Falsche LMM-Temperaturkompensation? Oder nur simple Physik? | Posts 16+

Hi,
Where are the 40% in the correction factor?

Quote:

LuMa(50°C) = LuMa logarithm value * ((LLT + 273) / (50 + 273))^0.6

You're changing the exponent to 0.6, which makes the "correction factor" dependent on the LLT value, but it changes it very little or not at all. For example, if the LLT is 70°C, the change is only 3%.
hg
Herbert.

Herbert wrote:

Du änderst die Potenz auf 0,6, das macht am "Korrekturfaktor " abhängig vom LLT Wert wenig bis nichts. Nimm mal 70°C für die LLT, dann sinds 3%.

Trotzdem verlaufen die breiten gelben Linien mit dem Exp. 0,6 unabhängig von der realen LLT (im Rahmen meiner Logs) praktisch immer auf gleicher Höhe, im Gegensatz zum 1. Diagramm with Exp. 1.
Even when I normalize the logs with an exponent of 0.6 to an LLT (Lower Limit Temperature) of 30°C or 100°C, the wide yellow lines shift uniformly and still remain at practically the same height.
This shows me that the Exp. 0.6 term in the formula largely eliminates the LLT dependence of the LuMa/LD relationship .

Do you mean the thick and the thin yellow lines?
Herbert.

Herbert wrote:

Do you mean the thick and the thin yellow line, respectively? Herbert

The thin line shows, as a reference, the uncorrected log value division LuMa/LD-Actual * 100, including the full influence of the LLT.
What's more important to me is the thick line (in both images), because that's the normalized line from which I want to remove the influence of the LLT – which, apparently, was successfully achieved in the second image.

Quote:

because the correction factor (LLT + 273) / (50 + 273) apparently overcompensates for the effects of the LLT on the volumetric flow rate.

The correction factor doesn't fully compensate, the problem lies in the different measurement locations. The LLT (charge air temperature) is measured in the charge air cooler, while the LMM (mass air flow meter) measures the cold intake air. Considering the entire LD (charge air) system, the charge air is not as hot as the LLT sensor indicates.

That would be my only explanation that comes to mind spontaneously. Otherwise, the formula is correct, even if it assumes an ideal gas.

Hi Blackfrosch,

Blackfrosch wrote:

LLT is measured in the LLK

. This is correct for the SMICs of the G4 and 9N platforms, as the sensor is located in the outlet airbox on the hotter rear side of the cooling net. However, my diagram comes from the BUK, whose LD-LLT sensor is located approximately ½ m behind the FMIC, and by that point, the initially temperature-stratified charge air should have already mixed into a largely homogeneous temperature distribution due to several pipe bends.

Quote:

Looking at the entire LD system, the charge air is not as hot as the LLT sensor indicates.

In the BUK, the maximum LLT in 3rd gear is approximately 5,000 meters above sea level (üAT), and in 4th gear, it's around 12,000 meters üAT (the heat sink effect of the FMIC significantly delays the LLT increase; at maximum speed, the LLT stabilizes at around 20,000 meters üAT). Where should the lukewarm charge air, after the measurement point, lose a significant amount of temperature as it travels through the engine bay, which is already warm, and into the hot engine block?

Quote:

The LMM measures the cold intake air. That would be my only explanation that spontaneously comes to mind.

Nevertheless, the LMM (Lambda Measuring Module) can only measure the mass flow, which is limited by the temperature profile after the compressor and the displacement (x charge pressure). The intake air, as a base temperature, should not significantly affect the normalized intake air mass (e.g., normalized to 50°C).

However, the fresh air drawn in is heated before the intake valves close, which, due to the resulting pressure loss, leads to a stagnation effect of the mass flow. And this mass flow stagnation effect increases with the amount of heat added to the intake air within the engine block. Let's assume, for simplicity, that the wall temperature of the intake passages and cylinders is constantly at 90°C, the same as the coolant temperature; then, the aforementioned stagnation effect would be...
a) At 90°C, the LLT (Low-Level Trigger) is not present (because the LLT is not changed until the inlet valves close).
b) the lower the LLT is below 90°C, the stronger the effect.
This explanation, in my opinion, logically leads towards the observed phenomenon that, with a decreasing Lower Limit Temperature (LLT), not much more mass enters the cylinders than one would expect based solely on the LLT value... and it seems to me that the aforementioned correction with the exponent of 0.6 is probably just a coincidence for my BUK logs }.

Okay, I've tried again to bring the thick yellow curve to the same level as the logged LLT, but again without using any exponents for the factor (LLT + 273) / (50 + 273).

Under the assumption that...
a) The surfaces in the intake manifold and behind the intake valves would have an average temperature of 70°C (which would heat up the entire air charge when the engine is running at a very slow speed, before the intake valve closes).
b) At normal speeds, the air charge heats up by a maximum of 20% of the difference between the logged long-term fuel trim (LT) value and the specified 70°C, until the exhaust valve (EV) closes.
A suspected LLT (Low-Level Turbulence) pattern, representing the air currently being drawn into the engine, is located above the log data and is shown in red and dashed lines in the image.

Should I input these suspected LLT values into the normalization formula?
LuMa "50°C" = LuMa log value * (estimated LLT + 273) / (50 + 273)
Then, the thick yellow curve resumes its course independently of the logged LLT, remaining at a largely similar level.
-> In my opinion, this is a practical and simple calculation model for quickly estimating the influence of leakage losses on the load-torque relationship (as a permeability value for the entire gas flow path), especially when comparing different chargers on the same engine block based on log data.