Turbo Outlet Pipe Design

Hello Gents and Ladies,

This is officially my first post on this forum and even though I didn't plan on writing my first post with regards to something that I heard I just had to share it amongst the audience to see what your thoughts were.

I was at MTQ which is a predominant turbo charger repairer/ stockist here in Perth Western Australia when the manager Tristan walked over to me. Tristan wanted to share some info with me as I was there to pickup some Turbine housings that I had modified/ machined up.

He told me that he had an Engineer from Garrett come visit a little while back by the name of Graham Luke (not sure if this is the correct spelling of graham). Graham stated that he found that when coming of the back of the turbo outlet and introducing a 90 degree bend as soon as possible actually helps promote power generation. Why this is is due to the fact that as the exhaust gases are normally swirling out post turbine the bend eliminates this and causes the gases to exit out much faster as they now follow a straighter trajectory.

For example, a 4 meter long exhaust system which is generally quite straight at the post turbine location will take longer for the exhaust gases to flow out due to the fact that the gases are following a spiral path created by the turbine wheel i.e the gases actually travel something closer to 15 meters. Now if that same exhaust system has a 90 degree bend post turbine it will force the gases to take a straighter pathway. So in a way this 90 degree bend deflects the incoming spiral motion of the gases to a more straighter trajectory.


This to me sounds very backwards and I don't understand it at all. Is there someone here that can Quantify or validate such a dare I say "theory".

It's an interesting conjecture.
I wonder if there's a research paper about it somewhere.

In my limited experience with turbo Diesels, there's usually a 90 degree elbow adjacent the turbo exit anyway, just for installation reasons, e.g. to clear a bulkhead or some engine part.

As a practical matter, if you tried to use a straight pipe after the turbo, you'd need to add a bellows or two, to allow the engine to shake normally on its mounts without fracturing the tube or overstressing the turbo flange. Bellows add a substantial amount of backpressure. Turbos don't like backpressure.

( ... except V___o Marine turbos, which are specified to have a high _minimum_ backpressure, well above the maximum spec for any other turbo. I think the factory folks may be misreading the turbo factory curves, but that's another discussion. )


Mike Halloran
Pembroke Pines, FL, USA

There is also a school of thought which supposes that a sudden increase in diameter right after the exducer helps for the same reason, i.e. break up the swirling flow. I have not seen any data in support of this.
A large diameter exhaust pipe is beneficial in its own right anyway, for reducing velocity hence friction loss.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

MikeHalloran,

As we speak I'm actually, through contacts within Australia trying to see If I can track Graham down as this result from what I've been led to believe has been obtained using University resources.
I just hope I kinda get lucky and am able to find some sort of documentation as I personally feel that there isn't enough been done generally on the passive element side of things when putting together an efficient setup.

My theory is that If your going to bother doing something you might as well do it the best you possibly can. It doesn't take much more effort getting passive elements correct if the knowledge you have is quantifiable.

As for packaging and clearancing restraints, On a rotary engine for example there is far more flexibility in terms of being able to achieve a straighter smoother flowing system due to the tiny nature of the engine.

Hemi,

In regards to your remark relating to the general school of though. I came across a post by a Garrett engineer which specified that sticking in between 7 and 12 degrees post turbine will achieve desirable pressure drop conditions. Here is a excerpt from his post.

Jay Kavanaugh, a turbosystems engineer at Garret,
"As for the geometry of the exhaust at the turbine discharge, the most optimal configuration would be a gradual increase in diameter from the turbine's exducer to the desired exhaust diameter-- via a straight conical diffuser of 7-12° included angle (to minimize flow separation and skin friction losses) mounted right at the turbine discharge. Many turbochargers found in diesels have this diffuser section cast right into the turbine housing. A hyperbolic increase in diameter (like a trumpet snorkus) is theoretically ideal but I've never seen one in use (and doubt it would be measurably superior to a straight diffuser). The wastegate flow would be via a completely divorced (separated from the main turbine discharge flow) dumptube. Due the realities of packaging, cost, and emissions compliance this config is rarely possible on street cars. You will, however, see this type of layout on dedicated race vehicles.

Some older Cat marine engines included a separate cast iron tapered elbow, which is a nice piece, but it won't be seen on newer marine engines because it wasn't jacketed, and class society rules about hot surfaces have gotten more restrictive.

Most marine Diesels turbo exhaust flanges basically must have a step change to a larger elbow (unless you're tooled up to make a tapered elbow), because an elbow matching the bore size is too small. The straight tapered diffuser should work very well, but there are bending moment limits on the turbo exhaust interface, so you can't cantilever a lot of pipework off of it.



Mike Halloran
Pembroke Pines, FL, USA

I wish I'd thought to try the hyperbolic trumpet (and found a way to make it) discharging into a plenum for exhaust plumber's nightmares like the Cat C18.


Mike Halloran
Pembroke Pines, FL, USA

Mr. Kavanaugh's statement describes a classic low loss diffuser, which, inthge absence of any data to the contrary, I would agree is also the best design for a turbine outlet diffuser.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

Thanks for the replies guys, as yet I'm still in search of some facts or data pertaining to the original post about the 90* pipe section. I have flicked out a few emails but as yet have received no replies. I wonder if by chance just by using a diffuser would actually result in the out-flowing gases to loose some of their helical pattern and follow a more direct path..

