Turbo oil seals - a solution?

[Kranz]

Quite a few people seem to be having turbo oil seal problems recently, where oil leaks past the turbine seal and produces puffs of smoke in the exhaust.

 

While this isn't a terminal failure in itself, it may lead to occasional oil pressure loss at the bearings which could lead to increased bearing wear. This increased wear will, over time, lead to turbo failure.

 

So, what is causing the seals to leak??

 

Current theory is that decatting reduces the exhaust back pressure in the turbine, so that the pressure differential between the turbine housing and the oil at the bearings puts more strain on the seal andcan allow oil to pass throught the seal to the turbine side.

 

This theory is probably correct, as the turbine seals were designed to seal against a set pressure differential i.e. something like 60 psi oil pressure in the bearings and 45psi in the turbine = 15psi more pressure in the bearings (the actual pressure it has to seal against).

Decat the exhaust and the turbine back pressure may drop to 15psi??? That would make the pressure it has to seal 45psi (60-15), which is three times greater.

 

BUT, is there anything else that is contributing to this problem????

 

Looking at the layout of the internals of the turbo its easy to see that the oil enters under pessure at the top and is fed to the bearings. The oil then passes throught the bearings and drains out throught the central void into the turbo housing. From the turbo housing it drains out of the drain hole and into the return pipe then to the sump.

 

Its commonly known that if the return pipe joins the sump below the oil level then the oil will not drain from the turbo properly and the seals will leak due to the excess pressure in the turbo housing.

 

This got me thinking. Oil not draining properly causes the seals to leak???

 

So another cause or influencing factor for the seals leaking could be due to oil hang up (not draining fast enough) in the turbo housing causing pressurisation of the housing?

 

What could cause this though???? Possible causes are:

 

* Oil level too high - Unlikely as a factory return pipe will exit into the sump well above the oil level, but its something to keep in mind.

* Return pipe too narrow or blocked - Again, factory sizes on factory parts should be ok unless there's s large amount of sludge buildup. The turbo can get the oil very hot and poor quality oils can build up sludge and carbon in the return pipe. Maybe worth a clean out?

*Oil viscosity - A heavier than stock oil will possibly drain slower, but it may also be less likely to drain past the seals. Swings & roundabouts?

* Oil feed volume too high - At stock levels this should be no problem, but with a high pressure or volume oil pump it may influence things, especially when combined with one or more of the above factors.

 

And finally....... Crankcase pressure, or should I say turbo housing pressure?

 

If the pressure in the turbo housing is higher than atmospheric pressure, this will have an influence on the oil path out of the housing. It may well aid oil drainage out of the return pipe, however if you consider the oil leaving the bearings and pressurising the turbine seal, a positive pressure from the turbo housing will increase the pressure differential on the seal as this can be considered almost a closed system.

The turbine housing can be pressurised by crankcase pressure and also by boost pressure leaking past the compressor carbon seal.

As the carbon seal on the compressor side wears it will allow boost air to leak into the turbo housing. On a stock turbo running 10 to 12psi this won't be much, but with hybrids running 20psi this will be quite a lot!

 

So perhaps by decatting the exhaust, having a partially blocked oil return pipe, fitting hybrids and upping the boost etc the turbos just can't cope?

 

A possible solution is to reduce the pressure inside the turbo housing by relieving the pressure at the return pipe connection with the turbo, and venting this pressure to the stock engine breather system (which should have a partial vacuum).

 

If a tee is added to the top of each oil return pipe and then a large diameter hose goes vertically to a large expansion chamber (to allow any oil to seperate out and drain back) then vertically up to a catch tank and then to the breather system, then the pressure in the turbo housing should be relieved, even if the return pipe is flowing oil badly/backing up with oil.

 

This way a partial vacuum may be possible in the turbo housing, reducing the load on the turbine seals and possibly preventing oil leakage?

[edge]

Good theory there mate!

maybe a diagram to show?

[Kranz]

Good theory there mate!

maybe a diagram to show?

 

Totally cr*p diagram done in MS Paint!!

[edge]

So basicly the return to sump pressure is reduced by catch can/air in-pcv system. They say the simple theorys are normally the best ones. Nice one

[SUPRALOOPY]

Interesting..I'll keep my eyes on this idea.

