Go to my Home page.
Go to my Joy of High Tech page.

The Joy of High Tech


Rodford Edmiston

Being the occasionally interesting ramblings of a major-league technophile.

Fascinating Subject

      Following the last large earthquake to cause serious damage to a California city, officials were surprised to find that buildings with the recently recommended, welded connections in their steel frames were more damaged than otherwise identical buildings with bolted or riveted connections. Meanwhile, structural and civil engineers who had to deal with the damage were muttering dire threats under their breaths in re. meddling politicians who wouldn't listen to them when they explained the difference between theoretical and practical.

      A good welded joint is stronger than a good bolted or riveted joint. The problem is, getting a good welded joint is very difficult, while good riveted joints are fairly easy, and good bolted joints easier yet.

      The vast majority of welders go for a smooth, consistent bead. Unfortunately, neither smoothness nor visible consistency are proper measures of how good a weld actually is. Even more unfortunately, welders strongly resist even the suggestion that you can't judge how well their weld will actually hold by how pretty it is. They don't want to go to the courses offered to teach proper welding techniques (Because they already know how to weld, dammit!) and if forced to go ignore what is taught when they get back to work.

      To digress a moment, as with drivers, the majority of welders think their skill is above average. The greatest threat to this country is not terrorism, but hubris. A complete inability to consider even the possibility of error means a complete inability to prevent or correct mistakes.

      Riveting has been around for thousands of years. Hot riveting has been around for centuries. Both involve hammering or squeezing a smooth bolt of malleable metal against a solid backing, in the process shortening the rivet and drawing what is being connected together. The shaft also expands to fill the hole, adding rigidity. Some forms of riveting start with a plain shaft and round both ends off at the same time when applied. Others have a premade, manufactured head on one end.

      Hot riveting differs from the earlier version in that the metal is heated to make it expand, and as it cools the contraction further tightens the joint. This contraction produces a significant force, around 60,000 Newtons per square centimeter, enough that the bolt doesn't actually shrink as much as it would if unconstrained but remains strongly in tension. The friction between the flat steel surfaces thus clamped together adds substantially to the resistance of the joint to moving.

      Note that for steel rivets the temperature is not enough to significantly soften the metal, but only to expand it. Rivet metals are chosen largely for their tolerance for cold forging. Once steel replaced wrought iron, rivets were made from such low-carbon steels as ASTM A141. Modern alloys such as ASTM A502 are stronger, but still compounded and heat-treated to be ductile. They still are usually weaker in tension than the parts being joined. Since the most common mode of failure in riveted joints is the tearing of the metal between the holes that's not a problem.

      Hot riveting is labor intensive, requiring at least four members per team. One member heats the rivets, carefully judging when they're the right temperature. He then takes the properly heated rivets and throws them, one at a time, with long-handled tongs to the second member of the team, the catcher. The catcher catches (naturally) the thrown rivet in a can and, with another set of tongs, fishes it out, makes sure it's clean then places it in a hole. The riveter and bucker then set to work. The bucker's job is to provide the solid backing with a bucking bar, which is pressed firmly against the head of the rivet. The bucking bar has a cavity in the end mirroring the shape and size of the head, to ensure a firm, even pressure. The riveter then hammers - manually in the old days, with a pneumatic, hydraulic or electric rivet gun more recently - the other end of the bolt, rounding it off to create the field head. A more recent development involves a clamp-and-squeeze hydraulic press which also eliminates the need for a bucker. However, this device can only be used where it can reach both ends of the rivet at once.

      The process then repeats. Over and over, thousands of times even for a small steel-framed building.

      Yes, this process is tedious and not without risk. But with a minor amount of training a crew would soon get into a rhythm of smoothly heating, throwing, placing and riveting, making good, solid joints over and over.

