Don't take what I said the wrong way. Changing gears at 3000 RPM won't cause the fracture. The reason it snaps when changing gears is because the fracture is already there and large enough for the change in acceleration of the rotating assembly to "break the camel's back"

oh boy, after reading a couple of pages in this thread...then I have to skip to the last page because you guys start to turn a good thread into a shit....Ladouche needs to stop posting up csrt4/bmw dynos and graphs and shit...can't compare this 2, there aren't many caliburs out there anyway.

back to the thread, I just got an AP from corbs, install the new denso plugs, load up stage 1 93 map....then I got kr like crazy around 3k rpm, a lot of hesitation/like boost lose..occur when kr...
my car before all this have 56k miles runs like a charm, last dyno have huge torgue with only 268whp....and about 320 tq, 56k with original oem FOMOCO spark plugs..lol
until now, after installing the AP...run like shit...literally like shit, will give a try longer with diff maps and see.
been monitoring everyday and it seems Lex is on to something, I do get kr around 3rpm no matter what gear, lately more so on 5 n 6.
I have cobb sri/inlet/ap/ and ets tmic for power mod..

This is what i was referring to when I mentioned manufacturing defects NOT design problems.

"what's wrong with a high compression ratio? "

This is a great question.
The issue is what compromise do we want to make? There is a curve of compression ration vs thermal efficiency (which means both power and economy at once) which is roughly the same for all passenger car engines. IIRC, in the range from 8-10:1 you gain just over 3% per full point of compression. As you raise compression the EGT goes down a bit, because more energy was used to push the piston rather than being wasted as heat.

For a car mfr faced with CAFE and customer demands, 3% from ONE design change is a big deal. They can even shrink the engine 3% for the same performance and get a another fraction of an mpg (too small to notice for us, important for them) to factor into their CAFE. They want to run the engine up to the edge of detonation, and use KR to (usually) keep in pulled back from that edge. If it turns out that too many engines come back for warranty, they just change the tune and reflash everyone. Speeds seem to be sort of beta versions for stuff that will later go into popular models. They want to find problems and get them fixed before the issue migrates to their volume models, where a reputation for blown engines or drivelines would be fatal.

But for us one less point of compression means significantly more boost before detonation, because high CR and boost affect cylinder pressure in different ways, and it turns out that you can get much more power, though with less efficiency, with lower CR and more boost. We want to stay out of high load detonation 100%: it's highly destructive and will kill a gas engine, it's just a question of when. There's a big difference between the factory writing off 1% of a short run of engines in exchange for an extended test and an owner losing their engine and paying 100%, since modded engines generally get no warranty from Mazda.

In short: Detonation Bad.

High CR vs More Boost
Let me use an extreme, not realistic example to show the boost/CR tradeoff.
Let's say we have a small 4 cyl engine with DI (such as ours). Compare high and low CR.

Baseline: Our engine, stock except CR lowered to 8:1 with dished pistons designed for DI.
i) 8:1 CR, 93 octane, timing adjusted within 4 degrees from maximum hp. Add boost until detonation begins, then back off the boost.
ii) 13:1 CR 93 octane, same deal.

Approximate results. These numbers are to show the forest, not the height of each tree.
i) Before boost, it has lost roughly 4-5% in power and efficiency compared to stock. This engine can handle more boost than stock before detonating, very roughly 20% more. That's at least 20# of intercooled boost and probably well over that. At 25# it should make 250% of it's unboosted output, or 235% of it's NA power at stock compression.

ii) 13:1. This engine produces roughly 13% (not the exact figure) more of both power and economy compared to our 8:1 example. At low rpm there may also be an extra gain from forcing exhaust out of the cylinder. It also sends less heat out the exhaust.
Because of diminishing returns from high compression, we get less than 8% gain compared to the stock, 9.5:1 engine.
Now we add boost. At 13:1 we're way past pump 93 on carbed or port injected engines, but with DI and small cylinder we get to push it a bit. Most engines need race gas at this level. Boost? I'd like to know how much boost at 13:1 CR produces cyl pressure equal to our stock engines at stock psi again I don't have the *exact* numbers. I'm going to say not much, because we're right near detonation in NA now that we have 13:1. Let's be generous and boost it at five psi, gaining 30% over NA. Personally I think we'd gain a vented block at just 5psi here, but let's put the 13:1 mill in the best light.

