Dale FZ1

The value of using a multi-nozzle WI setup is something that I’ve looked into. There is a system commercially available, which requires something that looks similar to a fuel rail, but carrying individual water nozzles per inlet runner. Done right, it looks GREAT – check the photo of a custom manifold: Technically it's superior to the single nozzle, so worth investigating. Problems came up in matching water:fuel delivey ratios, in the rated flow delivery of commercially available nozzles, since they need to be 6 times smaller than a single correctly sized nozzle. After a good hunt around, there was one fogger suitable, but at a high cost. Then the issue of fabricating and fitting a water delivery rail, and retaining the nozzles within some bosses welded into the manifold came up, along with a suitable pressure regulator referenced to boost. It was another case of cost:benefit analysis. Is there a significant water mist distribution problem from using a single fogger that justifies the complexity and cost in going for multiple fogger? Perhaps at levels somewhere above Mafia’s it would be important – 700hp plus? Fuel flow rates at that output would make it easier to match the fogger size. Just an idea. With knock, I use the word "explosion" as a best-fit option. To me, it signifies the lack of CONTROL of the combustion event. Like you say, it happens across a range of fronts, rather than smoothly emanating from the spark plug. Super-rapid combustion is not the ideal event, because it needs to happen over a period of time in order to PROGRESSIVELY do work/apply force to the piston over a certain amount of degrees of crank rotation. Mafia's comment about steadily applying the force to components makes sense here. The strength of stock conrods is definitely an unknown, but we are agreed that lack of hammering from sharp peak pressures during knock/detonation will safely take them and stock bearings much further. Pistons too – the main reason I see for using forged slugs is because they are much more robust and will take the hammering of slight detonation for much longer without breaking ring lands. So without knock, the stock cast pistons should safely see good service for Mafia. It was probably a poor choice of words to describe WI as an “add-onâ€, because it’s not something I believe. What I do believe is that it should be considered an essential integrated support system in the same vein as manifold design, combustion chamber design and cam specification . The only real negative I can see is the need to monitor your reservoir levels. Even that isn’t hard when low-level lights are available, and if your reservoir is sized to cater for more than one tank of fuel.

We're on the same page here. Note I used the word "limited". It's not that they're NOT available, but a case of to what extent. The castings are evidently made available by ATP only, so it seems they have done something off their own back. The freshalloy discussions seem to highlight the troubles those blokes are having getting exactly what they want - order a particular housing and another turns up without explanation? Doesn't sound good considering they are living in the land of consumerism let alone here. My comment was aimed more at the extent to which Garrett, IHI, Mitsubishi, B-W or whoever have focused any efforts at producing (for their own turbochargers) this style of housing for the sizes needed to cater for Mafia's performance targets. I'd like to be able to call my Garrett (or other brand) dealer and order a specific turbocharger, with that specific housing and know they can deliver it in a certain time frame rather than wait on an elastic calendar, with the chance that they will send something with the wrong A/R. It smells of VERY limited production runs of those ATP castings, with most being accounted for very quickly. Cost-wise I agree. I wonder if someone (Mafia?) could do a direct side by side costing for what he's running, vs the split pulse setup. For me, this is an essential part of the cost:benefit analysis in deciding what to run.

Split pulse is a good thing. Trouble is the limited commercial availability in the size range required means product knowledge and resulting recognition of the benefits is low. So we see little discussion. As per previous, it was employed in the 80s on a system engineered for the Falcon. From memory it was claimed at the time that the regular firing intervals of a 6 cylinder lent itself best to exhaust tuning and the split pulse, with the wider intervals of a 4 cylinder offering less technical and practical advantage. Execution of the AIT concept may have moved on, but the idea is still the same. If it was not a low volume/high cost option it would be much smarter to use one. A proper tuned/split manifold is not cheap either. Toyota has used split pulse/twin scroll turbines on the GT4 Rally Celica, and other models too. I suspect the difficulty/cost factor in getting the castings right may be a factor in a lack of corporate interest in supplying the general aftermarket, hence the question to Discopotato03. Perhaps IHI, Mitsubishi or Borg Warner have something already on the market in a suitable size for the performance target being discussed?

