How would you treat this?

[Vicoor]

This is a picture showing what I will refer to as the short side "corner" because as you can see there really is no discernable radius.


this is an exhaust port on a 22RE

My question is how would you treat this?

[Vicoor]

Flyin' by the seat of my pants

Gosh, I know it's ugly, but I can't believe you folks haven't thrown something out yet.

So far what I've done is built up with putty about 1/4" thick so that I could create a radius. this helps flow through the .200"-.300" lift range .400" is about the same and .500" is about 8% less.

I think it drops off at higher lifts because of the reduction in cross sectional area.

I'm thinking that next I would want to open up the port roof to get back the cross section and that once I'm there the flow should be improved across the board?

[gofaster]

You are on the right track. Raising the roof is the next logical step. along with making a nice radius on each side of the guide boss. Start at the "valve side" of the guide and work out with progressively larger diameter cutters (or stones, or cartridge rolls) to make the largest and smoothest radius practical. Make the guide boss smooth and round, not angular or sharp. When you work on the roof, keep in mind the thickness of the head under the valve spring seat. It's no fun to see sunshine through your roof!
You can also widen the port in this area if it won't cause you to hit water, oil, or pushrod openings. Keep in mind that you want all transitions in the port shape to be gradual, no sudden wide or deep areas. They cause turbulence which effectively restricts flow. Air does not want to turn, but it can be coaxed to follow your port if your shapes are right.
Here's a link (not the one I was looking for, but it'll do) that demonstrates what is happening at the short turn, and pretty much anywhere you have a direction change, bump, etc. in the port.


It looks like you have fairly heavy pressed in guides, similar to the engines I work on. You can push the guide out, and cut the guide boss down to a thickness equal to the guide diameter before you "teardrop" it (the boss). Before you put new guides in, you can turn a taper on the intake guide, and blunt the nose of the exhaust guide. There's too much heat in the exhaust port to use a fully tapered guide, it may burn away.

P.S. :

"Gosh, I know it's ugly, but I can't believe you folks haven't thrown something out yet."

Patience! you may not get an answer to a post for days or weeks, people on this forum are busy with their own problems, but they always come around the forum to check in when they have time. When it comes to dispensing information, this is one of the most generous groups on the net.



Edited By gofaster on 1184160230

[Vicoor]

Here's a picture of how I filled in behind the valve seat with epoxy.


I also worked on the combustion chamber some and smoothed out around the guide.as well as raising the port roof just slightly. Of course I don't have pics of those details.

all of these changes has brought the flow up from 67% to 83% on my sloped manometer scale. About 24% increase. This is in line with my expectations.

[Vicoor]

OK,
I've ben looking at the link you posted on the COANDA effect.

What I see that could be applicable is that the air flow seems to "cling" to the surface of the curved piece. I imagine that if that basic shape were used on the short side radius then it would help to prevent turbulence?

Does this apply to any airspeed? What about supersonic? (like I'll ever get there)

[Vicoor]

Here's before and after shots from the chamber side





Critique?

[Greg]

I did a fair bit of work on one of these last year and it did produce some serious power, it was also turboed. I used a set of ebay stainless o/s valves in it which moves the seat out a bit and gives you a bit more to play with in regard to the short turn. Is that epoxy in the exhaust port? I still haven't seen anything that will stand up to the heat.

Have you had a look at a 20R head? they are apparently a lot better head but not readily available in Australia so I gave up on the idea. The engine ended up making over 400hp at the wheels through an auto so it is possible to get them working ok. I had some flow figures but the flowbench computer blew up so I cant drag them out for you.

[gofaster]

You are definitely doing the right things. If you have time to search the forum, type in coanda, and see what pops up. If you don't find more info here, try the same search on this site:


On one of these forums, there is a link to a great live video clip of a demonstration of the coanda effect in action. It is very enlightening. You'll see that the curve will have the effect of pulling the air around the corner.

[Vicoor]