1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Help with a wave equation derivation

  1. Jan 7, 2014 #1

    mysearch

    User Avatar
    Gold Member

    Hi,
    Apologises if I have submitted this issue into the wrong Math forum. However, I was wondering if anybody could help me with 2 steps in a derivation of an equation. Simply by way of background, the derivation is linked to formation of a superposition wave subject to a Doppler effect

    [1] [tex]\phi = Ae^{i \left( ct+r \right)k/\gamma \left(1+\beta \right)} - Ae^{i \left( ct-r \right)k \gamma \left(1-\beta \right)} [/tex]

    So based on [1], the derivation states that [2] can be obtain by a process of multiplying exponents, rearranging and factoring.

    [2] [tex]\phi = Ae^{ik \gamma \left( ct + \beta r \right)} \left( e^{ik \gamma \left( \beta ct+r \right)} - e^{-ik \gamma \left( \beta ct+r \right)} \right) [/tex]

    However, so far, I have only managed to create a bit of a mess as I haven’t found a way of rationalising the different expressions in the exponentials. Would really appreciate any help that might confirm that step [2] is valid from [1] and/or any pointers as to how it is done. Thanks
     
  2. jcsd
  3. Jan 7, 2014 #2

    maajdl

    User Avatar
    Gold Member

    Equation [1] was obtained by applying a Lorentz transformation from a 'stationary frame' {x',t') to a 'moving frame' {x,t}.
    If you go back to the 'stationary frame', you will easily see the equivalence of [1] and [2] .
    The equivalence of [1] and [2] in the moving frame is then obvious.

    note1: there is a small typo in [1]
    note2: remember the Lorentz transformation: x' = γ (x + β c t) and c t' = γ (c t + β x)
     
  4. Jan 8, 2014 #3

    mysearch

    User Avatar
    Gold Member

    Many thanks for the helpful reply in #2. Your suggestion about collapsing [1] back to the stationary frame allowed me to spot a mistake I was making with my exponential algebra, it has been awhile! The following equations are primarily for confirmation and for future cross reference. However, I would like to know exactly where you think the ‘small typo’ in [1] lies as it is correct to the source. As suggested, switching back to the stationary frame, i.e. v=0, allows the following substitutions:

    [0] [itex]\beta=\frac{v}{c}=0[/itex]; [itex]\left(1-\beta \right)=1[/itex]; [itex]\gamma=\frac{1}{\sqrt{1-\beta^2}}=1[/itex]; [itex]\gamma \beta=0[/itex]

    This allows [1] to be reduced to:

    [1] [itex]\phi = Ae^{i \left( ct+r \right)k/\gamma \left(1+\beta \right)} - Ae^{i \left( ct-r \right)k \gamma \left(1-\beta \right)} [/itex]

    [1a] [itex]\phi = Ae^{i \left( ct+r \right)k} - Ae^{i \left( ct-r \right)k} [/itex]

    [1b] [itex]\phi = Ae^{i \left( \omega t+kr \right)} - Ae^{i \left( \omega t-kr \right)} [/itex]

    The second version is just to highlight the normal format of a standard wave equation within the exponential, which I believe can be transposed back into a trigonometric form using the following identity:

    [TI-1] [itex]e^{i \theta} = cos \theta + i sin \theta [/itex]

    However, using only the ‘real’ part allows [1b] to be written as:

    [1c] [itex]\phi = A cos \left( \omega t+kr \right) - A cos \left( \omega t-kr \right) [/itex]

    I don’t know where you think the typo was in [1], as the solution in [1c] looks OK to me, except I don’t understand why the two wave amplitudes are subtracted as this is a description of superposition wave. While I am still investigating this issue, I suspect it is because the following identity [TI-2] is used later in the derivation - see [2]/[3] below, although the substitution looks incorrect to me:

    [TI-2] [itex]sin(\theta)= \frac{e^{i\theta} – e^{-i\theta}}{2i} \Longrightarrow e^{i\theta} – e^{-i\theta}=(2i)sin(\theta)[/itex]

    [TI-3] [itex]cos(\theta)= \frac{e^{i\theta} + e^{-i\theta}}{2} \Longrightarrow e^{i\theta} + e^{-i\theta}=(2)cos(\theta)[/itex]

    Anyway, converting [2] to a stationary form using the substitutions in [0] above gives:

    [2a] [itex]\phi = Ae^{ik*1 \left( ct + 0*r \right)} \left( e^{ik*1 \left( 0+r \right)} - e^{-ik*1 \left( 0+r \right)} \right) \Longrightarrow Ae^{i \left( \omega t \right)} \left( e^{i \left(kr \right)} - e^{-i \left(kr \right)} \right) [/itex]

    [2b] [itex]\phi = Ae^{i \left( \omega t \right)}e^{i \left(kr \right)} -Ae^{i \left( \omega t \right)}
    e^{-i \left(kr \right)} \Longrightarrow Ae^{\left( i \omega t +ikr\right)} - Ae^{\left( i \omega t - ikr\right)} [/itex]

    [2c] [itex]\phi = Ae^{i \left( \omega t + kr\right)} - Ae^{i \left( \omega t - kr\right)} [/itex]

    So, as you indicated, the equivalence of [1b] and [2c] appears to be confirmed. However, I still don’t really understand why the derivation proceeds from [2] as follows, which appears to be dependent on [TR-2], but then seems to lose the complex value in [3]:

    [2] [itex]\phi = Ae^{ik \gamma \left( ct + \beta r \right)} \left( e^{ik \gamma \left( \beta ct+r \right)} - e^{-ik \gamma \left( \beta ct+r \right)} \right) [/itex]

    [3] [itex]\phi = 2Ae^{ik \gamma \left( ct + \beta r \right)} sin \left[ k \gamma \left( \beta ct + r \right) \right] [/itex]

    However, if [2] was changed to reflect a superposition of 2 additive waves it would presumably become:

    [2’] [itex]\phi = Ae^{ik \gamma \left( ct + \beta r \right)} \left( e^{ik \gamma \left( \beta ct+r \right)} + e^{-ik \gamma \left( \beta ct+r \right)} \right) [/itex]

    If so, the [TR-3] could then be used such that [3] would become:

    [3’] [itex]\phi = 2Ae^{ik \gamma \left( ct + \beta r \right)} cos \left[ k \gamma \left( \beta ct + r \right) \right] [/itex]

    Finally, using [TR-1], I would have thought that the remaining complex exponential could also be replaced by taking only the ‘rea partl’, such that [3’] would become:

    [4] [itex]\phi = 2A cos \left[ k \gamma \left( ct + \beta r \right) \right] cos \left[ k \gamma \left( \beta ct +r \right) \right] \Longrightarrow 2A cos \left[ \gamma \left( \omega t + \beta k r \right) \right] cos \left[ \gamma \left( \beta \omega t + kr \right) \right][/itex]

    Anyway, as stated, much of this is purely for future reference, but would appreciate the correction of the 'maths' on my part. I still need to take a closer look at the Lorentz transformation in this wave process and while I already have some issues to resolve within this derivation, it is presumably best if I raise these in another PF forum, e.g. general physics? Again, I really appreciate the help. Thanks
     
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook




Similar Discussions: Help with a wave equation derivation
Loading...