Adrian I think you're onto a good thing with the split pulsed housing and manifold idea. Like WI - nothing new though. Remember the AIT / Normalair Garrett Falcon XE? They were onto that idea and using it in the mid 80s to some good effect. The theory is good; if well applied, it will work. Our friends in the US are not doing anything new, just doing something that works, and doing it well. That's what so impresses me about engineering - you can take some completely different approaches to getting eerily similar end results. Again though, without side by side testing it would be hard to know which would be "better" overall. Would the potential increase in turbine efficiency and resultant combustion chamber scavenging outweigh the proven ability to get maximum ignition tuning with Mafia's spec? Maybe we could get some ballpark costing comparisons of the parts alone (not labour or tuning time) to see where the RB25 hot rodder's $$$ are going? Given that most of us are on limited budgets, I think that's a very important part of the equation. Just as importantly, is there any option outside of Garrett product that has the same features and quality you like?

I thought that noise in my ears was tinnitus, not drums rolling. :laughing-smiley-014: If I can clarify my understanding about temps, and the dynamics of what I was trying to explain, then I will. Rider: this is what I have deduced, but am happy to be put straight by anyone with a better grasp of what happens. Detonation is brought on by a range of different factors including combustion chamber design, fuel characteristics, excessive heat within the combustion chamber, excessive heat in the intake charge, incorrect spark plug spec, dynamic factors in operation such as backpressure, etc. Control measures are introduced either in the basic engineering of the engine or its support systems, in conservative fuel/ignition strategies, or as an add-on eg. WI, or add-in eg. higher octane fuel to obtain reliable performance. In a reliable combustion event, the fuel load is burned in a consistent, repeatable manner – not an explosion. Because of that, the peak temperature is consistent and repeatable. In a knock event, the combustion is not consistent or repeatable, and may not even originate from the spark plug. It is an explosion of sorts, and the temps and pressures within the chamber peak beyond the capacity of the engine components. Longer term failure looms large… Where WI helps (my understanding now), is in keeping the combustion chamber AS A WHOLE (not just the intake charge) cool. Water has a very high ability to absorb heat. It is this feature that makes the combustion event occur smoothly, avoiding the staccato rapid-fire type burn that happens when knock occurs. Hot spots within the burning mixture are in essence tamed down, and peak temps and pressures are avoided. As importantly though, there is heat removed from the piston crown, combustion chamber, and whatever is exposed to the combustion event. Less heat is retained in the environment where the next combustion event will happen in 2 more revs of the crankshaft. The average fuel load in a given inlet charge for a correctly tuned WI equipped engine should be lower (leaner), for a higher specific output. So I see the exhaust temps as otherwise likely to rise – BUT the water has controlled/contained the rate of combustion. So the cooling effect of the water probably has little net effect on gas temps. Long way of saying that for the engine tuned without WI but to its optimum (without knock) should exhibit similar exhaust temps and probably slightly higher mechanical component temps than an engine equipped with WI but tuned to its particular optimum. This part is not my theory – the 1942 tests by NACA confirmed it by empirical testing. Grandad's axe was sharpened up long ago. We just happen to be dealing with a slightly newer axe with a keen edge. FWIW, I would be very impressed with some data logging of EGT both pre and post WI setup and tuning. cheers

Rob, use maths to help your water nozzle sizing for max flow, then tune the system to get the optimal flow curve up to maximum. Calculate the progressive fuel flow through the rpm range via injector duty cycle against injector flow. Your water:fuel ratio is done as a mass (not volume) calculation, similar to air:fuel. Use specific gravity of each to convert to volumes that your WI nozzle(s) be rated with. Then it's a case of playing a bit to get the ramp rate to match the fuel. Not simple, but not exceedingly difficult either. And WI is unlikely to affect your EGT to any significant extent - it's about controlling combustion chamber temps rather than the temperature of combustion.

You could always try your Mercury dealer, or direct from the US. And to my knowledge they have for a long, long time been one of the most grunty ignition systems on the market.

