This is extraordinarily long so go get a snack or use the restroom before you try to read the whole thing.

This topic had so many questions it deserved its own thread to explain things better.

These sleeves were machined for me by Marcus Williams (turbostreetfighter) about 3 years ago. I told him what I wanted and he made them. Unfortunately he can no longer produce them. They are the result of many phone conversations with Paul Yaw. Paul originally was going to make me a set but he is very busy. He can do inserts (more on them later) but I thought it would be cool to have something original so I went to Marcus.

The factory exhaust sleeves suck pure and simple. There is a ton of turbulence through them. This is due to the lip between the port opening and the sleeve and the fact that the factory sleeve expands in area several times over a very short period. The POS aftermarket headers (all of them!) and lousy cast iron manifolds (turbo or not) we bolt to them don't help either. The goal of porting is to increase power by increasing airflow. Porting not only makes the exhaust ports larger but also increases the exhaust duration. Much like adding a longer duration cam with more lift on a piston engine. This is not always good though. The problem with the peripheral location (among other things) is that our timing is very long. The stock exhaust timing is really too long for maximum power but our ports are relatively small. The main downside to this location. By porting the exhaust ports we increase the timing even further as we enlarge our ports. The port is also enlarged to match the size of the sleeve opening. Great no more turbulence right? Wrong!!! We have improved airflow but it is still poor. We need to concentrate on how to get the most airflow we can from the smallest port we can. Velocity is important too. This is something that decreases on ported exhausts even though flow increases. The machined sleeves completely replace the factory sleeves. The port opening is based solely upon the shape exiting the rotor housing. I other words the shape you see from inside the engine is the same as that on the outside. Area and shape are maintained all the way through. This ensures low turbulence, greater flow, and higher velocity. Flow is increased around 30% while port timing stays the same and velocity increases! This type of airflow increase is typical of a ported stock sleeved motor but has the added benefit of greater velocity. Velocity is what spins turbos faster! You can use a much larger turbine wheel and a/r and still have less boost lag. A better flowing exhaust housing has less backpressure and therefore takes less boost pressure to get the same power levels. Less boost pressure = less heat. You get the picture. I will go into other details farther down the thread under the corresponding pictures.

Now for the downsides. Trying to get the factory sleeves out is a real pain. I just use a die grinder with a carbide bit and cut through the sleeve itself until I can see the factory pin. The I hammer it out. Takes about 10 minutes each. Took me 3 days to figure out how to do it though! The new sleeves have to be installed in the same manner. I threaded the rotor housing pin holes and used a slightly larger hole in the sleeve. Just screw in a set screw and voila you have an easily removable sleeve if you ever need to get it out. Now back on track. The sleeves are not machined out to the exact port size for your engine. You have to install them and port the sleeves out yourself. Marcus used a34 steel. Very hard steel. Harder than the factory sleeve. If you don't have a die grinder and a carbide bit you can not cut them. Don't even think about using a drill saw or a dremel. Save the dremel for polishing. Next problem. Not a single existing header or turbo manifold can work with them. The sizes and shapes aren't even close to matching. The new port openings are actually slighty lower on the bottom cornerrs than the factory ports were. If you bolt on an aftermarket system the flange will overlap part of the ports creating more turbulence than you originally had. Fabricating a custom exhaust is easy though and everyone can find someone competent enough to do it. Here's the big problem. The added velocity also has much more heat. If you are using the recommended 1 3/4" runners to the turbo you will melt your turbine wheel. Thats right. If it is not made of Inconel the blades will get so hot that they will actually fold over on themsleves. Inconel is the most durable and tolerable and would excel in this area however it is a $500 upgrade from Turbonetics. How do we get around that? Well the best way is to go up one size in runner diameter to 2". Wait a minute the stock size is 2" how are we getting any benefit? Remember to crunch down the runner to the shape of the port. Later pictures will show this. There will be a slight lip between the sleeve and the runner but not too much. Most of our velocity and lack of turbulence is actually still there. This slows the air down slightly through the runner and also decreases heat to the point where a standard turbo wheel can survive. Unfortunately turbo response decreases slightly. No matter though it is still infinitely better than the stock setup. I need a break from typing and I'm sure your eyes are about to fall out or your brain explode or something so I'll continue with more info under some of the pictures.

This picture really sucks. Trust me the port shape is even and symmetrical.

