Hi guys, this is the mad info splash that I've been promising for a while - I wrote this months ago except where is says
EditNow the impending wall-of-text is a summary of everything I've learned over the last few months. I'm not going to cite everything, because no. I will tell you that google is not the answer though, this is the kind of stuff you find in old books in these things called “libraries”. What I'm writing will probably not be any better than any blog or opinion in the sense that I'm not writing a verifiable list of sources it’s too much effort, and I did this for fun! All I can say is that the information in books is significantly better than what you find on the web for this topic. I suspect because it represents true engineering complexity and is only really used in industry applications.
To begin with, I want to highlight a few myths that I now understand:
• ITBs improve power
• Shorter is better
• Bigger is better
• Open air reduces intake drag
The above points are the common myths I came across online before starting, and just… no. No.
ITBs Improve Power
The first, and arguably most important point, is that ITBs (in their own right) ARE NOT a performance part. They are part of a performance system, but they do not contribute to the power output of the system. Of all the things I learned this surprised me the most. It turns out that the real reason that ITBs exist was to allow aggressive CAM profiles to run without overlap.
If you have a single throttle body, you need to all the cylinders have to run to common point. This caused significant issues at idle (such as when there is an accident, or in the pits) where the cylinder cross-talk could stall the engine. By running ITBs, each cylinder could run to atmosphere (or usually a pair of cylinders with carbies!) meaning no cross talk. This allowed you to run extremely aggressive cam profiles without the engine cutting out at low RPM.
As a side effect, driver’s noted that the cars responded better to throttle changes. This happens because the butterfly sits in the high velocity airstream within the intake. This means that the air takes less time from when the butterfly opens to when the change is observed within the engine – in other words, you improve the responsiveness of the engine!
These reasons meant that ITBs became a critical component of performance intake design. They allowed you to run very aggressive CAMS without negative effects, and improved the engine feel, but they do not contribute any power to the system on their own; but you wouldn’t have a proper performance system without them!
Shorter is BetterIntuitively, the path of least resistance is into an engine cylinder is the shortest distance – and that is correct. So intuitively, a lot of people believe that shortening the intake in order to reduce intake drag will improve performance, and you would be catastrophically wrong.
NA engines are witchcraft. They are basically resonant boxes that exploit the natural frequency of everything in order to run.I want you to understand that I don’t pretend to know everything, but I also want you to realise that fluids are a magical land of hate and pain and that NO ONE knows anything. It is all guess work and iterative design based on calculated guesses from past experience.
It turns out that the elasticity of air is a significant effector in intake design. Ever hear the term RAM intake? Well, it turns out the actual origins of the term are from some old Chevy Motor that had an air intake that looked like Ram horns. Basically for any given RPM, you want your intake to be a particular length to take advantage of resonant pulsing of the air along the length. This resonant pulse can force more air into a cylinder, because when it is timed correctly the pulse can travel down the length of the intake at the exact moment the valve is open. On this particular point I will actually expand later in regards to the OEM intake, because I actually understand the design now, and it is legitimately brilliant.
The origins of this pulsing behaviour is caused by the valves closing. When the valve closes on a cylinder, the air carries with it momentum, and keeps travelling forward. As the air hits the now shut valve, the incoming momentum compresses the air, which then decompresses sending a pulse back out the intake. If the intake is the correct length it will bounce from one end of the intake to the other. Designing your intake to take advantage of this is called ‘wave tuning’
Bigger is BetterNope. This one is a bit easier to explain though. As I said before, the moving air carries with it momentum. Physics baby! If you make the intake the right diameter, you will solve a best balance point between intake restriction and maximum momentum which will add up to maximum cylinder filling. As the cylinder slows down when it approaches the bottom of the cycle, the only thing filling the cylinder is the momentum remaining in the air. Too small an intake and you will restrict the mass flow of the cylinder, too big and the volumetric efficiency will be low. And if you’re using ITBs you will also reduce the velocity profile over the butterfly and reduce the throttle response.
