My 300rwkw Adventures

[discopotato03]

Forgot to add , a surprising number of twin scroll turbine housings are manufactured here in Australia and all you have to do is find one with similar (prefer fractionally smaller) turbine dimensions to the GT30 and have it reprofiled to suit . Obviously it helps if the flange is split T3 and a has a useful Area Radius Ratio .

ATP is a box company - they sell bits in boxes from a warehouse . I am only 99.99% certain that the .78 A/R turbine housing is a reprofiled GT32 and if you search the net everyone elso thinks so as well .

[Dale FZ1]

Toyota has used split pulse/twin scroll turbines on the GT4 Rally Celica, and other models too.

Perhaps IHI, Mitsubishi or Borg Warner have something already on the market in a suitable size for the performance target being discussed?

Did a bit of searching, and came up with

IHI - VF36, VF37, VF38. Used by Subaru on EJ20 , EJ25 engines, rated up to 350-370 crank hp

MHI - TD-05, maybe TD-06. Used on 4G63 EVO rally car, rated up to 370 odd crank hp.

BW - K-04, rated to 250 crank kW.

The trend is to build them in size and power capacity to suit sub 2.5 litre engines, but specifically to build early/strong torque and get the smaller engines performing like big ones.

The MHI units are spec'd with titanium-alloy turbine rotors, and thin-wall turbine housings, and apparently cost/durability issues have pulled them back from wider use outside the EVO Lancer range.

Everything I could reasonably locate was rated 100-150hp lower than Mafia's requirements. Can anyone else find more?

[Dale FZ1]

Dale I've been thinking about the costing of the twin scroll system and depending on what you pick the bits up for it may not be anything like 6.5K . With manifolds its been seen that the std lump of iron can support 300 odd Kw so all its really missing is the facility for an external wastegate (1) . I seem to remember ages ago JMS did a project car I think called OneEightVia (reverse panel Sileighty) and they managed to weld a section onto the RB20's cast manifold to mount an external wastegate .

I think welding techniques have developed to the stage that this would not be the end of the word to do reliably . In the US they do it often to production manifolds and if done properly they don't seem to fall to bits . You may have also noticed that HKS's cast low mount RB20/25 manifold has a divided gate mount as well . Either of these gets the whole shooting match on the side of the head so thats the manifold and gate for what 8-1200 ? A ball bearing turbo from Garrett to support 300Kw /400Hp is probably going to owe you 1500-2000 depending on taxes etc . You won't get one out of the box with a twin scroll turbine housing but they can be supplied with them if required . The idea is to not buy one with the expensive Ni Resist GT housing but to get the TS one instead - it could be line ball or even cheaper depending on the cost of the TS Diesel housing and the machining cost to reprofile it for the GT30 turbine and cartridge .

I get the feeling that the numbers chase is lurking again with the question being how little can 300Kw cost . My point is I reckon the twin scroll system done properly will give the RB25DET more of the larger N/A engine type power curve than a single scroll one and not lose any top end . This is a difficult thing to achieve but from an RB six point of view I think its a cheaper option that removing and gutting/replacing the bottom end in the search for larger capacity torque . Sure at very low revs it won't pull like a 20% larger RB30 but it wont cost like one either .

BTW I'm not nocking what Mafias does , great effort and learnt by doing it himself .

Cheers A .

It’s a big ask to get 2.5 litres to produce torque like a 3 litre on and off-boost. But Mitsubishi have done a great job with their 4G63 which is equipped with a split pulse TD-05 turbine. The torque numbers quoted from about EVO 7 onwards are outstanding (equal to R34 GTR), and from low rpm too – just what you want to punch a rally car hard out of corners. The thing with the TD-05 is they appear to have a significantly wider range of housing A/R options and it’s possible (at least for MHI) to tailor spec to exactly meet their performance needs. Going on their results, there is no reason to expect that an RB25 couldn’t produce some very impressive results using the same approach and thinking it all through.

I admire your never-say-die attitude in trying to get the absolute best out of the RB25, and agree that up around this 400rwhp mark an external gate setup is more desirable. The trend in high performance petrol-spec housings and impellers is to cope with extreme temps – somewhere around 1000+ deg C. Not so sure that a diesel spec housing will have the right metallurgical qualities to resist cracking, and trial is the only thing that will tell if the manufacturers can’t/won’t. Welding mild steel to a cast iron manifold is certainly do-able, but I’ve had plenty of rejection from skilled boilermakers approached to do just that.

