B3 - Quadratic equations in unfamiliar forms

B3 - Quadratic equations in unfamiliar forms

Choose a quadratic-solving method, recognise quadratic structure inside a less familiar equation and return to the original variable. Keep exact answers and use domain restrictions to decide which candidate roots really solve the original problem.

Choosing a quadratic method

Can an equation with a fourth power have four real solutions, even though a quadratic has at most two? The equation x4−5x2+4=0x^4-5x^2+4=0 will give us all four. Its secret is that it is quadratic in a different expression.

First recall the smaller problem. In 2x2+x=62x^2+x=6, move every term to one side before solving. It becomes 2x2+x−6=02x^2+x-6=0, so a product equal to zero can reveal the roots.

You already know

A quadratic is ax2+bx+cax^2+bx+c, with a≠0a\ne0. Factorisation, completing the square and the quadratic formula are alternative ways to solve its equation. The discriminant tells you whether its roots are real.

Factorisation is efficient when simple factors are visible. The zero-product rule says that a product of real numbers is zero only if at least one factor is zero. It is not a rule for a product equal to a nonzero number. This rule is not in the formula booklet: learn it.

Quadratic formula

For ax2+bx+c=0ax^2+bx+c=0, with real coefficients and a≠0a\ne0,

x=−b±b2−4ac2ax=\frac{-b\pm\sqrt{b^2-4ac}}{2a}

Real roots require b2−4ac≥0b^2-4ac\ge0.

Not in the formula booklet: learn it.

The formula is useful when rational factors are not available. Keep brackets around negative coefficients and around the whole denominator.

Worked example 1: solve by factorisation

Question. Solve 2x2+x=62x^2+x=6.

Plan. Rearrange to zero, factorise, then set each factor equal to zero.

2x2+x−6=0rearrange(2x−3)(x+2)=0factorise2x−3=0orx+2=0zero productx=32orx=−2solve each factor\begin{aligned} 2x^2+x-6&=0 && \text{rearrange} \\ (2x-3)(x+2)&=0 && \text{factorise} \\ 2x-3=0&\quad\text{or}\quad x+2=0 && \text{zero product} \\ x&=\frac32\quad\text{or}\quad x=-2 && \text{solve each factor} \end{aligned}

Why this step? Rearranging to zero is what makes the zero-product rule applicable.

Answer. x=3/2x=3/2 or x=−2x=-2.

Check. In the original equation, 2(3/2)2+3/2=62(3/2)^2+3/2=6 and 2(−2)2−2=62(-2)^2-2=6.

Change just the right side from 66 to 55. The convenient rational factorisation disappears, so use the formula instead.

Worked example 2: exact roots from the formula

Question. Solve 2x2+x=52x^2+x=5, giving exact roots.

Plan. Rearrange to zero and substitute a=2,b=1,c=−5a=2,b=1,c=-5 into the quadratic formula.

2x2+x−5=0standard formx=−1±12−4(2)(−5)2(2)substitute coefficients=−1±414simplify the discriminant\begin{aligned} 2x^2+x-5&=0 && \text{standard form} \\ x&=\frac{-1\pm\sqrt{1^2-4(2)(-5)}}{2(2)} && \text{substitute coefficients} \\ &=\frac{-1\pm\sqrt{41}}4 && \text{simplify the discriminant} \end{aligned}

Why this step? The irrational square root belongs in an exact answer; rounding it would change the roots.

Answer. x=(−1+41)/4x=(-1+\sqrt{41})/4 or x=(−1−41)/4x=(-1-\sqrt{41})/4.

Check. If x=(−1+s)/4x=(-1+s)/4, where s=±41s=\pm\sqrt{41}, then 2x2+x=(s2−1)/8=52x^2+x=(s^2-1)/8=5.

AQA permits calculator solutions for an ordinary numerical quadratic, without a compulsory written formula substitution. Exact form, a prescribed method, or a request to justify a result still has to be honoured. Displaying algebra is useful because an incorrect answer can still contain creditworthy method.

