Individual Cylinder EGT Testing

[Enki]

Large images.

As you can see, there's a curb/lip inside which is what you hear the tube hitting against when you shake the manifold. Also, the inner diameter might be as tight as 35mm, since my micrometer can't quite get in there far enough (it does hourglass slightly towards one side).

Hope this helps gents.

[Spectrix]

Originally Posted by Ziggo I am curious as to the compression results by cylinder - possibly playing a shell game with the coil packs. You can see how much of a difference timing makes, any extra triggering delay will have an impact. 2* of timing at 6k rpm is a difference of .05 milliseconds.

There might also be some merit in swapping around the individual injectors in case either of them are under or over performing.

How do you find time to eat and sleep between all this logging?

[djuosnteisn]

Originally Posted by Enki Large images. As you can see, there's a curb/lip inside which is what you hear the tube hitting against when you shake the manifold. Also, the inner diameter might be as tight as 35mm, since my micrometer can't quite get in there far enough (it does hourglass slightly towards one side). Hope this helps gents.

When you think about two cylinders funneling through this section with exh pulses very close to one another... the EGT data isn't too incredibly surprising.

But i would expect to see the difference between 1 & 2 and 3 & 4 to normalize at lower flow rates (lower loads). According to Clint, they don't.

I'm wondering if it's a sensor calibration issue or something.


Clint, how difficult is it to access these sensors? Any chance you'd be able to swap the cyl 4 sensor with cyl 1 and see if the delta moves with the sensor?

[Dano]

Tappin

[86AmishMs3]

Also I think we might be getting ahead of ourselves as well. Clint is there any way we can do a validation of the sensors. You know as well as I do that measurement system validation is pretty important. Cylinder 1 RTD might be sloped differently...who knows.

Now i'm not saying it would be easy to validate the probes in excess of 1000F...but can you switch cyl 1 rtd pretty easily and see if the results are the same or if they change.

Don't want to be a debbie downer...but this is just something that worries me...that we're trusting results if we're not sure that each sensor is within +/- 1% at 1500F

[phate]

Heh, I'd likely have to pull the manifold back off, but I'll see if I can squeeze cyl4's out (number 1 is super easy). And it's even less fun to pull the manifold now because I have to pull the studs from around cylinders 3 and 4....I won't be able to do anything major like that for at least 2-3 weeks. I'm in the process of moving right now.

[86AmishMs3]

Originally Posted by djuosnteisn When you think about two cylinders funneling through this section with exh pulses very close to one another... the EGT data isn't too incredibly surprising. But i would expect to see the difference between 1 & 2 and 3 & 4 to normalize at lower flow rates (lower loads). According to Clint, they don't. I'm wondering if it's a sensor calibration issue or something. Clint, how difficult is it to access these sensors? Any chance you'd be able to swap the cyl 4 sensor with cyl 1 and see if the delta moves with the sensor?

Guess we're on the same page too

[djuosnteisn]

Originally Posted by 86azms3 Guess we're on the same page too

I can't read... i just rented the movie :friday:

[86AmishMs3]

gimme the data and I'll do a Type 1 gauge study...hoping for a cpk>1.33

[phate]

There is a pretty consistent temperature difference between cylinders 1 and 2...would you guys accept swapping those two probes as a first step towards sensor validation? We can investigate further if the temp Δ follows the sensors.

Those two are easy!

[86AmishMs3]

Originally Posted by phate There is a pretty consistent temperature difference between cylinders 1 and 2...would you guys accept swapping those two probes as a first step towards sensor validation? We can investigate further if the temp Δ follows the sensors. Those two are easy!

I would say if you do that then cylinder 2 is now the hottest...then that's pretty damning evidence that cylinder 1 rtd has a high bias.

[djuosnteisn]

Originally Posted by phate There is a pretty consistent temperature difference between cylinders 1 and 2...would you guys accept swapping those two probes as a first step towards sensor validation? We can investigate further if the temp Δ follows the sensors. Those two are easy!

Yeah, that's fine. Just something to show whether or not the trends follow the sensors or the cylinders.

