Rod bolt tightening

[HJG]

The initial 20ft lbs is to seat the joint, the following angle is to tighten the fastener to the required stretch/pre load while eliminating inconsistencies from friction.

100% this.

Usually you want the torque to be as low as possible, but crucially enough to seat the joint.
Some trial and error in the freedom of OE development world makes it easier to determine.
One random real world example on a 14,9 M8 fastener is 12Nm 'snug', followed by 65deg.

[David Redszus]

So, how do you calculate angle (if not given) for head and main bolts?

Build a fixture that will allow your bolt to be tightened while able to measure under heead to nut length.

As a reasonable approximation, use the following formula:

Torque angle (deg) = bolt length (mm) * 0.002 * 360/pitch (mm)

Example:
Angle = 100mm * 0.002 * 360/1.5mm = 48 deg

Formula does not consider crush of washer, clamped surface or nut deflection.

[Rick!]

Bolt tensioning using torque is almost useless. Unless the components have been tested and verified regarding tensile,lubrication, and friction. Almost never possible with aftermarket components.

Angle torque or stretch are the only methods that should be considered in critical areas such as an engine.

So, how do you calculate angle (if not given) for head and main bolts?

Set up a dummy assembly so you can measure the bolt length, did it with crownwheel bolts after a loosening problem.

Ever seen the hollow diesel head bolts, I think they have a pin inside to measure to the head, either way it's measurable then ?

Atlas Copco DC Torque Wrench with their controller and software.
Torque to failure a statistically significant amount of real or simulated joints with the target fasteners, graph out each test, identify the initial torque value and additional angle, retest again using the same amount of joints with the torque-angle recipe and measure stretch physically or with an ultrasonic tester, adjust as needed and call it golden. Easy peasy.

[frnkeore]

Angle = 100mm * 0.002 * 360/1.5mm = 48 deg

I have a question, what part does yield or tensile strength play in the above formula?

[amc fan]

When you tighten a rod bolt to get the desired stretch how high of a torque reading do you get to achieve proper stretch? Does it ever go to 60-70-80ft.lbs. or more using the recommended lube?

[Tom68]

When you tighten a rod bolt to get the desired stretch how high of a torque reading do you get to achieve proper stretch? Does it ever go to 60-70-80ft.lbs. or more using the recommended lube?

As I posted earlier I've been up around 70 ft.lbs with 3/8 Carr bolts and arp lube, but I doubt you'd get to 50 with any lesser materials. Maybe 50 lbs if you didn't lube a standard bolt.

[Rick!]

When you tighten a rod bolt to get the desired stretch how high of a torque reading do you get to achieve proper stretch? Does it ever go to 60-70-80ft.lbs. or more using the recommended lube?

Calculating Stretch - note, stretch is directly proportional to clamped length.
Calculating Torque angle to achieve bolt clamp force.
Torque required to achieve the clamp force.

Since aftermarket rod bolts don't really fit on the "Grade" scale and are made out of ridiculously high strength alloys, one can deduce that the bolt's Young's Modulus, E, increases a bit and a higher torque is required to create the desired stretch. That is one of the reasons that high zoot rod bolts are "necked down" to promote a controlled (engineered) stretch. Stretch is the quantity that maintains clamp force during the mechanical events and environmental (thermal) changes, the usual joint relaxation, and for ever so slight embedment. Relaxation and embedment usually are due to excessive compressive stress under a bolt head or washer caused by high clamp forces.

[digger]

Bolt tensioning using torque is almost useless. Unless the components have been tested and verified regarding tensile,lubrication, and friction. Almost never possible with aftermarket components.

Angle torque or stretch are the only methods that should be considered in critical areas such as an engine.

Nonsense, millions of bolts in an engine have been successfully tightened using nothing but a torque wrench without issue.

[Tom68]

Angle torque or stretch are the only methods that should be considered in critical areas such as an F1 etc engine.

Nonsense, millions of bolts in an engine have been successfully tightened using nothing but a torque wrench without issue.

Fixed.

[David Redszus]

Bolt tensioning using torque is almost useless. Unless the components have been tested and verified regarding tensile,lubrication, and friction. Almost never possible with aftermarket components.

Angle torque or stretch are the only methods that should be considered in critical areas such as an engine.

Nonsense, millions of bolts in an engine have been successfully tightened using nothing but a torque wrench without issue.

