Faced the header flange; in this case a side cutter worked better than a fly cutter. The flange was warped really bad, nearly 0.125″ which explains why some of the original profile was lost. After that, holes were drilled and threaded for the exhaust gas temperature (EGT) sensors. The holes look kind of goofy but it’s where they had to go on the different primary tubes in order to end up with the sensor tips equidistance from the head. The Interweb has all sorts of conflicting information on how far from the head the sensors should be placed. The more reputable information seemed to suggest 1.25″, so that’s what was done. Also, the same sources noted that typically, the sensors are used only for tuning or calibration and not left in place. On the other hand my brother said EGT sensors are always left installed in aircraft, so I’m not sure about the context of the original information.
In the last two pictures, that giant lint ball with legs is my buddy Lee’s cat, and there’s Midi on his way to see Mom and her dog, and my brother’s Boston Terriers. They have a great time chasing each other around the yard.
Some late news: Right now Lulu, the book printer, is running a 20%-off everything sale, discount code “BASKET20.” It’s only good through midnight Sunday (tonight) but if you miss it keep checking; I’m pretty sure they’ll be running discounts over the next month or so due to it being tax (refund) season.
I posted pictures of the failed parts on a forum. It’s always hard to to know how knowledgeable the responses are, but surprise, the theories aren’t much different than before, though the most popular suspicion is that a rod bolt broke.
In other news, the manifold is getting there but still needs ports for the EGT sensors, some touch-up welding, and having the flange fly-cut to eliminate heat distortion. Hopefully the way the single wastegate is adapted to the twin-scroll setup will work. I’m a little concerned because the wastegate tube joins at nearly right angles, which is a big no-no in gas flow. The exhaust flow not only doesn’t like turning sharp corners, but the passing exhaust flow will actually try to pull air in from the wastegate. Don’t believe me? Try blowing across the top of a straw in a cup of water and watch what the water does – it’s how siphon-type paint sprayers work. Hopefully it’ll work out because it’s so much more compact than the last placement.
And lastly, my buddy Kane edited down Midlana’s most recent Streets of Willow event, recorded from outside the car.
Finally posting pictures that are a bit more light-hearted pictures from the last event. My buddy Kane brought out a pretty cool device called a Pixel Stick, which allows creating these neat images seemingly floating in space. We had a fun time other than the car nonsense. Sorry I didn’t generate big clouds of tire smoke on the skid pad… I left the smoke later in the day and in fact if you look carefully at the last picture, there’s a trace behind Midlana – very expensive smoke.
Pulled piston/rods 1, 2, and 4. The rods, pins, bolts, skirts, rings, ring lands, and bearings look great – to me. Some have said the crowns show evidence of high heat but I really can’t tell.
One thing several people have said is that running 11 degrees spark timing at 220 kpa and 6000 rpm (where it failed) is retarded – literally – meaning it’s far too little spark advance. Alas – and again – Interweb searches show two very different groups, one saying that it should be around 20 degrees and the other saying it should about 10-12, right where it was. I don’t know enough to know what’s right, and it’s very frustrating not finding more of a common answer, but it matters less and less since this engine isn’t repairable and the tune for the new engine will be closely reviewed after it’s back up and running.
Currently waiting for the rest of the exhaust parts so the header build can start. The idea is to do that first so that the head (currently serving as the exhaust manifold fixture) can be delivered to the engine builder for cleaning, inspection and reuse. Quoted lead time on the engine is 10-12 weeks, so there’ll be plenty of time to rework both the exhaust and intercooler.
I researched filtration for ethanol systems and am pretty sure the fuel filter isn’t ethanol-capable – never dawned on me that it wouldn’t be since it’s a paper element. An Injector Dynamics article on filtration was pretty helpful and noted that “…6 and 12 micron microglass, ethanol-compatible fuel filters did well in our testing and are recommended”, so a Fuelab 6-micron microglass unit will be added. Hard not to wonder if this was at the bottom of it all – don’t want to think about it.
On a related note, the fuel pressure regulator is also suspect for the same reason. I bought it about 5-6 years ago used and who knows how old it was then. The concern is that E85-compatible aftermarket parts didn’t really start hitting the market until maybe after this one was manufactured. As a precaution, a known-compliant regulator will be used.
In other news, during the collapsing intake fiasco an engineer deep within Garrett offered to help out. He noted that the engine would gain performance and lower boost (lessening the risk of detonation) by going with a larger 1.06 A/R turbine housing over the existing 0.83 housing. His recommendation was noted but put aside after the scoop failure was found. Now, with moving the turbo closer to the exhaust ports, a new exhaust manifold is needed. Because turbo merge collectors can be tricky to make with thick-wall pipe, I wanted to use a cast stainless collector. Turns out that it’s hard to find divided ones for T3 housings housings but they’re widely available in the larger T4 frame, so the switch was made. The idea is that the less-restrictive housing will offset the larger A/R (0.83 A/R isn’t available), and that by keeping it divided, lag won’t be increased.
It’s interesting the widely (wildly?) differing opinions from people who really know this stuff, the latest being comments that the plugs look good, with no sign of detonation. That was yesterday. Today the injectors and test results came back – note that I didn’t tell them what had happened or which one was suspect. Looking at the results sheet, well son of a… (by chance, #2 on the sheet happens to be cylinder #2). I’m very surprised, given that a 10 micron filter is immediately upstream, though I don’t know how large a particle the injector can pass without clogging, what if it’s only, oh, 8 microns – I’ll have to ask them. Additionally, I’m surprised they only flow 1400 cc wide open because they’re advertised as “1650 cc” injectors at the same pressure. Not sure what to do with them – sell them and get… what in their place? How would different injectors prevent this from happening again?
The experts have settled into two groups. One says the failure was “absolutely due to detonation, with signs of high heat.” The other says “no signs of detonation, the rod big end or a rod bolt failed due to a manufacturing flaw.” The picture of the piston bottom shows virtually no heat, unlike what the pro-heat camp suggests would be present as a dark or black area (the gray area is where the destroyed rod hit the piston). Then there’s being told that running 11 degrees timing is nuts, that it should be more like 20 degrees or so. Mind you, I didn’t pick the timing values, the tuner did, though I’m still ultimately responsible for keeping an eye on things.
At this point, examination of the evidence is pretty much done and I’m left with the dubious knowledge that about all I know for sure is that one injector had a very poor pattern. Time to move on with the new build, intercooler, and exhaust setup.
I just realized something – that my fuel filter and regulator may not be ethanol-compliant. Now wouldn’t that be an expensive lesson if that’s at the root of this?