What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

(Fastener) Torque. Not a debate

Freemasm

Well Known Member
If this comes off as a little preachy, sorry. There’s been some threads here where some don’t understand/want to debate:

Torque
Importance of
Proper methods
Application differences
whatever

Here’s a couple of examples:

https://vansairforce.net/community/showthread.php?t=69876
https://vansairforce.net/community/showthread.php?t=187303


Not getting into that again. In oversimplified terms, if you want to be safe and in compliance always follow the hierarchy:

Released engineering, OEM drawings
Released engineering, OEM specs (including service bulletins, etc.)
Industry standards (e.g. FAA Advisory Circulars)
Good Shop practice

Others could be added but are mostly subsets of the above. When in doubt or apparent conflict, inquire. When the aforementioned hierarchy is not followed, bad things can happen.

The following somewhat recent picture is from a Mooney that shed a jug over the UK (picture used with permission). It turned out as well as it could; no injuries and a repairable airframe. Could have been much worse obviously. A certified engine/installation so timing, CR, etc. should not have been contributors. While I haven’t personally inspected the fracture surfaces, it is most certainly a material fatigue issue; no necking near the fracture surface, flat perpendicular fracture surface. One of our very experienced NDEs agrees after viewing the picture.


The relevant OEM has released engineering for torqueing this joint; sequence, values, thread condition, etc. Vegas odds were this was not adhered to. So:

Tension application -> under torque condition -> cyclic load-> metal fatigue -> rupture.

The upside is there’s a good bit of design margin and builder expertise consideration in our builds thanks to the airframe OEM, Vans. That’s not always the case across the wider OEM spectrum. Failure to adhere to specs eats into that margin. Follow the released engineering. Build safe.
 

Attachments

  • cylinder stud failure.jpg
    cylinder stud failure.jpg
    92.8 KB · Views: 497
Some years ago, a friend who flew a BN Islander for a skydiving club lost a cylinder in the climb - came out of the cowling.

Landed.

They said - OMG, what are we going to do ?

He said - "I'm off to Manchester to get it fixed".

Grabbed his bag, started the good one and took off.

Those were the days - Ehhhh :D
 
I use the two torque methods:

The German method: Gut Und Tight

The mechanic method: Tighten until it breaks off then back off a 1/4 turn.​

All RV builders should down load this Aviation Circular (20MB) 646 Pages

Subject: ACCEPTABLE METHODS, Date: 9/8/98 AC No: 43.13-1B
TECHNIQUES, AND PRACTICESAIRCRAFT Initiated by: AFS-640 Change: 1
INSPECTION AND REPAIR

TABLE 7-1. Recommended torque values (inch-pounds).

With that said after turning a wrench for 40 years (DIY'er mostly) I know when and where and how I can estimate tightness or torque with a wrench or socket/ratchet. I am not going to break out a torque measuring tool for every screw and bolt.


The salient factors:
Having the data
Having a tool to measure torque
Using the data and tool mentioned previously.

Torque may be uber important. Torque may be less important or alternative methods used in less than critical places. Measuring the torque may not be practical. Screws holding a non structural item may be "estimated". I own two German vehicles, VW TDI Diesel and BMW Motorcycle. I have the shop manual for both. There are LOTS of torque numbers for almost everything in these manuals down to small screws. Not that you will under torque, but small fasteners can be over torqued easily. I have shop manuals for other products and vehicles. Little to minimal torque data is given.

I bought the VW car used. It just had the 40K service before the dealer sold it as CPO "Certified Previously Owned". The fuel filter (LARGE on a diesel) was leaking. 2 of the 6 fasteners were stripped. The manual requires 24 inch pounds of torque. The auto dealer mechanic stripped them. I have changed the filter several times since over 10 years. I bought a small 1/4" drive dial gauge for low torque fasteners. Don't use a 1/2" drive torque wrench capable of 120 ft-pounds to torque to 24 inch-lbs.

TORQUE wrenches can lie, get out of caliber, fail. I have 3 torque wrenches, dial for lower torques, beam for general use and click set for higher torque. There is also TORQUE + angle, typically for stretch bolts in head, case and main bearing fasteners. There are nice digital torque wrenches that will torque and tell you the amount of rotation or angle. I would like one but they cost a good bit. I don't do many stretch bolts.

