2

So I have the equations

  1. $$m_1 v_1 + m_2 v_2 = m_1 v_1^\prime + m_2 v_2^\prime \tag{1}$$
  2. $$\frac{m_1v_1^2}{2} + \frac{m_2v_2^2}{2} = \frac{m_1{v_1^2}^\prime}{2} + \frac{m_2{v_2^2}^\prime}{2} \tag{2}$$

and need to find $v_1^\prime$ and $v_2^\prime$.

I first I solved for $v_1^\prime$ in eq. 1, which is

$$v_1^\prime = \frac{m_2v_2-m_2v_2^\prime+v_1m_1}{m_1} \tag{$1^*$}$$

Then I solved eq. 2 for $v_2^\prime$, which is

$$v_2^\prime = \sqrt{\frac{m_1v_1^2-m_1{v_1^2}^\prime+v_2^2m_2}{m_2}} \tag{$2^*$}$$

Then I plugged $v_2^\prime$ into $v_1^\prime$, to solve for $v_1^\prime$ without the $v_2^\prime$.

And I got stuck at this point:

$$v_1^\prime = \frac{m_2v_2+v_1m_1-\sqrt{m_1m_2v_1^2-m_1m_2{v_1^2}^\prime+v_2^2m_2^2}}{m_1} \tag{3}$$

Now, I know both solutions from a CAS program:

$${v_{1_1}}^\prime = v_1$$ $${v_{1_2}}^\prime = \frac{m_1v_1-m_2v_1+2m_2v_2}{m_1+m_2}$$

But, when I square both sides of the equation, I only get the first solution, so there must be a $\pm$ somewhere after squaring (that's what I think). And the $\pm$ should be infront of the square root term after squaring, but it does not give the right answer. At least I think you should square, because how would you isolate $v_1^\prime$ then?

If one were to put the $\pm$ infront of the square root after squaring, you would get:

$${v_1^2}^\prime = \frac{m_2^2v_2^2+v_1^2m_1^2\mp(m_1m_2v_1^2-m_1m_2{v_1^2}^\prime+v_2^2m_2^2)}{m_1^2} \tag{4}$$

For the first the case where the square is negative, you get the first right answer, but for the second (I did not do this manually), you get

$$v_{1_{2,3}}^\prime = \pm\sqrt{\frac{(m_1^2 v_1^2 + m_1 m_2 v_1^2 + 2 m_2^2 v_2^2)}{(m_1^2 + m_1 m_2)}}, \tag{5}$$

which would imply $m_1m_2 = m_2^2$, which can not be the case, because they are independent of each other. There would also be 3 answers, which is also wrong.

After that it should be easy to plug in the $v_1^\prime$ solution into $v_2^\prime$ from the second equation.

So, what am I doing wrong here?

Also, is it just me, or does eq. 3 look awfully similar to the quadratic equation?

  • Once you have $(1^*)$ just plug that $v_1'$ into the second equation and you get a quadratic in $v_2'$ alone. – dxiv Jan 04 '22 at 00:54
  • In (2), Do you mean $v_1' ^2 $ or $v_1^2 \prime $? (The first is square of $ v_1'$, the second is another value for $v_1^2$.) – Calvin Lin Jan 04 '22 at 01:29
  • To solve this, Hint: Show that $v_1 + v_1' = v_2 + v_2' $, and hence solving the linear equation gives $ v_2 ' = ( 2m_1 v_1 - m_1v_2 + m_2v_2) / (m_1 + m_2)$. – Calvin Lin Jan 04 '22 at 01:34
  • The sum of moments (kinetic energies) of one pair of objects is equal to the sum of moments (resp. kinetic energies) of another pair of objects. Why do you think there are only two solutions? Looks like there are infinitely many (see my answer). – markvs Jan 04 '22 at 02:46
  • @markvs "Looks like there are infinitely many" $;-;$ It is not clear what part this refers to. The system has in fact two solutions, as your answer also shows. – dxiv Jan 04 '22 at 03:43
  • @dxiv: In my answer we have a system of two linear equations with 4 variables. Plus there is a solution where $v_1=v_1'$. What two solutions? Sorry did not notice that only $v_1'. v_2'$ are unknowns. Yes only two solutions. – markvs Jan 04 '22 at 03:46
  • @CalvinLin It should actually be ${v_1^\prime}^2$. – scientist man Jan 04 '22 at 10:55

1 Answers1

2

Assume $v_1\ne v_1'. m_1,m_2\ne 0$. From 1, $$m_1(v_1-v_1')=m_2(v_2'-v_2). \qquad (1)$$ From 2, $$m_1(v_1-v_1')(v_1+v_1')=m_2(v_2'-v_2)(v_2'+v_2). \qquad (2)$$

Hence $$v_1+v_1'=v_2+v_2'. \qquad (3)$$

Adding $m_1\cdot (3)$ to $(1)$, we get

$$2m_1v_1=(m_2+m_1)v_2'-(m_2-m_1)v_2 \qquad (4) $$

Together (3), (4), gives a general solution.

Edit. Since by the OP only $v_1', v_2'$ are unknowns the system (3)+(4) has only one solution. Plus the solution $v_1'=v_1, v_2'=v_2$, we get two solutions.

markvs
  • 19,653
  • Arguably, OP is asking about what is wrong with his approach, as opposed to how to solve this. (That's why I left it as a comment, not as an answer.) – Calvin Lin Jan 04 '22 at 01:37
  • OP's approach involves $\sqrt{.......}$ and complicated expressions. That is what is certainly wrong with it. Perhaps there is something else, I did not check. My answer would not fit as a comment. So I cannot convert it to a comment even if wanted to (which I don't). – markvs Jan 04 '22 at 01:40
  • @markvs Well, think up to the point i got, it is right, because the CAS system also came up with the right answer when I plug that into it. I Just want to find out, what it does to resolve the square root and isolate $v_1^\prime$. And yes, maybe brute forcing this isn't the right way to do it, but, since I want to learn something from this, I would like to know where my error lies (maybe it lies somewhere else, but by process of elimination, it should be here). Also, I guess it would be nice to know how you came up with this solution, since I don't think it's just pattern recognition, is it? – scientist man Jan 04 '22 at 12:51
  • In your ($2^*$), it should be $\pm$ unless you know that the LHS is positive. – markvs Jan 04 '22 at 16:29