Factors Effecting Performance
Turbo Mini Forums

Following on from my recent simulation work on optional camshafts, I thought it would be interesting to compare a few example 998 turbo builds and break down the specs to see how much of a contribution the specific components appear to be making the overall power output.

As a starting point, I've taken Almichie's 998 Turbo build that is an excellent example of what can be achieved with minimal expenditure. His spec is:

12G295 head
R5 T2 turbo
Mirage exhaust manifold
Std cam
Std MT Inlet

This was rolling roaded at 100hp on 13psi boost.

Then I've taken my 998Ti, spec as follows:

12G295 head
R5 T2 turbo
Custom exhaust manifold
MD274 cam
Tuned length inlet manifold

This was rolling roaded at 120hp ish on 13 psi.

Then we have the TT2 5-Port

12G940 head - 33/29 valves
GT1752 turbo
Std MT exhaust manifold
MG Metro cam
Std MT Inlet

Rolling roaded at 130hp ? on 13 psi - NMS Oct09

I think that the only thing that I've left out of the above is the intercooler which I've assumed to be of similar contribution in each case.

Then there's my simulated St2 build:

12G295 head - 33/29 valves
GT1752 turbo
Tuned length exhaust manifold
MD274 cam
Tuned length inlet manifold

Simulated at 150hp at 13 psi. Based on a calibrated model of the 998Ti.

So here's my breakdown ranked in importance:

1) Turbo - 10-15% improvement over the T2
Not just because of the compressor but the improved and larger turbine. The compressor is just a few points more efficient which can be offset by the intercooler. Biggest difference is the larger turbine allowing the engine to rev and make the most use of the head and cams.

2) Tuned length inlet - 10-15% improvement over the standard inlet manifold
Using pulses in the inlet makes a significant improvement on the 998Ti and is explained by the simulations.

3) Head - 8-12% improvement using a head with 33/29 valves over standard.
Not hugely significant. The simulation of a 12G940 with35mm inlets gave a lower power output.

4) Camshaft - 5-10% improvement using a MG Metro or MD274 camshaft over standard
As shown from the recent simulation exercise.

5) Tuned length exhaust - 5-10% improvement over the standard MG Metro manifold.
Based on simulation and RR numbers of 1275 users.

Most significant concept to come out of this is that the turbine and the inlet manifolding play a far greater roll than head and cams, which is where we have previously concentrate our efforts.

Also the above is equally applcable to 1275s.

Fire away :)

Edited by Paul S on 13th Feb, 2012.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Lots of interesting stuff to consider there mate thanks

Stupid phone!

Edited by theoneeyedlizard on 13th Feb, 2012.

In the 13's at last!.. Just

Not sure how useful this is to you Paul, but...

My old engine was RR'd at 169 nanas by the previous owner. I changed nothing but the head and turbo and it made 203 on the same boost.

Standard head - Benross head
T3 - Gt1752

Unfortunately there are other factors involved as it was a different rolling road , different down pipe and I had the head skimmed loads to bring the CR up from 7.9:1 to 8.6:1.

In the 13's at last!.. Just

thats interesting, looking forward to seeing how mine ends up over the next few months, with a 12g202 head, gt1544 or 49 turbo, standard cam, cr 9:1, around 10psi, escort rs cooler. hoping to see 90ish. fingers crossed.

one day boost will be mine!

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Interesting read Paul. It will be interesting to see how the st2 performs lower down the rev range with the slightly larger turbo. I'm guessing the longer runners should help with the lower down torque off boost? I have a gt1549 for my build but have no idea how this will work out as there are no maps for it (I think the ar's were similar to the t2 with perhaps a slightly larger turbine)

When are you hoping to get the st2 up and running?

I expect it be around 500rpm later spooling up, which might make driving tricky, although I will use gear based boost control to stop any wild kicks. I've geared the engine to rev 10% higher so, speed wise, it will not be spooling up much later at all.

The runners are longer than before and the compression ratio higher, so it should not be too bad off-boost. The 998Ti is very torquey.

The St2 engine is going together over the next few weeks, but I have to fix a car to put it in :(

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Can you get the Sprite back on the road or does it need major welding on the sills?

I've made some progress with mine - engine now rebored, crank ground and balanced. Head to refurbish next - hopefully get some more done over the next few months.

Gear based boost control :) - can you do this with ms3 using the speed sensor? That could help with prolonging the life of the gearbox (perhaps not such an issue with your sc setup)

Edited by minivan63 on 13th Feb, 2012.

Sprite needs welding, but we still have the Racing Flame on the ramps, so that has to be done first. Alternatively get the Studio 2 painted, but that means transfering subframes from the Sprite.....

MS3 does the gear based boost from a VSS signal. Using a hall sensor on the crown wheel.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

I wonder why you are not mentionning the CR of the different setups given it's importance (after all, CR is the "boost" of N/A engines ...)

Or maybe they are identical ?

Rusty by nature

I'm going slightly ot here but what are you picking up on the crown wheel to get the vss signal - or is this external like the miglia?

Just out of interest can you attribute any gains to the efi or does this just improve economy/drivability

Paul on the sim your using can you set different afrs on the inners / outers would be interested just how much is to be gain by optmising them

Yes i moved to the darkside

Instagram @jdm_brett

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

TT is 8.3:1

Fastest 998 mini in the world? 13.05 1/4 mile 106mph

www.twin-turbo.co.uk

So, Paul, are we talking the same potential of improvement on N/A setup ?

And would the recipe be the same ?
=> shall I think about long runners (à la Maestro setup) on my MD 256 / 12G295 / 9:1 / Cooper Freeflow / 998 ?
8% is certainly a good improvement !

Rusty by nature

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Maestro inlet on a mini would need a pretty good power bulge

If something is worth doing, it's worth doing half of.

Actually the concept was more "carb pointing to the front and 180° runners to the side" (rather than down à la Maestro)

If I recall correctly opposite waves tunning was used by Mazda on a diesel engine under the name Comprex.
Only tunable on a particular RPM thought ...
And wave tuning a siamese head is not really likely easy to achieve ...
Ho well ...

Edited by Advantage on 15th Feb, 2012.

Rusty by nature

The pressure waves that help improve the VE happily co-exist in the siamese port with the waves from the adjacent cylinders.

Waves from the adjoining cylinders do not cancel each other out or interfere, they just ride across each other.

Edited by Paul S on 15th Feb, 2012.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

So, what is a good length inlet for a HIF44 with a 12G295 head? Can this simulation tell you that? Or is it engine specific?

http://www.turbominis.co.uk/forums/index.p...tid=469104&fr=0

You could try 300mm long runners and a large plenum with the carb feeding the plenum.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

What do you mean by the Carb feeding the plenum? I assume that's not swapping them around (Carb -> Plenum -> Inlet Manifold -> Head)?

http://www.turbominis.co.uk/forums/index.p...tid=469104&fr=0

Glad your posting all this, makes for very interesting reading as my next engine will be getting tuned length inlet and will be doing what i can with the exhaust with the room available. Trying to optimise everything as much as i can to spool the turbine i plan to run, as with my old engine i always saw significant power gains swapping to even a slightly larger turbine housing, so trying to flog that horse a little more :) even if i kept the same compressor, a change in the turbine always made a notable difference, certainly i feel tuning the turbine to be more important than the compressor. so long as the compressors in its efficiency island you cant really go wrong, and there's plenty of examples and data on calculating that.

now i cringe when i think of old school "hybrid" turbos where they kept the same turbine with a bigger compressor with the nieve view if increasing power but not loosing out on lag.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.