Can your pistons do this!

[JR @ CFS]

Hope you enjoy this!

http://web.student.chalmers.se/groups/formula/2007/FSAE/K640_IMG_0232.JPG

http://web.student.chalmers.se/groups/formula/2007/FSAE/K640_IMG_0245.JPG


Here are some pictures of our broken engine. The engine is a 2004 Fazer with a compression ratio of 12.2 and 600cc. We are using an Autorotor supercharger and the highest boost we were using was 0.45 bar. The fuel we were using was E85 but during winter there is only around 70% Ethanol in it. So the knock resistance is lower than the orignial one. We couldn't notice knock tendences at all.

We saw that something was wrong when the engine started to blow white smoke out of the crank case ventilation. The engine could still run so we started the engine several times but as soon as we put load on it it started to blow smoke. After removing the sparkplugs we saw that one was melted and one cylinder was wet. We looked into the cylinder and then we could see that the cylinderwall was damaged. Then we removed the cylinderhead and you can see the result on the pictures.

So, anyone out there in FSAE land have any ideas on the issue? What we found strange was it only happened to one cylinder! Why does it happen only to one cylinder and all the other cylinders were still good? Maybe the compression ratio is too high and it looks like the temperatures are very high in the combustion chamber.

We have received a few opinions from different people and are wondering if anyone else has had the same problems in testing?

[JR @ CFS]

Hope you enjoy this!

http://web.student.chalmers.se/groups/formula/2007/FSAE/K640_IMG_0232.JPG

http://web.student.chalmers.se/groups/formula/2007/FSAE/K640_IMG_0245.JPG


Here are some pictures of our broken engine. The engine is a 2004 Fazer with a compression ratio of 12.2 and 600cc. We are using an Autorotor supercharger and the highest boost we were using was 0.45 bar. The fuel we were using was E85 but during winter there is only around 70% Ethanol in it. So the knock resistance is lower than the orignial one. We couldn't notice knock tendences at all.

We saw that something was wrong when the engine started to blow white smoke out of the crank case ventilation. The engine could still run so we started the engine several times but as soon as we put load on it it started to blow smoke. After removing the sparkplugs we saw that one was melted and one cylinder was wet. We looked into the cylinder and then we could see that the cylinderwall was damaged. Then we removed the cylinderhead and you can see the result on the pictures.

So, anyone out there in FSAE land have any ideas on the issue? What we found strange was it only happened to one cylinder! Why does it happen only to one cylinder and all the other cylinders were still good? Maybe the compression ratio is too high and it looks like the temperatures are very high in the combustion chamber.

We have received a few opinions from different people and are wondering if anyone else has had the same problems in testing?

[Dr Claw]

are the red X's a metaphore for the current state of your engine/pistons? In that case, im sure our pistons do that at some point durring the year http://fsae.com/groupee_common/emoticons/icon_smile.gif

[Jonathan S]

Melted spark plug = pre-ignition from sustained REEEAALY high temps, detonation. Bad stuff...

Maybe check your fuel injectors to make sure they are all flowing evenly. Are you logging O2? Were you logging water temp - was your engine running hotter than normal?

Sounds to me like you may want to spend a little time on a dyno.

[JR @ CFS]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">

Sounds to me like you may want to spend a little time on a dyno. </div></BLOCKQUOTE>

This did happen on the dyno :S

[Wesley]

Probably detonation - but the way that broke makes me think the bore was a bit on the tight side and the ring ends butted from thermal expansion, breaking off the ring land.

The detonation didn't do the destroying, just allowed the piston to get hot enough to get rid of your ring gap and they butted, which caused the galling of the cylinder and the missing ring land. Then you had a piston bit flapping around that messed up your spark plug.

That's my theory.

(And excellent title, BTW http://fsae.com/groupee_common/emoticons/icon_biggrin.gif )

[Marshall Grice]

yeah we've done that on our gixxer motors on the dyno before. We were running it hot (water temp, not on purpose) and hit a lean spot on the map. that spot on our motors seemed to just be the hottest spot and was the first to melt. Melted piston scores the cyl. wall.

