Torque applied to a bolt for belt tension. 3

[Yannis121]

Can anyone help me on the following calculations?

I have a belt & pulley system and a mechanism to create tension (please see attached picture).
I know from the manufacturer how much tension I should apply on the belt (that's ~100Kg = 220lb).

The tension is increased by tightening the nut No1 shown on the pic, with a torque wrench, so that the plate moves right horizontally.
How much torque should be applied on the nut No1, in order to have the required tension?
(The bolt is mild steel normal, 12mm diameter).

From some online calculator the given value is ~38Nm. [URL unfurl="true"]http://www.futek.com/boltcalc.aspx[/url]

Is that correct?
Can someone advise?

 

[dvd]

The only way that works is if you can calibrate against known torque/tension.

You might investigate the following, there are multiple sources for similar products:

force-deflection-gauges

sonic-tension-meter-and-accessories

 

[btrueblood]

Assuming the belt wrap is close to 180 degrees, you need 2*T force from the screw, or about 5 N-m by my simple calc's.

tension = 220 lbs = 980 N (note no use of stupid mass units for force here)
Torque = .2*D*(2*tension) = .2*.012m*(2*980) = 4.7 N-m

 

[Compositepro]

Belts are usually tightened to where, when you apply, say, 20 lbs. of force to the center of the span between the pulleys, you get X amount of deflection. With experience most mechanics can eyeball this. Using a torque wrench will not work well in practice, and I have never seen it done that way.

 

[btrueblood]

I've done it with a torque wrench, and it's at least as accurate as the deflection method - though I'd engineered it that way. The belt deflection requires you to know the force applied...how many mechanics carry a spring scale around with them? But, yes, most belt manufacturers have a calculation to give the deflection vs. force numbers, and their tech support people don't like the bolt tension method as it's not often applicable.

 

[MikeHalloran]

Turn the assembly around.
Use the screw just to retain and guide a helical compression spring, selected to apply the desired force just barely off coilbind. The force near coilbind is relatively predictable for a given spring design.
Use shoulder screws or captive features instead of clamp screws.

Better, find an appropriate spring operated tensioner that's already in production, and use that.




Mike Halloran
Pembroke Pines, FL, USA

 

[btrueblood]

Be careful in selecting a spring type tensioner; you don't want to use a spring that is resonant within the operating speed range of your machine.

 

[jszymkowski]

Your mechanism may benefit from using a bolt manufactured by SmartBolt; it has a built in indicator that indicates when a specific tension has been applied to the bolt. I am not a representative of SmartBolt, but I have found some very useful applications for their product.

 

[Yannis121]

Thank you guys for the help!

Btrueblood, please help me understand which formula you use.

T=Kdf

T (Nm)
K=0.2
d (m)
f (N)

T= 0.2 * 0.012 * 980 = 2.352 Nm

Why do you double it?

EditOk, you mentioned 180 degrees angle
So if I will apply force on my torque wrench on 90 degrees angle, I will need half the torque.)

Is it so easy?

Regards,
Yannis

 

[btrueblood]

The tension in the belt wraps through 180 degrees around the pulley, thus the total force on the pulley is 2*tension. This is the force that the screw needs to generate.

 

[TED7]

Is your bolt dry? Will it be rusty one day? Will it be sprayed with a lubricant because it has become sticky? Torque isn't a very useful measurement for what you need. I fitted a car wheel once to the manufacturers recomended torque and I could still wobble it by hand - it's a very subjective way of measuring how tight a threaded assembly is if you don't have absolute control over the thread conditions.

 

[Drazen]

Axial force produced by screw turning depends not only on torque applied to bolt, but also on friction resistance at both nuts, so you actually cannot use any simple formula to calculate it. Even complex formulas need actual data about friction resistance, which depends on mating surfaces conditions (nut-bolt contact surfaces).

You can get reliable data only indirectly - by measuring torque and measuring tension of belt by other means like deflection until you reach torque value to get satisfactory tension, but in real maintenance life you will need to make direct tension checks if you want to make sure that you did good job (as the condition of mentioned surfaces change over time).

So, your torque data can help only in rough pre-tensioning.