NDT suitability on a connection

[trainguy]

All,

I am evaluating a connection between a steel bracket and an aluminum support structure, fastened by steel blind rivets, and in use for around 30 yrs, on the underside of a railcar. The original design drawings show mica paper called out at the interface, but I can see corrosion at the perimeter of the steel bracket.

Is there a method of non-destructive testing that would help quantify how much of the aluminum is corroded and where, esp. near the rivets, etc.? We are trying to avoid removing the bracket in this evaluation. The goal is to determine if the joint as-is still meets original structural design requirements. Joint is loaded primarily in shear, as a bearing type connection, because the rivets do not provide high clamping.

Any help would be appreciated.
tg

 

[racookpe1978]

Has it been in use for 30 years, or is intended for use for 30 years - but is showing accelerated corrosion now?

 

[trainguy]

It has been in use for 30 yrs. I am not sure when the observed corrosion first appeared. On a sample of around 5 recently inspected cars, this appears on approx. half of the joints, i.e around 10 of the 20 joints inspected.

tg

 

[MintJulep]

Sheer joint, with little clamping and mica between the faying surfaces?

Sounds like the carrying capacity should be about zero.

Are you sure that you understand the load paths correctly?

To answer the question that you asked, no. There is no NDT inspection that will reliably do what you want.

 

[trainguy]

MJ,

It's not a slip-critical connection (i.e. friction joint). There is some clamping, but not as much as in a torqued bolted joint. Rivets are actually bearing against surrounding mat'l - hence the "bearing type" description. Drilled hole size is 0.413. Rivet (Huck BOM)diameter per catalogue range: 0.393-0.411, so a relatively snug fit.

I can confirm that there is mica, and there is shear, and it's held since 1980. Thanks for the info on NDT - much appreciated.

Any other ideas, short of removing it, grinding to sound metal and measuring thickness loss? This would be major...

tg

 

[dhengr]

Trainguy:
Side bearings (cages and wear pls.), jacking pads, sill steps, center plates, all might have the same problems. I assume these railcars have an aluminum underframe and some misc. steel specialties. This issue always bothered me in these situations and I never arrived at a totally satisfactory solution to your problem/question. The few times we ever did this, I think we used some sort of a fairly thick asphaltic paint on the parts, and let it partly set up before applying the parts. This part of the inspection report is going to be plenty subjective and filled with caveats. A little corrosion is showing up, only on about half of them; not too bad an indicator. How bad is the corrosion, probe around at the joint a bit to see if you can get a feel for the extent, on the exterior. Is there still good bearing btwn. the two parts, at the faying surface, is the mica still fairly intact? Tap on a few of these and see if you get any relative lateral movement btwn. the two parts. I don’t mean a tack hammer, but I don’t mean a 10lb. sledge hammer either; to try to get a feel for any looseness in the rivet bearing holes. Are these hot driven rivets, not huckbolts? Wacking all connections with a good sized hammer would probably show any looseness, a good indicator to look further. Load test a few of them to 1.1 not 1.8 load factor. Then say, ‘to the best of your judgement, they seem to be in pretty good shape for 30 years of service.’ Cut the worst one off carefully, and examine it internally. Ask a real good x-ray tech. what he would see/indicate if he shot right down the axis of the rivet; the steel rivet, a very tight (or loose and junky) circle at the edge of the hole, and the steel or alum. of the two primary parts. I just don’t know how that would indicate on the film. Is this an inspection to extend the life of the car? I’d be more concerned about the underframe and body conditions and details. While the specialties are not unimportant, they do get a little more attention by the carmen in normal walk arounds.

 

[Ron]

If you only want to check thickness loss, then you can use an ultrasonic thickness gauge. Further, someone who is very adept with weld flaw detection using an ultrasonic flaw detector and an angle probe can probable find anomalies or indications around the rivets. Wave transfer could occur in the mica sheet that will make interpretation more difficult. I would suggest heating the connection with a heat gun to drive off all moisture before attempting the ultrasonic testing.

 

[racookpe1978]

Sounds like they both anticipated, and then "tried" to prevent metal-metal contact at the pin. Failed to prevent it, but tried. Problem is, the pin does act as a pin, and in a "loose" sliding-type mobilized and rattling and bouncing pinned connection like a railcar, that pin IS going to move and touch the other walls of the pinned holes. Despite mica - which sounds like it workded actually pretty good on the flat, but can't help inside the hole, and the paint - which will wear through.

Good thing is that the pinned hole will be strongest in shear in the wall of the large plates; bad thing is that the metal-metal corrosion will be in the weakest link (the pin itself) inside the hole at the smallest surface area touching. I don't think a plastic (non-conducting sleeve) around the pin will be effective since it will be less resistance to simple wearing forces than the pin.

Could you replace the Al pin with stainless? Something less further apart on the galvanic corrosion difference scale?

 

[trainguy]

Thanks for the input so far guys.

Some more details:

Orig. rivets are Huck BOMS, which are zinc coated steel rivets, not hot driven but relatively tight fit.

And dh, we're not talking "specialties", we're looking at truck to carbody connections (ult. design load 250k), and body bolster to side sill connections - major.

Finally - passenger railcar so smooth ride - some fatigue, but doesn't usually govern design.

Client is asking to put off repairs for some time period TBD based on priorities I propose. It's like a passenger car variation of an extended life assessment.

tg