I was just watching a new highway overpass being built (Rt. 128). The steel I-beams are connected by plates with bolts aatching them to the beams. Why don’t they use rivets anymore? isn’t rivetting faster and stronger?
It’s mostly due to cost.
High strength steel bolts are cheaper than they used to be and can do the job well enough.
If you are installing rivets, someone has to heat up the rivets, then two guys (one on each side of the joint) have to install the rivet before it cools down.
Compare that to using bolts, where one guy can just carry a bunch of bolts in a pouch and can install them all by himself. You don’t need anyone constantly feeding him anything quickly, so there’s no additional man on the ground and the logistics are easier. Bolts are also a lot easier to install and don’t require much in the way of training and skill to do properly. Since it’s a less skilled job, the pay for it is less.
So there’s a huge labor cost savings with bolts.
See this demonstration of a rivet being installed. Note that there are 3 guys involved, one bringing over the hot rivet and two installing it.
Compare that to this video where there’s only one guy on each end of the beam installing the bolts. The video is about column climbers, so you’ll want to fast forward to somewhere around the 3:40 mark or so to see them actually install the bolts.
Is there any mechanical or structural advantage to a riveted joint - say if cost was no issue? Can a rivet joint do something a bolt joint can’t?
It just seems like a correctly engineered bolted joint could everytime be as good as a riveted connection. Then heap all the advantages of bolted joints on top of that. Notwithstanding there is some degree of skill required in installing bolts (ensuring correct alignment, surface cleanliness, correct torsion applied to the bolt / joint), the whole riveting process seems archaic and unnecessary.
Then again I’m a relatively young engineer who has never actually seen a riveted joint so don’t know if there are intrinsic advantages. But I have seen some mighty bolted connections.
Bolts can also be quality-controlled. They are made in a factory under constant conditions. Rivets are heated in the field to god-knows-what temperature and hammered who-knows-how-many times and you don’t really have any way to know if contaminants have been introduced into the rivet alloy when hot or if the rivet’s structure is compromised when it’s hammered or if it is cooled too quickly. With bolts you know you’ve got a million of em all exactly the same and they can even be torqued to a specified value every single time with great consistency.
I would say that bolts have gotten cheaper and welding has gotten better. Between those two rivets are obsolete.
It’s to make it easier for jet fuel to melt them.
This doesn’t specifically answer your question, but is related. Read the Wikipedia entry for the Citigroup Center, a skyscraper in New York City. After it was built, some additional engineering analysis showed that it might not withstand certain high winds. Part of the problem was that it used bolted joints on its load braces instead of the welded joints that were originally specified.
Not sure about riveted joints, but there seems to be a known difference between bolted and welded.
This.
Plus - bolted joints usually attach do at welded brackets and flanges. This type of weld area can be overdesigned without increased cost and welded in a controlled environment.
-
Bolt metallurgy can be significantly better than the hot rolled beams so fewer bolts are needed.
-
Riveted joints had to be overdesigned and required special skill, more preparation, and more time to assemble properly which all equal significantly more cost with increased design risk.
I just wanted to add that most structures I see (admittedly, we don’t have tons of skyscrapers in Austin) are constructed from concrete and rebar. I am guessing that there is an upper limit to how high these can be built.
Rivets can take load in direct shear. If you have a structure that has joints that only see a shear load, or for which it is impractical to install a threaded fastener (e.g. attaching sheet metal together, or a sheet layer onto a tubular structure) rivets may be advantageous. Ideally, threaded fasteners take loads only in tension (with certain exceptions like shoulder bolts) and have to develop enough clamping force to ensure that the friction from clamping exceeds shear forces on the joint. Bolted joints also provide a controllable spring force (i.e. the bolt itself acts like a spring with the joint material in preload) which offers controlled flexibility and damping, versus the less controlled preload and tendency toward plastic deformation of a rivet. If a bolt stretches or fails it can be easily replaced in situ; riveted joints often need extensive repair or replacement of the attaching member.
