For specificity, I assumed n point mass beads, each with 1 unit of string length between. The last (outermost) will be bead 1, next inwards bead 2, and so forth I immediately realized that not only is each bead at a different radius, but also a different velocity, ω*ri, with ω the angular velocity of the car and all the beads. Let v and r apply to the outermost bead then ω = v/r.
The string supporting the outermost bead is at angle atan(v^2/rg) from vertical to balance forces including centripetal force. The tension in the string resolves into m*v^2/r horizontally and mg vertically.
Moving on to bead 2:
r2 = r - sin(atan(v^2/rg) = r - (v^2/rg)/√(1+(v^2/rg)^2)
the vertical component of T supporting bead 2 must = 2*mg
the horizontal component of T supporting bead 2 must be mv2^2/r2+mv^2/r =
mω^2r2+mω^2r = mω^2(2r - (v^2/rg)/√(1+(v^2/rg)^2)))
so each successive centripetal force on the bead gets a bit stronger and the angles a bit farther from vertical, but beyond writing the second angle as an atan, I'm not seeing how to get the overall curve. It looks to me that the algebra explodes without major simplification.
Perhaps the answer is a continuous solution with no beads and distributed mass in the string. Since I always fail at deriving the catenary, I'm sure that problem would stump me too.