Are Certain Injector Positions More Prone to Black Death?

[talkinghorse43]

Thinking more about the occurrence of injector seal leaks and hold down bolt failures (aka black death), it seemed to me that most cases reported in this forum were from injectors near the front of the engine. If this is in fact the case, then this might mean that the working hypothesis that black death is due to low quality fuel (low cetane) and overpressure spikes during combustion might not be the correct hypothesis. So, to find out if most cases reported were from injectors near the front of the engine, I embarked on a slog down memory lane. I say slog because the only way I could be sure I had found most of the reported cases was to look at every post containing the word "injector" - there were 153 threads containing that word; now 154. I might have missed some because the search engine does not find "injectors" when you search for "injector". Anyway, the results are listed below (in no particular order) with the poster's screen name and the injector # involved:

Steady Eddie - 2
topless - 2
hkpierce - 1 (DAD reading - not Black Death) [HKPierce edit]
cerickson - 2
abittenbinder - 1
solera - 1
abittenbinder - 1
xenox2 - 1
brandon - 2
mean-in-green - 1,2,3
SteveD - 2
t-man - 1,4
05highroof - 1
abittenbinder - 2

Altogether, 14 reports of the black death, and you don't have to use statistical analysis to see that my observation about the black death usually occuring at the front of the engine is correct. Admittedly, this is probably just a small sample of all that have occurred and the sample could be skewed, but the results are pretty overwhelming. Personally, I can't imagine why poor fuel would cause the front injectors to leak and not the back. Unless someone can come up with other data, or explain how poor fuel could force this result, I'd say we need to find another hypothesis about why black death.

If this does hold up, then I can think of another hypothesis that could explain this black death observation and that's rusting of the injector body resulting in locking the injector to the bore in the head. This would force the injector to move in lock step with the head as it expands and contracts with temperature cycling, resulting in either lifting the injector off the copper ring seal, or compressing the copper ring seal and causing a leak on cooldown. This rusting has already been reported in this forum to be more prevalent in the front injector wells by sikwan and jim_wildgoose. Also, inspection of my own engine shows that to be the case as well.

 

[Altered Sprinter]

Consider the fuel factor of unburnt fuel.Have you taken into evaluation as to testing the spray patten of the fuel eg. mis alignment of spread rate to required volume>the spray patten of the injector. Good points though
PS first and rear injectors are as bad as each other it depends on which engine HP output. for the US side number one seems to be the obvious one, but rear also has shown to be a troublesome one as well.
Richard

​

 

[talkinghorse43]

Consider the fuel factor of unburnt fuel.

I don't have the foggiest idea of what you're writing about here, please explain your comment.

Have you taken into evaluation as to testing the spray patten of the fuel eg. mis alignment of spread rate to required volume>the spray patten of the injector.

Please explain how the spray pattern could impact the occurrence of the black death in the front of the engine.

PS first and rear injectors are as bad as each other it depends on which engine HP output.

How does engine HP output enter in?

for the US side number one seems to be the obvious one, but rear also has shown to be a troublesome one as well.

Except for MIG & t-man, I haven't seen any other reports of rear injector involvement - have you?

 

[mean_in_green]

...I embarked on a slog down memory lane.

Well ten out of ten for taking the time to do this, and what a curious thing to have realised.

 

[Ciprian]

Very interesting observation. I opened my engine cover just to check it out and no black death yet (knock on wood) at 261k, but I see some rust down there, and more of it in the front of the engine around injector 1 and 2.

 

[talkinghorse43]

Anticipating that we might finally find that rust is a problem, I've already treated my injectors by injecting about 1cc of clean engine oil into each well (will inspect soon to see if this amount is adequate). From a rust breakdown/lubrication standpoint, a penetrating oil would have been better, but those I'm familiar with have pretty high vapor pressure and would probably disappear quickly at operating temp. Oil should also keep more rust from forming. Also, I used Amsoil's synthetic 5w40 ECF (had a partial qt on hand and no other use for it) because I wanted something that would be heat stable and not polymerize with time into a gluey mass and be as bad or worse than the rust. Probably, over time, it too will vaporize and disappear, but I'll just continue to dose it as needed.