[Nima]

See this:
Queens FSAE: Everything Breaks (http://engsoc.queensu.ca/formulacar/enginepages/october192006.htm)
It may help you.

[Dallas Blake]

I dont think what happened here is the same thing that happened to me, mine was simply a broken plug that was ejected from the cylinder, no damage resulted and I actually put another 50 hours on that engine and it still runs strong.

This situation looks like a lean burn with detonation. As for why it only happened to one it could be any number of reasons, different charge distribution, different combustion temperature caused by variations in the exhaust from cylinder to cylinder, seized thermostat, air pocket in the cooling system with the engine continued to run for an extended period of time.

Were you logging AFR or EGT at the time? were either of them abnormally high? What about water temp or intake air temp?

[Grant Mahler]

Do you have an inline fuel filter?

Might I suggest you see what it has in it, as well as flowbenching your injectors to determine if one is clogged...It would be unfortunate if you blamed this on the fuel maps and then blew up another motor due to a clogged injector.

[Dallas Blake]

Out of curiosity what does your plug look like for the cylinder? is it white? chalky? are there signs of pitting on the electrode? also how far up the electrode has the heat passed?

[JR @ CFS]

Sorry lads, in the midst of exams, will reply as soon as we can. Cheers for the comments. We really appreciate the help and will keep you updated on the progress. The engine is back on the dyno today and will run tomorrow, so need to check a few of the above before we start again.

[MC Harry]

Thanks for your help!

We were monitoring EGT and water temp. In our dyno we have a close water curcuit. During the first days we had several times problems with air in the system.

We are using a VEMS and the problem is if you try to modify the VE map or the spark map you need to open another window. In this window there is no water temperature and the main window is not updating.

When the first smoke came from the crankventilation the water startet to boil. Some seconds later we stopped the engine. The water temperature was over 100?C but I don't know if this was a result of the melted piston or if the engine was already hot because of air in the system. Data logging was unfortunately off.

The damaged cylinder is close to the water outlet of the engine. This might also be the reason why it broke there.

The ignition at this time was very advanced and that was maybe another reason for this.

The sparkplug from this cylinder was melted. electrode almost gone... :-)

Tomorrow we will start the new engine. Now we got 4 sensors for EGT. Hopefully we can compare the cylinders and the AFRs are equal.

[Mechanicaldan]

what is the design of your intake? Can you post a picture? There can be a significant difference in the amount of air getting to the different cylinders, depending on the intake design.

It was probably due to too much ignition advance, and being a little lean in that one cylinder. It's really hard to hear detonation on these little engines when they are spinning at high RPMs.

Are you running a wide band O2 sensor? It's pretty much a necessity. The individual temperature sensors for the exhaust will help.

Tuning is an art form, and not easily mastered. I highly recommend downloading the AEM Power EFI Basics manual here: http://forum.aempower.com/forum/index.php?action=dlatta...opic=14527.0;id=2888 (http://forum.aempower.com/forum/index.php?action=dlattach;topic=14527.0;id=2888) You'll have to register on the forum to download it, but it's well worth it. There are a lot of tuning basics that apply to ANY ECM, not just the AEM. There are ignition tables that are a good starting point for pent roof engines, along with spark plug information, Air/Fuel ratio charts, and LOTS more VERY usefull information. Good stuff. Here's the link to the post if the above link doesn't work. http://forum.aempower.com/forum/index.php/topic,14527.0.html Download the "AEM_EFI_Basics_V1.3.pdf" file

[Pete M]

OK, a few things. You're running 0.45 bar boost on 12.2:1 compression. Not in itself insane, but what were your intake temps like? Do you do any form of intercooling, or are you relying on the E85 (E70...).

Second, as others are saying, it's likely that the cylinder that failed was running leaner than the others. Two easy causes of this are poor distribution of air into the runners due to intake geometry and clogged injectors. If your ECU can do any form of injector test, do it and make sure all the spray patterns look right. If not, take them to a place to get them flow benched and make sure they're all flowing similar amounts.

Another thing that can cause it is a boost leak. If you have a leak in one of the runners, that cylinder will get less air and will skew your AFR readings rich. You'll then trim fuel out and may damage the other 3 cylinders. This is particularly important to be careful of when you're running dodgy test intakes.