Compared to 8:1 NA we have
13% gain NA or 1.13, and then 30% more w boost, giving 147% of the base motor (or 47% more than 8:1 NA). Crediting the base motor with 140hp, that would be
205 hp at 13:1 and 5psi compared to
330 hp at 8:1 and 25 psi

Held to the same octane, the low CR more boost combo makes more power, much more. Changing CR by one point rather than five has a far smaller effect, but you can see which direction we're going.

The 13:1 engine gets better mpg and has lower EGTs. It's got no margin of safety at 5psi, and I suspect its already detonating towards the scrap pile. It will go boom on pump 93 if you put 15psi through it, fully forged or not.
The 8:1 engine can handle >25psi with a smart tune and enough fuel.

These numbers are to illustrate the tradeoff of higher compression vs more boost. They are NOT intended to be accurate, just ballpark. Perhaps someone with the formulas and firsthand knowledge can fill these in, but I can show the basic point.

Interesting that your car did not like the AP. My AP is on the way and I have essentially the same mods as you + race pipe and CDFP. On the stock tune I never get part throttle KR on my 08.5. I have predictable WOT KR with temps around 100 degrees here. Shitty Canadian fuel has most of the blame.

I would try the 91 maps first as the timing curve is more tame.

well, octane here is 93, gona keep trying diff 93 maps first, all esle fails then I'm gonna get rid of the AP, lol ..my gut tells me my engine doesn't like it...its running fine without it..lol

We have 94 octane here in Canada - but it knocks more than with 91 octane just over the border. I won't even get into that shit.

Yes, CR has diminishing returns in forced induction motors. Lower compression motors end up making more power through boost, so you are correct here.

its nice up there nowadays, good cool weather, here in FL its hot and humid,

okay, I will do as you say and try 91 map and see...

Well if your numbers are correct, even ballpark, than definitely 8:1 is the way to go instead of 9.5 that we have stock.

Of course, 13 is exaggerated, I am not sure if there are even NA petrol engines with 13:1.

Theory is completely sound, unfortunately the real world works differently. Imagine the car magasine bitching out Mazda would get over "turbo lag" and "no torque" if they gave us an 8:1 compression shortblock. Even DI can't help you there. It's just a little engine and no matter how small you make the turbo or how high you stack the static CR, it's not gonna torque off the line like a V8, which everybody wants...with 30 mpg too, of course.

It's a good balance stock or close to it but, you're right, you're in a small box when you try to stretch its output, without major changes.

Did this theory address why no motors have blown at WOT (save for the few that had WG lines pop off and boosted to >30psi)? Is it cause the rpms are higher (which is shorter stress periods in cylinders) and more fuel to cool things off?

Engines can blow off WOT for any # of reasons, here's some:

Too lean part throttle fuel mapping

Previous damage hanging by a thread

...and my favourite, overloading the motor with high boost in high gears at too low an rpm.

Then why don't ours ever blow that way, minus the WOT at low rpm. How come we don't see failures WOT at high rpm?

I was thinking about this as well. I think it would go a long way to remove the BS AND put a light weight flywheel in. That would equate to a lot less stress on the engine. aprox. 50lbs less sounds good to me.

We have seen some bent rods and some blown engines at wot or on the way to WOT that were outright compressive failures. The rod likely had no damage prior and simply was overwhelmed by compression forces. These are the "my line popped off and I made 30psi failures"

A rod that lets go during accel/decel lets go at that time because they are points where the rotating assembly is accelerated. F = MA remember. However, the rotating assembly goes through these transition points all the time on a normal motor without failure. So what can you conclude?

