Drilling hole in steel fork steerer

[Brucey]

OK, but stem bolts usually break because they are low-cycle fatigued in bending. In this respect two skinny bolts is often going to be worse than one fatter bolt, because the bending strength of the bolts is very greatly reduced, going as it does with the cube of the diameter. The worse the fit of the parts, the greater the bending loads are on the bolt(s) as they are tightened. How the bolt holes are drilled and how the bolt head is seated also influences this. In fact you are often (nearly always) much better off fitting studs with Loctite into the threaded assembly, and using some kind of nut on them.

Comparing M7x1, M6x1, and M5x0.8 bolts

M7 x1 -default bending strength.................. working cross section in tension = 36mm^2

M6x1 58% of the default bending strength .....working cross section in tension = 25mm^2

M5x 0.8 34% of the default bending strength......working cross section in tension = 17.6mm^2

So 2x M5 screws theoretically can exert the same clamping force as one M7 bolt, but each M5 screw is about three times more likely to break in bending. M6 screws are almost twice as likely to break in bending as M7 screws.

cheers

[MikeDee]

One can introduce a crack in a bolt by dropping it on a concrete floor. Two bolt clamp and four bolt face plates is what Brandt recommended. Thats what I try to use, although I have some two bolt faceplate stems that I use.

[Brucey]

One can introduce a crack in a bolt by dropping it on a concrete floor....

not the sort of bolts that are normally used on bicycle parts, I think.

cheers

[MikeDee]

One can introduce a crack in a bolt by dropping it on a concrete floor....

not the sort of bolts that are normally used on bicycle parts, I think.

cheers

When I was working, I had a number of new socket head screws cut, sectioned, etched and mounted; foreign and domestic fasteners as part of an evaluation of foreign fasteners that made their way into our supply, in spite of the rules we had to use domestic only. Many had thread cracks. The metallurgist said categorically that he did not cause the cracks. For whatever it's worth, I do not trust that these bolts/screws are without flaws and that redundancy is desirable.

This also applies to seatposts as well. Two bolt designs are better than one. Brandt described where a friend of his was riding no handed eating a banana, and the seat bolt broke and he fell backwards on the rear tire. He was hospitalized and had to have anal reconstructive surgery. Makes me cringe just to think of that. I had an American Classic seatpost that had a small bolt clamping the seat. It broke on me, but luckily when I picked the bike up by the saddle. I don't use that seatpost anymore.

[Brucey]

The manufacture and use of fasteners is a whole science by itself. What baffles me is that otherwise intelligent folk will insist on designing/making/buying parts that have fasteners that are poorly specified for the task in hand, in clamps etc that generate loadings that the bolts are not designed to withstand.

Doubtless you would prefer two pathetic M5 bolts to a single M7 bolt, even though I have already demonstrated that the M5 bolts are three times more likely to fail in low cycle fatigue when they get bent as they are tightened (which they almost invariably will do if/when the parts don't fit very well). In fact it is worse than that; in service any clamp that sees bending loads (eg a stem for example) imposes almost all the load on a single bolt of a pair at some points; they don't always share the load at all well.

A very common mistake is to use 'a stronger bolt' (e.g. 12.9 grade or higher) in the (usually mistaken) belief that it will be 'better' in some way. If manufactured correctly such bolts might perform better in a simple tension test, but

a) such tests are of questionable relevance in relation to typical bike parts and
b) the manufacture of such bolts is rather more difficult and
c) actual bolt failures in service are only weakly related to the nominal strength of the bolt and may occur for other reasons.

If dropping bolts on the floor routinely caused bolts to crack then we'd all be dead already; they see worse knocks than that during handling and shipping.....

Typical bolts that are used on stems and seat posts are 8.8 or 10.9 grade, and manufactured with rolled (not cut) threads. The steels that are used for this are not subject to severe risk of temper embrittlement, they are not so lacking in ductility that they are likely to crack during manufacture (or in service) they are not subject to a high risk of SCC or hydrogen embrittlement, and they are usually of adequate strength and ductility for the task in hand. That is not to say that such bolts cannot ever be defective or fail in service; it is just very much less likely than with nominally higher strength grade bolts which are intended to be loaded to the same proportion of their nominal strength.

A bolt with cold-rolled threads will have beneficial residual stresses and a better aligned grain structure in the thread profile than otherwise. The benefit of the residual stresses lasts until you bring the bolt up to yield, at which point it is largely lost. So it is obvious that you shouldn't overtighten bolts. What is less obvious is that a good number of (usually poorly fitting) clamps cause bolts to see such high bending loads that they approach (or exceed) yield in bending as they are tightened; this lies at the root of maybe nine out of ten of the failed bolts that I see on bike parts, hence my earlier suggestion that replacing bolts with captive studs and nuts is usually a better idea.

The other thing I'd suggest is to read the small print; when you buy some lightweight parts they have a rather stringent 'best before' date. For example a Thompson seat post is intended for a 10000 mile life. IME that is OK in benign conditions but is otherwise rather optimistic; if you ride on salted roads you could break one sooner than that. You would be amazed at how some folk react when their parts break; these are often (rather frou-frou) parts that are meant for a short life even under good conditions; they are (often decades) past their sell-by date, they have been abused (eg by the bolts being repeatedly overtightened to the point that they are obviously deformed) and they have been sprayed with salty water and not cleaned. "Why did it break?" they say, amongst other wailing and gnashing of teeth etc. Often a better question might be "how did it survive this long?"....

cheers

[MikeDee]

The manufacture and use of fasteners is a whole science by itself. What baffles me is that otherwise intelligent folk will insist on designing/making/buying parts that have fasteners that are poorly specified for the task in hand, in clamps etc that generate loadings that the bolts are not designed to withstand.

