How to design the best small 16g turbine inlet for max flow and WG?

[Join]

Hi!

I am currently rebuilding a s16G TD05H turbo. In the process I am doing a portjob on it. But for the turbine exhaust inlet, a guy at work that do 3D programming and modeling is going to help me machine that part.

He also have a Flow program that he uses, but I want to get some more input from you guys with the 3D skillz and flow knowledge how to get the best design.

I want the internal wastegate hole to be efficient enough to prevent boostcreep, but the main flow to get the quickest spool and highest performance is to be as little interrupted by vortex(es?) as possible.

Also it would be nice with some input on exhaust pressure, temps and airflow. Let's say 8000rpm's to have a solid startingpoint.


Just want the better design!


Anyone?

I'll post some flowreports in a while, need to scan them!


John

 

[pboglio]

Sounds cool. I'd spend more time in the exhaust manifold itself, specifically the collector area where the divider sits and how the runners join each other. This has far more effect on spool than just enlarging the turbine inlet. People think porting the turbine housing gains power/spool, but its an after affect of enlarging/reshaping the manifold collector.

You can't do the CFD with steady state flow, cause thats not how a motor works. All the spool gains are from reducing interference between runners as they try to dump into the turbine housing, not from just reducing pressure losses.

 

[90laserRSfwd]

I'd say concentrate on doing some short side radius work around the waste gate port. And maybe clean up the casting inside the turbine housing. Other then that, like mentioned above, the exhaust manifold is the problem WAY more so then the exhaust housing.

 

[Join]

Thanks!

My exhaust manifold is the oversized SBR manifold with a 65mm ported outletfrom SBR.
The O2 housing is the EVO housing/w. 60mm hole from turbo. Turbo have 50mm outlet rounded out to 60mm to fit.

Is there other things? shape of the hole in the intake port?

 

[DSMunknown]

Sounds cool. I'd spend more time in the exhaust manifold itself, specifically the collector area where the divider sits and how the runners join each other. This has far more effect on spool than just enlarging the turbine inlet. People think porting the turbine housing gains power/spool, but its an after affect of enlarging/reshaping the manifold collector.

You can't do the CFD with steady state flow, cause thats not how a motor works. All the spool gains are from reducing interference between runners as they try to dump into the turbine housing, not from just reducing pressure losses.

I would like to see you elaborate on this.

 

[Join]

+1

 

[Join]

Sounds cool. I'd spend more time in the exhaust manifold itself, specifically the collector area where the divider sits and how the runners join each other. This has far more effect on spool than just enlarging the turbine inlet. People think porting the turbine housing gains power/spool, but its an after affect of enlarging/reshaping the manifold collector.

What I want to do, is not just enlarging the intake port, but make the "perfect shape" to gain the highest possible flowrate from that part alone. Within the limits of that housing.

If you generalize it a bit, if you have a nozzle that spray on a turbine wheel with X mm2 size you get a certain spool. If you re-shape the nozzle, but keep the X mm2 size, you can change flow speed or width or a little where it hit the turbine-wheel and get a little different result. What is perfect? That's why I ask of help from people with experience and the proper software to do calculations on flow from 3D drawings.

Our program can only calculate one intake and one exit at the time, so simulating a open wastegate high-flow situation is not that easy.

You can't do the CFD with steady state flow, cause thats not how a motor works. All the spool gains are from reducing interference between runners as they try to dump into the turbine housing, not from just reducing pressure losses.

Why? I have a known amount of volume that I need to get through at a known time. There will be pulses, but the volume is consistent?

A vortex can steal a lot of area and steal space from the flow area, and that is what I don't want to happen.

Am I way off here?

I got a couple flow reports from a very basic drawing. One with straight hole into the housing from wastegate, and one with 10mm fillet around, too see the difference then. But I cannot scan them today, forgot them at work..

 

[pboglio]

I would like to see you elaborate on this.

Not sure what your asking for? The main flow losses in the manifold are from the exhaust streams intefering with each other at the collector itself. The evolution from log manifold, to 4 runner manifold with a flow divider (1g/2g), to the fully divided collector into the turbine housing (later EVOs) kind of illustrates this. What you can do to a stock casting is limited. Possibly retaining more heat in the manifold as well. Thats about it on stock parts.

