Nut Torque Calculation

[turelli2]

I have a 2.125 inch diameter aluminum hub with a pitch of 18 tpi and was wondering how I figure out the amount of torque I need to apply to my steel nut. We've seen numbers in the area of 50-60 ft-lb but I don't know how they get these numbers.

[turelli2]

I have a 2.125 inch diameter aluminum hub with a pitch of 18 tpi and was wondering how I figure out the amount of torque I need to apply to my steel nut. We've seen numbers in the area of 50-60 ft-lb but I don't know how they get these numbers.

[exFSAE]

Be an engineer. There are very basic equations in your textbooks to determine torque required to load a threaded shaft to a certain preload force.

Better question to ask yourself is how much preload force you think you need.

[turelli2]

Thank you, exFSAE, it is always nice of you to give such constructive advice.

I have done the calculations and have figured out the preload already. I was looking to see what other teams use for the torque to make sure I was in the ballpark. The equations may be fairly basic, but determining the friction factors is not.

We have just torqued to 215ft-lbs resulting in 6000lbs of clamping force.

If anyone has information that is constructive (as opposed to exFSAE who often just takes up space) please reply.

[Peter Pan]

If lubrication is used friction factor (K) can be assumed to be between .12 - .18 . This is information that is found in your Shigley's Mechanical Design book.

[Adambomb]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by turelli2:
I was looking to see what other teams use for the torque to make sure I was in the ballpark. The equations may be fairly basic, but determining the friction factors is not.
</div></BLOCKQUOTE>

What would lead you to believe that any other teams who post here have it right? One year the fellow who assembled our suspension thought that our tapered roller bearings in our front hub would be happy with about 200 ft. lb of preload. I'm sure he thought it was a great idea, and may well have even done some analysis on it too. The nut and spindle lived at least.

[exFSAE]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by turelli2:
Thank you, exFSAE, it is always nice of you to give such constructive advice. </div></BLOCKQUOTE>

I'm truly glad my advice has been of service. I thank you for the kind compliment.

+1 to Adam. One thing I learned in FSAE is to pretty much completely disregard what any other teams are doing, as there's about an equal chance of it being fundamentally sound.. and totally absurd. Scratch that.. higher chance of it being ridiculous.

Which is a good segue... 6000 pounds of clamping force? Does that not seem excessive? How much tire force do you anticipate trying to pull the wheel off the hub? No more than 300-400 pounds? I'd be particularly wary given you have aluminum threads, and many FSAE students are monkeys when it comes to using tools. Seized or cross-threaded nuts on centerlock hubs is no fun.

600 pounds of clamping force is probably more than sufficient.. or 1000 if you want to be overkill. Light torque on the wheel nut with a positive lock (e.g. clevis pin through castled centerlock nut) and you'll have no problems. Wiping the threads down with a rag and lubricating them with something like Starrett M1 helps too, in preventing threads from seizing on burrs and assorted crap.

If you don't have an effective positive lock.. maybe you need more to prevent the wheels from coming off. But, I'd rather use the preload purely as preload, and use another mechanism to make absolutely sure the nut can't back off.

[Ockham]

If you know the tensile strength of the wheel studs, you can determine how much clamping force the bolt exerts by measuring its elongation while applying torque. The short answer is that there really isn't a formula for this, since measuring is relatively easy, and the design of wheel nuts varies widely. Basically, determine the stress/strain curve of your bolts, their load-bearing cross section (based on thread minor diameter), and apply torque until the measured elongation is equal to the strain required for the right clamping force.

[AxelRipper]

keep in mind theres always the torque to yield test... take the fastner and a piece of material and fasten it in a vise such that you can torque on the bolt/nut. just keep tightening it with a torque wrench until it fails then subtract 5-10 (ish) ft-lbs from your yeild value.

[cjanota]

I helped out with the calculations.

For most of our max load calculations we put half the weight of the car (with driver) on the wheel and assume coefficient of friction of 3 (Lots of down-force this year, 1.5 for a non-aero car.) That is 900Lbs side force on a 10 in radius resulting in 9000in-Lbs torque pulling the tire in the direction of positive camber. The hub has a 1.5in radius so the force needed at the center line to overcome that moment is 6000Lbs (9000in-Lbs/1.5in) Our threads should be able to hold about 60,000Lbs before stripping (2.125 Dia. threads) If we were to use a clamp force less than that, the top part of the hub/wheel mate could peel apart this separation would rotate around with the wheel leading to wear on parts and possibly fatigue failure.

