RB25DET Custom Intake Manifold/Plenum Development

[johnnyjohnny]

Hi all,

This is the start of my R&D on the development/modification of the RB25DET intake manifold and plenum. In stock form it flows "fine", and from what I've read can support around 400bhp (have only read this and have not independently tested it). I'm going to be looking for about 400whp, so in it's current state it won't quite cut it.

Time for mad scientist activity!

A few caveats/assumptions/general comments first:

1. As some of you probably already know, I'm not one to accept things on blind faith, marketing hype, nor believe in "biggest is best" type claims. Numbers and scientific process rule. Butt dynos and regurgitating bad advice don't.

2. Intake manifold theory is similar for N/A and boosted motors. It might be a little different, but physics and its associated laws still apply.

3. Components should be chosen that meet the specific needs of the car and driver. My car is my DD and will be used for HPDE and road circuit lapping - no drag races, drifting or streetracing. This means that a log plenum (Greddy, eBay knock-off, etc.) with short fat runners is NOT APPROPRIATE. I will include a detailed explanation as to why this is the case for those who don't know, or who think they know better.

4. I'm doing this because I can and I find it interesting. If you think it's a useless exercise, too bad - I'm doing it anyway

OK, on with the research.....

[johnnyjohnny]

Here are some measurements of the stock RB25DET manifold from my '93 R33 GTS25T.

• Plenum chamber volume = 2250cc
• Intake runner upper diameter = 50.8mm (note shape of runner at plenum becomes a tall rectangle)
• Intake runner upper area = 2026mm2
• Intake runner length = TBD
• Throttle body ID = 63mm
• Throttle plate area = 3117mm2 (does not take into account the space taken up by the throttle bar and side of throttle plate)
• Throttle body cfm = TBD
• Q45 Throttle body ID = 81mm (contrary to what people have listed, it is not 90mm)

Based on these measurements, it can be seen that Nissan was looking to ensure that the GTS models had a broad torque curve to maintain off-boost response at the expense of horsepower at high RPMs.

From an intake manifold theory perspective, the following will be used in the design:

• Plenum volume should be around 1.25x - 1.5x engine displacement for a turbo application.
• Air flow capability of throttle body, throttle body flange inlet, IC plumbing and IC should be equal.
• Based on my application, runner design should be similar to stock configuration.
• Weight should be kept to a minimum, where possible.

Next....intake theory, and why log-type manifolds are not "always the best".

[johnnyjohnny]

On plenums and runners

An important first consideration is to plan out the powerband location along the rpm range. This is key to determining almost all intake manifold design elements. As my car will be used for daily driving and HPDE/lapping on road courses, I need a broad power range with good off-boost response and low-end torque. A road race car only spends about 5-10% of the time at peak RPM, unlike a car setup for the drag strip or drifting.

I could describe in my own words the basic theory about large vs small plenums but honestly, I couldn't come up with a better description than the following excerpt. This is based on long held theory, racing experience, dyno and flow bench testing. This is not new information - many of the people who developed and refined the theory are already dead ( SY). The theory applies to both normally aspirated, and boosted engines.

"When an intake valve opens, the piston begins to draw air into the cylinder. In order to get into the cylinder, the air has to accelerate. Since air has mass, it requires force to accelerate. That force derives from the low pressure area created by the piston inside the cylinder.

With a small plenum, the mass of air gets moving quickly when subjected to the pull of one cylinder. This is good for throttle response and for low RPM torque, as the air attains more velocity and momentum than it would if there were more of it.

With a large plenum, the air is harder to get moving, which hurts low RPM torque, but once it is moving at a decent velocity, is carries momentum with it, which helps keep air flowing into the engine. This is good for high RPM power. The large plenum also helps decouple the resonant properties of cylinders, making them more independent."

Similarly, runner length, area and taper all affect the HP and torque curves of a motor. For a given intake port shape, the cross-sectional area and length of the runner affects the location of an engine's torque peak in the RPM band.

Because dry air behaves as a compressible elastic fluid, there is a very complex set of conditions called Helmholtz resonation that goes on inside the intake runner and plenum. In a nutshell, once air has entered a runner, it does not simply stop when the intake valve is closed shut and wait for the intake valve to re-open. It bounces back up towards the plenum. It encounters the larger mass of air in the plenum as well as the back wall, and a second wave goes back down the runner. This continues back and forth a third and fourth time, and so on. Tuned runners make use of this theory by ensuring that the length of the runner is such that the wave is coming towards the valve as the valve is opening at a certain RPM, thereby forcing even more air into the cylinder. Obviously, the benefits are much more pronounced on a normally aspirated motor, where the air is at around ambient pressure. However, air still resonates under pressure, and therefore the theory is still applicable for boosted motors.

