Burned P/C TRACK CAR
 

A fresh 3.0, twin plug, crank fire, 12.0 to 1.0, 110 fuel, MFI engine, my best engine that I had just freshened up. Felt very strong for the first few laps and then began to deterioate. By the time I pulled into the pits she was missfiring. 3 cylinders involved, #s 1,3 and 4. #1 being the one most effected. The injectors all spray well and I had the pump on my test bench right before installing it. My ignition is Electromotive XDi and timed at 22,7 and 1 degree giving me 30 total. Running this system for 4-5 years with no issues. I noticed that the headbolts had loosened a bit, but I have experienced that on other rebuilds right after the initial outing. I have to think something has gone south in the ignition box. Any thoughts. Thanks, http://forums.pelicanparts.com/uploa...1742247117.jpg
http://forums.pelicanparts.com/uploa...1742247137.jpg

Ugh, sorry to see this.

How loose where the cyl head nuts? Just lost a little torque, or spinning loose?

I'm guessing you're looking at detonation.
Detonation causes abnormally high cylinder pressures and heat which would explain loose head studs.
Ignition malfunction is a pretty good guess.
I'm wondering why you're running cylinders with a CE rings instead of flat tops or Ni-resist rings with that much compression?
What cams are you running? Did you calculate dynamic compression?

There is a stand-alone knock sensing system you can use on analog cars. It triggers a flashing light that you can put on your dash. Made by a company called Link. Doesn't pull timing or anything, but it lets you know to back out of it if detonation is happening. I'm putting one on my 911 vintage racer to go with the rebuilt engine.

Thanks for all the response. Head nuts had lost 2-5 foot pounds. Henry, those were the cylinders I had at the time. I have a set of 3.2 KS which I will send out for Nikasiling but have run these cylinders for 4-5 years with no issues. Elgin cam 306/288 100 centerline which is not as hot as the 314/296. Did not calculate dynamic compression. Again, is same motor that I have been running for a while with one exception. Went from 3.2 heads to a set of 78 SC heads reconfigured by Xtreme heads. Bill said an additional 30 HP and I said I don't know if I want an additional 30 as I think in its previous form was probably making 325-330. Maybe the CE cylinders can't handle it. A way to monitor detonation I probably need. I am 80 and ran out of projects 7-8 years ago and turned to this addiction and it has been expanding with much fun and related problems but I knew that would happen. Bob

Don't the pistons and heads get a sandblasted look from detonation? The head looks pretty smooth still, but pic is not so good. How about top of piston's?

I know little to nothing about the science of detonation. Other than you have explosions rather than a controlled burn which can rapidly raise the cylinder head temp and cause things to start coming apart. Running too lean or too advanced in the ignition. The piston and rings don't seem to have been compromised surprisingly and the surface is smooth. Have just machined .010'' from the damaged head and appears another .002 will be required. Will need to add a shim and cut .002'' from the other two heads. Right now I am trying to decide if the CE style cylinders cannot take the stress. I have a set ready to go. Thanks, Bobhttp://forums.pelicanparts.com/uploa...1742324113.jpg
http://forums.pelicanparts.com/uploa...1742324127.jpg
http://forums.pelicanparts.com/uploa...1742324143.jpg

Bummer R
Looks and sounds like it was detonating. You changed the heads but didn't recalculate compression it seems(CC the new heads and piston)?

