Corner tolerance is a value defined by you. It is used to decide whether a point where two wireframe objects meet is a corner point of a surface curve. To explain this graphically, suppose we have two curves A and B as shown below:
Point P is where they meet. Corner tolerance is used to decide whether:
- Point P is a corner point of a surface curve where curve A defines one surface curve and curve B another.
- Point P joins curves A and B to define the same surface curve.
Corner tolerance is compared with an angle between the two curves. This angle, which we will call C, is defined as follows.
Two tangent vectors are defined at point P: one for curve A 
and the other for curve B 
.
Angle C is the angle between these two vectors.
If the corner tolerance is greater than angle C, then point P is classed as a corner point. This implies that curve A defines one surface curve, and curve B another.
If the corner tolerance is less than or equal to angle C, then curves A and B are taken to be a composite curve which passes through point P.