 

[hkpierce]

While I found a picture, I was not reporting any trouble with mine. In fact, the picture of mine is of a healthy set of injectors: http://www.sprinter-source.com/forum/showpost.php?p=8290&postcount=4

 

[talkinghorse43]

After my slog down memory lane, with all the injector info fresh in my head, I began to realize that we could probably estimate, by simple calculation, the combustion spike pressure required to lift an injector off its seal and cause it to leak. My reasoning here is that if the required pressure were at all reasonable, then the poor fuel and combustion spike hypothesis could possibly still have legs (still can't personally imagine how it would select the front injectors). If not reasonable, well, that would be one more nail in the coffin. I'll lead you through this calculation here step-by-step, but first, I need to credit Doktor A (through his informative posting) with providing the key info I will use in the calcs.

The injector hold down bolt is a 6mm dia, metric grade 8.8 bolt and it's torqued to at least the material's yield stress of 640 MPa. Assuming it has been torqued to just the yield stress, the stress on the bolt is 92,821 psi(640 E06/6.895 E03). The total load exerted on the pawl holding the injector against the copper ring seal and the head by the hold down bolt is 4067 lbs (92821 psi x 0.0438 sq in). The pawl is supported on the head on an embedded ball and the hold down bolt is about equidistant between the ball and the injector bearing surface, so the load holding down on the injector is about half of the total load, or 2034 lbs.

The diameter of the injector hole in the head into the combustion chamber appears to be the same as the diameter of the hold down bolt bore above the socket threads, or 8mm. So, the pressure in the combustion chamber would have to rise to a sufficient level such that the cross sectional area of this hole would generate enough force to counteract the hold down load of 2034 lbs. Dividing 2034 lbs by this cross sectional area of 0.0779 sq in yields a combustion chamber pressure of 26,106 psi required to lift the injector and make it leak.

This pressure seems unreasonably high for a pressure above the piston, more like injection pressure. Just for grins, I'm wondering what the torque on the crankshaft would be with this level of pressure? The wiki says the bore of my 612 engine is 88mm, so that's a total force down on the piston of 246,000 lbs (26,106 psi x 9.43 sq in). The wiki also says the stroke is 88.4 mm, or the moment arm is half of that or 1.74 in. 1.74 in is 0.145 ft. Torque at that load is 246,000 lbs x 0.145 ft, or 35,670 ft-lbs (eat your heart out Duramax!).

These numbers seem wildly out of whack to me, but I can't see what could be wrong with either my calcs or assumptions at this point (someone please check!). If they are correct, then no way could bad fuel cause an injector to lift off the seat.

Again, if correct, the only way I can see for sufficient force to be generated to overcome the hold down load is through constrained expansion/contraction. These are atomic scale movements and can generate enormous forces if not free to move. The injector would have to be stuck in the bore in the head by some mechanism. Of course, I'm still thinking rust.

 

[talkinghorse43]

While I found a picture, I was not reporting any trouble with mine. In fact, the picture of mine is of a healthy set of injectors: http://www.sprinter-source.com/forum/showpost.php?p=8290&postcount=4

Right, it was the picture you found.

 

[abittenbinder]

Thinking more about the occurrence of injector seal leaks and hold down bolt failures (aka black death), it seemed to me that most cases reported in this forum were from injectors near the front of the engine.

If this is in fact the case, then this might mean that the working hypothesis that black death is due to low quality fuel (low cetane) and overpressure spikes during combustion might not be the correct hypothesis.


If this does hold up, then I can think of another hypothesis that could explain this black death observation and that's rusting of the injector body resulting in locking the injector to the bore in the head.

Jon, I have observed and previously reported on typical progression of injector seat leakage from forward cylinder position rearward. There have been some, but as you observed few, exceptions.

Having removed many fuel injectors for a multitude of reasons, the most common being actual injector operation issues, seat leakage issues, or as part of general engine dis-assembly, I can report I have never encountered corrosion severe enough to cause seizure of the injector body in its bore.

Any corrosion I have seen on the injector's shank was superficial but I rarely see even that. The injector body is composed of high grade steel and is machined so as to allow only a narrow (approx 10mm wide), raised circumferential step (near top of bore) to contact the cyl head bore. Even for this narrow profile step, the clearance is generous, approx 0.10mm clearance to bore.

The remainder of the injector body has over 0.5mm bore clearance and I have found that any difficulty in injector removal has always been from baked deposits due to leakage.

My thoughts on causation and remediation of injector seat seal leakage have evolved somewhat. I now view seat leakage as analogous to human aging. If you live long enough, your chances of prostrate problems increases. I believe low grade, chronic seat leakage is likely inevitable given enough engine operating hours. If ignored, serious and/or premature cases progress to "black death".

I have found evidence of low grade seat leakage (defined as-not visible in the injector gallery) in many moderate to high mileage engines.