[phate]

Sounds good, 1 and 2 it is. I doubt it happens tonight, but mmaaaayyyybeeeeee I can squeeze it in tomorrow night.

[blackmicamps6]

is there an award for best new thread of the year?

[Lex]

When you let off the gas and injectors shut off the engine is just pumping air working against the vacuum created by the closed throttle body. So air is the only thing flowing over the probes and their temperature is dictated by this and their surroundings (+ any error in readings).

My bet is that the surroundings and proximity to the hot turbo has more to do with this cooling pattern than anything else.

[aggierandy]

I love all the experiments you guys come up with to test your theories. Phate, you're gonna be busy for a while.

RapaTapTap

[Easter Bunny]

is there a Super Gearhead Status?

[Roddiy]

@phate;

I hate to be picky but this has been bothering me:

On the very first picture you posted, it looks like the probes on cyl. 1&2 are farther from the exhaust opening (deeper into the runner) whereas the probes in cyl. 3&4 are almost at the flange.


Based on the picture with the red hot exhaust stream, it would make sense that they would pick up a hotter measurement than 3&4, as peak heat happens at the meeting point.



My suggestion is to try to get the probes evenly close to the flange so the distance between the probes and the actual exhaust stream is the same.




Just for clarity sakes, measure the distance between the opening of the exhaust mani (the flange part) and the probe, If I'm right cyl 1&2 will be farther than cyl 3&4. You get me?

[phate]

Placement measurements are in op

Originally Posted by phate I tried to keep the placement of the probes even between cylinders, and as perpendicular with the runner as possible. These are the measurements from inside each runner. I'd say they're within .050" Cyl 1: Base- 1.50" || Tip - 1.50" Cyl 2: Base- 1.45" || Tip - 1.25" Cyl 3: Base- 1.35" || Tip - 1.50" Cyl 4: Base- 1.35" || Tip - 1.45"

^^This was to the mating surface on the flange.

[Roddiy]

Originally Posted by phate Placement measurements are in op ^^This was to the mating surface on the flange.

Oh ok, I read those as the distance between the probe tip to the material.


So this confirms what I said, I honestly think a more accurate set of results would be reached if the probes shared the same distance to the flange surface.

[phate]

By all means, you are welcome to try your hand at this. I seriously doubt that 1/4" difference in placement is the cause of the temperature variance readings between the cylinders.

[blackmicamps6]

Originally Posted by phate By all means, you are welcome to try your hand at this. I seriously doubt that 1/4" difference in placement is the cause of the temperature variance readings between the cylinders.

agreed, for temperatures around and above 1000 degrees, 1/4" isn't going to be a difference of +/-25 degrees... i'm no gearhead but in an inclosed area like that the temps should not make a substantial difference.

[Roddiy]

Originally Posted by phate By all means, you are welcome to try your hand at this. I seriously doubt that 1/4" difference in placement is the cause of the temperature variance readings between the cylinders.

I really am inclined in giving a hand at this, my engine will be pulled sometime this summer for a rebuild and if I can find a spare exhaust mani to work with I'm definetely interested in doing something similar.



MAJOR EDIT:
I was looking at the whole thing wrong, lower EGTs = higher combustion chamber temps.



Edit:
@phate;

I hope you didn't feel insulted at the way I suggested the issue, I think you're doing God's work, I'm just trying to figure stuff out like the rest of us. Again, I'll be trying my hand at this as soon as I can as well.

[phate]

I'll put some notes in the OP about how I would place the sensors, IF I were to do it again. The stock turbo location really kills ease of mani install with my EGT placement, which is why I had to pull the studs. If/when I pull the manifold again, I'm going to use bolts, rather than studs.

^^Done. If I think of anything else, I'll add to it.

[Roddiy]

Originally Posted by blackmicamps6 agreed, for temperatures around and above 1000 degrees, 1/4" isn't going to be a difference of +/-25 degrees... i'm no gearhead but in an inclosed area like that the temps should not make a substantial difference.

Perhaps it's not ONLY the distance, but also the angling of the sensor (the discrepancy between Tip distance and Base distance on the chart above) that could be throwing off results.