Adrian Definitely a good point re: measuring exhaust manifold pressure. That would answer most questions, and no doubt pose a few more of their own. Sounds like you have something waiting out in left-field there. Keep the 3076 turbo, ditch the W-M Injection, run even leaner mixtures, and get even more efficient mass-air throughput? Sounds interesting. I hope it can be achieved through economic means? Wacky fuels? Fabricated manifolding? Exotic or aggressive engine internals? One of Dr Dowker's Energy Polarisers (remember those??) Time to stop teasing cheers

Piston engined hot-rodders usually go for cubic capacity for bulk power and/or jungle juice race fuels to make their engines produce the sorts of specific power outputs (kW/litre) that the rotor motors do on a regular basis. I'm not going to pooh-pooh them - they are great bits of gear but not necessarily my cup of tea. So yeah, they've got the tricks a bit more sorted. I believe the two compression types are termed adiabatic and isothermal. The first conforms to the normal gas laws in which P1V1/T1 = P2V2/T2, where the second does not. Isothermal allows for compression without temperature rise, and is evidently where the efficiency comes from when spraying pre-compressor. I do not have a strong science background, so the understanding here is only of the layman's type. I think there's room for different methods of actuation, pumping and metering of water. The key to making it work is correct metering, matched to the fuel mass being drawn into the combustion chambers. Anything else boils down to budget and length of penis arguments. And yes, there are plenty of reports of WI equipped engines having zero carbon deposits when pulled down after many hard miles. What Mafia has shown is that he has a system that allows for maximum efficiency of a given combination. He has been able to use a smaller turbine A/R that (with actual back-to-back comparison) gave a streetable spool-up, allowed the compressor to give pretty much maximum effort, and the turbine has not effectively choked the top end or been allowed to prevent optimal ignition timing from being used. AND he can do it SAFELY using an AFR that should lead to detonation via conventional tuning but allow for reduced specific fuel consumption while doing it. Sounds like this setup offers pretty much all anyone would want, and there are lessons to be learned. Long live WI? You bet.

It's not only the rotor-heads who look at alternatives. Introducing the water pre-compressor is getting very tricky, and beyond my area of knowledge. There's some different laws of physics at play once you do that, and all indicators are that it works even more efficiently. There are also some tales (whether originating amongst old wives or not, I don't know) about that suggests the water will abrade the compressor over time. Mafia will continue to reap the rewards for his efforts every time he drives the thing I would say; and hopefully for a long time.

Quite a valid point about the NOx emissions, but ATM the main point of the discussion is about full load timing. Don't want to get off-track, but I'd say emissions testing will be within a fairly tightly defined range away from full load. Most importantly, a properly engineered (and tuned) water injection system will have the water introduced as an ultra fine mist that turns to vapour (gas) prior to and partly during the compression event. Steam is as compressible as any other gas. The key to tuning a water injection system is to understand what is happening (and needed), and progress slowly. Same as mapping fuel and ignition. The most commonly reported occurrence from running too much water is that the combustion event is overly inhibited, and power lost rather than rods bent. It would be a different situation altogether if the nozzle stopped making a mist though. Like Mafia said, use the right gear to get the right results.

I looked long and hard at your results, and over time they helped me to understand what this 6 bladed internally gated 3076 would do on a RB25. Running a free floating (external gate) type housing is probably going to give higher overall turbine efficiency, and the better flow of an external gate should eliminate the slight boost creep Mafia is showing. I'd say it might run away slightly if he tried for 20psi. Power-wise, my maths say he is pushing around 54lb/min of air @ 2.40 PR, allowing for pressure drops through the inlet tract. That puts it at the extreme right edge of the compressor flow map. To get more power requires more mass air flow in an efficient range, and running extra boost would (if my calculations are right) push the plotted flow requirements upwards and further to the right. He would get a little more, but consistent with the law of diminishing returns as the inlet air becomes progressively hotter and less oxygen-rich. Without changing engine internals or manifolding, I'd say maybe (at best) another 10rwkW @ 20psi will pull it up.

Here's the range of Garrett CHRA that use a 76mm compressor : 60mm turbine combinations 6 blade 700177-5007; 76mm 56T comp + 60mm 84T turbine 7 blade 700177-5010; 76mm 48T comp + 60mm 84T turbine 700177-5018; 76mm 52T comp + 60mm 84T turbine 700177-5012; 76mm 56T comp + 60mm 84T turbine 700177-5013; 76mm 60T comp + 60mm 84T turbine Depending on what you wanted, you can get a few different options on housing size/A/R for both sides + internal or external gated turbines. Both internal or external gated housings have A/R roughly of 0.63, 0.82, 1.06 (give or take) but I'd take a punt to say they are very different internally.