 

O.K. I said earlier that all stock manifolds, aftermarket headers, turbo manifolds, etc. suck. Let me defend that statement. We all know that the factory pieces are best used as paperweights and doorstops. The aftermarket pieces are better and give us better performance but are still far off from perfect. Here's why. Even with the sleeves you still need the airflow after the port to be smooth. Testing (by Paul Yaw) has shown that the exhaust port runners need to extend a minimum of 3" straight out of the engine before making any turns. Look at a picture of his headers and you'll see this. If they are any shorter than this there will still be a disruption of airflow within the port even though our sleeves are smooth. Every aftermarket system turns immediately after the flange. Look at race car exhaust runners at the engine (any type of major race car series) they all go straight out for a couple of inches before turning. The next problem with the n/a headers is length. They are all too long. The best tuned length is about 20-24 inches for a short primary and around 100 inches for a long primary. Racing Beat's headers are too long and turn immediately out of the engine but the collector design is good and they are even length. I have their long primary system on my GSL-SE and lets just say it is going to come off very soon. Aftermarket log style cast iron turbo manifolds should never be used. Cast crap. Lousy flow, uneven length, but they are tough and fairly cheap. Just have one made and make sure it doesn't turn for the first 3".

I've had many people ask how to taper the shape for the first 3" without serious turbulence. This is easy and the following pictures will show this. The big turbulence problem is due to the very rapid area increase of the stock sleeves. The taper on my runners is gradual with no area change (minimal for all you die hard math junkies!). The following pictures are of the beginning runner pieces. They are exactly 3" long. These are what I weld to the exhaust flange. Anything is is fair game after this point.

 

Here is a second view. Obviously these are not yet welded onto the flange.

 

A third view of the taper on the runners. They are exactly 3" long.

 

O.K. one of the greatest questions I've gotten is if there is a way to improve port flow without installing new sleeves or taking the engine out. The answer is a very pleasing yes. Paul Yaw made a set of inserts for his 12A road race headers. They slide into the existing sleeve and reduce the expansion rate at the top of the port. His gain on the 12A engine was 28% more airflow. He made the exhaust header pipes the same shape at the opening to blend with the insert. You still can't use any currently available prefabbed pieces with them. I am going to find another machine shop that can make the sleeves so they will be available shortly. Pricing is unknown at this time but I will find a way to make them affordable. I will also make sure they include the proper exhaust flange.

One other downside. These will not work with 2nd gen. n/a exhaust ports due to the naziesque diffusers in the factory sleeves. The machined sleeves only fit '86 and up turbo housings as well as 3rd gen. cosmo etc. They will not fit pre '86 13B housings due to sleeve size differences before this point. An entirely different size would have to be machined and demand is too small for them. Inserts would be much easier to fabricate for this application.

I'm sure I haven't answered every question but I'm sick of typing and have to wake up in a few minutes!

 

 

Rotarygod:

thanks heaps for posting that info up about the sleeves one question, what sort of material would your recomend for the little wing thing that fits into the port like what paul yaw made for his headers, would say a peice of 3 or 4 mm steel be able to handle the temps etc. Also what sort of exhaust temps would you expect to see.

cheers

thanks again

Lance

 

I honestly don't know but it had better be able to survive to almost 2000 degrees without losing its shape!Can't be stainless steel since stainless has a high coefficient of expansion and it would physically change shape with temperature differences. I would think regular steel would disintegrate impressively but I don't know. Remember Marcus was melting exhaust turbines after less than 1 day of use! I'll do some research and find out what works best. If I could get a hold of Paul I would ask him what they are. Who knows the answer may surprise me. At a guess I would say use Inconel. That would work. Good luck on getting it though. PITA to weld too.

 

Titanium??

 

I think this is a good idea if you want to keep the factory timing of the exhaust port, for me this is far far too restrictive, it does not allow the engine enough overlap.

I take the sleeves out all together and open the port to the maximum possible, This gets rid of the very big expansion you are talking about, it also wides the power band of the engine, alot.

My turbo street port, with maximum size exhaust ports have a flat peak torque spread of 1500 rpm (within 97% of maximum torque) you do not get this with the very restrictive timing of the factory exhaust port, it is very typical in a restricted timing (exhaust rotary) to have a distict torque peak especially when you look at engines making over 500bhp in 13B form that are of non bridge or larger variety.

I would never run the restrictive mazda exhaust timing, after using an earlier exhaust opening and later exhaust closing specification.

