Todd, While it is true that forks involved in catastrophic failure typically bend backwards, this is a result of either direct frontal impact, or impact on a part of the wheel (say, 5 o'clock, in an extreme case) that is not normally in contact with pavement. It is the direction of the force, or the need for the bike to change direction in order to get over the obstacle that results in the more horizontal vector that bends the fork (and sometimes the frame) and, when this happens, it is typically noticed at the upper end of the fork blades, the crown, or the lower part of the steerer.
But what about the minor road imperfections that result in the "road feel" experienced in normal riding? While larger cracks and small potholes might result in force vectors that drive the fork back, vibrations caused by normal road surfaces are created at the very bottom of the wheel and logic indicates that the vector here is almost completely vertical. Take a look at the design of a Schwinn knee-action spring fork, and you see that the spring absorbs shock when the fork bends FORWARD, not back. http://www.oldschwinn.com/
I am not arguing that the flex in a fork does not take place in its upper half, but I don't accept that normal forces on a fork push the fork ends backward, in spite of what fork failures may indicate.
Steve Barner, Bolton, Vermont
>Date: Wed, 24 Dec 2003 10:49:30 -0600
>From: Todd Kuzma <tullio@TheRamp.net>
>To: <classicrendezvous@bikelist.org>
>Subject: Re: [CR]The shape of fork bends in the old days
--Snip--
>
>Your example above about the Raleigh fork illustrates this. You state "with
>the wheel pushing straight up on that . . ." However, the fork does not
>flex upward over bumps. It flexes rearward. The fork is cantilevered off
>of the frame and its pivot point is the crown. The steerer must bow in the
>opposite direction (forward) to permit the movement at the crown.
--Snip--
>Todd Kuzma
>Heron Bicycles
>Tullio's Big Dog Cyclery
>LaSalle, IL 815-223-1776
>http://www.heronbicycles.com
>http://www.tullios.com