Yellow Bullet Forums banner

1 - 13 of 13 Posts

·
Registered
Joined
·
2,298 Posts
Discussion Starter #1 (Edited)
I was bored and roughly modeled an intake runner from port to valve. BASIC SHAPES, just to get an idea of how the airflow acts. At first glance it looks like the straight shot keeps the most velocity, but is kind of uneven. Am I completely off here? :smt102

Port normal to both openings:



Port normal to valve:



Port normal to intake:



Straight Shot:

 

·
Registered
Joined
·
1,781 Posts
Yes, the valve, valve guide/boss, seat angles, and csa at different points, etc......
 

·
Registered
Joined
·
4,182 Posts
Also, discharge past the valve seat. That would be the most important.
what is the angle of deck?
or are we just looking at the port without bountries?

Very cool though, I like the topic!
I just got some GM 352 Raw casting's Im going to explore. Lets see where this thread goes.

T-flow
 

·
Registered
Joined
·
4,798 Posts
Is that with the Solidworks flow analysis tool?
Looks like Solidworks to me, .. don't think you'll
get very much "solid" flow data from this.

You have to model the entire port and plenum, ..
then model the cylinder with the valve placed in the proper place, ..
besides putting a valve, seat etc in there, .. in order to gain
any data that might remotely identify a trend.

curtis
 

·
Registered
Joined
·
1,123 Posts
I used that Cosmos(?) a couple times but I didn't like the fact that you had to close both ends of the device. You couldn't just have an open ended thing thing that you were trying to test.
 

·
Registered
Joined
·
7,271 Posts
Even though these are not useful for anything in the real world notice in the first model the velocity gradient looks goofy around the ssr and the lsr. May have the wrong sign? Something to watch for as you refine your models.
 

·
Registered
Joined
·
396 Posts
Cosmos is cool to get an introductory idea, but its pretty worthless for actual, useful data. Its kind of a joke in the CFD world. CFD requires a lot of post-processing and analysis, which can't be performed in Cosmos. In addition, the last time I used Cosmos (several years ago) there was no way to check residuals for convergence


There's way too many other effects going on to just look at a velocity gradient in a chamber, a cylinder head is not just air moving through a duct (as we all know, flow numbers for a cylinder head don't tell the whole story). Choked flow at the valve, the flow is compressible, shock waves in the intake runners and bouncing off the back of the intake valve, sonic/supersonic flow in the exhaust, and on and on....
 

·
Registered
Joined
·
7,271 Posts
Cosmos is cool to get an introductory idea, but its pretty worthless for actual, useful data. Its kind of a joke in the CFD world. CFD requires a lot of post-processing and analysis, which can't be performed in Cosmos. In addition, the last time I used Cosmos (several years ago) there was no way to check residuals for convergence


There's way too many other effects going on to just look at a velocity gradient in a chamber, a cylinder head is not just air moving through a duct (as we all know, flow numbers for a cylinder head don't tell the whole story). Choked flow at the valve, the flow is compressible, shock waves in the intake runners and bouncing off the back of the intake valve, sonic/supersonic flow in the exhaust, and on and on....
Fluent....
 

·
Registered
Joined
·
396 Posts
Fluent is great, Star-CCM+ is also, just quirky.

All I was saying is Cosmos leaves a lot to be desired and there's a lot of other factors that have to be looked at beyond what the OP did....
 
1 - 13 of 13 Posts
Top