Wingman, noooooooo!

Whatever the cause, I'm very sorry to see this.

 

Whew, lot to unpack here. I'll start with the events leading up to and during the weekend, and then get to yalls(very well founded) concerns.

On the dyno had some serious misfires when upping the boost. Felt exactly like an ignition breakup. Was kinda weird as I've been told/seen stock coils do fine till ~800whp or so, a good bit more than I was putting down. As I've been rather short on time lately, I skipped all diagnostics and went right to the nuclear option: IGN-1A coils. Borrowed said coils off a friend's dead BP and grabbed a plug wire kit off of summit that made building wires pretty painless. Did a super ugly wiring and mounting job, took it for a rip, and was happy to see no more misfire, even cranking it all the way up to ~310kPa.


And then threw it on the trailer, and headed for Road Atlanta. 13 loooong months since this car or me has been on a track, so needless to say I was a good mix of nervous and excited to be back out. How rusty was I? Well, it wasn't until I was loading the truck and found a box of brake pads... and remembered at that point I needed to swap my meh wilwood street pads out for my track specific PFC's! That would have been an unpleasant surprise after 3-4 laps.
Anyway NASA SE has started splitting their TT groups into an "A" and "B" group... "A" group being the slowest half of the pack, and "B" the top half. Since I had no times logged in the last 12 months, I got defaulted to 21st out of 24 cars in the slow group. Fine, whatever, it was freezing out anyway and this was going to be a shakedown session regardless.

How cold was it? Well ambient temps were ~33* or so, and track surface temps were below that. I wasn't about to run my sticker hoho's in those lethal temps so I tossed on my set of VR-1's and left the car on wastegate power. Regardless, the track was so slick I came within INCHES of putting it into the wall.. ON THE OUTLAP

Yup, thats me... weaving around at the tail end of the group to build some tire temp. Gave it just "big toe" amount of throttle and

WHOOPP **** THATS THE WALL FUUUUCK ok ok we good no one saw that ACT NORMAL KEEP WEAVING

So yeah. Rest of that session I was super cautious about putting power down/leaning hard on the tires, for obvious reasons. I was fastest in session, beating a TT4 car by... less then a second Not that anyone set a "real" lap, entire field was 8-10 seconds off pace.
For all other purposes, the session went well. The car had no abnormal vibrations above ~130mph, temperatures and pressures stayed in check, and it felt pretty solid and confidence inspiring, even with the complete lack of grip everywhere.

Once I got back to the paddock, the issues started. I immediately noted a metric buttload of oil in the drivers side of the bay. Oh, catchcan overflowage, my favorite nemesis, had returned! I mopped it up and wrapped the breather in pigmat to help soak anything else that came out. With the temps climbing to a more reasonable ~50*, I swapped to my set of hoosiers and headed out for session #2. I had wheezled my way into the fast group, but due to some confusion in grid was sent out last instead of the appropriate slot in mid pack. Due to some not so heads up traffic and a few cars going off, I didn't get any clean laps and was mostly frustrations of traffic navigation, but the bits of clear track I got the car felt GREAT at full boost.
As I was coming into pits at the end of my session I noted something odd... my EGT sensor was reading ~190F. Even at idle temps are ~700F, and on the dyno I saw temps tickling ~1850F. I chalked the reading up to a bad sensor- maybe I had cooked it? Coolant was a little hotter than normal, but nothing crazy, and AFR's were still fine. I didn't think much of it past "bad sensor", but I think this was the first indicator for me something was wrong.
Inspections in the paddock revealed I had again spat oil over the bay, so I started to wonder if it was overfilled. I had added some extra the night before to help fight oil slosh, as is normal, but the day before had been hectic and I didn't really remember how *much* I had added. I jacked the car up and drained some out- of note here, what came out of the pan was ONLY oil. No water.