Also, what do you mean by "very advanced" ignition? You have to be extremely careful with ignition timing on boosted engines, especially when you are running such a high compression ratio. Do all your initial testing on retarded ignition, we usually run at 5-10 degrees off final when we're not absolutely sure, such as with a new engine or a new intake, etc. We then very carefully put in more timing, one degree at a time, doing loaded ramp runs. When you see an area plateau, stop advancing that area. Make sure you are making good power gains when you increase advance. You need to weight up whether an extra 0.25 kW is worth pushing the engine just that little bit closer to the knock threshold. I've heard some methods such as putting in timing until power starts to fall, then backing off a bit, this is pointlessly dangerous. Just stop when your gains are minimal. And listen for detonation, but keep in mind that it doesn't have to be audible to be destructive.

We comfortably run 19 psi of boost (230 kPa absolute) with timing pushed all the way to the plateau, and can hold that boost until the inadequate cooling system can't handle the load. This is on 98 octane unleaded. It is doable, you just need to be extremely careful. I'd probably have another think about your compression ratio too though if i were you. All our initial turbo testing was done with a much lower compression ratio than what we ended up on.

Pete

[Grant Mahler]

Pete M

Sorry for OT but 19 psi on what motor/compression? I presume thats a turbo so what does it taper to at redline? I hadn't heard of an FSAE car boosting that high...

I actually disagree with your suggestion to reduce compression ratio. Lowering the compression allows higher boost levels, while hurting off-boost performance. Given the weight, gearing, and power output of these cars, (as a driver) I would prefer a steady torque output from 8-13k over something much peakier (as would occur by dropping from 12.2:1) to say 10:1). Also, dropping compression will hurt you on economy. Power can be made with less boost (less heat, smaller intercooler etc) by using higher compression, if you can tune it correctly.

Just my $0.02. Correct me if I am wrong.

[Pete M]

The motor is a Honda CBR600 F4, the compression is something i'd rather not comment on. I'm happy to talk about trends, etc but i hope you'll understand if i'm not willing to divulge specific details. Suffice it to say it's high, higher than most rules of thumb would suggest is safe for the boost we run. These engines seem to be able to take a lot more punishment than most give them credit for, so long as you don't do anything stupid.

We've never blown up an engine due to a legitimate pushing of the limit. We've lost a couple over the years to mistakes (3, i think). These were mainly lubrication related, but we did have a spectacular failure involving melted pistons last year (so i sympathise with the thread originator). It was a fuckup though, nothing more, and we haven't had any problems since.

We hit the restrictor below 7,000 rpm, and from then on, manifold pressure declines. By redline its virtually ambient. It gives the engine a very peaky torque curve (96 N.m peak, or 71 ft lbs for you yanks) but the drivers don't seem to mind the drivability. The slight turbo lag acts as a smoothing effect and the flat power makes gear shifts much less crucial.

Compression ratio is a tricky subject. It relates more to efficiency than to power, but of course increased efficiency increases power. Adding more air and decreasing the efficiency slightly can still yield a net power increase though, often substantial. Go too far, however, and your fuel economy and off-boost performance both go to shit, as you mentioned. It's a balance.

[MC Harry]

Thanks Mechanicaldan for the link.

Our intercooler consits of 2 laminova intercooler cores. With the intercooler the manifold temperature is around 50?.

Very advanced ignition was more than 50? between 8 and 9k rpm. Now we will be careful with the advanced timing. We focused too much on the EGT without checking the consequences for the engine.

[Pete M]

50?C is ok, 50?F is awesome. http://fsae.com/groupee_common/emoticons/icon_smile.gif

50? of timing, however, is fairly insane under boost. Assuming you had your timing reference set up properly, so that number is accurate, you were likely well above MBT. You'll find that the plateau likely occurs about 10 degrees below what you had it set to (depends on engine of course, some are more advance hungry). Remember, increased timing may decrease EGTs but it also makes conditions in the cylinder much more extreme.

