!! This new way uses a generating method and consists of creation of micro-planes or micro-lines that belong to involute surface.

This may happen because involute surface has some characters:.

-If you imagine to cut the gear with a plane tangent to base cylinder, you will obtain some straight lines costanly sloped as base helix angle; There will be more than one parlel lines for each side, and their distance (in trasverse and normal directions) is equal to respectively base trasverse pitch and base normal pitch.

While tangent plane rolls over base cylinder, the lines move (but remain equally spaced ad oriented each other) according to the very simple rolling law over base cylinder:

If the cut contains the whole central teeth, as for gears with a few teeth, you will find that the distance between first left-side line and first right-side line of central teet is equal to trasverse (normal) base thickness.

This happens also if you imagine anyway to continue involute profile forward base circle for any number of teeth.

Then (and this are probably the most importan carachteristic) the teeth surface immediately close to section is always normal to our imaginary plane, and this surface is anyway always convex, so that may be substituted by micro-planes tangent to real surface, with a small positive error.

A daily show of this simple rules comes from span measurement: just imagine to substitue plates with a milling tool and place it at the first contact.

I defined two different machining modes, each with its own advantages:.

CONTINOUS MODE:.

-The gear is placed on a rotary table (B-axe).

-The milling tool is placed on a square-head indexed as base helix angle.

-Its edge lays constantly in Z axe on plane tangent to base cylinder.

-The gear rotates (B) and mill moves (X) according to rolling law; In each moment a circular arc substitutes the real generating line.

-Final form error depends upon milling path width, tool diameter and mimimum curvature radius (innest X position).

In case you wish to have a constant feed along involute profile, it is necessary to fit X-Feed according to (fig.26A) (fig.27A) (fig.28A) :

It is also necessary to verify that this value doesn't become too large, as curvature radius becomes very small.

(fig.24A) (fig.04U) ¼/A> (fig.05U) (fig.19U) (fig.20U) .

MULTI-PLANE MODE:.

-The gear is placed on a rotary table.

-The milling tool is placed on a square-head indexed as base helix angle.

-Both gear and tool are moved in any position that respects rolling law.

-In this X-B fixed positions the milling tool moves along Z and creates a small plane tangent to real involute surface.

-Final form error depends upon the number of planned cuts and minimum curvature radius (innest X position).

-A special case is made by first rough machining by a thin milling tool (fig.33A) (fig.34U) .

Both methods may be considered as a special condition of modified rolling, in case tool pressure angle becomes ZERO.

Toolaxe index = base helix angle.

Zero-offset X : Centerline of rotary table.

Zero-offset Y : where you want, but better on an intermediate section where it is possible to center a pre-existing spacewith.

Zero-offset Z : Centerline of rotary table.

Zero offset B : Centerline of tooth or spacewidth, with their own delta-X contribution; If center of tooth is used, for B=0 then X cut position must already be equal to half trasverse base thickness.

Tool position T along Z:

Rolling law during movement:_

Helix law for different Y-levels:

Feed fitting law , only in continous mode:

Distance of eventual 2nd rear tool, that operates over a close teeth