OT - Odd Failure of Brass Pipe Fitting Domestic Use

I installed a new house water filtering system in March of this year. The system worked fine until July when both gauges of the first filter assembly developed a leak. I initially thought the gauges had failed and just bypassed that whole filter assembly. A few days later both gauges on the second filter assembly began to leak and then I bypassed the second filter assembly. I used cheap gauges, MCM part #4089K61, since this is just a 50psi potable water application. Yesterday I installed one better quality gauge, MCM part # 4003K11, as a test drive and it was clear that the pipe bushing was leaking and not the gauge. After inspection a crack was visible on all four bushings.

The system typically operates at ~ 45psi, but I have seen the gauges bottom out a few times at greater than 60psi. I haven't looked into this issue and was never aware of it since I didn't pay much attention to the gauge on the pressure tank. The pressure tank gauge may not be functional, I got tired of this project and moved on, but I need to replace that gauge as well. I cannot remember how it compared to the filter gauges when I saw them bottomed out. This is why I suspected the gauges in the first place, maybe I had damaged them with pressure, this seemed unlikely as I doubt it could have been much more than 60psi.

The day the first set of bushings failed I was filling a kid swimming pool with 1000 gallons. I filled it off and on throughout a 6 hour period. The system got fairly cold during this, whatever well water temp is in northeast Indiana, enough to make a good deal of condensation on all the pipes.

The pictures of the two fittings show the cracks on each. I machined a flat on the one and cleaned him up to highlight the crack. All four bushings cracked in the same manner, not all exactly in the same location relative to the hex. The bushings are MCM part# 4429K422. I disassembled these fittings with two 6" adjustable wrenches, I didn't go after these like a 200lb gorilla.

My only guess is a defect in the material, although this seems unlikely, but maybe they had a large enough run of poor material, that I could get four that all had the issue. This type of failure is the worst case if this was a second floor bathroom as the fittings did not leak for four months.

That may have been from a bad bar during manufacturing, where the mill rolled it improperly or [shit happened]. It doesn't sound or look like a crack due to improper installation.

Time for a call to MCM to let them know they should pull the remaining stock from that part number and purchase lot.

It's possible it happened during manufacture if they rammed a dull tap into the hole. I would notify MCM. You're not likely to get anything other than four replacement bushings, but MCM seems to care about the quality of what they sell, and I'm sure they will track it back to the supplier.

Dennis

Modelman said:

It's possible it happened during manufacture if they rammed a dull tap into the hole. I would notify MCM. You're not likely to get anything other than four replacement bushings, but MCM seems to care about the quality of what they sell, and I'm sure they will track it back to the supplier.

Dennis

Click to expand...

The middle picture doesn't seem like bad threads to me, they look pretty clean and well formed. I doubt that's from a bad tapping job, among other things the crack on the last fitting pic is through the heaviest section of the hex, not the thinnest. That, to me, indicates bad stock with a longitudinal seam in it.

I suspect they were simply over tightened.

steve-l said:

I suspect they were simply over tightened.

Click to expand...

By two six inch adjustable wrenches that weren't "gorilla'd"? And do the threads look torn and galled? Not to me...

Coworker found a gas leak in a new range. Same kind of deal, crack/void in a brass valve body casting iirc, but it might have been some other fitting.

That's bad barstock.

Limy Sami said:

That's bad barstock.

Click to expand...

Straight-line crack, thick section. Damnbetcha.

Hell - it's "captive", both OD and ID. The tubes and housings wudda failed first - not even its threads - if it had been wrench-monkeyed. As it was, they minimized the rate of leakage. Threaded plastics actually work in these apps, too. Line pressure in this case is trivial.

Not an expert but observe that the crack is through the heaviest part of the fitting and does not extend all the way through to the external threads of the thinner-wall section. Suggests that tightening the gauge wedged apart the hex section. As that upper portion wasn't constrained by being inside the next lower section of pipe it fractured. The portion with the exterior threads was constrained by the pipe it was threaded into and didn't propagate the crack.

Only you can say how much torque you applied, but it at least looks like it failed only during the installation and not previously.

I've done a far amount of wrenching and my internal torque wrench is reasonably calibrated. What is odd is that it held for four months, then two of them failed nearly simultaneously. I would have thought that if I over torqued them they would have split on installation. I appreciate the amount of stress that the tapered male fitting puts on the female bushing and I have learned the hard way to baby the bushings that are sequential in sizing, i.e. 3/8" NPT male and 1/4" NPT female, this fitting is fairly stout.