Open Air Reduces Intake DragOkay, this is seriously one of the more magical components of intake design.
Remember how I said the origins of ITBs was to allow aggressive CAM profiles to run to atmosphere? Well, it turns out that in reality, having groups of cylinders running to a plenum (which I will hence forth refer to as a pulse chamber and you will understand why soon) almost always performs significantly better than open atmosphere, especially with overlap. The only exception would be extremely high RPM motors (over 12 000 RPM) where it can be difficult to design an efficient pulse chamber for high velocity flow.
The reasoning behind the open air solution, is that there is less drag – again, this is true – but remember that magical resonant black magic thing? Yep.
Pulse chambers are too hard for me to properly understand in my short time learning, let alone to properly inform you of their operation – there is a lot going on.
- First, a pulse chamber can restrict the air flow, but increase air velocity (and hence momentum) contributing to cylinder filling. This works on the same principal described earlier.
- Next it acts as a resonant chamber with its own resonant modes to assist with cylinder filling. Like a RAM intake, pulsing within the chamber can send air back down the intake to an open cylinder with more momentum.
- Third it can provide a velocity over the intake trumpets which induces another resonant effect. Know how you blow over a bottle and it hums? That is Helmholtz resonance. The velocity stream (which you can see in the simulations once I do a proper write up on those) can induce resonance in the intake tract (if everything is perfect) assisting the RAM effect.
- And finally, it allows the backlash (the air sent back up the intake from a closing valve) to assist in filling an open cylinder. This works by briefly increasing the pressure inside the pulse chamber (hence why an intake plenum is actually a pulse chamber) when the positive pressure wave hits the chamber at the same time that another cylinder is partially open. The increased pressure is naturally sent down the intake with the negative pressure from the open valve.
Below are animations of the final design for the airbox that was used. The incoming airstream never has to sharply change direction to supply any of the cylinders and the momentum of the incoming air ensures that Cylinder 4 is not starved.
https://www.youtube.com/watch?v=SDHJF4_CFhAhttps://www.youtube.com/watch?v=nGDdkidqnHISo What Happens if you Add it All up?If you have lots of data on your engine, know all the different velocities, and understand your intake behaviour:
You will design your intake to be the correct length such that resonant pulses are combined with the correct volume for the pulse chamber so that it resonates at the same RPM meaning that the positive pressure pulses combine at the same time that another cylinder is opening, so that the whole intake is in positive pressure compared to atmosphere so that it slams down pressurised air to the open cylinder through the intake that is the right diameter ensuring high velocity ingestion for maximum cylinder filling. If you do this all perfectly, you can end up with anywhere between 2~10 PSI in your pulse chamber at the ideal RPM – you can actually run you engine intake at positive pressure by exploiting resonance even though it is entirely fed by the open atmosphere. The reason this works is that the momentum of the air heading into the pulse chamber via the MAF prevents the air escaping back to atmosphere allowing the pulse chamber to remain pressurised. Keep in mind, this is an ideal scenario made for one specific engine at one specific RPM.
It turns out that
a real ITB kit is the cherry on top of a perfectly tuned induction system. If you made a perfect induction system, you would still benefit from having ITBs rather than a throttle between the MAF and the pulse chamber because your ITB kit will sit right inside the highest velocity profile.
So the OEM system…?It’s all about torque, and it is genuinely brilliant. I actually went to the (considerable) trouble of modelling up the DISA intake for simulation, and I learned a lot from it.
The older M42 intake is designed to provide maximum power at ~7000 RPM by using an intake length of ~320mm. The throttle runs into the centre of the pulse chamber to try and reduce cylinder starvation. It’s simple, straight forward and top-end performance focused.
The DISA intake from the E36 cars – holy shit is that some brilliant design and is all about improving real-world driveability.
