Tighten nut on clamp T-bolt "snug": What the heck is "snug". 4

[WKTaylor]

We use the term "sung" for installing parts "tight-but-not-too-tight". Most of us have a "feel" for what snug is.

HOWEVER, recently, several of us got together to define "tighten-snug" for a 0.250-28 plain nut. None of us had a clear definition for the mechanic who wants to now what "snug-torque" is... or what a snug fit of parts is.

How do You define the term "snug" as related to torque and mating-part fit???

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

Unfortunately, in science what You 'believe' is irrelevant – "Orion"

 

[WKTaylor]

All… some more info...

See attached photo of damaged duct section. The effect of clamp T-bolt over-torquing is pretty obvious.

The evident pinching/flattening and rectangular [sharp-cornered] creases are enough to cripple/fail the duct per a theoretical analysis and testing program accomplished for critical hot-air/high-pressure pneumatic ducting of another acft, that is critically reliant on this system. I think the only saving grace, RE our duct damage, is that the clamps, in addition to damaging the duct, also provide some strain restraint to keep the distorted wall from continuously flexing [fatiguing]; and that the system is statically pressurized without significant airflow during most of the flight.

NOTE.
This same ducting section, and adjacent ones are, are experiencing significant flexible bellows damage. This raises a larger issue in my mind, IE: how these duct sections are connected to other duct sections and mounted to the engine/airframe. My prior experience with pneumatic ducting with strain-relief bellows, like this, is “how they are assembled together makes a huge difference in assembly and localized/operational strain”. I found that loosely clamping duct sections together, then loosely installing structural clamps/mounts, then slowly/methodically tightening the duct sections together and to-the-mounts-to-the-structure, usually brings the entire assembly into a minimal strain “best fit” situation. However, this some-what long-drawn-out procedure requires the maintenance tech to mix-in “maintenance art” with the raw-procedures listed in the MM… and NOT rush the job. Unfortunately this might not be the case here… along with the improper clamp-torquing procedures.


Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

Unfortunately, in science what You 'believe' is irrelevant – "Orion"

 

[MikeHalloran]

Yeah, the slow, gradual tightening is how it's done on yacht exhausts, too. Further, the tubes are typically suspended by rubber washers and helical compression springs, and not rigidly tied to the boat anywhere, both for reduction of conducted noise, and to minimize the installed stress in the tubing.

That bleed air duct appears to be made of foil, or something close.

... Which suggests that those clamps are _way_ too strong, and way too stiff, and assume the wrong shape when overtightened. If workmanship and uniformity could be assured, you'd be better off securing the duct with lockwire alone.

Have you looked in a Tinnerman catalog?

Mike Halloran
Pembroke Pines, FL, USA

 

[WKTaylor]

Follow-up NOTES.

Duct tubing is 0.035-inch WT X 3.0-inch OD Inconel 625 annealed.

The duct analysis, I mentioned above, was actually run for similar OD ducts... but ~1/2 that wall thickness [0.016--0.020-inch WT] made from a different heat resistant alloy [HRA]. The study suggested that creased dent stress concentration was far worse as the duct wall was thickened [stiffened]: the thinner wall tends to allow duct to attain a "rounder-shape" with lower strain; whereas the thicker wall duct retains its distorted shape due to much higher stiffness. In-all-cases, sharp corner dents [shown] have the highest KT relative to "pure" length-wise or circumfrential dents alone.

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

Unfortunately, in science what You 'believe' is irrelevant – "Orion"

 

[Compositepro]

You could put springs under the T-bolt nuts and tighten to a set spring compression.

 

[MikeHalloran]

Actually, with the proper springs, you could set up the 't-bolt' with no threads to adjust, and a quick-release overcenter latch. ... thereby controlling the installation with design and manufacturing tolerances, not assembly procedures.

ISTR a detail like that in a Voss catalog.


Mike Halloran
Pembroke Pines, FL, USA

 

[tbuelna]

wktaylor-

Thanks for the photo and the additional information.