Their basis of concern was not knowing exactly what the mixture of metals were used in makeup of the manifold, as that affects the ability of the bodies to fuse effectively without the join cracking. And that is even with using the proper cast iron welding techniques. None of them said it couldn’t be done, just that they couldn’t guarantee their work long-term.

I’ve not tried to source a HKS low-mount manifold, but comments around suggest they are out of production and fairly rare/expensive (is this true??). Logistical and financial hurdles notwithstanding, I reckon that the split pulse manifold spec would be great, and there would have to be a great opportunity for a skilled pattern maker working in a foundry to produce a ‘foreigner’ if a HKS unit was made available. Wouldn’t like to pay full freight for one though, given what current workshop charge rates are.

That numbers chase is something we can’t get away from, and for this discussion I can overlook immediate product availability. I don’t expect Mafia to lay his exact costs open, but a ballpark of turbo with IW GT30 housing (~2100AUD?) and custom dump (~500AUD?) + the ~850AUD for the water injection system would add greatly to this side of the equation.

One thing Mafia has in spades is balls and perseverance so hats off to him. Plenty of reading and understanding of the theory, then the willingness to have his results and experiences discussed for the benefit of all. At this stage he’s well out in front, having installed pretty much the turbocharger spec that was discussed about 12 months back when his first upgrade was installed.

Out of time tonight, more later.

[Cubes]

Dale...

GT30 IW from Slide $1859 with actuator delivered.

[Adriano]

Dont know if it is of interest, but the 7mgte supra motors run quite a large split pulse exhaust housing which may be able to be adapted to a gt30?

Good to see porsche have got the vvt housing, hopefully we can see some soon, they run a KKK turbo i think

Edited May 15, 2007 by Adriano

[discopotato03]

Dale the trick with welding these manifolds is partly getting the weld material right for the high nickle (sp?) iron alloy . The parts I've seen welded used stainless steel fittings were TIG welded and the welds actually look really neat .

You'll find that EGT's are not going to be higher actually lower with properly sized and developed components .

Out of time more tomorrow , cheers A .

[Dale FZ1]

Dale the trick with welding these manifolds is partly getting the weld material right for the high nickle (sp?) iron alloy . The parts I've seen welded used stainless steel fittings were TIG welded and the welds actually look really neat .

You'll find that EGT's are not going to be higher actually lower with properly sized and developed components .

Yes, that was the trick that the boilermakers I've spoken to said was the problem. The right material isn't known, and while they can make the weld look neat there is no strong guarantee that the mix is correct. So the advice I got was that even after heating the manifold red hot and THEN welding, and THEN following a proper cooling process, there is a good chance of there being a problem with cracks after a while. I don't have enough technical knowledge to do other than repeat what I was advised.

The JMS experiment obviously worked, but that was a few years ago. I wonder if that car is still running, and wearing that manifold? Undoubtedly someone will have a manifold that they've modified successfully, so I'm not going to say it can't be done - but it might be big task to find someone who can do it properly.

EGT comments were looking at the difference between diesel vs petrol. At full load this could be something like 650-700 degC vs 900-950 deg C pre-turbine which is significant. The other heat related issue I would see is RETAINED heat within the housing itself. No doubt they are robust, but there are reasons why they use different cast iron alloys for high temp applications.

As a matter of interest, who/where are these split pulse housings available, and are they a direct fit to the GT30 cartridge? Typical cost?

At this stage, I found 6Boost style (non divided) manifold ~ 1400AUD, Trust style manifold ~ 1800AUD, and Full Race style ~ 2800. There would have to be a good 500-600 to have the OEM manifold modified and faced if you went that way.

Typical external gate should be ~ 600 - 800AUD and another 300 - 500 to get all the exhaust + gate pipework done.

So depending on the reliability of manifold mods, it might be possible to get within distance of Mafia's costs. The question is though, how good is the tune with this split pulse setup going to have to be in order to get within distance of his knock-free tune and inherent reliability that goes with absence of knock? And is the potential for lower turbine inlet pressure with the bigger A/R split pulse going to help combat knock as effectively across the range?