Let's also wait for the sensor switch and whatnot, it could have something to do with it.

[Tokay444]

Originally Posted by Lex When you let off the gas and injectors shut off the engine is just pumping air working against the vacuum created by the closed throttle body. So air is the only thing flowing over the probes and their temperature is dictated by this and their surroundings (+ any error in readings). My bet is that the surroundings and proximity to the hot turbo has more to do with this cooling pattern than anything else. Here's something to try - do a pull on a nice empty road and then turn the engine off and leave the key in the ON position and keep your foot on the floor to open the throttle body and facilitate fresh air to enter the motor and see if it has any effect on the EGTs during decel.

with the engine not turning over?

[Lex]

Originally Posted by Tokay444 with the engine not turning over?

Hmm, no the engine would have to turn over which means it will start back up. Good catch. I guess there is no way to have the injectors shut off with the throttle body wide open and engine turning over due to DBW.

This was a silly test to identify how much airflow through the motor without actually firing affects the EGT readings during coasting. This could give us a better insight into how the exhaust manifold and the probe locations with respect to the collector affect the EGT readings under load.

[atvfreek]

Originally Posted by Roddiy I really am inclined in giving a hand at this, my engine will be pulled sometime this summer for a rebuild and if I can find a spare exhaust mani to work with I'm definetely interested in doing something similar. I wouldn't think 1/4" is such a big deal either, but most conclusions up until today suggest that cyl #3 is the hottest, but your results are showing otherwise, so I'm looking for the discrepancy here, and the only one I can find thus far is the distance between probe/surface material. Edit: @phate; I hope you didn't feel insulted at the way I suggested the issue, I think you're doing God's work, I'm just trying to figure stuff out like the rest of us. Again, I'll be trying my hand at this as soon as I can as well.

This test is testing egt's. We all assumed that #3 would be hotter. As you raise ignition timing, egt's get lower and combustion chamber temps go up. So, if #3 is leaner, faster flame speed, lower egt's and higher combustion chamber temps. Egt's can kill exhaust valves and do give a very good idea of what the engine is doing, but with higher combustion chamber temps the pistons are taking the heat/beating. Which imo is why we are seeing the most failures in 3 and 4.

Sent from your couch

[Boosted Beluga]

So lower isn't always better? To an extent of course.

[atvfreek]

Originally Posted by Boosten Ya(josh) So lower isn't always better? To an extent of course.

Obviously to an extent. Running lean across the board will raise egt's and combustion temps. You need that happy medium. Keep some of the hp in your pocket type of thing.


I'm liking where this discussion is going with the EM design too. As I'm thinking it starts to be a limiting factor, so hopefully other tests will follow thank you again Clint for taking the time to do all this!

Sent from your couch

[djuosnteisn]

Originally Posted by Boosten Ya(josh) So lower isn't always better? To an extent of course.

Lower can be worse than higher more often than not. Even detonation will lower EGTs. The trick with EGTs (because they are so dependent on where they're mounted) is to identify a "known good" value, and tune for that.

[phate]

As a general, simple rule:

Richer AFR: Cooler combustion chamber, cooler EGT's
Leaner AFR: Hotter combustion chamber, hotter EGT's

^^Another way of saying that is that AFR affects both in the same manner (both up, or both down)

More Ignition Advance: Hotter combustion chamber, cooler EGT's
Less Ignition Advance: Cooler combustion chamber, hotter EGT's

^^Ignition advance shifts where the energy is used. With more advance, there is more time spent pushing the piston down, thus less energy is expelled through the exhaust.

[Dano]

Originally Posted by atvfreek This test is testing egt's. We all assumed that #3 would be hotter. As you raise ignition timing, egt's get lower and combustion chamber temps go up. So, if #3 is leaner, faster flame speed, lower egt's and higher combustion chamber temps. Egt's can kill exhaust valves and do give a very good idea of what the engine is doing, but with higher combustion chamber temps the pistons are taking the heat/beating. Which imo is why we are seeing the most failures in 3 and 4. Sent from your couch

Exactly this. EGT and cylinder temps are 2 different things. the overwhelming evidence proves cyl 3 and 4 get fubared

The difference in my bore measurements between 1-2 and 3-4 were not even close. Deformation was rampant in 3-4 and 1-2 looked ok

Is this coolant heatsoak related or something else. Who knows.


