Le 4 mai 2021, la plateforme Yahoo Questions/Réponses fermera. Elle est désormais accessible en mode lecture seule. Aucune modification ne sera apportée aux autres sites ou services Yahoo, ni à votre compte Yahoo. Vous trouverez plus d’informations sur l'arrêt de Yahoo Questions/Réponses et sur le téléchargement de vos données sur cette page d'aide.
Number of divisors?
Hmm, the "similar questions" presented by Yahoo seem to indicate this one might be simple for some of you. Ah well, here goes:
Let n be a positive integer such that 5n has 36 positive divisors and 3n has 40 positive divisors.
How many positive divisors can 15n have?
How many values of n are possible?
[I have a solution which I believe is correct, but it's far from elegant; as such, whoever gets the most elegant solution will get best answer.]
2 réponses
- absirdLv 5il y a 1 décennieRéponse favorite
Spoiler:
I'm getting 15n has either 45 or 48 positive divisors and that n = (3^3)(5^7) or is of the form n = (3^2)(5^4)(p) where p is a prime not equal to 3 or 5. I'll see if I can write up a solution later. The solution I have involves some casework so I suspect there may be a slicker way to solve it.
Addtion: Let's do some casework!
Case 1: 5 does not divide n
Let's first write n in the form n = (p_1)^(e_1) (p_2)^(e_2) ...(p_n)^(e_n) where the p_i's are primes and the e_i's are positive integers. Define d(n) to be the number of positive integer divisors of n. Then d(n) = (e_1 + 1)(e_2 + 1)...(e_n + 1). We can write 5n = 5(p_1)^(e_1)*...*(p_n)^(e_n), so
d(5n) = (1 + 1)(e_1 + 1)(e_2 + 1)...(e_n + 1) = 2d(n) and
d(5n) = 36, so
d(n) = 18.
If 3 does not divide n, then we we could do a similar analysis to find that d(n) = 20, which contradicts the above statement that d(n) = 18.
If 3 does divide n, then we can write 3n = (p_1)^(e_1)*...*(p_i)(e^i + 1)*...*(p_n)^(e_n) where p_i = 3. In this case d(3n) = (e_1 + 1)...(e_i + 2)...(e_n + 1) = 40. Let's compare this with the similar eqn for d(n) above by dividing them:
d(3n)/d(n) = (e_i + 2) / (e_i + 1) <= 3/2 < 40/18 where we determine (e_i +2) / (e_i + 1) <= 3/2 by knowing the rational expression is monotonically decreasing over the positive integers. Here again we reached a contradiction.
We conclude that if there are any successful values of n, then 5|n for those values of n.
Case 2: 5 does divide n.
If 3 does not divide n, then we know from the analysis above that d(n) = d(3n) / 2 = 40/2 = 20. Assuming 5 does divide n we can write 5n = (p_1)^(e_1)*...*(p_j)^(e_j + 1)*...*(p_n)^(e_n) where p_j = 5. It follows that d(5n) = (e_1 + 1)...(e_j + 2)...(e_n + 1) = 36. Let's divide the d(5n) eqn by the d(n) eqn:
d(5n) / d(n) = (e_j + 2) / (e_j + 1) <= 3/2 < 36/20.
Here also we reach an impossibility, so we don't have any solutions when 3 does not divide n. Now we must consider the condition where 3 does divide n (the last possibility).
If there are any n with d(5n) = 36, d(3n) = 40 we must have 3|n and 5|n. Thus, we can write
n = (3^a) (5^b) (p_1)^(e_1)... (p_n)^(e_n) for postive integers a and b. We will work more closely with the following eqns:
d(3n) = (a + 2)(b + 1)(e_1 + 1)...(e_n + 1) = 40.....(1) and
d(5n) = (a + 1)(b + 2)(e_1 + 1)...(e_n + 1) = 36.....(2).
Observe that when we divide these expressions we get (a+2)(b+1) / (a+1)(b+2) = (ab + 2b + a + 2) / (ab + a + 2b + 2) = (s + b) / (s + a) = 40 / 36 > 1 so b > a (where s = ab + a + b + 2).
Let's analyze (1). Observe that 40 = (2^3)(5). We wish to split 40 into pairs, triples, etc. of factors such that each factor is at least 2. We can do this in six ways:
40 = (2)(20) = (4)(10) = (8)(5) = (5)(2)(4) = (2)(2)(10) = (2)(2)(2)(5). Now we must consider the six subcases where d(3n) is split under these conditions.
(i) (a+2)(b+1) = (2)(20). If a+2 = 2, then a is 0 which we can't have. If a+2 = 20, then a > b which we can't have either. I will skip this analysis in the subsequent cases since it's pretty quick. The restrictions are that a, b, and the p_i are positive integers and that a < b.
