Click on a thumbnail to see the full-size image
Nuts About Bolts
By Brendan Baker
The late Smokey Yunick, Hall of Fame race car mechanic, engine builder and car owner, once said that fastener technology used to be "cut and try." In those days if there was ever a failure the philosophy was, "go one size bigger." In today’s world of high performance engine building, luckily, this is no longer the case.
Ol’ Smokey employed a fair share of black magic when it came to building engines. His innovations were inspirational to a whole new breed of speed doctors who followed in his path. And these new doctors of racing and high performance have since turned art into science, more or less. We were fortunate enough to speak to a few of Smokey’s disciples about the finer points of sealing a high performance engine.
There have been several books dedicated to fastener technology alone. One such book, Nuts, Bolts and Fasteners, by Carroll Smith, is practically a bible among racing mechanics and engineers. Smith has had great successes as a crew chief/team manager on some of racing’s greatest teams, including Carroll Shelby’s Le Mans winning Ford Cobra Team. We caught up with Smith one afternoon while he was at his horse ranch in California to hear what he had to say about anything – but fasteners was the topic of the conversation.
Smith, who is also a consultant for ARP in Santa Paula, CA, is a firm believer in head studs as opposed to bolts. "Studs allow for more even distribution of stress on the engaged threads of the female end. Nobody uses bolts in performance engines," Smith affirms. "The joint between the cylinder head and block is totally dependent upon the clamping load of the bolts, which depends on: A) their strength and B) their preload. This joint is not subject to fatigue, unlike connecting rod bolts. And, even though it’s faster in production engines to use bolts, head studs should be used instead because you get more even distribution of stress on the engaged threads in the female end."
It boils down to a question of preload or torque – "that’s where the problems come into play," says Smith. Connecting rods use the stretch method to measure strength, but it’s not as simple when applied to head studs/bolts. "It can be done, but it’s very difficult to do.
"Basically you’re dependent on your faithful old torque wrench, which has to be recalibrated quite frequently," he says. "What you find, however, is that the actual amount of stretch per unit of torque is very much dependent upon how clean the threads are."
"There is definitely a relationship between the torque and the preload," says Ron Hukari, a racing engineer with SPS Technologies, an aerospace fastener manufacturer in Jenkintown, PA. "The preload is the force the bolt clamps the joint together with. Preload holds the engine together. Torque, however, is just the mechanism to get the preload."
According to Hukari, engine builders rely on the torque value too heavily to achieve the preload. "Even NASCAR people aren’t measuring preload directly," he says. "There is no way to easily measure preload on a head bolt/stud, at least one that’s inexpensive. The most popular way to measure it is with ultrasonics. But not even Formula One teams, and their huge budgets, go completely to that extreme," admits Hukari. Some Formula One teams have experimented with ultrasonics in test engines, and even then, they only measure one head stud. These ultrasonic devices use sound waves to measure the thickness or stretch of a fastener.
"With connecting rod bolts, you can use a micrometer to measure the elongation between both ends of the bolt to see how much it stretches. Consequently, this gives you a better picture of the preload," says Hukari.
Chris Brown, director of specialty products for ARP, says there is definitely a difference in sealing requirements with a high-performance engine. "Obviously the difference is you’re asking for more performance than that of a stock engine, so you’ll have much more load on the cylinder head."
According to Brown, this load is due to the amount of pressure each cylinder pumps out through the head with every revolution of the crankshaft. The amount of force required to seal the head gasket is called the clamp load. And each head bolt must hold a certain amount of force to maintain the joint between the head and the block.
With a performance engine, the clamping loads are higher because the pressures are higher. And with the higher pressures, you’ll need to use stronger fasteners than what’s provided by the manufacturers. According to Brown, it’s not uncommon in racing and performance applications for people to use high compression ratios and forced induction systems such as superchargers and turbochargers. And many of these configurations can run up to 40 lbs. of boost, which is an extreme amount of pressure on the cylinder head and block.
To this end, Brown and legendary engine builder Joe Mondello of Mondello Technical School in Paso Robles, CA, agree that it’s best to use a torque plate when honing the block because higher stress loads will affect the cylinder bore, depending on the mass around the bolt holes. And when honing with the torque plate, they both recommend using the same fasteners that will be used in the final installation because this allows you to better duplicate the final results.