1. The rotating assembly has locked at that point due to hydrolock or mechanical lock.

2. The rod developed a gradual crack. Due to imperfections in manufacturing and high torque output, a crack is formed in the rod and grows eventually letting go during these transition points.

I find #1 less likely than #2, but when the last guy that blew his engine tells me the hot side of his intercooler was filled with oil I do wonder. Remember that people blow doing stupid shit they don't mention as well so take it with a grain of salt.

So what have we learned here?

The pure MECHANICAL design of the rods and motor is fine for the power levels we are making on these engines given the rods have no flaws and nothing strange locks up. This is what DCR has been saying as well. If the actual mechanical limit was exceeded we would not have the whoosh, laloosh, darksun, and many others running in the 300-350whp/350 ft lbs of torque range for a long time. Pretty much everyone would blow and they would blow at WOT.

HOWEVER, something locking the assembly OR the more likely manufacturing flaws of the rod COMBINED with high cylinder pressures makes some of the "weaker" motors let go. This is why the failures happen at random times. You may be lucky or you may not be. High torque, especially high torque at low RPM will push you over the edge if you have a weak motor. So take your chances - or better yet tune and driven sensibly.

The SRT4s, DSMs all had similar issues, but after the car being around for 5+ years the weak ones were weeded out and the people that lasted were the ones that treated the car with respect and had a proper calibration.

STIs blow ringlands at stock power yet some make 450whp safely.

Don't be surprised by this. Parts have a tolerance. Let's say that at stock power 1% of cars blow. At 50% over stock power 4% of cars blow. At 100% over stock power 40% of cars blow - etc.

Big thanks, and i couldn't agree more.

^ +1. And I hate people who do "^" and "+1" but it seemed really appropriate here.

I think the high load at relatively low RPMs is part of the FUBAR'd engine equation.

(I do wonder if some of the compressive failures (bent rods) have been a result of pre-ignition though (due to carbon buildup breaking off in the IM or oil build up or whatever))...

This is why I am limiting load in the 2500-3000 RPM range to approximately the stock TRL levels.

It is also why I am injecting copious amounts of 119 octane meth starting at 2.8v on my MAF (0-1 PSI in most gears).

Here is my latest TRL (torque requested load config):

http://i77.photobucket.com/albums/j6...2SF93v104d.jpg

I do plan on tweaking my wastegate % to try to eliminate boost spikes (I occasionally get spikes to ~20PSI with the standard Cobb settings).

Hopefully, these are sufficient to keep my engine together for a long time...

I agree that keeping the IM clean is a good idea. I wonder if putting some sort of engine safe carbon disolving additive into the Water/meth would help?

I am mostly adding ~100% meth these days.

Methanol is a pretty damn good solvent, but it is fairly polar, which might limit its ability to dissolve non-polar oils. I suppose if we added some xylene or something like that it would be a bit more likely to dissolve the oily stuff better.

While I cannot draw any conclusions without repeated removal of my IM, I would imagine that significant buildup would be difficult with healthy and frequent meth injection.

One company here in NC (rotorsports racing) coats the inside of the RX7 IMs with a teflon coating... they claim it prevents garbage from sticking to the walls... I have no idea how well that works.

http://www.rotorsportsracing.com/per...intakecoatings

i didnt wanna read 20 pages, if i go to atr am i simply just moving all my load tables over one or 2 spots to the right to start my load 1000 rpm later????

