Redesigning an injection pump camshaft lobe and follower...

[peterblais1]

Hello there guys. I've got here a high pressure fuel pump which runs directly on a camshaft with a small flat tappet lifter / cam follower. The factory has gone so far as to DLC coat them, but regardless the life is short. When people put a larger fuel pump on there, all hell breaks loose.

The concept is to convert it to a roller follower and possibly look at changing the cam profile a little bit as well. One issue is that the pump has an "inlet valve" so to speak which is electronically controlled and mapped to the engine via load mapping. In low load situations, the valve opens, allows the pump to draw in a full load of fuel, but does not close immediately as the pump starts its way up... It allows fuel to blow back out the valve until the right displacement is reached, THEN closes and the pump pumps the rest. This is a piston style pump going from ~70psi up to ~1800.

A little overshoot is not a big deal as it is closed loop controlled and there is also a return system. The electronic control is to reduce wasted energy.

So, I sourced a suitable roller lifter which will be "easy" to manufacture a retrofit kit for, and had the camshaft read in with both the factory cam follower as well as the roller. I took that lift data and calculated velocity and acceleration throughout. It appears that the factory grind is a constant acceleration and that there is a LARGE discontinuity causing nearly (probably infinite) acceleration... See the attached acceleration plot...

My question to you guys is- should I just convert the factory profile to roller- That looks like it might not quite be possible anyways- or do I try to "fix" the ramp up and down? The roller follower is heavier then the factory follower, plus the pistons in the pumps these guys are using are also heavier- I don't know that MORE infinite (if there can be such a thing ) acceleration is a good thing... My cam design books are saying this profile is totally unacceptable, I'm a bit surprised that a major OEM would have this, but I am not a "cam guy" so what would I know.

Charts:

http://intengineering.net/images/CAM_LIFT.pdf

Lift

http://intengineering.net/images/CAM_VEL.pdf

Velocity

http://intengineering.net/images/CAM_ACC.pdf

Acceleration

 

[PJGD]

Very confusing; not enough information here. What is the application? What is the fuel? No diesel injection system I know of peaks at 124 bar. If it is a diesel injection system, don't be side tracked by cam design rules for gas exchange valves; injection cams rely on a high "jerk" factor.

PJGD

 

[peterblais1]

Sorry about that... I am just leery about posting too much because the competition is out there working on this same project.

It's not for a diesel injection pump, it's a high pressure pump for a newer gasoline direct injection setup. It pumps to a rail, electronic injectors take it home from there.

The cam events aren't even synchronous with injection events (3 strokes of the fuel pump per cam revolution on a 4 cylinder).

FWIW, I started working on simply taking the OEM lift curve and reworking the same displacement to be a roller camshaft... It appears it may not work- It's looking like it lifts so quickly initially that the lifter radius may be too large to follow the cam. I need to double check my work again, my brain went to mush after a while.

 

[BrianGar]

@ Peter, I wouldent be too leery, Im not setting out to improve or modify this item in particular but Im 99% sure I know exactly what make(s) your referring to just by reading it. So Maybe you could post up a bit more more info in order to get your design problems solved.