The CFD analysis would have to be run using a time dependant function as pressure in the runners is going to vary throughout the exhaust cycle, simulating flow reversals, along with simulating the interactions between cylinders, wall heat transfer, etc. It is a type of anlaysis you'd see in a master's or doctoral thesis. And I'd point out it would have to be done by somebody who already has a DEEP understanding of 4 stroke engines.

I use FEA extensively as a design engineer and I can tell you its easy to **ck up even basic structural anlaysis just by making the wrong assumptions. I've used CFD on a limited basis at work, mostly for pressure loss analysis in steady flows. I've used CFD extensively for simulating the exit side of the turbine and the interaction with wastegate flow, and saw some interesting things that illustrated gains in terms of pressure loss reduction. But we are talking radically simplified assumptions. Mostly for determining trends.

 

[the_mork]

Some of this stuff involving the collector and exhaust interference is discussed in the turbo education thread, basically what you are trying to do is reduce interference where two or more cylinders are trying to dump exhaust gas at the same time. When one set of exhaust valves opens and shoots exhaust into the manifold and towards the turbo you end up with a pressurized system which will be used to spin the turbine and is generally a very good thing but as soon as the next cylinder opens its exhaust valves then you have two pressure sources with one outlet. The pressure in the cylinder will be greater than in the manifold so the exhaust will come out but now we are increasing the pressure in the manifold to the point where it starts to impede exhaust flow out of the second and subsequent cylinders. The key of the game here isn't to try to increase flow (though this is generally good) but instead to focus on velocity. Your other option besides increasing velocity is to go to a continually more separated setup such as a true divided collector setup that many modern turbo applications use. The gain in the dual collectors is that each cylinder is competing with only one cylinder that isn't fired consecutively such as 1&3 and 2&4.

According to the turbo education thread we can make a manifold "boost gauge" which will help us figure out what sort of pressures we are seeing but IMO its use is relatively limited. This all came up in a discussion about volute efficiency as functions of size and shape but it applies to a limited degree to collector shape also.

As an aside collectors are very important on well designed n/a cars and less so in turbo cars because of a phenomenon termed as scavenging in which there is actually a low pressure "pulse" which when tuned via runner length will actually create low pressure spots just outside of the exhaust valves to literally suck the pressurized air out of the cylinder.

At least this is how I understand all of this to work. I'm sure some of the people in this thread know these things better than I do and might even find reason to correct me but for those who are following this thread I thought it was worth discussing.

http://www.dsmtuners.com/forums/turbo-system-tech/199534-turbo-education-thread.html

 

[Join]

Thanks for the thread, I would never found it by myself..

So basically I would try to create a teardrop shape in the inlet for best efficiency? Like this?

You must be logged in to view this image or video.

Here's the E3 16G inlet:

You must be logged in to view this image or video.

That one look more like my inlet.

Something in this shape, within the housing limits?

You must be logged in to view this image or video.

I will alter my turbo's A/R by enlarging the intake size? Increasing A/R? And mabye slightly lower spool up? But hopefully some higher top-end due to higher flow?

This is my air setup: K'N airfilter, stock 2G MAF, gasketmatched S16G compressor housing, 2" hard piping to stock SMIC, 2.5" hard piping to 3" Mustang TB, MAGNUS SMIM, SBS Stg.II Race head w/FP2s, SBR oversized cast manifold-ported to 65mm outlet, and then into the machined intake on the hotside of the turbo-outlet thrumpet rounded from 5cm to 6cm outlet, 6cm O2 housing w/2.5" gasketmatched outlet, 2,5" to 3" RRE all stainless DP, RRE High-flow 3" cat, Tanabe 3" cat-back system.

I guess a slightly higher A/R will be only good, from what I can tell from that thread.

From what you say here, the constant flow can only point to a trend in flow differences? So I will have to try to make the best "trend" for both main flow and wastegate flow to get an idea how it will work out, but I will not be sure how it works out in the end. And it might have a limited effect.

Interesting! Please correct me if I got something wrong.

And, how much material should be kept to prevent heatspots in the material and risk of cracking?


John

 

[pboglio]

Yes, constant flow will still illustrate flow loss differences between designs. A round cross section is ideal off course. The difference between a round cross section and oval or tear dropped has got to be small in terms of flow losses, given the same cross sectional area. What is going to have massive flow loss differences is changing the cross sectional area as this will push your choke flow up.