How does this apply to intake runners? Simply put, longer, narrower runners favor lower RPM's because they have a lower resonant frequency, and the smaller diameter helps increase the air velocity. Shorter, wider runners favor higher RPM's because they have a higher resonant frequency, and the larger diameter is less restrictive to air flow.

Based on my particular application, I'm sure you can now determine why the following, commonly used intakes are not adequate for me:



There is also a well-known issue with these intakes, in that cylinder #5 and #6 tend to run leaner, as the natural path of the airflow is directed towards the end of the plenum and down those runners (more air, less fuel=lean), leaving #1 and #2 with a reduced airflow (less air, more fuel=rich). These conditions can be compensated, to a certain degree, through tuning, however they still remain an issue to contend with.

Also problematic is the shape of the plenum. The back wall of the plenum is not equidistant from the intake ports. Based on the above description, unless the intake runner is a different length (which it is not in these designs), then the resonant frequencies are going to be different across all cylinders.

Finally, through computational fluid dynamics testing of this design (not by me, but can be found on other sites - I'll post the primary source shortly), it can be seen that the airflow is clearly inconsistent across each runner, and turbulence is not well controlled.

So we move on....

[johnnyjohnny]

BTW - Please note that I do not state that log-type plenums don't work, or suck, or are worthless. They have short, wide runners and relatively large plenum volume. Before someone responds back saying that Nissan (in the case of the RB26 plenum) or Greddy probably did their homework and wouldn't sell something that doesn't work, re-read the theory and application.

HINT - These plenums have a large internal volume, and short, wide runners.

All intake manifold designs, ALL OF THEM, are by definition a compromise. This is true for many engine components. There is no engine design that does everything best, and there are very good reasons for this!

[WowWee]

Well done. Anxious to hear more. Thermo fluid engineer?

[johnnyjohnny]

Well done. Anxious to hear more. Thermo fluid engineer?

Ha, sadly no. I wish I did something as interesting as that.

[johnnyjohnny]

OK, first thing I’m going to consider are the intake runners.

As we know, longer and narrower runners place peak torque, and the torque curve, more in the mid rpm range, at the expense of top-end power. As my car is set up for DD and HPDE (road circuit), in which case only about 5% of the time will see peak RPM, I want my peak torque range between 3200 and 6000 rpm.

The current runner design on the RB25DET motor actually looks quite good. Using Vizard’s equation for intake runner diameter, we can see that Nissan was right on at 50.8mm.

Rd = 25.4(SQRT[(RPM x D x VE)/3330])

Where,

Rd = runner diameter, in mm
RPM = target rpm for peak torque
D = displacement, in litres (2.5)
VE = volumetric efficiency (95% or .95)

So, for the power band 3200 to 6000, we can use the following results:

@3200, Rd = 38.4mm
@4000, Rd = 42.9mm
@5000, Rd = 48mm
@6000, Rd = 52.5mm

This range coincides roughly with the factory Peak Torque/Peak HP range (3200 – 6400) for the Series 1 Rb25DET motor. Though Nissan sought to boost low-end response with this motor in comparison with the RB26DETT, they designed the runner to favour the upper RPMs of that range and hit the 250hp mark. Of course, since the RB26DETT used ITBs, there was no resonance tuning used at all, so it’s not really a straight comparison.

The surface condition inside the runners is ‘OK’ – could stand to be smoothed and I’ve heard there are performance improvements of getting it extrude honed. I haven’t seen any hard numbers however, so the difference might be just butt-dyno derived. As the injectors are placed close to the intake ports, we don’t need as much turbulence to mix the air-fuel mixture, so we can focus on air speed and volume. This is satisfactory for my purposes, though I need to be careful that any smoothing of the runner does not increase the diameter further.

I haven’t measured the length yet, so I don’t know exactly how they compare with Vizard’s equations on runner length. It will be interesting to see how close Nissan engineers were in their design. Upon visual inspection of the runners, they were quite careful to ensure that each runner was the same length right up to the plenum.

[johnnyjohnny]

Finally got a measurement of the intake port length for an RB25DET head, which came out to 76mm (about 3"). So, using a total length of 438mm for the stock runners, we can see the following harmonics at play:

For 2nd harmonic, RPM range is from 6951 to 8436 with a pulse strength of 10 percent
For 3rd harmonic, RPM range is from 5223 to 5969 with a pulse strength of 7 percent
For 4th harmonic, RPM range is from 4072 to 4554 with a pulse strength of 4 percent

It is common for relatively low rpm motors (below 10-12k) to tune for 4th or higher harmonics. It's interesting to see that the factory rated peak torque, which occurs at 4800, falls between two harmonics. However, if you look at a dyno plot of a basically stock RB25DET (see mine below), there is an earlier bump in torque at 2900rpm, which actually corresponds to the 5th harmonic. I'd theorize that the large, 180* bend in the intake runner, and the runners being slightly larger in area than optimal, is the main reason for the tuning to be a bit off for the 4th harmonic. Based on the discussion above, though low-end power was a consideration, Nissan still wanted to see respectable upper HP numbers.



