Can we prove these bounds for this fraction involving surds?
Question
The variables x and y are connected by the relation y=axn, where a and n are constants; y=3 when x=4 and y=2 when x=9. Find the exact values of n and a.
Express a−ba+b as a fraction having a rational denominator. Show that, if n is a positive integer, then n(n+2) lies between n2 and (n+1)2. If n is a positive integer, use the results in (i) and (ii) to find in terms of n two consecutive integers between which n+2−nn+2+n must lie.
Suggestion
The question asks us to get rid of the square roots in the denominator; what method might we use?
Can we multiply a−b by something so that we lose the square roots?
What is n(n+2)−(n+1)2?
The question tells us to use results we’ve already proved… like many exam questions!
Solution
By substituting the information we’re given in the question, we obtain the equations
3=4na
and
2=9na.
Dividing the second by the first to eliminate a gives 32=(49)n.
To work with this, it will be more convenient to write everything as powers of 2 and 3.
That is, 22n+1=32n+1, which can only be the case when 2n+1=0 (as 2 and 3 have highest common factor 1).
Therefore, n=−21.
Now we can go back to equation (1) or equation (2) to find that a=6.
Alternatively, if you know about logarithms…
Taking logs to base 10 with our starting equations (any base will do, in fact) we get log3=loga+2nlog2,log2=loga+2nlog3.
Subtracting these gives log3−log2=2n(log2−log3), and so n=−21.
Substituting back into one of our starting equations gives a=6.
If we multiply both the numerator and the denominator by a+b, then the fraction becomes (a−b)(a+b)(a+b)(a+b)=a−b(a+b)2=a−ba+b+2ab. This has a rational denominator.
Note n(n+2)−n2=2n>0, so n(n+2)>n2.
We also have n(n+2)−(n+1)2=n2+2n−n2−2n−1=−1<0, so n(n+2)<(n+1)2.
So n(n+2) lies between n2 and (n+1)2.
If n is a positive integer, use the results in (i) and (ii) to find in terms of n two consecutive integers between which n+2−nn+2+n must lie.
By using (i) we get n+2−nn+2+n=22n+2+2n(n+2)=n+1+(n+2)n.
Now we know (n+1)2>n(n+2)>n2, which implies n+1>n(n+2)>n.
So adding n+1 to each part of the inequality, we get 2n+2>n+1+(n+2)n>2n+1,
and so n+2−nn+2+n lies between 2n+1 and 2n+2.