[math-fun] strange factoring in primes
hihi, all - rich asked about consecutive primes p, q, r with p*q*r = n^3 + 1 for some integer n - i thought i'd take a crack at it without requiring consecutive, and found prety easily 211 * 227 * 241 = 226 ^ 3 + 1 = 11543177, which leadsd me to think that rich's expectation (that only 7*11*13 = 1001 works) is correct BTW, it is easy to show that the middle prime is n+1 more soon, cal
On 8/1/06, Chris Landauer <cal@rush.aero.org> wrote:
hihi, all -
rich asked about consecutive primes p, q, r with
p*q*r = n^3 + 1
for some integer n - i thought i'd take a crack at it without requiring consecutive, and found prety easily
211 * 227 * 241 = 226 ^ 3 + 1 = 11543177,
The sequence of n such that n^3 + 1 = p*q*r with p,q,r distinct primes starts out: 9, 10, 12, 13, 21, 25, 30, 34, 36, 40, 46, 52, 66, 76, 81, 90, 96, 118, 126, 130, 132, 142, 144, 154, 165, 172, 177, 180, 193, 196, 198, 204, 216, 226, 228, 238, 240, 246, 250, 256, 262, 268, 273, 282, 294, 312, 333, 336, 345, 346, 366, 370, 372, 378, 393, 400, 406, 408, 420, 436, 438, 442, 457, 462, 466, 477, 478, 496, 501, 508, 513, ... So, as you can see this happens very often. The above sequence is not in OEIS (maybe I'll add it) but the supersequence not requiring p,q,r distinct is A115403. which leadsd me to think that rich's expectation
(that only 7*11*13 = 1001 works) is correct
BTW, it is easy to show that the middle prime is n+1
The middle prime need not be n+1, for example 2*5*73 = 9^3 +1 but 5 does not equal 9+1 more soon,
cal
Jim Buddenhagen
Almost certainly n=10 is the only case where n^3 + 1 is the product of three consecutive primes. As others have noted, n^3+1 = (n+1)(n^2-n+1). As noted, the middle prime must be n+1. Taking p = n+1, let p-i be the preceding prime, and p+j be the following prime. Then this equation can be rewritten as p(p-i)(p+j) = p(p^2-3p+3), which simplifies to (j - i) p - i j = -3p + 3, or p = (i j + 3) / (j - i + 3). Since the denominator is at least 1 (j and i are both even, so it is not zero), this means that p is (at least) approximately the square of (either of) the prime gaps around it. While it is unproven, it is almost certainly true that prime gaps are asymptotically considerably less than this. Franklin T. Adams-Watters
participants (3)
-
Chris Landauer -
franktaw@netscape.net -
James Buddenhagen