IanM

My 951 suddenly lost power and started to run like cr@p during a track day in Portland back in August. After trailering it home (thanks to Rich S), I did a leakdown test and found 100% leakdown on numbers 1 and 3, and 60% leakdown on number 2.

I finally got around to pulling things apart last weekend. Looks like I've burned 3 exhaust valves! Car had been running great up to then, rock solid 11.8:1 a/f ratios under boost (I'm running a Vitesse MAF / chipboard / piggyback), running 15psi (creeps up to about 16psi at redline - due to a low restriction turbine/exhaust combo - even with my 44mm V-band Tial wg it seems that exhaust would rather keep spinning up the turbo than head out the wg). No obvious signs of detonation. Did not blow the headgasket.

My theory at this point is that it was the result of detonation, likely due to: running 91 octane gas (I run 94 octane at home), and the fact that my car does burn some oil through the rings on the track (about 1/2 quart per 25 min session), which reduces the effective octane. I was running a bottle of NOS octane booster at the time, but apparently that didn't help. I've done lots of track days with this car, but this is my first experience with burned valves. Must have got pretty hot in there!!

Has anyone else ever had this happen? Thoughts?

derekvg

i think it's the oil usage.
oil on your valves makes them alot hotter.

i've already burned multiple exhaust valves due to that.

alxdgr8

Octane booster is only good for like 0.1 points...so you basically had 91.1 in the tank. It's a scam.

eclou

wow that looks like my head did. Only 15# boost with the VR chips and they still burned up? I'm going to start bidding on LS2 pullouts....

IanM

The head had been completely rebuilt in 2007, less than 10,000 miles on it (and about 15 track days).

I agree about the octane booster, that was the first time I'd run it. But I had read a pretty convincing article on it, so figured it was worth a shot. The results claim the NOS Racing Formula raises octane from 91 to 93.7. I don't think so, based on my field results.

http://www.europeancarweb.com/tech/0...ted/index.html

eclou

FWIW mine burnt up even with a mix of 50/50 104/93 octane

IanM

Did you determine that yours were the result of detonation?

951_RS

Hmm, definitely interested in learning the cause of this. I also find it really interesting that the nos octane booster supposedly raises 91 octane gasoline to over 93 octane... I'd always thought it was a very small gain.

Premier Motorsp

The way to burn exhaust valves is to have extra hot exhaust gas. You get extra hot exhaust gas when you have the ignition point too late, which is commonly called retarded timing.

16psi and 91 Octane on a racetrack is going to knock. The knock sensor then does its job and then the DME selects a later ignition point. Another possibility is that chip is mapped to prevent detonation using street fuel and runs 5-10 degrees advance at WOT. With the ignition point that late, the exhaust temp gets crazy, and the valves burn.

With the late ignition, the burning fuel/air charge does not transfer as much energy to the downward moving piston. This 'unused' energy is in the form of heat, and that heat goes out the exhaust valve once it opens.

If you 'fix' this problem by using exhaust valves made of heat resistant metal, the problem will then move to the exhaust valve guides, which will seize to the valve stems.

Chris Cervelli
Spline Technologies

IanM

Thanks Chris. In future I'll bring some cans of higher octane gas with me, and/or will turn the boost down to 12psi or so if I'm stuck on track with 91 octane. I love Portland International Raceway, but you can't buy race fuel there during normal track days.

I would also like to be able to monitor these conditions before they cause damage in the future. Obviously monitoring a/f ratios with my wideband O2 sensor is not enough to prevent it. If you were to recommend a means to help prevent this, would you recommend monitoring for detonation, EGT's, or both?

Does anyone have any experience with the Turbo XS Knocklite?

http://www.turboxs.com/more_info.php?ID=212

I know some of you have been using a knock counter, but I would prefer a bright light and/or a buzzer to warn me of knock when I'm on track. With the Knocklite, would you: (1) wire it into a separate dedicated knock sensor, (2) wire it into the raw signal from the oem knock sensor, or (3) wire it into the knock output from the KLR?

fast951

Having a engine failure is no fun, sorry you have to go through this. It sounds that you have multiple items working against you at the same time, hopefully this will help you determine the causes.

- Low octane level (91oct) to start with. You add oil in the combustion chamber, and now you have even lower octane level. Low octane is not boost friendly, could cause knock. Knock triggers the knock sensor which causes a reduction in ignition. Riding the knock sensor and retarding the timing for a extended period of time will cause the EGTs to go up, which could burn a valve.

