Toyota GR86, 86, FR-S and Subaru BRZ Forum & Owners Community - FT86CLUB - View Single Post - Estimated HP gains

serialk11r · 01-10-2012, 05:47 AM

Okay I thought of some stuff. So the reciprocating mass' kinetic energy is more or less recovered (minus friction) when the velocity hits 0 at TDC/BDC. However at the peak velocity it has some amount of kinetic energy. I think the effect on acceleration would be the average kinetic energy stored, but this seems wrong in some way...since the net is 0. If we assume sinusoidal motion and that the pistons actually store some energy (which is bad but not that far off) then it would be pi/4 times peak kinetic energy, which is very significant.

What I'm getting is that it doesn't actually affect transient response or acceleration much because the reciprocating motion is effectively smoothed out by rotational inertia. The effect on acceleration would be very minimal, since the reciprocating motion absorbs energy from and puts energy into the rotating motion. Sounds pretty crazy, but it makes sense. Although heavier pistons and rods would increase vibrations and have a noticable impact on smoothness, which manufacturers seem to solve by using heavier flywheel.

So another question remains, how much do lighter internals reduce friction as a proportion of total friction? I read some paper yesterday that showed something like 1% overall friction reductions with smaller journals alone, which suggests actually reducing bearing loads would do more. Loads increase as a 2nd degree polynomial function of engine speed +the constant effect of cylinder pressure assuming the typical linear relation between normal force and frictional force, lighter mass would be responsible for a similar decrease in the coefficient of the x^2 term, except that half the time the cylinder is exerting an opposite force so that makes things a bit complicated, doh. From typical FMEP vs. rpm graphs it's pretty hard to tell what proportion of the load is non-linear...if there were no cylinder pressure then it would be a straight line, but there is something complicating it. help?

EDIT: Actually I guess loss to oil pump and such would explain the "higher degree" losses. Bearing friction seems to account for about 20% of overall friction (more if water pump is electric), and piston skirt friction would go down a little bit so 25% seems like a safe guess. If reciprocating mass is reduced 20% then overall engine friction would go down about 5%. If engine friction is worth 10% of engine power/torque at full load, then this would be a 0.5% increase in efficiency. If friction is 20%, this is a 1% increase. How does this compare to real data or better models? Oh and of course, how much weight can you trim using 300M rods

01-10-2012, 05:47 AM	#178
serialk11r Senior Member Join Date: Mar 2011 Drives: '06 AM V8V Coupe Location: United States of America Posts: 5,279 Thanks: 285 Thanked 1,075 Times in 759 Posts Mentioned: 13 Post(s)	Okay I thought of some stuff. So the reciprocating mass' kinetic energy is more or less recovered (minus friction) when the velocity hits 0 at TDC/BDC. However at the peak velocity it has some amount of kinetic energy. I think the effect on acceleration would be the average kinetic energy stored, but this seems wrong in some way...since the net is 0. If we assume sinusoidal motion and that the pistons actually store some energy (which is bad but not that far off) then it would be pi/4 times peak kinetic energy, which is very significant. What I'm getting is that it doesn't actually affect transient response or acceleration much because the reciprocating motion is effectively smoothed out by rotational inertia. The effect on acceleration would be very minimal, since the reciprocating motion absorbs energy from and puts energy into the rotating motion. Sounds pretty crazy, but it makes sense. Although heavier pistons and rods would increase vibrations and have a noticable impact on smoothness, which manufacturers seem to solve by using heavier flywheel. So another question remains, how much do lighter internals reduce friction as a proportion of total friction? I read some paper yesterday that showed something like 1% overall friction reductions with smaller journals alone, which suggests actually reducing bearing loads would do more. Loads increase as a 2nd degree polynomial function of engine speed +the constant effect of cylinder pressure assuming the typical linear relation between normal force and frictional force, lighter mass would be responsible for a similar decrease in the coefficient of the x^2 term, except that half the time the cylinder is exerting an opposite force so that makes things a bit complicated, doh. From typical FMEP vs. rpm graphs it's pretty hard to tell what proportion of the load is non-linear...if there were no cylinder pressure then it would be a straight line, but there is something complicating it. help? EDIT: Actually I guess loss to oil pump and such would explain the "higher degree" losses. Bearing friction seems to account for about 20% of overall friction (more if water pump is electric), and piston skirt friction would go down a little bit so 25% seems like a safe guess. If reciprocating mass is reduced 20% then overall engine friction would go down about 5%. If engine friction is worth 10% of engine power/torque at full load, then this would be a 0.5% increase in efficiency. If friction is 20%, this is a 1% increase. How does this compare to real data or better models? Oh and of course, how much weight can you trim using 300M rods Last edited by serialk11r; 01-10-2012 at 06:55 AM.