To calculate the torque, the coefficient of friction between the threads and the nut bearing surface are needed. We assumed the two coefficients to be the same. They actually have an extremely large impact on the torque required. A CoF=0 leads to a torque of 68ft-lbs, Cof=0.17: 215ft-Lbs, CoF=0.6: 6000 ft/lbs! We just used an angle test to get the static friction coefficient for the stainless steel nuts on AL with lubricant. We got a value of 0.17.

We were a bit concerned about the SS on AL threads. We thoroughly cleaned and inspected the internal and external threads and used anti-seize as a lubricant. We have had them on an off a few times now. We cleaned and inspected the threads again and have found no signs of damage or wear.

[exFSAE]

Find the fastest time from any skidpad event last year... downforce car or not. From the skidpad radius, work out how many G's they were pulling. Don't think it will be anywhere near 3.0. I'd even be surprised if it were near 2.0.

On a good day with hot tires and a lot of downforce, I'd suspect max cornering of these cars closer to 1.8.

In any event.. IIRC Formula Atlantic cars with similar size hub threads, use around 250 ft-lbf of torque. That's a car that weighs more than double, with heaps more downforce and cornering capacity, and larger radius wheels. Seemingly something isn't adding up here.

[cjanota]

3.0 is probably quite conservative. I wouldn't rely on the skidpad times though. A larger radius corner would be at faster speeds and therefore have a high downforce. It also doesn't account for transient conditions. We have seen acceleration spikes of 2.0g on our accelerometers on a non-aero car.

[Pete Marsh]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by exFSAE:
Find the fastest time from any skidpad event last year... downforce car or not. From the skidpad radius, work out how many G's they were pulling. Don't think it will be anywhere near 3.0. I'd even be surprised if it were near 2.0.

On a good day with hot tires and a lot of downforce, I'd suspect max cornering of these cars closer to 1.8.

In any event.. IIRC Formula Atlantic cars with similar size hub threads, use around 250 ft-lbf of torque. That's a car that weighs more than double, with heaps more downforce and cornering capacity, and larger radius wheels. Seemingly something isn't adding up here. </div></BLOCKQUOTE>


I don't know about that. We use a higher nominal load for our designs than that, and torque them up to over 300Nm. I have seen fretting between the wheel and hub, indicating the pre load is not sufficient. In our case, it is usually caused by dirty threads and/or insufficient tightening, but still it can't be that conservative.

The diameter of the hub bearing area has a big influence on the pre load required. Maybe the F Alantic has a bigger hub, perhaps around 100mm, as opposed to the 70mm typical in FSAE designs.

Pete

[exFSAE]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by cjanota:
We have seen acceleration spikes of 2.0g on our accelerometers on a non-aero car. </div></BLOCKQUOTE>

Yes, except those are not real. Given that accelerometers ultimately are not exactly at the CG, when the sprung mass undergoes its initial roll motion this shows up as a 'false' extra linear acceleration in your sensor. Cars don't magically gain extra lateral capacity in transient maneuvers.

Interesting comment though, Pete. All I can say is from personal experience, on a car with a 1.25" diameter hub thread (~30mm), IIRC we didn't use much more than 35 ft-lbf torque on the nuts... we hit the spec and then whatever more was needed to line up the clevis pin hole. The clevis pin through the castled nut and hub worked fantastic in preventing the nut from backing off, and there was never a hint of fretting or wear on any of the components. Comically enough the hub was one of the most robust, monkey-proof parts on the car despite the fact Cosmosworks claimed in the worst case load scenario the FOS was 0.98. There's more to that obviously.

I suppose I could be wrong.. it's been a few years.. but I distinctly recall working it out and requiring very little preload.

Certainly more preload between hub and wheel doesn't hurt assuming nothing goes awry.. but I'm just very very wary of aluminum hub threads on one of these cars, even if they're hard anodized. Eventually if any of the bare aluminum wears through or shows and the thing galls up... Bad Things happen and you have to scrap a hub assembly. Seen it happen more than once.