Ultrasonic testing of assembled fasteners

[Tmoose]

Two of 90 36 mm dia bolts have failed (broke under the head) after ~ 2years in service. Replacing the two broken bolts is a big project. Replacing all 90 is ~ 25 X big.

I'm thinking some ultrasonic testing might help indicate if we need to go this way, if some severely cracked ones could be found. Could possibly grind the exposed end of the bolt, and use a low profile probe. Access is tight.

Any hope?

Thanks,

Dan T

 

[metengr]

Yes, ultrasonic testng using a straight beam should work. I would contact an NDT company and make sure they can support this.

 

[stanweld]

UT has effectively been used for over 30 years for insitu inspection of cracks in bolting. Back in the mid/late 70's fasteners in compressors let go with catastrophic results. We examined all of our bolts with UT and the examiner that we employed for the work went on to perform similar examinations for a number of other gas transmission companies.

I devised blind tests to verify the examiner's competence. He did find a number of cracked bolts, some of which were near end of life.

 

[CoryPad]

I agree, ultrasonic testing is ideal for this kind of work. If your photograph shows the actual part and conditions, then it looks like an easy project for someone well-versed with this technique.

 

[desertfox]

hi Tmoose

After you ultrasonic the bolts and assuming you have to replace some more, what steps have been taken to ensure it doesn't happen again, ie whats caused the failure after two years? reading your first post it doesn't sound like your company expected this failure of the bolts.

desertfox

 

[Tmoose]

" expected this failure of the bolts."
Our experience list shows we have sold hundreds of this type mill over the years, from 2 or 3 domestic manufacturers, with no record of broken bolts.
We bought this mill ..wink.... wink..... "overseas" .

Their website says they have over 400 of this type mill in service.
The manufacturer said to an engineer who speaks their language that they had some bolts break "several years ago." I have asked several times via e-mail what caused those reported bolts to break, and received no reply.

What follows is a list of questions that I have I asked the manufacturer several times, and to date have received no reply.

1-The body of the one broken bolt we have is extremely rusty. I sent pictures and asked if they expect this level of corrosion in service. (Our metallurgist said the corrosion of the fracture surface was too rusty to analyze - sheesh)
2- The Washer under the nuts shows sign of minimal contact ( a narrow ring) with the mating part. The holes shown on the drawings for the mating parts would be appropriate for a 43 mm bolt, but a 39 mm bolt was used. I asked if they expect this narrow contact, how thick the washer is supposed to be, and what is the flatness tolerance for the spot face under the washer
3-We got a copy of the bolt drawing after the bolt broke. The material is nominally Grade 8/12.9. The assembly drawing for the factory assembled (over there) tube and bearing trunnions describes hydraulic tensioning to a value appropriate for a grade 2 bolt. I asked how the tensioning value compares to the working loads.

There are other issues with systems on this mill, and many, probably most of the replies to my technical questions have been similarly ..... disappointing. For instance, the mill is driven by a girth gear and pinion. The pinion is supported by SKF double row spherical roller bearings using tapered adapter sleeves drilled and tapped for hydraulic removal, mounted in mill manufacturer supplied custom pillow block housings and seals. The O&M manual has no lubrication procedure or maintenance schedule. We eventually got a procedure ( the bearing housings block the W33 groove and holes in the outer race) But the OEM's reply to my request for a lube schedule started with they would ask SKF, and recently ended with "SKF recommended twice a year, but it depends on rpm, service, etc." Well, the OEM designed the seals, so I'd guess after 400 mills they might have an idea how effectively they seal against dust and dirt in a typical plant, they designed the mill for a particular rpm, and with a customer base of 400 I'd expect they'd have a little feedback, and been asked this question before, or could go ask their largest happiest customers how often they grease the danged bearings.

So, I think when this mill was purchased a few years ago no one would have expected bolts to break. Today, I am not really surprised one bit.

 

[desertfox]

Hi Tmoose

Oh dear sounds like a bit of a disaster to say the least, however given that the holes for the bolts etc are to large, it sounds to me as if you need to replace all the bolts and possibly of a larger size so don't waste money on checking for cracks.
I presume the OEM's warrenty as lapsed?
Good luck I think you need it.

desertfox

 

[unclesyd]

Based on your post you have myriad of problems to work around. You have a fastener that reality is an unknown material , The bolt hole hole is too big for for the size fastener used, an improper washer, and improper tightening of said fastener.

We routinely UT check bolts on our remaining Clark Air
Compressors. We started this many years ago so all
replacement bolts, studs, have one end ground to accomodate the UT transducer. Our problem originated from improper installation, low torque. We were able to find small cracks emanating from the thread root. We took the approach that any crack was bad crack

and that fastener would be replaced. A material change was made for the studs, unknown to B7 with a 2H nut. As these fasteners were mainly from the cylinder jugs each time we replaced a jug or where the nut landing area was compromised it was refaced.

As posted above you will have to screen your UT technicians as I've had several that were not good at testing bolts.. I don't know what transducer they using now, but initially we used the same transducer that was used for measuring elongation in a fastener.

On our remaining engines we now use a nut style tensioner to eliminate any problems with tightening. We have used tensioners in areas where there is close quarters that could affect tightening of the nut.

http://superbolt.com/index.php

 

[Tmoose]

"improper tightening of said fastener"

>>If<< someday I can actually confirm the listed preload was solidly in excess of the applied loads I'd consider the specified tightening adequate.

I have seen a few applications where fancy bolt materials have been used at much lower than grade-appropriate torque/stretch/pretension (similar products, pretty highly engineered components, from other OEManufacturers with much different quality, engineering and manufacturing cultures). I reckon sometimes the material is there because that's what's commercially available in that size bolt, or for other properties like corrosion resistance. >>When<< we replace these bolts, we will use US name brand socket head cap screws in a comparable inch size (detailed for extra corrosion protection and control, with Conan-esque flat washers) because they are in-stock, and a huge bargain compared to duplicating the originals.