Improved Tundra-Jet Valve Array

[Bruce]

One of the important things to consider when designing the valve retainer is that the reed material should not be allowed to deform beyond its elastic limits.

If it is allowed to deform beyond theses limits then it will fail very quickly for several reasons:

1. Metal fatigue will cause cracks
2. Permanent deformation (curling/bending) will affect the seal against the valveplate
3. The force with which the materials natural spring holds the valve against the plate will be changed (and hence the open/close times.

It's amazing what a difference just a mm or two can make in respect to the amount of "lift" on a valve and the life of that valve.

For maximum valve life, the retainer should always be a curve, not just a flat plate with a bend at the root.

The exact shape of that curve is something that depends on the shape of the valve leaf itself and the shape of the opening it covers. Such things are generally not worth calculating for a hobby-engine so a simple constant radius is usually adequate.

The one thing you need to check with that engine is that you have enough valve area compared to the tailpipe area. Remember that a valve-grid array is a whole lot more efficient than the 90-degree valve setup you're now looking at.

To work out your effective valve area with the new setup is not very simple but you can probably say (using a rule of thumb) that it's equal to somewhere betwen 0.3 and 0.7 of the area of the holes in your valveplate -- depending on things such as the amount of valve-lift, the exact size of the holes (reynolds numbers), etc.

To be safe, I'd want to have inlet holes of at least 70% the area of the tailpipe -- preferably more because you can always choke down the total inlet size with a venturi on the front if you need to.

[larry cottrill]

Gentlemen -

There is another approach, taken in the venerable Dynajet, that works well, at least in fairly small size valves: I believe that the retainer is NOT designed to take the reed hitting 'all at once', but rather that the valves 'roll out' onto the retainer without kinking, and with the bending zone gradually shifting. You may remember me picturing this a year or so ago with this drawing [the flexure is exaggerated for clarity].

L Cottrill
----------

[Bruce]

Indeed, this is the way the retainer should work. If the valve were to hit all areas of the retainer simultaneously you'd get enormous additional stresses set up because the air between the two would not exit cleanly but form little pockets of high pressure where it was trapped.

This is the same phenomenon that produces splitting of the valve-edges when there's too much overlap on the valve seat. The goal is to promote smooth (yet rapid) evacuation of all the air between the two surfaces -- this is best achieved when they contact at a reasonable angle.

[Mike Kirney]

Thanks for you commentaries, Bruce and Larry. I'm going to try running the jet with the reeds just pop-riveted in place at first and see how it goes. Except for the absence of the valve retainer, my valve design pretty much follows Bruce's advice. If a valve retainer is required, I think I will try to find some half-round rod and rivet that in along with a new set of reeds. I've made all the reeds and I'm about half-finished making the valveplate. This 16 ga. steel I'm using is much harder to work with than the 22 ga. I was using before, but I think it will help make a good seal between the duct and the valveplate when everything is bolted down. I won't start thinking about a gasket unless the first test goes unsuccessfully.

[Bruce]

You're definitely going to need a retainer (movement-limiter) of some kind or those valves will fail in a few short seconds (probably just a single-digit number).

The problem is that, left to their own devices, the valves will tend to bend right at the root -- an area that will already be weakened by the hole for the rivet.

If these valves have any kind of "spring" to them and you drilled that hole with a regular twist-drill, there will already be microcracks around the hole and one (or more) of those cracks will propogate right through the valve causing fracturing.

Remember that, left to their own devices (unlimited in movement), the valves will tend to "flutter" violently as the air flows around them - mainly due to the karman vortex that forms over the edge.

One of the jobs of a good retainer/restricter is to limit the effect this vortex has on the valve movement when open.

There's a lot of science behind something as seemingly simple as a reed valve and its design.

[Mike Kirney]

Now I'm thinking I will put a strip of flat 22 ga. steel down the rivet line at the back of the valves before i fasten them in place. I could even bend up a semi-circular piece of 'channel' to mimic the curved washer in a petal-valve system. How about this idea - what if I riveted the reeds in place, and then brazed between the rivets to the edge of the reeds, so that bending stress would be spread across the width of the reeds, instead of being concentrated around the two little holes? It's not hard to melt away and then re-apply the bronze when reed-changing time comes.

[Mark]

You're definitely going to need a retainer (movement-limiter) of some kind or those valves will fail in a few short seconds (probably just a single-digit number).

The problem is that, left to their own devices, the valves will tend to bend right at the root -- an area that will already be weakened by the hole for the rivet.

If these valves have any kind of "spring" to them and you drilled that hole with a regular twist-drill, there will already be microcracks around the hole and one (or more) of those cracks will propogate right through the valve causing fracturing.

Remember that, left to their own devices (unlimited in movement), the valves will tend to "flutter" violently as the air flows around them - mainly due to the karman vortex that forms over the edge.

One of the jobs of a good retainer/restricter is to limit the effect this vortex has on the valve movement when open.

There's a lot of science behind something as seemingly simple as a reed valve and its design.

I have several different kinds of pyramid reeds off of boat motors and other two stroke engines. Often, right at the base of the pyramid where the reeds are held down, just a tiny, tiny distance from the attachment point there is a tiny channel running the length of the base where the reeds all are pinned down. Yet the flex point seems to be further up so that this is something perplexing to me. It must provide for something but what? The channel is perhaps just 2 mm wide and very shallow. Perhaps it is an air buffer or flex/strain relief but the bending point of the reeds takes place further away and it is hard to imagine the little ditch is critical, but I see this design repeating in the various pyramid combinations, from a two petal per side or larger reed cages of 5 reeds per side. The way the retainers hold the reeds, it presses the reed flat against a flat surface, again, up from the minuscule ditch which is perhaps as wide as this letter n and about the same in depth but just about like this letter v for example, a v groove . Now imagine the .010ths resting on top of the v. How much flexing would this afford? Somehow someone discovered something here.
Mark

[Ivar]

by the way Troy, I assume your jumbojet project is forgotten.

[Mark]

Gentlemen -

There is another approach, taken in the venerable Dynajet, that works well, at least in fairly small size valves: I believe that the retainer is NOT designed to take the reed hitting 'all at once', but rather that the valves 'roll out' onto the retainer without kinking, and with the bending zone gradually shifting. You may remember me picturing this a year or so ago with this drawing [the flexure is exaggerated for clarity].

L Cottrill
----------

Yet, to break the mold, I have some harmonica-looking bank of reeds off a Johnson outboard I think it was, and instead of the French curved retainers on either side of the pyramid bank, they hover over the length of the reed and then at the tip they, (the retainers), dip down with a lip. So no curvature is used only a rounded small lip to stop the tips of the reeds from swinging open further. Strange but true, they also honk nicely if you blow into the harmonica-like cage. Basically they look like typical reed banks like a V-1 yet with this twist in philosophy.
Mark

[Troy R. Legner]

by the way Troy, I assume your jumbojet project is forgotten.



I think about it from time to time. I do not have much time to research it. I ran into fuel system design problems. The valve grid was very nice.
Troy L.

[Hank]

Hello- Looks like a winner.
As Larry mentioned make sure you've got enough effective intake area.
I used 40% of frontal area for my first engine and the duct looked like a well used tube of toothpaste after a few seconds. (I have a 70% Radial Valve array in the works now.) Using thin metal (.025) didn't help much, either. The next duct is 19 gauge (.042) 310 SS.
You mention that you were not going to use a valve stop. Please do.
Yours looks like a simple, effective design. Have you run it yet, Tundraman?

Hank