Failure of anchor rod via stripped threads 4

[Ingenuity]

I recently repaired a 60 ton center-hole hydraulic ram (for post-tensioning bar stressing), and to verify that the new seal kit was indeed holding I did a recent load test. I grabbed a 1" dia x 105 ksi F1554 galv anchor rod with heavy hex nuts and washers (I usually use a 150 ksi PT bar but not on this occasion).

Anyway, got to about 3000 psi of gauge pressure (about 36 kip of load) and the threads stripped, pressure dropped back to less than 2,000 psi then I increased it again, and every increment to about 3,000 psi I continued to strip the threads.

So got me thinking that what I thought was 105 ksi materials appears to be only 36 ksi (1" dia tensile capacity is min 35 kips). Opps - and why I usually use 150 ksi PT bar with non-UNC threads.

But I was surprised that the it did not fail at the root of the threads. I have always assumed that the mechanical threads were stronger than the tensile failure of the thread-included section.

The separation you see in the photo is a grinder cut (not section failure) because I had to remove the bar from the test fixture.

Load was applied essentially monotonically, except for the stepping increment after first thread failures.

I learnt something today, besides keeping better account of anchor rod grades.

 

[desertfox]

Hi Ingenuity

You are correct the bolt should fail in tension before the threads fail, I am surprised the threads stripped assuming of course that the nut and bolt had the correct thread engagement and similar mechanical properties. Is there any possibility that the anchor bolt suffered some defect in manufacture like hydrogen embrittlement?

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[Ingenuity]

desertfox:

Thanks for the confirmation of expected mode of failure.

The anchor rod is new and has never been in service so EXTERNAL HE is therefore not a probable cause, but I am somewhat familiar with HE (not so much with anchor rods and bolts, but with 270 ksi prestressing strands and wires), so my (limited) understanding is that during the galvanizing process it can cause diffusion of hydrogen and hence possible embrittlement, as INTERNAL HE.

I had assumed that HE would cause premature section fracture (not thread stripping), similar to what happened on the SF- Oakland Bay Bridge anchor rods, although that failure was due to EHE.

 

[XR250]

I believe the tolerances on hot-dipped hardware are looser so the nut can actually thread on over the the thick layers of zinc. I imagine that would reduce the capacity. Is this a thing?

 

[JLNJ]

I thought they re-thread the nuts after galvanizing. Maybe the re-threadding was done improperly and the nut fit too "loose".

 

[dauwerda]

Nuts are tapped oversized after galvanizing (here is more info on that), this should not cause this type of failure though.
Do you know where you got the anchor material from?

We once had 1-1/2" diameter A325 bolts breaking in the field when being pretensioned. After a very thorough investigation by the supplier and manufacturer (including chemical analysis of the bolts) it was determined that a short (maybe 3ft) length of incorrect stock had been left in one of the machines (something about a demonstration had been done for a field trip or something) and not removed prior to the bolts being manufactured and was inadvertently used in the production.
They replaced all of the bolts in the lot and payed for everything (including labor of replacing bolts already installed)

I would be calling the supplier and having some sort of investigation made. This type of failure could be catastrophic in the field.

 

[dik]

Nope... and lose all that nice sacrificial zinc...

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[IRstuff]

Did the nut threads strip as well? It almost looks like the nut was larger than it was supposed to be.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[azcats]

How is/was the fit b/w the nut and the rod? Every now and then you come across some that feel 'looser' than they should.

Doesn't look like great engagement on the left side where it appears the failure started.

 

[azcats]

At 36 kips aren't you over yield on that bar (if it's 36 ksi). Any odd thread engagement things happen as you start stretching the bar and loading threads individually?

 

[dik]

I always thought that threads were designed so the failure would be through the threaded part and not the thread. It seems the threaded part did not fail.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik

 

[azcats]

I've heard the same. I'm really suspecting poor thread engagement b/w parts.

 

[dauwerda]

At 36 kips aren't you over yield on that bar (if it's 36 ksi).

He said it was F1554 Grade 105, so should be 105 ksi yield and 125-150 ksi ultimate.
The effective area of a 1"Ø anchor would be 0.61in² at the threaded area. So at yield, it should be able to resist 64 kips. And even then, the failure should be yielding of the cross section, not shearing of the threads.

(I thought they re-thread the nuts after galvanizing)

Nope... and lose all that nice sacrificial zinc...

They are re-tapped oversized after galvanizing.
The zinc coating on the threads of the anchor rod (or bolt) protects the threads of the nut.

 

[EdStainless]

I would start by checking the threads on the rod and nuts (un-tested samples) for size.
And by testing the strength of the rod (not relying on the treads).
Even though the threads on galv are cut oversized they should still be plenty strong.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[azcats]

OP suggested/suspected the rod was 36 ksi.

 

[Ingenuity]

I have 3 sources for my anchor rods - HILTI, PORTLAND BOLT and my local steel merchant. The first 2 manufactures/suppliers I have all have mill certs as they are custom lengths and all 105 ksi. I have purchased from my local steel merchant in standard 10' lengths and it is always grade 36 - basically your off-the-shelf, typical all-thread rod.

For this exercise I was a bit (a lot?) careless and grabbed a left-over piece from my remnant shelf and ASSUMED it was 105 ksi, but it could not be with such a failure.

The nut fit up was good, without excessive 'play'.

The nuts are somewhat now mechanically 'swaged' on to the rod, but I don't believe the nut threads have stripped. I shall try and remove and visually check and report back.

I am going to repeat the test using verifiable MILL CERT material, same 1" diameter and I shall report back.

I am still surprised that the threads sheared/stripped the way they did.

We recently installed 105 ksi rods (PORTLAND BOLT material with traceable mill certs, 1-3/8" dia custom, 48" long etc.) to an aggregate crusher support, and preloaded the rods with hydraulic jacks, and they have been in continuous service for 6+ months with no thread issues.

 

[Ingenuity]

==> UPDATE

1. I was able to cross-match the sheared/stripped anchor rod back to a remaining segment - and placing both together they were 10' long, and the material is confirmed as 36 ksi.

2. I removed the nut from the stripped threaded rod. The threads are stripped/sheared too:

3. I had a 16" length of 1" dia x 105 ksi rod with mill certs (rods, nuts, and washers by HILTI), and tested the rod to failure at approx 80 kips, with a cup-and-cone tensile fracture through the rod:

Ductile necking and elongation of >10%:

4. I then went back to another sample of 36 ksi rod (same batch as that with sheared/stripped threads) and when I installed the 2H heavy hex nuts there was a bit of 'play' so I switched to a nut with less play and a coupler with near-zero play. I load tested to falure at approx 35 kips, with cup-and-cone tensile fracture:

Ductile necking and elongation >10%

​

==> CONCLUSION: It appears that the excessive 'play' of the 2H hex nuts on the 36 ksi rod was the cause. Admittedly this is with n=1 sample, so certainly not a statistically valid conclusion.

 

[desertfox]

Hi ingenuity

Good detective work, I agree with your conclusion.

“Do not worry about your problems with mathematics, I assure you mine are far greater.” Albert Einstein

 

[XR250]

I wonder if manufacturers actually chase the threads after galvanizing or just make the nuts larger to begin with? I certainly have had my fair share of these nuts feeling loose.

 

[dik]

My experience is the latter... they have special 'nut' designations for the thread clearance.

Rather than think climate change and the corona virus as science, think of it as the wrath of God. Feel any better?

-Dik