Oh, yeah? Well I’m streaming the talks on two monitors and doing no work!
10 Likes
One model - the Passat from 2011 - 2014 use the Urea injection. The others (Golf, Jetta) do not.
The 2015 + passat is an entirely new engine (EA289 engine code) and also uses Urea and has a different emissions setup.
I read “Urethra injection” here and winced.
6 Likes
Far more interesting than I expected.
In a nutshell:
Urea + water (“AdBlue”) is injected to convert NOx to nitrogen & water. (This has to be refilled every so often, like windshield wiper fluid.) Too little injected, and you get incomplete conversion (but better than not doing anything). Too much and you emit ammonia, which is VERY bad.
So complex software, and several different engine models (displacement, temperature, thousands of parameters are modeled) figure out what needs to be done. Different models are used because some models don’t work under certain conditions (high temperature, high power…)
(How complex? You think your tachometer is simply reporting the RPM of your engine? It’s actually taking RPM plus 20 other variables, massaging it through 12K worth of dense code to give you a number that’s more useful than the actual speed.)
To determine how much AdBlue needs to be injected, the computer monitors intake temperature, outlet temperature, mass flow, NOx concentration in, NOx/NH3 concentration out, oxi-cat temperature and many other things.
But sometimes this model doesn’t work, and may inject too much urea, so there’s an alternative model, which is much simpler, uses fewer input parameters, but this model ALWAYS under-doses.
So both models are always calculated, and then one output is selected, and there’s code to select the model. Some of the selection parameters are thing like:
*atmospheric pressure < -32768 hPa (this can never happen)
*air temperature > 3276.8 K (we’re not driving on the sun)
*air temperature < 0.1 K (we’re not driving in outer space)
*engine temperature > -3276.8 K (this is ALWAYS true)
(!!! below absolute zero !!!)
If one or more parameters is true, the alternative model is used.
So the alternative model is ALWAYS selected.
So instead of dosing an expected 2.5 liters of AdBlue per 1000 km, only 0.6 liters are used. Owners like this because they don’t have to get AdBlue refills as often, but there’s almost always more NOx being emitted than if the proper model were being used.
BUT… sometimes the regular model is being used, so there must be more to the selection algorithm…
There are some weird unlock criteria based on parameters like
*engine & fuel temperature > 15 C
*atmospheric pressure > 920 hPa (<~750 m above sea level)
*driving profile must follow given limits defined by 7 piecewise linear min-max curves of distance driven since time of motor start.
Interestingly, this unlock criteria routine is named “Acoustic Function”, though it has nothing to do with sound.
Now, when emission testing is done, the car is put on a dynometer which allows it to run without moving. The standard driving test of effective distance versus time are shown below based on the NEDC (New European Driving Cycle). (Note that no one would drive a car like this.) They also do things like over-inflate tires, tape up all gaps, add light diesel to the oil to reduce viscosity, etc.
Note the overlay below. This is how the car “knows” that it is is being tested for emissions. (This is one of the true keys of the scandal.)
Now if you were very careful to drive within these prescribed limits, the dosing will actually work properly, but as soon as you exceed the limit, the alternative model takes over and does no dosing at all , and you get high emissions. This is exactly what the presenter did, and you can see at the big red circle the dosing stops completely when the limit is exceeded.
So there you go…
The first half outlines the socio-political environment of why and how car companies deal with pollution mandates, and the fact that every process decision at every level in the automobile business is highly regulated from both within and without, and documented at every stage, so it is absolutely impossible for a “rogue engineer” to have snuck this in.
So when they say things like “we have no findings on the involvement of the supervisory board, or the management board presented” that means that they in fact DO have findings on the involvement, they simply weren’t PRESENTED.
And what they mean by “involvement” isn’t what you think. The supervisory board “supervises”. The management board “decides”. So by definition they aren’t “involved”. Truthy. Very truthy. It was perpetrated by the corporation.
16 Likes
Thanks you for this summation. You deserve all the Likes.
2 Likes
Does ‘playing the video on a tablet while doing something else on a computer’ count as watching it on a second monitor?
This is a really good video, and the two halves form a nice complement. The first half drives home the concept that this was intentional on the part of the leadership, and the second half shows some of the nitty gritty behind how the cheating actually works.
I’d say “de facto yes”.
1 Like
I think it’s well worth listening to both speakers. The first guy gives some great context, helping with the understanding of the pressures car manufacturers are under, the fudges that are already allowed in emissions testing, and how the development process is thoroughly audited and the ‘small group of rogue engineers’ theory is bullshit. He really understands this stuff well, and it shows.
1 Like
Very much this ^^^ Another vote for listening to the first half from me.
1 Like
So in the end I did watch it, very much worth the time! Watching both halves is recommended.
All the while I was thinking the cars would drive slower with their new firmware update. Why else would they have a “cheat” in their code if it does not improve the cars specs (like the 0 to 100 time or the torque output?) in a meaningful way while it sneaks around the environmental demands.
It seems however that this is not the case. Can anyone explain why they did not just apply the right program? Is it really because they did not want to bother their customers with refilling a tank every 1000 kilometers? That seems like a small gain compared to the large scale fraud it required…
My understanding (from a friend who has a VW TDI and is fairly knowledgeable about cars) is that properly applying the pollution controls does in fact reduce fuel economy and performance. It won’t be the vehicle that was advertised…
This topic was automatically closed after 5 days. New replies are no longer allowed.
