Saturday, August 1, 2009
Bolts: Shear and Tension, not Torque
Bolts are bolts. Anti-seize goes on before every nut. I don’t need to use a torque wrench; after many years doing this, I can feel when it’s tight enough. I always use a torque wrench, so I know every bolt is tensioned properly. Just flip the washer over if it has an indentation in it. Aircraft are all way overbuilt; everything has redundancy, too.
All those statements are wrong, and those assemblies that depend on nuts, bolts, and threaded studs need to be accorded proper respect.
Bolts keep things in place, whether they are used in shear (where the bolt stops the movement of something 90 degrees to itself) or in tension (where the bolt or stud holds things together with its head or a nut). Structural bolts display certain characteristics — they are stronger in shear than in tension; they are strongest when properly tensioned; they need to match the corresponding threads; and their sizes are governed by their strength requirements.
Specialty bolts, for instance banjo bolts, may be sized to satisfy other requirements. This article will focus on structural, tension assembly applications.
Wear and Tear
Bolts, nuts and washers, like all mechanical devices, wear out. Washers wear out with a single application of clamping force. These are the softest parts of the assembly, and their deformation is a calculated part of proper assembly. Never reuse washers in critical applications. Nuts, as well, are designed to wear faster than bolts. Though most of us will reuse nuts as long as some lock action seems to be working, nuts are not as strong in their second and subsequent uses as in their first. Bolts are meant to last the longest and can be re-used, as long as they do not exceed certain mechanical and environmental parameters. Bolts that have experienced overheating, over-torque, or that have been subjected to bending, have had their threads bottomed, or have visible damage, should be replaced out of hand.
Torque Ratings: Overrated?
In an effort to get the proper clamping force (tension) on a bolt, mechanics have relied on torque wrenches — from simple torsion bar units through more-sophisticated spring-loaded "click" units, to today’s electronic digital wrenches — the most accurate and precise tools generally available. Duane Vallejos, torque training coordinator for Snap-On Tools, says that even if your torque wrench is operating properly, the tension in the assembly could be far off the mark. It’s not the tool; it’s you!
"Inertia plays a big part in how a torque wrench clicks," he says, "and we’ve found that the average tech over-torques a fastener by 20 to 30 percent." The higher torque ratings tend to have the bigger errors.
The digital torque wrenches greatly reduce, but don’t totally eliminate that operator error. "And merely holding the torque wrench incorrectly — or when an unusual angle forces you to use an unfamiliar grip — that can throw you off, a lot."
Experienced mechanics all "know" they can feel the right tension, and experienced mechanics are nearly all wrong, nearly every time, Vallejos says. "Friction and lubrication are so often overlooked," he adds. To standardize friction, he recommends chasing the threads for every assembly, for both new and used bolts and nuts. This includes chasing the female threads in the engine case before reinstalling hold-down studs.
Many fastener manufacturers now also recommend a "torque-turn" technique, as used for years on new sparkplug installation, oil filter installation, or when setting up hydraulic valve lifters. Use a torque wrench to some set torque, say, 30 lbs-ft; then add 60 degrees of turn. Of course, each assembly, and each application, has its own spec.
Unlike GA aircraft inner tubes, for instance, one size isn’t universal in application. Don’t substitute fasteners. A harder, "stronger" bolt may be a disadvantage, for several reasons. It may transfer undue stresses to other parts. If used with mismatched nuts and washers, it may not provide the correct tension.
Had the torque continued to increase, it would have snapped, and extraction would have been a problem.
It may be more susceptible to shaking. There may be metallurgical problems like anodic reactions or damage to the clamped parts, and the physical size or thread length, etc, may prove problematic in assembly or subsequent clearance.
Never use rusted or otherwise-damaged fasteners, even if they’re new. Scratches help concentrate stresses; rust is insidious.
