[AkaJabari]

I remember talking about this issue almost a year ago when ECU tuning for this car was still in its infancy. Specifically, concern over DI temps at WOT. Is arghx7 still lurking in here somewhere? I hardly ever check this board much anymore.

I've been lurking in this thread for awhile now and had some ideas about the
torque dip...

Taken from the 2006 study of Toyota's development on the
D-4S system for the 2GR-FSE
http://www.ft86club.com/forums/attachment.php?attachmentid=3868&stc=1&d=132605279
1

4.4 FINAL INJECTION STRATEGY ADOPTED FOR
PRODUCTION
VEHICLES
The final strategy of D-4S system for introduction to

production now will be discussed. Simultaneous injection
is utilized to
optimize combustion at lower engine
speeds. The mixture formation using 100%
DI at higher
engine speeds is sufficiently homogeneous and
combustion
efficiency is at a level comparable to
simultaneous injection due to the
piston speeds. For this
reason simultaneous injection is not required at
higher
engine speeds. Another consideration when
simultaneous injection
is utilized at higher engine
speeds, the injector-tip temperature of the DI
injector
becomes too high promoting injector deposits and
degrading the
durability of the injector. According to the
research [9], a DI injector has
issues when the injectortip
temperature reaches approximately 150°C. For
this
reason utilization of simultaneous injection is limited.
Figure 16
shows the DI ratio of the utilization area for
simultaneous
injection.

I'd imagine this is the reason for the torque dip at ~3400
RPMs. I.E. cyclinder temps are getting dangerously close to that 150C threshold
with both injectors running full load at that engine speed. Rather than flirt
with premature direct injector (DI) failure, the engineers chose to shutoff the
port injectors a bit sooner than would be optimal. Raising the RPM limit where
they scale back should take care of the torque dip but DI longevity will more
than likely suffer. To what degree, who knows.

Originally Posted by arghx7

Well, that article was written about the earlier Toyota V6 engine, the 2GR-FSE. On that engine, the port injectors shut off before the torque dip occurs. The torque dip occurs as the cam phasing (and resulting overlap) ramps down.

On the FA20, based only on the graph I've seen from the FR-S service manual, the port injection comes back right near redline. This implies several things:

1) for whatever reason, on a stock FA20 engine the direct injector tip temperature is within specification to achieve durability targets. They figured out how to keep the injector from overheating even with port injection at high rpm.

2) the addition of port injection at high rpm is most likely to provide additional fuel because the direct injectors either can't supply it or are better off not doing so

There could be constraints in direct injection timing, direct injector flow rates, or fuel pressure. As the injection pulsewidth lengthens on any injection system, by definition you are moving your start-of-injection and/or end-of-injection timing. With port injection you have much longer time for mixture formation; it's all going to end up on the back of the valve anyway, at least for conventional strategies. With direct injection, as you lengthen the pulsewidth you can have mixture formation problems or problems with piston and valve wetting. This leads to smoke, particulate emissions, and oil dilution.