Tricky limit

jakeisthesnake

New member
Joined
Mar 25, 2011
Messages
8
Hello Folks!

I am working on my asking-for-help technique, so let me know if the following problem isn't clear.

Any ideas for how to take the limit as t goes to 0?

gif.latex

gif.latex


Thanks! If you want to know why I am trying to solve this, see my longer post.
Jake
 
progress

So I made some progress! The denominator simplifies to:

attachment.php


I isolated all of the terms with sin(t), so now I am trying to solve lim as t -> 0:

attachment.php


any tips for here?

Thanks as always,
Jake
 

Attachments

  • CodeCogsEqn.gif
    CodeCogsEqn.gif
    2.9 KB · Views: 11
  • gif.jpg
    gif.jpg
    25.9 KB · Views: 11
I don't think there's anything that can be done to make it spectacularly easy. You just have to simplify every expression you can, and I think at some point you'll need to use L'Hopital's Rule. You say you're interested in the limit as t approaches 0, but make no mention of any of the other variables approaching a value, so then they're all treated as constants. Furthermore, since it's just four (admittedly very complex) terms being subtracted, I'd be inclined to tackle each one separately. Consider the numerator of the first fraction:

R(cos(θ)cos(t)sin(t)sin(θ))+R(θ+t)(sin(θ)cos(t)+sin(t)cos(θ))A(sin(θ)cos(t)+sin(t)cos(θ))\displaystyle R\left(cos\left(\theta \right)cos\left(t\right)-sin\left(t\right)sin\left(\theta \right)\right)+R\left(\theta +t\right)\left(sin\left(\theta \right)cos\left(t\right)+sin\left(t\right)cos\left(\theta \right)\right)-A\left(sin\left(\theta \right)cos\left(t\right)+sin\left(t\right)cos\left(\theta \right)\right)

After applying angle addition identities, that mess simplifies to:

R(cos(θ+t))+R(θ+t)(sin(θ+t))A(sin(θ+t))\displaystyle R\left(cos\left(\theta +t\right)\right)+R\left(\theta +t\right)\left(sin\left(\theta +t\right)\right)-A\left(sin\left(\theta +t\right)\right)

Then play with the denominator:

sin(ϕθ)cos(ϕθ)sin(ϕθt)cos(ϕθt)=tan(ϕθ)tan(ϕθt)\displaystyle \frac{sin\left(\phi -\theta \right)}{cos\left(\phi -\theta \right)}-\frac{sin\left(\phi -\theta -t\right)}{cos\left(\phi -\theta -t\right)}=tan\left(\phi -\theta \right)-tan\left(\phi -\theta -t\right)

Apply the following identity:

tan(a)tan(b)=sec(a)sec(b)sin(ab)\displaystyle tan\left(a\right)-tan\left(b\right)=sec\left(a\right)\cdot sec\left(b\right)\cdot sin\left(a-b\right)

And you can simplify the denominator down to:

sec(ϕθ)sec(ϕθt)sin(t)\displaystyle sec\left(\phi -\theta \right)\cdot sec\left(\phi -\theta -t\right)\cdot sin\left(t\right)

Now, use L'Hopital's Rule on the fraction, and see what you get. Then tackle the other three fractions in a similar manner. I'd make a special note that the numerator of the fourth fraction has no t terms anywhere, so that makes things even easier. Good luck.
 
Use ksdhart's simplification and simplify a bit further:
Numerator:
N = cos(ϕθ)cos(ϕθt)[Rcos(θ+t)+(R(θ+t)A)sin(θ+t)]\displaystyle cos\left(\phi -\theta \right)\, cos\left(\phi -\theta -t\right)\, [R\, cos\left(\theta +t\right)\, +\, (\, R\left(\theta +t\right)\, -\, A)\, sin\left(\theta +t\right)]
Denominator:
D = sin(t)\displaystyle sin\left(t\right)

Now take the individual limits as t goes to zero:
N cos2(ϕθ)[Rcos(θ)+(RθA)sin(θ)]\displaystyle \to\, cos^2\left(\phi -\theta \right)\, [R\, cos\left(\theta\right)\, +\, \left(R\, \theta\, -\, A\right)\, sin\left(\theta\right)]

D \displaystyle \to 0.

So, unless there is something in the relationships between the other items [R, A, θ\displaystyle \theta and ϕ\displaystyle \phi] the limit is infinity.
 
Thank you ksdhart and Ishuda!

ksdhart - Yeah, it looks like I will have to do some l'hopitaling. I hadn't considered that in the angle addition form so that was a great idea.

Ishuda - I think that I will have to take the other terms of the fraction into consideration because the final answer should be some non-zero, non-infinite function of theta. There is a chance that the steps that lead to taking this limit have a error, but I have checked several times, so I don't think so. (there is a link to my original post that describes the larger problem that I am working on.)

I'll post back here with any more progress (hopefully this weekend.)

Thanks again,
Jake
 
Top