Avoid oil or anti-seize unless it’s part of the assembly specification on threads or contact surfaces, as this drastically changes the clamping force exerted by any given amount of torque! Vallejos notes that "only about 10% of the torque applied to an assembly translates into clamping force," so reducing the friction allows more of the applied torque to translate into greater force, over-stressing the fastener and possibly compromising the assembly.
Common Sense Practices
Use your head. "Checking the torque" isn’t the same as "tightening it some more." Looking at a bolt head isn’t "inspecting" it. Once when changing a tailwheel assembly, the bolt that held the leaf spring was invisible inside the fuselage tubing. When applying torque to remove the bolt, the head snapped off. The bolt was completely rusted through, inside the tubing.
Every quality system lets through a bad part once in a while, and counterfeiters often do a good job of producing "look-alike" inferior parts. When crooked manufacturers and their government officials cooperate, "certified" bad parts can end up in our hardware supply.
Years ago, the writer was torquing the main bearing caps on a racecar engine, using new, very expensive bolts from a reputable house. The torque went up to 30, then 45, then 60, on the way to full torque. However, one bolt was turning more than the others. After stopping and backing the bolt out, it was soft, and had stretched.
Had the torque continued to increase, it would have snapped, and extraction would have been a problem. Worse, if the rated torque had been a lower number, say 45 lbs-ft, the soft bolt would have remained in the engine and the crankshaft would soon have destroyed the engine. That’s expensive on a racecar — and it’s real trouble in an aircraft!
Spindle/Axle Nut Life Cycle Service Deficiencies
Wheel, axle, and bearing loads all cause wear due to many factors, including, smooth pavement as a secondary taxi. Aircraft ground operations, taxi, tow, takeoff and landing all impose wear. While the aircraft is in the taxiing and takeoff mode, a static load dominates it. Whereas, while the aircraft is in takeoff and landing mode, it is dominated by impact and dynamic loads. The bearing, axle, tire and break wear is directly related to a key variable: endplay. Endplay is caused by axle nut preload torque settings. Even on fresh, smooth pavement, the off-optimum torque, inherent in the conventional axle nut, causes wheel chatter, which transmits adverse loads to the bearings and axles. The conventional axle nut design is locked in place by a castle nut and cotter pin with inadequate indexing. The cotter pin can only be pushed into one of six holes in the castle nut that is aligned with the keyway or pinhole in the axle (60 degrees apart on the circumference). This causes the castle nut preload torque to become non-optimum if it needs to be overly tightened or overly loosened (as in most cases), for the cotter pin to be inserted.
Excessive bearing loads result due to two cases. The first, according to Leslie Weinstein, CEO of True-Lock, is excessive endplay. The worst case in axle nut looseness comes when the castle nut is backed off after establishing bearing preload by the use of a torque wrench by almost a full 60 degrees to ensure cotter pin insertion alignment. For standard 16-pitch threaded axles, this results in 0.010+ of an inch lateral slack on the spindle axis. The result is excessive bearing wear, as well as wheel chatter or axial floating. Scored axles from the bearing rotating on the axle, damage to the bearings and premature brake and tire wear are a result of excessive wheel endplay. Bearing manufacturers specifies an endplay of 0.001 to 0.005 thousands.
The second case is excessive preload. The ideal bearing to wheel fit should have a gap equal to the film thickness of the lubricant. Hence, over-tightening the axle nut by 30 degrees would cause the bearing to be too tight allowing them to overheat from their lubricant film extruding out of the gap. This will rapidly cause further damage and a chain reaction of damage to further components.
A solution by the company True-Lock employs an infinite variable locking mechanism that can be set to the optimal torque according to Weinstein. "The True-Lock system eliminates the conventional axle nut systems’ inherent un-optimum torque setting characteristics imposed by the use of the castle nut and cotter pin. This not only provides a smoother running wheel bearing, but also a controlled preload torque specification for maintenance manuals and reduces parts count," says Weinstein. The makers of this system say it reduces maintenance inspection time, extends the life of high load and stress mechanical parts, and saves maintenance costs.