Diesel(rücklauf)temperatur beim PD | Posts 16+

According to your assumption, the VL (Vorlauf - supply) line to the ZK (Zirkulationspumpe - circulation pump) should be approximately 50 - 60°C warm.


Um, that's not an assumption!
That's what was said during a PD training session that I attended.
If someone told me something that was nonsense (which I don't believe), I can't change that.

However, the values displayed by 'Dieselmartins' indicate that the temperature is already relatively stable.
When dealing with the VP (presumably a piece of equipment), it's important to remember that it's fully exposed to the cold airflow. 43°C (approximately 109°F) is quite high in those cold ambient temperatures.

The 'trial run' should, however, produce at least some results.
I think you should probably use the filter housing for the temperature check instead.

goettmann wrote:

The "trial run" should, however, produce results to some extent. I think you should rather use the filter body for the temperature check.

Okay, here's the result:

The return trip (after approximately 2 minutes of predominantly rolling motion on the last section, with the engine running warm after a 21 km journey) was uncomfortably hot in idle - I would estimate it to be around 50-60°C.

However, the initial temperature was significantly lower, estimated to be around 40°C.
In particular, no difference could be felt between the angle adapters on the tank and the engine sides.
If a temperature mixing process had occurred in the DiFi unit, the inlet and outlet temperatures of the primary flow would have had to be different.

My observation: In my car, it seems that above approximately 50°C of coolant temperature, all diesel fuel without additives bypasses the diesel injection filter (DiFi) and returns to the tank.

The return trip (after approximately 2 minutes of predominantly rolling motion on the last section, with the engine running warm after a 21 km journey) was uncomfortably hot in idle - I would estimate it to be around 50-60°C.

However, the initial temperature was significantly lower, estimated to be around 40°C.
....

Okay, I just drove a G4 with an AJM engine. I observed similar behavior to what you're experiencing, but the temperatures were about 10-15°C higher than you estimated. I measured it with an infrared thermometer (coolant temperature = 51-58°C; oil temperature = 70-85°C).
Unfortunately, I can't say whether the thermometer is an hourglass or whether it's accurate. However, it is calibrated.
I measured the angles at the main distribution board (ZK) and the distribution fuse box (DIFI). There was only a small difference measurable between the DIFI and the main supply cable at the ZK. The fuel inside the pipes might be slightly warmer.

A mixture is definitely occurring. I don't believe your fuel in the tank reached your estimated 40°C (let alone the 50°C that I measured).

EDIT: I can't find my VP37 training materials anymore... that's where the mixing process was explained.

goettmann wrote:

I measured it with an infrared thermometer (VL=51-58°C; RL=70-85°C).

With equipment like that, it would be interesting to perform a cross-check by reading the coolant temperature using VAGCOM. I tend to trust the VAGCOM value more than a newly purchased thermometer, especially if it's not properly calibrated.

(Therefore, I estimated my perceived load level to be in the range of logic values corresponding to "no motor load".)

Quote:

A mixing will definitely take place.

The ultimate test for this would, in my opinion, be the comparison of the fuel tank and engine-side vapor lock (VL) valves of the direct injection (DiFi) system. Have you directly compared these points?

...
A mixing will definitely take place.
The ultimate test for this would, in my opinion, be the aforementioned comparison of the fuel tank and engine-side VL (Vent Line) stub of the DiFi. Did you directly compare these points? }

I couldn't take accurate measurements. The inlet nozzle was slightly cooler (2°C). However, a meaningful measurement is difficult due to the lack of 'mass' (nozzle length). With the G4, the hose starts directly at the nozzle. I can probably only get rough estimates with the infrared thermometer. The hose, on the other hand, was noticeably colder (25°C).

goettmann wrote:

Bimetal is correct. Basically, this bimetallic switch measures the housing temperature. This should correspond to the VL temperature within a few degrees. Function and structure, see: http://f23.parsimony.net/forum49387/messages/112895.htm

Hmmm... when I look at the image in the link and try to imagine the assembled valve, then...

