Detroit Diesel Heat Exchanger Failure

Do you have any interface with domestic water, hot water tanks etc ?

The only thing I could conceivably think of would be A LOT of crankcase gases in the coolant system. Fuel in the coolant system from a fuel cooler failure or something and the heat ignited it? I really do not know as this is a new one on me.

Is it possible to tell which side failed ? salt or fresh.

can you post pictures?

Hi,

Do you have any idea what the temp was just before it fell apart?

Did the Hi Temp Alarm go off after the failure?

I would test the Hi Temp Alarm Switch and Temp Sender before re assembling and hoping everything was good.

Hi,

Is it possible that the SW outlet was somehow blocked or closed?

The Jabscos are a positive displacement pump and do not take kindly to running against a closed valve.

The whole system might have got to a pretty good old pressure and as there is no relief on the SW side, the weakest part failed and the ensuing explosion of water flashed off into steam to an extent on contact with hot surfaces and the flying parts caused some damage.

Also, if the SW goes to the exhaust you would have had this entering the fray once the system was open via the spray ring or whatever injection device is used on your application adding to the mess.

The Jabsco Impeller would not look good after 35 mins of running against a closed valve.

That's impressive.

It looks like your seawater flow failed and the coolant overheated, did not relieve or alarm and the tank ruptured. As K1W1 suggested, the hot coolant flashed to steam and created an overpressure that seemed like an explosion.

The FW side is sealed tightly but the cap should have relieved at least part of the pressure, and your hi temp alarm should have worked. A flow switch in the raw water side might have been a good investment given the propensity of Jabsco impellers to fail.

That tank is a fairly thin casting and had probably been subjected to quite few pressure cycles over its lifetime so I would chalk that bizarre failure up to metal fatigue and bad luck given a bit of a boost by inoperative protective systems.

What you need to determine next is what damage may have occurred by the overheating as it must have been fairly substantial. Also, find the impeller blades before they contribute to another overtemp condition.

15 minutes interval between checking gauges seems like a recipe for disaster.

I would carefully check th exhaust system before running that engine after the HE is replaced as things must have gotten very hot in the there with no flow.

Impellers rarely fail suddenly, either they fail over a time period or something else makes them fail suddenly. Sometimes, slowly rising gear temperature will be the first thing to show limited raw water flow or an impeller which is shedding bits and pieces. At least this has been my experience on cat 3412s

Has any cooling system maintenance been performed on that engine? Strainers? I ve had two instances where a raw water pump failed to reprime after the system was opened. In one case, the coolant never got hot enough as we were running slow on the way out but a cat mechanic down below touched the pump cover and realized it was burning hot (sea trial after maintenance). I now check the raw water pump cover for temperature after most cold starts and after haul out or strainers clean up and monitor coolant and gear temp for even rise after start up.

That damage looks to me like what you would expect from freezing. I agree with Marmot, it is impressive.

The damage is impressive I agree. BUT the shape of the piece that came out. It really looks like that piece was cut out previously and welded back and ground smooth and painted for some unapparent reason. It's strange how the edges are so straight and not jagged. Unless the part that stayed happens to be a thicker part of the casting. I've never seen anything like it on a DD heat exchanger housing.

It is hard to tell from the pictures.

I really would like to seem some close ups of the cracks and breaks.

The edges of the metal do not appear to have failed from excessive interior pressure.

I would expect a failure of a hose clamp or hose before the HE fails from high pressure.

Is it possible that metal fatigue or electrolysis cause the metal to fail at a fairly low pressure?

How effected was the floor above? Is there significant damage and movement of the floor?

Impressive

15 minutes between checking gauges is impressive, not checking water flow after "opening up the system" is even more so !.

I would check the lengths of the bolts that secure the pipe support on the front of the heat exchanger. If these are too long and are "bottoming out" they could cause stress cracks in the casing.

Although the photos are kind of hard to tell but there are small stains on the edge of the broken piece but not all the way across which may indicate cracks have been there and unfortunately during this run the casing let go.

Most temperature senders will not alarm if there is no water and in one case I was involved in the DD got so hot the wiring harness went up in flames, burnt all the way up to the fly bridge, Halon system went off, shut downs did not work and engines sucked the good old halon in resulting in two new engines and two new generators.

The cause in that case was the complete failure of a fresh water hose which dumped all the coolant, no water no alarm

Good luck.

Please send close up photos of the welded casing which failed lol

It is that invisible welding and the disguising of the grinding on the inside that really gives away that process isn't it.

Its Cast Iron which is actually quite brittle if mis treated, beat one with a Sledge Hammer and you will see similar results.

Welding indeed. It was obviously done by laser proton stir welding in a NASA vacuum chamber them micromachined internally in order to disguise the old repairs and make the failure look exactly like the type one would expect to see when a thin casting with a large flat surface is over pressurized.

I'm quite familiar with cast iron. Cast Iron is pretty strong when done right, especially the new MIM process. It definately is not nearly as strong as a piece of forged metal, but you couldn't create a tank of that shape out of forged metal back in the day, or at least a cost effective one. With today's CNC machines it could be done fairly easily out of forged, but you'd still have to have a weld somewhere due to it's shape. However cast iron does suffer from porous castings and weak castings when not done right, which is evident on the top edge where it's rusty and appears water has almost been seeping through the casting there for years. I said it almost looks like it was cut and welded. The other question is why didn't a rubber hose blow first or the relief valve on the cap, but I guess the casting was weak to begin with and 2 decades of age set in. I've never seen a 16v71, but have seen many 16v92's and even a 24v92 which was an engineering disaster on DD's part.

Woww I’m extremely worried now Al-Qaeda infiltrates in Detroit engines…I will put some extra surveillance on my 6-7's

How cold was it when you started? When this happened were you still in a river? Water temp?

With allot of imagination please bear with me;
Old stressed tank.
Water temp coming up. Expanding cast.
Then a blast of real cold sea (river) water. (cold eddi) Contracting cast.
Then warm water again. Expanding (exploding) cast/tank.

It may not have been allot of pressure, just allot of thermal flexing.
Note the heavy cast areas are intact. Slower to react to quick temperature changes. The thin areas that could have taken a drop or hit 20 years ago or even a cast blem, then a drastic thermal.
Q C on this tank when new involved what? Eyeball and it held water. Nobody x-rayed this after it was made. No audio or ping on it.
Yes modern technology is a whole different world on HEs now. There was not much high tech (todays standard) around when this tank was made.
Find another one, bolt it on and go back out.
As these ole Detroits (and others) pick up more years, we may start seeing more bazaar failures like this.