Jean

WARNING: Only read if genuinely interested in the truth about your numbers, otherwise , please jump to the next thread

Where do we draw the line, day after day Hp and torque numbers thrown around left and right are becoming a real joke. So how does the customer establish whether he got what he paid for, or he was made to believe so by some dyno mumbo jumbo? Read on....

Engines have a specific efficiency that basically is the limit of the torque output that the engine is able to produce.

The simplest way to measure this efficiency is through something called BMEP (Click me) , which is widely used by engine builders, from factory engineers to F1 engineers and is used to benchmark against other engines, or against different modifications performed in-house.

Other than hardware and succesful programming, major drivers of BMEP are boost and compression, however at a fixed boost and compression, the best engineers in the world would struggle to improve same engine BMEP by 5% + on any engine that is already very performant, without involving expensive (in $$$ millions) R&D and re-engineering. Typical heads/valves, air intake, cams, exhaust, turbos, IC, etc.. might improve by 5% or less that BMEP at the same boost levels.

For reference, an outstanding BMEP on a race engine is considered to be anything above 200, with the top F1 engines standing at around 230. This is at atmospheric pressure (no boost).

So here is some BMEP data based on Porsche factory numbers and let's see if the practice supports the theory, all numbers are adjusted for atmospheric pressure:



It looks like it does…N/A engines are more efficient than turbo engines, race engines are more efficient than street engines, newer engines are more efficient than older engines. You can work out the percentages from above.

Obviously open exhausts, straight through intakes, valves, aggressive cams and mainly, higher compression on N/A engines is key to driving BMEP upwards.

Therefore considering Porsche data, a 997RSR engine improved by 3% vs. the Cup, about 8% vs. the 997 GT3, and 11% vs. the CGT. No wonder you can extract 130HP/Ltr from its atmospheric engine, this puts it very close to F1 territory at same RPMs.

Now to the heart of the matter, how do I know if my engine has the torque I have been told it has? Quite simple indeed..

How much maximum torque can a 993TT engine have at 1.1 Bar of boost?

Max. TQ= (BMEP X displacement)/150.8 x Absolute pressure
Max. TQ= (156 X 220)/150.8 x 2.1 = 478 lb.ft of torque <---This is the maximum REAL TQ you can see on your engine at 1.1 Bar

Here is a chart that will help you situate graphically where your engine should stand, you decide how accurate are your numbers afterwards...



Morale of the story, If you are seeing at 1.1 Bar more than these numbers with bolt-ons-- or over, say, 5-7%, with internal extensive mods (including heads, valves, turbos, itake , exhaust etc..) you might want to question the data provided to you.

Hopefully this will be taken as an eye opener to RL users here, rather than a confrontation to any tuner.

TomF

Interesting read. Thanks Jean, you are a wealth of information!

TB993tt

Jean
Thanks for taking the time to write this. Some questions:

Since 3.8 993tt engines tend to have different compression ratios (I'm not sure but don't the Mahle 3.8 kits comes at something like 8.3:1 from Mahle ?) can you adjust the torque/boost chart to show a 3.8tt engine with 8.5:1 C:R ?

Surely the Torque/boost chart fails to take into account heat which at over ~1.1bar becomes the biggest issue preventing these torque numbers being produced ?

Also, I'd be interested what the BMEP is for a CGT tuned with headers/exhaust/map to 690PS @8100rpm and 648NM @5750rpm - I'm interested how these numbers pan out with the 5% variation for "tuning" ?

For our 993tts the best way for me for seeing how powerful the engine is is to read how well the engine can pump the air ie to measure its maximum intake of air kg/m3 at maximum power, this number can easily be read from Hammer or other system reader -this is the number which JS has been pestering me for when I get my tt back - Is this good science ?

Edit
One more thing - can you work out what the BMEP would be for one of those 9M N/A cars with the super duper heads ?

Jean

Tom, thanks, hopefully it is clear.

TB

Certainly, larger displacement and compression ratios change this equation, I can work out the compression impact by translating it into boost equivalent, it will just take sometime (and energy ).

The larger displacement will proportionally increase these numbers i.e. 3.746/3.6 ltrs= about 5.5% more. Compression has a direct impact on the numbers, this is the main reason N/A engines have better BMEP, other than the rest of the hardware, and finally twin plug also impacts these numbers, how much I cannot tell, but certainly does. All in all I would think that a 10-12% improvement is easily very likely in your engine, with all the above, a higher compression and GT2 EVO heads and cams.

The air measurement and its relation to torque has been discussed before, it is the BSAC that I sometimes refer to, for our engines, about 8.5 hp/lb/min of air..

TT Surgeon

It's all about volumetric efficiency!
There's an old book on the subject, "The Chevrolet Racing Engine", by Bill 'Grumpy' Jenkins. It's a bit dated, but a great read.

kennyboy

Very interesting Jean, thanks for the info.