1. According to IMO, the thin "Knackfrosch" (a type of valve) alone determines whether the RL (return liquid) flows back into the "DiFi" (likely a specific tank or system) or continues into the tank.

2. A continuously variable mixture (in the sense of a control mechanism that truly deserves that name) does not seem possible, but rather only two fixed, predefined states.
In this case, the volume fractions directed towards the DiFi (differential pressure filter) and the tank do not necessarily have to be 0 and 100%, respectively. Instead, in the "hot" state, the internal cross-sections could be designed such that, for example, 80% flows to the tank and the remaining 20% flows into the filter (this is just an arbitrary example).

3. There are a few millimeters of insulating plastic casing between the DiFi flange and the "Knackfrosch" device, which, on the other hand, should be constantly flushed with coolant from the RL diesel engine in both switching positions.
This suggests that the switching, in my opinion, depends almost entirely (99%) on the actual temperature and only 1% on the differential temperature.

What do you think about that?

Hmm... when I look at the image in the link and try to imagine the assembled valve, then...
...

The valve shown there is available in several versions from VW, not just the one shown. Therefore, I wouldn't necessarily apply that information to the PD engine.
Basically, the structure is the same.

A continuously variable mixture (in the sense of a regulation that truly deserves that name) does not seem possible, but only 2 fixed, predefined states.

Yes, I agree. However, that shouldn't be a problem given the tolerance required there. Let's just keep the standard tolerance at 20°C.

In this case, the volume fractions in the direction of the DiFi (differential pressure) and the tank do not necessarily have to be 0 and 100%, but in the 'hot' state, the internal cross-sections could be designed such that, for example, 80% flows to the tank and still 20% flows into the filter (an arbitrary example).
Yes... I agree. There are probably various versions for that reason.

Therefore, the switching, in my opinion, depends almost entirely (99%) on the actual temperature and only 1% on the differential temperature.
No agreement... at least not 100%.
While the bimetallic strip is heated from above by the return flow (RL), it is simultaneously cooled from below by the supply flow (VL) coming from the DIFI. This creates a mixed temperature, and as a result, the valve opens or closes regardless of the higher RL temperatures.

I estimate the influence of the RL temperature to be around 60-70%.

I'm not entirely sure if this part actually switches at a specific temperature, but rather at a temperaturedifference.
As far as I remember, that's how bimetallic strips should behave.

For example, you have a supply temperature (VL) of -10°C and a return temperature (RL) of 70°C, and the system mixes them. If the supply temperature now rises to 50°C, the difference (e.g., 20°C) is exceeded, and the mixing ratio is adjusted to a minimum. The cold fuel (-10°C) cools the system down again, and the process starts over.

If we were to extend the idea of the temperature difference, the flow rate to the tandem pump would largely depend on the return line temperature. This would also solve the problem of the 'unknown' supply line temperature.

goettmann wrote:

While the bimetallic plate is heated from above by the hot water (RL), it is simultaneously cooled from below by the cold water (VL) coming from the DIFI. This results in a mixed temperature, and the valve opens/closes despite higher hot water temperatures (RL).

Well... regarding the mechanism, it probably doesn't necessarily have to be a "click frog" that only switches between two positions. It could just as well be a "softly" bending bimetallic strip, which would allow for a continuous distribution of the incoming RL between the filter and the tank.

But I can't agree with Dr. about the cooling system coming from the DiFi.

Plastic and standing diesel (located between the DiFi and Frosch sections, with the exit towards DiFi closed) act as almost perfect thermal insulators. Therefore, very little heat can escape from there, while the hot RL constantly "overpowers" the system with heat (comparison: a voltage divider with a 10 Ohm resistor on top and a 1 kOhm resistor on the bottom).
The (soft?) frog itself can be only a few millimeters thick, so it is immediately heated from the top (RL), and a "warm cushion" forms underneath it, directed towards the bottom (DiFi).

In a mixed operating mode (between its end positions), the frog would even be completely surrounded by the hot rail, so that contrary to the hot diesel current towards the DiFi , no "cold" would reach the frog and influence its behavior.