Doubtless you would prefer two pathetic M5 bolts to a single M7 bolt, even though I have already demonstrated that the M5 bolts are three times more likely to fail in low cycle fatigue when they get bent as they are tightened (which they almost invariably will do if/when the parts don't fit very well). In fact it is worse than that; in service any clamp that sees bending loads (eg a stem for example) imposes almost all the load on a single bolt of a pair at some points; they don't always share the load at all well.

A very common mistake is to use 'a stronger bolt' (e.g. 12.9 grade or higher) in the (usually mistaken) belief that it will be 'better' in some way. If manufactured correctly such bolts might perform better in a simple tension test, but

a) such tests are of questionable relevance in relation to typical bike parts and
b) the manufacture of such bolts is rather more difficult and
c) actual bolt failures in service are only weakly related to the nominal strength of the bolt and may occur for other reasons.

If dropping bolts on the floor routinely caused bolts to crack then we'd all be dead already; they see worse knocks than that during handling and shipping.....

Typical bolts that are used on stems and seat posts are 8.8 or 10.9 grade, and manufactured with rolled (not cut) threads. The steels that are used for this are not subject to severe risk of temper embrittlement, they are not so lacking in ductility that they are likely to crack during manufacture (or in service) they are not subject to a high risk of SCC or hydrogen embrittlement, and they are usually of adequate strength and ductility for the task in hand. That is not to say that such bolts cannot ever be defective or fail in service; it is just very much less likely than with nominally higher strength grade bolts which are intended to be loaded to the same proportion of their nominal strength.

A bolt with cold-rolled threads will have beneficial residual stresses and a better aligned grain structure in the thread profile than otherwise. The benefit of the residual stresses lasts until you bring the bolt up to yield, at which point it is largely lost. So it is obvious that you shouldn't overtighten bolts. What is less obvious is that a good number of (usually poorly fitting) clamps cause bolts to see such high bending loads that they approach (or exceed) yield in bending as they are tightened; this lies at the root of maybe nine out of ten of the failed bolts that I see on bike parts, hence my earlier suggestion that replacing bolts with captive studs and nuts is usually a better idea.

The other thing I'd suggest is to read the small print; when you buy some lightweight parts they have a rather stringent 'best before' date. For example a Thompson seat post is intended for a 10000 mile life. IME that is OK in benign conditions but is otherwise rather optimistic; if you ride on salted roads you could break one sooner than that. You would be amazed at how some folk react when their parts break; these are often (rather frou-frou) parts that are meant for a short life even under good conditions; they are (often decades) past their sell-by date, they have been abused (eg by the bolts being repeatedly overtightened to the point that they are obviously deformed) and they have been sprayed with salty water and not cleaned. "Why did it break?" they say, amongst other wailing and gnashing of teeth etc. Often a better question might be "how did it survive this long?".... [emoji57]

cheers

The major flaw with your argument is you can't ensure that there is not a pre-existing crack in the fastener with these cheezy commercial grade fasteners. It's also tricky to figure out if the bolt is loaded in bending if the fit seems pretty good. In that regard, redundancy is preferable. If the bolt breaks the stem or handlebar doesn't come undone and you don't crash.

I know you don't always agree with Brandt, maybe because of the "not invented here" philosophy, but this is what he wrote on the subject "For safety and reliability, get a stem that has two screws on the steer tube end and four on the front plate so that a one screw failure will not let the handlebar go free."

[Brucey]

I'd add a caveat to that which is "unless the screws are M5 ones where they clamp the steerer". In which case you should look for another stem.

FWIW I have 'very often' seen broken M5 bolts in seat clamps and handlebar/stem clamps. I have 'quite often' seen M6 bolts break in similar situations (seat binder bolts that are badly designed for example). But I don't think I have ever seen an M7 bolt break in a stem, even if there is one bolt in a steerer clamp and/or two bolts on a detachable face plate.

The 8.8 and 10.9 bolts that you describe as 'cheezy' have a known tensile strength and recommended torque settings. For a whole host of reasons they are far more reliable bolts than ostensibly stronger grade ones if they are made to a cost. Apart from pure stupidity, what breaks them is if the parts don't fit well; it is in fact usually easy to see if this is the case; the clamp has to be deformed appreciably before it grips the steerer or handlebars. In extremis you can actually see the bolts are bent.

FWIW I don't often use them myself, but I accept M5 bolts on 31.8mm faceplates because that is pretty much all you can get. The reason I am not keen on the 31.8mm handlebars is that

a) they are needlessly stiff and
b) they are usually so thin walled that they are easily damaged where they are clamped.

The damage typically arises because often there is a slight mismatch between the diameter of the clamp and the that of the bars; the bars are not held securely and the rider (understandably) overtightens the bolts. Even with M5 bolts the end result is usually crushed handlebars, just waiting to crack.

cheers

[Gattonero]

An alternative to drilling a hole in the top cap might be to have a spacer on top of the stem and drill a hole in that (or cut/file a slot in the edge of the spacer). The wire would then exit horizontally, which should be neater and might also help a bit with minimising water ingress through the hole (although I would still try to improve that by using a bit of silicone sealant or similar).

That's what I did many years ago with a wired cyclo computer, a 5mm spacer over the forks (which is ok even with a carbon steerer, as long as the expander bung is proportionally long) with a slot filed on one side. The end of the fork's steerer protruding 2mm in the spacer and <1mm of bulge from the top-cap had left a good 2mm that was more than enough for the wire to run through.