Just Another Valveless Front End

Moderator: Mike Everman

Bruce
Posts: 95
Joined: Fri Oct 17, 2003 10:47 am

Re: something similar

Post by Bruce » Fri Nov 07, 2003 9:55 am

[quote="Mike Everman"]Here's something along the same lines that might be machinable in mirrored halves, maybe a chance at a mini since the NC contours will be so smooth.
got sidetracked.[/quote]

Just to prove that great minds think alike (although they also say that fools seldom differ ;-) your latest concept looks just like half the coanda-effect engine I proposed at: http://aardvark.co.nz/pjet/coanda.shtml

Bruno Ogorelec
Posts: 3542
Joined: Sat Oct 04, 2003 7:31 am
Antipspambot question: 0
Location: Zagreb, Croatia

Post by Bruno Ogorelec » Fri Nov 07, 2003 12:32 pm

Ha-ha-ha.... Right you are.

Mark 'Thixis' once (or maybe even often) said that there's nothing new in the pulsejet field -- everything you can think of has already been proposed by someone or other between 1905 and 1995. I have searched through the history of the field pretty extensively and I can say that I almost agree.

The real scope for innovation may well be in actually building the damned things. It is amazing how few pulsejets (especially valveless ones) have actually been built. Fewer still have been tinkered with with any kind of determination, systematic approach and honest practical engineering. I think the total count is about two -- Lockwood/Ecrevisse and Escopette.

These engines are very. very sensitive to small details, meaning that they probably offer great scope for fine tuning. But, practically no one does that.

To give you an example -- the Cosworth DFV engine formerly very popular in Formula 1 Grand Prix racing developed about 320 HP when it first appeared in 1966. By the time it was phased out of Formula 1 more than 15 years later, it was developing well over 600 HP -- yet still looked basically the same, even internally. At that point it was supercharged and went racing in the Indy CART series and its powered almost doubled and it served for another decade. Still, if you dismantled one of teh first and one of teh last and compared the components. they did not look that terribly different.

Some stories of 2-stroke engine development are at least equally amazing.

There a complete absence of such stories in the valveless pulsejet field -- with the partial exceptions of the two I mentioned. Who knows how far a valveless pulsejet can be pushed if you approached it with determination, flair and some money.

Bruno

Mike Everman
Posts: 4934
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

Post by Mike Everman » Fri Nov 07, 2003 3:15 pm

"fools seldom differ" (LOL) Glad you showed up, Mr. Simpson! that layout should look like your Coanda design, your design sparked me on all of this a few weeks ago. My problem with the cylindrical implimentation was I thought that the forward exhaust would be disinclined to crunch down and out the central pipe. I am having a great deal of fun with 2d sections in rectangular construction.
My researches range widely, and while the old saw that "there's nothing new under the sun" is amusing, it's rarely (ulimately) true in any field. I do however live in the fear that as my memory capacity diminishes, each "new" idea I have is merely something I saw once. Fortunately I can remember to last week. I've also been watching Larry Cottril's attempts with great interest.
Bruno, rotor tip jets is why I've found myself here, with smaller units as pseudo turbo-props for ultralights the goal; though now the focus is on making some noise and thrust, and perhaps, if I'm very lucky, making an effeciently running engine.
Mike

Stuart
Posts: 83
Joined: Fri Oct 24, 2003 9:35 pm

Post by Stuart » Fri Nov 07, 2003 3:43 pm

Bruno says - "No velocity. No velocity head"

That isn't true Bruno. In fluid flow there are a couple types of "head": Velocity, PRESSURE, enthalpic, and gravitational. A momentarily stagnated adiabatic flow still has at least a pressure head - which translates back into a velocity head.
I'm writing an automated airplane designer in java, useful later when you guys get ready to bolt a p-jet onto some wings

Bruno Ogorelec
Posts: 3542
Joined: Sat Oct 04, 2003 7:31 am
Antipspambot question: 0
Location: Zagreb, Croatia

stop-go-stop

Post by Bruno Ogorelec » Fri Nov 07, 2003 4:22 pm

Hmmmm.... Stuart, we could argue about this, I guess but it makes little sense. I am not an expert in the field. I'd get lost soon.

