Problems going 600+ HP with CA18DET?

[zeppelin101]

Just trying to ask at what point people think an engine has crossed over into numbers for numbers sake. 400bhp/litre is the highest output engine I've attempted to drive on the road, it was crap.
Do you have a link to either paper, they sound interesting.
Have asked before, but has anyone seen actual proof of an RB26 making over 1000bhp?

Sent from my 5051X using Tapatalk

I don't have a link, I have access to photocopies that a colleague made of the originals years ago though Suspect I'd get a slap on the wrists for handing those out, SAE charge for everything!

Anything over 175 hp/l or so starts to cause trouble imo. Whether its reliability or usability or cost, compromises have to be made. I mean its completely possible to build a 1.8 or 2.0l to make 450hp and have a monster power band with boost everywhere that lasts 100k miles... but it'll cost you an awful lot of money and be complex.

[R3K1355]

To limit the insane terminal speeds they have reduced the strip to 1000 foot from 1/4 mile 1,320 feet.

That was done some time ago, like you said for safety reasons after several nasty crashes.
A few places still do the full 1/4 mile, but NHRA events are 1000ft or less.

I wouldn't be surprised if 1/8 mile starts to be used in the future.

[skyshack]

How (un)reliable are FQ360/FQ400? 180/200bhp/L.

[CNHSS1]

To be honest that's a good point. My CA runs 360hp 200hp/l and was built by prev owner as a fast road car BUT it wasn't just turn the boost up to 11, had forgies, Pauter rods, bigger than needed injectors and had lots of time being mapped properly, not an eBay chip. Did 6-7000 road miles and has been thrashed on track for last 4 years by me. I reckon 500hp would see it needing an annual 'refresh' new shells, rings, oil pump etc but would be ok. 700+ and I don't think it's going to last 5k road miles a year.
You have to bear in mind the DET is based on late 70s technology, it's architecture is CA18 which evolved from pov spec van and pickup engines thru ET 8v Silvia and then added a twin can head. Evo engines have evolved too of course.
Personally I think it's the advances in turbo and edu design that has made a big difference to engine longevity. A friend builds rallycross engines and he said cossie YBs went bang on track at 500hp within an hour or two, but now with more efficient turbo, better ecus and mapping, the boost comes in earlier and more progressive so they don't need to be chasing massive revs to get power which is usually the cause of bits letting go.
I think Nissan realised that a bit of extra capacity helps though, hence SRs went to 2litre and RBs were 2.5 litre and above. A 500hp 2.5l engine will be less stressed than a 1.8 500hp motor so think we have to look objectively. Tones engines have gone bang on occasion, that's always the painful bit of development, but the knowledge filters down and everyone benfits and more reliable, longer listing CAs are the result :-)

[zeppelin101]

How (un)reliable are FQ360/FQ400? 180/200bhp/L.

Requires servicing every few thousand miles and over fuels to bollocks to keep temperatures sensible as well as inconel in the turbine and fairly low compression.

Nearest current production 4 cylinder engine I can think of is the AMG A45 engine at 188hp/l. Again low compression, excellent boundary conditions (look at the air box route for instance compared to other contemporaries) and the turbo is oversized on the hot side to limit exhaust residual / knock at high speed / high load. Which is why the low speed response is poor compared to other similar engines (2250rpm peak torque rather than ~1750rpm or less).

After that you're talking about exotics with more cylinders (Ferrari 488 / Mclaren 675) or maybe an M4 GTS is knocking on the door at 164 hp/l but that has water injection at high speed to control knock and combustion temperatures with 10.2:1 CR.

Turbo sizing is something that people in aftermarket are only just getting smarter on. It doesn't help that we are working on about 30% of the information available to an OEM in terms of instrumentation, but its worth looking at some of the turbos that are on OEM engines for certain specific outputs for an application. You'll find that they are all using much smaller hot sides for a given compressor side than you would get off the shelf - either in turbine housing or even wheel and turbine. Probably topics for another thread / day...

[cleanhands]

Never understood why tuners don't really scrutinise what manufacturers do, they're the ones with the budget. Japanese car tuners particularly have this weird thing for linear boost. If a turbo wants to give you 1.5 bar at 3500rpm don't waste it, chuck some fuel at it and hang on. All those with 2871s who were experiencing boost creep, your turbo is trying to tell you something. Vauxhalls and VWs have been running 10 second 1/4s for years with similar sized turbos.

[CNHSS1]

I think if you actually log the boost levels in modern fast oems the boots levels vary a lot through acceleration, the boost is 'managed much better than to give aftermarket best acceleration, not busting gearboxes or smoking tyres. With advent of traction control and launch control, an aftermarket setup would constany be fighting the TC. Analysing where the cars losing traction and backing out the boost momentarily then the car moves forward quicker and the TC only has to deal with variances in surface not an additional 100lb/ft of torque. The difference between mapping and track mapping is often case of less is more imho

[zeppelin101]

Boost does vary more with rpm in oem engine tunes but that is a function of developing a torque structure. The throttle position, wastegate, VCT, VVL and ignition are all working together to manage torque output versus pedal position at varying conditions. Bare in mind how much it costs OEMs to develop their ECU methodology and compare that to regular aftermarket stuff - not even in the same league. A Bosch Motorsport ECU is 10s of thousands of euros plus additional costs for any further controllers.

You don't need all that for a fairly low tech engine, but the fact is that the hardware is making 'some' performance and that learning can be used if the differences are understood between what you are building and what you are using as a benchmark.

Half the problem is that people will dismiss theory out of hand but actual understanding of what the various bits of the engine want/are doing is totally invaluable imo. I've built my career around it lol

[CNHSS1]

Agree you don't need all the tech, but most seem to be happy with mapping that just has enough fuel to ensure the engine doesn't go bang. On my last racer, we hit max but figs within 35mins of being in the Dyno cell, then Dave Walker spent 7 hrs mapping it to be faster on track. Map the boost solenoid to remove boost spikes so that the torque curve is as smooth as it can be. What looks like a small spike on a graph, is often 15-30lb/ft, and when on the limit of grip tahts enough to light the tyres mid corner. Not as good as an OEM ecu which is monitoring many more parameters but much better than just safe fuel air mapping.

I'dike to try fly by wire throttle body next, you can then map response and 'feel' which would be ideal for the wet map

[zeppelin101]

It gives you another option for anti lag without combustion too. And the throttle opening required to achieve torque changes with air flow so you can get better tip-out response (i.e. coming off the throttle) so that the engine doesn't want to "run on" at medium engine speed when feathering the throttle.