I still believe seat leakage to be related to cyclical stresses on the hold down hardware, whether it's bolt stretch, bolt breakage, or female thread stripping in the cylinder head bolt's bore.

Serious leakage (aka, black death) progresses more rapidly via leak induced erosion paths in the copper seal ring

Why the typical progression from front cyl of engine to rear cylinder? I surmise there is likely a slight but uneven distribution of peak combustion pressures/temperatures.

The peak combustion pressures in a Sprinter combustion chamber approach approx 175 bar (about 2540 psi). Slightly higher than average combustion pressure may occur in front cylinders relative to rear for a variety of reasons.

Possible reasons? Uneven coolant temps due to head casting design, slight variations (toward retarded) in injection timing due to minute torsional twist of crankshaft (CKP sensor is located at flexplate, not at the crank pulley end of crankshaft) under heavy loading, and EGR flow distribution are some, however abstract, possibilities that come to mind.

The ECU does have a "smooth running" feature which varies the fuel injected at each cylinder to equalize rpm of each cylinder, but this feature is shut off above 2800 rpm.

Prevention? After observing the much more common, low grade ( not visible above) seat leakage, I may begin recommending pre-emptive injector seat seal replacement as an adjunct to regular maintenance on high mileage Sprinters. But keep in mind I have the advantage of being prepared to deal with possible bolt breakage or stripped hold down threads in the extremely difficult to access, cyl head's hold down bore. Doktor A

 

[talkinghorse43]

Thanks Andy, more great info!

The injector body is composed of high grade steel and is machined so as to allow only a narrow (approx 10mm wide), raised circumferential step (near top of bore) to contact the cyl head bore. Even for this narrow profile step, the clearance is generous, approx 0.10mm clearance to bore.

Ouch!, the worst possible place for a narrow (to me) gap. Sticking here maximizes the potential amount of differential movement/force between the head and the injector during temperature cycling. Having been raised with the English system of units, I still have trouble picturing metric units. This 0.1 mm gap is 0.004 inches - remember that the thickness of the proverbial human hair is 0.003 inches. Seems to me it wouldn't take much rust at all to bridge that gap.

The peak combustion pressures in a Sprinter combustion chamber approach approx 175 bar (about 2540 psi).

If my estimates are correct, a cylinder pressure well below the capabilities of the OEM hold down system to contain - an order of magnitude (10x) below. If I read it right, this reference suggests that a fastening system that is never stressed near/above its preload is immune to fatigue.

http://www.rbwmfg.com/images/HelpfulHints2.pdf

 

[dronsin]

I am truly amazed at the technical competence of the discussions on the injector hold down issues that I have just read. Lots of facts and accurate interpretations, not some guesses and assumptions. The quantitative work by talkinghorse and Dr A are just what good engineering is......trusting math and physics to understand things that cannot be seen or related to.

I have been suspicious of the hold down fork. Using a leverage clamp with only a fraction of bolt load to do its intended job seems poor design....at least the clamp bar should be longer on the backside to get more like 75% which would add a great margin for this apparently sensitive detail. I have not looked closely to see if there is any space for an extended bar.

Your discussion of movements and internal stresses due to temperature changes, and differentials is truly an important issue. As a designer, it is confirmed that a properly tensioned bolt clamp application will put the fastener in a fatigue proof condition....provided the preload is maintained. ( and includes an analysis of the elasticity of gaskets and all components)

One of the designers methods to insure that is to use a LONG bolt which provides more stretch in an elastic manner to allow for some losses in seating of interfaces. That said, I could see a sleeve and longer bolt as a way to attempt to cure some of these leakage ills. One has to wonder how some short 'grip' bolts can remain tight!

Any takers?

Dale

 

[Eric Experience]

The injector clamping and thermal cycling is an issue. Also the type of seal is important, I am trusting that mercedes have chosen the correct alloy. If the seal was copper as some suggest then the aluminium head would corrode and oxidise where it contacts the copper if it had any water present, If a motor was washed with a pressure washer the front cylinder would get wet and cause corrosion, this oxide would force the injector up and stretch the bolt. BUT Mercedes have good designers and they would use an alloy that is compatible with aly and steel. The japs use an aluminium washer in this aplication, this causes corrosion of the aly and steel where the aly touches the injector. This copper aluminium problem was solved by a refrigeration company so they could use aly tubes in the freezer and copper in the compressor. This same alloy used in the splice is the ideal material to make injector seals from. I would be interested in hearing if the people who have had problems have washed there motors. Eric.