I agree with GeeTR - best to use the best components and get the best overall result. At this level though, I think WI is too valuable to dismiss.

I think we need someone with balls like Mafia who's willing to give it a run.

Edited May 15, 2007 by Dale FZ1

[Dale FZ1]

Dont know if it is of interest, but the 7mgte supra motors run quite a large split pulse exhaust housing which may be able to be adapted to a gt30?

Good to see porsche have got the vvt housing, hopefully we can see some soon, they run a KKK turbo i think

Variable vane possibly slightly off-track, but some stage in the future they will possibly see service on a Skyline. At this stage, the gear they are fitting to the Porsche wouldn't have quite the mass flow capacity to hit Mafia's power target I suspect.

KKK are part of the Borg Warner conglomerate.

Thanks for the info on the 7MGTE units. No doubt size would be a bit on the small side there too, but there could be options for other lower powered upgrades.

[discopotato03]

After re reading my last post I should have said exhaust manifold pressure and temperature are lower with properly sized and developed divided manifold/turbine housing components . I see elevated EGT's as a lack of tuning or control over AFR/Ign timing or restrictions in the system .

I get the feeling that the point some people are missing is that the engine evacuates its cylinders far more easily and completly with the split system . The big issue is that to get sufficient gas velocity to make the turbine responsive overall the A/R ratio will be compromised with the single scroll housing , this generally results in the exhaust manifold pressure rising out of proportion to inlet manifold pressure . The cause is restrictions now in the exhaust system and the fact that the non segregated exhaust pulses and pressure rise is more than happy to backflow into the cylinders when all valves are open trying to scavange the clearance volume (everything above the piston at the overlapping end of exhaust stroke/beginning of induction stroke) .

Back to the old detonation demon , as we know it's possible to defeat it with anti detonants be they high octane fuels or in Mafias case water/methanol injection . Good as they may be they don't stop the exhaust side pressure rise or reversion so progressively more horsepower is used trying to force the exhaust gas from the cylinders - the term generally used is pumping losses .

The well put together entirely split system will have more effective scavanging/lower charge temperature/lower combustion temperature/lower EGT/less crank power loss to pumping inefficiencies so its all cream . To get even close to this with even a split manifold/single scroll turbo means using a very large A/R turbine housing and then you lose the gas velocity to accelerate the turbine at low/medium engine speeds . Extreme methods to get around this are available such as ALS (anti lag systems) but I think you'll agree that this wouldn't go unnoticed in a road car .

Anyhow to each their own . I have the HKS manifold so with a suitable external gate and TS turbo I think it would work quite well . Sadly my resources are going into the daily driver ATM so some other will no doubt do it before I do .

Cheers A .

[wolverine]

disco, dale, mafia etc this is one of the great threads particularly the last post from disco.

anyone have a HKS manifold lying around for me....

[Dale FZ1]

Not an entirely valid comparison to make, but interesting to see what the freshalloy blokes achieved running the same compressor as Mafia, but with an internally modified 2.0 litre SR20 and the technically superior split pulse system. Whether the actual recorded outputs are closely comparable to Mafia’s is unknown, but I noticed that Mafia 397hp @ 18psi, freshalloy 410hp @ 16psi. Assuming the hp results ARE directly comparable, in many ways it validates 1. The capacity of the BCI-18 compressor and 2. The impact of improved volumetric efficiency through cam specs. The theory, and the practice says that bottom end and mid range improvements are there.

At this stage, until we see the results, there is no evidence to demonstrate whether it has any PRACTICAL on-road superiority to Mafia’s spec, especially when the cost is factored in. I’m all for finding technical improvements, but they have to equate to the on-road experience in a demonstrable manner. I’ll keep off the cost issue.

What I won’t keep off is the tuning issue.

We are in complete agreement that fine control of fuel and ignition are critical components of achieving the potential of the mechanical spec. WI – equipped engines can (and do) run more efficiently at leaner AFR similar to a naturally aspirated engine. They do not rely on extra fuel to cool and suppress detonation. The coolant (water or water-meth) allows significantly more aggressive ignition timing which is worth LOTS of reliable torque through the ENTIRE range. More aggressive ignition improves spool off full-load, and torque through all speed ranges. Whether the split pulse system offers that degree of improvement despite any reduced pumping losses is debateable – but I would say not. I would also say that the split pulse system would not allow as aggressive ignition timing, primarily to control the demon knock. Conclusion: WI equipped system offers better scope to achieve the optimal, knock free fuel + ignition tune, and it is probable that both will be more aggressive (read: power productive) than the non-WI equipped split pulse system could achieve.

