I have to prove that for a triangle, $$a\cos A+b\cos B+c\cos C=\frac{8\Delta^2}{abc}$$ where $a,b,c$ are the lengths of the sides opposite to the angles A,B,C respectively. I followed the following procedure for the LHS: $$ \begin {align} a\cos A&+b\cos B+c\cos C\\ &= a\left[2\cos^2\left(\frac A2\right)-1\right]+b\left[2\cos^2\left(\frac B2\right)-1\right]+c\left[2\cos^2\left(\frac C2\right)-1\right]\\ &=a\left[2 \frac{s(s-a)}{bc} -1\right]+b\left[2 \frac{s(s-b)}{ac} -1\right]+c\left[2 \frac{s(s-c)}{ab} -1\right]\\ &=2a\left (\frac{s(s-a)}{bc}\right)+2b\left (\frac{s(s-b)}{ac}\right)+2c\left( \frac{s(s-c)}{ab}\right)-2s\\ &=\frac{2s}{abc}[a^2(s-a)+b^2(s-b)+c^2(s-c)-abc] \end{align}$$ where $$s=\frac{a+b+c}{2}$$

I want to convert that last equation into Heron's formula: $$\Delta=\sqrt{s(s-a)(s-b)(s-c)}$$

But I'm stuck there. Any help please?


From the cosine rule:



$$\begin{align}a\cos(A)+b\cos(B)+c\cos(C)&=a\frac{b^2+c^2-a^2}{2bc}+b\frac{a^2+c^2-b^2}{2ac}+c\frac{a^2+b^2-c^2}{2ab}\\ &=a^2\frac{b^2+c^2-a^2}{2abc}+b^2\frac{a^2+c^2-b^2}{2abc}+c^2\frac{a^2+b^2-c^2}{2abc}\\ &=\frac{a^2b^2+a^2c^2-a^4+a^2b^2+b^2c^2-b^4+a^2c^2+b^2c^2-c^4}{2abc}\\ &=\frac{-(a-b-c)(a+b-c)(a+c-b)(a+b+c)}{2abc}\\ &=\frac{(b+c-a)(a+b-c)(a+c-b)(a+b+c)}{2abc}\\ &=\frac{(2s-2a)(2s-2c)(2s-2b)(2s)}{2abc}\\ &=\frac{8(s-a)(s-c)(s-b)(s)}{abc}\\ &=\frac{8\Delta^2}{abc}\\ \end{align}$$

  • $\begingroup$ That's great! :) I tried this method too. But I wasn't able to factorise the terms in your step 3. Can you tell me how you found the four factors? $\endgroup$ – Tejas Feb 5 '15 at 3:29
  • $\begingroup$ @Tejas If you want to get somewhere you can always cheat/work backwards. $\endgroup$ – Pp.. Feb 5 '15 at 3:37
  • $\begingroup$ @Pp.. Right! However, it would really help me knowing the method, in case I didn't know what lies at the other side of equality. $\endgroup$ – Tejas Feb 5 '15 at 3:43
  • $\begingroup$ @Tejas The method was: working backwards. What you were doing was more close to deducing the factorization because you took out the factor $s$ instead of multiplying it. When you do that you get a cubic polynomial (say cubic in $c$) and you could solve for $c$ and find the root $a+b$. This tells you that there is a factor of the form $c-(a+b)$. Being cyclic it tells you there are also factors $a-(b+c)$ and $b-(a+c)$. $\endgroup$ – Pp.. Feb 5 '15 at 3:55
  • $\begingroup$ To be perfectly honest - I used Wolfram Alpha to do the factorisation - in general, factorisation requires a stoke of genius or bloody hard work; I don't get that many strokes of genius and I don't like to work that hard. $\endgroup$ – Dale M Feb 5 '15 at 5:16


Use Law of sines $a=2R\sin A$ etc.

Now Double-Angle Formula $2\sin A\cos A=\sin2A$

Then Prove that $\sin(2A)+\sin(2B)+\sin(2C)=4\sin(A)\sin(B)\sin(C)$ when $A,B,C$ are angles of a triangle

Finally use $\triangle=\dfrac{abc}{4R}$ (Proof)


The polynomial $$\frac{a^2(s-a)+b^2(s-b)+c^2(s-c)-abc}{4}$$ vanishes for $c:=a+b$ and because it is cyclic it vanishes also at $a:=b+c$ and at $b:=a+c$. Notice that the same is true for $$(s-a)(s-b)(s-c).$$

In fact, for the second polynomial it is clear and for the first putting $c:=a+b$ in the first polynomial we get

$$\frac{2ba^2+2ab^2-2ab(a+b)}{8}\equiv 0.$$

Finally the two polynomials have the same leading coefficient $\frac{1}{8}$.

Therefore $$\frac{a^2(s-a)+b^2(s-b)+c^2(s-c)-abc}{4}=(s-a)(s-b)(s-c).$$

  • $\begingroup$ What is the meaning of ':='? Is it the same as '='? $\endgroup$ – Tejas Feb 5 '15 at 3:29
  • $\begingroup$ @Tejas Yes. I just put the colon to emphasize we are going to assign $a+b$ to where $c$ is in the expression. $\endgroup$ – Pp.. Feb 5 '15 at 3:36
  • $\begingroup$ But we can't have $c=a+b$ in a triangle, isn't it? $\endgroup$ – Tejas Feb 5 '15 at 3:40
  • $\begingroup$ @Tejas It doesn't matter. We are proving that two polynomials are identically equal (as opposed to equal only when the input are sides of a triangle). The idea is that you check that the two polynomials have the same roots and have the same leading coefficient. This is enough to say they are equal. It is a useful technique to remember for many proofs like this. You don't have to get dirty expanding everything. $\endgroup$ – Pp.. Feb 5 '15 at 3:43



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.