Watch out

Wrong: from (2x−3)(x+2)=6(2x-3)(x+2)=6, set each factor equal to zero. Right: first obtain a product equal to zero. The product in example 1 equals zero only after moving the 66.

Change the constant again and choose an efficient method yourself.

Completing the square

A completed square solves an equation by reducing it to one square equal to a constant.

The expansion (x−2)2=x2−4x+4(x-2)^2=x^2-4x+4 tells us that x2−4x=(x−2)2−4x^2-4x=(x-2)^2-4. If the first terms have a common coefficient, retain it outside the whole bracket so the compensating constant is scaled correctly.

Solve from a square

For real h,kh,k,

(x−h)2=k(x-h)^2=k

If k>0k>0, then x=h±kx=h\pm\sqrt k. If k=0k=0, then x=hx=h. If k<0k<0, there is no real solution.

Not in the formula booklet: learn it.

The symbol k\sqrt k is non-negative, but an equation asking which numbers square to a positive kk has both signs. This is the same repeated-root boundary seen on quadratic graphs.

Worked example 3: solve using a positive square

Question. Solve 2x2−8x+3=02x^2-8x+3=0 by completing the square, giving exact roots.

Plan. Form (x−2)2(x-2)^2, isolate it, then take both square-root signs.

2x2−8x+3=2(x2−4x)+3factor 2=2((x−2)2−4)+3complete the square=2(x−2)2−5simplify2(x−2)2−5=0⇒(x−2)2=52isolate the squarex=2±102both signs\begin{aligned} 2x^2-8x+3&=2(x^2-4x)+3 && \text{factor }2 \\ &=2\bigl((x-2)^2-4\bigr)+3 && \text{complete the square} \\ &=2(x-2)^2-5 && \text{simplify} \\ 2(x-2)^2-5=0&\Rightarrow (x-2)^2=\frac52 && \text{isolate the square} \\ x&=2\pm\frac{\sqrt{10}}2 && \text{both signs} \end{aligned}

Why this step? The −4-4 inside the bracket is multiplied by 22; overlooking this changes both roots.

Answer. x=2+10/2x=2+\sqrt{10}/2 or x=2−10/2x=2-\sqrt{10}/2.

Check. For either root, (x−2)2=10/4=5/2(x-2)^2=10/4=5/2, so 2(x−2)2−5=02(x-2)^2-5=0.

Change only the constant from 33 to 88. The square will now equal zero, so the two signs coincide.

Worked example 4: a repeated root from the square

Question. Solve 2x2−8x+8=02x^2-8x+8=0 by completing the square.

Plan. Complete the same square and find where it vanishes.

2x2−8x+8=2((x−2)2−4)+8complete the square=2(x−2)2constants cancel2(x−2)2=0⇒x−2=0square equals zerox=2solve\begin{aligned} 2x^2-8x+8&=2\bigl((x-2)^2-4\bigr)+8 && \text{complete the square} \\ &=2(x-2)^2 && \text{constants cancel} \\ 2(x-2)^2=0&\Rightarrow x-2=0 && \text{square equals zero} \\ x&=2 && \text{solve} \end{aligned}

Why this step? There is only one distinct number whose square is zero, so listing 22 twice adds no solution.

Answer. x=2x=2, a repeated root.

Check. 2(2)2−8(2)+8=02(2)^2-8(2)+8=0.

Completed-square form and the formula must agree: both are rearrangements of the same equation. Completing the square is particularly revealing when the right side is zero or negative.

Watch out

Wrong: (x−2)2=5/2(x-2)^2=5/2 gives only x=2+10/2x=2+\sqrt{10}/2. Right: both signs are needed because both positive and negative bracket values have that square.

Move the constant past the repeated-root case and use the square to explain the result.

A quadratic in another expression

A substitution names a repeated expression, turning an unfamiliar equation into a familiar quadratic.