Third session everything unraveled. As I rolled though paddock to grid coolant temps spiked to ~230F and refused to drop. I assumed my fan had failed, and watched it like a hawk while sitting in grid- eventually it came down to ~220F, and then we were rolling out. Coolant temps refused to drop. As I approached the T11 bridge, I had a call to make- go to WOT and see if it cooled at speed, or abort the session and head back into pits. And in that moment, the engine completely cut out on me. I yanked the wheel over and coasted into the pit late. Engine refired immediately and I got back to my paddock spot without issue.
Started digging into the issue. Coolant fan was working. Initially I thought I had a failed water pump, as even when turned up to maximum it wasn't pulling normal amounts of suction on the feed hose and sounded weak. Then I popped the coolant pressure cap, noted I was low, and proceeded to pour a full gallon of water into it before I saw level. The second I stopped pouring the level started going down, and then we all noted the billows of steam out of the exhaust... which 100% smelled of coolant.

At that point I put the cold EGT's, weak sounding water pump, eating of coolant, and high coolant temp symptoms all together. Coolant sensor was reading 230* AIR temps inside the water housing, the pump sounded weak because it was pushing AIR, and EGT was ice cold because it was being steam cleaned. As if I needed more confirmation, I pulled the dipstick out.
It was milkshake

****

I loaded the car on the trailer feeling pretty confident I could remove and reseal the water plate that I thought was leaking. I planned to take it home, drop the milkshake oil, put in fresh oil, and then run the car super hot(EWP perks) to flash off any remaining water, and hopefully make it back out for Sunday.
Picture of the offending part, which I will point out, while is engraved, isn't anywhere near the water port, only where the breather seals to the outside world.


I let the fresh RTV cure overnight. Woke up at 4am to refill the system. Left the oil drain bolt out to check for leaks. None found with a full coolant system, so thought I was good. Filled it with oil, and juuuuust before I started loading it on the trailer, decided to crank it over and make sure it sounded fine.
*presses starter*
CLUNK
"Huh that's weird" *presses starter again*
CLUNK
"maybe the battery is dead?" *attaches jump box*
CLUNK
"goddam it work you-"
CLUNK
*spams starter*
CLUNK CLUNK CLUNK CLUNK

At that point I kinda woke up and realized something else was up. Pulled the plugs and tried turning it over then. Engine proceeded to shoot coolant onto my ceiling. And kept shooting it up there as long as I held the starter down without slowing down. There was a leak so bad, it was actively draining into every cyliender of the engine.

****

At that point I sent a few texts out telling people trackside I wouldn't be returning and the motor was toast, and went back to bed. It was 5am and it was already too much of a day to deal with.

 

No hate, you're not wrong for the majority of that. I have zero background in design or engineering(mechanic by trade), and my QA consists of me looking at said item a second time and doing a quick visual go/no go evaluation on it.
It had been in the back of my mind that an RTV seal might not be the best idea due to the harmonics/stresses that part sees, but this was honestly a "I'll redo this after the one event this year" when I had 4 months to fiddle with things and fix whatever else broke during this track shakedown. I was not expecting such a drastic and complete failure.

I feel like I should mention, I'm haven't fully finished my evaluation on what the hell happened here. While the plate is the leading suspect, the RTV seal looked intact when removed, and the water had dropped below the level of it and the water pump. I cannot, at this time, rule out the possibility that I simply lifted the head/cracked a cylinder wall while on track(as EGT's point to this happening in my final lap), then warped the crap out of the head by running with no coolant for 15-20min.

The vast majority of this car was built with the mindset of:
Up until the middle of this year, it was frankly the only way I had a shot at doing any of this. For the last 4-5 years my job allotted me ample free time and steady hours, but not a lot of extra income. If I could spend 4-5 hours building something that cost $500, it was a value proposition I'd jump all over.
Due to some job and company changes, I've swung to the opposite side of that spectrum- a $500 part really isn't a big deal, but finding 4-5 hours to fab something is difficult and gets chipped away at over two weeks instead of a single day.