On the hex face of both fittings is a depression at the crack that you can feel with your scale, approximately 0.010" deep which also seems a bit odd.

Looking at the machined fitting again you can clearly see the crack at 180° and it runs past where the installed male fitting stopped.

I'll cut the one side and try to pry apart the crack to give a cross section view.

Update: I contacted McM and as always the customer service was excellent. They refunded the order without me even asking, I was just interested to hear what they thought. They asked me to forward pictures to their "fails" department which I did.

MCM has the best customer service of any large company I've dealt with.

If you want, try driving the bushing down a male fitting on its own, the wedging action should finish the split. Then you can forward pics of the lap (or whatever form the crack took) to MCM and post them here too.

Nice filter install. I’d sure get the new bushings and put the system back on line asap. I suspect a bad batch of stock too. But I’m not a failure analyst...
Joe

"...I used cheap gauges, MCM part #4089K61, since this is just a 50psi potable water application. Yesterday I installed one better quality gauge, MCM part # 4003K11, as a test drive and it was clear that the pipe bushing was leaking and not the gauge. After inspection a crack was visible on all four bushings.

The system typically operates at ~ 45psi, but I have seen the gauges bottom out a few times at greater than 60psi..."

Kyle, you never said what range gauges you bought. If your system is normally at 45 psi, you should use 90 0r 100 psi gauges. The rule is, use a gauge that ranges twice the expected pressure. You never want the gage to peg if you have any interest in reading the pressure and thinking the reading is accurate.

Here is a good set of recommendations for selecting pressure gauges: Selecting a Pressure Gauge - Winters Instruments

And I agree with the others that said the fittings were made from defective bar stock.

Larry

Just had 2 threaded brake line couplings break slightly enough to drip and slowly drain the system.

mark costello said:

Just had 2 threaded brake line couplings break slightly enough to drip and slowly drain the system.

Click to expand...

Uhh - are you on good terms with the wife? And is she handy with wrenches...?

I have had various brass tubes cracked in a similar manner. It seems to be stress from drawing. It is always longitudinal Sometimes there are multiple slits and the end of the tube flares out, like petals on a flower. No doubt a metallurgist can give a reason, but I can only guess.

Bill

Usually the petal fail is an extrusion, it’s gone on the die support (the core or mandrel is held in place by three wings if you like, sometimes called spreaders at 120 degrees)
When the metal flows around these supports to form the bore it fails to weld, can be wrong temp, pressure, or back end defect common to extrusion.
FYI the bore gets formed first flows round the webs and the outside diameter last, often called a port hole die.
In cast bar cooling practice can cause what’s called “inter granular macro segregation, it’s common on continuously cast steel too.
I’m not going to ramble on, you can etch the section of stock, it will often reveal this defect,
Mark

Milland said:

If you want, try driving the bushing down a male fitting on its own, the wedging action should finish the split.

Click to expand...

That is what I ended up doing but I twisted off the square portion of the male pipe plug and it did not split completely apart, the crack propagated further down the length of the fitting as you would expect. I had to saw the opposing side and then I could pull apart the fitting. The material seemed fairly malleable and it did not split apart with ease.



The material is clearly oxidized at the crack, I left the parts installed for about 3 weeks after the leak and they were exposed to water but 0 pressure.

L Vanice said:

Kyle, you never said what range gauges you bought. If your system is normally at 45 psi, you should use 90 0r 100 psi gauges. The rule is, use a gauge that ranges twice the expected pressure. You never want the gage to peg if you have any interest in reading the pressure and thinking the reading is accurate.

Click to expand...

You are correct Larry, I wrongly assumed the water pressure was around 35-40psi and I used 60 psi gauges. I bought 100 psi gauges to replace them. The gauge companies usually tell you the first 20% and last 20% of the scale is unreliable.

And I thought as long as you had a Scotch-Brite wheel and plenty of Teflon tape you can make leak free low pressure water, air, and gas connections. Is there a place to buy high quality off the shelf fittings these days? The big box stores seem to have nothing but low quality junk from China.

As for the original issue I agree with those voting bad bar stock. The pictures don't show evidence of over tightening.