I am asked to find the line integral for the following field: $$F = (e^{y^2}-2y)i + (2xye^{y^2}+\sin(y^2))j$$ On the line segment with points $(0,0),(1,2)$ and $(3,0)$. I have to do it with Greens theorem. This is the setup I have so far.

$$\int_C F \cdot dr = \iint_D \frac{\partial Q}{\partial x}-\frac{\partial P}{\partial y}\, dA$$ $$\frac{\partial Q}{\partial x} = ye^{y^2} \qquad \frac{\partial P}{\partial y}=2ye^{y^2}-2$$ Line from $(0,0)$ to $(1,2)$ is $y = 2x$. Line from $(1,2)$ to $(3,0)$ is $y = -x + 3$ $$\int_0^1\int_0^{2x}ye^{y^2}-2ye^{y^2}-2 + \int_1^3\int_0^{-x+3}ye^{y^2}-2ye^{y^2}-2 $$ $$\int_0^1\int_0^{2x}-ye^{y^2}-2 \, dy\,dx+ \int_1^3\int_0^{-x+3}-ye^{y^2}-2 \,dy\,dx$$ $$\int_0^1\left[\frac{e^{y^2}}{2} - 2y\right]_0^{2x} + \int_1^3\left[-\frac{e^{y^2}}{2} - 2y\right]_0^{-x+3}$$ $$\int_0^1\frac{e^{4x^2}}{2}-\frac{1}{2}-4x \,dx+ \int_1^3-\frac{e^{(-x+3)^2}}{2}-2(-x+3)+\frac{1}{2}\,dx$$ Let $u = -x + 3$ and $du = -1$. $$\int_0^1\frac{e^{4x^2}}{2}-\frac{1}{2}-4x \,dx+ \int_1^3\frac{e^{(u)^2}}{2}+2(u)-\frac{1}{2}\,du$$

$$\left[\frac{e^{4x^2}}{8x}-\frac{x}{2}-2x^2 \right]_0^1+ \left[\frac{e^{(u)^2}}{4u}+u^2+\frac{u}{2}\right]_1^3$$ $$\left[\frac{e^{4x^2}}{8x}-\frac{x}{2}-2x^2 \right]_0^1+ \left[\frac{e^{(-x+3)^2}}{4(-x+3)}+(-x+3)^2+\frac{-x+3}{2}\right]_1^3$$ $$\frac{e^{4}}{8}-\frac{1}{2}-2 \,-\frac{e^{4}}{8}-4-1$$ $$\frac{1}{2}-2 \,-5$$

But the answer key says the answer is $-3$. Where have I gone wrong?

  • $\begingroup$ wow this is one really tedious integral. $\endgroup$ – cgo Dec 16 '15 at 9:04
  • $\begingroup$ Try using $dx$, $dy$, $du$ in every integral you define. Also, use align and &= to facilitate for the viewer. $\endgroup$ – Guilherme Thompson Dec 16 '15 at 10:07

The problem lies in two places: the first one in $$\frac{\partial Q}{\partial x} = 2ye^{y^2},$$ but this is only algebraic, even though it made your life much harder.

The second one is much more sensible. Look at Green's Theorem $$\oint_C (P dx + Q dy) = \iint_D \frac{\partial Q}{\partial x}-\frac{\partial P}{\partial y}\,dA,$$ where D is CLOSED domain and $C = \partial D$. Therefore, the integral you are suppose to have is

$$\int_C F \cdot dr = -\left[\iint_{\triangle} \frac{\partial Q}{\partial x}-\frac{\partial P}{\partial y}\, dA + \int_3^0 F(t,0) \cdot (1,0) \, dt\right],$$ where, $\triangle$ is the triangular domain of the question, $((0,0),(1,2),(3,0))$, and the the last term is the integral to bound you domain. Note that we integrate from $3 \to 0$ in order to close the domain, and we need the negative of RHS, because the chosen parametrization, that needs to respect the right-hand rule.

So, let $$G(x,y) = \frac{\partial Q}{\partial x}-\frac{\partial P}{\partial y} = 2$$

We have \begin{align} \int_C F \cdot dr &= -\left[ \int_0^1\int_0^{2x} G(x,y) \, dy\,dx + \int_1^3\int_0^{-x+3}G(x,y) \,dy\,dx + \int_3^0 F(t,0) \cdot (1,0) \, dt,\right]\\ &= -\left[ \int_0^1\int_0^{2x} 2 \, dA + \int_1^3\int_0^{-x+3} 2 \,dA + \int_3^0 1 \, dt,\right]\\ &= -\left[ 2\int_0^1\int_0^{2x} \, dA + 2\int_1^3\int_0^{-x+3} \,dA - \int_0^3 \, dt,\right]\\ &=-\left[2\dfrac{1\times2}{2} + 2\dfrac{2 \times 2}{2} - 3 \right]\\ &=-3 \end{align}


The partial derivative of $Q$ regarding $x$ lacks a factor $2$.

  • $\begingroup$ This is not a real answer. It's more like a comment. $\endgroup$ – Guilherme Thompson Dec 16 '15 at 13:17
  • $\begingroup$ @GuilhermeThompson The question was "Where have I gone wrong?", so it is not answering this and not useful? Hm.. $\endgroup$ – mvw Dec 16 '15 at 13:25
  • $\begingroup$ Somehow yes, but even with the modification you pointed out, his method would still be wrong. As I stated, this was a much smaller mistaken, because it's algebraic... On the other hand, his notion about Green's Theorem was wrong, and this is much more relevant, don't you think, @mvw? $\endgroup$ – Guilherme Thompson Dec 16 '15 at 13:30
  • $\begingroup$ Someone else might identify another issue, would this justify diminishing the given effort so far? $\endgroup$ – mvw Dec 16 '15 at 13:48
  • $\begingroup$ Your answer was very incomplete, such that even adjusting the mistaken you spotted he would still get the wrong answer! That's all. $\endgroup$ – Guilherme Thompson Dec 16 '15 at 14:05

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.