One of their water pump plates is showing up tomorrow. My DIY wing mount plate will not be mounted on another motor.

Final photo before this thing rolled into the garage to sit on jackstands for god knows how long this time. I'd like to have it going again before hillclimb season starts in March but... that really depends on what I find inside the motor once I tear it apart, and what I decide I want to do with anything I can salvage.

 

Man... what a bummer. Glad you didn't put it into the wall on the outlap. I've done that in my sim racing league and it's pretty embarrassing. I'd have to imagine it would be 10x in your real car.

Seems like you're pretty in tune with the criticism/feedback you've gotten. While I'm not an expert by any means the RTV on that plate looks fairly intact.

I'm glad to see you're going to keep at it, some other people might throw in the towel and buy an NC (Sorry Z, I had to do it)

Two questions.
In your dyno video you had a bunch of steam coming from the engine bays during pulls, was that hot oil filling the catch can?
You mention a clunking when cranking over the engine. Was the engine completely locked up or was there debris in the cylinders?

 

First off, I sympathize with your **** sentiments.. It felt good to repeat it for you..

Secondly, great overall breakdown of the series of events (not being sarcastic...). I know I messaged you privately on this, but I wanted to share my detailed analysis based on your more specific summary for the rest of the group so that maybe someone here will learn something technical about root cause analysis of these systems that can be quite complicated to regress.

EGT's that cold are in indication something was wrong for certain. Always trust your sensors! EGT's can only monitor combustion performance upstream from them, so what could that cause be?

Well... based on what I read, it sounds like you were saying coolant temps were holding steady at around 230F. If by steady you mean, AN ABSOLUTE FLATLINE AND IT WONT BUDGE. That is extremely significant!

"Why is that significant, Matt?" well... I'm glad you asked. It is thermodynamically pretty much IMPOSSIBLE for a sensor to read a perfect flatline in a cooling system as fluid temp is ALWAYS moving during dynamic thermal inputs and outputs during changing vehicle operation. There are really only 2 circumstances where it can truly flatline. Case #1 is that the sensor itself has failed, but you were mentioning fluctuation in reading throughout the day so I take this as very unlikely (again, trust your sensors! They are more robust than you may think). Case #2 is that the flatline is real and that points to only ONE possible explanation. Your fluid is boiling at your sensor. As fluid undergoes phase change from fluid to gas during boiling, regardless of the energy input the liquid will stay at a CONSTANT temp while the gas rises in temp. Related aside, fluid temp sensors actually really suck at measuring air temp because the mass of the brass is a huge thermal heat sink that air just doesn't heat up and cool down that quickly. Fluid sensors will measure air temp, but in a very delayed manner (which still isn't flat). That's why IAT and fluid temp sensors differ.... anyways....

"But Matt, how could the fluid be boiling?" Well, let me keep going... Depending on exact formulation, most automotive coolants have a boiling point of around 218F (103C) at atmospheric pressure. Pure water is 212F (100C) @ 1 ATM. That temp is too easy to hit in common application, so we use pressure caps to increase the system pressure to 20+psi. Every 1psi of system pressure equates to an increase of ~3F in boiling point, so a 20psi pressure cap will add 60F of boiling point margin to your fluid, meaning that a coolant system that originally boiled at 218F @ 1ATM will now boil at 278F @ 20psi gauge pressure all because you put a pressure cap on it. I'm explaining all of this, because now we get to the fun part... what happens when the pressure cap has an issue or is faulty? well, generally pressure caps fail by only holding minimal pressure (<4psi) for a variety of possible reasons. Let's do the calculation at what the system boiling point at 4psi of gauge pressure is for coolant.. it is (4psi * 3F/psi) + 218F = 230F.... Does that number look familiar to you? It should. Read Wingman's notes that I quoted again.. I'll bet him a nice lunch that he wasn't reading an air temp, but rather, boiling coolant at the sensor.

