Sometimes it can be a tight fit installing a headset cup / race. Bike forums are full of people saying putting the cup in the freezer helps installation based on their anecdotal, uncontrolled experiments. Internet forums are also full of people saying it doesn't help at all, or you need to use dry ice or something else.

Aside from everyone's personal anecdotes, or what someone on the internet said one time, is there any evidence that freezing headset cups makes them easier to install? Does freezing them really shrink them sufficiently to make a difference?

  • Heating metal expands it slightly, cooling metal shrinks it slightly. Question is whether the shrinkage of a household freezer is enough to do anything measurable/useful. They do have to be a good firm fit else creaks.
    – Criggie
    Jul 13 '16 at 2:11
  • 2
    They defiantly contract, and if you're handy with maths you could work out by how much. But, what that will tell you is a handfull of micrometers. A more useful answer is the anecdotal one you don't want.
    – alex
    Jul 13 '16 at 3:39
  • 2
    This really shouldn't be necessary. Apply a thin smear of grease and use a proper headset press.
    – Emyr
    Jul 13 '16 at 9:19
  • 2
    Are you attempting to install a headset? As per @Emyr 's answer, are you even using the right tools ?
    – Criggie
    Jul 13 '16 at 9:24
  • Only works for metal frame , and your mileage may vary : caps quality, room temperature, frame quality. etc.
    – mootmoot
    Jul 13 '16 at 15:34

According to this link the thermal expansion of aluminum 6061 is 13 micro inches/(inch x °F). A headset has a diameter of 1.125 inches, and circumference of 3.5325 inches. Assuming you change the temperature of the headset from 70°F (room temperature) to 0°F (typical household freezer), you'll change the circumference by 13 x 70 x 3.5325 = 3215.5 micro inches. Which is .0032 inches. The diameter will now be smaller by .001 inches. I really don't think that will be enough to make a difference when installing a headset. Hopefully somebody can check my math/science to ensure I'm applying the information correctly.

Here's a nice chart illustrating the thermal expansion of aluminum from 0 K to 293 K (20°C). This seems to fit with my numbers from above that shows a decrease of 40 degrees (+20°C to -20°c) results in a change in length somewhere in the vicinity of 1x10-3 length units.

enter image description here

  • The geometry may affect the numbers a little but the conclusion would still hold. And by the time you've handled it and started to apply pressure it will have already started to warm up.
    – Chris H
    Jul 14 '16 at 7:18
  • Thank you for a thoughtful answer instead of just calling me dumb! This is great.
    – user21245
    Jul 14 '16 at 11:29
  • Even if you froze the headset cups to absolute zero, you'd see very little actual contraction. About .004 inches. I really think that is enough to make a difference. Most manufacturing techniques aren't that precise and you'd probably get more variation from different headsets off the same assembly line.
    – Kibbee
    Jul 14 '16 at 16:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy