proper lambda values before and after catalytic converter

Post Reply
puff
contributor
contributor
Posts: 2961
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

proper lambda values before and after catalytic converter

Post by puff »

when googling for that question, I found a table stating that before and after the converter both sensors should be reading 14.7. looks a bit confusing.

it is mentioned that the second lambda sensor is used to control the efficiency of the catalytic converter, but how exactly do we interpret those readings? what should be the sensors readings to say that the converter has failed?

The sensor itself is called O2 sensor, so will it detect CO2? Or in this case it is considered bonded and not being counted by the o2 sensor?
puff
contributor
contributor
Posts: 2961
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: proper lambda values before and after catalytic converter

Post by puff »

no ideas?
User avatar
Dron_Gus
contributor
contributor
Posts: 450
Joined: Wed Nov 13, 2013 1:11 pm
Location: S-Pb
Github Username: dron0gus

Re: proper lambda values before and after catalytic converter

Post by Dron_Gus »

Catalic errors sounds like "too slow reaction" or "too fast reaction". AFR after catalytic should be smoother than before.
puff
contributor
contributor
Posts: 2961
Joined: Mon Nov 11, 2013 11:28 am
Location: Moskau

Re: proper lambda values before and after catalytic converter

Post by puff »

shall we switch to Russian?

на самом деле вопросов несколько больше.

я слаб в физике-химии. тот o2, который считает датчик - это именно o2, или o2 в составе co2 тоже?

не очень понятно, как это 14 частей воздуха к одной части топлива - формула топлива тут какая получается?
и почему датчик стоит после - ведь топлива уже не должно остаться (должно сгореть) - тогда каким образом определяется это соотношение?

откуда берется эта периодичность 0-1 в узкополоснике? я наивно предполагал, что эбу пытается удержать равновесие, а т.к. абсолютного равновесия не бывает, оно и прыгает то больше, то меньше.

too slow - в смысле частота? smoother - больше на характеристику амплитуды похоже. в любом случае не очень понятно, как работа катализатора может привести к изменению сигнала.

если отталкиваться от того, что датчик считает именно свободные o2, а реакция горения (дожигания) как раз приводит к тому, что этого o2 становится в итоге меньше? или там какие-то особые условия, и реакция горения в нейтрализаторе не такая, как в цилиндрах?


(кажется вот тут начинает потихоньку доходить - или не начинает)

что-то мне подсказывает, что правильнее теорию объяснять словами, что мы по составу выхлопных газов судим о том, насколько эффективно сгорела смесь. а эффективнее всего она горит, когда массовое соотношение остаточных углеводородов и o2 в выхлопе – 14,7. или я тут неправильно сказал?


но если карта такая, что нам нужна более богатая смесь, и на входе ШДК будет показывать условно 0,9, то на выходе из нейтрализатора что он будет показывать?
User avatar
Dron_Gus
contributor
contributor
Posts: 450
Joined: Wed Nov 13, 2013 1:11 pm
Location: S-Pb
Github Username: dron0gus

Re: proper lambda values before and after catalytic converter

Post by Dron_Gus »

Слишком много вопросов.
Узкополосный сенсор просто детектирует наличие несреагировавшего o2 в выхлопе. Есть-нет. Поэтому он постоянно туда-сюда скачет пытаясь нащупать грань. Которая в районе 14.7:1.
Широкополосный это узкополосный + насос. Который может закачивать o2 снаружи или наоборот выкачивать. Током насоса регулируется его производительность, а полярность - направление работы. Вся система все так же пытается узкополосный датчик загнать на грань 14.7:1. Только при наличии насоса это можно сделать в широком диапазоне AFR, просто добавляя или откачивая кислород.
За катализатором обычно стоит узкополосник. Если он переключается слишком часто - значи катализатор потерял способность накапливать кислород.
mck1117
running engine in first post
running engine in first post
Posts: 1493
Joined: Mon Jan 30, 2017 2:05 am
Location: Seattle-ish

Re: proper lambda values before and after catalytic converter

Post by mck1117 »

(attempting to continue the conversation here based on exactly 0 knowledge of Russian, plus Google Translate :D )

CHAPTER 1: Redox (reduction-oxidation) reactions, and your engine.
Oxygen sensors in general measure whether the atmosphere of the exhaust is reducing or oxidising. (warning: vast oversimplification incoming) If the environment is oxidising, there's an excess of oxygen. In a reducing environment, the opposite is true: it's oxygen deficient.