It's been our experience (with the F4) that the exhaust valves in these motors are fairly resilient little buggers. We've never damaged them. Be more worried about your pistons, you'll fail them first. Don't go boosting up on 10 degrees of timing though, or you'll likely fry something exhaust related. Use a safe number like 25-30 degrees or so, common sense. Avoid detonation at all costs. Boosted engines are very unforgiving and you have to be careful.

The good news is it sounds like a tuning fuckup rather than something wrong with the way you were doing your hardware. Use a more conservative method when tuning ignition next time and you should be fine. I'd still check all the stuff that people have mentioned though. Do the other 3 pistons look brand new or is there some damage done to them as well? 1 piston failing before the others isn't uncommon, but if only one piston is damaged, it might point to some kind of imbalance (air, fuel) that you should really investigate.

[John Stimpson]

Very good advice all around, Pete!

[bigwig]

I'd suggest "investing" some time/money into some sort of way to monitor detonation. The cheapest and easiest way is a mechanical det can(hit up google). Then you can take the next step and use a knock sensor and hook it up to your laptop and monitor via monitoring frequency and sound. The best method would be a fancy dancy electronic device that uses a knock sensor and filters out all the "other" noise and tells you what is or is not detonation. If you are using it for tuning, just let your ears/brain decide what is or is not detonation. Your brain/ears are pretty good at it.

[Eahab El Naga]

Although at first detonation seems to be the cause, which can be caused by a variety of problems, Intake temp, cylynder temp, spark plug temp, lean condition, or mapping, It seems to me that a piece of debris had its way with your engine. Since the cylinder next to it looks flawless, unless, you are using two tables of igniton timing (which i doubt) and different spark plugs on each cylinder (which i also doubt) that rules out Intake temp, spark plug heat range, and spark timing. So that leaves either a bad fuel injector, of a piece of debris. And from my experience, detonation seem to not be so concentrated in one are like that and seem to be more of the sand texture.

Good luck

Eahab

[KU_Racing]

This is an interesting post, and there is a ton of good advice here. I would just like to chime in with my 2 cents on the 'use your ears and brain' knock detection technique. I would recommend that anyone who has the time and expertise to set up a knock sensor or buy a professionaly engineered system do so. The reason I suggest this is twofold:

A) Some people can hear knock very early, but some people seem to have ears that cant detect it until its really loud. If the engine guy from your team has explained away an explosion with "well... I couldnt HEAR any knock..." than you know which side of the fence he falls on.

B) suppose for a second that there is a 1-10 scale of knock- 1 being 0 knock at all times, 10 being extreme amounts of loud, piston melting/ringland cracking/money evaporating preignition or detonation. Your ears, if you are very sensitive, start to hear knock at about level 6 or 7. A OEM calibrated knock sensor can hear it at maybe 3 or 4. Damage, however, starts to occur at level 1. Im not saying that everybody needs a knock sensor; im a firm believer that intelligent tuning practices will prevent damage. Im just saying that I wouldnt recommend a "bump timing until you hear knock, then back of 3 degrees" tuning method.

Just my 2 cents.

[bigwig]

I'm going to disagree with just about everything you typed. First and foremost the human brain is a MUCH more powerful at processing information than a computer. Computers are fantastic for doing one task over and over, but as far as thinking, they can't. Detonation is incredibly easy to hear when the sound is focused on and amplified. This is all a det can does.

If you try to hear detonation while sitting in a room with the engine roaring or while driving down the street, it more or less won't happen until you have incredibly violent detonation. BUT if you stick your ear over the engine, you can hear significantly lower levels of detonation. If you use a det can, you can hear incredibly low levels of detonation which you will never get with a knock detection system.

The biggest flaw with electronic knock detection systems is the piss poor job they do with filtering out the other noise. Very few systems do a good job. There is a lot of tuning of noise based on rpm and specific engine qualities. It incredibly difficult to accomplish via electronics. Light detonation often gets drowned out by the other noise and the electronic knock detection device won't pick it up. Your ears on the other hand will notice SOMETHING. Then you can change the timing to see if that something is detonation or just engine noise. Again, your brain is significantly more powerful than electronics.