I am not saying that there is no 'head'. Some of the velocity head gets converted into static pressure, which is a different kind of 'head'. I am saying that there is no 'velocity'.

The potential energy of the air coming in gets converted into something else in the engine no matter what you do. Pulsejets use a conversion sequence that is slightly different from the sequences found in other jet engines.

One of the unique aspects is that fresh air loses almost all velocity before being ingested. It has to be re-accelerated once the intake commences. That is not true for other jet engines.

(It gets stopped once more before ignition -- another unique feature -- but that's another story.)

The point is that the streamlining of intake passage sections need not follow the same logic as in other jet engines. For instance, a right-angle turn in an intake passage is an anathema in all jet engines, but not necessarily in a pulsejet. (Of course, the right angle must be at the right place.)

If a fast air stream is braked to a stop (as it is in a pulsejet intake) it does not matter in which turn it will be re-accelerated again. It can be drawn to the side, or even backwards, at the same cost in energy.

This is an over-simplification, of course, but true enough for simple discussion.

BTW. the same goes for the exhaust. When combustion commences, the charge in the combustion chamber is at rest, statistically speaking. Products of combustion get accelerated from standstill. Again, that means that the direction in which the exhaust tube is pointing does not matter. It does not matter whether the intake and exhaust directions are aligned or not. The pretty streamlined lines of pulsejet models are irrelevant to engine efficiency. They only matter in terms of outside air drag.

Mike Everman
Posts: 4934
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

Adjusted design

Post by Mike Everman » Sat Nov 08, 2003 2:14 am

OK, it now has all the proper areas and lengths of segment to be equivalent to a Kentfield with 4" dia combustion chamber. Inlet lengths got us out of the vortex issue as now they cross.
I'll list some dimensions soon, have to go see Matrix 3 in two minutes.
Best Regards,
Mike
Attachments
concept3.jpg
real proportions
(20.96 KiB) Downloaded 797 times

Bruce
Posts: 95
Joined: Fri Oct 17, 2003 10:47 am

Re: Adjusted design

Post by Bruce » Sat Nov 08, 2003 3:31 am

[quote="Mike Everman"]OK, it now has all the proper areas and lengths of segment to be equivalent to a Kentfield with 4" dia combustion chamber. Inlet lengths got us out of the vortex issue as now they cross.
I'll list some dimensions soon, have to go see Matrix 3 in two minutes.
Best Regards,
Mike[/quote]

There are a few points to watch with this design:

1. The fresh intake charge will likely be a lot warmer than it really ought to be -- partly because there will be some rebreathing of previously expelled gasses and partly because one side of the intake tubes doubles as the compbustion chamber. Ideally, with a pulsejet you want the intake charge to remain as cold as possible until the air/fuel is ignited.

2. As drawn the fresh intake charge enters the combustion chamber at points very close to the exhaust point. This means that there could well be a significant "dead area" of gas stagnation between that point and the front of the chamber. Both the Lockwood and Kentfield engines benefit significantly from having the intake/exhaust flows at opposite ends of the chamber and one only has to compare their performance with the trusty old Chinese valveless to see that this contributes to efficiency and power.

3. The way the ducting is currently designed, the effective lengths of the intake tubes will be *longer* than their actual length -- this will result in a loss of efficiency and power.

However, none of these problems is insurmountable so I suggest you carry on -- well done!

Mike Everman
Posts: 4934
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

Post by Mike Everman » Sat Nov 08, 2003 6:59 am

1.)
I agree, Bruce, I don't like the pre-heating of the incoming charge either, it's bothered me on all I've seen and the way it panned out, I've created excessive surface area there to get the proper(?) augmentation, that is my guess is the augmenter entry to be around 3x the cross jet area. That's a PADOMA (pulled it directly outta my well, you know.)

