PRC Air Inlet Tracts for Turbo's

[Mark]

I don't know if anyone gives a cr.ap on this subject but I've been doing a bit of research on what people try, and theoretically what is best.

Basically, if you've got a turbo with a restrictor in it - and the restrictor can only flow 210 cfm, then does it matter how convoluted the air inlet tract is, just so long as it flow the same cfm as the restrictor wants at any speed?

Is ram-air important in a turbo car as it is in a N/A car?

Is inlet tract length as important in a turbo car as it is in a N/A car?

One gentleman said to me that it was crucial to pay attention to the length of the air inlet tract in front of the turbo because this would contribute to where the peak torque was occurring etc. A good example of this was that Michael Thompson's, now Windus, PRC-spec WRX (possibly the best example of a PRC-spec rally car in Australia?) still features an inlet tract length similar to that of normal bugeye WRXs. Just everything is a bit oversized...

Then you go look at some pics of old Evo WRCar's and they seem to suck all their air from the hole in the bonnet - which I thought was a low pressure, high speed area (though probably cool in temperature and away from water splashes - as per another person suggested) - but all the air flow is perpendicular/normal to your air filter.

Another gentleman reckons to run an airbox open at the both ends, to keep the airflow moving the whole time with the filter in the middle, rather than stalling the air.

The best I can figure out is to have a cold air box totally insulated from the engine bay, have the right angle bend of the airflow prior to the filter rather than after it because the air is suspectible to turbulence post-filter just to the restriction/pressure change.

I'm curious on peoples thoughts who thought about this before? Also, would be interesting if anyone can find WRCar bonnet photos with enough detail to see how they are doing the air inlets and filter setup...

Thanks.

Mark

[Matt Dyne]

I was looking at a few things.

Subaru changed their inlet design in about the Ver 2 WRX's...previously (as also in the Legacy and RX) the turbo had an elbow with air ducting around a 90 degree corner before going into the turbo...later WRX's utilised a straight in approach with better results...leading me to think that air going into the turbo can affect performance. The new path even went under the hot inlet manifold which you would assume made the air potentially a bit hotter, but that was ignored in favour of the air entrance to the turbo.

Looking at photos of the underbonnet of an ST205 TTE spec, the airbox was absolutely massive...it took up the entire LHF guard cavity (normally reserved for the batter) and even went a bit up the strut tower...CF of course.

My car runs a ram air inlet from behind the bumper up into the air box. My airbox is quite large too. I reckon, (although untested) that my airbox is possibly pressurised at high speeds...probably requires investigation...simple manometer would discover this...unsure of any benefit if any.

I've heard all sorts of things about air stalling...big air filters etc.

Unfortunately, I think any gains would be minor, but you'd really need back to back dyno time to quantify the changes.

I am more interested though.

FWIW, I utilise a 100mm OD restrictor and air pipe...it's normal truck air piping. I had the conversation with MB on the merits of other sizes too as the original pipe was something like 150mm.

Matt

[Spac]

The following is the thoughts of a non-fluid-engineer... In other words, I'm no expert, but I'm sure I'm better than some of the people you've spoken to...

There are three issues: Restriction, air intake temp and pulse tuning.

Pulse tuning is non-existant in front of the turbo. It may make a tiny difference when you're totally off-boost and the compressor is not spinning, but how often will that be?
The rest of the time, the rotating compressor blades will effectively work as a one-way valve - so anything upstream of the compressor is out of the equation.

Heat is fairly easy to avoid - draw the coolest air possible. This is usually as simple as partioning off a section of the engine bay, and ducting cool air from the front of the car into it. Biggest likely hassle is avoiding water and dust being sucked in too - those rice boy "cold air scoops" that are made from 5" air-con ducting hanging below the front bumper are obviously not a good idea.

So then you only haveto look at minimising restriction, and possibly even scoring some ram-air intake if you're really keen.

To minimise restriction is usually dead easy - make the pipe as big and straight as possible, and make the filter as big as possible. Given that you're then choking the airflow down through a restrictor, you won't NEED anything huge, but there's no performance drawbacks in going too large either.

Ideas like making the engine draw air from a moving body of air, have been proven to make less power than if they are drawing from a still body of air. That suggestion is essentially the opposite of the well proven ram-air theory and is obviously not a good idea.

[buzzard]

My thought on this are that the standard setup is fine, though i would
like to do some testing with a maometer to back up my thinking.

Seems to me if it is larger than 34 mm then leave it as is. I was having this conversation with a customer yesterday. Also very few people do back to back testing making it difficult to determine the effect of changes.