Basically, don't have silly load targets at 2500 and 3000 RPMs on your TRL tables. The same goes for the boost targets.

thanks for this post .. this info is very iimportnat to all our feellow speed owners

i just made all my load tables very linear down low.
stage2 93 fmic map was this before i touched it. you can see the load hase a huge jump down low. this is an around about from what i remebver it was.
2000 2500 3000 3500 4000 4500
1.15 1.46 1.80 1.83 1.83 1.83

now this is how my loads start on all my gears load tables. oviosuly loads after that are diff per gear but they all start like this now since it is noticable less in each gear... hope this helps save my rods
2000 2500 3000 3500 4000 4500
1.15 1.30 1.45 1.60 1.83 1.83


For lex and other knowledgable guys.. does that seem like a good start or do u think i can add more boost for the 3500 kinda to have the map look like this....
2000 2500 3000 3500 4000 4500
1.15 1.35 1.55 1.83 1.83 1.83

^ How does boost ramp up now? Do you get any spikes above 20psi below 3000 RPM?

I would say that you are probably fine in the last table. However, you will probably quickly notice that you will be way over the aims in the 2000-3000 RPM ranges on datalogs (like 1.6x at 2500 vs the 1.35 aim). I think this is partially due to the WG spring being ~10 PSI and we can get higher loads at low RPMs than you have in your last table even with ~10 PSI.

My guess is that even at 0%wgdc you still have the 10(I think its 10, maybe its 8?)psi spring for the wastegate. So you are asking for loads that you would see below the cracking pressure of the wastegate spring.

When I was in limp mode it ran about 14.x psi - others have said the same. Can't swear it was "0" WGDC, but one would think.

its an 8 psi spring but you will hit above that depending on modifiications to the intake/exhaust tract.

Here is an example of what I am talking about:

http://i77.photobucket.com/albums/j6...R/082409-4.jpg

This was with my previous tune version that had dampened load/boost dynamic tables. I now have more aggressive ones.

Above target load and WGDC low to zero. I expect to see much more of this as the weather gets cooler. This doesn't happen when it is 80°F out.

On a side note though, I had been running 15-17 PSI for two winters prior to tuning in the 2250-3000 RPM range without issues, so I doubt that ~11 PSI is a big deal at these RPMs even though the calculated load is 1.8-1.9. Just my gut instinct though.

That log is pretty scary man, 0%wgdc & 1.9 calc load @ <2500 rpm = alot of stress on your motor IMO. This is exactly why some will claim BT are safer for these motors. This and also the cooler more efficient charged air near redline and reduced back pressure.

I tend to believe (right or wrong) that this motor should have no problems with ~11 PSI at low RPMs. Spiking to ~20 PSI would be likely very different story at the same RPM range.

I wish we knew exactly how the ECU was determining load. I ASSuME it is equating load with torque with some formula related to g/sec MAF flow, but beyond that it is a bit of an enigma on the formula nuances...

Why the fuck aint this a sticky ?

GOOD WORK MAN!

what is the real reason people are breaking rods??

and what is the best way to solve this?? i mean forged pistons and rods i guess makes the most sense. right now i have intake, inlet, full tbe, fmic, and v105 map. what would be my best bet in order to be safe? I plan on running meth soon, but i mean should i really invest in pistons and rods to have my engine last longer?

just my .02 on a subject i'm not too knowledgable about, but I think if your gonna be making close to or above 300 whp on this car, buy some forged rods, buy some pistons..... use the money you've been saving for those new rims and save your engine.... if you're worried about it; better than spending 4g's on a new block

i think the when people start having their intake mani ported/polished (possibly head) so it flows evenly before getting heavy into the mods we will see a whole new number of hp that can be made on a stock bottom end before failure

If this were the case, wouldn't you expect to see the blow trend favor the more highly modified cars, rather than randomly thrown rods across the whole platform?

I'm to the point where i just don't want to replace a blown engine. It's become too real of a possibility lately.

So fuck it, if the rods are what keeps breaking, pauter here i come. I still believe there is an unusual reason for our misery, but i don't like the idea of just sitting here while we all come up with one wild idea after another. I'm sure we'll eventually figure out what the problem is, but i'm just not down to be one of the test samples.

Lex, if you want me to send you 4 good stock rods in the coming months for analysis, i'll def do that, but don't be offended when i say i don't want to send you a broken one.