I've learned to not believe in coincidences in this industry. I am very interested in what you diagnose, but based on your description so far my bets are on this rough order being the strongest likelihood.
1. I think it's possible that you could have lifted a head or compromised the head gasket surface during the 310+kPa excursions while testing coils. The misfires you had on dyno also seemed pretty harsh. I'm not sure how long you did runs with misfires, but (while uncommon) you can warp heads by misfiring bad/long enough as temp isn't consistent across the head.
2. With a compromised gasket surface, pressurizing the breather system is possible and would cause the catch can issue (your first noted issue)
3a. Further operation would likely further compromise the gasket surface and cause the observable EGT temp if coolant leaked into the cylinders.
3b. In a similar manner, it's possible that the combustion pressure starts pressurizing and overheating the coolant. Hard to say if 3a or 3b would be first (or if they are both happening on different cylinders simultaneously)
4. A weak pressure cap combined with 3b behavior would cause an observable ~230F flatline as the coolant is now boiling at the sensor.
5. At this point, **** is going downhill fast... Coolant and oil could be mixing at gasket surface (creating the first milkshake). Coolant is boiling while also being further pressurized by combustion fumes in the water circuit. That circuit may have found the water pump plate as the path of least resistance to relieve pressure, therefore creating a second path for coolant and oil to mix (while the boys are already in the yard with milkshake #1).. The water pump quickly runs out of fluid to physically pump as combustion gases and evaporated water fills the system..
6. Leak paths at pump plate and gasket surface won't let you refill it without instantly draining to the oil pan or cylinders and the water pump sounds weird because it doesn't have fluid in it because of aforementioned leaks.

I'm 100% willing to eat my words on this and be wrong... These puzzles are not easy to solve even with data logs, but the above story is the most feasible chain I can think through based on the order of everything you've said. If the custom pump plate started this all, you would have just drained all coolant into the oil pan and that would have been the end of it. It could possibly explain the breather puking out, but it wouldn't explain the EGT observation and I don't think that explains the holding constant temp at the coolant temp sensor either.

I'm happy to hear you bought that Tractuff plate. As additional insurance for the future, I'd recommend running a coolant pressure sensor in conjunction with coolant temp. Those two readings are very powerful signals to monitor together. 230F flatline at 4psi is more concerning than peaking 250F at 18psi as just one example of where monitoring just temp only tells part of the story. Hope this breakdown helps...

 

Jeez, sorry to hear Wingman. Sounds like a shitty turn of events.

My only advice would be I share the sentiment of some of the others in this thread. I know you wanted to get the car running and hit an event this year, but what you did in such a short period of time is a very tough feat. Next go around I'd say just take your time a bit and ease into it. You changed a whole lot of things around, brand new engine, and did a full send after not too much shake down. While many may joke around with the "just send it" line, you have a much smaller margin of error pushing gobs of power on a new engine (and other changes). There's a fine line between being a p**sy and being prudent.

As Padlock mentioned, more sensors to monitor things is always nice. A coolant pressure sensor with a dummy light is definitely a must for your car. With how cheap sensors are and how hard you're pushing this motor it probably wouldn't hurt to consider a few others like say a crankcase pressure sensor so you can monitor more things. I probably fall more on the p*ssy side of the line, but I'd also say easy into the boost levels next go around and see how the car does/feels. Do a day at 10 psi, check logs, make sure everything looks good, then 15, etc. Easy to say this stuff after the fact so I'll shutup. Hopefully the damage isn't too bad and you're back up and running soon.

 

I just want to comment here again, how awesome it is to have actual engineers chiming in with their best shot at an analysis of events after something like this. Exactly why I love this site.

In for teardown pics from the Wingman. For science!

 

Same. That theoretical autopsy, while not necessary what happened, is still super interesting to read. I get so tired of people on the internet who have no hands on experience regurgitating the same information over and over. It is useful for certain things, but I'd take 1 post from someone with experience over 100 posts of people pretending that they know what they're talking about.

 

Looks like the big brains here have beaten me to everything I had worth saying.