Here's a real world example of what I'm talking about. I had a 13g turbo setup on my 91 GS. I switched to a 16g with 7cm2 housing. I tried ALOT of combinations and for kicks I slapped on the 13g unported exhaust manifold with its TINY flow collector. Now that flow collector cross sectional area was still bigger than the turbine volute diameter of 7cm2. But what it did flow wise was shocking. I could not accelerate the car past 80 mph what so ever. The flow losses in the collector and resulting miss match with the 7cm2 housing where MASSIVE and creating a choked flow situation at extremely low mass flow rates.

So your willing to sacrifice spool for topend? Then just buy the HAFE bolt on 8cm2 turbine housing for TD05H chra and call it a day:

Slowboy Racing :: Turbos :: Subaru Turbos :: SBR :: Hafe Performance Cast 8cm Turbine housing

If I was going for every h.p. I could squeeze out of the EVOIII 16g, then something like an 8cm2 housing would be worth it. The TD05H turbine housing drops off with a v-band clamp and 4 bolts. I'd swap back and forth for street or strip duty.

I think that turbine housing is for the Subaru. Still, I remember Forced performance selling 8cm2 turbine housings for DSM's.

 

[Join]

Yes, constant flow will still illustrate flow loss differences between designs. A round cross section is ideal off course. The difference between a round cross section and oval or tear dropped has got to be small in terms of flow losses, given the same cross sectional area. What is going to have massive flow loss differences is changing the cross sectional area as this will push your choke flow up.

Yeay, I guess, but I also think a need to get the flow moving efficient into the turbine to get the axial movement, so I'll just play with some flow models with the help from the guy at work..

So your willing to sacrifice spool for topend?

Due to the very quick Small 16G spool, it can, if I must, be a little reduced. But I guess because the higher airflow from new cam's, O2 housing, 3"exhaust and so on, will make up for a higher A/R? I love the quick spool, don't get me wrong!

If I was going for every h.p. I could squeeze out of the EVOIII 16g, then something like an 8cm2 housing would be worth it. The TD05H turbine housing drops off with a v-band clamp and 4 bolts. I'd swap back and forth for street or strip duty.

I think that is no option, I just want to play around with this housing..

Thanks for your replies!

 

[Join]

Some of this stuff involving the collector and exhaust interference is discussed in the turbo education thread, basically what you are trying to do is reduce interference where two or more cylinders are trying to dump exhaust gas at the same time. When one set of exhaust valves opens and shoots exhaust into the manifold and towards the turbo you end up with a pressurized system which will be used to spin the turbine and is generally a very good thing but as soon as the next cylinder opens its exhaust valves then you have two pressure sources with one outlet. The pressure in the cylinder will be greater than in the manifold so the exhaust will come out but now we are increasing the pressure in the manifold to the point where it starts to impede exhaust flow out of the second and subsequent cylinders. The key of the game here isn't to try to increase flow (though this is generally good) but instead to focus on velocity. Your other option besides increasing velocity is to go to a continually more separated setup such as a true divided collector setup that many modern turbo applications use. The gain in the dual collectors is that each cylinder is competing with only one cylinder that isn't fired consecutively such as 1&3 and 2&4...

...As an aside collectors are very important on well designed n/a cars and less so in turbo cars because of a phenomenon termed as scavenging in which there is actually a low pressure "pulse" which when tuned via runner length will actually create low pressure spots just outside of the exhaust valves to literally suck the pressurized air out of the cylinder.

At least this is how I understand all of this to work. I'm sure some of the people in this thread know these things better than I do and might even find reason to correct me but for those who are following this thread I thought it was worth discussing.

http://www.dsmtuners.com/forums/turbo-system-tech/199534-turbo-education-thread.html

So, when the exhaust mainifold was mentioned already, I'll post mine!

Here on top is my Slowboy Racing oversized cast manifold. Ported outlet to 65mm and some work down where the collectors come together.

You must be logged in to view this image or video.

Same type have been used on a 600+hp app by SBR.

It's all good, but I have seen pictures on several ones that have cracked, so I purchased a EVO exhaust manifold to have in stock until the SBR one crack.


But I think the EVO OEM need some work done. I will have to make it 65mm at the outlet like the SBR one to fit my exhaust housing on the turbo.
Mabye keep the divider to prevent interference between cylinders, but try to open it up to make room for more flow.


I hope a "expert" will chime in on this!

Here are a couple of more pictures of different angles.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

John

 

[Join]

Hi again guys!