[JohnA]

I can see how such a setup might help turbo oil flow on an engine with increased crank pressure.

 

However this is not addressing the problem of smoking after a full decat - that is because of the decat upsetting the air pressure balance on each side of the oil seals on the exh side.

If it is similar pressure on both sides (stock setup) then there is low pressure difference, hence no incentive for the oil to seep through.

But typically after a decat there is higher pressure on the inside compared to the outside, so some of the oil is 'sucked out' (or pushed out, same thing depending on your frame of reference). The quantity of oil seeping through is small, but then again it only takes a few droplets to create a smokey exhaust.

[Kranz]

So basicly the return to sump pressure is reduced by catch can/air in-pcv system. They say the simple theorys are normally the best ones. Nice one

 

In a nutshell, yes.

 

Came across a Y piece in the Dodge Daytona VNT (the worlds first gasoline VNT production car in 1989) oil drain. Couldn't work out what it was for at first, but it looks like towards the end of production it was fitted and a pipe run to the airbox.

After asking a few questions of people in the know it turns out that it was a production modification to reduce oil leakage past the turbine seal.

Aparently any oil getting onto the VNT mechanism caused it to turn to carbon and stick in one position. If it stuck in the maximum boost position

[JohnA]

...Aparently any oil getting onto the VNT mechanism caused it to turn to carbon and stick in one position. ..

True, that is the main reason VNTs typically fail. (Corky Bell's babies)

We had two members in the Honda Blackbird members forum with VNT turbo kits, and both failed from this. Didn't live long.

Ease of fitting (no wastegate or oil feed/drain) and the promise of boost down low. In reality it was a series of failures and low-boost operation at the peak of the dream.

I was thinking of going that way initially, then decided not to bother...

[Kranz]

I can see how such a setup might help turbo oil flow on an engine with increased crank pressure.

 

However this is not addressing the problem of smoking after a full decat - that is because of the decat upsetting the air pressure balance on each side of the oil seals on the exh side.

If it is similar pressure on both sides (stock setup) then there is low pressure difference, hence no incentive for the oil to seep through.

But typically after a decat there is higher pressure on the inside compared to the outside, so some of the oil is 'sucked out' (or pushed out, same thing depending on your frame of reference). The quantity of oil seeping through is small, but then again it only takes a few droplets to create a smokey exhaust.

 

Now we're talking John

 

What if (yes, this is one of those what if scenarios) the oil pressure isn't directly causing the oil leaks, but the oil backup in the drain tube and turbine housing causing high pressure in the housing due to poor drainage?

 

The oil flow from the turbo bearings MAY then be forced through into the exhaust turbine due to the oil backup.

 

Do single turbos or aftermarket twin conversions suffer this problem? They have the same oil pressure feeding them, and probably lower turbine housing pressures. They have the same design of piston ring seal.... so why don't they leak?

 

Could it be due to a specific problem with the internal volume of the turbo housing, the amount of boost blowby from the compressor side, the volume of oil flowing through, the size of the oil return pipe that affects the stock turbos?

 

I'd give it a try, but mine don't smoke as I still have cats

 

Its just a theory..... from supporting evidence found by accident!

[Kranz]

True, that is the main reason VNTs typically fail. (Corky Bell's babies)

We had two members in the Honda Blackbird members forum with VNT turbo kits, and both failed from this. Didn't live long.

Ease of fitting (no wastegate or oil feed/drain) and the promise of boost down low. In reality it was a series of failures and low-boost operation at the peak of the dream.

I was thinking of going that way initially, then decided not to bother...

 

Those were the ill fated Aeromotive fully sealed ball bearing VATN turbos weren't they?

 

I'd rather trust a Garrett

[Stevie Boy]

What sort of pressure would you say oil leaves the turbos? Is there not a possiblity oil would just fill this catch can and then just get drawn into intake breather? Might need to take another look at your diagram!