Yes they have, back in the old days when fasterner technology was much less advanced. Back then good mechanics could even "feel" the torque limit without using a torque wrench. They also ran carbs and points. So what? No more wire wheels. No more vacuum wipers. No more hand crank starters.
Most savy techs have left the past behind...where it belongs.

Since friction torque is exactly what it sounds like, and friction can not easily be determined, bolts were often larger than they needed to be to accept the inaccurate friction torque methods.

I have a question, what part does yield or tensile strength play in the above formula?

Tensile is not a factor used in the angle torque method. Neither is friction of the threads, washer, nut or compression of clamped materials. There is little or no correlation between clamping force and friction torque. Unless, accurate testing is performed using the exact bolt materials, sizes, finishes, lubrication to determine the friction coefficient.

Besides the uncertainty of frictional surfaces, the other significant objective is to determine how much clamping force will be needed for a specific joint load. Then a bolt can be selected that will produce the necessary clamping force at near the tensile limit. Now it becomes a simple matter to determine how much bolt stretch will be required and lastly, at which turn angle the required bolt stretch will be reached.

[Tom68]

No more hand crank starters.

Don't dis the crank, can get you up a sand dune when the trucks broken down.

Crank.jpg

[hoodeng]

Into the fifties even sixties in some places, in smaller workshops the tension wrench was kept in the Foreman’s office only to be bought out on specific occasions, if a worker asked for the tool he would be queried on its application, if the boss thought the necessity trivial it stayed in the box and the worker would be told to get on with it. Even in the early seventies the inch pound wrench was kept in the Foreman’s office strictly, only to come out for trans bands.

As manufacturers were becoming more aware of their products miss service in the field they created specification lists in more than just part fitment and maintenance, the tension wrench now moved from the foreman’s office to the tool store for everyone that required it to have it handy, this started the habit of fitters getting their own wrenches hence better outcomes on service.

In early service books we were told to tighten crankpin nuts very tight and pinion nuts just tight, now they had a specification, identifying two types of fitters, one that thought the specification was nowhere near tight enough, and the guy that thought he was going to strip it if he went any further.[Another thing, I have never understood the wriggle method when checking valve clearances].

Now we are spoilt, anyone can buy at a reasonable price a wrench that will do all torque specs and degrees all in one, there are no more excuses to not set fasteners to what the manufacturer specifies. I know the next line will upset some, but, if the manufacturer is satisfied the application of the fastener will be serviced adequately with a torque spec, do just that. If he has given stretch specs, do that. If he has given a degree setting, do that…He will be satisfied with what you have done.

When I was an apprentice, I recall a couple of guys in the heavy diesel shop rotating a bar around a dog on main nuts in a loco engine with another guy watching the dial indicator. I also used to do Deutz BF6L413FCR engines in an earlier life, the apprentice got dizzy when it was head bolt time, the spec was, pre torque 30ftlb +45°+45°+45°+30° these things would sing if you tapped them after setting.

[David Redszus]

I have a question, what part does yield or tensile strength play in the above formula?

Perhaps we jumped a bit ahead of ourselves. Let's back up a little.

How much clamping force do we need to keep a joint from coming apart?
More than the forces that are trying to take it apart.

Bolts come in various grades which determine their strength; called yield load.
SAE 2 = 43500 psi
SAE 5 = 95700 psi
SAE 8 = 136300 psi

We are really only interested in the higher grade bolts SAE 8.

The maximum clamping force that can be exterted by a bolt is determined by the yield x bolt area.
The bolt area will depend on whether it is fully threaded, partially threaded or has a reduced shank.
Assume a 3/8 -24 SAE grade 8 with a shank diameter of 3/8".

Max clamping force = yield (136,300 psi) * bolt area (0.110 in2) = 14,993 lbs

If that clamping force is too small, a larger bolt should be used.
If the clamping force is too great for the the material being used (avoid crush), a
smaller or lower grade bolt can be used.

The bolt should be torqued to the yield limit just at the point of leaving the elastic range and entering the plastic range. A very close approximation to yield point can be determined by bolt stretch equal to about 0.2% or 0.002 times bolt length.

Bolt length, diameter and thread pitch are then used to determine angle torque.
Angle = Length x .002 x 360/Pitch (in inches)

For a 4 inch bolt. 3/8-24 we have:
4.0" x .002 x 360/0.0419" = 68.7 degrees.

