Given a k-unfolding of a function f, the bifurcation set is the collection of points q in Rk such that F( . , q) has a degenerate singularity. A useful fact is that if F is a k-versal unfolding, the bifurcation set is an invariant of the singularity type of f. This fact is useful for studying objects like evolutes ([P2]), Gauss maps ([BaGM1], [BrGT1]), and binormal vectors.
If we have a versal 3-unfolding, we can actually see the
bifurcation set. The following singularities can be expressed using
two variables u and v, and they are versally unfolded
using three (or less) parameters (the ti's).
The pictures of the bifurcation sets follow. (The fold, or
A2 singularity, is unfolded by one parameter, and so
technically belongs on this list. But its bifurcation set is just a
regular surface, and so it does not look very interesting. This list
is a subset of the list of all singularities types of codimension less
than 6, found by Thom. This result can be found in many sources; I
would suggest [G].)
Singularity |
Normal Form |
Unfolding |
Cusp (A3) |
u4 + v2 | u4 + v2 + t1 u2 + t2 u |
Swallowtail (A4) |
u5 + v2 | u5 + v2 + t1 u3 + t2 u2 + t3 u |
Elliptic Umbilic (D4-) |
u2 v - v3 | u2 v - v3 + t1 u2 + t2 u + t3 v |
Hyperbolic Umbilic (D4+) |
u2 v + v3 | u2 v + v3 + t1 u2 + t2 u + t3 v |
![]() |
![]() |
Cusp | Swallowtail |
![]() |
![]() |
Hyperbolic Umbilic | Elliptic Umbilic |
Bifurcation sets of the four singularities