At least our RB25 hot-rodders who work on a budget have options. Go for the technically superior turbine spec with fabrication or manifolding sourcing issues, or go for a technically less advanced (I won’t call it inferior) turbine that bolts-up and get to work giving it an aggressive knock free tune with the WI system as an integrated component. Then spend the leftover on event entries Again, it’s great that there ARE options, and it depends where the owner’s priorities and finances are.

At this stage I think we’ve tossed the issues of WI, compressor, turbine, and exhaust manifolding around as aspects of Mafia achieving the 300rwkW mark as far as need be. Mafia can you tell us whether you’ve looked at the inlet manifolding as a hurdle? And you still haven’t told us what the car is being used for.

Edited May 16, 2007 by Dale FZ1

[discopotato03]

If I can find it I'll get a link to one of Geoff Raicers (FullRace Geoff here) posts where he talked about developing a Honda 4 cyl drag engine using a Garrett GT4088R . I may not get the numbers exact but you'll get the picture .

I believe it started with a single scroll .8x A/R turbine housing (possibly the Precision Turbo one) and power production was up somewhere in the 800's range . He progressed through native GT40R divided housings in I think .95 and 1.06 A/R ratios (obviously with one of his own manifolds) and power I think rose while boost pressure went from 40 to (can't remember) to around 24 PSIG . Thats pretty incredible when all thats changing is the manifold gate/s and turbine housing . Its pretty obvious to me that the turbine inlet or exhaust manifold pressure has dropped away markedly allowing the cylinders to be charged and blown down much more effectively . This is an extreme example but the principles haven't changed . The bottom line I believe is that if you can't effectively empty the cylinder (actually I should say clearance volume) at the end of the exhaust stroke you can't effectively fill it on the induction stroke can you ? Where there is hot exhaust gas there can't be cool charge air and fuel so you lose some effective cylinder capacity , whats worse is that the spent gas is not combustable and will dilute the once clean charge air AND its hot so will pre heat it costing you charge density - hot air is less dense than a cooler air so its mass is less inc its oxygen content . Lower mass means less so less to burn fuel with to develop pressure to drive the pistons down on the power strokes .

Anti detonation strategies can be effective but there is power to be found beyond best mean torque timing when there is restrictions in the system . Remember when SK talked about ported heads and improved cam timimg and their effects on gas throughput vs boost pressure ? More of the same only its a bit further south beyond the exhaust ports BUT these things radically change the world the engine see's .

I'm out , peace all A .

Edited May 16, 2007 by discopotato03

[Dale FZ1]

I’ll take the last word before bowing out of the turbo aspect of this thread.

Make sure the information is complete. I couldn’t deduce any change other than in turbine A/R, which still dynamically affects the turbine inlet pressure. Here is the link for all to see: http://forums.freshalloy.com/showthread.ph...6028&page=2 There is absolutely no doubt that effective cylinder scavenging through low turbine inlet pressure is important. Note that the maximum boost pressure reduction was also associated what were admitted to be “aggressive” camshafts. with The cited case though, is EXTREME. 300 more hp, from 39% less capacity, using a significantly larger framed turbocharger? That is in another league guys.

Agreed, it highlights the importance of reduced backpressure. That was identified very early with Mafia’s setup, regards my comments about “effective volumetric efficiency”. ie. better scavenging through reduced backpressure results in better cylinder filling = more power. 50rwkW gain at the same boost pressure validated that.

The sticking point for me – and there is NO absolute right answer – is that at this particular power level, for this particular engine spec (internally bone-stock) the technically elegant split pulse system probably can’t stretch its legs well enough to show a practical superiority on-road especially once the tuning aspect is factored in. At best, I could only surmise that the split pulse system would see a couple of psi drop to run the same output.