 

I've noticed at work that on Twin/Bi-Turbo'd model 911's the turbo chargers fitted have small sumps connected directly as exit/drain path from the turbo, from which the return line/pipe leaves the small sump from the bottom, which is then I believe aided by a scavange pump run mechanically off the engine, can't remember if their run off ends of inlet cams or from intermediate shaft i'll have have a better look when i get one in again.

 

Didn't really think of possible benefits off having this oil scavanging from the turbo's on my Supe (which has both cat's removed! )just know thier an added chore when carrying out an oil change as each turbo sump is meant to be drained as well as the crankcase and dry-sump tank. Seems pointless when they only hold a ml or 2 but gets done anyway.

 

Eitherway i doubt a cam driven scavange pump would be feasible on the Supra, but perhaps a small inline collection tank from turbo with small belt driven scavange pump of some form to draw oil away from turbo and feed back to the sump? What sort of temps does oil exit the turbo charger i assume an electric pump of some kind would be out of the question?

 

Not exactly realistic sounding suggestions i know but nothing else comes to mind that would actually improve the dilemma.

[Chris Wilson]

The reasons these 911's have a scavenge for the turbo drains are:

 

A: They are dry sumped anyway

 

B: They are a flat engine, so there is no fall for gravity to return oil from the turbo drains to the sump anyway.

 

None worn turbos don't smoke on MKIV's anyway, jut make sure they aren't worn out, or on their way to being worn out

[JohnA]

What sort of pressure would you say oil leaves the turbos? ..

practically zero.

The oil feed is only pressurised, the return is purely gravity-dependent.

[JohnA]

...They have the same design of piston ring seal.... so why don't they leak?...

Hybrids etc often opt for a double-ring seal on the exhaust side, exactly for this reason.

 

OEM stuff is designed around the stock exh backpressure, and when it's gone then the pressure balances are disturbed.

Note that I am not talking about oil pressure here, rather air pressures on each side of the seals.

[Digsy]

I'm in two minds about this.

 

Some info I have from Honeywell (a.k.a. Garret AirResearch) says that as long as the crnakcase pressure is not positive and the oil drain is within +/-30degrees of vertical with a decent sized pipe, then oil drainage should be fine. I have had cause to send them some crankcase depression figures in the past and they seemed happy with a typical depression of 1 to 4 millibar at part throttle (the same engine produces about 10 times thi at WOT).

 

Provided the oil drain pipe is clear and the WOT breather is closed then surely pulling this kind of vacuum shouldn't be too difficult? The more power a tuned car puts out the higher the volume airflow past the WOT breather connection to the intake pipe will be, thus the higher the vacuum, which should help (provided blowby is under control).

 

Since in this setup the vacuum source for the depression at both ends of the drain pipe is ultimately the same, but the crankcase end has to pull on a much bigger volume and work through the whole breather system, I think it would create a pressure difference in the wrong direction up the oil drain tube, which is to be avoided.

 

The oil separator whould have to be bloody good too. I can see this setup pulling over a LOT of oil.

 

Expecting that much oil to drain back past against the airflow is a no-no. You would definately need a seperator with a dedicated return to sump pipe on it.

[Stevie Boy]

The reasons these 911's have a scavenge for the turbo drains are:

 

A: They are dry sumped anyway

 

B: They are a flat engine, so there is no fall for gravity to return oil from the turbo drains to the sump anyway.

 

None worn turbos don't smoke on MKIV's anyway, jut make sure they aren't worn out, or on their way to being worn out

 

Of course, now you mention it and having discussed this set-up ages ago with another Techie, we did come to the conclusion it was cos of the flat engine set-up not allowing downwards path for oil return and the fact there was no sump as such, except what oil sits at the bottom of the crankcase.

 

My Supra's turbo's don't smoke as yet (touches wood) but despite being lectured "Singles" is the way fowards I don't wanna go down that route, though I have recently been considering getting hybrids purely to increase my confidence when pushing the car hard that i may over stress the ceramic turbines too far even though i'm told 1.2bar should be fine you never know.

 

Having seen a tale or 2 of club member hybrids failing after relatively little mileage this is a little off putting, I wondered if this was down to this theory on oil seal failure due to additional stress on oil seals at higher than stock boost levels. Would oil scavenging of some form aid this theory?

