NAS1351 Heads Strip Below Installation Torque

[JCorsico]

thread2-442113

Resurrecting this old thread, I have a batch of NAS1351-3 (#10-32) socket head cap screws that are drilled for lock wire. The heads strip at a torque value that is below the desired installation torque value. Specifically, they strip at about 65-70 in-lbs (we were attempting to torque to 75 in-lbs). For reference, the Unbrako recommended installation torque for #10-32 socket head cap screws (190,000 psi tensile) is 91 in-lbs. An NAS1351 bolt should have comparable strength.

This is not an issue with the installation tooling, as we've carefully measured the hex bits and they all have acceptable dimensions (many bits are even 0.0005" or so oversize). It seems to be a problem with the fastener.

I then looked at the SPS catalog that Will previously provided in the above thread, and it includes a section where drilled head bolts from a competitor stripped easily. SPS's theory was that the defective bolts were annealed to allow drilling the lock wire holes, and then the heads were not re-heat treated.

I suspect that is the problem with my batch of bolts. Thanks Will!

Jon

 

[WKTaylor]

JC...

Whoooa Tiger... 65-to-70-in# torque for 10-32 male fasteners is waaaaaay above any 3/16-32unjf threaded install that I have ever heard about. I'd also be very surprised if Your aren't damaging nuts or inserts at the same time...

Checked a bunch of torque tables and a corporate document. Typically the NAS1351-* [and high strength screws] would be installed with a set-point torque of 30 or 35-in#... up-to-40-in# if turning the bolt head dry.

However, coincidentally... these same tables listed typical torque values for 10/32 [5/16-Dia] of 60-to-85-in# for removable bolts and screws... higher for tension-rated permanently-installed bolts [ONLY, not screws].

Please provide a copy of the torque tables being used. Oddly the procurement spec for NAS1351... FF-S-86... did not have mandatory test-torques that are common for typical aircraft bolt procurement specs... I'm aware-of.

As a kid I constantly broke 3/16 bolts by over-torque because I was not careful/measured... and didn't understand how 'little' 25-in# torque actually was 'by feel'.

Regards, Wil Taylor
o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[rb1957]

he does have unusually strong fasteners ... 190 ksi

another day in paradise, or is paradise one day closer ?

 

[JCorsico]

An NAS1351 (or NAS1352) fastener is basically the aerospace equivalent of a high quality commercial socket head cap screw. As RB1957 notes, the ultimate tensile strength is 180,000 psi, which is much higher than most "regular" aerospace bolts. Thus, you can torque them much higher than "regular" aerospace bolts.

For example, the recommended installation torque for a #10-32 Unbrako socket head cap screw (which is 190K ultimate; 170K yield) is 91 in-lbs. This figure is coming from Unbrako (which was owned by SPS at the time that I got a copy of their engineering guide - so I'm guessing they did the math right!). The Unbrako figure is designed to generate 120,000 psi of residual stress in the screw threads, which is well below yield.

 

[JCorsico]

Please provide a copy of the torque tables being used.

Will - take a look at the SPS catalog that you linked too in the earlier thread. Page 8 contains the torque table. Their recommended seating torque for a #10-32 is 76 in-lbs (which should be de-rated to 57 in-lbs for cadmium plated fasteners; but if you had zinc plated fasteners, they want the torque at 106 in-lbs, which is wildly above what my batch stripped at).

 

[WKTaylor]

JC... I am unconvinced.

From the certification paperwork received with these NAS1352 parts: what company made these bolts? What is the alloy [steel] [CRES] [HRA]? Per the NAS1351 Embedded table, these alloys all have different tensile strengths... CRES is ~(-)1/2 that of the LA steel parts.

IS the manufacturer's ID marking impressed on the head [REF MIL-HDBK-57]… and does it match the paperwork?



Regards, Wil Taylor
o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[verymadmac]

I won't use the unbreako numbers unless I established the same torque valve with some verification (OEM manuals, Calculation).

Have you only used a torque wretch to establish the bolt preload, try using anothere other method to determine fastener stress (bolt extension maybe). It could be as simple as the surface is lower friction on the NAS bolts.

Also have you looked at SAIB HQ-14-16 for fastener issues, you might need to adjust the torque for the fastener material.

https://rgl.faa.gov/Regulatory_and_...8925ec2f1c86257cc800611b60/$FILE/HQ-14-16.pdf

 

[JCorsico]

Wil - I don't know what to tell you!

I have a batch of 50 NAS1351-3 fasteners, alloy steel, cadmium plated, drilled head. We successfully torqued about 35 of these fasteners to 75 inch pounds, but approximately 12 of them stripped at that level. The SPS catalog that you posted previously shows a picture of this exact type of failure, and cites a rationale for the failure (head not properly heat treated after drilling for lock wire). Seems plausible that I've encountered the same issue. But who knows.

The head is stamped with "BB", which seems to correspond to the certs which said they were made by B&B Specialties in Anaheim in 2017.

Jon

 

[3DDave]

You could do a hardness test to find out.

