TSB EG-0127T-1014 (SW update to avoid cylinderhead gasket failure)

In perspective, I use and monitor an OBD2 reader constantly displaying those water and oil temperature and 10 other parameters every time I fire up the system. That said, it only reports what the OBD2 sees.

 

Better probably to say what we can't measure. We get production engines, delivered with one temperature sensor in one location.

They most likely have fairly detailed finite-element models of their blocks and heads, able to model the heat-moving effects of the coolant at various pump speeds and the heat production from combustion and friction at various loads. Beyond that, they may well have heavily-instrumented test-stand engines with temperature readings from numerous places.

Exactly, and the OBD2 sees only what the ECM sees, the reading from one sensor located at the water outlet from the cylinder head. (And what the combination meter sees, which is the reading from one other sensor located in the return path from the exhaust heat exchanger, but that's not what the ECM is relying on.)

And the very crux of this TSB seems to be that when the pump isn't running fast enough, the temperatures throughout the engine won't be kept uniform enough by the flow of coolant; the trickle (probably an exaggeration, but gets the point across) of coolant will get quite hot as it passes through hotter spots, lose a lot of that heat as it passes through cooler spots, and end up looking ok by the time it passes the sensor at the water outlet.

They probably have the resources to demonstrate and quantify that. Our OBD-II readers do not.

The ECM's algorithm for choosing the water pump speed is probably not just like a simple thermostat, watching to see the coolant temperature go up and then reacting to that. The ECM has much more complete information available to it, like the exact amount of power it is producing with the engine, and how much heat production corresponds to that. So there's no reason they would have to only wait to see the effects in the coolant temperature and react only then.

Chances are whoever designed the original control algorithm had not spent enough time modeling the heat distribution through the engine in the so-called "severe driving conditions" the TSB refers to.

Oops.

 

Battled the procrastination monster this morning and finally had them update my ECM to 34715700, which is the version those two US TSBs both recommend for a US 2010.

I suspect this version includes the tweaked water pump algorithm, though of course they don't come out and talk about that in the US TSBs the same way they do in Europe. And it's later than the purge-valve-kablooie fix version, so I shouldn't have to worry about that any longer either.

Of course nothing changes the 172k miles already put on under the old water pump algorithm, but any time is probably better than later.

One thing I hadn't thought of, but important to people who live where inspections happen: don't get an ECM reflash too close to when an inspection's due. The reflashed ECM, not too surprisingly, doesn't remember its old monitor results, so it won't be inspectable until all the new ones get stored.

 

any differences (seat-of-the-pants) while driving?

 

The idle rpm during initial warmup may be increased a little. It was turning 1472 this morning until the coolant temperature hit 30℃. Other than sounding a little different during that first minute or so, it didn't seem to affect driving at all.

I'm sure the seat of my pants wouldn't notice any change in water pump rpm.

 

Did you just take the car into your dealer and say "Hey could you update my software to version 34715700" and it's easy as that?

I try never to visit dealers if I can ever help it, so I don't have any relationship with any of the dealers for the variety of cars I own.

 

That's pretty much what I did. More specifically, I said "when can you get me in to update my ECM to calibration 34715700" and we agreed on 10:30 Tuesday, and I showed up then with the hatch area tidied up and the cover already off over my 12 volt battery, because I know they have to hook up a supplemental power supply there for the reflash. And I pulled the fuses for the electrical mods I added that they wouldn't know about.

The other thing I did was read through either TSB (T-SB-0103-12 or T-SB-0027-16) to make sure I asked for the right new calibration ID for my car. The 34715700 that I asked for is only the right one for my car; you should check the table in either of those TSBs to see which one goes with your car.

 

Very helpful. I have a 2010 so it looks like I need the same software as you. Despite my aversion to dealers maybe I'll call and find out what they'd charge.

 

What would the change be then, runs the water pump more/sooner/faster?

 

That seems to be the gist of the TSB published in Europe.

As our US regulatory body seems happy to accept TSBs that don't say what got changed, I'm only speculating that the update I've applied here (to a version that's simply mentioned in two different TSBs from around the same time and mentioning related symptoms, but without saying a word about what's changed in the update) is effectively the same one.

In any case, it's at least newer than the other update that addressed the backfire issue occasionally blowing hoses off people's manifolds and damaging their purge valves, so I should be able to rest a little easier about that anyway.

 

@ChapmanF

Did you have to have an engine code for them to perform this update?

 

This seems like it would be much more likely to be a cause of head gasket issues than the EGR to me.