Has anyone come across anything in regards to the original post 1. I am eager to get to the bottom of this as in the near future I will be fabricating my own dump pipe and system and wanted to squeeze out every ounce of efficiency.

Little 13B rotaries need all the help they can get in a way, inherently by having a peripheral type combustion cycle they do tend to produce a fair bit of gas speed.

That may be so, but whatever kinetic energy exists at the entry to the exhaust port is not available at the turbine outlet.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

It would be interesting to know more details behind "helping to promote power generation."

A Google search for Mr Kavanaugh leads to mostly references back to a single post that includes the tapered exit.

In the first 2 editions of Hugh MacInnes "Turbochargers" books (Red covered published in 1971, and Deep Purple 1984) he discusses turbo turbine exit conditions in the "Exhaust" chapters. In both editions The stated goal was to kill the swirling flow.

In 1971 he said "the best exhaust system leaving the turbine seems to be the one which expands rapidly to convert the swirl into turbulent flow." See attached image. That section also suggests that in the unfortunate case an elbow MUST be used near the turbine outlet it should be a long radius type to reduce pressure drop. That follows typical published data for pressure drop in air and water piping where each 90 degree elbow brings the same pressure drop as many feet of straight pipe.

In 1984 the tapered housing discharge image was re-used, but the text was modified a bit. When swirling "...the path for the exhaust gases will be considerably longer than if they were coming out axially. For this reason it is desirable to break up the swirl and change the gases to turbulent flow as soon as possible after they leave the turbine housing. One way to do this is to have a sharp diffuser angle on the turbine housing. Once turbulent flow is established it is not important for the pipe to be that big....". Then standard recommendations for a gradual taper to the working diameter are made to avoid creating extra backpressure.

Hemi & Tmoose thanks for the replies.

hemi, could you elaborate on what you stated a little please. I may have a bit of an idea on what you are saying but I would rather hear it from yourself rather than me guess away. I'm tending to think that if you are using a split pulse turbine housing you would have close to the same amount of kinetic energy traveling through the individual runners leading all the way up to it. At the end of the day a rotary engine has peripheral exhaust ports and usually only two of for that matter and as a result runners can be made pretty short i.e approx 6 inches so the gases exiting the exhaust port are "still in expansion?" due to continued flame propagation. Wouldn't that contribute to a slight increase in volume of gas charge?? I don't think your referring to the parasitic losses that the compressor imposes on the turbine are you. Another wonderful thing about rotaries and short runner manifolds are that they can be made dead even and centered between the two exhaust ports making it easier to create something tuned length without having to go to a great deal of trouble.

Tmoose,

With regards to the exert from Mr Kavanaugh's page you kindly shared, If I was to go out of the 7 - 12 degree taper as stated by Graham Luke (Garrett Engineer) I would be increasing my back pressure so it would be contradictory to what Mr Kavanaugh states?. So at the end of the day how well does a certain angle work to act as a diffuser and not induce high pressure zones? Does it actually diffuse anything? probably not. I'm feeling kinda confused.

I wonder if by placing a passive style diffuser section within the inside of the dump pipe starting section that went against the axial pattern created by the turbine i.e a axial pattern traveling opposite to the turbine rotation by stitching in some curved sheet. or even following straight geometry down the length of pipe for a particular distance wouldn't that in theory act as a diffuser without reducing flow?

Since this topic is about optimizing the pipe configuration downstream of the turbine, I was just pointing out that the kinetic energy upstream of the turbine is pretty much irrelevant to this discussion.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

Hi Rot8,

The Kavanaugh and Luke quotes were supplied by you. I only knew to look up Jay Kavanaugh because you mentioned him. From his (original post?) quote at the head of the trail he did not seem concerned with swirl at all. Just basic not-too-steep tapered diffuser theory for best pressure recovery. I have not looked at many turbos, and only one for a diesel (early 80s vintage RotoMaster for an olds 350) but I'd say the diffuser in the exhaust housing is not likely to be much more than 2 inches long, which with the "maximum" 14 degree included angle would provide no more than 1/4 inch larger diameter at the exit.

I only quoted stuff from books by Hugh MacInnes that I have on my shelf. It was all pre 1984. I figure if he put it in a book, he meant it, and believed it to be true at least on the day the book was published.
Hugh specifically stated twice he was concerned with killing swirl, just like Graham Luke allegedly was according to your contact at MTQ.

So, now we seem to have 2 votes for the importance of killing turbo exit swirl (MacInnes and Luke) and one vote for focusing on classically shallow discharge diffuser angles to try to prevent losses due to flow separation. And, if we can believe what is published in the HP Turbocharger books, at least one (MacInnes) worked developing turbochargers for multiple companies (TRW, Rajay, and Rotomaster), admittedly in the early years, at up to the chief engineering level, and has his name on a small string of patents related to turbocharger plain bearings starting in 1962 at TRW. Experience like that does not make someone all-knowing, but I would not ignore what they have to say, either.

I would not speculate very far based on the VERY limited info available, but I believe it is >>possible<< that MacInnes and Luke found the losses due to unchecked swirl were large, even larger than the simple pressure recovery from a too sharp transition to an exhaust pipe of sufficiently large nominal size. And, if Kavanaugh's total relevant work experience was in the diffuser development lab, and that showed there were improvements when diffusers were shallow, but another group were the ones to test aggressive or experimental swirl killing features, he might only be able to report on the goodness of shallow angles.