Do the math , if the head reconditioner took off .010 or .020" , you may have lost several CC's of head volume, that will make quite a difference in compression. Also your math on the ignition timing: 22.7 plus 1 is 30 , can you clarify? As a rule of thumb, when we twin plug an engine like this , we generally need 5-6 degrees less ignition advance, vs a single plug to get to max tqe, pk hp. . So generally speaking, a race engine like that single plug may achieve max tqe and Hp at roughly 30-31 degrees, so 25 degrees of advance twin plug we commonly see. This is where Porsche Motorsport would set the factory RSR 3.8 sprint & original 3.8 Std engines we have seen, run and tuned . Of course there are exceptions and not all engines built the same run identically, but your advance of 30-31 would be way way too much advance.
You can check those damaged, and your other pistons and check the top ring land gap between top ring and the piston land with a feeler gauge and see if all pounded out, .004" is the limit tested with a new ring., but with your existing rings you'll get an idea of condition. New, this should be at or just under .002". You rod bearings also are likely pounded out with the detonation.
You also can send your pump out to whomever you use the service and calibrate it and check each cylinder flow is correct and even. On the last OG 3.0 RSR engine we rebuilt Gus at Pacific gave us our new injectors back(we supplied new streeter injectors) with each cylinder position noted to be more even.
Ideally, you need to be checking lambdas each side on a new engine(we log these ALL the time if EFI) , and use some form of high quality audio knock detection when optimizing your ignition timing at initial set up. I know it isn't always convenient but finding a Dyno and savvy operator with necessary equipment will save a bunch of time and money in the end. If you cannot do this be very conservative and stick within known realms.

Kevin
GAS Motorsport

If all things are equal, the only way to make more horse power by head configuration (IE: porting or combustion camber reduction) is to improve cylinder filling (volumetric efficiency or compression).
Increased volumetric efficiency will increase dynamic compression.

As stated in my original post, the engine is twin plug and 12 to 1 compression. The XDi ignition timing is configured as initial, 3000 and 8000. 22* plus 7* plus 1* equalling 30*. AFR are high 11s-low 12s. This engine has been in action for 3-4 years and on this refresh it got a diff set of heads and pistons to match so the compression stayed the same. On paper, the heads should have added power, but enough to blow up the motor? Don't know and that is my conundrum, what caused it? I will rebuild it, drop the timing to 28*, keep the compression at 12-1, test with a timing light, probably get a knock and head temp sensor. Again thanks for all of the response. Bob

Was the original engine set up on a dyno by chance?

Burned up
 

Bob
You need to actually calculate compression by getting volume numbers for your new piston and heads. We have returned likely 20% of the custom pistons we order because the manufacturer totally missed on some critical detail, either way too much compression, or way to little. Hard to explain why but they mess up too, like all of us.

If head folks report you are going to gain 30 HP, they think they have achieved a ton more air flow(I suspect way to rosey a prediction unless you old heads were absolute junk- totally possible w someone who has no idea what they are doing ported them however) and logically that means in keeping with a 12:1-x A/F ratio, you are going to need a corresponding bunch more fuel in those areas. On a dyno monitoring lambda(A/F) you would have seen this and adjusted accordingly.

Setting an arbitrary number like 28 degrees is a tiny step in the correct direction I believe, but these lumps are expensive, and to not dyno it and set it up properly, tune it for fuel and ignition, is very risky(as demonstrated) . You don't want any more ign advance than what makes peak power and Tqe, and really want to build some safety margin in above that by taking out a wee bit more advance in some areas for when things get really warm etc. We have that ability w Motec (any decent EFI system actually but shameless plug for me being a Motec dealer in North America for 12 yrs or so) and when head temps creep up we pull timing. You don't have that luxury w MFI. If you cannot dyno it at bare minimum get a high quality Lambda meter always displaying in the car (I can link you to one and you can thereby check fueling is in the "ball park" , dual lambda one on each side would be better not sure if avail. in a meter display) , set timing at 25 , and hopefully you get lucky. Realize, we NEVER do this and dyno , and tune every engine we build, customer's and our own, but trying to give you a chance here. These are way more fun to enjoy on track at RPM than work on, so hope I have convinced you .

Last thought, also verify w a timing light (easy to do on a dyno or briefly free revving to say 5500 RPM -oil warmed up already!) that at rpm and load at Peak Tqe , that ign advance is where you believe it is.

beST

Kevin
GAS Motorsport

MFI? Do as Kevin says and get a standalone Lambda sensor (or better yet, 2) and a knock retard system. Then religiously observe the lambda and "adjust" the MFI as necessary. 25 degrees is a good place to start, 23 might be better. Leaving a tiny bit of hp on the table is better than torching a cylinder. And if you're too "busy" when driving, set up a camera to watch the lambdas under load. Too lean and too much advance will quickly destroy any engine. Oh, and what kind of fuel did you have in it? 12:1 is race gas only territory.

