A New Type of Hot Air Engine

[VincentG]

For those of you following my other thread on the development of an LTD engine, this is a breakaway experiment from the results of my (admittedly limited/low tech) testing. Find that discussion here... viewtopic.php?f=1&t=5506

Due to the fact this forum does not allow post editing, and I will make mistakes/miscalculations, I will start any subsequent post with ***** if I am issuing corrections. I am not formally educated on any thermodynamic principles so feel free to chime with corrections if you are.

I had identified what I believe to be major shortcomings of the standard Stirling engine configuration. It seems to me these shortcomings are present in all three main types of Stirling engine. They are as follows:

• The hot and cold gas are never truly separated, and thus are always trying to reach equilibrium

• The displacer does not shuttle gas fast enough to reach max or min cylinder pressure when needed

• The displacer consumes power, yet offers no real efficiency gain when compared to a free piston engine

I will start by explaining my design goals, and then move on to explaining each cycle of the engine. It may seem slightly complicated, but it should be quite simple to understand if I explain it well enough.

This is simply an experiment, and as such it may be a total failure, but I have a 3d printer and cnc milling machine for hobby type use that makes projects like these relatively quick and easy.

The first prototype will be pretty rough, as I try to adhere to the philosophy of "fail fast, fail cheap". There is no sense putting lots of time into an unknown design principal when quick and dirty will get you 90% of the way there.

[VincentG]

Design goals:

• First and foremost, the engine should be constructed of readily available materials and with parts and processes that are available to the average well equipped builder

• The piston must function as the displacer and timing control device, eliminating unnecessary parts

• The hot gas must be contained until it is called upon to do work

• The "ignition" events must happen within a short crank rotation angle to provide controllable Otto cycle like "combustion"

• Use the energy normally expended on displacer movement to run a partial heat pump cycle to hopefully boost efficiency (think of a supercharger consuming power but allowing for more power overall)

• Offer a high degree of tunability, such that if the idea works, each cycle can be fine-tuned for specific use applications/peak efficiency

• Use a minimal amount of moving parts

• Take advantage of the very small amount of heat needed to raise the temperature of air

• Utilize hot air to effectively heat more air

The following will be a description of each cycle of operation based on a rough sketch. Any hard numbers written down are subject to change, I just have to start somewhere with the design. The sketch is not drawn to any proportions, it is just to provide a basic overview. Please take some time to go over all of the components to familiarize yourself with the design.

If this ends up working well, I would like the design to be fully open source. Perhaps we could all have biomass or solar fueled home power generation.

[VincentG]

Cold Stroke/Primary Compression:

The piston is on its way towards TDC, driven by atmospheric pressure after the final cycle of operation(expansion stroke). The gas will begin to compress as the piston nears impact with the poppet valve. This stage of compression will be roughly 1.5:1 on my prototype. It of course may be varied based on testing results. I believe it should be relatively low, as to not cause much heat buildup and thus resistance to this stage of compression. It should be thought of as more a displacement of air towards the hot end. The working cylinder may or may not have active cooling, depending on testing results.

compression stroke.jpg (256.38 KiB) Viewed 5965 times

[VincentG]

Air transfer to the hot air mixing chamber:

The piston has made contact with poppet valve at roughly 15 degrees BTDC. The primary compressed air is now shuttled into the preheated mixing chamber through the now exposed ports, where pressure will begin to rise. The movement of the poppet valve sealing disc has agitated the air, hopefully causing turbulence and thorough mixing. The insulated cylinder head and closed poppet valve has thus far kept the hot air contained.

The poppet valve spring pressure should exceed the maximum obtainable cylinder pressure so it does not open before the primary piston impinges upon it.

The secondary piston is nearing entry into the secondary cylinder for the next compression/ heating phase.

poppet valve opens.jpg (259.4 KiB) Viewed 5963 times

[VincentG]

Secondary compression:

The poppet valve has effectively fully opened. At roughly 10 degrees BTDC, the secondary piston has been driven into the secondary cylinder, who's I.D. is a small amount larger than the O.D. of the former. The Piston stops just thousandths of an inch short of bottoming out, causing an extreme spike in air temperature that is allowed to slip around the secondary piston and further heat air in the mixing chamber. The diameter of this piston is much smaller that the working piston, so the resistance should not be detrimental to operation, but perhaps the gains from the delayed application of superheated air will be significant.

The timing of this event can be adjusted as needed, along with the opening of the poppet valve in the first place.

I believe with this cycle, and the subsequent expansion cycle, it may be possible to self start this engine(for a short while) with no external heat source, just from rotation alone. I am, of course, not claiming perpetual motion, but utilizing the effects of a heat pump to add efficiency to the engine.

secondary compression.jpg (264.14 KiB) Viewed 5960 times

[VincentG]

The exhaust phase:

This is a key cycle to the intended operation of the engine. The timing of the exhaust port should be delayed as long as possible to extract full expansion out of the gas. The purpose of the exhaust port is two fold.