This angle torque is measured from the pre-load torque point, where all the slack and clearances have been taken up. This should not include compressing a head gasket.

The advantage of the angle torque system is that it does not require the knowledge of thread, washer, head or nut friction, or of surface finish and lubrication. Friction does not matter. But, it does require the correct yield point force, and correct bolt diameter. A fully threaded bolt is quite different than a reduced shank bolt.

With a little bit of work, an excel spreadsheet can be constructed to accomodate a range of bolt grades, diameters, lengths, and thread pitches. To confirm clamping pressure, a product called FujiFilm can be used.

[Rick!]

Clamped length is the important dimension as opposed to bolt length.
If I bolt together two pieces of 16 gauge steel with a 4" bolt, the joint is going to vibrate apart or thermal cycling will loosen it quickly due to a lack of bolt stretch as a result of the 1/8" clamped length.

Stretch, torque, angle.

Normally measure stretch, with stock unmachined end bolts I use torque.

Got an engine with unmachined ends on the bolts and it calls for angle after 20 lb/ft.

Now this will apply to torqueing them as well. How do they know if the bolt is tight, if you change bearing brands or worse the bearing packager changes supplier, you don't know whether your bolt is tight or whether you used up most of that energy just crushing the bearings into place. Especially doing it with angles starting from such a low reading.

I think I'll angle tighten them, mark them, back them off and see what torque value they're at when they get back to the mark.

OEMs or Aftermarket Rod/bolt suppliers cannot possibly accommodate accommodate variability due to the myriad bearing suppliers' products.
If the bearing mfr makes bearings that don't have the right eccentricity and don't measure a greater clearance at the parting line, that's on the builder to identify that and make the necessary corrections. If torquing the rod bolts to "spec" cause the mating surfaces to deform, then the rod material is suspect.
The solution to those issues is to use known good, properly sized rods and known good quality bearings and measure everything before final installation.
Measuring the loosening breakaway torque is not a very accurate way to determine what value the fastener was "torqued" to.
Again, I refer to the Atlas Copco method and measure tightening breakaway torque; tightening breakaway torque measures the spike in torque and friction to start the fastener rotating and then the data will show the valley after breakaway and the true static torque the fastener had.
This is fundamentally impossible using ordinary hand torque tools.
Keep in mind that there is a finite amount of "torquings" one can apply to one fastener set before it's time to renew them.

[Dan Timberlake]

For my money Getting the absolute most out of every fastener is not time or money well spent.
Not every application requires torque-to-yield.
My boss said last week that the (aluminum) torx fasteners that hold on his BMW's oil pan are torque to yield. In my head I said WTF!!!! .

For decades HoloKrome and the boys included generic torque specs for their SHC screws based on something like this -
"These tightening torque values are 75% of the torque required to yield the screw, and apply only for the (lubrication, etc) conditions listed below."

Many if not most engineering calculations are based on information or estimates that are very possibly and even of necessity no "better" than +/- 10%.
I believe 'most every formally designed product will have a calculated FOS ( Factor Of Safety) . Some standards for gantries, lifting eyes and airplanes include a minimum FOS. Those factors are not 1.1 or 10%. They are values like 1.5, 2, or even 3.

We require certs for most ferrous materials we buy. It is pretty common, even typical for the ultimate tensile strength, yield strength, and my best friend elongation to exceed ASTM specifications significantly.

Decades ago the late great Joe Mondello described a "torque cycling" procedure for best results tightening rod and main bearing bolts and head bolts etc.
LO! and Behold! in the modern era ARP apparently found and then quantified something similar.
https://arp-bolts.com/i/t/PreloadGraph.gif

Someday I''d like to see if the first angle torque knocks down the asperities to a safe, mechanically stable condition while it is successfully ignoring surface friction.
If starting With .006" or so bolt stretch, on the 14th lap the loss of a few .0001 inches stretch to creep/collapse of imperfect surface high spots might result in a measurable and maybe even significant loss of preload.

One of the improvements claimed and provided on the Chevy "pink rods" was an improved surface finish on the rod and cap parting faces.
Some may remember the standard SBC rods of that era had a dreadfully torn broached surface on the parting faces (and in the vulnerable bolt head notches) . When rebuilding those plain vanilla rods it sometimes took several extra cuts just to get decent surface cleanup.