"Don’t go by what is in the ‘book’ (repair manuals) for clamping," Brown explains. "Who knows where those torque specs came from?" If you’re using specialty fasteners that have much higher strength than OE fasteners, then these torque specs will have no relevance. "These specs could have been lifted from an OEM bulletin somewhere," warns Brown. "And you’ll find out pretty quickly then, if you snap a bolt or stud, that it was wrong."
Most specialty bolt manufacturers provide detailed instructions with their kits explaining how to install, and torque their bolts, as well as what type of lubrication to use. Some even include charts showing how much clamping load to use.
Very few engine builders know how much clamping load they need, says Brown. In many cases, they’ll measure the clamp load by using torque values, and this, in turn, creates some confusion. “Say you torque a 3/16˝ bolt; you can torque it to 100 ft.lbs. if you don’t put any lubricant on it because it generates so much friction. But, if you put a dab of moly lube on the threads and under the head, you may only be able to torque it to 30 ft.lbs.," Brown explains.
Some of the confusion surrounding torque and preload comes from not knowing the difference between them. Torque is just the twisting force. It’s an index number and does not equal load. You also have to compensate for the friction coefficient of the oil. Therefore, torque values should be accompanied by the specified lubricant for the most consistent results.
Carroll Smith agrees that friction plays a vital role in achieving proper stretch and clamp load. And if you use a different lubricant than the one provided by the bolt manufacturer, Smith says, "good luck." It’s critical to use the correct lube because it was designed for the specific bolt and application in many cases, especially in high performance applications. Many high performance bolt manufacturers include the lube with the bolt kit when you buy them.
"Obviously, the more slippery the lube is, for the same amount of torque you’ll stretch the bolt further, and vice versa," says Smith. "Proper lubrication of head bolts doesn’t really matter as much in a vintage 289 Ford – but if you’re assembling a true high performance engine, you’ve got some real pressure inside of those cylinders.
"If you’re running a 13:1 compression engine at 9,000 rpms, you’re trying to lift those heads right out of there. If you’re running a 9:1 engine at 4,000 rpms, you can practically hold it on with nails! A little difference is huge; the devil is in the details," Smith quips.
"Thirty weight oil (a common bolt lubricant used in many rebuilding facilities) is a little more inconsistent because it goes away fairly quickly under extreme pressure applications, like high performance engines are,” says ARP’s Brown. A 7/16˝ head bolt can, for example, generate a clamping load around 14,000 lbs. – a significant amount of stress.
SPS Technologies’ Hukari, says, "with sealing, a very important thing for guys who aren’t necessarily building engines for NASCAR or Formula One, is getting a good, consistent product and installing it correctly. Metallurgically the yield strength of the bolt needs to be consistent. If you have a problem with heat treating and don’t get enough yield strength, when you go to tighten to a certain torque, you’ll find the bolt doesn’t have the preload you think it does and you’ll have a leak.
There are several things that affect clamp load, according to George Lorimer, a former engineer with GM’s Powertrain Fastener Lab. In an article by Lorimer posted on www.boltscience.com, he states that one of the problems with clamp load is: "the second you take your wrench off the tightened head, it starts to decrease." Furthermore he claims that head bolt clamp loads can vary up to 40 percent when torque control is used. This "relaxation" effect can vary from bolt-to-bolt, thus the need for a proper tightening strategy.
Some of the factors that Lorimer says influences the relaxation effect are: Surface finish, Temperature - high and low, Type of joint: Gasketed (soft), Material of joint (hard or soft), Combination of metals (steel/iron vs. aluminum), Initial preload, Joint strength (spring rate), Bearing area under bolt head, Vibration, External loads to the joint.
Decreasing the variation of clamp load requires a good tightening strategy. One way that has proven effective is to "cycle" the fastener. Torque cycling bolts can reduce the relaxation effect by up to 50 percent or more.
DO THE TIGHTEN UP DANCE
Torque cycling – in essence, "breaking in" the stud or bolt – helps "ease" the threads. At the microscopic level, fasteners have peaks and valleys in the threads. Thus, when torquing, you have to overcome a certain amount of friction first.
According to the experts, new fasteners have a higher friction level because they have not been burnished. New bolts need to be seasoned with the threads it goes into. After this process, the grain will compress at the surface, which reduces friction. The proper torque cycling procedure is to tighten and loosen about five or six times, but only to 50 percent of the final torque value each time.
"You stabilize the friction coefficient after repeated tightenings," explains Brown. "This, correspondingly, increases the clamp load because the bolt or stud will stretch more. You’re overcoming less friction. More of your torque value is going into stretching the fastener, less into overcoming friction," ARP’s Brown explains.
