Back in 1976 when I build probably one of the first
turbocharged Honda Civics with a 1200cc engine east of the Mississippi - I used
water injection from a windshield wiper bag and it worked fine - once I got it
working.
While installing the water injection "system", unbeknownst
to me, water had dribbled out of the water tube injection end down into the
intake manifold. The car was on a stand which had a slight slope to the
left.
In any case, after getting everything installed I couldn't
wait to test it out. I got into the car and fire up the engine - there was
an immediate "bump" following by a pining sound coming from the engine.
The pining sound clearly was in sync with the cars rpm. I killed the
engine and got out to investigate.
I check the compression on all four cylinders and
found the down slope cylinder to be lower that the rest. I dropped the oil
pan to view the engine and remember thinking how ingenious the Japanese
engineers were to put a curve into the connecting rod to get the best thrust
angle. Then the light bulb came on as I noticed the other three connecting
rods were straight.
I had re-confirmed the scientific fact that water as a
liquid is essentially incompressible - water had leaked into the cylinder and
when I fired the engine the piston tried to compress it. The forces
sheared the compressing ring land off the aluminum piston destroying the
piston and had forced the connection rod to curve to such an extend that
the curved part was now slapping on the bottom of the cylinder opening each time
the crankshaft went around.
Fortunately I had not tried to drive the car or I almost
certainly would have cause the rod to break and destroyed the
engine;
Probably not as likely with a rotary engine, but I think I
would check very carefully any water injection design and
installation.
Ed