As I look at it, there is some cooling of the intakes going going on, the sides of them being part of an augmenter, after all. Looks like an advantage over a plain Kentfield or any other valveless's pipes which are statically cooled only convectively. I'm not sure I've seen any of the designs force-cooling the intake pipes, so actually the more I look at it, the more defensible it becomes. I'll chew on your observation a bit more.

2.)
I see what you mean, I'll attempt to move the intake flow forward, or barring that turn it more than 90deg and direct it forward; very good point.

3.)
I designed the intake pipe length to be 5.3" at the neutral axis, the recommended length if straight. I hope what you say is true, I'd like to make their length shorter to help with #2 above. Are you talking about turbulence in the bends making them seem longer?

I mentioned before about the augmenter side being one side of the CC, with thermal expansion of the augmenter flow a hoped-for addition to thrust. Anyone have thoughts on whether this augmenter path should be constantly diverging or straight or what?

Mike

Mark
Posts: 10813
Joined: Sat Oct 11, 2003 10:14 pm

Post by Mark » Sat Nov 08, 2003 10:29 am

Mike Everman wrote:1.)
I agree, Bruce, I don't like the pre-heating of the incoming charge either, it's bothered me on all I've seen and the way it panned out, I've created excessive surface area there to get the proper(?) augmentation, that is my guess is the augmenter entry to be around 3x the cross jet area. That's a PADOMA (pulled it directly outta my well, you know.)

As I look at it, there is some cooling of the intakes going going on, the sides of them being part of an augmenter, after all. Looks like an advantage over a plain Kentfield or any other valveless's pipes which are statically cooled only convectively. I'm not sure I've seen any of the designs force-cooling the intake pipes, so actually the more I look at it, the more defensible it becomes. I'll chew on your observation a bit more.

2.)
I see what you mean, I'll attempt to move the intake flow forward, or barring that turn it more than 90deg and direct it forward; very good point.

3.)
I designed the intake pipe length to be 5.3" at the neutral axis, the recommended length if straight. I hope what you say is true, I'd like to make their length shorter to help with #2 above. Are you talking about turbulence in the bends making them seem longer?

I mentioned before about the augmenter side being one side of the CC, with thermal expansion of the augmenter flow a hoped-for addition to thrust. Anyone have thoughts on whether this augmenter path should be constantly diverging or straight or what?

Mike
All I know is that the more variables you introduce, the difficulty becomes exponential in tracking down which is the way out of the maze you've gotten yourself into. Changing one variable on a pulsejet can often change several other things you planned on being a constant or given. I would start out using as few basic shapes as possible in your design, building something complex straight off is a big "time" risk. It would be a lot of work for a long shot.
A lot of unexpected things can happen, metal can buckle, and the distinctive resonant combustion we all know and love can shadow you, I've been so close to something working without even knowing it until months or years later, but unless you know it is there, you might walk right past your lurking phenomenon without knowing it, just as mankind missed discovering it many times in the distant past. A large body of knowledge and a paramount amount of patience is required to tackle big problems that are introduced when you impede the airflow too much or dampen feedback loops from too draggy a design. What channel often seems good or fine in one direction, exacts a penalty of equal degree in the other reverse air flow half cycle.
Look at a ramjets, they don't have any choice but to be kept "simple" and oh so streamline, lest drag rear its ugly head.
Mark

Mike Everman
Posts: 4934
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

simplicity

Post by Mike Everman » Sat Nov 08, 2003 3:07 pm

Mark-
Isn't that the way? I have always said (quoted?) that simplicity is the hardest won aspect of any system. Simplicity isn't just less parts, but less variables. Just yesterday I was conducting a design review with two of my engineers and it turned in to that very discussion: minimizing variables. The system in question is simple, but even a few little things can can add up to a huge probable cause matrix when it starts acting strange.
I am obviously lightyears away from the engine's variables being reduced to fuel flow, but that's the lofty goal.
I am going to have to experiment with these intake pipes; need to gage mass flow when cc pressure is posetive and negative, because it should be obvious to everyone that moving from straight round pipes to rectangular with two bends (all welded together, with only area and length considered) has got to be a recipe for grief. And the flat sidewalls are going to require greater thickness that I would prefer, and may have detrimental or beneficial vibration modes that slightly change the volume, and not necessarily at the combustion frequency! >:-O
mike

Dave
Posts: 227
Joined: Sat Nov 08, 2003 7:03 pm
Antipspambot question: 0
Location: Buffalo, NY

RE: Simplicity

Post by Dave » Sat Nov 08, 2003 8:04 pm

Mike

How about a simple change to the shape of your intake tubes?

