Thoughts on Torque & Tightening for Critical Fasteners

By WS Hyatt, TC 4926

I am concerned that lay D.I.Y. List  members who look up bolt torque specs in our Special Files and/or Tech articles such as http://www.mg-tabc.org/library/torque.htm can get led astray by thinking that any aftermarket fastener sizes listed can be torqued to the specs on chart(s) and thus it will be adequate for use per application in chart without regard to the grade of the fastener.  Fastener grades can vary widely and can be hard to identify. Torque figures mentioned in the Special Files only relate to original grade of OEM fastener(s) & not generic aftermarket fasteners which could be one of any number of unknown grades. The torque specs noted on charts such as above are not reliable unless original OEM 60-70 year old fasteners are still installed.

NOTE: THE DISCUSSION BELOW ONLY APPLIES TO CRITICAL FASTENERS ( e.g. big end and small rod ends, main bearing caps, head studs/bolts, flywheel/clutch bolts, rocker stand bolts, timing gear bolts, cam assembly bolts, internal diff. & tranny bolts, et. al.) not fasteners for compression joints with gaskets, excepting head fasteners. I think most would generally agree that proper fastening together of any critical components is a function of the “clamp” (preload) required to hold the items together beyond max. projected loads that are trying to separate the joints. This number was originally derived by qualified engineers from well established engineering principles, experience, and subtle intuitive nuances that engineers on the right hand side of the bell curve always seem to master.

Once the clamping loads required to keep parts together are empirically established, then the various mechanical aspects necessary to engineer such fasteners (material, diameter, thread pitch/ shape, thread engagement length, etc) can be determined with some reliability. It all boils down to creating a reliable fastener(s) that keeps the joint together while providing adequate predicted fatigue life protection. A tall order, that if not gotten right, will reflect poorly on the designer’s choices & end users wallet.

So to reiterate, the goal is to tighten/stretch the critical fastener(s) to the point where the residual stress (preload) in the fastener will always exceed the stress produced by the forces trying to separate the joint that is being clamped rigidly together. Keep in mind that fasteners exposed to high levels of cyclic stress can be expected to fail eventually, as applied cyclic loads approach the inherent tensile strength of loaded fastener(s). Over time, these fluctuating cyclic forces stretching & relaxing the fastener will cause the fastener to fail from fatigue. All of these are problems that the original design engineer must resolve.

Our problems center around confirming that any aftermarket fasteners used to replace NLS OEM original fasteners are up to the task. This is a very slippery slope without knowing the various engineering spec parameters of the original fasteners. Various MG links give original torque specs for factory OEM bolts/studs, but omit the fastener grade that torque figure relates to. Unfortunately some take these torque numbers as gospel for the fastener size noted w/o considering that any presently available “x” size generic aftermarket Whitworth or metric fasteners are unlikely to match the alloy/grade or tensile strength as the originally supplied OEM Factory spec fasteners, and may not be up to the task for the application. Torquing critical aftermarket fasteners of unknown alloy/strength per published XPAG OEM spec is fraught with danger! Just ask anyone who has broken  rocker arm stand bolt!

So, how to reasonably determine an appropriate preload clamp for a particular stud or fastener of unknown alloy or quality? Note that the preload/clamp determined per below is only for the particular fastener being tested, which quite likely will not be the same grade that the original design MG engineer speced for the fastener application.

Torque wrenches may give unreliable results mainly due to unknown and unaccounted for friction factors at male and female thread interfaces and under nut and bolt heads. These friction factors can result in high artificial torque readings that do not achieve target preload clamp, often shearing the bolt before the correct torque reading to give the desired preload is achieved. Ouch! Typically, if fasteners are assembled dry, up to 50 percent of applied torque is used to overcome friction between nut and bolt head bearing surfaces, and up to 40 percent additional friction can be expected from thread interface friction, leaving a marginal 10% torque available to preloading the bolt.  One can see that it would be easy to over torque the bolt & break it trying to achieve the target preload.  Measuring bolt stretch is a more appropriate means of achieving proper mfg. preload clamp. To mitigate friction and achieve required preload at the torque required, typically bolt threads and the surfaces under nut and bolt heads are lubed with the particular  assembly lube specified by fastener mfgr. Then they are tightened to specified preload by stretching the bolt a percentage of tension required to reach the yield strength of the fastener, insuring a known amount of preload clamp is accurately achieved.  (The amount of stretch of a new bolt over its new length is specified by bolt engineer for the application.) Fasteners are usually stretched 60%-80 +%  of Y.S., though some current OEM critical assemblies spec one time usage bolts stretched to nearly 100% of yield to achieve the max clamp potential of the fastener. This must be done with a very accurate tool as a fastener stretched to .0005” over Y.S.  must be considered failed and discarded. Due to consumer torque wrenches potential inaccuracy, but mainly due to unknown friction factors (w/o a lot of calculations and testing), DIY fasteners installation stretch percentages are typically set to 60%-80+% of yield for safety margin.

