This post of Boris Bukh mentions amazing Gustav Herglotz's integral
∫10ln(1+t4+√A15)1+t dt=−π212(√15−2)+ln2⋅ln(√3+√5)+ln1+√52⋅ln(2+√3).
I wonder if there are other irrational real algebraic exponents α such that the integral
∫10ln(1+tα)1+tdt
has a closed-form representation? Is there a general formula giving results for such cases?
Are there such algebraic α of degree >2 ?
Answer
Here is a list of some of these integrals:
∫10log(1+t2+√3)1+tdt=π212(1−√3)+log2log(1+√3)∫10log(1+t3+√8)1+tdt=π224(3−√32)+12log2log(2(3+√8)3/2)∫10log(1+t4+√15)1+tdt=π212(2−√15)+log(1+√52)log(2+√3)++log2log(√3+√5)∫10log(1+t5+√24)1+tdt=π224(5−√96)+12log(1+√2)log(2+√3)++12log2log(2(5+√24)3/2)∫10log(1+t6+√35)1+tdt=π212(3−√35)+log(1+√52)log(8+3√7)++log2log(√5+√7)∫10log(1+t8+√63)1+tdt=π212(4−√63)+log(5+√212)log(2+√3)++log2log(3+√7)∫10log(1+t11+√120)1+tdt=π224(11−√480)+12log(1+√2)log(4+√15)++12log(2+√3)log(3+√10)++12log(1+√52)log(5+√24)++12log2log(2(11+√120)3/2)∫10log(1+t12+√143)1+tdt=π212(6−√143)+log(3+√132)log(10+3√11)++log2log(√11+√13)∫10log(1+t13+√168)1+tdt=π224(13−√672)+12log(1+√2)log(5+√212)++14log(2+√3)log(15+√224)++14log(5+√24)log(8+√63)++12log2log(2(13+√168)3/2)∫10log(1+t14+√195)1+tdt=π212(7−√195)+log(1+√52)log(25+4√39)++log(3+√132)log(4+√15)++log2log(√15+√13)
No comments:
Post a Comment