I just cannot imagine how clogged EGR passages, either in the cooler or in uneven distribution in the manifold, could cause failures of the head gasket. At least not on a car with a knock sensor. If there was ever detonation the ECU would pull timing immediately to take care of it. Even if it was only in one cylinder. If there was truly uneven EGR flow causing uneven AFRs, the ECU would still not allow detonation to occur. And if it did, we would hear it.

Hot spots in the cooling jacket seem much more likely - especially since the non-hybrid versions of these motors don't have any HG issues. Those cars don't stop and start the motor all the time, which means they get hot and stay hot rather than going through continuous heat cycles.

 

Does the car monitor individual cylinder's Air Fuel Ratio's?

 

No, I just mentioned I'd seen the newer ECM calibration mentioned in two different TSBs and the one in my ECM was older, and asked them to update it.

I've never been very convinced by the arguments that involve heating, but the ones involving detonation can be found discussed in the literature easily enough. The Toyota engineers may have imaginations more vivid than yours, because they have also programmed the ECM to pull the timing preventively whenever it is aware of an EGR valve or flow problem. They do not seem to think it a good idea to just wait to detect detonation and then react to it ... at least not when they have evidence of a condition that's going to cause it.

 

No, but uneven AFRs won't blow a head gasket as long as it doesn't detonate, which the knock sensor does monitor.

Uneven AFRs would show themselves by lowering the MPG and rough running as well. If you were at WOT and had 20:1 in two cylinders and 10:1 in the other two, the 20:1 cylinders would be detonating so it would be pulling a bunch of timing (and thus reducing power and MPG). Meanwhile you'd be flushing fuel down the drain in the other two.

 

Maybe it's good to remember that the whole magic of EGR is that it doesn't change AFR. The engine can be fed an air/fuel mixture right around the desired 14.7:1 ratio, and then by introducing exhaust, which behaves as neither air nor fuel, there is simply less of the air/fuel mixture in the cylinder, but its AFR is unchanged.

The dilution with exhaust spreads the reacting air and fuel molecules farther apart, and slows the advance of the flame front. The combustion happens more gradually, over more of the piston's stroke, but it's still complete combustion, without too much air or too much fuel left over.

 

It could certainly affect AFR if more air is displaced by exhaust gas - which I think is Mendel's point that if one EGR passage were completely clogged, that cylinder would get more air (oxygen) than the other cylinders, despite getting the same amount of fuel. And that could cause a lean condition in one cylinder. It's just that the knock sensor would immediately hear this and pull timing until it was gone.

 

It's not even specifically Mendel's point; I've made the point about detonation from insufficient EGR in a cylinder myself. And it's not my idea either, it's in the literature for SI gasoline engines with EGR.

It just isn't because of a change in air-fuel ratio. It's because of the loss of the EGR effect of spreading the air and fuel molecules farther apart using exhaust, which is neither air nor fuel.

And, again, as to whether detecting the detonation after the fact with the knock sensor and pulling timing then is good enough, you're not disagreeing with me but with the Toyota engineers, who definitely programmed the ECM to pull timing if it is aware of any EGR delivery problem, even before the knock sensor detects anything.

Trouble is, the ECM only monitors the EGR valve and overall flow, and is not aware of any difference in flow through the four manifold passages.

 

Thoughts:

1. Of the intake EGR tributaries to the individual ports it's usually cylinder one that clogs first, and the head gasket failures are usually at the wall between cylinder one and two. Correlation?

2. When an individual cylinder is running lean, does detonation (premature combustion of the air/fuel mix ignited by glowing carbon if I"m not mistaken) happen immediately, or could there be chronic over-heating of that cylinder, punctuated by occasional detonation.

3. P0301 (misfire on cylinder 1) seems to go hand-in-hand with the head gasket failures on gen 3..

 

Again: "running lean" is kind of a spurious term to come up in a discussion of EGR.

The whole point of EGR, without which the engineers would not be using it, is that it allows adjusting the spacing of the air and fuel molecules in the mixture, without making the mixture richer or leaner. The mixture stays the same, the spacing and flame propagation speed changes.

When the engine is running tuned for a certain % EGR and doesn't get it, detonation results because of the change in propagation speed, not because of a change in mixture.

We're sort of off on a double OT now; this thread is about TSB EG-0127T-1014, which is about the water pump control issue that Toyota believes accounts for their head gasket concerns.

Then we got off onto EGR, which is not about the water pump (and has plenty of other threads where it's on topic), and then we got off onto AFR and "running lean", which is not about EGR.