Keep in mind ONE SMOOTH torque motion to the set value without stopping is preferred. If you stop you have to overcome large static friction and will likely get a bad torque. Stopping may make the torque wrench read incorrectly. This does not mean you can stop at a set lower torque and come back and apply final higher torque.

THOU SHALL NOT TOUCH THREADS... often torque is dry (no lube). If you lube the threads (oil from skin) you may over torque. However other times they are wet (as specified). If there is no specific torque use the standard.or
 
Last edited:
I would not say often in general, only as specified.

THOU SHALL NOT TOUCH THREADS... often torque is dry (no lube). If you lube the threads (oil from skin) you may over torque. However other times they are wet (as specified). If there is no specific torque use the standard.or

If anything really needs proper tension and is obtained by torque then they MUST be lubricated. Connecting rods, case bolts, cylinder base, heads on any other engine. Tension variation vs torque varies widely if dry fasteners are used.

To the barrel base failure - - did you know that Continental does not allow RTV on the bases as enough material remains in the joint that the fastener(s) will fail in fatigue?
 
If anything really needs proper tension and is obtained by torque then they MUST be lubricated....Tension variation vs torque varies widely if dry fasteners are used....

Bill, while that's true, the torque values for lubricated threads are lower than for dry threads, and that needs to be accounted for. We can't simply dab some lubricant on a thread and then tighten it to the dry torque.

Dave
 
Bill, while that's true, the torque values for lubricated threads are lower than for dry threads, and that needs to be accounted for. We can't simply dab some lubricant on a thread and then tighten it to the dry torque.

Dave

I did not mean to imply that we dab anything Dave. For all core engine, transmission final drive and equipment threads there is no dry torque, as the grade 8, or higher, fasteners are required to be torqued to the specification, lubricated or something will fail. This is because if one attempted to reach the design fastener load dry with excessive torque, the threads will typically fail before reaching the proper stretch. I came out of the heavy duty diesel and equipment world and nothing is torqued dry and it is rare that a fastener is used in a shear function. When I first started building was surprised that airframe fasteners were all dry and nearly everything is a shear load. Completely different design philosophies at work there. Engines, all, are back to the tension being required to keep parts from moving, that takes stretch and lubricated threads.
 
Hey Bill,

I’m glad you brought up the RTV jug/case joint issue. How many (fractions of?) mills of RTV you think still exist between those surfaces under that much clamping force? When it bakes away in short time, that infinitesimal (for lack of better word) strain reduction and subsequent stress relief can allow a reversing load on the fasteners.

People like to reference anecdotal cases where lack of adherence to released engineering, and/or other methods or procedures is acceptable. The “evidence” is the fact that failure didn’t occur immediately. Then, a jug separates from a power plant.

Following the aforementioned hierarchy in some cases, is far more critical to safety than many will ever know.
 
Lots of long, critical fasteners in modern automotive engines (like head bolts) are torqued to a low value to seat and then turned 1/2 to 1 1/2 turns further rather than using ultimate torque. This results in more accurate fastener stretch and clamping force.

Race engines often use machined spots on rod bolts and these are torqued until the fasteners reach a specific length, measured with a mic, indicating proper stretch.

Conventional torquing results in ballpark stretch due to differences in lubrication, and thread friction.

The top 2 methods are mostly/completely unaffected by differences in thread lubrication and friction.
 
Last edited:
If this comes off as a little preachy, sorry. There’s been some threads here where some don’t understand/want to debate:

Scott, I sense your frustration re the application, or lack thereof, of proper torques for fasteners. Good on you for raising this important issue... however futile it may be. However it's the Experimental category and in the end builders will do whatever they like. Many of them are in too much of a rush to finish the plane to worry about reading and educating themselves on correct aviation building practices. Many are building simply because they figure it's the cheapest way to achieve their real end goal which is flying.

The problem is that within a few more posts this thread will evaporate into the ether. It will simply be washed away like footprints in the sand at high tide. And then inevitably down the track another similar thread will be started and the same information (and disinformation) will emerge again. It's perpetually groundhog day on VansAirforce.

I've been warning VansAirforce members for years about the very real dangers of cracking their canopies by applying inappropriate solvents, and by unsatisfactory treatment of edges and holes. But I can tell that I'm largely pissing into the wind.

The reality is that the Experimental building world is a passing parade, with builders constantly entering and departing.
 
Last edited:
You have to watch this video to truly appreciate how complex and demanding the head bolt torquing on this Subaru engine really is. Just amazing.

