improving flow for the 3FE’s top end (17 Viewers)

[RockDoc]

Okay, time to get back to some turd-polishing updates.

The plaster cast of #4 is done. It turned out well enough to get a nice cross-sectional view of the valve area, but I don't think it will be good enough for flow testing (especially since the ever important valve guide and guide boss area didn't hold up to removal of the mould). So I'll show the pics of that, and give casting one more try. If it doesn't work with a change in the geometry of the pourings, then I'll likely just go straight to working on the head. I'm actually pretty happy with the general shape of the stock casting anyway, so most attention will focus on the guide/boss area, and cleaning up the contacts between casted surfaces, machined surfaces and inserts.

So the cast...

Here's a pic of the 3-pour block before separating it. For the middle cast I tried adding dryer lint to the plaster to give it some more strength. It did that, but in the end, it wasn't needed, and seems to have somewhat hindered the ability of the plaster to take the detail of the mould.

Some shots of the intake side. Full size, valve closed, halfway to full lift, and at full lift (~0.377"). The valve guide should impose into the port/bowl transition area, but didn't hold up to the removal of the mould.

The same, closer view of the valve area (there is some distortion in the photos from the short focal distance)

Now a couple sketches of the intake valve area at half and full lift. You can see that there is a very sharp edge on the short radius that will lead to flow separation from the floor of the port as it transitions into the bowl area. This will need to be altered to smooth things out and keep the flow attached. You can also see that there is no valve masking at partial lift, but there is some on the short side of the chamber at full lift (would be more dramatic with a high-lift cam).

Now some pics of the exhaust valve side. Not so concerned with this side, as exhaust pressure should go a long way in overcoming any blips in the flow path. Things will be generally cleaned up, and polished to try to keep heat in the exhaust and out of the head. (ditto for the chamber)

Broad view:

Close view:

More color commentary to come when I need my next break from drafting lithologic logs.

 

[rover67]

this looks like a really fun project... i'll continue to watch!! great work!!

 

[RockDoc]

this looks like a really fun project... i'll continue to watch!! great work!!

Thanks

 

[RockDoc]

Okay, some semi-random thoughts to drop in the thread. A bit of clarification of what I want out of this project, and some observations on the stock fitment of things.
In no particular order:

1) I'm going to assume that since I'm working with a fuel injected set-up, the fuel in the intake charge will be sufficiently atomized so that the fuel will behave as an essentially passive component of the intake charge rather than a separate phase that must be coaxed to stay mixed with the charge. This will negate a few things Vizard suggests to help shear deposited fuel off the walls of the runners and chamber that would work against optimum air flow.

2) The head does appear to be a "down onto the valve" type as opposed to an "across the valve" type (referring to the path that the bulk of the intake charge takes in making its way through the bowl and past the valve head). In the one cylinder checked so far, the flow seems to be slightly bias towards the "near" side of the valve. This *may* change some with streamlining the valve guide and boss, and using valves with an undercut stem.

3) On of the tricks Dalton suggests is the use of "anti-reversal ratchets" in the intake and exhaust tracts to prevent reverse flow. These "ratchets" pretty much negate the possibility of pressure wave tuning.... but are supposed to provide for a wider power band, and are offset the negative effects that a longer duration cam has on low rpm performance. (see images below)

Taking a good look at the relative sizes of the port openings on the head and the manifolds, it is apparent that the stock set-up already has a weak form of such ratchets. The down-flow side of the interface (head for intake, manifold for exhaust) is slightly larger at the mating surface, then constricts back down to about the size of the port on the up-flow side. I suspect this was done to ensure that slight misalignments don't provide an impediment to flow, but likely offers a bit of benefit in the form of these anti-reversal ratchets.

The port size in the gasket is very tight for the intake valves (i.e. you can't open the ports up, as there is no extra room in the gasket opening), but offers lots of working room with the exhaust ports.

4) Removing the vacuum bag gives the flow bench much more suction (vacuum and flow) at the test face, much more than I was anticipating. Adding a touch of dish soap the the manometer fluid to break the surface tension makes it behave much better inside the tubes.

5) When the mods to the head are done, I'll have the head shaved an amount that will allow it to be suitably used on either a 3F or 2F block (future 2FE build?).

6) The head will be installed with the stock cam, but some of the features of the build will help to maintain low rpm performance if I bump the cam later. (like the ratchets) I'll try to make sure that flow is good a bit past stock lift too.

7) When I have to compromise between peak power and usable low rpm power, I'll err on the side of low rpm power.