Now before I start working on the exhaust housing and spare exhaustanifold, how much material should be kept in wall thickness to prevent the parts from cracking?


John

 

[dsm-onster]

I don't like that sbr manifold. It could have done 900whp, but a better division would have netted better results. There is no real divider between the 2-3 pairs and the 1-4 pairs. A better division of pulses helps a turbo that is being maxed out. When you push the limits of a turbo, your backpressure skyrockets. Having more separation between the pulses keeps power from falling off due to exhaust gas reversion. More energy is used for the turbine wheel to spin the compressor and a higher VE results because of less exhaust gas back flow. In other words, the division helps get more flow per rpm, especially up top; and give more energy to the turbine so that it can spin a compressor that needs so much more energy to convert to boost instead of just making heat, due to the fact that you're pushing its map and are out of it's efficiency range. Simply put, if you want the most velocity and the smoothest transition, simply by the seal ring for the evo3 manifold. It will fit right in the 7cm^2 small 16g turbine housing. A big fat collector will do similar to what pboglio described (suddent stem from a 5cm^2 outlet to a 7cm^2 inlet). It will slow gases down and creat more turbulance. There is a big difference between the runner area and the collector area. This is not good. Having that difference as small as possible will be best. An FP race manifold does a better job than a stock /evo manifold because it has longer runners. Any equal length tube runner manifold is better than a cast manifold with short runners and not even a good division between 1-4 & 2-3.

Exhaust gases reach the turbine wheel in pulses. One pulse leaves the turbine before the next can reach it. This is one reason why the evo 8 turbo and all twin scroll turbos do so much better for spool yet have enhanced flow characteristics. They keep the collector area closer to the port area in size. The individual runners see pulses. You don't see guys going to 7cm^2 area runners do you?

Are you going to increase the volume of the entire 7cm^2 volute? Just tearshaping the inlet alone will do little to nothing for higher flow. And, there's not enough meat for you to make a centered volute of the 7cm^2 housing. This is the achilles heal of the MHI turbine housing. The BEP bolton turbine housing and the FP bolton turbine housing have a centered volute. The PTE bolton housing does not. You would be better off just swapping to a td05h bolton BEP housing for a couple hundred, than trying to force better results out of the 7cm^2 MHI turbine housing. You'll have to port match the turbine housing to the larger SBR outlet. Or you can be smart, keep velocity up and do a better job dividing the pulses, and use the evo3 manifold and just port the step to match the bep housing.

I saw good results going from a 7cm^2 td05h 18g to a bep housing. More power up top. Unnoticeable loss in spool. Evo 3 exhaust manifold. I've run one on my 60-1 and holset h1c. I've been satisfied. It's the turbine wheel for the 60-1 which was the problem. A T31 is tiny for a 60lb/min compressor wheel. A t350 turbine wheel would have netted less back pressure and the ability to reach the compressor potential, even with the bolt on housing. Because the bolton housing has a better volute design than say a pte uncentered volute of similar size.

 

[Join]

I don't like that sbr manifold. It could have done 900whp, but a better division would have netted better results. There is no real divider between the 2-3 pairs and the 1-4 pairs. A better division of pulses helps a turbo that is being maxed out. When you push the limits of a turbo, your backpressure skyrockets. Having more separation between the pulses keeps power from falling off due to exhaust gas reversion. More energy is used for the turbine wheel to spin the compressor and a higher VE results because of less exhaust gas back flow. In other words, the division helps get more flow per rpm, especially up top; and give more energy to the turbine so that it can spin a compressor that needs so much more energy to convert to boost instead of just making heat, due to the fact that you're pushing its map and are out of it's efficiency range. Simply put, if you want the most velocity and the smoothest transition, simply by the seal ring for the evo3 manifold. It will fit right in the 7cm^2 small 16g turbine housing. A big fat collector will do similar to what pboglio described (suddent stem from a 5cm^2 outlet to a 7cm^2 inlet). It will slow gases down and creat more turbulance. There is a big difference between the runner area and the collector area. This is not good. Having that difference as small as possible will be best. An FP race manifold does a better job than a stock /evo manifold because it has longer runners. Any equal length tube runner manifold is better than a cast manifold with short runners and not even a good division between 1-4 & 2-3.