Joe Mondello also points out a critical but often overlooked point: "Once you torque cycle a fastener, the nuts and washers must go back to the same stud because they will be married to that particular bolt or stud. When you go ahead and torque them down after following the recommended cycling procedures, you’ll be at about the fifth cycle, and you’ll be at the maximum clamp load and maximum distribution of the bolt or stud."
Torque-to-yield bolts in OE applications are usually only used in the most critical areas like heads, rods and mains, say these experts. For the manufacturers it comes down to a cost savings issue, but that’s not the case with racers because their cost is on a smaller scale. It is a way of wringing out every last bit of clamping load out of a relatively low strength bolt.
Torque-to-angle is a better method of tightening – it’s more accurate than straight torque. With straight torque, friction impedes consistent clamp load values. So you have to overcome a certain amount of friction in order to achieve load. "Friction is very hard to predict, too, even with special lubricants to try and control it," says Brown.
Brown believes that the inconsistency of straight torque is why torque-to-angle is a better method. "With torque-to-angle, you typically have very low torque values and then a rotation through a certain angle based on the helix or thread pitch to achieve a specific amount of load," he explains.
"With our fasteners, because the strength is much higher, the idea is that we select material that will generate the correct amount of clamping load at 75 percent of the yield strength of the fastener. With torque-to-yield you’re stretching it to the point just before it breaks." The advantage of very high strength specialty fasteners is that they reduce the chance of failure by not being as close to the limits of the material, say both Brown and Smith.
"With fasteners it’s a question of knowing just a little bit," says Carroll Smith. "You don’t have to know metallurgy, but you need to know the facts." Although it wouldn’t hurt to know the metallurgy, the important thing is to know how to get the best, most consistent clamp load.
High compression race engines, and engines with turbos and blowers, are dealing with much higher lift-off forces than production engines. And with the most common cause of head gasket failure being due to incorrect or uneven clamp load, the head gasket seal is even more critical in high performance applications.
Today high technology engines and years of metallurgy research and development have made facts easier to come by than they were back in Smokey Yunick’s day. Back then racing and performance engines were more of an art than it is today. Things like clamp load and toque-to-yield were not even thought of yet. But if anyone did think about it, you can bet that it was ol’ Smokey.
This article is dedicated to the memory of Smokey Yunick, owner of The Best Damn Garage in Town.
– Henry "Smokey" Yunick, May 25, 1923 - May 09, 2001.
- Nuts, Bolts and Fasteners, Carroll Smith ISBN: 0879384069
- ARP, Automotive Racing Product’s 2003 Catalog & Tech Guide.
- Threaded Fasteners, by Bill McKnight, February 2001, Automotive Rebuilder, can be download from Engine Builder Magazine’s Web site, www.babcox.com/editorial/ar/ar20134.htm).
To better understand the importance of clamp load we refer you to an article in February 2001 Engine Builder (formerly Automotive Rebuilder), by Bill McKnight, director of training for Clevite Engine Parts. It illustrates the forces on the head joint and the strength required of today’s torque-to-yield (TTY), stretch bolts. McKnight says that the clamp load required to seal an engine is equal to three times the lift-off force.
An example of lift-off force for a 4.250˝ bore race motor with 1,400 psi firing pressure is 19,861 lbs. This means that 9,930 lbs. of force is needed per bolt in a 6-bolt pattern (19.861 x 3 divided by 6). This is the initial load needed on each head bolt to seal the head gasket. There is a bit more to it than this, but you get the general idea, says McKnight.
Here are some tips from Fel-Pro’s Chief Engineer, Jerry Rosenquist (Spring ’03: NASCAR Performance magazine), to help ensure proper clamp load of the head and block seal:
- Make sure head bolts are straight, clean and threads are undamaged.
- Remove black oxide coating on threads and washers. Use a wire brush on threads and a fine sandpaper on the washers. This coating, found on many high performance fasteners, has a high coefficient of friction.
- Clean all holes and remove any debris. Run a bottoming tap in bolt holes that are dirty or damaged.
- Always use the manufacturer recommended lube with the fasteners. Some moly lubes are more slippery than others and affect torque readings as much as 15 percent. Also, lube bolt heads and washers with the same lube.
- Make sure that the washers are positioned correctly (rounded or chamfered side up), and check that there is no debris under them.
- Use the manufacturer recommended torque sequence.
- Torque cycle bolts in incremental steps using a smooth motion. Herky, jerky movements can give false torque readings. Double check final torque readings, too.