Big bolts…
In my hand right now I am holding a bolt.
More specifically, it’s an Allen-key bolt (sometimes called a ‘cap screw’) that’s 30mm long and 10mm in diameter. It uses a metric thread.
It’s a high tensile bolt, which means – in plain terms – that it’s bloody strong.
I’ve just stepped over to the digital scales – it weighs 28 grams.
Now the reason that the bolt was sitting on my desk is that a moment or two ago I took it out of my pocket. And the reason it was in my pocket is that I’ve just stepped in from my home workshop, after finishing for the evening.
I’ve been building ‘Chalky’, my recumbent, full suspension touring bike that I hope to be one of the best human-powered touring machines in the world. Best for me, anyway.
And these 10mm, high tensile, Allen-headed bolts are used in lots of places in the anatomy of Chalky.
Lots and lots and lots of places.
In fact, two are used in the rear suspension arm pivot (as through-bolts in the sealed ball bearings), two are used to bolt the seat in place (making the seat removable for folding – Chalky has to be able to be flown internationally); two are used in the ball-bearing steering handle bars pivot; and no less than eight are used in the front suspension system.
That totals 14 bolts – or nearly 0.4kg! Add the weight of the nuts (usually brazed into place as fixed threads) and the weight grows to – well, I don’t know, because I don’t have a nut in front of me and I am too lazy to walk out to the workshop to get one.
But even taking just the weight of the bolts into account, that’s a simply massive amount of weight – in fact, something like 8 per cent of the weight of the entire machine.
Which is why I am looking at the bolt – and thinking.
If I went from 10mm bolts to 8mm bolts, everything would probably be fine – and the weight would be less. Six millimetres bolts? Probably OK in most of the uses.
And then I mentally back-pedal.
Some of the bolts are used in suspension systems where the arm has a lot of leverage over the pivot. On a bike, there should be very little lateral forces, but what if there are? On my previous trike I think I calculated forces of the order of 800kg (ok, for the pedantic, 7800 Newtons) acting on these suspension bolts. I wasn’t worried, because these 10mm, high tensile bolts are very, very strong.
So I am in a bind. I wear the 0.4kg of bolt weight – and never worry in the slightest about failure of the bolts, whatever conditions the machine is ridden in.
Or I slim down some of the bolts (oh yes, that also means carrying two sizes of bolts as spares, not just one…) and when there’s suddenly a need for 30kg of water on the back, or I hit a huge bump at 80 km/h, I shit elephants… wondering if the bolt is going to be up to the task.
But I’d be instead much happier thinking of all those hefty bolts doing their work with contemptuous ease…
Ten millimetre bolts, you’re all staying!
on March 3rd, 2009 at 4:50 am
I remember hearing about Colin Chapman (I think), the original Lotus designer. He was queried about why his race cars had a maximum of 3/8″ shaft size bolts. His answer was that a single high tensile 3/8″ bolt in double shear could lift a london double decker bus, so why does he need anything stronger than that? This was challenged, and he held a demonstration, with a single bolt holding the entire weight of the bus.
However we are also talking about the cars that experienced many failures due to lack of durability.
on March 3rd, 2009 at 7:41 am
Careful, the trunks are pointy! Have you looked into aircraft grade bolts? The price goes up again but they are always trying to save weight
on March 3rd, 2009 at 8:17 am
Move to 10.9 grade M8 will give back some safety margin. Also drilling the centers and across the flat (internal in this case) takes a lot of weight out. It is time consuming however.
Maybe you could drill out the centres to different diameters, rebuild chalky and try to break the bolts?
on March 3rd, 2009 at 11:22 am
Aren’t most high quality cap head bolts supposed to be a 12.9 grade anyway?
Though some pedantic people may argue that the picture shown is actually a ‘screw,’ as it has no plain shank, and others would want to know if it will be used with a nut or a threaded structure, but that’s really neither here nor there.
Apart from a larger drilled bolt having a different geometry, if it has a same cross sectional area from having a drilled centre, compared to a smaller bolt, then simple calculations will have them being the same strength given the same materials.
on March 3rd, 2009 at 12:56 pm
What you need is some way to measure the loads on the bolts. Then you assemble using the 10mm ones, and use the vehicle, hard, for 6 months.
