I wanted to measure adhesion to my glass bed, so I made this small force probe, to figure out the force required to push it over.
Calibrate it on a milligram scale (pressing it down onto the scale with x-carriage/z-axis. Each 0.1mm step will give you a fairly reproducible increase in the pressure exerted by the probe as long as you don't give it time to relax. I got +/- 10% on repeated measurements.
Use the printer axes (I used the X-carriage) to move the small distances to generate force and bend the scale. To avoid the probe softening with the heat from the bed or extruder I kept it covered with aluminium foil when not in use.
Pushing my tower at 55mm above the base I got readings of around 0.5-1N using sugar/water/detergent coating on my glass bed (but very wide SD, so further control of coating factors is required). I haven't formally measured ABS applied to glass with ABS juice (acetone with dissolved ABS), but adhesion with this seems more consistent even without any real control of application thickness, so is on the to-do list.
See my post at http://forums.reprap.org/read.php?4,156005,216866#msg-216866 for more details.

To print it correctly you'll probably need to play around with the SCAD file to get wall thickness correct. Even so neither Skeinforge nor Slic3r do a great job of it, and some of the parts are 'supports' to encourage a reasonable quality print, and should be cut off after printing.
An arm of 0.5mm thick by 5mm high will distort about 2.5mm to exert 1N (1 Newton is equivalent to 100g) of force. Unfortunately the ABS tends to relax if the probe is held in position for any length of time, but readings seem to be reasonably repeatable.