Innovate sells a 2-channel AFR meter. Check with the usual suppliers.

I am running an AFR meter and see high 11s to low 12s. 110 race fuel. The piston build sheet says the dome is 47.5 cc which should give me a little over 12 to 1. They I think were basing that on a 87.6 cc head. My heads are 86.6, I measured them. I calculated the dome as 48.9. By the time I removed material from the valve pockets and small relief for the plug tips, my dome was down to 47.34cc. Approx 12.09 to 1.0. So far as depending on the speculative additional HP numbers I wasn't. Not that it wasn't a potential improvement over what I had, but what I had was an unknown in terms of its flow numbers. Bob

How old is the AFR meter sensor? The lead in race gas damages the sensor and it loses accuracy.

Burned p/c
 

I think you can test the sensor with a volt meter. Would want to do it under load or install another one and compare. As long as they are cycling in a normal manner, I want to trust them. Have dealt with numerous failed ones that began to act erratically.
Cycling in a normal manner may not necessarily mean they are accurate though. MFI is not known for having lean issues unless you have a leaky check valve. Doubt if all would suddenly fail together. Fuel pump, filter and fuel console check valve have to be checked. The problem effected the whole motor, some cylinders more than another. Will put everything back together, confirm timing with a timing light. If that checks OK will still install my backup XDi unit. Of all the problems I thought I might have to deal with, this isn't one of them. Plan to have head temp and knock sensor installed by launch.

With leaded fuel, the failure mode might be different than what you would see on a car running unleaded. Not sure the electrical signal test is valid. With gradual lead buildup, the sensor slowly becomes less sensitive to inputs and may give false readings. I think Bosch says something like 50-100 hrs of use with leaded, assuming a properly tuned engine. Could be faster if running rich or lean, etc.

What Bosch says about leaded gas and 02 sensors

More about sensor degradation and failure

It sounds like you've got a solid setup, but the issue with misfiring on cylinders 1, 3, and 4, especially with #1 being the most affected, could indeed point to an ignition issue. Given that the Electromotive XDi has been reliable for years, it's possible that the box may have developed a fault, especially after some heat cycles.

Sorry to hear about your troubles, there's a lot of good advice above. If it's a true 12:1, that's definitely a lot of compression even running 110. Your dynamic will be slightly higher with the milder cam too. I second the mentions that 30 degrees is a LOT of timing for twin-plug. I run 24 in my 11.5:1 3.0 twin plug race engine (built to HSR rules). It did make a little more power on the chassis dyno with 26-27, but I wanted to be conservative especially since I'm running carbs. The pics are of the plugs in an engine that ran lean, didn't even run it a full lap. After that I installed an AEM afr meter. With sunoco purple (110), the O2 sensor only lasts ~5 race weekends.
http://forums.pelicanparts.com/uploa...1743212901.jpg
http://forums.pelicanparts.com/uploa...1743212901.jpg

Ugly.

Yep that motor was really unhappy

Damn SV - Flame thrower. How long did that run and it wasn't misfiring like crazy? I have never seen plugs that bad even on Turbos that went kapputt.

R lane You didn't share which A/F meter you were using? Also, given what you have seen a new sensor is on order no matter what at this time even if a high quality device is on hand. In our experience the Bosch Lambda 4.9s last at least 50 hrs, often 80-100 w leaded fuel. Never buy from a cheapie internet source(goes for everything on these engines) , loads of Chinese knock offs out there.

Also check your rod bearings are okay, or you will be right back here with even greater damage...and how did ring land gaps on top ring measure?

Lastly, trusting your piston sheet is a mistake, actually CC your dome displacement volume, sometimes they are quite off from intended design as I shared.

Good luck

Kevin
GAS Motorsport

Hey Kevin, yeah it was a bad day. Rolling race start and down the front straight at summit point. It was down power, so after turn 1 i limped it back to the pits but the damage had been done. Amazingly those plugs were still firing. The HVC coil is a heck of a coil.