 

[talkinghorse43]

I am truly amazed at the technical competence of the discussions on the injector hold down issues that I have just read. Lots of facts and accurate interpretations, not some guesses and assumptions. The quantitative work by talkinghorse and Dr A are just what good engineering is......trusting math and physics to understand things that cannot be seen or related to.

I have been suspicious of the hold down fork. Using a leverage clamp with only a fraction of bolt load to do its intended job seems poor design....at least the clamp bar should be longer on the backside to get more like 75% which would add a great margin for this apparently sensitive detail. I have not looked closely to see if there is any space for an extended bar.

Your discussion of movements and internal stresses due to temperature changes, and differentials is truly an important issue. As a designer, it is confirmed that a properly tensioned bolt clamp application will put the fastener in a fatigue proof condition....provided the preload is maintained. ( and includes an analysis of the elasticity of gaskets and all components)

One of the designers methods to insure that is to use a LONG bolt which provides more stretch in an elastic manner to allow for some losses in seating of interfaces. That said, I could see a sleeve and longer bolt as a way to attempt to cure some of these leakage ills. One has to wonder how some short 'grip' bolts can remain tight!

Any takers?

Dale

The more I think about the situation, the more I am comfortable that MB designers provided their usual bulletproof design here - at least as far as resisting combustion forces (not so for rust). A 10x safety factor is substantial. But, there are still some questions that could be put to rest (more or less) if we could get more data. The hold down bolt is fairly long, but, for one, I would like to know what the length (in a typical installation) is between the top of the threads in the head and the bottom of the bolt head. For another, I would like to know the thickness of a typical copper seal ring before and after a typical injector installation. Also, I would like to know some dimensions of the hold down pawl, specifically the distances from the center of the ball socket to the center of the hold down bolt and from the center of the hold down bolt to the contact of the pawl with the injector body. These pieces of info should allow a much better estimate of the actual preload of a typical installation. Other hold down system measurements that would allow us to get a better picture of how preload could change during thermal cycling would be good to have, but would require much more work. I have already gone through the mental exercise of examining this hold down system for response to thermal cycling and it seems to me that its elements were designed to be immune to thermal cycling so that preload is basically unaffected. Creep of the copper seal ring over time is another possible problem (unlikely in my view), but maybe that could be answered by a literature search. If we could show that hold down preload is always much higher (absent rust) than any conceivable combustion forces generated, then the observed fatigue failures would have to come from another source.

The design change you suggest is a possible solution, but I think there is much more to be learned about the capabilities of the existing system before we venture into uncharted territory, Anyway, if it turns out that rust and injector sticking in the bore is the root cause, no hold down system could be guaranteed to resist those forces.

 

[abittenbinder]

Ouch!, the worst possible place for a narrow (to me) gap. Sticking here maximizes the potential amount of differential movement/force between the head and the injector during temperature cycling. Having been raised with the English system of units, I still have trouble picturing metric units. This 0.1 mm gap is 0.004 inches - remember that the thickness of the proverbial human hair is 0.003 inches. Seems to me it wouldn't take much rust at all to bridge that gap.

Jon, you're missing the point here. 3 points, actually.

The long injectors need to be supported laterally in their tall bores. The engineers minimized the surface area of that toleranced upper support step on the injector housing. The minimal surface area prohibits binding/seizure in the bore.

Your barber shop analogy aside, the injector body narrow step to bore clearance of 0.10mm is quite generous (in context of engine design) particularly when combined with the expansion characteristics of the aluminum bore and the corrosive (spalling induced) expansion resistance of the alloy steel injector housing.

Most importantly, and this is direct from the trenches, there is no corrosion visible nor any increase in diameter (due to corrosion) present in the injectors removed to repair seat seal leakage. This includes my own large data base and the fleets I consult with. Doktor A

 

[talkinghorse43]

Jon, you're missing the point here. 3 points, actually.

The long injectors need to be supported laterally in their tall bores. The engineers minimized the surface area of that toleranced upper support step on the injector housing. The minimal surface area prohibits binding/seizure in the bore.

Your barber shop analogy aside, the injector body narrow step to bore clearance of 0.10mm is quite generous (in context of engine design) particularly when combined with the expansion characteristics of the aluminum bore and the corrosive (spalling induced) expansion resistance of the alloy steel injector housing.