I gotta say, between the events that happened, constructive criticism from the engineers on here, the response, and breakdown of the weekend, this has been the best reading material I've had in a while. For whatever that's worth.

Also, I think we've all "been there" at some point to some extent, but waking up to run your motor at 4am then going back to bed at 5 after finding it hydrolocked... f*ck. I felt that so hard. That must have seriously sucked.

Additional cats given for starting your first post with almost binning it on the out lap haha. Way to own that sh*t

 

I’ll 2nd that, great reading here, sorry you’re the one that had to learn from it all!

as for the cold tires, your tires wrap around 4 ovens, pre-heat them. Drag your brakes like crazy, this creates a lot of heat and warms the tire from its core out. Weaving is great to keep surface temps up and prevent the core from cooling as much, but it’s not a great way to warm the tire, if that makes sense.

and it might be a bit safer

 

What does this sensor tell you that others already don't considering he's vented to atmosphere through a catch can?

 

I'm interested too. AFAIK, OEMs (Ford at least) have been using these for a handful of years to tell if all of the crankcase vent/pcv components are connected and sealed. We get check engine lights/codes on our cars when the oil cap or dipstick aren't fully seated, or if there's a substantial leak from a cracked valve cover. Funny enough, check engine lights and codes come on most often for the sensors failing themselves, but I digress.

Even with a VTA system, I'd imagine a well-placed crankcase pressure sensor would be useful to tell you if you're getting excessive blowby. I'll await the engineers to one-up my response.

 

Ya, I think it'd be helpful to check for excessive crankcase pressure and it could also tell you if the existing crankcase ventilation isn't sufficient. I've seen some honda builds that have high crankcase pressure and had to up their VC vents to -12 from -10an. Sensors are cheap and if you're really pushing the limits just seems like something simple worth monitoring imo, if even just for a little bit. Heck, it probably isn't a terrible idea to see what the crankcase pressure is on some highpower BP builds as well, just something not many people run.

 

80% of the reason I stay here is the depth of knowledge and wide variety of fields. You ain't getting this kind of input from Facebook.
1. Hot gas coming from the catch can. Not oil, but condensation from E85.
2. Hydrolock.


Alrighty everyone, bust out your bingo cards for this round of "what went wrong in Wingman's engine?"
Looks like my leading contenders are "blown headgasket", "lifted head" "cracked cylinder"... and of course, don't forget to mark your free spot of "Just LS swap it already"




Uh so... yeah I found the issue as soon as I pulled the turbo off.




Uh... well, "cracked cylinder head" was NOT on my bingo list! I did several double takes to fully process that, infact, that was NOT carbon build up, but was a full blown, deep crack.

Cleaned up and magnified a little:



"Uhh... Wingman, that sucks and all, but a crack in the exhaust port doesn't explain how water was getting into the cylinders-" HOLD THAT PHONE RIGHT THERE, DO I GOT A DEAL FOR YOU BUDDY!

Yeah... its cracked all the way though into a water jacket. Simply holding the head at an angle and dribbling some water into the water ports is enough to have it dripping leaky faucet style. When hot and with pressure on it... no doubt this becomes a flood.


On the bright side... headgasket looked perfect, zero signs of lifting, blowthough, or stress in the gasket or head. The block and bottom end are MINT (I hit the HG with some brake clean, so ignore the bubbling in the black layer)



I tore the rest of the motor down all the way to bare block and inspected everything. No signs of damage. Crank is smooth, bearings have only idle wear on them, lots of milkshake residue, and rust marks everywhere the rings were sitting, but it's 100% salvageable. Since its torn down anyway, I'll do new bearings and rings and dingleberry hone the walls to get a new surface for the rings, but the bottom end is going back together and back in service.