 

[WKTaylor]

Per MIL-HDBK-57H

BB = B & B SPECIALTIES INC., 4321 E. LAPLMA AVE., ANAHEIM, CA 92807

B&B specialties

https://bbspecialties.com/

CALL (714) 985-3000 or FAX (714) 993-0119[highlight #EF2929][/highlight]
sales@bbspecialties.com / sales@gsaerospace.com

[highlight #EF2929]BE ASSERTIVE... CALL B&B Tech support...[/highlight]




Regards, Wil Taylor
o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[enginesrus]

Maybe give another calibrated torque wrench a try. Some times its the simple things that are the problem.

 

[WKTaylor]

JC... fastener math...

[47] Bolts@75-in#... [35] 'head-not sheared' + [12] 'head-sheared' = [32] compromised/over-stressed + [12] failed-for certain!!!

Regards, Wil Taylor
o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[MintJulep]

For reference, the Unbrako recommended installation torque for #10-32 socket head cap screws (190,000 psi tensile) is 91 in-lbs.

No. It's not.

The recommended seating torque in the tables in the Unbrako catalog is for plain, unplated fasteners.

You say that your fasteners are cadmium plated.

f you read the notes below the table you will find: "For cadmium plated screws, multiply recommended seating torque by .75.

91*.75 = 68.25

 

[JCorsico]

[47] Bolts@75-in#... [35] 'head-not sheared' + [12] 'head-sheared' = [32] compromised/over-stressed + [12] failed-for certain!!!

Yes, we removed all fasteners from this batch and are purchasing replacements.

And to be clear - it's not that the head has sheared off. It's that the internal hex recess has stripped (the bit spins).

 

[mfgenggear]

Jcorsico

have the fasteners metlap evaluated for embrittlement, seems from previous post high strength fasteners are not processed correctly with proper post bake after chemical treatment.
at least this can be eliminated if not the cause. also verifying the hardness with met lab for the correct heat treat & for correct hardness.
and a spectral analysis if the correct material was manufactured. if machined in the soft state a cross hole can be a high stress riser during heat treatment.
and if not processed correctly will propagate cracks, depending on the material used.
ooops: if the hex hole fails it appears to be lack of hardness. hmmm

 

[WKTaylor]

JC... OK, so the internal hex is 'rounding-out' not the head shearing-off. Hmmmm…...

Per NAS1351-3-* the hex recess depth [dimension 'T'] is 0.090 minimum... VERY shallow! AND... the end of the recess is ill-defined.

For comparison of hex-recess depth-of-engagement... as close as I can get...

NAS1351-4-* hex recess depth T = 0.120-min
MS20004-* hex depth = [D]-[E] = 0.250-0.112 = 0.138-nominal... and the end of the recess appears better defined.

Per A-A-2932 KEY, HEXAGON SOCKET HEAD SCREW AND KEY SETS SOCKET HEAD the edges simply have to be deburred... to any extent. Hmmmm…

Did You inspect the ENDS of Your hex key wrenches for any excessive edge radius that could inadvertently be reducing your hex engagement depth up-to-20%... with an already shallow hex recess???

Regards, Wil Taylor
o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[mfgenggear]

on standard socket head grade 8 bolts rarely ever strip out
except as noted by will, worn tool

 

[JCorsico]

Following up on this, we purchased another batch of NAS1351-3 fasteners from a different distributor. But this time, we ordered them without the lock wire holes...

This batch was also manufactured by B & B Specialties. And guess what - this batch doesn't strip!

We were able to torque the fasteners all the way to 90 in-lbs, which is well above our desired seating torque. We used the same torque wrench and the same bit that caused the prior batch to strip.

Thus, our theory is that either (1) the fasteners with lock wire holes weren't properly heat treated after drilling the holes (like that shown in the SPS catalog), or (2) the lock wire holes remove enough material from the hex recess that the remaining material cannot withstand the installation torque.

As Wil noted above, the hex recess is pretty shallow. The lock wire holes are pretty large in relation to the depth of the hex (it's a 0.050" hole). Thus, we suspect that theory #2 is correct.

 

[WKTaylor]

Hmmmmmmm…

Curiosity is killing me: what did B&B tech support have to say about all this?

Also... I had presumed that lock wire holes were necessary for 'security'... were they not really needed, after-all???

Regards, Wil Taylor
o Trust - But Verify!
o We believe to be true what we prefer to be true. [Unknown]
o For those who believe, no proof is required; for those who cannot believe, no proof is possible. [variation,Stuart Chase]
o Unfortunately, in science what You 'believe' is irrelevant. ["Orion", Homebuiltairplanes.com forum]

 

[JCorsico]

We have not had a chance to talk to B&B tech support yet.

And correct - the lock wire holes were there because we were going to lock wire the fasteners... So we are now evaluating alternatives. Either, keep the lock wire holes and reduce the installation torque so the hex recess survives. Or keep the installation torque and come up with some alternative locking mechanism. I think we are going to reduce the installation torque (it was selected based on the strength of the fastener, rather than the strength required for the joint), but TBD.