If u=x+1u=x+1, then (x+1)2=u2(x+1)^2=u^2. Both occurrences must change together. After finding uu, return to xx: the question has not asked for the new variable.

Quadratic in a function of the unknown

An equation of the form af(x)2+bf(x)+c=0af(x)^2+bf(x)+c=0, with a≠0a\ne0, is quadratic in f(x)f(x). The substitution u=f(x)u=f(x) gives au2+bu+c=0au^2+bu+c=0. For example, x4−5x2+4=0x^4-5x^2+4=0 is quadratic in x2x^2, because x4=(x2)2x^4=(x^2)^2.

The substitution's range matters. A bracket such as x+1x+1 can take any real value, whereas x2≥0x^2\ge0. This substitution process is not in the formula booklet: learn to recognise the repeated expression.

Worked example 5: a repeated bracket

Question. Solve (x+1)2−5(x+1)+4=0(x+1)^2-5(x+1)+4=0.

Plan. Put u=x+1u=x+1, solve the quadratic, then undo the substitution.

u2−5u+4=0u=x+1(u−1)(u−4)=0factoriseu=1 or 4zero productx+1=1 or 4return to xx=0 or 3subtract 1\begin{aligned} u^2-5u+4&=0 && u=x+1 \\ (u-1)(u-4)&=0 && \text{factorise} \\ u&=1\text{ or }4 && \text{zero product} \\ x+1&=1\text{ or }4 && \text{return to }x \\ x&=0\text{ or }3 && \text{subtract }1 \end{aligned}

Why this step? Keeping the bracket intact avoids expanding away the structure you can already solve.

Answer. x=0x=0 or x=3x=3.

Check. The original bracket is respectively 11 and 44; 1−5+4=01-5+4=0 and 16−20+4=016-20+4=0.

Now change the repeated expression from a linear bracket to x2x^2. Undoing this substitution can produce two original-variable roots from each positive substituted root.

Worked example 6: a fourth power hiding a quadratic

Question. Find all real solutions of x4−5x2+4=0x^4-5x^2+4=0.

Plan. Put u=x2≥0u=x^2\ge0, solve for uu, then take both signs for each positive square.

u2−5u+4=0u=x2(u−1)(u−4)=0factoriseu=1 or 4both non-negativex2=1 or 4undo the substitutionx=±1 or ±2both signs each time\begin{aligned} u^2-5u+4&=0 && u=x^2 \\ (u-1)(u-4)&=0 && \text{factorise} \\ u&=1\text{ or }4 && \text{both non-negative} \\ x^2&=1\text{ or }4 && \text{undo the substitution} \\ x&=\pm1\text{ or }\pm2 && \text{both signs each time} \end{aligned}

Why this step? Two roots in uu need not mean two roots in xx: the substitution u=x2u=x^2 combines opposite values of xx.

Answer. x=−2,−1,1,2x=-2,-1,1,2.

Check. The original expression is zero at x2=1x^2=1 and x2=4x^2=4, so each positive and negative pair works.

This answers the opening puzzle. There is another small surprise: a quadratic in x2x^2 can have three distinct real solutions. For x4−x2=0x^4-x^2=0, the substituted roots are 0,10,1; the first gives only x=0x=0, while the second gives x=±1x=\pm1. Count original values, not signs written on paper.

Watch out

Wrong: “u=1,4u=1,4, so the solutions are x=1,4x=1,4.” Right: solve x2=1x^2=1 and x2=4x^2=4. Always write the equation that undoes your substitution.

Later, the same structure will appear with trigonometric and exponential expressions; their ranges will decide which substituted roots survive. Here the square substitution is enough.

Domains and admissible roots

An algebraic candidate becomes a solution only when it satisfies the original equation and its domain.

For u=xu=\sqrt x, record x≥0x\ge0 and u≥0u\ge0. A negative uu cannot be a principal square root, even if it solves the transformed quadratic. For u=1/xu=1/x, record x≠0x\ne0 and u≠0u\ne0. These domain facts are not in the formula booklet: apply the definitions.

