Redhot exhaust manifold

[guuskolo]

Hello,

in the past year I installed a Haltech ECU, with all necessary components for controlling the engine.
after properly calibrating, and testing for a few months, I installed a turbo from an Audi TT (K04-22) on the engine.

Currently, everything is running fine and the car feels great. (I am running 0,5bar (7psi) of boost with standard compression (9.5:1))

Yet I have one problem. My exhaust is getting quite hot...
https://pasteboard.co/JIddpI1.jpg

This picture was taken after some rather normal driving. No highway, a bit of 80Km/h back-road driving, almost not under boost.

Now I did the following things:
- timing: everything atmospheric pressure and under is BMW m10 standard based (slightly advanced even) everything under boost is based on audiTT except slightly retarded (Audi has compression 9.0:1, Bmw has 9,5:1)
- lambda: lambda under light load is 1, lambda under higher load ( -20kpa --> 0Kpa) is 0.95 to 0.85, lambda under boost is 0.85 to 0.70
- valve lift: compression test gives reasonable numbers compared to Haynes, so should be fine
- valve timing: never removed timing chain so i expect this is still standard.

The only thing I am still planning to do, is making the injection maps for lambda 1 everywhere (when not under boost), because it is sometimes running into lmbd 0.95-0.90 under low load. This might cause some after burning.. But again, I am not sure it will solve anything.

My question to you guys: am I missing something? Or is a slightly glowing exhaust normal?


Note: my exhaust manifold is a log-type. The headers are the same size as the ports on the head, and the pipe towards the turbo is the same size as the turbine inlet. So while a log-type is not optimal, I don't think it is restricting .

[reggid]

did you check the ignition timing with a timing light to verify ?

[guuskolo]

Yep.

If i make it spark at TDC, (advance = 0) the crank pully mark will line up with the aluminium pin sticking out the front.

[Lemon98]

Is the manifold made from a suitable material?

[guuskolo]

Stainless steel grade304, 3mm thick.
Yet that should make a difference wether it glows or not. Only when it will stop glowing hehe..

[Turbo-Brown]

It's worth measuring the exhaust gas temperature I'd say.

Metals will glow red around 600C, can't see the image unfortunately so can't see how red it is.
Couldn't tell you what a normal EGT is but sure the internet will!

Worth determining if there is a problem before trying to fix it!

My first suspicion if the EGT is high and you have AFR between 14.7 and 12.5 would be ignition timing.

Rich AFRs give cooler EGT, weaker mixtures tend to give higher EGT.

People often believe that retarding the ignition timing will lower temperatures. It will lower chamber temps, but retarded timing taken too far will send EGTs rocketing since the combustion event can still be taking place when the exhaust valve opens, if taken to extremes.

The real danger there is that the heads fall off the exhaust valves.

You mention that you've copied the Audi timing?
The Audi presumably has a 5 valve head whereas your M10 has only 2 valves.

You'll see a big difference between timing maps comparing 2 valves to multi valve setups, would perhaps have a look at some typical examples and compare to your maps.

Been a really long time since I mapped an engine, but I recall 2 valve engines needing way more advance than multi valves as mixture motion is better with multi valve heads.

[guuskolo]

It's worth measuring the exhaust gas temperature I'd say.

Metals will glow red around 600C, can't see the image unfortunately so can't see how red it is.
Couldn't tell you what a normal EGT is but sure the internet will!

Worth determining if there is a problem before trying to fix it!

My first suspicion if the EGT is high and you have AFR between 14.7 and 12.5 would be ignition timing.

Rich AFRs give cooler EGT, weaker mixtures tend to give higher EGT.

People often believe that retarding the ignition timing will lower temperatures. It will lower chamber temps, but retarded timing taken too far will send EGTs rocketing since the combustion event can still be taking place when the exhaust valve opens, if taken to extremes.

The real danger there is that the heads fall off the exhaust valves.

You mention that you've copied the Audi timing?
The Audi presumably has a 5 valve head whereas your M10 has only 2 valves.

