Tightening connecting rod bolts while measuring bolt stretch provides a much more accurate method of achieving proper bolt preload and clamping force. Professional performance and race engine builders have long realized that the correct approach to tightening connecting rod bolts is to stress the bolts into their "working" range of elasticity, but not beyond. Since OEM connecting rod bolts may vary in terms of their ideal torque by as much as 10 lbs./ft. from batch to batch due to variations in heat treating and materials, if the concern is to arrive at both peak bolt strength as well as maintain concentricity of the rod big-end, the rod bolts should be measured for stretch instead of simply tightening until the torque wrench hits its mark. In simple terms, in order to measure bolt stretch, first measure the total rod bolt length (from the head surface to the tip of the shank) in the bolt's relaxed state. Then monitor bolt length as you continue to tighten until the specified amount of bolt stretch has been achieved. The difference in length indicates the amount of stretch the bolt experiences in its installed state. For the majority of production rod bolts, stretch will likely be in the 0.005" to 0.006" range (this is a generalization; always follow the specific stretch specified by the bolt maker). If it's difficult to achieve enough stretch, the bolt is probably experiencing too much friction that is preventing the proper stretch (requiring lubricant on the threads). If stretch is excessive, the bolt may have been pulled beyond its yield point and is no longer serviceable. While an outside micrometer may be used to measure the rod bolt length, the most accurate method is to use a specialty fixture that is outfitted with a dial indicator. Excellent examples of this gauge include units from ARP, GearHead Tools and Goodson Shop Supplies. GearHead's bolt stretch gauge features a heat-treated aluminum frame (with a very handy thumbhole) with a specially modified dial indicator with sufficient spring tension to hold the gauge firmly to the ends of the rod bolt. The indicator can be rotated for right- or left-hand operation and the lower anvil is adjustable to accommodate various bolt lengths. Goodson Shop Supplies also offers a rod bolt stretch gauge (P/N RBG-4) featuring spherical points for consistent and repeatable readings, and can also be rotated for right or left hand operation. Also, ARP offers its own bolt stretch gauge (P/N 100-9941) designed with 0.0005" increments, with a heavy spring and ball tips. There is a debate among some engine builders regarding the validity of measuring rod bolt stretch, due to potential compression of the rod material as the rod cap is clamped to the rod. While this may be a consideration, the use of a stretch gauge remains the best and most practical method of accurately determining bolt load. Connecting rod bolts can be viewed as high-tensile springs. The bolt must be stretched short of its yield point in order for accurate and, most importantly, repeatable clamping of the rod cap to the rod. Improper or unequal bolt clamping force can easily result in a non-round rod bore. Stock, or production, rod bolts typically offer a tensile strength of approximately 150,000-160,000 psi. However, due to variances in bolt production, tolerances can be quite extreme with peak bolt stretch occurring anywhere from, say 0.003" to 0.006". If the installer uses only torque in the attempt to achieve bolt stretch, he runs the risk of unequal rod bolt clamping loads due to the potential inconsistencies between bolts. High-performance rod bolts are manufactured to much tighter tensile strength tolerances. ARP, for instance, calculates each and every rod bolt for stretch, and the bolt packages include reference data to that effect. The instructions actually recommend that a specific amount of bolt stretch should be achieved on each bolt (ARP cites 190,000 psi as its nominal, or base tensile rating, with actual ratings much higher in some applications). How can unequal/inadequate rod bolt tightness affect the connecting rod big end bore shape? Let's cite an example: If one technician reconditions the connecting rods using torque value alone to tighten the rod bolts and another technician who is responsible for final assembly uses the bolt stretch method, the final result can be out-of-round bores. This is because of frictional variances that will be encountered. As a result, the assembler using the stretch method may achieve a higher clamping load on one or more bolts as compared to the loads imposed when the rod reconditioner torqued the nuts without regard to actual bolt stretch. When a bolt is tightened with dry threads, as much as 80 percent of the torque can be exerted because of friction as opposed to bolt stretch. In a high-volume production rebuilding facility, technicians may not have the time to measure for bolt stretch. However, a slower-paced operation that is attempting to obtain maximum accuracy (for a race engine, as an example), is far better off using the stretch method instead of relying only on the torque method. A set of connecting rod bolts' instructions may list both a torque value and a stretch range, effectively giving you a choice of methods. Yes, tightening only to a specified torque value is quicker and measuring bolt stretch requires more time, but the best results will be achieved by measuring bolt stretch. So, unless you're in a rush, take the time to measure stretch, tightening each rod bolt to the recommended stretch range. It's all about the quest for precision.