As you may have guessed, when an engine is running lean, there's an excess of oxygen in the exhaust. When it's rich, there are lots of chemicals in the exhaust that really want to strip oxygen from somewhere.

More on this here: https://en.wikipedia.org/wiki/Redox

Now let's apply this chemistry knowledge to engines. See the following diagram:

Image

When lambda is <1 (engine is rich), the carbon monoxide content is very high. When it's >1 (engine is lean), it's very low, because carbon dioxide is produced instead. (The emissions of oxides of nitrogen, on the other hand, are caused by high combustion temperatures - that's why it increases until around lambda = 1.1, then decreases again as lean misfire causes temperature to reduce).

CHAPTER 2: what does this have to do with your O2 sensor?
Oxygen sensors have a piece of ceramic in them through which ions can diffuse, which creates a current (it's just flowing charge - electrons or not). A narrowband sensor just exploits this effect: High voltage (around 1v) means that there's a lot of oxygen trying to diffuse through, and low voltage (around 0v) means there isn't. Since it provides very little granularity about how much oxygen is trying to flow across the cell, the characteristic curve of the sensor is very very steep at the lambda=1 point. A wideband O2 sensor simply adds the hardware to be able to sense how hard oxygen is trying to push across the cell, and in which direction.

The reason that vehicles with narrowband O2 sensors exhibit oscillation is that the sensor only provides information about whether lambda > 1 or not. So the computer slowly richens the mixture until lambda < 1, then slowly leans it out until lambda > 1 (and repeats).

CHAPTER 3: Hey, wasn't the original question about my catalytic converter?

Just like burning gasoline is a big redox reaction, so is your catalytic converter. Modern three-way cats do three reactions simultaneously:

1) Reduction of NOx to elemental nitrogen and oxygen:
NOx -> Nx + Ox
2) Oxidation of carbon monoxide to carbon dioxide:
CO + O2 -> CO2
3) Oxidation of hydrocarbons in to CO2 and water:
CxH4x + 2xO2 -> xCO2 -> 2xH2O

(note that steps 2 and 3 need oxygen to function - that's important)

It turns out that since these reactions are sort of at odds with each other (#1 prefers a reducing environment, while #2 and #3 prefer an oxidising environment), they all only work when you're right near the stoichiometric point. However: you get better catalyst performance if you oscillate the air-fuel ratio at 1-3hz. The large surface area of the catalyst can store significant amounts of oxygen, so it builds up during lean periods, then is consumed during rich periods.

(note: these reactions happen during combustion in the engine as well - it just doesn't do a very complete job of it. That's why you need a cat in the first place.)

CHAPTER 3b: So what's the downstream O2 sensor for/what's it supposed to see?

The ECU knows that the cat can store some oxygen, and it exploits this fact to check if it's working. Without a cat, one would expect a similar signal out of both O2 sensors: they should both oscillate more or less together, with a tiny phase shift. With a cat, the upstream sensor should oscillate as expected, but the downstream should either have a significant phase shift, or not oscillate at all. As the catalyst wears out, its capacity for storing oxygen is reduced, so the downstream gas composition will slowly become more like the upstream composition.
User avatar
md_sanci
Posts: 69
Joined: Thu May 30, 2019 6:13 am
Location: Romania
Github Username: shadowm60
Slack: Alex M
Contact:

Re: proper lambda values before and after catalytic converter

Post by md_sanci »

I was searching on this topic on how to make my car safe since i am running a CAT.
Ok, right now it is not yet fully ready, so rich and lean is there, at least the cat is not overheating and the smoke is not stinky.

I was wondering if it would be possible the following situation.
1. Set AFR table to 14.7 when not in the power band.
2. with the help of Lua change target AFR with 2Hz period to fluctuate from 14.2 - 15.0 and keep close loop.
3. do all of this only when engine is hot?

Could this work as a simple control? i mean driving in the city it is usually ok since AFR will fluctuate, but on the highway 30min doing the same speed with the same load.... that would make problems at some point.
Post Reply