Thoughts on this failure.
I was told waaaay back when I started this that EGT's would be an issue. I immediately thought of the valves so moved to a full stainless valvetrain to protect against valve melt. I ran this motor on E85 and as rich as I could stomach(11:1 at peak boost) to keep that hot hot engine breath down. But seems I've straight up run into the limit of what this head design will tolerate. With EGT's cresting 1850F during the long pulls of Road Atlanta's back straight, and a turbo pushed past its efficiency zone, I have no doubt that the combination of high EMAP and high sustained EGT's was more then this restrictive single port design could stand up to. Once a crack started, it rapidly would have flashed coolant, dooming that cylinder head. At this point, I do not believe my water plate was the initial failure point, if it even failed at all(examinations of the RTV patterns have led to me revising this initial trackside hypothesis).

I'll soon agian have a fully built, serviceable K24 bottom end. If only there was a way to both reduce EGT's and make the exhaust more free flowing... Like... I dunno... maybe give each cylinder its own individual exhaust port, instead of cramming all four into less square footage then a $500 New York apartment? Hmmm....





Inevitably, someone is going to ask: "Why don't you just turn down the power some, you can't hook up till 4th gear anyway. If you turn it down some it might actually last more then 3 laps and you can get a year of enjoyment out of it"

 

This thread continues to be the most entertaining read and (unexpectedly) house some of the best memes I’ve seen in weeks.

Damn, the rollercoaster continues. Glad to hear the bottom end isn’t completely hosed. Also god damn, way to make quick progress. At the rate you work, I’m sure this thing will be back up and running in no time.

 

Good to hear your block looks alright. That's some great news. I don't know the power differences between the K24Z and A heads but I'd think it's pretty substantial and it allows a proper tubular manifold that flows well vs a log. I'm running the same turbo on my F22 with an equal length tubular manifold, and it makes almost 390whp at ~9psi on pump 93. On e85 and 18+ psi I'd expect to be close to 500whp. You might give up a tiny bit of initial spool up but should make noticeably more powa. At a track like Road Atlanta it should work well.

 

Yeah that's fookin hot. 1850F = 1010C
From the OEM Honduh world we had a steady state limit of 905C and a transient/failsafe condition of 950C for the L4(K20C*) and V6 turbos(J30AC). They glow pretty bright at 950C for the record.

We also did a lot of research on thermocouple size, because the sim guys would always bitch about sensor lag during highly transient test modes (like WOT pulls or engine warmup characteristic). If you're running a big honkin 1/8" EGT K-type thermocouple, then the response time is a few seconds slower than a 1/16" thermocouple. So unless you were near steady state like on the back straight approaching vmax, you can consider the EGT reading to mean "at least 1010C".

It would make sense the crack was in the exhaust manifold if it was exhaust temp related. There is a big water jacket underneath the integrated exhaust mani, so you would have the highest thermal stress there. And looks like it started at one of the port bifurcations, which would be a stress riser.

 

Damn, I guessed a compromised head gasket but didn't think to consider the head being compromised itself, which gives near identical symptoms.. nice diagnosis and thankfully it seems relatively straightforward to rebuild this one.

Doing some googling... it's interesting to see this being a "known" issue. Going to the K-series experts themselves, they note "Areas of the port that are close to water jackets are left alone to prevent cracking with high exhaust temperatures and big boost loads."
https://4pistonracing.com/collection...-cylinder-head

Not to be a bad influence, but their 4-port design is still on sale.... just saying
https://4pistonracing.com/collection...port-race-head

 

Wow, I wasn't aware of that being an issue on the Z heads. The Ford Ecoboost 4 pots have a similar issue, so I'm not sure why I didn't think about it. Their failures are usually a little less dramatic though and just end up with a bunch of steam coming out the exhaust. I'm glad to see the bottom end is in good shape though.

 

I don't think anything was going to hold together reliably at those temps. Sure the integrated exhaust mani is more sensitive, but the exhaust valves and turbine wheel are also well over their temp limits. So you really need sodium filled inconel exhaust valves. And then your limit is going to be turbine wheel, which are only rated to ~950C.