• It forces a reduction in cylinder pressure to ensure the start of the expansion stroke

• The weight of the reed valve can be tuned to hang open just long enough to function as a snifter valve to allow the engine to ingest a fresh charge of air to keep from gassing itself out

In my testing with the LTD engine, a reed valve may not even be necessary, just a tuned orifice that allows exhaust out, and a small amount of fresh air in. It must not be so large that it allows too much air in so as to limit the negative pressure provided by the expansion stroke.

exhaust.jpg (132.15 KiB) Viewed 5958 times

[VincentG]

Expansive cooling:

This is perhaps the most important phase to reach a high efficiency. The piston after traveling to BDC, has pulled a yet undetermined amount of vacuum in the cylinder, and as such has hopefully started to cool the insulated cold sink (not shown in sketch) to below ambient temperature levels. This is a parasitic phase, but I have hopes the trade off will be worth it in the form of a much higher delta T.

If a higher compression ratio is a net positive to efficiency, than maybe the same will hold true for this stage. As always, the timing and amount of crank angle devoted to this cycle can be altered. If it has no net benefit, it can be dispensed with all together and replaced with a traditional cooling system of ground water or ambient air temperature.

expansive cooling.jpg (133.03 KiB) Viewed 5957 times

[matt brown]

Atkinson 750k.png (148.5 KiB) Viewed 5953 times

Attached pic shows a low end Atkinson cycle where the expansion ratio is greater than the compression ratio. The neg work value is due to direction of rotation (of cycle) as reefer rather than engine. This site is fun, but their program is a bugger to use at times and often ends up with such limitations. In this program, the red curves are adiabats and the green green curves are isotherms. Note lower right corner indicates that these adiabats are for diatomic gas (dry air would be very similar). So, putting up with this botched direction, adiabatic compression 2>>1, isochoric heating 1>>>4, adiabatic expansion 4>>>3, and isobaric 'cooling' 3>>>2. Note that isochoric heating 1>>>4 is slightly under volume 1, but close enough to your pitched volume values.

Pt 2 is 300k and Pt 4 is 750k, but if you stare at this PV plot, you'll see that the adiabatic expansion 4>>>3 can be raised or lowered 'over' the adiabatic compression 2>>>1 and the cycle still works; it just varies the work out depending upon heat in AND a 'variable' expansion ratio.

I've spent many years scheming such Otto cycle engines, especially low end versions. Your poppet valve is what's known as a bash valve, and similar to ball & pin version in toy airplanes with disposable cartridges. Your cylinder design is commonly called double-diameter or stepped. To minimize backwork on main piston when activating valve, you'll need a 'clearance' volume around underside of valve plate whereby O-rings seal transfer ports when valve plate is closed but allow 'upper' chamber pressure to largely balance pressure. With careful design, only a very light spring would be req'd (mainly for startup) since the upper chamber pressure would tend to keep the valve closed (due to valve plate area differential).

[matt brown]

I like the exhaust valve, but this begs the question...where's the 'intake' valve ??? My schemes usually favored staggered ports like 2 cycle ICE, and scavenged to common crankcase like ICE. However, you could also go with intake thru piston crown (a la WWI Gnome radial). As for heat input, Tom has an interesting idea he calls the hot potato, but I can't remember where it is onsite.

Your biggest issue will be getting heat in, and I doubt your current scheme will work since it's the 'typical' counterflow scheme (back & forth from main chamber) vs uniflow (from main chamber to secondary then secondary to main). I think Tom's hot potato could be uniflow or counterflow, and surely worth checking out before you continue.

[matt brown]

This type of ECE doesn't require much thermo knowledge since we're all familiar with similar ICE. And at this low end, you don't even have to worry about the PVT values per adiabatic index I posted. Instead, most values will be intuitive enough to guide you except one: mass transfer. Your current scheme is a reservoir scheme and 'simply' swapping gas volumes will be more challenging that it appears. Since this hinges on 'constant volume' heating, it presupposes that you'll be swapping the same MASS volumes, and have to achieve this continuously, consistently, and fairly quickly. Thermal lag schemes nix this mass volume issue, but that's about all they solve. The more you study this, the more you'll understand why everything is at a standstill...

[VincentG]

I had thought I had some level of novel idea. But thank you for that insight. At least I know I am on the right path. As usual, your numbers make my head hurt.

I had wondered the effects of the hot chambers influence on the poppet valve. i had thought it possible it may end up evacuated of air over time and need a snifter valve of its own.

