Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Parameterized curve vs. Integral(Please look at my result)

  1. Feb 10, 2009 #1
    1. The problem statement, all variables and given/known data

    Hello just found this forum and have question if I have done my homework correctly

    Anyway let [tex]\theta: I \to \mathbb{R}^3[/tex] be a parameterized curve.
    Next assume that [tex][a,b] \subset I [/tex] and put [tex]\theta(a) = l[/tex](1) and [tex]\theta(b) = m[/tex]

    Then show for any constant vector [tex]\vec{u}[/tex],[tex]|\vec{u}| = 1[/tex].

    that the inequality

    [tex](m-l) \cdot \vec{u} = \int_{a}^{b} \vec{\theta(t)} \cdot \vec{u(t)} \ dt \leq \int_{a}^{b} |\vec{\theta'(t)}| dt[/tex](1)

    3. The attempt at a solution

    If I look at the dot-product of [tex]\frac{d}{dt}(\vec{\theta} \cdot \vec{u}}) = \vec{\theta(t)^{\prime}} \cdot \vec{u(t)}[/tex]

    Next by Fundamental theorem of Calculus leftside of (1) is true.

    The inequality however is a bit tricky for me. I seem to remember an inequality
    which states that the dot product of two vectors a and b can be written as
    [tex]\vec{a} \cdot \vec{b} \leq |\vec{a}| \cdot |\vec{b}|[/tex] which transfered to Our situation gives
    [tex]\vec{\theta'(t)} \cdot \vec{u(t)} \leq |\vec{\theta'(t)}| \cdot |\vec{u(t)}|[/tex] and since [tex]|\vec{u}| = 1[/tex], then this is equivalent to [tex]\theta'(t) \cdot \vec{u(t)} \leq |\theta'(t)| [/tex] and finally by the rules regarding our Riemann Integrals [tex]\int_{a}^{b} f \leq \int_{a}^{b} g[/tex], then the integrals [tex]\int_{a}^{b} \ldots dt \leq \int_{a}^{b} \ldots \ dt[/tex] is also true.

    There is a second part as well

    If I set [tex]\vec{u} = \frac{l-m}{|l-m|}[/tex]

    Then show that [tex]|\theta(b) - \theta(a)| \leq \int_{a}^{b} |\vec{\theta(t)^{\prime}| dt[/tex]

    If I insert [tex]\vec{u}[/tex] into lefthand side of the inequality I get [tex](l-m) \cdot \frac{l-m}{|l-m|} = |l-m|[/tex] since [tex](l-m) \cdot (l-m) = |l-m|^2[/tex]

    From this it should follow that the length of the curve from l to m is a straight line joining these points. But how exactly does it do that?
    Is it something to do with if I on my curve draw a straight line from l to m and use this line as a side in a triangle and by using pythagoras then claim that line previously mentioned always will be shorter than the arc length from l to m?

    Cheers Superboy
    Last edited: Feb 10, 2009
  2. jcsd
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook

Can you offer guidance or do you also need help?