According to this, the incoming RL temperature alone determines the mixing ratio; the DiFi temperature has practically no influence.

Quote:

I'm not entirely sure if this part actually switches on at a specific temperature, but rather at a temperaturedifference. As far as I remember,

bimetallic strips should behave in this way. The movement is caused by the different thermal expansion coefficients of the upper and lower layers of the bimetallic strip, even with a completely homogeneous temperature distribution in the sheet.

An insulating layer between the alloys is not necessary, as it is not required to create a significant temperature difference.

Quote:

If we were to extend the idea of the temperature difference now, the flow rate to the tandem pump would largely depend on the return line temperature. This would also solve the problem of the "unknown" supply line temperature.

According to my understanding of the frog diagram, this is not the case (reasons see above):
For example, at a return temperature of 60°C, everything flows directly into the tank, regardless of whether the differential temperature (DiFi) or the supply temperature (VL) is -10°C or +40°C.

"Only as long as the engine oil is cold (and the engine load is low) such that the coolant temperature is below the trigger temperature of the thermostat, will the engine oil be warmed by mixing it with the coolant in the oil cooler; in my opinion, the heat leakage from the coolant hose into the oil cooler can be safely neglected in this case."

Well... regarding the mechanism, it probably doesn't necessarily have to be a 'click frog' that only switches between two positions. It could just as well be a 'softly' curving bimetallic strip, which would allow for a continuous distribution of the incoming RL between the filter and the tank.

Possible, but I'm more inclined to believe in a 'hard' switch.

But I can't follow your explanation about the cooling system coming from the DiFi.
...

Then, find another explanation for the fuel preheating.

We are discussing facts that are present in every description (this was the case even with older models). And these facts can even be confirmed through VL measurements (see the logs from 'Dieselmartin').
Ultimately, how it's implemented or how it works doesn't really matter.

You can see a preheating temperature of 60°C on the 'Dieselmartin' device (I measured 50°C yesterday). If the mixing valve completely releases the RL fuel into the tank (as your theory suggests), that would mean the 60°C warm fuel is coming directly from the tank. I find that hard to believe.

goettmann wrote:

Then, find another explanation for the fuel preheating.

The standard loop over the return
If the supply line (VL) is too cold, the return line (RL) temperature is also low. The control valve opens the RL outlet to the mixing valve (DiFi), and the supply line is heated.
If the VL (variable liquid) and RL (return liquid) are warm enough, the frog in the RL will start to reduce the preheating, and more cold diesel from the tank will flow into the VL, etc.

A Knackfrog with abrupt switching between full VL heating and zero VL heating would, in any case, make significantly less sense than a continuous control system, especially with a PD motor...

Quote:

We are discussing facts that are present in every description (this was the case even with the older versions). And these facts can even be confirmed through VL measurements (see the logs from "Dieselmartin"). How it is ultimately implemented or how it works is, in the end, completely irrelevant.

Yes and no... The starting point was the precision of the diesel density compensation for the PDE opening duration when the vehicle speed is not detected and a significant mixture between the high-pressure and low-pressure sides in the tandem pump.

As long as the frog is operating within its active operating range, the VL temperature can be considered approximately constant and therefore used as a reliable correction factor.

From the cold start to that point, the VL temperature can, at best, be calculated as an estimated value, and the actual injection mass is therefore defined with considerable uncertainty.

However, a different meaning of a strong mixture of volatile liquids (VL) and reactive liquids (RL) in the pump has become clear to me:
Reducing the temperature fluctuations that the diesel engine experiences directly after combustion, between maximum torque and maximum power, would alleviate the strain on the VL (variable load) temperature control system. Perhaps only mixing can prevent these control oscillations.

Quote:

If the mixing valve completely releases the return fuel (RL) into the tank (as your theory suggests), that would mean the 60°C warm fuel is coming out of the tank.

You are forgetting the heating that occurs as the fuel passes through the VP (fuel pump), the ZK (common rail), and the PDE (high-pressure pump) – and the control loop that, after a surge of cold diesel from the tank, will naturally reactivate the fuel pre-heater (VL) as soon as the return fuel temperature (RL) drops sufficiently.

