Inactive Tutor answered 02/06/18
Tutor
New to Wyzant
y'=(x^2-y^2)xy, with y(1)=2
First equation.
y'=x3y-xy3 (This is, so called, Bernoulli equation)
y'-x3y = -xy3 (multiply both sides by y-3)
y-3y'-x3y-2=-x
Now we make substitution u=y-2, where u=u(x) is a new unknown function. Derivative of u: u'=-2y-3y', -u'/2=y-3y'.
Therefore
-u'/2-x3u=-x
or
u'+2x3u=2x (this is first order linear equation) (**)
We solve this equation and then go back to y
First this : u'+2x3u=0. If you solve this you get u=C e^(-x4/4), C=const
Assume that C=C(x) (that is, for time-being not constant but function of x) and substitute into equation (**).
If you do that, you find
C(x)= C0-e^(-x4/4) and
u(x)=C0e^(-x4/4)-e^(-x4/2).
remind that u=y-2
So
y-2=C0-e^(-x4/4)-e^(-x4/2).
y=1/(C0e^(-x4/4)-e^(-x4/2))^(1/2)
You have y(1)=2. So
1/4=C0- e^(-1/4)-e^(-1/2)
C0=1/4+ e^(-1/4)-e^(-1/2)
Plug in C0 and you have that solution with initial condition y(1)=2.
Note:I could made mistakes in details. If you go through every step yourself, you might find inconsistencies (if any). But, say, the approach is correct.
Of course you can solve it numerically as suggested Andy C. Numerical solution is good. This is in addition to that.
Inactive Tutor
Alright. The equation we have is
y'=(x^2-y^2)xy.
Nothing special, I just multiplied xy by "whatever is inside the parentheses": (x^2-y^2)xy= x^2*(xy)-y^2*(xy)=x^3y-xy^3.
So you have
y'=x^3y-xy^3
I hope it is clear now. Feel free to ask if you have more questions.
In principle there is a more "compact way" to solve the equation but might be not so clear.
Report
02/06/18
Julia K.
ah makes sense. With all my diff eq hw I seem to keep getting stuck over simple algebra lol thank you Bobosharif!
Report
02/10/18
Inactive Tutor
You are welcome Julia!
Report
02/10/18
Julia K.
02/06/18