Every engine is different, and if I recall correctly these bolts were single use on my STi. Impressive amount of engineering has gone into this process. Seems surprising that with all that work they still had problems with leaking head gaskets for so many years. Engines are simple in theory, but complex in reality!

I think this thread is great to help raise awareness of bolt torque.
 
Stress and Strain are linear and relative to each other while the material (fasteners here) are in their elastic range.

As Ross pointed out, some fasteners in car engines utilize the "turn" method versus the "torque" method. This is more common in other parts of the world and works better with longer fasteners versus shorter ones. Strain being fairly easy to accurately measure and back calc as it is deformation/unit length; thus, long bolts stretch more under the same stress/strain than short ones as stretch more under the same load. The Turn method also lends itself to the more precise tolerances of a power plant. I'll bet anyone any amount of money the turn values specified were validated via measuring the related fastener strain during development.

As Clarkefarm added, the video is a great illustration of a required technique I'm also sure was validated via some sort of stress/strain measurement during deveolopment. The best point made by both of the aforementioned gentlemen?

Both follow the procedure/recommendations of the OEM.

When the ink on my degree was still wet, I was an engineer on the Space Shuttle Main engines; Field engineer basically. There is a scary amount of energy and operating conditions involved in those machines. As thrust/weight ruled, every joint was custom engineered. In general, weight was kept down in fluid systems by utilizing relatively thin flanges and more fasteners. The flange surfaces were neither very rigid nor flat. They "rolled" flat when torqued. Torqueing in increments was essential, the last torque cycle never more than 10% of final total. Silver plating of threads was sometimes used and limited to two torque cycles. Strain of studs/bolts in critical joints (power head to main combustion chamber and turbopumps) were always verified via ultrasonic measurement. As the flanges were not very rigid as mentioned, any adjustment to one fastener would affect the strain of others. It often took multiple work shifts to get a single joint to pass QC. I digress and succumb to the way-back machine but it gave me an acute appreciation of many things as the consequences of insufficient engineering or poor workmanship were huge.

As many are visual learners, I lifted a graph from the net everyone in engineering school had to familiarize themselves with early on. I marked-up a couple of points for illustration and pasted below. You can get there by various means; torque, turn, stretch but the targets only move in rare exception. Hey, want to win a bar bet? Bet someone who believes themselves to be mechanically savvy which identical fastener gets a higher torque value; one in tension or one in shear? You'll enjoy more free adult beverages than you purchase. Few are aware or consider fatigue.

I'll state again;

Follow the released engineering from the OEMS down.

How's that for a morning rant? Build safe. Cheers boys.
 

Attachments

  • Stress strain.png
    Stress strain.png
    111.1 KB · Views: 203
Last edited:
Head bolts

Very interesting. I don't know squat about the engineering so I trust the tables. I have torqued quite a few head bolts. Several were "torque to yield". We couldn't figure out why the head gasket kept failing. Yep. We reused head bolts. I also learned a lesson, long ago, about hydraulic locking. Bolts were supposed to be coated with oil then left to drain overnight. I dutifully followed, but didn't blow risidual oil out of the case holes. Yep. Snap! Oh Bother!
 
Some years ago, a friend who flew a BN Islander for a skydiving club lost a cylinder in the climb - came out of the cowling.

Landed.

They said - OMG, what are we going to do ?

He said - "I'm off to Manchester to get it fixed".

Grabbed his bag, started the good one and took off.

Those were the days - Ehhhh :D

As a publicity stunt Aero Commander once removed the prop from one side and flew coast to coast. This was back in the 50s as I recall (yeah, I'm that old).
 
Bolted joints are one of the oldest and most misunderstood technologies known to man. What makes it worse is the torque-tension relationship is very predictable in theory, but unfortunately very unpredictable in practice. As a result, many different joint designs result depending on:
- fastener
- materials
- criticality
- accessibility
- design margin
- design style of the particular design engineer
- etc.

you can theorize and argue all you want, but at the end of the day, as frustrating as it may be, follow #1 Engineering drawing, #2 OEM spec, #3 Industry spec.

All that said, many joints on machines, in this case airplanes, do not have that much design work behind them. Only the critical ones, like on engines, props, primary structure, stuff like that. Cowl screws for example don't have that much analysis or thought behind them because they don't need it.

in any case, this thread started out with an engine picture, so I think it is primarily targeted at that. For all those fasteners there should be specific lube and torque specs that should be followed precisely.

Tim
 
Back
Top