P.S.:
Of course, our discussion has hardly any practical value.
However, in terms of understanding the VL temperature regulation, its relationship with the PD density compensation, and considering potential exotic errors, it has certainly been helpful to me (and hopefully to some other readers as well).
Your link to the image of the RL controller was, incidentally, very helpful .

Then, find another explanation for the fuel preheating.
The standard loop over the return
If the supply line (VL) is too cold, the return line (RL) temperature is also low. The control valve opens the RL outlet to the mixing valve (DiFi), and the supply line is heated.
If the VL (variable liquid) and RL (return liquid) are warm enough, the frog in the RL starts to reduce the preheating, and more cold diesel from the tank flows into the VL, etc.

Now you're saying exactly what I was trying to make you understand before.

A Knackfrog with abrupt switching between full VL heating and zero VL heating would, in any case, make significantly less sense than a continuous control system, especially with a PD motor...

Agreement.
Therefore, I also doubted that the 'spring frog' depicted in the image could be directly transferred to the PD. The one from the PD might have a completely different operating characteristic.
. . . The starting point was the precision of the diesel density compensation for the PDE opening duration when the vehicle speed was not detected and a significant mixture between the high-pressure and low-pressure sides in the tandem pump.

Okay.
There are three mixing points: at the plate-and-frame heat exchangers (for cooling and steam bubble separation), in the tandem pump (for cooling and steam bubble separation), and in the DIFI (primarily for heating the process liquid).

As long as the frog is operating within its active operating range, the VL temperature can be considered approximately constant and therefore used as a reliable correction factor.

Yes, I agree.

From the cold start to that point, the VL temperature can, at best, be calculated as an estimated value, and the actual amount of fuel injected is therefore defined with considerable uncertainty.

That's why I suspect that the temperature correction is only applied starting from a certain kilowatt temperature.

However, a different meaning of a strong mixture of volatile liquids (VL) and reactive liquids (RL) in the pump has become clear to me:
Reducing the temperature fluctuations that the diesel engine experiences directly after combustion, between maximum torque and maximum power, would alleviate the strain on the VL (variable load) temperature control system. Perhaps only mixing can prevent these control oscillations.

That's exactly what I've been trying to explain all along.
If everything goes off track, you'll make the silliest mistakes (jerks --> regular oscillations).

...In doing so, you forget about the heating that occurs during the VP cycle...

I assumed a temperature of around 10°C. That would also be consistent with my measurement.

...
However, in terms of understanding the VL temperature regulation, its relationship with the PD density compensation, and considering potential exotic errors, it has certainly been helpful to me (and hopefully to some other readers as well).

I just wanted to express that the temperature control systems in HD injection systems (including CR systems) are no longer as straightforward as they are in naturally aspirated or turbocharged engines.
I'm sorry that this discussion has devolved into basic research. I'm not entirely innocent in that.
I also hope that it will be helpful to some people.


It's ironic that while modern cars are packed with electronics, a fuel temperature control system can suddenly be implemented in a completely traditional way, without all that fancy technology. However, the electronics blindly rely on the proper functioning of this 'stone-age' system.

goettmann wrote:

Now you are saying exactly what I wanted to make clear to you before.

But now, with full conviction

Quote:

I'm sorry that this discussion has turned into a fundamental research topic. I am somewhat to blame for that.

But it was 99% factual and fair.
And because (unfortunately, due to such lengthy dialogues) this is no longer a given, you will receive a big thank you .

Quote:

However, the electronics "rely" blindly on the correct functioning of this "Stone Age" system.

Hmm, if a developer reads this, this system (as far as I know) which is generally very reliable, will probably be replaced by a vulnerable, high-tech digital gimmick in one of the next engine generations...

Hm, if a developer reads this, this (AFAIK) generally very reliable system will certainly be replaced by a vulnerable, high-tech digital gadget in one of the next engine generations...

Super disaster with a diagnostic port?

Please don't!!