T. KORAKIANITIS

Piston assembly dynamics 

 "Korakianitis" is pronounced phonetically
email: tk@mecf.wustl.edu  link: [ 183 kB audio wav]

May 1999 -

   The following table of links is a site navigation map  

T. Korakianitis' main page
  
Research in piston engines		   
Research in turbomachines		  
Engineering practice		   
Opinions on education
Archival publications

Do your numerical computations on valve trains with hydraulic lifters (lash adjusters) match experimental observations or engine experience? What numerical integrator are you using? Your computations may be converging in what appears a correct periodic solution, but does that solution represent nature? If not, why not? How can you be sure? How can you improve your model to compute what experiments tell you you should be computing?

HOW DOES ALL THIS AFFECT ENGINE EMISSIONS?

The figure shows converged computations with implicit and explicit integrators for a journal bearing with eccentric mass. For this case there is an analytic solution with which we can compare. The performance of explicit Euler solvers... Our solver predict the correct trajectory with crank-angle increments (time steps) from less than 1 degree and up to 120 degrees. We need to limit the time step in order to resolve the trajectory rather than accuracy or stability of computation. How does your numerical computation compare?

Include here two or three figures from journal bearings showing integrator performance. Delete from figures any hints on what is the suitable integrator.

Our models for valve train dynamics include:
lifter with disk or spherical check valve;
push rod or overhead cam;
lubricant-cavitation models;
dynamic models with loss of contact and impacts between several components;
the effect of surface asperities; and
other aspects not available in the literature.

Include here the figures of two lifters, two valve trains, and one figure with valve lift with bounce.

 

Click the left arrow  to return to T. Korakianitis' main page