GM 6.6 Duramax turbodiesel 3

[MKhan01]

Hi,

I am doing research on a turbo vane sensor for the 6.6 GM Duramax. The sensor rides on a cam which entail reads the position of the vanes. I am curious to know how hot that sensor will get. If the EGT is about 1200F, and the turbine housing is 1200F then wouldn't the sensor which is bolted to that housing be 1200F?

Also a generic question, what temperature does the turbine shaft reach at the bearings? Typical engine oil is 250F maximum before breaking down. Does the temperature drop down from 1200F to 250F at the bearings using oil as a form of coolant and lubrication?

btw I am a mechanical engineering student!

Thanks.

 

[moon161]

I'd say you have
1) a heat transfer problem. If the EGT is reasonably the high temperature, there's a gradient that the heat follows away from the hot side of the turbo.
2) a homework problem, maybe. If it's a homework problem, it's not for here.

 

[MikeHalloran]

There are thermal gradients everywhere.

When the turbo is working properly, the exhaust gas at the exhaust manifold interface may reach 1300F, but the exhaust gas exiting the turbine may be at 800F or less. The turbo turns heat into mechanical work.

The bearings are roughly midway between the turbo, which is of course hot, and the compressor, which is much cooler. If the bearings haven't coked, they are probably running within the oil's limits.

I am not familiar with the Duramax, but I have seen quite a few turbos running in anger. The various linkages attached to turbos are generally attached by bolts in cast bosses, linked to actuator/sensor/whatever by thin steel stampings with relatively low thermal conductivity (and smallish cross sections).

A proper thermal model of something attached to the turbo needs to account for a modest flow of atmospheric air flowing over the engine, which of course makes the gradients steeper. So your sensor is probably only a few degrees warmer than the air in the engine compartment.


Mike Halloran
Pembroke Pines, FL, USA

 

[MKhan01]

@Moon161. No this isn't a homework problem.

@mikehalloran. So even when the turbodiesels have EGT's of 2000F, the bearings and the shaft are within the oils limit below 305F(typical)? I have seen photos of complete turbos turning red, that sure has to effect the turbo vane position sensor and the oil temperature.

How would I calculate the heat loss from the exhaust to the shaft at the bearings?

 

[MikeHalloran]

Perhaps it would be educational for you to do a one-dimensional heat transfer problem, like predicting the junction temperature of a single transistor. That involves knowing or estimating:

The power dissipation at the junction.
The thermal resistance of:
- junction to case
- case to mica wafer
- through thickness of mica wafer
- mica wafer to heatsink
- heatsink interface to fins
- boundary layer of air flowing over fins
- ambient air

It's just a resistor summation problem, with resistors representing a lot of different actual things. Once you know the summmed resistances, you can calculate the junction temperature for any given power input, and the various interface temperatures that you can measure for checking.

Doing the same thing for a turbo is pretty much the same exercise, except in three dimensions, one of those side problems that textbook authors leave 'as an exercise for the reader'.

In the case of the turbo, there are two flows carrying away heat; the air flowing over the turbo, and the oil cooling the bearings. Plus of course radiation to its surroundings as it approaches brighter colors. Well, not to mention the heat that's converted to mechanical work in the turbo-expander, which substantially cools the exhaust gas.

My point is, the bearings and the vane sensor, and all the other components of the turbo, are not at one temperature; every part eventually stabilizes at some temperature, depending on the heat flows arriving and leaving. An accurate model is not going to be simple, and will probably comprise thousands of elements.







Mike Halloran
Pembroke Pines, FL, USA

 

[moon161]

Just thinking about the oil, that will cool the bearing, and also have a short residence time in the bearing.

 

[MikeHalloran]

I am used to marine Diesel turbos, which commonly have an engine coolant channel in the bearing housing. That also carries away some heat, in addition to what the oil carries away.

Terrestrial turbos more commonly rely on oil and air to cool the bearings, and I think synthetic oil is also recommended for that reason.


Mike Halloran
Pembroke Pines, FL, USA

 

[durablack2]

Just about every modern passenger vehicle engine I've seen in the past few years has a coolant and oil cooled turbocharger (may be different for the Duramax). Automotive companies have done a great job at keeping these things reliable - under full load conditions the coolant rise out of the turbo is similiar to the coolant rise out of the block - coolant temp into the turbo is the same as the coolant temp into the engine. Oil going to the turbo is already as hot as the oil in the block, so my understanding is that the coolant pulls more heat than the oil.

