4.5 Airbox details

[Fex2005]

Morning

Does anyone know what the pressure drop in in Intake plenum on the 4.5 AJP?

I have seen a thread on here, that mentions the airflow into the plenum at around 4m3/minute, does that seem a correct assumption?

Cheers

[Tanguero]

It will depend on the engine speed.
Each cylinder in the bank is filled and emptied for every 2 revolutions of the crank.
Total capacity 4.5 l
One bank 2.25 l
Per crank revolution 1.125 l
So at idle 900 rpm = 1012.5 l/min which is as near as drat it 1 cubic meter/minute

[itiejim]

Planning some clever airboxes?

I was looking at mine earlier, I don't think that a set of velocity stacks would be a bad idea, though getting some 70mm ones and working out the right length is another matter...

[Storer]

To avoid air starvation you would need to work on 9m3 per min per side at full throttle. There would be a bit of "ram effect" at high speed though.

[Fex2005]

Thanks, im working off 9m3/m per side at the moment.

The reason for my question is that I am experimenting with various airbox designs. I have removed my old airboxes and now have molds to make some new carbon fiber ones next weekend. Quite pleased with how the mould come out (see below).


2 Part mold open


The closed mold


I am also knocking together a set of new carbon backplates, but testing various designs (including velocity stacks) in solidworks.

One interesting thing the CFD is showing, is that the first intake is getting slightly starved of air compare to the other 3. There is also quite a bit of turbulent airflow within the airbox.







Im now experimenting with adding a central carbon fin to the airbox to help stabilise and channel the air to reduce the turbulence, I am also experimenting with velocity stacks as well as increasing the fillet on each cylinder intake.

Im quite new to solidworks and CFD, so if anyone has any suggestions it would be appreciated!

Edited by Fex2005 on Thursday 8th October 01:31

[HarryW]

Daft question but is that cfd showing the air being rammed in from the inlet/air filter, or is it showing what happens when the 4 throttle bodies pull the air through, a subtle but import difference to me. Although I agree some form of baffle above No 1/2 may help steer more towards the early cylinders.

It also shows that if you have your injectors flow tested you put the higher flowing ones in 7&8 cylinders as there is no individual injector duration trimming with the MBE.

[Fex2005]

HarryW said:

Daft question but is that cfd showing the air being rammed in from the inlet/air filter, or is it showing what happens when the 4 throttle bodies pull the air through, a subtle but import difference to me.

You are correct, I had the inlet and outlet the wrong way around, I have re-done the CFD based upon the throttle bodies pulling the air through the airbox at 9m3/min. I have updated the images on my original post.

Although better, the first cylinder still seems to be getting a bit starved (air following curvature of the rear wall) and there is quite a bit of turbulence.

The images shows how velocity stacks will extend into the main bulk of the airflow path.

Edited by Fex2005 on Thursday 8th October 01:36

[a1rak]

are your new carbon airboxes going to be a copy of the factory fiberglass ones or have you made the new moulds to your own design. Also, have you designed then to fit the 4.2 cast manifold or the 4.5 hoses. ?

[Redmist336]

You're making a huge assumption that flow into each cylinder is uniform and constant. Unfortunately it's no where near the truth. Instantaneous flow charastics will be completely different with only one or two intake valves open at any one time. Then there is reversion and helmholtz to take into account. Both causing a backwards pressure wave of air fuel back into the airbox. Each factor also being very RPM specific. You may find that the one cylinder you presume to be starved of air to be "supercharged" by a backward pressure wave of air/fuel mix from another cylinder valve closing, or at lower RPM, reversion.
Don't forget that there are also at least two helmholtz frequencies to take into account. One for each intake runner, which will be damn near impossible to calculate given the delamination of air/fuel through the excessive taper of the intake runners, and one for the airbox neck.
Your baffle will certainly alter if not destroy any helmholtz effect.
In addition you are talking about 100% VE for your flow rates. Not likely on a production, even a TVR engine. Fuel, as a volume, has also been discounted.
Please don't see me as knocking down your efforts, which I think are simply brilliant! I just want to make you aware that it's nothing like as easy as it first appears. Hence the need for many many hours on the dyno with as many airbox options as possible.

Edited by Redmist336 on Thursday 8th October 04:30

[ridds]

Personally I think the airboxes are too big. All the engines I've worked on and used a 2.5x rule have never responded well to large volumes and have preferred a smaller capacity airbox.

tbh the current airboxes aren't great and there's a lot that can be done to improve them just by remaking them (move the bell mouths off the back plate by a few mm, angle them into the airflow ever so slightly) They may not be perfect and you may not achieve over 100% VE (which is possible, think I've seen 110% on one intake setup I tested) but it's a step in the right direction.

[Fex2005]

Redmist336 said:

You're making a huge assumption that flow into each cylinder is uniform and constant. Unfortunately it's no where near the truth. Instantaneous flow charastics will be completely different with only one or two intake valves open at any one time. Then there is reversion and helmholtz to take into account. Both causing a backwards pressure wave of air fuel back into the airbox. Each factor also being very RPM specific. You may find that the one cylinder you presume to be starved of air to be "supercharged" by a backward pressure wave of air/fuel mix from another cylinder valve closing, or at lower RPM, reversion.
Don't forget that there are also at least two helmholtz frequencies to take into account. One for each intake runner, which will be damn near impossible to calculate given the delamination of air/fuel through the excessive taper of the intake runners, and one for the airbox neck.
Your baffle will certainly alter if not destroy any helmholtz effect.
In addition you are talking about 100% VE for your flow rates. Not likely on a production, even a TVR engine. Fuel, as a volume, has also been discounted.
Please don't see me as knocking down your efforts, which I think are simply brilliant! I just want to make you aware that it's nothing like as easy as it first appears. Hence the need for many many hours on the dyno with as many airbox options as possible.

Thanks thats certainly a lot to consider! I was aware of the helmholtz effect, but as you say, it soon became apparent that its almost impossible to include this in the CFD, hence the concentrating on the air quality and smoothness through the airbox only.

The intention of the CFD model is not intended to give me 100% perfect results and solution, I am more playing around with various designs, which should give me a few ideas for prototypes for dyno testing. The idea is to produce 3 prototype airboxes (and various backplates) from the most promising designs. These will then be tested on a dyno.

a1rak said:

are your new carbon airboxes going to be a copy of the factory fiberglass ones or have you made the new moulds to your own design. Also, have you designed then to fit the 4.2 cast manifold or the 4.5 hoses. ?

The moulds are for standard 4.5 airboxes. I will be using these to produce a set of std carbon boxes for the car this weekend (current ones are almost knackered), I will be then using the moulds to produce a set std of fiberglass boxes which will be chopped about and used as development prototypes along with various backplates.

The CFD airflow model above is for standard 4.5 airboxes with short induction pipes as my current set-up.

Edited by Fex2005 on Thursday 8th October 15:47