1
$\begingroup$

Duplicate Disclaimer: At the bottom

This question is mainly about heuristics; I would like to learn how to think when doing these problems, and I am just using $\sqrt{a-\sqrt{a+x}}=x$ as a case study.

I am very interested in becoming better at algebraic manupulation; but many times I get stuck and don't know what to do. There are many many ways to go forward which all look equally good to me but I don't know which to choose, and how far to go on each way.

As an illustration I will try to solve $\sqrt{a-\sqrt{a+x}}=x$

I can easily transform this into $(x^2-a)^2=x+a$. At this point I try different things:

Expand:

This yields $x^4-2ax^2-x+(a^2-a)=0$. Since factors must multiply to the constant term, I try $a$ and $1-a$ but they don't work. At this point I recall that it is possible to factor a quartic by adding and subtracting a clever value, and then factoring by grouping. But I don't know which terms to group and how many terms should be in each group. At this point I start wondering if expanding was bad and I shouldn't have thrown away the nice factored form.

Substitutions:

Back at $(x^2-a)^2=x+a$. There is nothing obvious to do (barring expanding) so I try substitutions:

$1)$ Eliminate the $a$ on the left by $x=y+\sqrt a$ or by $x^2=y^2+a$. This yields $(y^2+2 y\sqrt a)^2=y+a+\sqrt a$ or $y^4=\sqrt{y^2+a}+a$ respectively, which don't seem any nicer.

$2)$ Eleminate the $a$ on the right by $x=y-a$, which doesn't lead to anything nice.

$3)$ Scale: $x=ay$, or $x=\sqrt ay$ which again doesn't help.

$3)$ Translate and scale $x=ay - \sqrt a$ which doesn't lead to anything nicer either.


I could try to go farther on each of these ways, but I never know if I should stop because it's a dead end.

How can I improve my ability to solve algebraic manipulation problems?

Duplicate Disclaimer: This is not a duplicate of this because answers just stating the solution would suffice for that question. Such answers would not suffice for my question, because I am interested in an explanation of the heuristics involved in solving such questions in general.

$\endgroup$
3
$\begingroup$

Sugestion: Try to replace the role of variables. Solve the equation on $a$ first. That aproach helps many times.

From $x^4-2ax^2-x+(a^2-a)=0$ we have:

$$ a^2-a(2x^2+1)+x^4-x=0$$ so we have $$ a_{1,2} = {2x^2+1\pm (2x+1)\over 2}$$

$\endgroup$
  • $\begingroup$ and from where do you know, that this "Solutions" fulfill your equation? any conditions on the parameter? $\endgroup$ – Dr. Sonnhard Graubner Dec 12 '17 at 18:11

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.