The exhaust valve also functions as the intake, at the closest source of negative pressure I could envision, the piston at BDC. I also imagined a remote cold air sink accessed by a separate intake port as you suggested. I thought it might be added complexity for no appreciable gain.

Can you explain what you mean by constant volume heating? The primary piston will be reducing the volume at the same time the secondary compression stroke is adding heat.

I do value your insight, but level with me for a second. When you say stand still, where do you mean? The industrial/government sectors? Because as low level as this may be to them, where is there even a remotely advanced level Stirling available to the publc. The sunpulse seemed like an amazing machine, and was as basic as they come in design. The vast majority of the interested public would be ecstatic to produce 500-1000W from biomass. We are not looking to power a multi decade space mission to Uranus. No pun intended.

[matt brown]

Can you explain what you mean by constant volume heating? The primary piston will be reducing the volume at the same time the secondary compression stroke is adding heat.

Common ICE Otto cycle has adiabatic compression, constant volume heating, adiabatic expansion, constant volume 'cooling'. In this manner, the common ICE is a gas spring with a fixed volume ratio with heat added after compression, but where various heat inputs may be added (pending input temperature and time). IOW the Otto heat input process is independent of the volume ratio. Most engines (ICE or ECE) have inter-related thermal & mechanical relationships.

With your scheme, you may be compressing the gas in the main chamber (and the secondary chamber) but the transfer between these chambers will determine output & efficiency. An ECE Otto would have heat input (after compression) at constant volume, but multiple chambers & volumes complicates this. Rereading your pitch, I see that you're scheming to heat the secondary volume at constant volume while the main volume is compressing, and then mix both volumes when the valve opens. Overall, this doesn't loose (much) heat energy, but mixing (for ex) 100cc primary 'warm' volume with 100cc secondary 'hot' volume will not yield as much power as 100cc 'hot' volume. It's better to compress the primary volume while heating the secondary volume, then exchange volumes without mixing, whereby the hot 'secondary' volume expands while the 'primary' volume is_now_heating.

Tracking the mass of each gas volume will remain the issue. The gas will move to equalize pressure, but temperature differences mean different relative mass: take a 6' sealed pipe and pressurize with gas, then heat one end to 600k while keeping the other end at 300k. The pressure will be the same thruout pipe, but the relative mass (density) will be twice as high at the 300k end as the density is at the 600k end. No biggie for 'static' gas in fixed volume pipe, but major issue for dynamic flow in gas circuit of engine. We can easily measure PVT values, but gas mass is elusive, and sinks many schemes. The takeaway here: cold gas 'pools'

[VincentG]

Matt, as is often the case, there is a divide between the formally educated theorist and practical application in the real world. While I will not turn a blind eye to the experience and physics you present, I would like to offer a measure of effectiveness for hot air engines I have not seen often discussed.

We essentially have an unlimited source of heating and cooling, so I have little concern for BTU utilization efficiency. What I prefer to use as a metric is what I think of as delta T utilization.

For instance, with my LTD, I have rough numbers of ice water (40 degrees) and steam (210). Using a Gay-Lussacs law calculator, this gives me a maximum theoretical target of a 5psi pressure differential. I was able to more than double the MEP of my LTD with just improved displacer timing and movement to roughly 1psi (and this is with effectively zero compression ratio from the LTD). I have no doubts I could improve that number on the LTD. But it has inherent flaws that work against itself.

If I could achieve, let's say, 3psi (at same dT) with this engine, I would consider it a success, with a DTU(yea I'm making this up as I go) of over 50 percent. Scaling up in size and temperature should only increase this number.

That said, you have given me a great idea to implement a uniflow design. I had wondered if it would be better than counterflow. In my head, the counterflow would create a more rapid temperature increase, at the expense of pumping losses and standing wave issues.

[VincentG]

If I install a one-way reed valve on top of the large face of the poppet valve (on the hot side of the valve face), it will allow working cylinder air to be compressed into the hot chamber until the pressure of the hot chamber exceeds that of the working (constant volume heating perhaps?). Then when the bash valve is hit, there will be a uniflow back to the working cylinder.

Maybe this will also keep the mass volume of air up in the hot chamber.

[Tom Booth]

I have a YONGHENG high (4500 psi) pressure air pump that has a similar second stage compressor piston but I'm not sure of the internal workings. I think some port system connects the two cylinders, however the two pistons are joined as one unit. I haven't had it apart, but this image from an ad on eBay shows the duel low and high pressure piston

Resize_20230317_173100_0051.jpg (59.16 KiB) Viewed 5912 times

It generates an enormous amount of heat and requires using a water pump to circulate cold water (many people recommend using ice water) around the high pressure cylinder to prevent heat damage.