Burned p/c
 

svcetiquette, my condolences. You are running a 6 plug which I think might be a little warm. The motor is back together and will install today I plan. I cut the heads .015 which dropped the chamber to 83.8-84.2. Cut the top of the pistons to keep the compression at 12. I find measuring the pistons very tedias as the dome height measurement accuracy is very critical to getting the correct dome volumme. Replaced the rod bearings on #1 but the old ones looked just like the new ones. #1 piston and its ring gaps looked and measured fine. Bosche LSU 4.9 sensor. I now have a head temp sensor. Plan to get it to op temp, back timing down and check with a light and run it up and down the ave a few laps, let it cool and check the head bolt torque. If all appears well, will go up to RA for a PCA race in two weeks. will keep posted. Thanks for all the share. Bobhttp://forums.pelicanparts.com/uploa...1743426919.jpg

burned p/c
 

What did you find to be the culprit. Timing? You have backed the timing down and the problem appears solved? That is what is keeping me awake at night as there are a few things I can change, but is that my problem. I had an oiling issue with a motor 2-3 years ago that didn't manifest until I increased the power. So I wonder if this is what I am dealing with. An existing problem that didn't bring the previous motor down, but assuming that I have increased the power with this motor, is now a problem. Bob

31 degrees ignition advance at full throttle for a twin plug motor is way too advanced. Doesn't matter if you're running 110 octane, or your dynamic or static compression is as low as 10.0:1. You're causing the peak pressure of combustion to hit your pistons as it's still rising up in the cylinder, hammering on the pistons, cylinders, crank, bearings, etc. Imagine when you press down on a bicycle pedal for max torque. You want to press down in at 1:00 position of rotation, not at 11:30 - you'll hammer your knees. If you use a dyno, you can advance your timing until power levels and back down from there for margin, assuming fuel octane is not your limitation. When you advance timing on the dyno, when it levels, and goes down after that, you went too far. The suggested timing numbers given by others here are a good place to start.

svcetiquette, in the picture of your head it looks like there is a crack from the intake seat over to the spark plug hole.

An article on pre ignition vs detonation: http://forums.pelicanparts.com/porsche-911-technical-forum/250384-detonation-knocking-etc-3.html#post2232959

My last race engine (120HP per liter at the crank) had 12.5:1 compression using 110 and 112 leaded race gas. I ran no where near advance 30 degrees of timing...

Oh, NTK oxygen sensors handled leaded fuel better than Bosch oxygen sensors...

Installed a knock sensor kit but the controller appears defective. Fellow sent another by USPS and they apparently have lost it. Have a dyno appointment on Tues morning, so will make some runs without the benefit of knock control or being able to see when PI or det is occuring. I plan to set my timing at 24* and be armed with a stethoscope. Not sure for what I will be listening but will continue to raise the timing a bit at a time until HP has reached its terminous. I am thinking for my setup, 26* might be my limit. Prescribed procedures On how to know where to set ignition to avoid melt down here would be greatly appreciated. Thanks, Bob

Start at 20 deg, then keep advancing until the power drops off, then retard 2 degrees from there.

Installed a J&S knock sensor system and took it to the dyno yesterday. At 20* we had 267HP, added 2 degrees and had 287HP. 2 more degrees and 291HP. At the top of that run the low end of the LED bar flickered briefly. The LED bar has low, medium and high light intensity indicating the severity of the knock and that timing has been pulled back. Off to the track to see if we are safe. Bob

Nice result. So you ended up at 22 degrees?

My last run was at 24*. I think I am ok with that, maybe replace think with hope, but will I think bring it back to 22* as the extra 2* netted me but 4 additional HP. An interesting aside to backing off timing is that as you reduce that number, the piston finds itself further up the barrell when ignition happens. Meaning that it has less distance to travel when pushing against the force that ignition has just initiated. Bob

That’s a LOT different than the previous 30 degrees !!

Tested at a track night, 3 sessions. Engine felt good and no issues. 4 days at Summit Point and no engine issues. Have just pulled the valve covers to check valve lash and head torque, all good. Am at 22* timing and engine has good power. Don;t know if it is an improvement over the previous engine with ''my'' heads at 28* timing, but it felt good and it stayed together. Thanks again for all the counsel. Bob