The DISA manifold has two pulse chambers. The first is the same 320mm (peak power, DISA valve open) and the second is at 800mm to create peak performance at ~2500rpm! It does this by isolating each cylinder entirely by closing the DISA valve so that each cylinder runs
800mm before it can cross-talk with the pulse from another cylinder. Even smarter, they utilise the firing order of the engine so that they only need two runners to service all four cylinders to the furthest pulse chamber!
Edit: Since I originally wrote this you can actually see when the DISA switches and prevent the power drop off after peak torque is reached. Just before 4750 RPM power starts to drop from intake drag until the valve switches and the resonance prevents further drop.
Even better, the design is equal length – so no cylinder starvation, and in the first chamber they put a golf ball pattern in the casting to assist the air with the sharp direction change from the throttle body. When I can animate the simulation results I will come back here and link it. In the simulation you can see that the golf pattern prevent all the air from the throttle running into just one runner by breaking up the flow. Very clever. This is the model I used to simulate the OEM behaviour.
All this combines to an extremely turbulent air flow which significantly assists fuel mixing at low RPM PLUS the air-bleed injectors to also assist fuel mixing all adds up to better low-down torque. Clearly, it was on BMWs mind that street comfort was a big issue for their sporty engines.
So what about Rama's kit?Edit: I wrote this well before the results, and was too afraid to post it in case I was wrong – so now victory!
I don’t work for RHD. I’m designing this kit with Rama’s help because I wanted a decent ITB kit, and I wanted to learn something. I say this, because I just want you to realise that everything I say is honest because I have nothing to lose or gain on the success of the kit.
I have high hopes.
Rama, who has a lot more experience than I, is very confident that the kit will improve performance. I however, believe it will, but this is my first rodeo, so I am more just interested to see the results.
Our kit isn’t just an ITB kit, it is a properly designed, tuned-length induction system with a calculated guess at a pulse chamber volume, and all of it has been verified with CFD. I feel quite confident in saying that I expect there to be a significant improvement in top-end performance with this kit. The design is as smooth as possible for flow, and the castings are good quality. I honestly believe there will be a significant improvement in flow rate at the 4000+RPM range because of the much more direct intake design. Our intake should resonate between 6000~7000RPM utilising the same OEM 320mm short-path intake tract, but with MUCH less obstruction.
I think for the DISA intake, it is likely that there will be a loss in low-end torque, simply because it isn’t possible for us to replicate the 800mm length with all the convolutions that would increase fuel mixing down low. An interesting point though, our intake diameter is a few mm smaller than the OEM casting – but will be better matched to the cylinder head – which will affect the low RPM air velocity, which may mean that we don’t see much loss, or any
Edit: The tighter diameter prevented any loss from occurring!
Rama has selected the ITB diameter on his previous experience, and I’m not in a place to doubt it – though in time the Dyno will tell all!
I think that it is very likely that this kit will really open up the top end for the M42/4 engines, while doing very little to lose lower end performance. And I’m really looking forward to finishing up with the kit!
I’m anxious to see the results, because the design does follow everything I’ve learned (plus Rama’s experience) and I really do want to see positive results!
Below are the Dyno results on the day, 
https://www.youtube.com/watch?v=5W_ObbncUlsWe used a massively long 15" intake to get the results we wanted, and you can see that the peak torque actually occurs earlier than in the OEM setup and still doesn't choke out the engine at any point.
So yeah that is pretty much it, it sounds wicked and added power without any losses by doing everything properly. We used a tight diameter butterfly, long runners and placed the trumpets at an ideal angle to the intake stream. The OEM ECU working in closed loop was completely capable of adjusting itself and it makes real street power and by simply shortening the runner this whole system would be easily adjusted for any race engine.
This should really put to rest any myths or uncertainty surrounding this much fantasised mod and it should also illustrate
with evidence the problems with kits like the Dbilas and M3 ITB conversions
DON'T DO IT!!! 
I'll have a new muffler soon and we'll also see if that actually has any effect on power. I've heard claims that 15kW has been seen from just the muffler before, so here is hoping that I hit 95wkW!
