MHB What is the minimum value of a trigonometric expression with a given condition?

  • Thread starter Thread starter juantheron
  • Start date Start date
  • Tags Tags
    Minimum Trig
Click For Summary
The minimum value of the expression \( \cot^2(A) + \cot^2(B) + \cos^2(C) \) under the condition \( A + B + C = \pi \) is determined to be 1, achieved when \( A = B = C = \frac{\pi}{3} \). The discussion involves using properties of trigonometric functions and inequalities, specifically the Arithmetic Mean-Geometric Mean inequality. It is noted that while equality holds for the cotangent terms, it does not necessarily imply that \( A = B = C \) without additional restrictions. The optimization approach involves converting the problem into two variables and applying calculus techniques to find critical points. The findings emphasize the importance of constraints in determining the minimum values of trigonometric expressions.
juantheron
Messages
243
Reaction score
1
If $A+B+C=\pi$. Then Minimum value of $\cot^2(A)+\cot^2(B)+\cos^2(C)$ is
 
Physics news on Phys.org
jacks said:
If $A+B+C=\pi$. Then Minimum value of $\cot^2(A)+\cot^2(B)+\cos^2(C)$ is

What have you tried?

If you put \(C=\pi-(A+B) \) and substitute into the objective you get an unconstrained optimisation problem in two variables.

CB
 
Thanks Caption Black But I did not Understand What steps i do after that means after converting into 2 variable.

would you like to explain it to meWe Know that If A+B+C = \pi, Then \tan (A)+\tan(B)+\tan(C) = \tan(A).\tan(B).\tan(C)Which we can prove easilyA+B=\pi-C\Leftrightarrow \tan(A+B) = \tan (\pi-C) = -\tan (C)So \frac{\tan(A)+\tan(B)}{1-\tan(A).\tan(B)} = -\tan (C)So \tan (A)+\tan(B)+\tan(C) = \tan(A).\tan(B).\tan(C)Now Using \mathbb{A.M}\geq \mathbb{G.M}\frac{\tan (A)+\tan(B)+\tan(C) }{3}\geq \left(\tan(A).\tan(B).\tan(C)\right)^{\frac{1}{3}}\frac{\tan(A).\tan(B).\tan(C)}{3}\geq \left(\tan(A).\tan(B).\tan(C)\right)^{\frac{1}{3}}\left(\tan(A).\tan(B).\tan(C)\right)^3\geq 27 \left(\tan(A).\tan(B).\tan(C)\right)So \left(\tan(A).\tan(B).\tan(C)\right)\geq 3\sqrt{3}bcz \tan(A).\tan(B).\tan(C)> 0So \cot(A).\cot(B).\cot(C)\leq \frac{1}{3\sqrt{3}}Now again Using \mathbb{A.M}\geq \mathbb{G.M}\frac{\cot^2(A)+\cot^2(B)+\cot^2(C)}{3}\geq \left(\cot(A).\cot(B).\cot(C)\right)^{\frac{2}{3}}So \cot^2(A)+\cot^2(B)+\cot^2(C)\geq 1and equality hold when A=B=C=\frac{\pi}{3}
 
jacks said:
Thanks Caption Black But I did not Understand What steps i do after that means after converting into 2 variable.

