3

If the line $x+y-1=0$ is a tangent to a parabola with focus (1,2) at A and intersects the directrix at B and tangent at vertex C respectively, then find the value of AC.BC

First of, they have given absolutely no infomation regarding the parabola, leaving me with no choice but to assume some standard case

If it is of the form $y^2=4ax$, the axis of the parabola is $y=2$

So $$(y-2)^2=4a(x-x_1)$$

The equation of tangent to the parabola is $$y-2=m(x-x_1)+\frac am$$ Putting m=-1

$$y-2=x_1-x-a$$ $$x+y=2+x_1-a$$

So $$2+x_1-a=1$$ $$x_1=a-1$$

The distance from vertex to focus is +a So
As given $$x_1+a=1$$ x coordinate of focus given $$(a-1)+a=1$$ $$a=1$$ Then $$x_1=0$$ Now vertex=(0,2) U can try now finding directrix and its intersection point with tangent and intersection point with tangent at vertex then u get

                AC= square root of 2 
                BC=square root of 2 
   Now,
        AC.BC=2

That’s all I could do.

Aditya
  • 6,191
  • The equation for the parabola can be written $$(l^2+m^2)((x-1)^2+(y-2)^2)-(lx+my+n)^2=0$$ with an arbitrary directrix $lx+my+n=0$

    Experimenting in geogebra there seems to be an infinity of configurations with this parabola tangent to $x+y-1=0$ and for a few of them the value of $AC.BC$ seems to vary.

    – Jan-Magnus Økland Feb 27 '20 at 08:58
  • Or maybe the answer is $2$. – Jan-Magnus Økland Feb 27 '20 at 09:35

1 Answers1

2

A partial answer.

The parabola with focus $F=(1,2)$ and directrix $lx+my+l=0$ has equation $$(l^2+m^2)((x-1)^2+(y-2)^2)-(lx+my+l)^2=0$$ and experimenting in geogebra I found it happens to meet $y+x-1=0$ doubly in $A$: $(x,y)=(\frac{l-m}{l+m},\frac{2m}{l+m})$ (substitute the line $y=1-x$ in the equation for the parabola and check that it factors as $(mx+lx+m-l)^2$) so it's tangent there for all $l,m$ with $l+m\neq 0$.

The directrix $lx+my+l=0$ meets $y+x-1=0$ in $B:(x,y)=(-\frac{l+m}{l-m},\frac{2l}{l-m})$.

The axis is parallel to $ -mx +ly$ and passes through the focus, and intersects the parabola in the vertex which is $V:(x,y)=(\frac{m(m-l)}{l^2+m^2},\frac{m^2-lm+2l^2}{l^2+m^2})$ so the tangent at $V$ (which is parallel to $lx+my=0$ and goes through $V$) is $lx+my-m=0$ which meets $y+x-1=0$ in $C: (x,y)=(0,1)$.

Now $$AC\cdot BC=\sqrt{\left(\frac{l-m}{l+m}-0\right)^2+\left(\frac{2\,m}{l+m}-1\right)^2}\sqrt{\left(-\frac{l+m}{l-m}-0\right)^2+\left(\frac{2\,l}{ l-m}-1\right)^2}=\sqrt{2\frac{(l-m)^2}{(l+m)^2}}\sqrt{2\frac{(l+m)^2}{(l-m)^2}}=2$$ for $l\pm m\neq 0.$ When $l=m$ the points $A$ and $C$ coincide and $B$ goes off to infinity. When $l=-m$ the parabola degenerates to a double line $(x+y-3)^2=0$; which is not tangent to $x+y-1=0$ in the finite plane; $A$ goes off to infinity with $B$ and $C$ coinciding.

  • Your answer is right, but I fear it’s beyond my scope of understanding. As you can tell, I am still largely unfamiliar with math at my level – Aditya Feb 27 '20 at 13:23
  • @Aditya You've in your edit of the question found one of the infinitely many parabolas in my answer (the one with $m=0$). If you want a complete answer you need to argue for the constancy of the number 2 in the family of parabolas. BTW the reason I've said my answer was partial was that the family of parabolas was found experimenting with geogebra, and a clear way to find it was not provided. If you want I can put more effort into elucidating this point. – Jan-Magnus Økland Jun 15 '20 at 06:41