If critical fasteners are used, and the fastener suppliers’ recommendations for assembly lube are unavailable, give the fastener a fair chance of achieving its potential preload clamp at “x” torque and use a proprietary assembly lube. Tighten and release the fastener(s) in incremental steps for at least 10 cycles of tightening and releasing to reach full torque spec.  This will burnish thread interface and mitigate inherent friction so preload clamp is actually achieved at desired torque. Tighten a critical fastener dry at your own risk! It is most important to use the fastener mfgr’s. explicit directions  regarding assembly lube to achieve “x” desired preload at torque they specify. If this information is unavailable, ARP Ultra-Lube can be used,  so the full preload stretch target can be achieved within 5% with 1-2 tensioning cycles/pulls to full torque. http://arp-bolts.com/p/arpultratorque.php

A problem arises with studs and bolts in blind holes that cannot have their length elongation easily measured with a conventional bolt stretch gage when tightened,  to insure stretch required to a specific preload clamp is achieved.

However:
1.) Studs and bolts can be procured of known alloy, strength, and stretch/torque  spec. to achieve a particular target preload. I.e. fastener supplier will provide  a specific stretch figure in thousands  of an inch that will achieve desired preload clamp with assembly they specify at a particular torque reading.
2.) Preload washers are available (from SPS for one, very expensive) that will indicate when a specific stress or clamp is achieved. They are not re-usable.
3.) Studs can be axially drilled through the fastener neutral axis  so stretch can be measured with a depth gage. Not too practical, but a real solution
4.) For studs use a dial indicator to top of stud to measure stretch as nut is tightened. Use this method if stretch info is supplied with fastener. Much easier.

Additionally,  preload potential for unknown fasteners can actually can be measured with a reasonable degree of accuracy as bolts/fasteners stretch in proportion to tension loads in them.
All that is necessary is to:
1.) Make up a bolt test fixture of a block of steel deep enough to accept an unknown suspect bolt or stud and its nut(s).
2.) Drill a hole through the block of correct size to pass test fastener through
3.) Lubricate threads and underside of bolt head and nuts (using hardened washers) with appropriate thread lube e.g. ARP “Ultra-Torque”
4.) Measure the free length of the new stud/bolt to nearest .0005” and record.
5.) Torque the fastener until the Y.S of the fastener is achieved, ie until it stretches  beyond  its elastic range and into the plastic range, so when relaxed on release it has taken a permanent set and does not return to  original measurement i.e.  now measures +  .0005”-.001” over new bolt measurement previously recorded.  This fastener is now considered failed, and scrap.
6.) Note the torque reading required to induce yield failure. Take 75-80%  of this torque reading, and use that number torque reading for final assembly of fixtures using studs or bolts into blind holes to reasonably assure that appropriate preload  clamp is achieved. Thus one can determine a reasonable target  preload for that particular fastener in its particular alloy. Keep in mind that it is unlikely to match the MG factory preload spec and will only
achieve a reasonable preload clamp for that particular fastener.

Quality bolts have marks on their heads identifying their strength grade. (unless counterfeit, know your sources!)  (Click HERE for fastener grade chart) SAE bolts  range from unmarked grade 2, junk, grade 5-8 likely junk, grade 8 suspect if from unknown source.  Higher grades include AN, MS, & NAS bolts. Whitworth bolts are graded with letters,  see  HERE, or  HERE for  ISO metric bolts. ISO bolts are usefully  marked with both Minimum, Ultimate Tensile (UTS)  & minimum Yield Strength (YS)  grade markings on heads. These range from ISO 8.8 to ISO 12.9. The # to the left of the decimal is the UTS in Newtons per sq. mm. The number to the right of the decimal is the percentage  YS. is of  UTS.  (1 Newton per sq mm = 14,500 psi MOL)   So an ISO metric bolt with a 8.8 grade would have an UTS of 116,000 psi & a YS of 92,800 psi. An ISO 12.9 bolt would have an UTS of 174,000 psi and a YS of 156,600. Useful, as the Y.S. of the fastener in now known. Taking 75- 80% of the ISO 12.9 bolt YS would result in a fastener that could achieve known a preload clamp of 117,450 psi -125,280 respectively. The more serious question is whether or not this preload matches the  figure speced by the MG engineer for the original factory fastener. Using directions 1-6 will only determine a known preload that the fastener being checked can safely achieve, but not the original load speced by the MG factory engineer for the application!

Unknown or suspect bolts can be easily checked to a certain extent  by:

1.)  Spark test on grinding wheel. Harder, high tensile metals will throw a different spark than low strength soft metal fasteners. (Click HERE for Google image search or HERE for Wikipedia page images)
2.)  Garden variety SAE bolts typically have shanks that are .002”-.004” undersize, AN & MS bolts range from .001”-.0015” undersize, high quality NAS bolts are typically .0008”- .0015” undersize. Thus a graded NAS bolt with a .002”-.004”  undersize    shank  should be considered suspect and likely counterfeit. Buy your fasteners from a reliable source.
3.) High quality fasteners can be further identified, as they have an obvious rolled fillet radius where shank & underside of bolt head meet. Right angle corners are stress concentration points.

For further info see the SPS Technologies  http://www.spstech.com/home/  &  ARP  web sites. http://www.arp-bolts.com/

WSH