Rod-Bolt School

Uncovering The Truth About Rod-Bolt Torquing And Bolt Stretch

Jim Resnick Dec 1, 1997 0 Comment(s)

Step By Step

Tom Smith, manufacturing foreman at ARP, uses the company’s rod-bolt stretch gauge. It’s best to get a gauge with a stiff spring inside; Smith recommends a 2.5- to 3-pound spring so that the chance of erroneous readings is minimized.

You can also mark off the stretch spec with movable hands on the outside of the gauge. Here, Smith’s reached the spec for this bolt: 0.007 at 0.0072-inch bolt stretch.

Using a conventional torque wrench, Gail Trent of American Speed brings up the rod bolts of this LT4 to torque spec three times. The first two polish and burnish the bolt threads, the nut surface, and the bearing surface between the nut and cap. By a third torquing, everything’s polished and slicker than after the first. It’s lost the friction that gives the initial reading on the torque wrench. That third, final reading is accurate to within 0.0002 inch of bolt stretch on ARP bolts using ARP moly lube, according to ARP’s Tom Smith.

When you’re using a torque wrench alone, the rod-bolt manufacturer’s lube (or another specified lubricant) should always be used. With other lube, your torque readings of say, 50 lbs-ft may match the manufacturer’s, but since the friction coefficient of the two different lubricants probably doesn’t match, the bolt’s stretch may not be on spec, regardless of the torque reading.

The nut’s bearing surface must be lubed along with the bolt and nut threads, too.

Here are the three stages of bolt life. At left, the stock Pro Series bolt has not been used. In the middle, the same type of bolt has been stretched on a machine. At right, separation and failure occurred at the designed location in this Pro Series bolt.

Bolt failure or build failure? Well, obviously, the bolt broke, but only after the nut backed off the bolt, leading to lots of damage. Had the bolt been properly torqued to its correct spec, this might not have happened.

Here is ARP’s torture tester, where bolts are sacrificed—but not before a graph of strength is logged as the bolt is pulled apart.

Engine-building myth shattered: Connecting-rod bolts rarely fail. Or, more accurately, they’re rarely to blame when engines have what is usually referred to as rod or rod-bolt failure. More often than not, the nuts back themselves out slowly, due to incorrect torquing procedures. Which begs the question, “What’s the best way to torque rod bolts?”

The best way (but one that isn’t always done), according to ARP’s manufacturing foreman, Tom Smith, is with a rod-bolt stretch gauge. This gauge is a beautiful device for the pro engine builder and bucks-up hobbyist. But is it feasible for the homebound engine builder? There’s an appropriate tool for every task, and connecting-rod bolts go best with a rod-bolt stretch gauge. We picked Smith’s brain for the full picture on rod bolts.

“Yes, you can quite reliably tighten rod bolts with a torque wrench, but this process assumes a few things. First, you must use the specified lubricant to give accurate torque readings. Second, you must use an accurate wrench; without insulting every home mechanic, not everyone has one. Third, a specific procedure in torquing the bolts is required with most rod bolts for an accurate job.”

Mostly, though, the strongest reason for using a rod-bolt stretch gauge is simple from an engineering standpoint. At a certain amount of bolt stretch, a given alloy and a given bolt diameter will create a specific preload. Assuming that a certain torque on the nut will create that preload is not always precise. It may be accurate enough, or even very accurate, but the bolt threads, the nut face, and the cap’s nut surface all affect this friction. Each bolt is designed to stretch a specific amount when it sees a certain loading. With a torque wrench, you’re only measuring the friction, not the preload.

Let’s dive into the three different methods of torquing rod bolts:

Torque Angle

Theory: Using an angle gauge attached to your torque wrench, you approach, say, 30 lbs-ft and then go 30 degrees clockwise past that 30–lbs-ft point. At 20 threads per inch, 30 more degrees will come close to the designed rod-bolt torque spec.

Reality: This procedure doesn’t take into consideration the compression, especially at the cap and rod parting line, and that the internal thread will give a little. If you want to hit the spec within 0.0005 inch, and you’ve got 0.001 fudge factor between the various compressions going on in the process, you’re already way off. This also varies from rod to rod. In real life, it doesn’t work out well.

Smith continues with the torque-angle problems:

“Another big variable involved in using torque angle is that most torque wrenches are not really accurate at 30 lbs-ft. Most torque wrenches’ base reading is 25. In the end, you’re better off using the proper lube and torquing the bolts three times, rather than doing the torque-angle method, but it’s still not as accurate as using a stretch gauge.”

Torque Wrench Alone

Theory: At a certain torque reading, the rod bolt will achieve the designed stretch and preload.

Reality:

Tightening the bolt once to the torque spec only overcomes the friction between the various surfaces and doesn’t preload the bolt sufficiently. A first and second tightening will polish the threads on the bolt, the nut, and the bearing surface between the nut and the cap. This also helps to ensure that the bolt is fully seated in the rod. So by a third torquing, everything’s polished and slicker than after the first; you’ve lost the friction that gave you the initial click (or reading) on your torque wrench.

Smith continues:

“We know that the third torque is the most accurate to create a certain amount of stretch. If you leave the bolt on the first torquing, you could be 0.001 inch off, whereas on the third you’re probably within 0.0002.

“After two years of research, we came up with a slicker lube with a lower friction coefficient than just about everything on the market,” Smith states.

“From rod No. 1 to rod No. 8, if you torque it three times to spec, you’ll be within 0.0005 inch of the exact torque spec 99 percent of the time. The torque required to tighten a rod bolt to a certain stretch is a designed-in factor reflecting the diameter and alloy of the bolt, but you cannot take away the fact that it’s indirect.”

Rod-Bolt Stretch Gauge

Theory: This measures the actual stretch of the rod bolt, giving the most accurate information on the bolt’s preload, since no other variables factor in the measurement.

Reality: This measures the actual stretch of the rod bolt, giving the most accurate information on the bolt’s preload. However, this gauge is yet another tool to buy, and it’s generally not cheap—around $100 to $120. And if you plan to build only one or two engines over the next few years, you may feel money is better spent elsewhere. Rod bolts have one of the few open-seat bolt holes in most engines. It’s therefore one of the few areas that actually allows a stretch gauge to be used.

“Once you get into a stretch gauge,” Smith cautions, “there’s one thing that must be checked. Even on a $100 gauge, you might get a light spring in it. A stretch gauge needs a 2.5- to 3-pound spring. You often have to buy one separately, because many gauges come with springs that have too little preload; they won’t hold its indicators tight in the rod bolt’s dimples. There’s also poor repeatability with the soft spring. You’d have to take apart the gauge to install the stiffer spring.”

People often use a ½-inch piece of tubing under the spring in a 1-inch-travel gauge so that the spring is even more compressed than normal. Also, the tips for the indicators should seat themselves reliably and repeatedly in the rod-bolt dimples. Very small radiused tips are more reliable than points in that respect.

If, over the course of several years, you plan to build two or three engines, a stretch gauge is a worthwhile investment, because it will help alleviate the most common bottom-end failure in all engines—improper rod-bolt tightening that leads to rod-bolt or cap failure.

COMMENTS

TO TOP