8) At this point, I don't think I'll play around with turning valves, I'll just for with the new SBC stainless ones, and have them ground optimize flow over/around them. (3 angle grind on the seat, face, top-cut and back-cut on the valves)

Some estimated improvement numbers as suggested in Vizard's books:

Bumping compression from 8:1 to 9:1 should yield ~2% power increase. The 3FE starts at 8.1:1, and I don't think I'll bump to quite to 9:1 given the short duration of the stock cam

A respectable port, polish and valve grind job should yield 5-10% power increase. (I'm hoping this is a bit conservative given that the 3FE head isn't really a competition or hi-po head, and may have more room for improvement)

A cold air intake set-up can yield up to 9%, depending on the stock set-up you are improving on (I suspect the stock 3FE/62 set-up is actually pretty good, so I expect something lower would be the case here)

Cleaned and "balanced" injectors, working over the AFM and throttle body - ???? Others here have reported that the throttle body overbore really make things come alive. I hope that's true.

Over all, I think I'll be happy if I get something like 15%, and notice better performance in the 2000-2500 rpm range, as that seems to be what's missing for daily driving duties.

Anyway, enough yakking for now.

Basic idea of the anti-reversal ratchets. “A – For the moment ignore the changes in pressure and the anti-reversal effect and imagine the gas flowing at a fairly high speed. Notice the shape of the streamlines and how at (a) the effective port diameter is slightly smaller than at (b). The boundary layer at (a) is at lower than mean pressure and possibly turbulent, however, the main flow through the centre is slightly faster than you would expect and if happening at the right time this could be more useful as it keep the gas at high speed. Now slow the flow down as at B and the effect is more pronounced and even better as it can give you more of a usable rev range. This of course only really happens when the throttle is wide open, for example when accelerating. But the really interesting thing happens when we get to C: if the gas speed is high enough, the boundary layer thins down an a form of shear takes place with the gas flowing straight over the ratchet leaving the pocket relatively stable and therefore not restricting flow. If you take it a step further and actually introduce a swirl as in D the effect becomes even more pronounced at lower speeds. This can give you a much more tractable engine or allow even more radical cams. Blah blah blah trade secrets to be withheld and such”

The Dalton-proposed layout of anti-reversal ratchets for the exhaust tract. (a) on the back of the valve, (b) just behind the seat insert, (c) just inside the port, (d) utilizing the mating surface between the head and manifold/header. I plan to nix (a), perhaps cut a minimal one at (b), and work them in as at (c) and (d). For the intake, I'm thinking one at the manifold/head interface, and one where the two pieces of the manifold meet.

Possible ways to situate a ratchet at the exhaust port with different manifold/header constructions.

 

[Randy88FJ62]

Could you explain the first figure in post number 104 in greater detail. I don't understand the different turbulent areas in the 4 comparisions and what's driving the differences.
-Randy

 

[RockDoc]

Could you explain the first figure in post number 104 in greater detail. I don't understand the different turbulent areas in the 4 comparisions and what's driving the differences.
-Randy

The caption given is what Dalton gives with the figure, mind you with out the bulk of text, it may be a missing a bit of background explanation. If the flow is analogous to water flow in a channel, or over a rough stream bed (what I am more familiar with) then the turbulent areas represent zones of flow separation, where the intake flow (or exhaust) has pulled away from the boundary. I assume those turbulent areas would have a back flow circulation (as would be the case with water flow), but the main point I get out of it is that depending on the rate of the intake flow, the degree of flow separation changes.

With low flow (low rpm, cases A and B), the flow separation increases and squeezes the flow into a smaller cross-section, this mimmicks the effect that a smaller runner/port would have in increasing velocity of flow for a given rate/"discharge". By bumping up the velocity of flow at low rpm, you aid the ram effect for squeezing as much intake charge as possible into the chamber as the valve is closing and the piston is turning around into the compression stroke.

When the rate of intake flow is high (higher rpm, case C) the flow shears across the "anti-reversal ratchet" with little disturbance. As a result, you (apparently) get full flow characteristics that don't deviate from a smooth-wall case. When you need the full cross-sectional area of the runner/port, you get it.

Case D doesn't get much explanation from Dalton. He basically says that by cutting the downstream end of the ratchet to have a slight wall or lip, rather than an angular interface as with the other sketches, you exaggerate the constriction of the flow to get even higher velocities. This, he says, opens up the tantalizing possibility of "pressure wave tuning", which he doesn't expand on, saying he has to hold on to some trade secrets. If I was to guess, he's talking about flow going super-critical, where the flow is so fast in that squeezed zone, that pressure communication is lost between up-flow and down-flow. This would mean that upstream flow doesn't start to decline until after the constriction effect has collapsed, and the valve is already closed. As a result, you would get very high pressure in the valve bowl as the valve is closing. Again, this is based on my understanding of water behavior though, and not so much an understanding of gas flow.