Exhaust gases reach the turbine wheel in pulses. One pulse leaves the turbine before the next can reach it. This is one reason why the evo 8 turbo and all twin scroll turbos do so much better for spool yet have enhanced flow characteristics. They keep the collector area closer to the port area in size. The individual runners see pulses. You don't see guys going to 7cm^2 area runners do you?

Are you going to increase the volume of the entire 7cm^2 volute? Just tearshaping the inlet alone will do little to nothing for higher flow. And, there's not enough meat for you to make a centered volute of the 7cm^2 housing. This is the achilles heal of the MHI turbine housing. The BEP bolton turbine housing and the FP bolton turbine housing have a centered volute. The PTE bolton housing does not. You would be better off just swapping to a td05h bolton BEP housing for a couple hundred, than trying to force better results out of the 7cm^2 MHI turbine housing. You'll have to port match the turbine housing to the larger SBR outlet. Or you can be smart, keep velocity up and do a better job dividing the pulses, and use the evo3 manifold and just port the step to match the bep housing.

I saw good results going from a 7cm^2 td05h 18g to a bep housing. More power up top. Unnoticeable loss in spool. Evo 3 exhaust manifold. I've run one on my 60-1 and holset h1c. I've been satisfied. It's the turbine wheel for the 60-1 which was the problem. A T31 is tiny for a 60lb/min compressor wheel. A t350 turbine wheel would have netted less back pressure and the ability to reach the compressor potential, even with the bolt on housing. Because the bolton housing has a better volute design than say a pte uncentered volute of similar size.

So, basically you tell me I should get a exhaustgasket-ring for the evo from mitsubishi and be better off with that? That is a lot less work!

If I want to make it better for higher rpm and more lb/min I need to get a BEP housing for my S16G? If I will get higher flow with that one, and you recommend it, it's hard to not think twice about it. Do you know where I can find some good pictures of that housing?

Is this the correct housing?

TMM1 Mitsu Inlet, Mitsu CHRA TD05 Internal Stainless.

Is there a good relieable webshop that do international shipping?

Thanks alot,


John

 

[pboglio]

That pic of the SBR manifold says it all. Please put the die grinder down, put the CFD software away and just read this: Visual Frequently Answered Questions - Home Page

I'd skip porting the runners at the head flange as it is also worthless unless its to remove casting slag or polishing maybe.

I'll break it down into what I KNOW works on a small 16g. Your SBR manifold has practically no flow divider. I tried this 13 years ago and there is nothing good about it unless you like lag. That collector diameter is massive, there is no need to go past 60mm. I went from a 57mm diameter ported collector to a 60mm diameter and saw zero gains during back to back testing, this was near the 350 h.p. mark.

Slap on the EVO manifold with a good port match or with the fire ring and you'll be infinitely happier. Do not touch the turbine volute unless its to polish it and/or eliminate the fire ring step, thats it. Some may argue but if you want topend without a major increase in spool lag, just clip the turbine wheel. My 15* clipped small 16g/7cm had full boost by 3400 rpm. Considering the moderate mods I had at the time, the top end was impressive.

I hate to oversimplify things, but a significant increase in lb/min is only going to happen by cranking the boost up higher.

 

[Join]

"Put that die-grinder down, and step away from the viechle!"

I never touched the SBR manifold. This is how it came from them. Gasket mached.

I guess the EVO manifold will be the best choice then. I have heard the SBR crack, so it's a good thing I got the other one!


I ask questions first. Then I do it.

Thanks for the very valueable info from you guys!


John


Edit: The one thing I'll do then, is to make a large as possible radius on top of the wastegate hole, to kind of "glue" the flow to that side to make the wastegate entrance more efficient. I got new FP2's and a Magnus SMIM, so the larger airflow will possibly create a boost creep problem, so I make that radius and call it a day!

 

[the_mork]

The other thing that you can look at is focusing on the other side of the turbo. A mild ramp on the compressor outlet may help reduce turbulence and increase flow slightly if it is matched to a J-pipe, this is one of the services that RRE offers with their 16gs. One word of caution though is that you will have a reduced gasket surface and no way to use the oem gasket so if you knick the outlet face or have a slightly warped flange on the J-pipe it will be harder to get a good seal.