Friction (underhead and thread area contact) is a variable that is difficult to control precisely. The best way to avoid friction variables is to use the stretch method when tightening rod bolts. The stretch method allows you to accurately control the all-important bolt preload, independent of friction. Each time a bolt is tightened (to value) and loosened, the friction factor is reduced as the mating surfaces (threads and underhead) "wear" in. Eventually the friction becomes fairly constant during future tightening. Considering this "evening-out" of friction, when installing a new rod bolt where the stretch method cannot be used (because of available space for the stretch gauge, etc.), the bolt should be tightened and loosened several times before trying to achieve final torque value. While the number of tightening/loosening cycles depends on the lubricant being used, ARP recommends when using its lubricant, five loosening/tightening cycles is sufficient. If a bolt is tightened using straight torque, you may not necessarily achieve the desired pre-load due to the variable of friction. Since we can't predict the frictional loss, measuring rod bolt stretch provides the most accurate method of ensuring that the clamping loads will be both sufficient for the task and that each pair of rod bolts will achieve equal loads. Bolt stretch is generated by a number of factors, including tensile strength and mass (the length of the bolt being stretched). The effective diameter of the bolt contributes to this. For example, let's consider two 3/8" x 1" bolts. One features a 1" long shank with threads on the full 1" of the shank length. The other bolt features 3/4" of shank length that is full-diameter and smooth, and only 1/4" of thread length at the tip. The bolt with partial thread will stretch less because the shank area between the head and nut engagement area has a thicker cross section. The partial-thread bolt will have a .375" diameter shank, while the all-thread bolt will have only a .324" shank (due to the smaller root diameter inside the thread path). ARP, to cite one example, calculates the stretch number for every bolt. On the spec sheet that is included with every bolt set, this stretch goal is listed, in addition to a torque value, but the torque value should be used as a guide only. ARP does not want the installer to use a torque value as the final indication of bolt stretch. Rather, the bolts should be individually measured for stretch, to assure that each bolt is installed at its optimum strength. While we cannot control the reaction of the connecting rod base material, at least consider the potential compression of the connecting rod material itself during bolt clamping. As the bolt is tightened, the head of the bolt will tend to embed itself into the rod, slightly compressing the stock material of the connecting rod under the bolt head. Production rods are typically softer, allowing the head of the rod bolt to sink into rod until the material under the bolt heads "work hardens" under compression. ARP recommends that the bolt stretch is based on the bolt itself and not on the compression of the rod since we can't accurately predict what the rod does in this state. Since too many variables exist in terms of rod bolts and connecting rods, we can't draw any generalized conclusions regarding ideal connecting rod bolt stretch. However, to use the Chevy smallblock 350 as but one isolated example, ARP typically looks for an installed bolt stretch of .0063". Since each engine/rod/bolt application differs, we cannot assume that ideal bolt stretch would be the same for any given application.


Fatigue failure is frequently caused by local stress risers, such as sharp corners. In bolts, this would correspond to the notch effect associated with the thread form. It is well known that maximum stress in an engaged bolt occurs in the last engaged thread. By removing the remaining non-engaged threads, the local notch effect can be reduced. This leads to the standard configuration used in most ARP rod bolts-a reduced diameter shank and full engagement for the remaining threads. Providing a local fillet radius at the location of the maximum stress further reduces the local notch effect. The reduced shank diameter also reduces the bending stiffness of the bolt. When the bolt bends due to deformation of the connecting rod, the bending stresses are reduced below what they would otherwise be. This further increases the fatigue resistance of the bolt. The direct reciprocating load is not the only source of stresses in bolts. A secondary effect arises because of the flexibility of the rod big end. The reciprocating load causes bending deformation of the bolted joint (rod to rod cap). This deformation causes bending stresses in the bolt as well as in the rod. These bending stresses fluctuate from zero to their maximum level with each crankshaft revolution.