Tappin

[Roddiy]

Originally Posted by atvfreek This test is testing egt's. We all assumed that #3 would be hotter. As you raise ignition timing, egt's get lower and combustion chamber temps go up. So, if #3 is leaner, faster flame speed, lower egt's and higher combustion chamber temps. Egt's can kill exhaust valves and do give a very good idea of what the engine is doing, but with higher combustion chamber temps the pistons are taking the heat/beating. Which imo is why we are seeing the most failures in 3 and 4. Sent from your couch

I'm a fucking dumbass, you're right, lower EGTs in cyl. 3&4 means higher combustion chamber temps which means more strain on the rods/pistons.


I suppose a goal for safety would be to pull timing (like @phate; did) from those cylinders to reduce chamber temps and even then out.


^ that was just me rephrasing what's going on, it helps me understand.



Your post made it all click for me. Carry on.

[phate]

Originally Posted by Roddiy I'm a fucking dumbass, you're right, lower EGTs in cyl. 3&4 means higher combustion chamber temps which means more strain on the rods. I suppose a goal for safety would be to pull timing (like @phate did) from those cylinders to reduce chamber temps and even then out.

No, not necessarily. We know that timing has a significant affect on power production, so pulling timing simply to bring EGT's closer together is a bit premature.

It could more easily be an AFR issue (hardware/software) since we see more drastic shifts in EGT from changes in AFR - not to mention the engine doesn't run as smoothly with pulled timing in 3 and 4. It could be sensors, it could be the exhaust manifold, it could be all sorts of things.

We're at the point of needing to identify why they are different, then go from there.

[Roddiy]

Originally Posted by phate No, not necessarily. We know that timing has a significant affect on power production, so pulling timing simply to bring EGT's closer together is a bit premature. It could more easily be an AFR issue (hardware/software) since we see more drastic shifts in EGT from changes in AFR - not to mention the engine doesn't run as smoothly with pulled timing in 3 and 4. It could be sensors, it could be the exhaust manifold, it could be all sorts of things. We're at the point of needing to identify why they are different, then go from there.

Ok that makes sense, then the next step(after the sensor switch) should be to gather data with a ported intake mani?

[Dano]

Lol

We are at the variable elimination stage.

This will all come out in the wash as they say.

Still thread of the year in the making here.


Tappin

[Tomas]

There is a lot of misinformation regarding the relationship between AFR and EGTs here.

All else being equal EGTs are the highest at Stochio. They are cooler when you run richer mixtures AND also cooler when you run leaner mixtures. It's a curve that peaks at Stochio.

EGT vs A/F gauges | EGT Tuning | Exhaust Gas Temperature

[phate]

Originally Posted by Tomas There is a lot of misinformation regarding the relationship between AFR and EGTs here. All else being equal EGTs are the highest at Stochio. They are cooler when you run richer mixtures AND also cooler when you run leaner mixtures. It's a curve that peaks at Stochio. EGT vs A/F gauges | EGT Tuning | Exhaust Gas Temperature http://www.msextra.com/doc/ms3/pix/EGT_Temp_AFR.png

I understand that, but I don't expect to be (or anyone else to be) running leaner than stoich at any point, lol. [Maybe I was unintentionally close with the 14:1 trial, though.] It was just a quick and dirty rule in the context of what we'll actually be running.

[Mfinlay04]

Originally Posted by Tomas There is a lot of misinformation regarding the relationship between AFR and EGTs here. All else being equal EGTs are the highest at Stochio. They are cooler when you run richer mixtures AND also cooler when you run leaner mixtures. It's a curve that peaks at Stochio. EGT vs A/F gauges | EGT Tuning | Exhaust Gas Temperature

I don't believe many people are too interested in crushing EGT and as Clint hinted at if your running stochi or leaner at WOT you have more problems than this thread can fix.

Sent from my Galaxy Nexus using Tapatalk 2