Does anyone know how these Porsche BMEP figures compare to other manufacturers?

Jean

TB

The modified CGT that you mentioned results in a BMEP of 207, a 10.5% improvement over factory numbers provided the dyno figures are exactly per factory..

As far as compression, displacement and internal modifications are concerned, I believe the best way to answer these questions is by observing how the numbers change at the factory level and trying my best to isolate the changes:

1- Displacement impact:

993 3.6 Ltr: BMEP: 167
993 3.8 Ltr.: BMEP: 170

Displacement increase: 4% BMEP Increase 2%

997 3.6 ltr: BMEP: 187
997 3.8 ltr: BMEP: 192

Displacement increase: 4% BMEP Increase 2.4%

2- Engine internal/external modifications (no displacement change)

An almost entirely redesigned engine with internal improvements in intake, internal airflow, variable cam timing, improved head and head cooling, freer flowing exhaust, updated programming, lighter internal rotating mass and finally compression bump from 11.3:1 to 12:1 yield the following changes:

997 3.6 Ltr: BMEP: 187
997GT3: BMEP: 205
Displacement increase: 0% BMEP Increase 9.5%

We need to keep in mind that this chart does not mean these are the figures that are achieved, there are other limitations such as turbo sizing, heat etc.. this is the Maximum torque that can be achieved at every boost point by this engine.

As for the 9M heads, if I have torque numbers I can calculate that no problem!

KennyBoy, the latest Ferrari F430 has a BMEP of 197, which is a bit less than the 997GT3. However we are not comparing same factory numbers so there could be some slight variations due to dyno testing methodology, but should not be meaningful.

kennyboy

So taking all this in to consideration what we 993tt owners need for big torque without con rod bending boost is.......the 9M heads!

Pure speculation but given what the 9M heads can do on a NA 993 perhaps we could assume some heads tailored for the TT could increase BMEP to around 172.

This would be: (172 X 220)/150.8 x 2.1 = 527 lbs

Jean

Ok Just found the data for the 964 4.0 ltr Ninemeister. It shows a torque of 400 lbs.ft for a 4.0 ltr engine. Engine reported stock, other than the 9M Heads/cams I believe.

9M 964 4.0 Ltr. BMEP: 247
964 3.8RS: BMEP: 172
Displacement increase: 6.7% BMEP increase: 41%

TB993tt

Jean
Thanks for crunching the numbers

Re the CGT, could the large improvement following tuning reflect the relatively low state of tune (considering the componentry spec) of the base V10 unit ?

The 9M numbers look a bit wonkey - looks like some chassis dyno torque numbers you are working to there ?

Jean

TB

I cannot answer you concerning the CGT but it is unlikely driven by a factory conservative state of tune. The numbers look high for those mods. Then again we are comparing a V10 to Flat 6s, maybe they behave differently.

The 9M numbers were taken from here:

https://rennlist.com/forums/showthre...ighlight=heads

They do seem aggressive.

The 3.6 Head conversion reported about 315 lbs of torque on the car tested by Total 911:

BMEP 9M 3.6 Ltrs: 216
BMEP 964 3.6 RS:165
Displacement: 0% BMEP difference: circa 31%

AVoyvoda

Outstanding Jean, outstanding. Read the posts several times to fully understand them.

One question: Everything else remaining equal, a 4 litre engine with an BMEP of say 200 will have twice the torgue of a 2 litre engine with the same BMEP?

Am a little confused, as you mention "The larger displacement will proportionally increase these numbers i.e. 3.746/3.6 ltrs= about 5.5% more" but then you show 3.6 to 3.8 differences where they don't.

Jean

Avoyvoda.

Yes based on the formula for BMEP, if 2 engines have the same BMEP, the displacement change will be proportional to the torque change. In your example, theoretically speaking torque will be double on the 4ltr engine, that is how the formula works.

As far as the Porsche torque data published and its relationship with the difference in displacements, the BMEP numbers are based on the actual measured torque data, so it will for sure vary vs theory.
It could be many things such as the intake size in relation to the larger displacement, the filling of the larger cylinders, cam timings and/or cams in relation to the larger displacement, heat, and also ECU mapping? I wouldn't know which one for sure, you need an engine dyno and controlled conditions to know exactly which one it is..

PramTT

Good info! Thanks Jean


PramTT
lowly BMEP P93T but with MAHA dyno verified numbers....

Felix

Not sure I understand how engines like the Carrera GT, BMW M3 and McLaren F1 make about 80 ft lbs/liter and the 9M engines makes 100. Something doesn't compute. Even the 3.6 engine making 315 ft lbs of torque is hard to believe with 88 ft lbs/liter.