loveinla
- 11
- 0
The solution is so horrible as to be essentially unusable. Maple manages to find four roots, by finding and solving the equivalent fourth-degree polynomial. One of the roots isloveinla said:View attachment 216933
Here a, b, c > 0, and a > bc.
Can anyone find the solution of k as a function of (a, b, c)? Thanks.
I'm not sure that I would call that useless. A computer doesn't get tired.Ray Vickson said:The solution is so horrible as to be essentially unusable. Maple manages to find four roots, by finding and solving the equivalent fourth-degree polynomial. One of the roots is
r1 =
-15/58*b*c+15/29*a+1/2*(1/4*(30/29*b*c-60/29*a)^2-44/87*b^2*c^2+40/29*a*b*c-26/29*a^2+1/87*(1126*b^6*c^6-3132*a*b^5*c^5+1170*a^2*b^4*c^4+3582*a^3*b^3*c^3-2772*a^4*b^2*c^2-1026*a^5*b*c+1053*a^6+6*(-3807*b^12*c^12+67500*a*b^11*c^11-460278*a^2*b^10*c^10+1562622*a^3*b^9*c^9-2919687*a^4*b^8*c^8+3050688*a^5*b^7*c^7-1736700*a^6*b^6*c^6+783348*a^7*b^5*c^5-884757*a^8*b^4*c^4+841092*a^9*b^3*c^3-314046*a^10*b^2*c^2-7650*a^11*b*c+21675*a^12)^(1/2))^(1/3)-87*(-112/7569*b^4*c^4+28/841*a*b^3*c^3-68/2523*a^2*b^2*c^2+44/2523*a^3*b*c-23/2523*a^4)/(1126*b^6*c^6-3132*a*b^5*c^5+1170*a^2*b^4*c^4+3582*a^3*b^3*c^3-2772*a^4*b^2*c^2-1026*a^5*b*c+1053*a^6+6*(-3807*b^12*c^12+67500*a*b^11*c^11-460278*a^2*b^10*c^10+1562622*a^3*b^9*c^9-2919687*a^4*b^8*c^8+3050688*a^5*b^7*c^7-1736700*a^6*b^6*c^6+783348*a^7*b^5*c^5-884757*a^8*b^4*c^4+841092*a^9*b^3*c^3-314046*a^10*b^2*c^2-7650*a^11*b*c+21675*a^12)^(1/2))^(1/3))^(1/2)-1/2*(1/2*(30/29*b*c-60/29*a)^2-88/87*b^2*c^2+80/29*a*b*c-52/29*a^2-1/87*(1126*b^6*c^6-3132*a*b^5*c^5+1170*a^2*b^4*c^4+3582*a^3*b^3*c^3-2772*a^4*b^2*c^2-1026*a^5*b*c+1053*a^6+6*(-3807*b^12*c^12+67500*a*b^11*c^11-460278*a^2*b^10*c^10+1562622*a^3*b^9*c^9-2919687*a^4*b^8*c^8+3050688*a^5*b^7*c^7-1736700*a^6*b^6*c^6+783348*a^7*b^5*c^5-884757*a^8*b^4*c^4+841092*a^9*b^3*c^3-314046*a^10*b^2*c^2-7650*a^11*b*c+21675*a^12)^(1/2))^(1/3)+87*(-112/7569*b^4*c^4+28/841*a*b^3*c^3-68/2523*a^2*b^2*c^2+44/2523*a^3*b*c-23/2523*a^4)/(1126*b^6*c^6-3132*a*b^5*c^5+1170*a^2*b^4*c^4+3582*a^3*b^3*c^3-2772*a^4*b^2*c^2-1026*a^5*b*c+1053*a^6+6*(-3807*b^12*c^12+67500*a*b^11*c^11-460278*a^2*b^10*c^10+1562622*a^3*b^9*c^9-2919687*a^4*b^8*c^8+3050688*a^5*b^7*c^7-1736700*a^6*b^6*c^6+783348*a^7*b^5*c^5-884757*a^8*b^4*c^4+841092*a^9*b^3*c^3-314046*a^10*b^2*c^2-7650*a^11*b*c+21675*a^12)^(1/2))^(1/3)+((22/29*b^2*c^2-60/29*a*b*c+39/29*a^2)*(30/29*b*c-60/29*a)-16/29*b^3*c^3+52/29*a*b^2*c^2-56/29*a^2*b*c+20/29*a^3-1/4*(30/29*b*c-60/29*a)^3)/(1/4*(30/29*b*c-60/29*a)^2-44/87*b^2*c^2+40/29*a*b*c-26/29*a^2+1/87*(1126*b^6*c^6-3132*a*b^5*c^5+1170*a^2*b^4*c^4+3582*a^3*b^3*c^3-2772*a^4*b^2*c^2-1026*a^5*b*c+1053*a^6+6*(-3807*b^12*c^12+67500*a*b^11*c^11-460278*a^2*b^10*c^10+1562622*a^3*b^9*c^9-2919687*a^4*b^8*c^8+3050688*a^5*b^7*c^7-1736700*a^6*b^6*c^6+783348*a^7*b^5*c^5-884757*a^8*b^4*c^4+841092*a^9*b^3*c^3-314046*a^10*b^2*c^2-7650*a^11*b*c+21675*a^12)^(1/2))^(1/3)-87*(-112/7569*b^4*c^4+28/841*a*b^3*c^3-68/2523*a^2*b^2*c^2+44/2523*a^3*b*c-23/2523*a^4)/(1126*b^6*c^6-3132*a*b^5*c^5+1170*a^2*b^4*c^4+3582*a^3*b^3*c^3-2772*a^4*b^2*c^2-1026*a^5*b*c+1053*a^6+6*(-3807*b^12*c^12+67500*a*b^11*c^11-460278*a^2*b^10*c^10+1562622*a^3*b^9*c^9-2919687*a^4*b^8*c^8+3050688*a^5*b^7*c^7-1736700*a^6*b^6*c^6+783348*a^7*b^5*c^5-884757*a^8*b^4*c^4+841092*a^9*b^3*c^3-314046*a^10*b^2*c^2-7650*a^11*b*c+21675*a^12)^(1/2))^(1/3))^(1/2))^(1/2)
and there are three more like that.
And, of course, one problem is that the process of converting the problem go a polynomial-solving task can introduce extraneous solutions---solutions of the polynomial that are not solutions of the original equation. Telling which of the four symbolic solutions fails to solve the original equation is just about impossible with expressions as complicated as the above. However, looking at some numerical examples, we can just plot a graph of your ##f(k)## and see how many real roots it has. For example, when ##a=3, b=1, c=2## (which satisfies your restriction ##a > bc##) we have the equation ##F(k) = 0##, where
$$F(k) = 2-6k+19k^2 + (2-7k) \sqrt{k(5k+8)}$$
The graph of ##F(k)## crosses zero in just two points, ##k = 0.542990634## and ##k = 3.371514666##. The other two (complex) roots of the 4th degree polynomial are decidedly not roots of the original, unconverted equation---not even complex roots of ##F(k)##.
Sure, but you can ask a computer to get numerical solutions without that expression as well.FactChecker said:I'm not sure that I would call that useless. A computer doesn't get tired.
I agree. I see some large powers like a^12. I was half joking about the computer not getting tired, but not really sure if that formula would be useful or not..mfb said:Sure, but you can ask a computer to get numerical solutions without that expression as well.
I said "really messy", but that expression is even worse than I expected.