As for better flow on the turbine side I agree that there is nothing that can be done besides removing the step. As suggested you could go to a different aftermarket housing which might be of enormous help but could easily result in a lazy turbo where the larger more free flowing volute sacrifice spool time in order to give you top end but the turbo reaches its limits before the new found gains can be realized. Perhaps a better specimen for that sort of work would be a b16g or an e316g whose compressor wheels could more fully benefit from the increased flow. Also realize that unless you are close to the high end of your efficiency curve that increasing the volute size can only have detrimental effects.

 

[Join]

The other thing that you can look at is focusing on the other side of the turbo. A mild ramp on the compressor outlet may help reduce turbulence and increase flow slightly if it is matched to a J-pipe, this is one of the services that RRE offers with their 16gs. One word of caution though is that you will have a reduced gasket surface and no way to use the oem gasket so if you knick the outlet face or have a slightly warped flange on the J-pipe it will be harder to get a good seal.

I have worked a little on that side, gasket matched it. was about 2mm's on the diameter. But I,ll look into it a little more, because, as you stated the pipe is a bit bigger than the flange itself. I might just make a copper gasket or something instead of the OEM one.. And the flange is perfect, I have fixed that! Do I want to enlarge the diameter all the way in, to where the snail starts? If so I can use a drilling tool and mill it out in there. No drills in aluminium..

As for better flow on the turbine side I agree that there is nothing that can be done besides removing the step. As suggested you could go to a different aftermarket housing which might be of enormous help but could easily result in a lazy turbo where the larger more free flowing volute sacrifice spool time in order to give you top end but the turbo reaches its limits before the new found gains can be realized. Perhaps a better specimen for that sort of work would be a b16g or an e316g whose compressor wheels could more fully benefit from the increased flow. Also realize that unless you are close to the high end of your efficiency curve that increasing the volute size can only have detrimental effects.

That is what I'm thinking also. I would rather choose a Holset setup if I want more power, but for now I'll be happy if I can manage to get 350whp out of it!
My car is also fwd w/ no lsd, so testing will show if any more power is needed..

 

[dsm-onster]

So, basically you tell me I should get a exhaustgasket-ring for the evo from mitsubishi and be better off with that? That is a lot less work!

If I want to make it better for higher rpm and more lb/min I need to get a BEP housing for my S16G? If I will get higher flow with that one, and you recommend it, it's hard to not think twice about it. Do you know where I can find some good pictures of that housing?

Is this the correct housing?

TMM1 Mitsu Inlet, Mitsu CHRA TD05 Internal Stainless.

Is there a good relieable webshop that do international shipping?

Thanks alot,


John

That is the one. Here's some great comparison pictures compiments of Steve. You can see that the area at every point along the cone is larger with the BEP housing. It is centered around the exit to the turbine blades. This gives more flow, since there is less opportunity for the gas molecules to stumble on each other on their way out of the turbine. This also gives a faster spool for the same reason; because the gases can get to the turbine wheel to release it's energy there instead of loosing energy battling with itself.

I recommend a decent tubular manifold. Like the DNP mitsu outlet manifold. My brother and I both have had good success with the eBay version of this manifold and paid less than half the cost. They come with an outlet hole that is the same diameter as the outer diameter of a 2g manifold hole. If you just bolt on the 7cm^2 housing to this manifold, you will have a step out and then a step back in at the collector. You can port the 7cm^2 housing to take away the step back in. And port the manifold flange to take away the step back out. This will significantly increase the volume of the collector. Better idea is this. It will save you alot of time porting which will slow velocity any way.

Now, I still recommend the BEP bolton housing for your goal. But you will have alot of porting to do to get that tubine housing to match, as it's inlet diameter is as large as the outer diameter of the 7cm^2 turbine housing. It doesn't have a step in though. So the inlet is like a ported 7cm^2 housing. You can port the evo3 exhaust manifold to take out the step there and have good results. I recommend a larger a/r & better designed turbine housing with the evo3 manifold over a tubular manifold set up with the stock housing like in my blog. I think you'll get more power per psi and guarantee maxing out the small 16g.

Ideally, I would like to see a tubular manifold with the BEP turbine housing. I don't know your budget. But, for you to run the BEP housing in the best form, you will have to port the evo3 exhaust manifold step anyway. So porting a turbular manifold to match the BEP housing isn't too much more work.

