The Student Room Group

Official TSR Mathematical Society

Scroll to see replies

DFranklin
If your integral is of the form f(t)dt\int f(t) \, dt you are only guaranteed to get the right result with a substitution of form u = g(t) if g' is never zero (in the interval under consideration).


Why is that?
Basically you end up dividing by g'(t).
Reply 1582
Show that if a+b+c=0a+b+c=0 and a,b,c lie on the unit circle then a,b,c form the vertices of an equilateral triangle
SimonM
Show that if a+b+c=0a+b+c=0 and a,b,c lie on the unit circle then a,b,c form the vertices of an equilateral triangle

I'm not sure if this would be considered a stupid question, but by a + b + c = 0, do you mean the x-co-ordinates, or the y, or am I just showing how little maths I know?
Is it the sum of the x-coordinates and also (separately) the sum of the y-coordinates?
Aha, thank you.
SimonM
Show that if a+b+c=0a+b+c=0 and a,b,c lie on the unit circle then a,b,c form the vertices of an equilateral triangle


roots of unity?

I think you need to consider a,b,c, as vectors
Reply 1587
a,b,c are complex numbers

I can't think of a good hint which doesn't give it all away, I guess "is there something you can do WLOG"
if a+b+c = 0 then the circumcircle of the triangle formed by these three points is centred at the origin the WLOG we can considerer only a,b,c of unit modulus we can also WLOG rotate it as this conserves its shape. If we note that we can rotate it till one of the points, say for instance a, is 1; we now need only show if 1+b+c= 0 then 1,b,c is an equilateral triangle.
letting b = cos(t)+isin(t) and c = cos(s)+isin(s) and spiting into real and imaginary parts, Im: sin(t)+sin(s)=0 => sin(t)=-sin(s) => t=-s and Re: 1+cos(s)+cos(t)=0 => 2cos(t)=1 => cos(t)= -1/2 => t = 2pi/3 , s = -2pi/3, so the points must form an equilateral triangle. thus for any complex numbers a+b+c=0 a,b,c must be the points of an equilateral triangle QED
Not tough but in the MIT integration speed challenge - see how quick it can be done;

x4x2+1dx \displaystyle \int \frac{x^4}{x^2 + 1} dx
Reply 1590
DeanK22
Not tough but in the MIT integration speed challenge - see how quick it can be done;

x4x2+1dx \displaystyle \int \frac{x^4}{x^2 + 1} dx


x41+1x2+1=(x21+1x2+1)dx=x33x+tan1x+C\displaystyle \int \frac{x^4-1+1}{x^2+1} = \int \left ( x^2-1+\frac{1}{x^2+1} \right) dx = \frac{x^3}{3} - x + \tan^{-1} x + C
Reply 1591
DeanK22
in the MIT integration speed challenge


What's this?
That was x41x2dx\int \frac{x^4}{1-x^2}\,dx, surely?

(Pretty easy, to be honest - I suspect many people here could have done as well).
DFranklin
That was x41x2dx\int \frac{x^4}{1-x^2}\,dx, surely?

(Pretty easy, to be honest - I suspect many people here could have done as well).


Yeah, it seems more a test of how fast you can write. The contestants weren't impressively quick.
x(1+x)3(1+x+x2)12dx \displaystyle \int \frac{x}{(1+x)^3(1+x+x^2)^{\frac{1}{2}}} dx

Spoiler

DeanK22
Not tough but in the MIT integration speed challenge - see how quick it can be done;

x4x2+1dx \displaystyle \int \frac{x^4}{x^2 + 1} dx


I had this integral in a STEP question yesterday.
A problem I came up with and did whilst I was bored in physics:

By making a substitution of the form y=dxdt+kxy = \frac{dx}{dt} + kx, prove (without guessing the form of the solution) that the differential equation d2xdt2+bdxdt+cx=0\frac{d^2x}{dt^2} + b\frac{dx}{dt} + cx = 0 has general solution
x=Aeλt+Beμt,μλx = Ae^{\lambda t} + Be^{\mu t}, \mu \neq \lambda
or x=(At+B)eλt,μ=λx = (At + B)e^{\lambda t}, \mu = \lambda
where μ\mu and λ\lambda are the roots of the equation z2+bz+c=0z^2 + bz + c = 0.
Reply 1598
Eurgh, I prefer the argument using vector spaces tbh
I know nothing about vector spaces :p:

It was more interesting than whatever pseudo-quantum-physics crap we were supposed to be doing, anyway.

Quick Reply

Latest