Anyways, what we can see here is that Nissan engineers appeared to be applying the theory quite effectively. The runner area and length equate to peak torque where they intended (lower in the rpm range for better off-boost response and driveability). And it looks like they were catching some benefits from 4th and 5th harmonics to aid in cylinder filling.

What does this mean for me? Well, since I do not want to increase the rpm where peak torque occurs, I definitely can't go any shorter on the runners. Also, the existing runners are a little on the large side already, so I can't go larger without negatively affecting peak torque.

So, what will result is an intake manifold somewhat similar to the factory unit, with various dimensions tweaked to get my power curve exactly where I want. Also, from a packaging perspective, I want the throttle body facing forward so that I can run a FMIC without having to run the IC plumbing over any part of the motor. And finally, the runners need to be filled equally in terms of air density, velocity and turbulence. Oh yeah, it all needs to fit under the hood, too.

[maxgohan]

using your analysis,
what do you think of this option?


if I read this correctly, the length of the runners are the same of stock rb25 so that is a positive, however I'm concerned that the following condition may not meet,

"the runners need to be filled equally in terms of air density, velocity and turbulence."

thoughts?

[johnnyjohnny]

Correct - this is not an optimal design unless there is significant internal structures to properly guide the air in the plenum. See the first post for the explanation why this is the case. "Packaging" remains the biggest problem - designing a plenum that feeds the runners equally that also fits in the engine bay.

using your analysis,
what do you think of this option?

if I read this correctly, the length of the runners are the same of stock rb25 so that is a positive, however I'm concerned that the following condition may not meet,

"the runners need to be filled equally in terms of air density, velocity and turbulence."

thoughts?

[ToddJ]

Hey Johnny, what did you come up with? My goal is to shorten my fmic piping to the i/m without using an aftermarket intake manifold.

[mightyrocker]

Wow really great info on here. Excited to see what you come up with.

I am going for power goals just outside of the stock plenum "capacity" so I am a midst a plenum upgrade.

My only thoughts i had on these front facing "log" plenums is that those of us with fmics are drastically shortening our intercooler piping. Never thought to dig this deep.

Have you found that there is any restrictive or negative effect when the stock runners make the 180º turn into the cylinder head? Could you in theory tune the runner length and dia. inside the log to achieve the proper resonance for all 6 cyl?


Also what do you think of hypertune manifolds?

Sent from my SGH-I747M using Tapatalk 2

[Dragon Humper]

I would not take anything David Vizzard has to say on engine dynamics too seriously. 90% of the stuff in his books is over 25 years old and varies from slightly misguided to flatout, completely wrong. It's not all his fault, he did a lot of work but was basing HIS processes on theories and evidence dating back to the mid '60s + the equipment he had for most of his work was, again, 25 years old and the computers were 1/10 as capable as the phone I am typing on.

No NA engine today uses any of his rules of thumb and turbo motors never applied to any of his theories. Once you pressurise the intake all those resonance theories get chucked right out the window. Why? Because any change in pressure affects EVERYTHING, so unless you are never going to change your boost pressure, there is no point in tuning for resonance.

The Hypertune plenum is pretty much the way to go, if you have access to welding services for a good price the forward facing throttle body pictured above is no better or worse than stock. Hypertune + a slightly bigger turbo than stock with a good tune will get you all in boost at 3000rpm and pull a flat torque curve all the way to 7500. 400whp is a cakewalk.


Jon.

[mightyrocker]

Thanks for the input Jon.

Under that assumption, even in an n/a, one would still see very different manifold pressures and would therefore have to tune for 1-4 specific rpm ranges for each harmonic(as johnny was saying) at 1 specific manifold pressure. Being that there is even more pressure variance in a boosted engine, that window of rpm efficiency would become drastically smaller..... it seems kind of redundant.

Regardless, I think most people (mostly myself) are concentrated less on that fine of a level of plenum efficiency, as they are all relatively close in design, and more for equal distribution of air throughout the cylinders. I have been searching for a way to fix the notoriously lean cyl 5 and 6 with a specific plenum.

Jon, or Jonny, have you seen any numbers from the existing manifolds (hypertune, greddy, rips, Ku etc...) that stand out or put to test any of the theories above?

And Jonny, what are you planning on changing from the ones avail. on the market now?



Kurt