About focusing on the otherside of the turbo. I agree. this is VERY important to maximize your setup. You have a turbine that can push the limits of the evo3 16g compressor wheel right now. More important than the compressor outlet is the compressor inlet. It is much easier for a compressor to push than it is for it to pull. This is the reason turbo systems respond to speed density setups. Speed density does away with a major restriction the maf and allows intake piping better suited to your turbo to be used. It allows a bellmouth which would be difficult to incorperate into MAF intake piping. Upgradeing to a maf translator and orienting this to blowthrough or waiting until V3 dsmlink comes out will be a GREAT way to get more flow out of your compressor wheel. There will be minimal pressure drop from the airfilter to the compressor inlet. I strongly recommend waiting until V3 comes out since the website already stated that it will be out in mid december. The very first update to V3 will be a complete speed density solution. They need just a little more field time before they include this in V3. This will allow you to upgrade to a bellmouth entrance to the compressor inlet which greatly reduces turbulance and significantly increases the choke flow of the compressor. You can have a filter attached to the bellmouth as one unit sold by vendors or you can fabricate your own tube from the bellmouth to the filter. . .

 

[Join]

That is the one. Here's some great comparison pictures compiments of Steve. You can see that the area at every point along the cone is larger with the BEP housing. It is centered around the exit to the turbine blades. This gives more flow, since there is less opportunity for the gas molecules to stumble on each other on their way out of the turbine. This also gives a faster spool for the same reason; because the gases can get to the turbine wheel to release it's energy there instead of loosing energy battling with itself.

That is a big difference! Quick spool is what I like about the s16g turbo, can it be quicker?

I recommend a decent tubular manifold. Like the DNP mitsu outlet manifold. My brother and I both have had good success with the eBay version of this manifold and paid less than half the cost. They come with an outlet hole that is the same diameter as the outer diameter of a 2g manifold hole. If you just bolt on the 7cm^2 housing to this manifold, you will have a step out and then a step back in at the collector. You can port the 7cm^2 housing to take away the step back in. And port the manifold flange to take away the step back out. This will significantly increase the volume of the collector.

As much as I would like a tubular manifold, I will have to keep a stock appearing setup turbo-wise..

Better idea is this. It will save you alot of time porting which will slow velocity any way.

This is something I easily can make at work! Stainless?

Now, I still recommend the BEP bolton housing for your goal. But you will have alot of porting to do to get that tubine housing to match, as it's inlet diameter is as large as the outer diameter of the 7cm^2 turbine housing. It doesn't have a step in though. So the inlet is like a ported 7cm^2 housing. You can port the evo3 exhaust manifold to take out the step there and have good results. I recommend a larger a/r & better designed turbine housing with the evo3 manifold over a tubular manifold set up with the stock housing like in my blog. I think you'll get more power per psi and guarantee maxing out the small 16g.

It is tempting! But, how and where is the portjobs located on the housing? It CAN be bolted on directly, but will gain from porting, or do I have to port to get it to match?
And, the turbine wheel on the s16g, isn't it smaller than the outlet hole on that housing? Doesn't it need to be close to the wall?
Just trying to get an idea of the amount of work.. Can I use the EVO O2 housing and a internal wastegate this way?

Ideally, I would like to see a tubular manifold with the BEP turbine housing. I don't know your budget. But, for you to run the BEP housing in the best form, you will have to port the evo3 exhaust manifold step anyway. So porting a turbular manifold to match the BEP housing isn't too much more work.

Me too! My budget now is like water in a pond in the middle of Sahara mid-summer, after I have spent approx. 11000$ in new engine parts and a RRE DP, 3"Cat and Tanabe 3" cat-back. Also including 1700$ in machining my block 0.20" over, resurfacing and balancing of the rotating assembly here. Before a BEP housing, a new SMIC with better abilities to cool the charge air. I think. That will end up at approx. 620$ included shipping and taxes to me..

Porting and machining is easy, because I work at a machineshop and have about every tool needed for this type of work.

About focusing on the otherside of the turbo. I agree. this is VERY important to maximize your setup. You have a turbine that can push the limits of the evo3 16g compressor wheel right now. More important than the compressor outlet is the compressor inlet. It is much easier for a compressor to push than it is for it to pull. This is the reason turbo systems respond to speed density setups. Speed density does away with a major restriction the maf and allows intake piping better suited to your turbo to be used. It allows a bellmouth which would be difficult to incorperate into MAF intake piping. Upgradeing to a maf translator and orienting this to blowthrough or waiting until V3 dsmlink comes out will be a GREAT way to get more flow out of your compressor wheel. There will be minimal pressure drop from the airfilter to the compressor inlet. I strongly recommend waiting until V3 comes out since the website already stated that it will be out in mid december. The very first update to V3 will be a complete speed density solution. They need just a little more field time before they include this in V3. This will allow you to upgrade to a bellmouth entrance to the compressor inlet which greatly reduces turbulance and significantly increases the choke flow of the compressor. You can have a filter attached to the bellmouth as one unit sold by vendors or you can fabricate your own tube from the bellmouth to the filter. . .