I would like to see the pressure differential between exhaust and Intake with those small inserts, it would be very high in the exhaust, at least 1.5 to 2 times greater than the intake. The best performance on turbo charged engines is achieved when this pressure is near equal.

If you do not believe me then go look in the F1 world from the early 80's to late 80's and simply look at the size of exhaust manifolds and exhaust port runners on 500 to 900bhp engines, I am not intending this as a flame on anybody, but as a mechanical engineer I just would not ever consider running a set up like that with the heat load assosiated with a turbo, let alone a rotary turbo ! The timing restriction is also another point that limits the ability of the engine to make power via increased revs, instead of increased boost.

Just a different view for you guys

 

rotarygod: i think there are various grades of steel so i imagine the heat range of them would differ, i have to find out what ingrediants are in those that have a higher melting point

Riceracing: So you are saying you open up your port so it is the same size as that of the port after the sleeve is removed??? not sure if that makes sence how i worded it

Lance

 

rice racing:

was just curious weather you have done any sort of flow bench work with your porting?

cheers

Lance

 

Lance, yes I have.

I have made my own systems up using simple manometers and such, and quickly realized that testing some arbitrary air flow figures against certain ranges of pressure drop is only one side of the story.

My method of exhaust porting flows an astronomical amount compared to miniscule reduce factory ports and runner combinations.

Bit like when you test a 2" straight thru muffler V's a 4" Muffler with Venturi Inlets and outlets, you can get crazy flow ratings with very very little pressure drop. The main considerations here that I look for are exhaust timing and it's effect on engine speed, Volumetric efficiency, BSFC, and durability of the engine as a whole. The system of running such a small exhaust runner will limit engine speed, power, and peak torque "spread" mainly due to the limited Exhaust timing. I think they have touched on that it is extremly hard on turbine wheels ? I can see that just by looking at it.

I would like to see some dyno results of these smaller exhaust ports/runners out if interest. I personally think it would be very unreliable in any type racing rotary besides maybe a drag engine which is run for very short periods of time ?

 

Pretty much ! My exhaust port is the largest you can put in the rotor housing, before water leaks in or the apex seals drop out ! The area left once the sleevs are removed is blended smooth, they are very large, mainly to get the exhaust timing I wanted. My exhaust manifold uses an inverted funnel at its inlet, the theory behind my logic is to introduce a small amount of turbulence into the area of the runner, at which point laminar flow is established In the inverted funnel section through the rest of the header into the turbin section, I use this same principle on the down pipe, converting the swirling "turbulent flow" into a laminar flow by introducing turbulent via a large volume chamber right after the turbine outlet, then funneling down into my 3" system.

I get 566bhp (472rwhp on a dynapack dyno) using 19.9psi with a quite (for a rotary) road exhaust system @ 7500 to 7700rpm on pump gas, T2 9.0:1 engine, TO4 etc etc, I use the std (ported) intake manifolds and Throttle body. It works for me quite nicley I am not saying it is the only way to skin a cat, but The flat torque peak in the higher rpm ranges near peak power is something you cannot replicate with a side port (non BP) with near stock exhaust porting, You may make power at lower revs but it generaly has a distinct peak to it, which limits the usefull range of the engine.

My design criteria for my engine was maximum torque from 5500 to 8000rpm, almost 90% over that speed range, from 6000rpm to 7500rpm average torque is within 98% of the maximum. This is as a result of the increased Exhaust timing. Also the engine operating range is also increased which is good for a road racer such as my self.

Having said all this I am realy open to new stuff, despite what I have written, so please do not take it the wrong way, I would love to see some dyno runs of this stlye of porting/runner/manifold combination. There are some parts to the reduced runner that make sence to me, but the heat would be an uncurable problem from what I know of, "hint" that is why I run water Injection evey on my "large" exhaust ported engine.

regards

Pete

 

Thanks for the info rotarygod. You and Paul Yaw ROCK! WHen will his headers be availible? If he would ever answer his phone, he'd make alot more $$$$$ LOL.
If I cant get his inserts/headers, I think I'll hack up a pacesetter header to make my own setup. The RB turns WAY too early!

