Originally Posted by
Road Fan
All depends how accurate you need to be. All measurements are estimates. All results of calculations that involve measured data as inputs are also estimates. To know something exactly is an engineer’s or scientist’s will o the wisp (Webster’s definition: a delusive goal), which is never possible to achieve and might not be beneficial.
But how good is good enough? What dimension has the most control over your desired result (sensitivity functions with a Pareto chart)? What is the best technique for maximizing that measurement with tools on hand? Are the assumptions you may have to make reasonable?
It’s too complex a problem for anyone to just walk in and summarily dismiss another person’s method for a measurement.
1. I like to remove the fork, clean it, and prop it up on a flat surface (kitchen counter checked to be level ot at least flat) with a bare hub clamped in the drops.
2. Put a level on the steer tube and raise it so the steer tube is as level as you can see, with your level.
3. Block up the steer tube so it stays in that position.
4. Now your steering axis is as close to parallel to the ground plane as you can get it.
5. Now with a decent grade digital, dial, or vernier caliper you can measure the height of the hub axis, top of the steer tube, and bottom of the steer tube above the tabletop.
6. With basic arithmetic you can calculate the height difference between the hub axis and the steer tube center axis, to whatever number of significant figures you like. This is the fork offset. You should not have to do this frequently.
This is how I like to attack measuring trail for my bikes. I think taking off the fork is a good idea.
How accurate does it need to be? Accurate enough to at least distinguish a 40mm from a 45 or 50mm fork, otherwise, why bother measuring it?
And the technique matters, because if you are going to invest the time in trying to do it, what's the point if your method is so inaccurate that you could have just held a framing square up next to the fork and read the apparent rake right off the bottom. Sometimes we mistake really accurate components (lasers, complex calculations, dial calipers) with accurate measurements, but it doesn't work like that. If your measuring system is crude, it doesn't matter how straight a laser is.
If "whatever" is the level of accuracy you need, just use a framing square or T-square. If your eyeball is good at picking out the center of the stem top and lower headset race, you'll be as accurate with this method as anything else that doesn't use perpendicular triangulation.
If you want accuracy but don't want to remove the fork, measure to something perpendicular to the axis of the fork, like the BB. Much, much more accurate than anything else.
And if you want a super accurate rake number, pull the fork and use the better tested methods in this thread:
Correct way to measure a fork