I will look into how much I safely can take out of that compressorhousing outlet without making any holes or make it too thin!

I will give the compressor inlet the same treatment as well. But I understand that I must not enlarge the intake hole all the way in to the compressor housing, because the wheel needs to be close to the walls of the housing?

I have already a Eprom ECU with them.. I have some time to wait, I need to get the engine back in first, and run it in a little..


John

 

[the_mork]

If you move to dsmlink just for speed density I would not buy it until that update has been released.

The package should be worth it (I have no personal experience) even without speed density but because ecm tuning tends to develop software somewhat slowly it would be a shame to buy dsmlink just for speed density and then not be able to use it for a significant amount of time. This isn't a criticism of ecm but there are no guarantees for future products or release dates.


RRE ports their compressor outlets to ~1.5" and they sell a J-pipe to match. I have no idea if this is the optimal size or how to find that out but from what I remember from installing my turbo it was tapered all the way down to the entrance of the outlet. I also have not been impressed with the J-pipe used for that application since they use a larger pipe and make two cuts then press it down and weld the cuts back together and add a flange. Of the two that I have had experience with one leaked (they replaced it free of charge) and one had a very inconvenient angle for putting the upper bolt in, both of them had gobs of metal on the inside from welding. I do not know if there is a better option but I would imagine something could be done.

 

[Join]

If you move to dsmlink just for speed density I would not buy it until that update has been released.

The package should be worth it (I have no personal experience) even without speed density but because ecm tuning tends to develop software somewhat slowly it would be a shame to buy dsmlink just for speed density and then not be able to use it for a significant amount of time. This isn't a criticism of ecm but there are no guarantees for future products or release dates.

No, I have been thinking of DSMLink over a year now, with all the abilitities and possible tweaks to the ecu.
Speed density is only a bonus as I can see!

RRE ports their compressor outlets to ~1.5" and they sell a J-pipe to match. I have no idea if this is the optimal size or how to find that out but from what I remember from installing my turbo it was tapered all the way down to the entrance of the outlet. I also have not been impressed with the J-pipe used for that application since they use a larger pipe and make two cuts then press it down and weld the cuts back together and add a flange. Of the two that I have had experience with one leaked (they replaced it free of charge) and one had a very inconvenient angle for putting the upper bolt in, both of them had gobs of metal on the inside from welding. I do not know if there is a better option but I would imagine something could be done.

This is something I can fabricate and design as I do the work. I have the tools needed..

Thanks

 

[dsm-onster]

It is tempting! But, how and where is the portjobs located on the housing? It CAN be bolted on directly, but will gain from porting, or do I have to port to get it to match?
And, the turbine wheel on the s16g, isn't it smaller than the outlet hole on that housing? Doesn't it need to be close to the wall?
Just trying to get an idea of the amount of work.. Can I use the EVO O2 housing and a internal wastegate this way?

The inlet is formed with the same shape that your 7cm^2 housing will have after the inlet will be ported. Since the evo3 exhaust manifold has a step-out at the outlet, I wouldn't worry about it too much. A step-out creates much less turbulance than a step-in or a step-out and step-in. Gains will be minimal at best. Bolt her her on and go.

Yes you can use the stock wastegate actuator with their internally gated housing. It comes with a much larger flapper for creep control. You probably will actuall have problems with the flapper blowing open because it is so much larger. But I've found that shimming the actuator (at where it is fastened to the compressor housing) works good. And You can also increase the length of the flapper arm, since you work at a machine shop. This will give the actuator more leverage against the flapper when exhaust gases are trying to push it out. Expect for it to blow open. If you start loosing boost up top look there first.

The housing comes formed perfectly to the td05h turbine wheel (14b/16g turbine wheel). Outlet hole is some distance after the turbine wheel exducer. The outlet enlarges from the turbine exducer diameter to the outlet size of about what the evo3 o2housing is, if I recall correctly. . .