- Do you have a chip designed for 91oct or for 93oct? Running a map for 93oct on 91oct is not good, run it on the track like this is asking for trouble. Were you running the pump or race gas map?

- What % Ethanol was there in the blend? Some States have 10% Ethanol in their fuel.

- Have you run on the track with the same setup before? Anything unique for this track day?

- Going over 15psi on pump gas is not recommended even for street use. On the track, it is recommended to add a mix of race fuel even to 93oct fuel.

- You mention the turbo is experiencing boost creep. Have you checked the exhaust back pressure in the X-over pipe. High backpressure causes the exhaust to make it's way to the combustion chamber causing detonation.

- One of the spark plugs has a bit lighter color than the other three. Possibly a injector issue?

- Are you using one heat range colder spark plugs?

- Any sign of detonation on the pistons? (You will see some tiny holes in the surface area.)

- Since you have been using the same software for a few years now without a issue, What changed in the configuration? Increased oil consumption? Anything other than the low 91 octane? Same spark plugs as always?

- Since the head was recently redone (10000 miles). How wide was the valve seat? The wider the area the more contact area there is between the valve and the seat, the more heat transfer takes place.

The above is a list of items I would look into, hopefully you can narrow it down to the actual causes so it does not happen again.

fast951

Hi Chris. Don't you mean retard the timing 5-10 degrees, not advance the timing?

I fully understand what you are saying. Some chip resellers do retard the timing to allow higher boost on pump gas. Many brag about running 20-25psi boost on the track on pump gas.. Some 951 owners not knowing any better jump all over it and think they are on to something.

The software in use here has been tested and proven on the race track in 100s of applications. We don't retard the timing to allow the user to run higher boost on pump gas. We just advise to limit the allowed boost on pump gas.

slownrusty

I can relate to this experience, I was running on the Interstate late at night and dropped the engine down 4th and brought it to red line. My Turbo XS manual boost controller crept to 15-16psi (cold night) and within a few seconds, it sounded like I had rod knock.

I limped the car home and did a compression test at 2am and revealed 0psi on cylinders 3 and 4. I took the head off and I damaged 3 exhaust valves way worse than you did (large chunks missing), but the head gasket was perfect. This was on a 16valve head.

I replaced all my exhaust valves with Inconel valves (which my machinist sources through his connections), $30 each his cost and dialed the boost back a few psi. So far (2 years later) no problems.

Yasin

IanM

Thanks John.

Re: 91 or 93 octane chip, I guess I don't know for sure what you sent me. However, I'm sure when you and I were discussing this we chose the 93 chip, since I run 94 octane fuel in my car 99.9% of the time.

I just called the Union 76 gas station in Portland that I took my car to. According to the guy on the phone, they sell 92 octane - so that must have been what I was using (not 91).

I was running the pump gas map.

No idea on % ethanol.

Yes, I have run with this same setup before, minus the NOS octane booster.

Re: exhaust back-pressure, no I have not measured it. I started to notice a bit of boost creek after going to a big exhaust d/s of the turbo. I swapped out my 35mm Tial for a 44mm unit, and saw an improvement. If I run a cat, the creep disappears - that's why I deduced that the creep is due to the low restriction exhaust. It's easier for the exhaust to go through the turbine than dumping out the wastegate. I have been thinking about dumping from the wg to atmosphere. I'm also running a full Garrett turbo, w/ a Garrett 0.63 a/r T3 turbine hsg. It's physically larger than the #8 KKK turbine hsg, and I assume flows with less backpressure (although I do not have any data on it). I modified the crossover outlet from the KKK round outlet to the Garrett rectangular one.

I had my 52# Bosch injectors were cleaned and tested in 2007, were in excellent shape. Number 1 spark plug has been darker than the others for a couple years, presumably due to oil bypassing the rings on this cylinder. Note that number 1 experienced the most "impressive" valve burn.

Yes, I've been running NGK BPR7ES plugs for years. They are 1 heat range colder than stock.

The pistons look pretty good.

Yes, likely the oil consumption problem has gotten worse over the last couple of years. I've been pushing the old girl pretty hard on the track.

I have no idea about valve seat width. I use a local guru named Lance Hayward for my headwork.

fast951

Since you have been running the same setup for years and pushing it on the track without a problem, it sounds that the oil in the combustion chamber might be the cause. Of course running 92oct did not help, but I doubt it's the main cause.

The piston in the second picture could have a sign of detonation, however it's not clear. The piston will need to be very clean to be sure.

Have Lance examine the head. He's good!