Well take a bow, Mafia. While the turbocharger selection is obviously superior to your old one, not everyone is after 300rwkW. My maths said this particular GT3076 should support around 275-280 rwkW at slightly higher boost, but 1. My efficiency assumptions were out, and 2. Didn't account for water/meth What impresses me is the steps taken to get the absolute best from your combination. Conventional tuning methods would have hampered your outcome probably due to the onset of detonation of some type. Even dangerman4 has had problems combating early knock onset with his bigger turbine GT35. Your control is so efficient the timing was taken past the point of maximum torque WITHOUT knock rearing its head. I've read the actual research papers from the 1940s, and those engineers achieved the same results. Yours is just a modern testimonial to the effectiveness of the system. I recall you had some fairly significant knock control issues with the previous spec GT30 based high flow, that led to you fitting this water/meth kit. Obviously you developed a tune that sorted that one and eventually got you around 250rwkW. What sort of power gain did you observe through tuning and water/meth at that time? (basically I wouldn't ask you to wind this setup backwards by removing the water/meth). I will pass a further comment on fuel mapping with your system too. AFR tuning conventions go out the window once you've got a properly engineered WI system in place. Yes, your fuel system is a problem that needs addressing (side issue) but basically you should be aiming to progressively move your full load AFR to 13:1 - similar to tuning a naturally aspirated engine. You are no longer required to pump excess fuel through the chambers to try and cool things down.

Reviewing the compressor maps for the current 56T 6 blade + 0.60A/R vs his previous 52T 7 blade + 0.70A/R, it was clear that both have the same top end flow capacity, so could theoretically support 295+ rwkW. I am in no doubt that turbine efficiency and flow capacity is the basis for the increase of 45-50kW over his old unit, so releasing that log-jam of pressure in the exhaust manifold by using the GT30 turbine just freed the compressor to do its job efficiently. Having the time/inclination to fit a gauge would help, but it is obviously working a lot better than before. Being a production turbo, I'd agree that none of it is cutting edge - but it is affordable. Where there is some area for discussion is in the actual compressor spec vs the intended application, given that either way seems to be capable of producing the top end power numbers. In short, it's about the torque delivery. 6 or 7 blade? Interesting analogy from a propeller engineer helped clear the matter for me. More blades = more low rpm throughput, but less efficient at higher shaft rpm. To me, that spells easier/earlier boost rise but slightly hotter boost charge at higher engine rpm. Maybe this response = too brutal delivery? Mafia has expressed a view, but until someone tries the T04S/60mm GT30IW combo we won't know for sure. I believe that there is a place for it depending on the car's intended use. Old tech does not always = wrong spec. Grandad's axe with a keen edge will still remove your foot very effectively.

Almost, but given you have the GT35 turbine I would draw a distinction between the two. The difference in volumetric efficiency of the engine with decreased pressure in the exhaust manifold (pre-turbine) will give you even better scavenging, and to some extent the ability to run more max timing than (say) Mafia's GT30. Not much, but enough to make a difference. And 2 degrees could easily equal 15-20 rwkW gain/loss. That aside, I think he's obliged to let a couple of secrets out here on just what went on with his timing.

Darren, I would have to say that a properly engineered water (or water-meth) system is the key to hitting the results Mafia has got. FWIW, he did nominate the system type, but that's only the part of it. Sorting out the volume and rate of delivery relative to the mass-air intake charge, and the fuel mass is critical. Given the amount of time he's spent playing around with that setup. he's obviously got it optimised for the engine/turbo combination. There is some proper engineering literature around on these systems, but the essence of it is that (when properly matched) WI can improve the knock resistance of a given fuel by up to 10 octane (MON). And yes, in my view it is ALL about cooling the combustion chamber and controlling the rate of the combustion event. He's basically got the same end result as those guys who play around with Elf Turbomax rocket fuels and other similar brews, but by a different means. Rather than focusing on Mafia's actual max timing, it might be interesting to hear how many degrees he wound into the map, and what he observed on the dyno? That might give some idea of just what the water/meth did to his tune?

Very worthwhile question to pose, and I don't have a definitive answer. Still, the engine is now running at effectively double its rated output from the factory. It's a testimony to the integrity of the basic engineering, but refer back to my comment and the question as to where the car is used. Extended power loading such as racetrack environment would highlight any weakness quickly. Perhaps some of the guys who are doing lots of track work or racing their RB25s can comment? It's more about thinking where the weak points are, given a tune that won't break pistons and revs that won't stretch rod bolts or break the rods.

