Machine Design

Joint decisions

When comparing its importance to effort expended, few aspects of product development are more out of balance than joint design.

Dave Archer
Archetype Joint LLC
Orion, Mich.

An ultrasonic sensor measures tension of a 5/16-18 Grade 8 bolt installed in a part-handling system.

The screen shot is of a reference and test echo along with a bolt-tension reading of 1,563 lb. The bolt was initially torqued to 25 lb-ft, which in theory gives a 4,700-lb clamp load.

A screenshot of the fastening/joining screening utility from Archetype Joint.

Design decisions made early in product development are generally the most influential. But those involving fasteners rarely address whether a particular fastening method is best for the application, or if a joint is needed at all. Threaded fasteners, for instance, tend to be the attachment method of choice because they are easy to specify and viewed as inexpensive. In reality, fastener unit cost is about onetenth to one-third of total installed cost.

Perhaps a better approach is to select fastening and joining methods by a process of elimination, not from the refining of "established" methods. A software screening utility from Archetype Joint does just that. It looks at joint metrics including permanence, relative movement, and component materials, and flags incompatible options, leaving potential designs that warrant a second look.

This is important because most products contain nearly as many fasteners as they do other components. In automotive final assembly, for example, fasteners comprise nearly three-fourths of all installed components and consume about 66% of labor hours. No reliable correlation exists — across product classes — between fastener-to-component ratio and assembly time associated with fastening.

One reason: The true cost of fastening and joining may be largely hidden. Costreduction efforts typically start with a so-called Pareto analysis of the bill of material that emphasizes "high-impact" components. Such analyses may ignore fasteners because they generally account for less than 5% of total material cost. No surprise that fasteners are often the last items added to a new product's bill of materials.

Once a joint type is selected, consider conducting physical tests of the design before it's locked in. Though analysis and simulation are useful for the sizing and placement of fasteners and other joining methods, it is probably unwise to assume these tools can capture all the variables, in even simple joints.

Independent of the fastening and joining method employed, product failures often happen at the interfaces (joints) of connected components, not within the components themselves. For example, auto-industry estimates show fasteners are responsible for about 70% of warranty costs and 20% of product recalls. In aircraft structures, the main drivers of inspection and maintenance schedules are the potential for cracking, fatigue failure, and corrosion of riveted joints. And loose or improperly located hardware is cited as the most common cause of electrical shorts. Testing can help engineers identify such problems and take the appropriate action.

Fastening and joining methods are often straightforward in theory but can be complex in practice. For example, the basic relationship between the clamp load developed in a threaded fastener and installation torque is:


where T = applied torque, lb-in.; K = friction factor; D = nominal fastener diameter, in.; and F = clamp load, lb. However, a recent Air Force study found 76 variables that can influence K. Just 10 to 15% of the energy input during tightening contributes to clamp load, while thread and underhead friction absorb the remainder. This means that small variations in friction can significantly change clamp load.

The good news is recent advances in ultrasonic testing make it possible to directly measure tension in threaded fasteners, instead of relying on torque as an indicator of joint clamp load.

Case in point: After repeated fatigue failures of bolted joints in the field, one manufacturer used ultrasonic testing equipment to check bolt tension. The 5/16-18 bolts were initially torqued to 25 lb-ft, which in theory gives a 4,700-lb clamp load. But the tests revealed residual tension was only about onethird of initial levels. It turns out much of this loss was the result of a high preload needed to pull stiff mating parts together during tightening. Once the machine had run-in for a period of time, the joint "settled," causing further loss of clamp load. The fix: Tweak the fabrication technique to produce more dimensionally stable components that retain a safe percentage of clamp load.

In another case, tensile testing of single-lap-shear joints made in cold-rolled steel with various fasteners showed none were as strong as theory predicted. In application, offset and bending in the two coupons prevent joints from being put in pure shear. Moreover, failure in most of these examples was not caused by failure of the fastener or joined spot, but rather a tearing or deformation of the material around the joint, a common problem that is difficult to analyze and predict numerically.

Archetype Joint LLC

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.