Your last design used a shape something like the letter "C". That puts the entry point into the combustion chamber on the side and fairly far back.

By changing the intake tube shape to something that looks more like a question mark "?" the entry into the combustion chamber would move to a more conventional spot, the front.

One negative is that this arrangement will create some dead space to fill if smooth gas flow is to be maintained around the outside of the intakes before reaching the combustion chamber. Sorry, I do not have adequate drawing tools to show the concept.

Dave

Mike Everman
Posts: 4934
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

Laminar flow bends

Post by Mike Everman » Sat Nov 08, 2003 8:59 pm

Dave,
How's this for a change in geometry? (brings my computer to it's knees, but sure looks cool.) First cut at getting the turbulence out of the turns. Path widens in the turns toward the inside of the bend to slow the flow down, matching velocity with all points outboard. In fact, I feel that any turning tube should be trapezoidal or at least eggish in section, with the fat side toward the inside.
Forming these transitions is going to be a b___ch, but I think once the tool is made, I can press it in halfs and seam weld.
As to the entry point, there's no reason for the CC to be tucked right under the inlets, in fact now that I think about it, we should have some space for better inlet cooling; I'll be moving it down. I continue to like the side entry for superior mixing.
I'll upload a full assy sometime today.
Mike
Attachments
constant velocity bend.jpg
(25.82 KiB) Downloaded 670 times

Mike Everman
Posts: 4934
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

Laminar flow bends

Post by Mike Everman » Sat Nov 08, 2003 10:06 pm

I don't know if I'm on the right track or whether it's of third order importance... but it's an attempt at making any path through the bent intake be the same length. It doesn't account for the natural pressure increase from inside radius to outside and it's effect on velocities from centripetal acceleration, which must be considered, or just drink some wine and watch the beautiful women walk by...
Mike

Graham C. Williams
Posts: 571
Joined: Fri Oct 03, 2003 11:33 pm
Antipspambot question: 0
Location: England
Contact:

Post by Graham C. Williams » Sun Nov 09, 2003 1:42 am

Dear Mark.
This is starting to get a long way from the simplicity of form in your initial design. To my eyes at least that design has many good points. Viv and Bruce have made some very important comments that cover the possible difficulties.
When developing something a bit different it is important to be able to locate the variables effecting performance and to be able to make simple changes to these variables at the same time suppressing the effect they may have upon the others. I think Mark has alluded to this in his comments. So, please return to the initial simplicity and develop your design keeping in mind the need to make changes without having to rebuild.

What are you intending to do at the other end of the motor - the exhaust end? A careful design of the exhaust end could have a profound influence upon the flow in the induction system.
Best Regards
Graham.

Mike Everman
Posts: 4934
Joined: Fri Oct 31, 2003 7:25 am
Antipspambot question: 0
Location: santa barbara, CA
Contact:

Post by Mike Everman » Sun Nov 09, 2003 2:13 am

Graham-
Point taken, I admit I'm a bit mired in details now that I have some dimensions to start with. The first one was a lot simpler, though a good deal out of proportion. I think I'll be taking your sound advice and re-evaluate the formed sheet intakes; perhaps I can make the proportions work, and use fasteners so this part can be tweaked and re-installed (I moved to tubes because I was concerned about sealing against the CC).
As to exhaust, I'm once again using the 4-pipe proportions, and I have not given much thought yet to the flow through the augmenter jacket. Looking for ideas there, length, taper, I'm completely over my skis there.
Thanks,
Mike

Post Reply