Hope this helps.

 

[Randy88FJ62]

No one has ever come up with a perfect equation that explains turbulent flow. That’s why correlations are used based on Reynold’s Number and other non-dimensional variables. I think the thought of using reversal ratchets for preventing reverse flow is interesting. I would expect this effect to be used to mix the air fuel mixture one last time before entering the intake. All I see these chamfered indents doing is creating different turbulent flow patterns. I think that a good FEA program could knock out some theoretical flows and pressure differentials which would help you determine the geometry of these ratchets and proper application. I’m tempted to use the floworks program that comes with Solidworks. I think improving surface finishes would give better payoff than these ratchets but I know nothing about engine flow dynamics. This conversation is quickly pushing my limits on theory.

Thanks for the explanation RocDoc.

 

[Randy88FJ62]

I started talking to one of the engineers at work about ratcheting and he made the comment that ratcheting at the exhaust port is useless unless you race and the engine is coming apart a lot. He said that carbon buildup would fill the ratchets and negate their purpose unless cleaned on a regular basis.

He agreed with the ratcheting of the intake ports stating that it would prevent reverse flow as you mentioned above.

 

[RockDoc]

I started talking to one of the engineers at work about ratcheting and he made the comment that ratcheting at the exhaust port is useless unless you race and the engine is coming apart a lot. He said that carbon buildup would fill the ratchets and negate their purpose unless cleaned on a regular basis.

He agreed with the ratcheting of the intake ports stating that it would prevent reverse flow as you mentioned above.

Cool, there was a lot of carbon built up on surfaces when I took the head apart (but surprisingly it was the intake valves that had more build up on them, back to page one for a look-see). Dalton's book is more tailored for getting the most out of 4 bangers, and I suspect he was geared towards racing applications where tear-down would be common. Maybe not worth while to cut the ratchets inside the port, but it can't hurt to do it at the manifold mating surface. It's a pretty standard idea there I believe. Would polishing the exhaust ports be effective enough in keeping carbon off? Or are you still going to get buildup?

 

[Randy88FJ62]

I asked him about polishing the headers and he said that carbon buildup would negate that quickly. I think your best bet is to ratchet only where carbon buildup is not prone.

 

[RockDoc]

Gotcha, hopefully the deletion of the EGR will keep the intake clean. I should have taken more pic of the head before cleaning to get an idea of how much build up there was.

 

[iaintscared1969]

Teaser shot

Teaser shot for Mr. RockDoc

The new TB size is 74 mm with new seals and butterfly. Should make it scream...

 

[RockDoc]

Teaser shot for Mr. RockDoc

The new TB size is 74 mm with new seals and butterfly. Should make it scream...

Is it all ready to go? Getting that and the other goodies might be a good kick in the pants to get working again. Things have gotten busy at school again, and I haven't been getting much time in on the project. I made a second attempt at a plaster model, but it didn't turn out much better than the first, so I think I'll skip playing with a model and go straight to a moderate work over of the metal. Maybe this weekend I'll set aside the time to do all the before measurements on the flow bench and start working the metal.

 

[iaintscared1969]

Better get back to work Doc! Had a mix up on the 3809 but the correct ones should be here Fri. or Monday. Then I am shipping.

 

[Tinker]

I'm trying to muster the initiative to read this whole thread, but in the meantime may I make a really dumb statement? So just sending your head & manifolds to the Extrude Hone people is not the final answer: they can do flow but anti-reversion is outside the scope of abrasive sludge machining.

 

[FJ40Jim]

Extrude hone is good as far as it goes. But hydraulic sludge is not intelligent. The sludge doesn't ask "would it be better to remove material here or over there". It just removes material everywhere, especially from the most obvious restrictions. It does not take into account the cylinder bore or partially open valve effects.

My DD has an extrude honed intake, because the design of that manifold precluded getting inside it w/ a die grinder. For the easily accessible ports in the Cruiser head, might as well get after it w/ a carbide burr.

 

[ntsqd]

I'll echo the same, put the shape in the port that is required or desired, then Extrude Hone it if you want. The nature of Extrude Hone is to work the hardest on the most obstructive points and the least hard on the least obstructive points. Electro-polishing on a much larger material removal scale.
I would not Extrude Hone a wet flow port, only a dry flow.

 

[Moby]

I have a ported throttlebody. Not sure that it is of value on an otherwise stock engine. But with a whole package of mods (head porting, backcut valves, upgraded intake and exhaust...) it probably makes sense.

 

[Tinker]

Jim - What's your DD? And do you think the Extrude Hone did the job?

NT - By wet flow do you mean after the carb/injector?

 

[ntsqd]

Tinker,
Yes, exactly.