[JVJ]

Dear all


I have a doubt regarding how ignition source is being modeled in CONVERGE. In actual experiment spark is an instantaneous process. But when we model it we model it in 2 phases, break down phase and arc glow phase each have respective duration of .5 CAD and 10 CAD (followed generally from example case).

But is this correct since actual engine the whole process is completed in micro seconds. Also the distribution of ignition energy to a longer duration (10 CAD) results in lesser energy available per CAD. This actually decreases the chance of getting ignition right?
Also sustaining the energy source for a longer duration 10 CAD will enhance the chance of getting ignition (like in the case of GDI where combustion is very much depend on mixture distribution) and will be unrealistic. Also it can result in different start of ignition for the same ignition timing right?


By,
Jubin V Jose
Research scholar
NCCRD Automotive lab & IC engine lab
IIT Madras

[nitesh.attal]

The complete spark energy deposition is modeled as two square pulses of energy deposition to mimic the approximate energy deposition curve of a real spark. The Crank duration of these square pulse will depend on your engine speed (as spark duration needs to be converted to CA). The Tumble GDI example case does this for a crank speed of 3000rpm for an assumed spark energy of 40mJ spread evenly over short 0.5CA and long 10CA energy sources.

Originally Posted by JVJ

Dear all


I have a doubt regarding how ignition source is being modeled in CONVERGE. In actual experiment spark is an instantaneous process. But when we model it we model it in 2 phases, break down phase and arc glow phase each have respective duration of .5 CAD and 10 CAD (followed generally from example case).

But is this correct since actual engine the whole process is completed in micro seconds. Also the distribution of ignition energy to a longer duration (10 CAD) results in lesser energy available per CAD. This actually decreases the chance of getting ignition right?
Also sustaining the energy source for a longer duration 10 CAD will enhance the chance of getting ignition (like in the case of GDI where combustion is very much depend on mixture distribution) and will be unrealistic. Also it can result in different start of ignition for the same ignition timing right?

[JVJ]

Dear sir
How to divide 40mJ between .5 CAD and 10 CAD.


With respect to speed how we must vary the energy split and duration in CAD?


In actual experiment are we really getting arching for so long period of 10 crank angle degree?



Jubin V Jose
IIT Madras

[nitesh.attal]

Hello Jubin,


According to Heywood's book on ICE, the arc phase duration is of the order of 1micro sec while the glow lasts for a 1 ms. These are typical numbers but you can calculate the duration in CAD corresponding to 1micro sec and 1 millisecond (or the actual duration of the spark you have in the experiment) to split up the spark energy equally (40 mJ of total energy goes as 20 mJ in arc phase and 20 mJ in glow phase).

Originally Posted by JVJ

Dear sir
How to divide 40mJ between .5 CAD and 10 CAD.


With respect to speed how we must vary the energy split and duration in CAD?


In actual experiment are we really getting arching for so long period of 10 crank angle degree?



Jubin V Jose
IIT Madras

[JVJ]

Dear sir


That clarifies my main doubt. For selecting the sphere volume in which energy is released for break down and arc phase do we need to keep spark plug electrode gap as reference or is there any specific value we need to follow?

[nitesh.attal]

Yes, please use the spark gap as a reference diameter for energy source volume.

Originally Posted by JVJ

Dear sir


That clarifies my main doubt. For selecting the sphere volume in which energy is released for break down and arc phase do we need to keep spark plug electrode gap as reference or is there any specific value we need to follow?