For sure there is limited or no pulse tuning in front of the turbo. You will fid the pipe from the throttle body to the intercooler is where the action is. Ever noticed the large box in the right front guard on an EVO 2. It is hooked up to the pipe from the throtle body.

In conclusion what is a reasonable intake for a non restrictor setup is most likely a good setup for a restrictor.

Chris

[Mark]

Well, since I posted this morning, I went over to Banyard-Land and picked up my new restrictor and took across the entire inlet tract and flow benched them and then in a whole range of combinations. The short end of the stick is:
a) Grp. N restrictor flowed around 167 cfm.
b) Grp. A restrictor flowed around 210 cfm.
c) the whole standard inlet tract flowed 149 cfm without a restrictor in it.
d) The total system flow with a restrictor in, was 144 cfm.

Puts a torpedo in a few theories! So basically the stock inlet tract isn't good enough even for a restrictor let alone standard. Strangely, the majority of the things that I thought were going to be really poor performers such as the snorkel and airbox, worked fine compared with the piping.
The things going against are, roughly:
-90 deg bend post filter - as per predicted in my first post.
-convulated rubber pipes
-air flow meter turbulence.

There is some bizarre physics observerd at work- Question:
If you had an inlet tract system consisting of say two parts. Part A, flows 200 cfm. Part B, flows 100 cfm. Combine them. What will be the flow rate?
100 cfm - the minimum amount of flow - is the wrong answer. The answer will be something around say 80 - 90 cfm. Presumeably, the join/transition between the two parts is where it loses it. The length of pipe affects it as well, with boundary layer getting thicker for the longer length of pipe.

Chris, the large (resonator?) box that sits hidden near the bumper is post-turbo. I went to turf it and plug up the pipe in the last few weeks - but found a previous owner had already done it - just seems like an extra amount of pipe to pressurise...

Regardless, it was a interesting afternoon. A big thanks to Mark Banyard for the time and education - and the restrictor is a work of art!

Mark

[scooter]

Originally Posted by Mark

There is some bizarre physics observerd at work- Question:
If you had an inlet tract system consisting of say two parts. Part A, flows 200 cfm. Part B, flows 100 cfm. Combine them. What will be the flow rate?
100 cfm - the minimum amount of flow - is the wrong answer. The answer will be something around say 80 - 90 cfm. Presumeably, the join/transition between the two parts is where it loses it. The length of pipe affects it as well, with boundary layer getting thicker for the longer length of pipe.

When a fluid flows through any piping at all, there are losses due to turbulence, which usually increase roughly exponentially against the flow rate. These are tabulated in head loss curves that will show you what the losses in head (head=pressure, which is what drives the flow of fluid) for any given flow rate. The total head loss in a pipe is then calculated knowing the length of pipe - the longer the pipe, the more loss you will incur

Irrespective if you have a particularly restrictive section of pipe in a system, all the sections of pipe will contribute to losses incurred through the system. Thus there are significant losses through your 100cfm pipe, and negligble losses through your 200cfm section, but they the total loss is the sum of these two losses, hence the 80-90 cfm you achieved.

[Darren]

The question is a little open ended insofar as you have not identified in which direction the flow is occuring or why the flows are dfferent (are we to assume a different cross-section area alone contributes or are there additional geometrical considerations?) - Im no Fluid Mechanics expert but I do know that you wont gain much from the answers unless you fully consider the issues you are dealing with - Fluid dynamics is not an area a layman is ever going to grasp easily.

Obviously the losses at the junction and its geometry (taper or square) in respect to that flow direction will greatly effect the losses that have been discussed above. But if all the time and calculations required to analyse the problem was easy and exact then we would not have flowbenches! You are on the right track to experiment and I would belive 'trial and error' results from your flow bench before any "technical advice" offered from less than an experienced designer/engineer...and even then what sort of an engineer would dispute proven & reproducable results? .

[Mark]

It's all good. Turns out since I looked at the inlet tract issue and was having a chat with my dad, it turns out that he has a first class honours degree from Cambridge University majoring in Fluid Dynamics and Thermodynamics. The things you find out embarrasingly late in life

Mark

[Darren]

Originally Posted by Mark

it turns out that he has a first class honours degree from Cambridge University majoring in Fluid Dynamics and Thermodynamics.

He never told you! Looks like you've just won big time !!

[buzzard]

Generally when you flow a port, restrictor or intake piping you should quote a pressure. As such our Chev heads are quoted 358cfm@28 inHg for the intake port. Small flow benches are often not capable of creating a suuficient pressure to give meaningful results.

Chris