Single or twins

[GTSBoy]

I'm damn sure that if someone just put a 7670 or 7163 onto an RB26 they would have exactly the ~300rwkW screamer that you all pray for.

Ceramic coating and judicious use of heat shielding is a triviality that I won't even bother to go into.

Edited May 26, 2019 by GTSBoy

[joshuaho96]

Ceramic coating can be done on low mount manifolds as well, I've just never seen anyone make a single EFR low mount manifold. There's also the problem of retaining MAFs, which I'd prefer due to the ITBs of the stock engine. Speed density blended with TPS can work but it impacts emissions which is a relevant problem for me.

[BK]

This thread should be in forced induction really.

I retain twin z32 afm on the 33, but that's because I did it 15 years ago and wouldn't bother now if I was doing it from scratch. Having a 4 bar MAP is fine if you've got the ECU to run it over AFMs and do not impact the stock throttle bodies for emissions with a modern ECU. Low mount single on an RB26, why would you bother ? Having a high mount single 32 myself (currently moving to a 1.00 a/r twin scroll PT6466) and retaining the twins in my 33 HKS gt-ss / Garrett -9 is best compromise. I'd never go single on the 33 for the exact reasons mentioned in this thread, as I basically want it to appear stock as possible. That said, this turbo choice does tend to fall over at the top end - it's all about the low - mid range with gt-ss and I know that. Slightly laggier than stock ceramics, -5 does have more top end but with shitter response, but at the end of the day the more "complete" package for response vs power under the curve vs top end will always be a single now with what's out there turbo wise. Twins are a compromise these days and it is all about the balance of the delivery you can achieve with what you choose to run.

[joshuaho96]

I'm questionable on your statement regarding AFMs vs MAP with ITBs. From what I have seen the resolution of the sensor is for a small portion of the total engine load scale which then requires the remaining portion of the map to be reliant on the throttle position sensor. I just don't buy that throttle position is a reliable load signal, there are way too many factors that can change and throw it off.

Admittedly AFMs don't have the best transient response characteristics. The MAP could be used within the range that it works as an accurate load signal but after that the only thing that works would be the throttle position signal anyways.

[BK]

17 hours ago, joshuaho96 said:

There's also the problem of retaining MAFs

If you want to retain AFMs well that's fine. Why is it an issue for you on a single ? Plenty of single turbo GTR have run AFMs, my 32 when I bought it was setup this way.

MAP vs AFM debate is not really relevant in a thread about single vs twin turbo though and a bit off topic.

[joshuaho96]

1 hour ago, BK said:

If you want to retain AFMs well that's fine. Why is it an issue for you on a single ? Plenty of single turbo GTR have run AFMs, my 32 when I bought it was setup this way.

MAP vs AFM debate is not really relevant in a thread about single vs twin turbo though and a bit off topic.

The problem with most single turbo setups is that the intake piping isn’t really figured out very well, AFMs are sensitive to turbulence and reversion so I’m skeptical that a Y-pipe on the turbo inlet as a mount for AFMs and then some pods strapped to that makes for a good AFM signal.

 

[GTSBoy]

I'm not sure that any of this is relevant in any way. Shirley it's a no brainer to do the single AFM thing on Nistune, and no-one is seriously using PFC on big single GTRs anyway, right?

It's like saying I won't put a big single on a GTR because the stock 225 wide tyres can't handle it.

[BK]

Pretty much this. I will say though that my 32 when I got it from Japan was made to work with one afm on a single with a power fc pro. Put down 530kw with that setup, but have since moved on to 4 bar MAP with Haltech elite 2500.

I thought tuning via 4D when using a MAP sensor alleviated these light load and emissions control issues ?

 

[Dose Pipe Sutututu]

23 hours ago, joshuaho96 said:

Speed density blended with TPS can work but it impacts emissions which is a relevant problem for me.

Poor tune, poor integration of hardware such as EGTs and WBO2.

Emissions testing is also steady state.

[joshuaho96]

1 hour ago, GTSBoy said:

I'm not sure that any of this is relevant in any way. Shirley it's a no brainer to do the single AFM thing on Nistune, and no-one is seriously using PFC on big single GTRs anyway, right?

It's like saying I won't put a big single on a GTR because the stock 225 wide tyres can't handle it.

Even with a single AFM I haven't really seen single turbo builds that try to avoid turbulent flow into and out of the AFM, most seem to assume speed density instead which doesn't really struggle with these issues.

1 hour ago, BK said:

Pretty much this. I will say though that my 32 when I got it from Japan was made to work with one afm on a single with a power fc pro. Put down 530kw with that setup, but have since moved on to 4 bar MAP with Haltech elite 2500.

I thought tuning via 4D when using a MAP sensor alleviated these light load and emissions control issues ?

 

4D tuning is basically what I described, a blending of MAP and TPS based on throttle position. As you reach something like 20% throttle you go from a filtered MAP signal to TPS as the primary load source with a blending method to make sure the transition isn't noticeable.