Now look at the data. If this joint is getting 8kN loads sometimes, but this joint has never gone over 2 kN, then you can think about adopting a smaller bolt in joint #2.
This is like the old story of Henry Ford going around to scrapyards and doing measurements on the parts of old, worn out Model Ts. When he found a part that was never worn, even on the discarded old cars, he lowered the quality of that part in the current production vehicles (it was clearly overdesigned). Testing for fracture is much harder than wear however. You might sometimes reach 98% of the fracture load, without the bolt showing anything. The bushes and mounting holes might show some damage though, especially if made with that in mind…
The other approach is what they did with the 911 Porsche. Start as a 98 kW, 1000 kg design, and gradually increase the power, weight and tyre grip. When parts start to fail, beef them up, but not the others. Eventually you end up with a design that is close to optimal. But that is done best with someone else doing the high speed test driving.
on March 3rd, 2009 at 2:53 pm
Hi Julian,
In BMX we have similar issues with bolts.
There are three (well really four) options:
1. Go titanium. No sacrifice in strength. Heaps more expensive – especially if you have 400g in bolts!
2. Hollow bolts. Some companies manufacture hollow bolts for bmx, I don’t see why you couldn’t drill them out yourself. From what I hear there is no major sacrifice in strength – perhaps you could clarify?
3. Smaller bolt heads. You may notice there is a fair chunk of metal around the allen section. A number of bmx companies have machined down their bolts around this area to the bare minimum, while still being strong enough to torque down on. This saves a little bit of weight that adds up.
And the fourth?
4. Suck it up and deal with the weight.
I figure that while the applications are different, the excessive forces are similar. We jump up, down, and off things and exert massive amounts of unusual force on our bikes, and bolts never really seem to be a failure point, even when made out of exotic metals or drilled out.
Cheers,
James
on March 3rd, 2009 at 6:10 pm
Greetings fellow bolt nerds!
For through-bolting various bearings could also use shoulder bolts which have a ground section of the diameter. It really depends on how they’re loaded, for example if they’re in bending. One possible downside to overly large bolts is that clamped joints may be more prone to loosening?
Some catalogues have operating strengths and torques published which can be handy:
http://www.spstech.com.au/Engineer_guide.htm
http://www.unbrako.com/
The online catalogue thingy seems easier to follow that the pdf file which unfortunately only has some metric stuff towards the end..
on March 3rd, 2009 at 9:14 pm
Bolts or not the whole rig depends on a thin rubber tube holding the air in a tyre. I sometimes thnk about that as I tear down a hill… but not often 🙂
on March 4th, 2009 at 5:31 pm
You could use a finer thread bolt and then drill thru the center. But I wouldn’t.
An M10x1.5mm bolt with a 4mm hole thru the center would have less strength than an M10x1.0mm bolt with a 4mm hole because the depth of a metric thread is the same as it’s pitch. So an M10x1.0mm with a 4mm hole would have a thinnest wall section of 2mm (at the depth of the thread), but an M10x1.5mm would have a thinnest wall thickness of 1.5mm……. Not much meat, especially on such a small diameter.
on March 5th, 2009 at 8:37 am
A hollow bolt only helps you if the bolt is loaded in torsion or bending. Bolts are usually loaded in tension and shear. The bending load should be taken by what the bolt is holding together. Torsional load should be minimal apart from the preload applied with the spanner. The idea is that the thread translates that torsional preload into a much larger tensile preload based on an inclined plane.
I wouldn’t bother with hollow bolts.
on March 11th, 2009 at 8:34 am
Titanium and Aluminium bolts (and other fastners) are good to use in different applications.
It’s a different machine, but I managed to shave off a few hundred grams from my motorbike by using a mix of Ti and Al fastners. Al bolts are good to use in areas of low stress. eg. your water bottle holders, reflectors, etc. Ti bolts are better for use in high stess locations. eg. brake calipers.
On my bicycle, I replaced the front caliper bolts with Ti ones, but didn’t like the feel they gave, so I re-installed the standard steel ones. On the rear, I couldn’t feel any difference, so left them in place. Strength wise, they are interchangeable, but feel wise they’re not. I think it has something to do with the springiness of the alloys, with steel more springy than Ti.
Ti components cost a lot per unit; dollars compared to cents. Al components cost more than steel, but no where near as much as Ti.