The underlying issue remains Mafia’s early comment about “bang for the buck”. Mr. Full-Race obviously has business interests, but he rightfully indicated that simply installing a split pulse turbo on a split pulse manifold won’t necessarily get the desired results. http://forums2.freshalloy.com/showthread.p...scroll+question He’s saying “buy mine, at 2300USD + shipping”. For the money involved, I’d keep the edge on Grandad’s axe and run the WI and live with the compromise.

Meanwhile, maybe Wolverine should pick up that HKS manifold from Adrian, run the split pulse + WI, and put me back in my place with the results?

Top discussion, and Peace.

[Cubes]

For streeter if we want excellent low mid range we are never going to get away from high turbine inlet pressures.

If we do by upsizing the scroll it results in less usable average power. Yes it will make some good top end and all round be more efficient but that really doesn't matter when 90% of the time we are playing around between 3000-5000rpm.

Tune the setup accordingly; street you - need good boost response in the low load first and second gears.

Big scrolls reduce this response dramatically. On the dyno the smaller scroll vs larger scroll doesn't show whats going on in the low load gears.

From what I've seen dyno results and drivability on the street.

Up to 300rwkw the GT30r .82 is the way to go with stock exh and plenum. 300-400rwkw = gt35r .82 + a nice exh. manifold and plenum.

[discopotato03]

Nice , gives me more time to get around to building mine .

[Cubes]

You still looking at the 3ltr Disco?

I really think its the only way if you want some decent boost in by 2600rpm.

If your ever down in adelaide drop me a pm; you can take it for a little spin.

[Adriano]

Just wondering after al this, is my HKS rb26 ex manifold suitable for this sort of aplit pulse scavenging setup, as the runners arent the same lenghs?Another question Mafia, when you were tuning this, if you added more timing did it ping, or did it just lose power, all the rb's ive tuned ive found max torque for a given load is found just shy of detonation, which is the opposite of other motors ive tuned, whereby max torque is found well before detonation.

Edited May 17, 2007 by Adriano

[discopotato03]

Hi Cubes , can't say for sure . From a dollars and cents point of view its more practical to see what I can reliably screw out of the RB25 .

Cheers A .

[Busky2k]

Disco, I'm somewhat skeptical of the twin scroll turbos. They sound great on paper but the dynos that have been posted over at FA have been less than stellar. Have you got any hard proof that the twin scroll is worth the extra effort and more importantly, the extra money?

[discopotato03]

Busky personally I don't like dyno graphs because they're not a good indication of what an engine/car feels like in real world operating conditions . There is that much variation between different ones and all sorts of tricks used to make print outs look good or bad depending on what the operator sets out to do . Dynos are a valuble tuning tool don't get me wrong but there is this want to flash the hard copy around showing peak Hp/Kw figures , the fact that very few drive at that point has to mean something .

Gary once mentioned that he'd like to see dyno plots at half/three quarter/and full throttle loads - I guess to have a better overall idea of an engines power and torque characteristics . And torque that reminds me , you often see power figures quoted for turbo diesels and the Kw count looks pretty limp . But , your average death bringer punting his 3 tonnes of Cruiser/Patrol etc around is not exactly going slow so what gives ? Torque and lots of it but without our cure all (Kw being torque numbers x revs numbers) .

What this all means is that its entirely possible to have a road engine make a lot of torque over a reasonably wide range BUT if you put the vehicle on a rolling road dyno and stand on it (go pedal) it may not generate glowing Kw numbers .

So busky I don't know how to go about giving proof that it works . All I know is that everyone I can think of thats done it half right is over the moon with the results . The night and day bit has been said a few times so those individuals are impressed .

This may sound like I'm up myself but if you look hard at the four stroke cycle and do some research on what piston engines do and don't like - from a gas flow perspective - its difficult to draw any other conclusion . The split system will allow a piston engine to make more torque and over a wider range than the single system . Manufacturers of large slow reving diesel engines are taking advantage of it and have for some time . When you look at an engine with a short usable rev range its even more important to pull the bottom end torque up because there's not too many up "high" to play with .

Performance car manufacturers are using this system as well because I believe they are running out of ways to get respectable power without moonshot revs from small/light/compact engines . Look at it this way , if you made cars and you found a way to get more low down torque from your current engine design do you A) throw some better technology at it B) go back to the design room and start the larger engine design process .

Out of time now , cheers A .