You'll see a big difference between timing maps comparing 2 valves to multi valve setups, would perhaps have a look at some typical examples and compare to your maps.

Been a really long time since I mapped an engine, but I recall 2 valve engines needing way more advance than multi valves as mixture motion is better with multi valve heads.

Oh sorry, didn't know the picture wasn't loading, I've added the link!

Regarding AFR, it is definitely rich, did this for that chamber temperatures as you also state. But i was thinking maybe it is cause such slow burn-rates that this causes higher exhaust gas temperatures... it was just a thought.

your comment on 5-valve / 2-valve mixing makes complete sense btw.. I will be investigating this!

also, I installed a thermocouple and will post below

[guuskolo]

In the mean time i also installed a thermocouple POST-turbine, and did some measurements:
- 35kmh, 2nd gear, 2000 rpm(ish) --> 550DegC
- 50kmh, 3rd gear, 2000 rpm(ish) --> 600DegC
- 70kmh, ?nd gear, 3000 rpm --> 750DegC
- 100kmh, 5th gear, 2500 rpm --> 750DegC
- full load 0 - 140kmh peak (at 140kmh) --> 890DegC

note that this is POST turbine, these will likely be 100-200 Degrees higher in manifold because of higher pressure (and loss of entropy)

do these numbers look strange to you guys? i was a bit worried with the 750 degrees at 70kmh onwards... that just looks like it's too high. But i must say i don't have enough reference to judge.

[Turbo-Brown]

Maybe worth putting the thermocouple in the manifold, as you say, the turbo is powered by heat so the temperatures measured are going to be lower than in the manifold.

Would certainly encourage you to look at your timing

[guuskolo]

Maybe worth putting the thermocouple in the manifold, as you say, the turbo is powered by heat so the temperatures measured are going to be lower than in the manifold.

Would certainly encourage you to look at your timing

Yeah i know that would be better.
But we've had these thermocouples break (at work) in the past. Furthermore, this thermocouple has a maximum of 1000C°, which i will reach, which will make the measurement less usefull.
So pre-turbine i will not do..

Do you have any reference regarding the temperatures ?

[Turbo-Brown]

Must admit I've no idea what constitutes a reasonable EGT.

What type of thermocouple do you have?

K type goes above 1200C if you can lay your hands on one. Suspect if you've a K type with a lower limit it'll be the insulation or some other limiting factor rather than the usable range of the junction itself.

There's a vague hint about EGT on Wikipedia:

https://en.m.wikipedia.org/wiki/Exhaust ... ture_gauge

[fixedwheelnut]

Have you checked mixture or for air leaks? I have only seen this on cars running weak

[guuskolo]

Must admit I've no idea what constitutes a reasonable EGT.

What type of thermocouple do you have?

K type goes above 1200C if you can lay your hands on one. Suspect if you've a K type with a lower limit it'll be the insulation or some other limiting factor rather than the usable range of the junction itself.

There's a vague hint about EGT on Wikipedia:

https://en.m.wikipedia.org/wiki/Exhaust ... ture_gauge

It's a K-type. I thought they had their limit at 1000C° Whoops

[guuskolo]

Have you checked mixture or for air leaks? I have only seen this on cars running weak

It's got a lambda sensor post turbine, and the mixture is quite rich (0,85 - 0,70 under boost). So i don't suspect any issues there.

[Turbo-Brown]

The junction temperature is fine above 1000C, if the datasheet for your sensor says lower It's possibly some other limiting factor like insulation materials.

Bung it on the manifold and see what happens! They're cheap sensors !

As a quick experiment, you could advance your boosted timing map 10degrees and do another set of temperature logs.

[guuskolo]

The junction temperature is fine above 1000C, if the datasheet for your sensor says lower It's possibly some other limiting factor like insulation materials.

Bung it on the manifold and see what happens! They're cheap sensors !

As a quick experiment, you could advance your boosted timing map 10degrees and do another set of temperature logs.