MAP sensor with a normal single throttle body setup has lots of intake vacuum so you can clearly see what engine load is doing, I don't dispute that speed density works very well in that application. However with ITBs there's so little vacuum volume that the signal is going to be extremely noisy from the intake valve opening events and because all the cylinders are linked there's a big "vacuum leak" from the other cylinders that makes it really difficult to get good load signal from MAP past small throttle openings.

I do wonder if you could make speed density and ITBs work without TPS blending if you had a MAP sensor for each runner and only fed the ECU the lowest pressure of the 6 signals being sampled. That could work even better than AFMs.

1 hour ago, Dose Pipe Sutututu said:

Poor tune, poor integration of hardware such as EGTs and WBO2.

Emissions testing is also steady state.

Banking on WB02 isn't a great strategy IMO, my experience with O2 closed loop is that unless you're within ~5% before O2 fuel trim on the dyno it can cause the ECU to start chasing its own tail. California smog tests thankfully exempt AWD vehicles from the IM240 drive cycle dyno test but RWD/FWD cars can be required to do a proper smog test instead of just revving in neutral at 2500 RPM.

[GTSBoy]

36 minutes ago, joshuaho96 said:

Even with a single AFM I haven't really seen single turbo builds that try to avoid turbulent flow into and out of the AFM, most seem to assume speed density instead which doesn't really struggle with these issues.

Not sure I follow your concern. Speaking with my aerodynamicist hat on, all flows in all automotive systems of any concern to us are turbulent - as in, they meet the criteria of having a high Reynolds number. Certainly high enough that there should never be any talk about laminar flow. If what you are talking about is "well developed flow", then the sad fact is that there are no duct lengths in pretty much any automotive application that are long enough to provide that either.

The beauty of all this bad news is that it really doesn't matter, especially with the newer blade type AFMs. But even without the new ones, something like the classic Nissan hot wire AFM is just a hot wire anemometer. It is measuring the velocity of the air at just one point in the pipe, and via a "calibration" outputting a number that represents the total air flow through the pipe (the pipe being the AFM body). It really does not matter if the velocity profile across the pipe is even or not. It does not matter if it is high on one side and low on the other, or if it has swirl in it. That's because the relationship between the measuring location in the centre and the average flow remains pretty much the same regardless of the amount of air being drawn through, because the flow remains pretty firmly stuck in the turbulent regime (Re >>2000) for all likely flows. Thus the gross flow patterns remain the same.

Now, it is true that OEMs have usually gone to some effort to straighten out or balance the flow across the pipe. The Nissan mesh will put up a little pressure drop and thus will spread a biased flow profile back across the pipe a little bit, but not much, because it's actually bugger all flow restriction. You need to put up quite a resistance to cause a really biased flow spread out much. Other manufacturers have has structures to knock out swirl or otherwise straighten up flow that wasn't approaching the AFM in an axial direction.

All of these attempts are only doing so much. In reality, all the hand wringing that people have learnt over AFMs comes down to the fact that the relationship between average velocity and the velocity measured by the hot wire is part of the calibration, that the ECU expects to see, and the ECU then has another calibration over the top of that that says that X voltage = Y flow rate and then on top of that you have the numbers in the fuel table that can be moved up and down too. So they are well covered for being able to handle whatever differences in "calibration" the AFM might exhibit as the flow increases from idle up to max power. People wring their hands because they hear about Joe Bloggs who put a pod filter on his Nissan and it ran hella rich (or more likely, lean, blew his engine and that's why people learn to be upset by it). But that's only because the assumed calibration of the AFM was buggered by the change to the inlet conditions and the ECU was looking at a different number than it used to for the same air flow rate. All this is not a problem when you can retune the ECU. You just change the AFM calibration curve, or the values in the fuel table.

Turbulence is not a problem. Not one to be confused with the effects of reversion or the effects of changing the inlet pipe geometry without having the ability to compensate for it at the ECU.

[joshuaho96]

That's some neat info. I have seen cases where aftermarket intakes cause immense amounts of noise on the sensor and all kinds of other issues when not designed with MAF in mind which is why I'm reluctant to blindly have a go at it without any real aero/mech background.

[Dose Pipe Sutututu]

For me, putting it on the cold side, I.e. post intercooler works well, especially when using the R35 card style AFM. 

[Andrew Le]

Update guys got the car back 

 

Stock injectors were max out. Gonna bring it back in for injectors, fuel pump cam gear n cams 

 

and let’s crack 300+ 

[mr_rbman]

1 hour ago, Andrew Le said:

Update guys got the car back 

 

Stock injectors were max out. Gonna bring it back in for injectors, fuel pump cam gear n cams 

 

and let’s crack 300+ 

that's a ticking time bomb if you still have the stock turbos.............

[Andrew Le]

Sorry forgot to put it upgraded -7s 

[BK]

It does say GT2860 on the dyno sheet for induction, so I take it they've been upgraded now.

Edit: posted at the same time.

[Dose Pipe Sutututu]

Please tell him to use derived torque instead of roller torque, so your torque isn't showing some BS sky high number.

[niZmO_Man]

Hahaha that's an epic low-revs diesel engine

[Leroy Peterson]

My pulsar once made 950Nm of torque on a roller dyno and I ain't letting nobody take that away from me