18

I don't think I can drop $50-$100 for a torque wrench right now. What parts of a bike are the most critical to get the torque just right? Is there a way to make sure the torque is ok without a torque wrench? I know what would happen if you don't tighten enough. Too tight is probably better than not tight enough! Is there a way to make minor adjustments based on some sort of observation or feel?

milesmeow
  • 2,363
  • 5
  • 27
  • 36
  • Torque wrenches can be had for $25 usd if you're ok with beam types, which are arguably more accurate than cheap click types. – RoboKaren Feb 02 '17 at 16:52

5 Answers5

15

You can't approximate torque by feel. Loctite, grease, carbon compounds can all alter the 'feel' when you are tightening down bolts and if there was anyway to 'approximate the feel' these variables render it fairly impossible.

If you use ti/aluminum bolts, carbon components, high end aluminum do not cheap out, get the torque wrench.

If you tend towards stainless/steel bolts and less expensive components, you can forgo the torque wrench and just ensure you are working to even torque. For example, if it is a faceplate make sure that the gap between faceplate and stem is even all the way around.

Saying 'too tight is better than not enough' is not true. Too tight can create stress risers in a component (specifically I am thinking handlebar) that can lead to catastrophic failure down the road. Too tight on a square taper crankset (specifically the fixing bolt) can cause deformation of the taper fit and eventual loosening of the crank.

Hopefully that helps.

Noticed I didn't answer the question fully.

Faceplate/Bar interface Stem/Steer interface Crankset (whether pinchbolt or square taper style BB)

Those are areas where I would ensure proper torque. Seatpost is dependent on style of bike (ie. Madone seatMAST, or carbon post/carbon frame) as to whether torque is super important. Rotor bolts, brake mounting bolts etc... but with increasing importance as exotic materials increase.

tplunket
  • 3,739
  • 20
  • 14
  • 4
    That brings up another point. Any substance on the bolt alters the proper amount of torque for the bolt stretch required... Normally, torques are specified expecting the bolt to be "clean and dry", so loctite, etc, will make the specified torque value incorrect. Note that some bolts may have torque specified expecting a light grade oil on them! It's never easy... :-) – Brian Knoblauch Sep 23 '10 at 13:04
8

In my area, Community Cycles is a non-profit organization promoting bike riding. For a modest membership fee (less than the price of a torque wrench) one gets access to their bike shop's tools. Their site includes a list of other bike collectives in North America, which might be a good alternative to buying your own tools each time you need a new one.

Commodore Jaeger
  • 565
  • 1
  • 5
  • 11
6

Too tight and something will break, it may be the thread on the bolt, or much worse the thread on in the hole. Bolts will also break, sometimes leaving a very hard to remove bit in your frame.

Not tightenough and it will work loose, and then you know to do it a bit tighter next time.

(You don’t need to be spot on with torque, just not a long way over or under)

However wheels and brakes are not hard to do without a torque wrench, so a lot comes down to what maintenance you will be doing and how much your bike is worth.

Ian
  • 6,113
  • 4
  • 32
  • 52
2

You said

Too tight is probably better than not tight enough!

That is incorrect - Too tight is just as bad as not tight enough.

What can go wrong? When its too loose, things slip under pressure, or there's enough slack to allow small amounts of motion, which cause wear.

And this can happen when its too tight.

enter image description here

Yes, the eagle-eyed among you will notice I managed to break two rotor bolts while reassembling this.

enter image description here

Criggie
  • 124,066
  • 14
  • 180
  • 423
  • 2
    Exactly. There's two ways to buy a torque wrench. After you've wrecked an expensive part by over-tightening something, or before. The second way works out much cheaper. – stib Dec 30 '19 at 21:49
1

I mean to estimate something like that you have to get the necessary variables for the estimation:

  • torque (target torque to achieve)
  • wrench length (distance from which the force is being applied)
  • Force (force being used)

Once you have this from either experience or user manual you can then just use it in the torque formula:

Torque = distance * Force

Replacing distance with a 10cm wrench for example; replacing Force with common 1kg object (like 1 liters of water):

Torque = distance * Force becomes 1Nm = 0.1m * 1kg * 9.8m/s^2

So a 1kg object at 10cm make a little less than 1Nm

  • If you need 1Nm you have to use 1kg at 10cm
  • If you need 5Nm you have to use 1kg at 50cm, or 5kg at 10cm
  • If you need 10Nm you have to use 1kg at 100cm, or 10kg at 10cm
  • If you need 20Nm you have to use 1kg at 200cm, or 20kg at 10cm

this is just an approximation and you should consider your specific gravity

and also you should clean and re-oil/grease or otherwise do the pertaining maintenance before you try adjusting anything as this is how to be more certain the torque is the correct one

DavidW
  • 5,190
  • 5
  • 23
  • 40
Owlet
  • 11
  • 3
  • This is definitely "better than guessing" but is still plus-or-minus 50% in terms of accuracy. – Criggie Apr 22 '23 at 02:29
  • 1
    Hi, welcome to bicycles. What do you mean by "your specific gravity?" My specific gravity, like most humans, is approximately 1; unless you expect there to be androids in the audience the variance would be less than the error bars of the other measurements, and 1 could be universally approximated. Unless you're talking about the specific value of g, which again, unless you're addressing a hypothetical audience on the moon or Mars, could be approximated at 10, since the variation over the surface of the Earth is less than 1% and the 2% difference between 9.81 and 10 is not significant. – DavidW Apr 22 '23 at 02:32
  • 1
    @Criggie With a 10cm wrench, perhaps. But if you have a wrench with a .5 m handle and a set of graduated hand weights ("dumbbells") you could do substantially better. I would be willing to accept an estimation of a 20% error. – DavidW Apr 22 '23 at 02:33
  • @DavidW consider that a proper torque wrench is "±2-4% over 20% to 100% of full-scale" when its due for calibration. Within 20% with a Standard deviation of 1 (ie 65% of readings within 20%) might be optimistic. People are horrendously bad at estimating and errors multiply. – Criggie Apr 22 '23 at 08:17
  • @DavidW you are correct, I meant the specific value of g. Really I used it because it's part of calculating the Force in the Torque simplified formula and I like to use the most correct numbers when possible and round only the result. Since this is an approximation you'll notice a value of 10 for g when expressing the next few lines so as to give an approximated value as required.

    to other readers: Please note that many physics formulas already come ready for estimation use since they are derived from more elaborated formulations. So best to use accurate information from the get go

     – Owlet Apr 24 '23 at 18:21