On the aesthetic side, Ti and Al bolts can come in a range of colours.
Ti weighs around 60% of steel and I think Al weighs around 40% of steel.
Out of interest, Ti doesn’t conduct heat very well. I was trying to shorten one of the bolts I was using and when filing it down, I got one side red hot, yet was able to still hold the other end only about 20mm away.
on March 11th, 2009 at 4:59 pm
Greg, that’s interesting – how did the front brakes feel different?
on March 12th, 2009 at 7:47 am
It’s kind of hard to describe. When braking, it possibly felt a little firmer (which is good), but I could feel more high frequency(?) vibration. Maybe, vibration isn’t the right word, but it felt more rough and harsh when braking. The standard steel bolts feel smoother and less harsh.
I should say that it was the bolts that connect the callipers to the front fork that gave this difference in feel. The bolts that hold the calliper together I swapped and left as Ti – the callipers aren’t a mono-block design.
My brakes are a low-end Shimano hydraulic set up from 03 or 04, so different and newer designs may give different results. Like I said, I couldn’t tell the difference with my rear callipers.
Because you’re looking at doing something with your axle bolts, you may find yourself in a similar situation.
on March 12th, 2009 at 8:29 am
You may have seen this before, but it does illustrate how strong a few small bolts/studs can be.
http://www.mustangsplus.com/catalog/articles_lift_engine_08.pdf
on March 12th, 2009 at 10:01 am
Interesting link, Trevor. But I am never worried about the strength of bolts just in tension.
on March 12th, 2009 at 10:13 am
In theory, bolts are only supposed to be in tension (with the unavoidable torsion required to do them up).
In theory, theory and practice are the same, in practice, they aren’t.
on March 12th, 2009 at 12:29 pm
Some of the people commenting here seem to be confused between the terms “class” and “grade”. The term “class” is used to describe the relative tensile strength of a metric fastener. The term “grade” is sometimes used to describe the reltive strength of a fastener in any of the non-metric systems.
Class 8.8 and grade 8 fasteners, for example, are common fastener strengths. The metric class 8.8 fasteners are made from material that has 92000psi yield strength, and a grade 8 fastener(to SAE J429) has 130000psi yield (ref http://www.americanfastener.com/technical/grade_markings_steel.asp). Clearly they are different. The same can be said for the other class and grade markings.
Also, with regard to the earlier comment regarding titanium fasteners – Titanium is well regarded as a material for high end racing or aerospace applications. ie, high levels of maintenance are applied to vehicles in these categories. Titanium does not tend to perform well in spradic high/low load, low maintenace enviroments – ie push bikes, road cars. The nature of the material means that it will generate stress fractures that will eventually fail. Someone earlier mentioned applications in BMX racing. Titanium fasteners are often used in road and MTB racing also. They are also known to break fequently in MTB racing applications, so they tend to be replaced frequently.
All I am trying to say, is that for important fasteners such as Julian is implementing in his HPV, that in order to minimise weight, it is important to understand the load conditions that are applied to each fastener, and select appropriately!
on March 12th, 2009 at 1:37 pm
The metric classes Adam refers to above don’t actually have arbitrary numbers like 92 000 psi.
They are metric. A class 8.8 bolt has an ultimate tensile (breaking) strength of 800 MPa. And a yield strength of 0.8 times the tensile strength, or 640 MPa.
Likewise, a 10.9 bolt would have an ultimate tensile strength of 1000 MPa. And a yield strength of 0.9 times the tensile strength, or 900 MPa.
It’s only when this is converted into medieval units based on the thumb size of an english king that we get random numbers like 8.8 meaning 92000.
Typical Americans: making everything just that little bit more complicated for everyone…
on March 20th, 2009 at 12:34 pm
I race DH mountain bikes where the loads induced and the dynamic forces involved will be far higher than a recumburant bike. I’m struggling to think of even one application where there’s a 10mm bolt. All are either M4,5or 6. The only 10mm hard ware is the rear axle. The only failures I’ve ever experienced are in frame or tubes. To be honest I’d rather replace a couple of M5’s after a crash than fix the frame work a heavier bolt would’ve been attached to.
on March 20th, 2009 at 12:48 pm
Jamie, it’s a bit different when the bolts are the pivots of a long suspension arm. As I said, on a trike it’s not hard to get 800kg forces acting on the bolts.