Remember that it's always very important to compare apples with apples. Changed cams = altered breathing ability = altered volumetric efficiency. Upsized tiurbine A/R = altered breathing ability (less turbine inlet pressure @ given airflow) = altered volumetric efficiency Either change is going to allow the torque curve to carry through to higher rpm (developing more power) than one running stock internals or with the same A/R. I'd done a fair bit of mathmatical reviews of matching this particular 3076 to RB25 over the past 12 months, and it was interesting to see that Mafia had come to a similar conclusion that just under 20psi would get close to the best from that particular compressor + stock engine combo. Working on Sydneykid's rule of thumb of 80hp losses, the engine is likely to have about 480 crank hp, with the turbo capable of about 500hp. So maybe (and only just maybe) there is another 10-15rwkW hiding there if he wanted to bump boost up 1-2psi. I will confess that I didn't believe it would get to 300 however. A thinking man would be mindful of the integrity of stock conrods though, with extended loading increasing the chances of a bent rod at some stage I would suspect. BTW Mafia, is this thing a dyno queen, or a Drag Queen??

Well after dredging through a fair bit amount of old notes, I found that his first unit had a CHRA # 700177-5011. That made it a 52T T04S compressor, with the cropped 90T 56.5mm turbine. Mafia's dissatisfaction with that unit came (I believe) from a mismatch of compressor:turbine, after the RB spec turbine housing was fitted to it. And there was no real joy with the VG spec housing either. So, no. Not a 2876 based high flow.

Mafia, what is your view on the 0.63 A/R housing? Any reason you would go to the smaller size when there seems to be a preference amongst other owners for the 0.82 or 0.87 offered by Garrett/HKS in this turbocharger range? Is it possible that the smaller 0.63 is holding the engine back from its maximum potential at higher rpm? (and this is not a slag of "you haven't got the best out of your car") Just looking for some discussion on the specs you've settled on and how it equates to the driving experience.

Curiosity got the better of me, just had to play with some figures and see. First thing I discovered was that the pumping efficiency of the GT30 turbocharger unit (and this in many/most ways refers to the low backpressure of the turbine) gave significantly better effective volumetric efficiency compared to any of the small frame GT28 turbos I'd studied. Basically, less backpressure means the cylinders are able to better expel exhaust gases (ie. scavenging) so there is less dilution of the intake charge with remaining exhaust, and increased filling of the cylinders with fresh intake charge mass. Try around 110% effective volumetric efficiency for something like a GT2530 or my highflow, against 120+% calculated for this GT3076. That is a big gain, and means that the compressor just doesn't have to work as hard because overall conditions are better. BTW, the shape of those graphs is outstanding - the full load delivery (which is what the dyno is measuring) is very linear. Now, I know for a fact that there was much discussion a while back about the value (or otherwise) of the T04S compressor assembly used on your previous highflow. I also happen to know you have gone for a 6 bladed impeller option with this unit (CHRA 700177-5007; correct?). Not a fair back-to-back comparison of compressors, but how do you find the actual on-road response when driving from low rpm and load into boost and the hang-on-tight zone? Clearly the results you've got are very very good, not just down to the hardware fitted either. But I've got to ask the question if you have an opinion about this much discussed "proper" GT30 - is the gain due to turbine efficiency, and is there any noticeable downside to the compressor assembly you chose? I've got more questions, but out of time. It might be worth sharing with us what your exact performance target was, and what use you intend putting the vehicle to. Surely it's not a dyno queen?

I thought they all sounded like trucks. Looks like I will have to get up to Warwick in August and find out how time had dimmed the memory on the R31. Thanks for the early warning. Is there an easily accessible calendar of events? That time of year I will have to bring a supply of Bundy to ward off the cold though.

That's the great thing about the category - you're seeing and hearing a variety of machinery. And I reckon it's a good thing if different aspects of the different makes are what gets us going as spectators. The R31 as I said was something I followed as a fan during that era. Same as the Group C Bluebirds, DR30 and the R32. It was a bit of an interest kicked off by Howard Marsden's professional approach to running a team. Way ahead of his time, and the machines were different too. Two things count against me here though - Oran Park is a bit far to travel, and those bloody hearing aids keep buzzing when I turn them up... :laughing-smiley-014: Seriously, I hate it when there is some good motorsport that's just too far away.

No, so I used the values that were available in that spreadsheet. I believed them credible because of the extent of data available for a wide range of injectors. As per previous, while I don't understand what the value represents, I know that we have to input a value to correct.