One thing that caught my eye in the photo was that it appears the lower clamp location is located on a bend, but it may just be an optical illusion. It's a bit unusual to clamp in the middle of a tube bend, since all-metal clamps (like your example) are not really designed to accommodate the lengthwise curvature in the tube wall.

Since your clamp installation procedure must be based on using the existing components, as I noted above I think the best approach would be to tighten the t-bolt using a dimensional value. Looking at the sketch of the clamp installation you provided, it would seem easy enough to first take a measurement between the inside surfaces of the strap loops (the .52" reference dimension in the sketch) with the strap loops squeezed together using finger pressure. Then the t-bolt would be tightened to an easily controlled/measured lesser distance between the strap loops.

It is easy to see how tightening the t-bolt using a torque value could result in the damage to the tube that you show. Once the tube wall yields due to forces from the clamp, the wrenching torque required to tighten the t-bolt further should not increase much. The mechanic would have no way of knowing that the tube wall has yielded, and would simply continue to tighten the nut, probably wondering why the wrenching torque never reaches the spec value even after applying way more turns on the nut than should be necessary.

Lastly, you mentioned there were some stress failure issues with the metal bellows between duct sections. I spent some time working for a company that specialized in designing/manufacturing aircraft duct systems. One thing I learned during that time was just how complex designing a high-performance metal bellows actually is. They are designed for very specific operating conditions, including angular, radial and axial displacements. It is easy to imagine how excessive misalignment at installation of the ducts could produce overstress in the bellows.

Regards,
Terry

 

[3DDave]

I'm surprised at the design of the joint, given the range of temperatures. There is considerable thermally induced size change in the tube that isn't going to be followed by the bolt fast enough. It seems too stiff tangentially. Sometimes grabbing an item too tightly is worse than letting it rattle.

It doesn't seem that the design is as well investigated as it should have been. A better look at the thermal effects, especially transient conditions, would be worthwhile.

 

[tbuelna]

3DDave-

The effects of uneven thermal expansion is a serious concern with this kind of high temp bleed air ducting. It is critical not to over-constrain the ducting with the clamping. With a bleed air duct operating at up to 800degF attached to a nacelle structure that might be 700degF cooler, it is important to ensure the duct attachments allow for adequate thermal expansion without over-stressing the duct components.

 

[MikeHalloran]

With a bleed air duct operating at up to 800degF attached to a nacelle structure that might be 700degF cooler, it is important to ensure the duct attachments allow for adequate thermal expansion without over-stressing the duct components.

The subject half-clamp doesn't get anywhere close to meeting that spec.

</dead horse beat-down>


Mike Halloran
Pembroke Pines, FL, USA

 

[WKTaylor]

Yeah, I know: the clamp design/installation in my humble opinion s*cks. There is a matching saddle mount contoured to the the tube OD radius... but it only spans about 90-deg [+/-45 from center of contact]. There is no "flexibility" to this design for thermal and pressure effects; perfect clamp fit, not-too-tight/not-too-loose, is all-important... unlike flexible installations suggested by Halloran, 3DDave and tbuelna.

This design is way-down the road, and no redesign feasible [customer is not able without heroic effort... and other problems are far worse than this (another story)].

Thanks for great imput!

NOTE. I do like tbuelna's suggestion RE tightening the clamp T-bolt nut, until a thin feeler gage begins to bind-up between the tube OD and the clamp ID.

NOTE. SAE AIR809 Metal Dimensional Change with Temperature is helpful for calculating thermal expansion from RT to +/-xxxF!

Regards, Wil Taylor

Trust - But Verify!

We believe to be true what we prefer to be true.

For those who believe, no proof is required; for those who cannot believe, no proof is possible.

Unfortunately, in science what You 'believe' is irrelevant – "Orion"

 

[Compositepro]

1/8" of silicone sponge rubber between the clamp and the duct would do wonders.

 

[MikeHalloran]

Yep, it should burn real nice.


Mike Halloran
Pembroke Pines, FL, USA