The cost and complexity of installing large rivets on steel beam structures is a factor as well but from an engineering design standpoint, welded joints and threaded fasteners are a lot more controllable and inspectable. In practice the only place I’ve ever used rivets is joining face sheet materials to stringers or tubes below using pop rivets, and even then I would prefer welding or adhesive bonding.
Stranger
Dumb question time, but why are they preheated? Is it to make them more plastic/malleable during the riveting process? Or is it to increase the compression force of the joint (due to thermal contraction as it cools)?
At any rate, I once saw a video where they were doing the opposite… they would dip the rivet in liquid nitrogen before inserting it in the hole. (I think they were building a boat or an aircraft - not sure.) Why were they doing this? Were they doing it to make the rivet “tighter” in the hole once it warms up and expands?
Rivets are heated to make the shank longer; once inserted into the hole and the heads peened into place, the rivet cools and contracts, making a secure (if variable) preload.
I doubt you’ve seen a rivet inserted into liquid nitrogen, but this is often done with bearings and non-removable pins with the intent of getting a slight interference fit once the bearing/pin warms and expands into the bore. This is a very secure joint because it creates a reload around the entire circumference of the bore, preventing the pin or bearing from rotating or sliding axially even under extreme loads. However, if it ever has to be removed it has to be bored or drilled completely out.
Stranger
Is the reason shear loads are a no-no just that the threads could get damaged? Or is there something else?
A bolt has some play, because the through holes must be slightly larger than the bolt diameter. (Otherwise you can’t install the bolt). So if a bolted joint is subjected to enough force to overcome friction, it can shift slightly. This is a problem when precise dimensional stability is required, e.g. optical instruments.
The usual remedy is to drill through the interface near the bolt and installing a tight-fitting pin. But I’ve seen rivets used in such situations. (A rivet bulges as it is hammered in, so a riveted interface has no play.)
Yes, to both.
Probably aluminum rivets. LN2 is probably overkill, but there are aluminum alloys that age harden at room temperature and must be kept refrigerated after heat treatment and before installation. As they age, they become harder (stronger) and less ductile, so cracking on installation could happen if they age harden before they are installed.
For the same issue I just posted about above. The bolt is always slightly smaller than the hole it’s in. If the stress on the joint changes direction, the interface shifts slightly until the bolt is under shear stress in the new direction.
the other thing would be for cyclic loads; e.g. a building swaying back and forth in the wind. the “play” between the bolt and the hole it sits in means over time the hole can enlarge and/or the shank of the bolt can begin wearing away.
Because the presentation they’re watching is boring. Try adding more jokes, maybe varying your tone of voice a little? Add some anecdata that would be relevant to steel beam’s lives?
That is certainly part of the issue, but it is also that the threads create a natural stress riser that can be exacerbated in shear. A properly designed and assembled threaded joint should not permit any shifting or overcoming of the friction between of the joint members; even intermittent separation of the joint and loss of preload is regarded as a failure condition because of the load cycles it puts on the fastener. In a correctly assembled joint, the external tensile load is shared between the fastener (in additional tension) and the joint members (in a release of compression) which reduces load cycle amplitude and thus, sensitivity to fatigue. This same preload condition also dissapates vibration and shock loads, making a fastened structure more forgiving over a welded structure. Plain threaded fasteners can also be used in oversized holes or ovaled holes to permit large tolerances in the structure whereas rivets and shear fasteners need to be pretty tightly fitted, requiring precise alignment or match drilling of holes.
There are threaded fasteners that are designed to take load in shear, such as the previously mentioned shoulder bolt or the Hi-Lok pin. These are expensive machined components versus the more simple swaged solid or blind rivet, but they are more precise and capable of carrying tensile loads, which rivets have a very limited capability to do. They still require the precise alignment or match drilling of holes, which is expensive and time consuming compared to threaded fastener joints, but are desirable in certain high precision applications.
Stranger
I wonder if this bridge is held together with self-sealing stem bolts or rivets?