Most importantly, and this is direct from the trenches, there is no corrosion visible nor any increase in diameter (due to corrosion) present in the injectors removed to repair seat seal leakage. This includes my own large data base and the fleets I consult with. Doktor A

A corporation, like MB, would have the resources necessary to fully evaluate these issues (complete calculations addressing all aspects, microscopic studies, etc.), but we are mostly just owners trying to minimize repairs costs and any analysis we can do will just amount to hand waving at worst and circumstantial evidence at best. Personally, I think the rust hypothesis has enough legs for me to attempt to mitigate it by keeping engine oil in the injector wells and, possibly, try to do a better job of sealing the plastic cover to cut down on water intrusion. Others, of course, are free to make their own decisions.

 

[talkinghorse43]

Old thread, but resurrecting it to report that the black death just bit me too. About 5k miles after I last checked under the cover, we were returning from a trip to Nashville and when we pulled up at home; diesel smell in the cab. Went to the post office the next day and diesel smell again. Pulled the cover with the engine off and the smell hit me in the face - found some deposits around the base of #2. That one was replaced by the dealer under warranty @ 99k, so about 170k miles elapsed. The first photo shows the deposit. Thought about it for a while and decided to try to ask the good Doktor to handle it since I don’t have all the tooling that could be required (lucky that it happened on the way home and luckier yet that the good Doktor is local to me). Contacted the Doktor and he thought he could work me in. About 10 days later (no driving in the meantime), he called and told me to bring it in, but that I wouldn’t be able to watch because he would have to work it in at odd moments between other work. A couple days later, I picked it up and he told me:

• The injector was stuck in there good and resisted most of his attempts to get it out (heated it up to operating temp – wouldn’t come out, drove it hard a couple times with the hold down loose – wouldn’t come out, finally got it to move a little and was gradually able to extract by working it back and forth).
• Copper seal wasn’t grooved by escaping gases.
• Injector wasn’t rusted

In the context of this thread, this seal failure was near the front of the engine like most of the others. But, I have been adding effective (better fuel economy, less soot in oil & EGR) cetane boost religiously (the good Doktor’s guess is it’s fuel quality-related) and also have been maintaining engine oil in the wells (my guess is it’s corrosion-related). So, it shouldn’t have happened.

But, it did. I’m still thinking it might be corrosion and I’ve been noticing over the years I’ve been adding oil to the wells, that after sitting in the rain, water collects in the cover hold down screw recess right next to #2 (second picture shows the water after a rain, third picture (after repair) shows that a leak of water here would find #2 well). That observation hasn’t worried me ‘til now because I assumed the oil would stop any corrosion. The injector wasn’t rusted, so the oil stopped steel corrosion, but how about corrosion of the aluminum head? Since aluminum forms a stable oxide coating (doesn’t flake off like steel), corrosion of the aluminum head in the well could reduce the diameter of the well and seize the injector in place. If seizing occurs, then the copper seal is bound to fail due to the differential expansion/contraction of the head/injector. I’m still guessing here, but, since this occurrence, I’ve decided to pay more attention to the water collecting on the top of the cover to try to stop it from getting to the injectors.

So, the fourth picture shows a rain hat I installed to keep rainwater from dripping onto the engine cover. I haven’t had the intake manifold recall done on mine, so I don’t know if the new OM612 manifolds have the mounting holes I used on mine, but there are two 6x1 mm threaded holes I used to fix this piece of thin aluminum sheet I had in my scrap pile. I also added a short strip of Velcro to the cover to keep the sheet from flapping around. The fifth picture shows the hat after some miles of driving and days of sitting with off and on rain and shows that it has definitely been effective in diverting the rainwater. Time will tell if it will stop black death.

 

[jdcaples]

Thank you for the update!

I like the rain hat.

Will you continue to add engine oil to the injector wells?

I read your posting a couple of times but I'm not clear about what was holding the injector in the injector bore.

Maybe I should ask, "what two surfaces were creating the friction to lock the injector in place?"


-Jon

 

[talkinghorse43]

Will you continue to add engine oil to the injector wells?

I will continue since it seems to protect the injectors.

I read your posting a couple of times but I'm not clear about what was holding the injector in the injector bore.

Maybe I should ask, "what two surfaces were creating the friction to lock the injector in place?"

Andy B thinks it was the black death "varnish". I think it could have been the reduced diameter of the well due to corrosion of the alumimum of the head. If "varnish", it would be like glue holding it in the well. If corrosion, it would be the clamping force of the corrosion trying to grow thicker. Unfortunately, I wasn't able to be there to measure the diameter of the well to put this question to rest and I didn't want to encumber Andy with another task in his very busy schedule.

 

[northener]

I went back to post 1 and reread them all. I noticed in Dr A's number ten post towards its end the information that over 2800 rpm's the fuel pressure equalizer function cuts off.

My question is are engines given this workout more prone to black death? Paul