      Rivets are deceptively simple things. Construction rivets look like a threadless bolt with a smooth, round, domed head (a button head rivet) or a flat square head (a countersunk head, which fits flush in a square hole) on one end. But! The metal has to be capable of being heated - with the temperature judged by eye via the color - pounded into shape without fracturing and then cool in an uncontrolled way without becoming brittle and while retaining enough strength to hold the joint together. The amount it expands when heated and contracts after being riveted must be the right amount. Too little and the joint will be loose. Too much and the head will snap off or what is being riveted will deform.

      In spite of orange-hot rivets being thrown around, serious injuries directly from a dropped rivet are rare (though getting one in a sleeve or boot could cause some impressive branding). A more worrisome problem was that they could easily start fires. So could the burners used to heat them. But for a long time the metallurgy and manufacturing practices for bolts just weren't up to the task. However, that is no longer true.

      Modern bolted construction is fast, easy and safe. The teams are smaller, generally with just two members, and the work requires less training. Construction bolts or nuts usually have a "fuse" connection, which lets the driven head snap off at a predetermined torque, leaving the fastening head behind, properly snugged down. With one person placing the bolts and another using a power wrench to tighten them, bolted connections can be placed in the same holes as rivets, only much faster. Indeed, as far as the end product is concerned, bolts are simply slightly inferior rivets. For the consumer, they offer the added feature that they are easier than rivets to remove.

      In contrast to rivets, bolts are usually about as strong as the parts they connect, and can be much stronger. This is not to make the joints stronger; as mentioned above, joints of this type usually fail by tearing between the holes. Rather, it is a byproduct of producing bolts which will take and hold threads properly. These threads must deal with the concentrated forces caused by tightening, which are far higher, locally, than the overall tension stress put on the shaft as a whole.

      Rivets and bolts both require holes in the material being fastened. This is not necessarily a bad thing. The total amount of metal removed for the holes is small compared to that of the beams or plates being connected. The holes can be used to align the parts, through the application of tapered tools. And holes are often placed in components deliberately to stop cracks from propagating, anyway.

      What most people think of today as welding is actually a fairly modern process. Originally, welding involved placing two hot pieces of iron or steel together (usually with the addition of a flux to help prepare the surfaces) and pounding on them until they fused, becoming essentially one piece. Today, though, welding is an additive process. Whether accomplished by electricity or through burning gases, modern welding involves putting two pieces of metal tightly together, then using heat to fuse material from a welding rod to both pieces.

      Welded joints involve a different geometry from bolted or riveted ones. They don't go through the material, but only around the edges. They are therefore only applicable where the geometry of the connection produces a contact surface which is large enough to handle the stresses and strains on the joint. Given that, however, a good welded join is quite strong.

      Another problem with welds is that many construction materials undergo physical and even chemical changes at the temperatures needed for true welding. (Note that brazing, soldiering and so forth are not welding, and need lower temperatures.)

      Inert gas welding is required for some materials, such as most aluminum alloys. The area being welded is flooded with helium or something similar to keep the metal from combining with oxygen from the surrounding air.

      There is also the problem of needing a clean surface for the weld. Grit, grease, paint and other materials which get between the bead and the structural metal physically interfere with the bond. They can cause chemical changes, as well. Worse, such contaminants can vaporize and explosively escape, spattering molten metal through the surrounding area, including on the welder.

      Even when the parts being welded are not adversely affected by the process, and even when the job is done properly, welds still have drawbacks. There's none of the crack-stopping capacity you have with rivets and bolts unless you put holes in, anyway. And getting the job done properly isn't easy. Bitter experience has shown that welds - or at least a high-percentage sample of them on each project - must be inspected and the inspection requires more than just looking. To find whatever flaws might lie under the bead, dedicated examination equipment may be needed. With bolts and rivets any problems with a joint are generally obvious.

      If you know you have good materials going in and the people doing the work are properly trained, rivets and bolts can be passed with a visual inspection which is little more than cursory. Welded joints generally hide their problems.

      So please try to keep in mind the difference between theory and practice when making real-world decisions. ;-)

          This document is Copyright 2006 Rodford Edmiston Smith. Anyone wishing to repost it must have permission from the author, who can be reached at: stickmaker@usa.net