60$ is here because $60\approx 180/\pi$. –  Feb 18 '22 at 08:52
  • 1
    This is an heuristic used by engineers that allows for a quick mental computation. You can check that $0 \leq \sin(x)-x/60 < 0.01$ for $0 \leq x \leq 30^\circ$ if $x$ is given in degree, just by studying the function $x\mapsto \sin(x)-x/60$ and finding its maximum. – coudy Feb 18 '22 at 08:54
  • @StinkingBishop Thank youuuuu very much either. – Sebastiano Feb 18 '22 at 09:01
  • For the record, ${21 \over 101}$ is a slightly better approximation of $\sin(12^\circ) = \sin(\pi/15)$ obtained by the algorithm of continued fraction (and better than $224/1069$). – coudy Feb 22 '22 at 22:08
1

It's easy. First, we find the value of $\sin(18^{\circ})$.

Observe that $\sin(36^{\circ})=\cos(54^{\circ})\Longleftrightarrow 2\sin(18^{\circ})\cos(18^{\circ})=\cos(18^{\circ})\cos(36^{\circ})-\sin(18^{\circ})\sin(36^{\circ})$

$\Longleftrightarrow 2\sin(18^{\circ})\cos(18^{\circ})=\cos(18^{\circ})(1-2\sin^2(18^{\circ}))-\sin(18^{\circ})2\sin(18^{\circ})\cos(18^{\circ})$

$\Longleftrightarrow 2\sin(18^{\circ})=1-2\sin^2(18^{\circ})-2\sin^2(18^{\circ})$

$\Longleftrightarrow 4\sin^2(18^{\circ})+2\sin(18^{\circ})-1=0$

Easy to solve quadratic equations. it is obtained that $\sin(18^{\circ})=\dfrac{\sqrt{5}-1}{4}$. By Identity $(\sin^2(x)+\cos^2(x)=1)$. and remembering that $0<\cos(18^{\circ})<1$. we get $\cos(18^{\circ})=\sqrt{\dfrac{5+\sqrt{5}}{8}}$

Now, note that $\cos(18^{\circ})=\sin(72^{\circ})$, $\sin(18^{\circ})=\cos(72^{\circ})$ and $\sin(12^{\circ})=\sin(72^{\circ}-60^{\circ})$.

$\sin(12^{\circ})=\sin(72^{\circ})\cos(60^{\circ})-\cos(72^{\circ})\sin(60^{\circ})$

$$\sin(12^\circ) = {1\over 4} \sqrt{{5+\sqrt{5} \over 2}} -{\sqrt{3}(\sqrt{5}-1)\over 8}$$

With similar step even we can easily find value of $\sin(3^{\circ})=\sin(75^{\circ}-72^{\circ})=\sin(18^{\circ}-15^{\circ})$

and, we can get the approximation value of $\sin(12^{\circ})$ using Newton Method

Newbie000
  • 148
1

When an angle is narrow, the hypoteneuse is almost parallel with the adjacent edge and both are close to a pair of radius lines from the centre of a regular $n$-gon to two neighbouring vertices. This gives you $$\sin\theta=\dfrac{\text{opp}}{\text{adj}}\approxeq \dfrac{2\pi\cdot r/n}r=\dfrac{2\pi}n$$

$n$ here is how many edges your $n$-gon has, which in this case is $360/12=30$

Giving you $\approxeq\dfrac{\pi}{15}\approxeq0.209$ which is close to the true value of $0.2079\ldots$


If you wanted you could take $\pi$ to be roughly $22/7$ in which case you get the fraction $\dfrac{22}{105}$ which again is good to nearly three decimal places.

  • Thank you very much for your explanation. I have not understood the last part on $\pi$. Please can you extend the explanatation. – Sebastiano Nov 10 '22 at 20:14
  • 1
    Originally to get $0.2079$ I divided $3.14159263\ldots$ by $15$. In fact I used a calculator to do the division, which you asked us not to do. Now $3.14159263\ldots$ is a decimal approximation to $\pi$. If you wanted to do it without a scientific calculator, you can use $22/7$ which is just another well-known approximation to $\pi$ - this time it's a rational fraction and not a decimal one. So if you put that into the fraction $\pi/15$ you can get your final answer ($22/7)\div 15=\frac{22}{105}$ as a fraction. You can even do a long division of $105$ into $22$ and you would get $0.20\ldots$ – it's a hire car baby Nov 11 '22 at 09:54
  • 1
    Here you go, the pics on this page show you how when the angle is small, the triangles are very nearly right angle triangles because the triangles are isosceles with two angles very close to right angles. https://www.maa.org/external_archive/joma/Volume7/Aktumen/Polygon.html – it's a hire car baby Nov 11 '22 at 09:59
  • 1
    Thank you very much for your new explanation- I will give you any points. – Sebastiano Nov 11 '22 at 12:43