Worked example 7: reject a negative square root

Question. Solve x−x−6=0x-\sqrt x-6=0 over the real numbers.

Plan. Put u=x≥0u=\sqrt x\ge0, use x=u2x=u^2, and reject any negative value of uu.

u2−u−6=0u=x≥0(u−3)(u+2)=0factoriseu=3 or −2candidate valuesu=3reject −2<0x=32=9recover x\begin{aligned} u^2-u-6&=0 && u=\sqrt x\ge0 \\ (u-3)(u+2)&=0 && \text{factorise} \\ u&=3\text{ or }-2 && \text{candidate values} \\ u&=3 && \text{reject }-2<0 \\ x&=3^2=9 && \text{recover }x \end{aligned}

Why this step? Squaring u=−2u=-2 would produce x=4x=4, but 4=2\sqrt4=2, not −2-2; it would not undo the substitution.

Answer. x=9x=9.

Check. 9−9−6=9−3−6=09-\sqrt9-6=9-3-6=0; the rejected candidate x=4x=4 gives 4−2−6=−44-2-6=-4.

Change just the sign before the square root. The admissible substituted root changes too; the domain rule does not.

Worked example 8: the same domain with a changed sign

Question. Solve x+x−6=0x+\sqrt x-6=0 over the real numbers.

Plan. Use u=x≥0u=\sqrt x\ge0, solve the new quadratic and return to xx.

u2+u−6=0u=x≥0(u+3)(u−2)=0factoriseu=−3 or 2candidate valuesu=2reject −3<0x=22=4recover x\begin{aligned} u^2+u-6&=0 && u=\sqrt x\ge0 \\ (u+3)(u-2)&=0 && \text{factorise} \\ u&=-3\text{ or }2 && \text{candidate values} \\ u&=2 && \text{reject }-3<0 \\ x&=2^2=4 && \text{recover }x \end{aligned}

Why this step? The sign condition applies to the defined variable uu, independently of the signs in its quadratic equation.

Answer. x=4x=4.

Check. 4+2−6=04+2-6=0. Squaring the rejected −3-3 would give x=9x=9, which instead gives 9+3−6=69+3-6=6.

For a reciprocal example, 2/x2+1/x−1=02/x^2+1/x-1=0 with x≠0x\ne0 becomes 2u2+u−1=02u^2+u-1=0, where u=1/x≠0u=1/x\ne0. The factors (2u−1)(u+1)=0(2u-1)(u+1)=0 give u=1/2,−1u=1/2,-1, so x=2,−1x=2,-1. Both satisfy the nonzero condition. Multiplying by x2x^2 is an equally valid route on that domain.

Squaring needs an additional check because it can lose sign information. From x+2=x\sqrt{x+2}=x, squaring gives x2−x−2=0x^2-x-2=0, with candidates 2,−12,-1. The original right side must be non-negative; −1-1 fails, while 22 works. Squaring preserved a true solution but also introduced another candidate.

Watch out

Wrong: solve the transformed quadratic and accept every root. Right: apply the substitution's range, undo it, then check the original equation. Multiplication by a known nonzero quantity is reversible on its domain; squaring does not preserve sign information.

In the next question only the constant changes. Keep the non-negative square-root condition when solving the new quadratic.

Quick Recap

  • Rearrange a quadratic to zero before using the zero-product rule; choose factorisation, completing the square or the formula efficiently.
  • A positive square gives both signs; a zero square gives one distinct root; a negative square gives no real root.
  • Name the repeated expression, record its possible values and undo the substitution after solving.
  • Check admissibility and the original equation, especially after squaring or clearing denominators.

The opening fourth-power equation has four real solutions because each positive root of its quadratic in x2x^2 gives two opposite values of xx: −2,−1,1,2-2,-1,1,2.