 

These guys make some nice looking headers. They cater to the IT and EP crowd if I'm not mistaken. Still quality looking headers.

www.sdjmotorsports.com

 

Ok, I think I understand this so correct me if I'm wrong. You keep the port shape to the outer end of the rotor housing. At that point then you taper it open and at a distance of 3" from the housing/header interface it will be the same size as the header tubing? And is that 3" straight a minimum? If you can go longer is there further benefit? And I would think that the largest radius turn you can make will benefit the gas velocity (I'm speaking of the 90 turn towards the back of the car).

 

Carl: you do understand it correctly. Everything you typed made perfect sense.

Rice: That's pretty interesting how you do your exhaust ports. I have a housing that I did that to just for the fun of it. I always wondered how well it would work. I don't want increased port overlap though on my engine since I need to pass emissions testing. Horsepower goals for me were anything over 425 at the wheels with no more than 15 psi of boost. Easily accomplished.

The Renesis takes care of many of the problems of a peripheral exhaust. It has less timing, but much more area which is what we are all trying to get.

 

WOW!! Rice that is exactly how i have my race car setup!! my exhaust ports are HUGE and i had the insides ceramic coated along with the headers too!!my reasons for this setup is because i run an auto and i need a wide torque curve for the trans brake and launch. what kind of EGT's do you see? i am running methanol and my probes are exactly 7" from the ports. for some reason my EGT's stay at 900c no matter what .
at any rate i second what Rice says about the powerband, on my race car i run a step header to keep the 76 spooled!! Rice just as you said at a certain level the inserts do become restrictive but i would say at anything under 400FWHP it works great.

MWW

 

MMW & Rice Racing:
what about if you port the exhaust, but not quite to that extent yourself and rice do, would it be an advantage to run a sleeve to the same size as the ported exhasut port. The reason i ask is i have a comso engine that has had the exhaust ports opened up and the sleeves removed though there is a still a step where the sleeve was. would it be an advantage to make up a sleeve to suite the port i have? my manifold runners are 2 1/4 though they bend straight after the port, my exhaust housing is a t-04 style 1.32 with a p trim, thoughts please

cheers

Lance

 

Interesting stuff to think about here!

 

I am thinking along the same lines as lance at the moment, I am about to put together my semi BP motor, and having been taught by the 'RICE RACING' school of exhuast porting techniques my exhuast ports aren't small, but they're not to full RICE proportion. actually they can be seen here.



I have altered my stock sleeve abit to suit the port shape/size, and what i am thinking about doing is getting a piece of material similar to the stock exhuast sleeve welded into the top part of the stock sleeve, this will mean the sleeve only gives a small amount of expansion. from there using 1 3/4inch piping for the 13inch exhuast manifold runners............ the spool up should be stupid with turbo spec bridged secondary ports, but will my divided 1.15a/r P trim exhuast side handle it or not? for what its worth, the exhuast port exit has a sectional area only fractionally smaller than that of the 1 3/4 inch pipe.

idea's Pete?

Cheers
Shane

 

this will definately work, the sleeves that i originally made are made so that you can port out as much as you want. but i will tell you that using 1.75 tubing will burn your turbine wheel in a matter of minutes, so i would suggest tapering the port out to at least 2" or you could make the first bends and some straight out of 2" and then taper to 1.75 at the collector to gain some velocity. my race car header goes from 2.5 to 2.25 to 2.0 at the collector. anything is better than the stock liner that are on the engines especially the RE engine!!



here is some more food for thought: what if the linings were made of aluminum instead of steel? i have been running no liners on my race car for about 6 months now and have had no complications, granted i probably dont have 5 miles onthe engine but no complications have occured yet. the aluminum is way more machine friendly and i could even CNC taper port the liners alot quicker and nicer out of aluminum if they would last. what do y'all think? my ports are also ceramic coated which may be why they have not disentgrated yet but if i could CNC port them to a specific size the liners could be coated afterwards.


MWW

 

I would think anything which smoothens out the turbulence would be beneficial. It doesn't matter how big your ports are just get rid of that turbulence. Everyones goals are slightly different.

Interesting idea on the aluminum liners. I would be afraid of them melting but then again thats what the housings are made of and many people have removed their sleeves. If graphite would hold up I know where to get that in rods of various thickness. Small rocket engine nozzles are made from graphite. Just a thought.

Marcus didn't you just extend the header tubing into the sleeve through the flange before?

 

you can do that with the 1.75 tubing but i was also melting turbines at that time, i dont know how far the 2" stuff will go in to the flange. the 1.75 will go about .900 into the sleeve.



MWW