Also your OEM calibrated knock sensor comment is completely invalid. Most aftermarket electronic knock detection systems use OEM Bosch flat response knock sensors. The sensors are plenty fine. Its just the level of filteration of the other noises that generally are not focused on. Because power levels on OEM setups are rarely on the brink of safety and they are tuned ridiculously safe so the need for a finely tuned knock sensor system is not a focus. They just need something just incase. But the sensor its self is more than capable. Its just the level of noise filtering done on OEM applications is just not that thorough.

The end result is a mechanical det can costs $15 and takes an hour including the trip to home depot to build. If it doesn't work well, then you haven't lost much. I can promise you that you won't regret spending the $15 or the hour.

[KU_Racing]

The bottom line for my argument is that now matter how good your ears may be, they will not ever be as good as a finely tuned high quality microphone set up to detect one range of frequencies at a very low intensity (knock sensor). Your brain is excellent at processing information, but your ears are not nearly as efficient at receiving it. You will never convince me or anyone who has done the amount of dyno testing and tuning that I have otherwise. This is not just my opinion, but the opinion I have confirmed after several discussions of this type with other people that have far more tuning experience than me or any other student who posts here. Knock is much more difficult to detect in an engine such as an F4i. The small bore, high compression, and especially high RPM levels all contribute to this; the problem is compounded when knock occurs in only one cylinder, which is usually the case unless something in the calibration is very far off the mark. Their are other benefits to an electronic system- a computer can think faster than anyone I know of, which can save an engine from certain death. An electronic sensor becomes much more powerful when it is wired into the ECU and given control over timing or fuel or both. Good luck getting your ECU to recognize the output from your det can and alter your maps accordingly.

I am not suggesting buying jo-bobs super 350 chevy knock detector system off of ebay. I am saying take the time to develop a system that actually works. I have no argument about the fact that typical systems are ineffective as a result of the low quality of the filtering. But with the right amount of design time or money spent, whichever you prefer, you can get a system that works consistently and accurately.

I should also add a slight correction to my last post- what I meant, and should have said, was a knock system using an OEM sensor with a high level of calibration. I agree that most OEM systems are tuned extremely conservatively. An OEM system will never be a plug-and-play, accurate system, considering the level of OEM knock sensor sensitivity as well as the effects of hardware variation on knock frequency.

One last thing: if you want to go push your ear against the valve cover of your engine while its at full load and 13,000 RPM, be my guest. Ill watch from behind the bulletproof glass of the control room.

[Pete M]

We've used a variety of methods to detect detonation, all of which eventually relied on a human ear and brain.

I'm very keen on electronics in general and would love to have a well-tuned knock sensor tied into the ECU to automatically back off on timing. We've never tested with a knock sensor though because, from what i know, you can't just bolt any old knock sensor to an engine, it has to be tuned. There is no stock knock sensor on the F4 (although i believe some of the larger bikes have them). So we're left with either retrofitting one from another engine or making our own. The thing is, the only way i can see to make sure it works (or tune it if it's DIY) is to make the engine detonate and see what the sensor says. I'm just not really willing to risk an engine on that.

Has anyone put a knock sensor on their engine and actually confirmed it worked? Did it actually detect detonation before it was audible anyway?

[KU_Racing]

There are non-destructive methods of generating a knock frequency in an engine for the purpose of testing a knock sensor setup.

at work we have an electonic knock sensor setup that detects knock very very early on, before it is audible. It has shown knock where the running conditions of the engine would lead the tech to believe there is none until the plugs were removed and the engine inspected. Tuning a knock sensor really isnt that difficult if the electronic filter is set up correctly. As BigWig said earlier in this thread, a reagular old bosch sensor is plenty capable with the right hardware to run it.

[VFR750R]

We did that a couple of times in 2003 on a YZF600. I feel the reason we burnt up a single piston and not all 4 was batch fuel injection with E85. Even though we flow check injectors, the timing of fuel delievery with such a high heat of vaporization fuel with low volitility creates fuel distribution variance.

During your testing were you past restrictor choke? Think about your superchargers pressure ratio after choke and where your operating point is on the efficiency map for that charger. Basically, what was your intake temps?