(ii) (a+2)(b+1) = (4)(10). This gives (a, b) = (2, 9).
(iii) (a+2)(b+1) = (8)(5). This gives (a, b) = (3, 7).
(iv) (a+2)(b+1)(e_1 + 1) = (5)(2)(4). This implies (a, b, e_1) = (2, 4, 1).
(v) (a+2)(b+1)(e_1 + 1) = (2)(2)(10). We don't generate any successful triples here.
(vi) (a+2)(b+1)(e_1 + 1)(e_2 + 1) = (2)(2)(2)(5). Here again we don't generate any successful quadruples.
Our list of possibilities is reduced to (a, b) = (2, 9) or (3, 7) or (a, b, e_1) = (2, 4, 1). Let's check which of these satisfy (2). I'll associate with these pairs (triple) their corresponding subcase above:
(ii) (a, b) = (2, 9). Subsitute into (a+1)(b+2) to get (3)(11) = 33 which does not equal 36.
(iii) (a, b) = (3, 7). Substitute into (a+1)(b+2) to get (4)(9) = 36, so this works.
(iv) (a, b, e_1) = (2, 4, 1). Substitute into (a+1)(b+2)(e_1 + 1) to get (3)(6)(2) = 36 which also works.
We have the following two forms for n:
n = (3^3)(5^7) or n = (3^2)(5^4)(p) where p is a prime not equal to 3 or 5. This answers the second part of the question. Now for the first part:
If n = (3^3)(5^7), then 15n = (3^4)(5^8) which has (4+1)(8+1) = 45 positive integer divisors. If n = (3^2)(5^4)(p), then 15n = (3^3)(5^5)(p) which has (3+1)(5+1)(1+1) = 48 positive integer divisors.
- Mugen is StrongLv 7il y a 1 décennie
for 5n,
36 = 2*2*3*3 divisors
for 3n,
40 = 2*2*2*5 divisors
let n = a^x * b^y * c^z, then n will have :
(x+1)(y+1)(z+1) divisors, where a, b, and c are all primes. and x, y, z ⥠0
i just take 3 variables, because both 36 and 40 have 4 prime factors. assuming the difference in the multiplicators (3 and 5) gave the fourth factors, all the other 3 came from n.
if there aren't any 3(s) or 5(s) in n, then
number of divisors of 3n and 5n = 2(x+1)(y+1)(z+1). but we know this is not the case, so in n, there must exist at least 1 "3" or/and 1 "5".
if either one is a factor but not the other, the number of divisors would be
2(x+1)(y+1)(z+1) against (x+2)(y+1)(z+1), assigning x to be the exponent of repeated prime factor. we have 5 in 40, that is an abnormality(?). so if this is really the case, we'll have to "eliminate" 5, by being one of the retorted (x+1) or (x+2).
x+2 could be 5, 6 or 10, and x+1 be 4, 5 or 9 (respectively).
then (y+1)(z+1) must be 8, 6, or 4 (in that order). we clearly see that this is definitely NOT the case.
this only mean 1 thing; 2 of a, b, c must be 3 and 5. suppose a = 3 and b = 5. now x, y ⥠1.
for 5n,
36 = 2*2*3*3 divisors
= (x+1)(y+2)(z+1)
= (z+1)[(xy+x+y+2) + x]
for 3n,
40 = 2*2*2*5 divisors
= (x+2)(y+1)(z+1)
= (z+1)[(xy+x+y+2) + y]
z+1 is either 1, 2, or 4. but 5 is certainly (x+2) OR (y+1). must be or. that will lead us to x = 3 OR y = 4.
and also, from
36 = (z+1)[(xy+x+y+2) + x] and
40 = (z+1)[(xy+x+y+2) + y],
we get y = x + [4/(z+1)].
the OR rule nullified (z+1) = 4. so z+1 is either 1 or 2.
if x = 3, y = 5 OR 7. <-- (edited : wrote 5 or 9 before)
if y = 4, x = 2 OR 0.
x = 0 is contradictory (we already established of 3 AND 5 being repeated primes). for (x,y,z), we have 3 bawai;
(2,4,1), (3,5,1) or (3,7,0).
from all 3 sets, (x,y,z) = (2,4,1) and (3,7,0) gave the right answers.
n = 3^2 * 5^4 * c^1
= 75^2 * c
= 5625c
or
n = 3^3 * 5^7 * c^0
= 2109375
answer = many (with 1 fixed value), as long as c is prime other than 3 or 5.
number of divisors of 15n
= (3+1)(5+1)(2)
= 4 * 6 * 2
= 48
or (for 2109375)
= (4+1)(8+1)(1)
= 45