 

Is it normal practice for improved oil seals to be put in hybrids? I was considering Turbo Technics, whats the conscensus on their hybrid conversions from J-Spec's to their Steel exhaust turbines and new compressor housings?

Any views on them Chris? They're the only turbo reconditioners that come to mind, i'm sure there's more out there.

[JohnA]

...I don't wanna go down that route, though I have recently been considering getting hybrids purely to increase my confidence when pushing the car hard

How is that? The stock turbos were made under strict quality control, while the 'hybrids' might well have been hacked in someone's shed. Or in a 'reputable' firm's bench using cheap (but out-of-true) machinery.

Given the cost of proper equipment for testing/modifying/re-engineering turbochargers, it is no surprise (to me at least!) that so many of these mongrel creations fail quickly when stressed.

 

Also don't forget that these 'hybrids' have already forced you to junk the stock sequential setup, so that advantage of the Supra's design has already gone out of the window.

What you're left with is TWO of everything (turbos, pipes, hoses, costs, heat emissions etc;) )

Boost threshold isn't any lower, either.

For an inline six of medium displacement a good single would make a lot of sense today.

 

that i may over stress the ceramic turbines too far even though i'm told 1.2bar should be fine you never know.

Come on, you do know.

It's not a question of 'if' but 'when'.

 

I wondered if this was down to this theory on oil seal failure due to additional stress on oil seals at higher than stock boost levels. Would oil scavenging of some form aid this theory? ..

not really, it is not oil seal failures that start the problems, more like the results of extreme centrifugal forces on the tips.

[dandan]

John,

 

I've had zero issues with my TurboTechnics turbochagers in 30,000miles+ of boosting between 1.0bar and 1.3bar.

 

The sequential operation is perfect, I have no oil seal issues, the car pulls very strong from 2000rpm to 5800rpm, EGT's (pre turbine) have NEVER gone above 840 degrees C and that's running 1.3bar.

 

I am not arguing the efficiency and performance benefits of a small single but that is just not an option for some people.

 

I don't know why people automatically rubbish the idea of ALL hybrids - my car pulls stronger and faster than either of our two vvti BPU's and has been doing so day in and day out for the past four years.

 

As with everything else, all hybrids are not the same.

[AndyT]

John,

 

I've had zero issues with my TurboTechnics turbochagers in 30,000miles+ of boosting between 1.0bar and 1.3bar.

 

The sequential operation is perfect, I have no oil seal issues, the car pulls very strong from 2000rpm to 5800rpm, EGT's (pre turbine) have NEVER gone above 840 degrees C and that's running 1.3bar.

 

I am not arguing the efficiency and performance benefits of a small single but that is just not an option for some people.

 

I don't know why people automatically rubbish the idea of ALL hybrids - my car pulls stronger and faster than either of our two vvti BPU's and has been doing so day in and day out for the past four years.

 

As with everything else, all hybrids are not the same.

 

Thank god for this. Who did you're set up DanDan and could you post a spec sheet.

[Guest coldflame]

i am haveing this problem as we speak i want to get it sorted as its kind of embarasing when you leave someone standing and then at the next set of lights they see nothing but smoke it gets on my nerves a little so any help on this would be grate i will even use the car as a guenipig to see what sorts it out please get in touch with help and ideas

[JohnA]

John,

 

I've had zero issues with my TurboTechnics turbochagers in 30,000miles+ of boosting between 1.0bar and 1.3bar.

 

The sequential operation is perfect, ....

Dandan,

 

so you have hybrids that exactly replace the stockers and keep the sequential operation.

 

Have you got any technical details of them? (A/R ratios, flow map, that sort of thing)

[Guest coldflame]

yeah thats correct but i have no detail about them

[JohnA]

That is usually the case;)

[Thorin]

I was under the impression that the vast majority of hybrid owners kept the sequential operation. Are you suggesting they don't?

[JohnA]

I was under the impression that the vast majority of hybrid owners kept the sequential operation. Are you suggesting they don't?

I'm not aware of any hybrid setup that successfully managed to offer larger exh A/R ratio (the main first-stage limiting factor) and keep the sequential operation.

 

The only one that I know of (it came close) was by a SF member.