EGT's during full load are very much the same in and out of the turbine, which is why catalyst protection (i.e. enrichment) is so important - this may be different for diesel engines since EGT's aren't as high in diesels as they are in gasoline.

A good starting point may be to look at several manufacturers who design such sensors and see if there are any heat ratings for them on their websites. You could even contact them, tell them about your research, and ask them what temperatures the sensor needs to be rated for and why.

 

[Lou Scannon]

EGT's during full load are very much the same in and out of the turbine

Say what?!!

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[SwinnyGG]

Actually the opposite is true. EGTs near idle or very low load can be very nearly the same value on both sides of the turbine. As load goes up the delta across the turbine will increase. Depending on the turbine map there may also be points of operation where the partial load, part throttle delta between EGT and Turbine Outlet Temp will actually be higher than the full load/full throttle delta is.

In short, its complicated- but the temperature delta across the turbine is a direct measure of the amount of energy being consumed by the turbine to accelerate it or keep it rotating. If the temperature delta across the turbine is zero, the turbine is providing no power to the compressor side.

 

[gruntguru]

In round figures the turbine temperature delta is a function of the pressure ratio. Turbine power will be approximately proportional to temperature drop times mass flow.

je suis charlie

 

[turbomotor]

I think durablack2 may be correct, and his info may be based on empirical data for a spark ignition engine. In many spark ignition engines at peak power, there is considerable (almost 50%) of the exhaust flow that bypasses the turbine expander through an integral wastegate port. So the measured T2t may be very close in value to the measured T1t, depending exactly where the thermocouple is located in the turbine outlet flow stream.

 

[Lou Scannon]

So, what pressure ratio are we talking about across the turbine/wastegate?

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[turbomotor]

It depends.

Turbine (expander) pressure ratio (or expansion ratio, depending which direction you are thinking) depends on the demanded boost level, which drives the bypass valve (wastegate) position.

At the risk of going a bit off topic (as I agree the topic can be addressed with a relatively straightforward thermal mapping exercise):

Turbine work must equal compressor work (if you allow me to assume that bearing loss is insignificant for the purpose of this discussion).

So mdot * Cp * Delta T (expander) = mdot * Cp * Delta T (compressor).

Delta T (compressor) = Tin (PR^((k-1)/k)-1) / cmpr eff

So expander work demand can be determined. Then one needs to iterate the expander mass flow and pressure ratio based on the (mapped) flow characteristic of the turbine expander.

So expander pressure ratio, expander mass flow, and wastegate flow are a function of boost demand and turbine nozzle size (A/R), which defines the turbine flow characteristic.

So it all depends on what size turbine nozzle (A/R) is on the engine, and what boost goal is dialed into the control system.

 

[Lou Scannon]

just give us a pressure ratio

"Schiefgehen wird, was schiefgehen kann" - das Murphygesetz

 

[tbuelna]

The temperature at the turbine shaft journal surface is what matters. And this surface is fairly well thermally isolated from the turbine wheel itself and turbine housing surfaces exposed to hot exhaust gas. It is fairly safe to assume that the flash temperature within the turbine side journal bearing hydrodynamic oil film does not exceed the capability of the engine lube oil, which is probably around 350 degF. The mass flow of lube oil thru the journal bearings is based on the cooling required to keep the bearing materials within safe limits.

 

[MKhan01]

Ok, so basically this sensor sits on a cam that connects to the vanes. The tip of the sensor is lubricated in oil, so tip is being cooled. I can see how the center of the turbo is cooled with oil along with the tip of the sensor. I don't see any lines for antifreeze. I have attached a pic.

Would anyone have a specific textbook where I can read up on the heat transfer of the turbo?

 

[moon161]

If it oil lubed and not a dry assembly, you're under the cook flash point of the oil then. Looking at the picture, there's a well already there. Can you just put take out position rader and put a temp sensor in there? You could just put a thermocouple through a bung and pot it.

 

[MikeHalloran]

Your school should still have a librarian.
Go to said person and ask for help finding the book you want.


Mike Halloran
Pembroke Pines, FL, USA

 

[tbuelna]

The (prox?) sensor shown in your drawing is likely installed with a tight fit into the turbo center housing. So the sensing face will likely be at a temperature close to that of the surrounding center housing material. The worst case temperature condition this sensor will likely experience is from heat soak after the engine is run at high load and rapidly shut down.