Wouldn't that quick experiment cause knocking in every possible position ? (I mean, 10 degrees is a lot)

[Turbo-Brown]

There's a lot of difference between the timing you run in a 2v and a multi valve engine.

Take your laptop along and change it back and don't just go for full throttle!

[guuskolo]

In the following link I pasted a picture of my current ignition map:

https://pasteboard.co/JIdcKYG.png

again:
Below atmospheric pressure (-100 - 0) is based on standard calibration BMWM10B18, with some timing advance.
Boosted partition is based on Audi TT map, yet slightly retarded for safety.

[Turbo-Brown]

Can't see the image again unfortunately.
Interesting article here:
https://www.bimmerforums.com/forum/show ... n-4v-heads

Pretty sure your car will take off like a scalded cat with more advance!

[guuskolo]

Can't see the image again unfortunately.
Interesting article here:
https://www.bimmerforums.com/forum/show ... n-4v-heads

Pretty sure your car will take off like a scalded cat with more advance!

f*** me again... do you happen to have a good image posting tool ?

This is a really nice picture! Also explains the twin spark setups (Alfa head suck )
He is throwing numbers around in the ranges of 10-15 degrees, but does mention they are hypothetical.
I will be doing some research on this. Maybe I'm able to find some M10 or M20 turbo ignition maps.

One other thing that is bothering me tough. When I'm not under boost, temperatures still easily reach 750C° (this seems very high still). While in these areas I am using standard BMW timing.

[BenHar]

[Turbo-Brown]

I'm very much not an expert on EGTs so be interesting to see some representative numbers for other engines.

One possible explanation is that at light throttle your compression ratio is really low so you extract less energy from the fuel.

Simultaneously, the turbo isn't doing much so that isn't extracting much energy from the exhaust so perhaps the EGT isn't all that high on light throttle, it's just that the turbo isn't skewing the results as much.

Really would recommend fitting your thermocouple to the manifold and reviewing the ignition map.

[DanThe]

Have you checked mixture or for air leaks? I have only seen this on cars running weak

I had an M20 engine in with a BTB manifold that was running massively rich, it had actually ripped the bores out and had piston slap! The manifold was glowing red and it had melted the clutch reservoir! There is too much fuel going in for a clean burn so as the remains get shoved out the head it continues to burn in the exhaust manifold causing the extra heat

[guuskolo]

Hello all,

I wanted to post some extra info, because I've been doing a lot of research into ignition timing.
First of all: The ignition timing chart I used (BMW original) somehow seems to be EXTREMELY RETARDED given the engine's specifications: https://pasteboard.co/JJIARcz.png
The Germans where either drinking too many weissbier, or they made a super safe map for shitty '80's gasoline. i'd guess both.

Secondly: The Megasquirt guys are great.
In the Megasquirt manual you will find clearly explained what timings should give you a good ball-park map. With reasonable explanations for the choices you have to make.
So for anyone making a new timing map, I recommend looking at this.
- http://www.megamanual.com/ms2/configure.htm#make
- http://www.megamanual.com/begintuning.htm#spark

and this brings me to my new map:
https://pasteboard.co/JJIHyPh.png

which is still very safe, but I will be doings some tuning soon

[Turbo-Brown]

Strange that, 80s petrol in the UK at least, was leaded and delicious and nicely knock resistant.

Really interested to see how advancing your timing affects EGTs.

Did you move your thermocouple in the end?

[guuskolo]

Strange that, 80s petrol in the UK at least, was leaded and delicious and nicely knock resistant.

Really interested to see how advancing your timing affects EGTs.

Did you move your thermocouple in the end?

Haha i'm sorry to dissapoint, no i didn't...

In this case i'd have to disasseble too much, for what seems (to me) to be just a slight delta in egt (and some increase in response time).

I will however update this post when i've got the new timing map! Would be interesting to see the new tempertures.
From looking at other people's data i'm expectimg something in the 700-800C° range (post-turbine)

[Turbo-Brown]

What would be really interesting is pre and post turbine Vs air flow and boost!

The turbo is driven by heat, the harder it works, the more heat it extracts