Documented practices for tack-welding hardware 1

[geesaman.d]

I work in rotating equipment where hardware loosening is occasionally a problem.

My company standard practice is to apply any means except welding to keep hardware from loosening. It seems like welding causes as many problems as it solves.

We have customers with specs who require welding of the hardware to prevent loosening. Is there a standard or documented practice that gives a framework for tack welding hardware? Or a standard that advises against it?

 

[Gr8blu]

Tack welding the hardware does three things. It adds thermal stress to the hardware, thereby changing the mechanical properties. It positively fixes the hardware to either another piece of hardware or to a fixed surface (think tacking a bolt head to a washer or nut, or directly to the surface the surface its threaded into.) And lastly, it has the possibility of creating a winding failure if - for some reason - the welding apparatus is improperly grounded. That last is why the rotating equipment group does NOT want to weld if at all possible - the first two reasons are why a user might suggest it.

Lastly - what about being able to take things apart again? If you have to grind off the weld, you're likely to require all new hardware and MAY damage the part as well.

Converting energy to motion for more than half a century

 

[3DDave]

Tack welds for that purpose seem like a great place to start fatigue cracks and I don't expect anyone to have a specification that suggests adding crack initiation sites.

The underlying problem isn't that the hardware is coming loose from turning - it is allowed to turn because it is already loose. Tacking hides that cause and will lead to hidden wear or fatigue failures elsewhere. It's more important to properly restrain the parts and maintain the elastic deformation that fasteners provide than it is make sure the nut is still in place when the fastener snaps loose. I have seen exactly that happen - loctite kept the nut in place on the broken-off stud.

To be clear - locking compounds are good for excluding corrosive materials and to help prevent side-to-side movement within the thread - but they don't provide a way to maintain a preload in a joint that isn't properly designed.

 

[geesamand]

It’s reassuring that I’m not the only one who thinks it’s a very poor solution for a host of reasons, and I’m sure we could each write our own thesis on good reasons to abolish the practice. Sadly, none of us are likely to change my customers opinions.

So I’m still on the hunt for published standards or documents that I can use as a 3rd party reference when customers insist on bringing it into a spec. Ideally it would scare them into allowing better design practices.

 

[3DDave]

Don't bother. Let them have what they want. If it works, they won't like any objections. If it doesn't - it belongs to them.

 

[BrianPetersen]

If they insist upon this method, then get THEM to specify what they want done - thus ensuring that any bad outcomes belong to them. It's not up to you to find out how to do something that shouldn't be done.

 

[3DDave]

ASME needs a new section with committees on various bad practices. Just document the worst of the worst. A recent example: using food-grade vegetable oil on automotive wiring harnesses as an installation lubricant so that mice are motivated to eat the insulation causing electrical failures and potentially electrical fires. Good job.

This way any time someone wants to justify a really bad idea they can ignore the titles, such as "Really Bad Ideas and Things Never to Do" and say it's per specification.

Some orgs did a bit of this already. I recall standards for soldering frequently have images of "not like this." That sort of effort could be the backbone of a whole new class of standards.

 

[geesaman.d]

Don't bother. Let them have what they want. If it works, they won't like any objections. If it doesn't - it belongs to them.

Ah, if only customers could be held accountable for their contributions to an outcome. If only the salesperson wouldn't spin a story about how this is a choice between losing a customer forever or showing support that lines up the next multimillion dollar PO from them.

It sure would be handy if ASME BPVC said something about this, as most of our customers are installing the equipment in an ASME coded vessel. That might get their attention. But sadly, BPVC has no coverage at all for agitators.

 

[desertfox]

Hi

This is not free but might help

https://www.epri.com/research/products/1003543

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

 

[geesaman.d]

Actually, that document is free. Thanks!

 

[EdStainless]

We had a written procedure that basically said that if either part was Q&T or if they were alloys that would usually require PWHT then you could only do it with a fully qualified procedure.
I recall one case where a customer insisted, and our sales dutifully added a &50k like item for the procedure qualification.
The customer decided that it wasn't that important.

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

 

[geesaman.d]

That's an interesting risk reduction method, and I always like the break out the cost of a risky option for a customer and let the money drive the answer.

In my case we're almost always playing with Zinc plated Grade 5 vs low carbon steel, zinc plated grade 5 vs 316, 316 vs 316. We tend to stay away from high heat treat grades. So the hardness of the welds is less of a risk than, say, weld distortion twisting the parts and changing the preload.

The EPRI document takes a nice dip in the water but it's not comprehensive. It does remind me of some good ideas - to propose welding flat bar against a flat, to supply pantleg washers and let them weld those, etc.