would you like to explain it to meWe Know that If A+B+C = \pi, Then \tan (A)+\tan(B)+\tan(C) = \tan(A).\tan(B).\tan(C)Which we can prove easilyA+B=\pi-C\Leftrightarrow \tan(A+B) = \tan (\pi-C) = -\tan (C)So \frac{\tan(A)+\tan(B)}{1-\tan(A).\tan(B)} = -\tan (C)So \tan (A)+\tan(B)+\tan(C) = \tan(A).\tan(B).\tan(C)Now Using \mathbb{A.M}\geq \mathbb{G.M}\frac{\tan (A)+\tan(B)+\tan(C) }{3}\geq \left(\tan(A).\tan(B).\tan(C)\right)^{\frac{1}{3}}\frac{\tan(A).\tan(B).\tan(C)}{3}\geq \left(\tan(A).\tan(B).\tan(C)\right)^{\frac{1}{3}}\left(\tan(A).\tan(B).\tan(C)\right)^3\geq 27 \left(\tan(A).\tan(B).\tan(C)\right)So \left(\tan(A).\tan(B).\tan(C)\right)\geq 3\sqrt{3}bcz \tan(A).\tan(B).\tan(C)> 0So \cot(A).\cot(B).\cot(C)\leq \frac{1}{3\sqrt{3}}Now again Using \mathbb{A.M}\geq \mathbb{G.M}\frac{\cot^2(A)+\cot^2(B)+\cot^2(C)}{3}\geq \left(\cot(A).\cot(B).\cot(C)\right)^{\frac{2}{3}}So \cot^2(A)+\cot^2(B)+\cot^2(C)\geq 1and equality hold when A=B=C=\frac{\pi}{3}

This is a different question from the one you asked in the original post. Please clarify what the question really is.

Also as this is posted in the calculus area of MHB I would expect to see some calculus.

CB
 
jacks said:
So \cot^2(A)+\cot^2(B)+\cot^2(C)\geq 1and equality hold when A=B=C=\frac{\pi}{3}

By the way equality holds when \( \cot^2(A)=\cot^2(B)=\cot^2(C) \) which does not imply that \(A=B=C \) unless you place some restriction on the allowed values of \(A,B\) and \(C\), which you have not done.

CB
 
jacks said:
Thanks Caption Black But I did not Understand What steps i do after that means after converting into 2 variable.

would you like to explain it to me

Given a function \(f(x,y,z)\) for which you seek a minimum subject to a constraint \(z=g(x,y)\), we look for the unconstrained minimum of \(h(x,y)=f(x,y,g(x,y))\).

Such a minimum if it exists is a solution of:

\[ \begin{aligned} \frac{\partial}{\partial x}h(x,y)=0 \\ \frac{\partial}{\partial y}h(x,y)=0 \end{aligned} \]

Applying this to the given problem of minimising \(f(A,B,C)= \cot^2(A)+\cot^2(B)+\cot^2(C)\) subject to \(A+B+C=\pi\) , gives us a condition:

\[\frac{\cos(A)}{\sin^3(A)}=\frac{\cos(B)}{\sin^3(B)}=\frac{\cos(C)}{\sin^3(C)}\]

and as \( \cos(x)/\sin^3(x) \) strictly decreasing over \((0,\pi)\) and periodic with period \( \pi \) this implies that \( B=A+n\pi,\ C=A+m\pi ,\ n,m \in \mathbb{Z}\). So:

\[ A=\frac{\pi(1-n-m)}{3} \]

(Not all of these critical point correspond to minima, in particular the cases where \(1-n-m\) is a multiple of \(3\) are not)

Now since \( \cot^2(x) \) is periodic with period \(\pi\) at a critical point we have:

\[f(A,B,C)= \cot^2(A)+\cot^2(B)+\cot^2(C)=3\;\cot^2(A)=3\;\cot^2\left( \frac{\pi(1-n-m)}{3} \right)\]

and a bit more checking shows that the mimimum is \(1\) and occurs whenever \(1-n-m\) is not a multiple of \(3\).

Note: if we had specified that \(0\le A,B,C \le \pi\) at the start we would have found that the critical point is \(A=B=C=\pi/3\) without much trouble.

CB
 
Last edited:
CaptainBlack said:
By the way equality holds when \( \cot^2(A)=\cot^2(B)=\cot^2(C) \) which does not imply that \(A=B=C \) unless you place some restriction on the allowed values of \(A,B\) and \(C\), which you have not done.

CB

Yes CaptionBlack
 

Similar threads

  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 3 ·
Replies
3
Views
2K
  • · Replies 2 ·
Replies
2
Views
1K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 47 ·
2
Replies
47
Views
7K
  • · Replies 1 ·
Replies
1
Views
2K
Replies
4
Views
2K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 12 ·
Replies
12
Views
4K
  • · Replies 20 ·
Replies
20
Views
1K