Gadz, no one pointed out that these are textbook isobaric values
VincentG wrote: ↑Thu Mar 21, 2024 12:44 pm
The actual observed total volume while at operating temperature is 3.1cc for a total gain of .5cc!!!!!
This corresponds to a delta T of less than 80k!! And that's what makes these things toys. Not for being small, but for being shadows of their potential.
"a man's gotta know his limitations"
VincentG wrote: ↑Thu Mar 21, 2024 12:44 pm
This is a pathetic DTU(delta temperature utilization), but an illustration of how much untapped potential these engines have. This volume ratio to actual thermal ratio would leave most scholars scratching their heads and traditional Stirling guys in a fit.
Indeed, ANY isobaric process must be carefully considered. If we consider EXTERNAL heating processes via an order of merit, then (1) isothermal (2) isochoric (3) isobaric. Without regen, isobaric input is lame (1) temperature increases per volume increase while work output per volume increase decreases (2) there's a lead/lag mechanical/pressure issue when completing the cycle that's especially nasty when closed cycle. The best isobaric eff cycles are low volume ratio Braytons, but nowhere to be seen due to low specific power.
VincentG wrote: ↑Thu Mar 21, 2024 12:44 pm
It's basically just beyond a self sustaining adiabatic bounce at this point. But it still runs at high rpm and makes pretty impressive power and noise.
Yep, more parlor trick than serious contender.
VincentG wrote: ↑Thu Mar 21, 2024 12:44 pm
Now, imagine a scale HTD engine with the same volume to thermal ratios but with an actual delta T of over 300. We don't just have to imagine, it can be built.
All it takes is massive perseverance and a little luck. Energy has consumed more man hours than any other human research endeavor, ahead of food and weapons.
Gadz, no one pointed out that these are textbook isobaric values
Yup but I hit 80% of target isobaric numbers on my LTD epoxy engine. So about the same 80k effective delta but from only a (plus or minus 10k) 100k external delta.
I'm still uncertain how to model LTD thermo. However, if there's a cold stroke, then it's a vacuum cycle, and this would align with no regen. The biggest problem I have with modeling this bugger is DP/PP volume ratio where if DP/PP=50, then there's little expansion work and almost as if (under this volume assumption) that 98% of input (49/50ths) is DP 'dead space'. The opposite view is to combine DP and PP volumes into a singular volume and ignore distinct heating/cooling issues or invent some creative buzz.
Assuming volume is closed and never leaks past PP or DP, we know:
(1) a thermo cycle must be maintained where Wpos>Wneg even if meager
(2) which rules out any 2 process cycle where Wpos=Wneg
(3) and leaves conv'l 4 process cycles, 3 process cycles, and maybe some unconv'l cycle/s
(4) any isothermal or isobaric process is slower than similar adiabatic process due to external heat transfer rate
(5) cycle rate (rpm) is due to internal vs external MEP differential
Returning to LTD as a singular volume, we can rule out distinct isobaric Ericsson and Brayton since both require different compression vs expansion volumes. Next, we can rule out strict Carnot and consider it merely an approximation between strict Otto and Stirling. This leaves Otto and Stirling as 4 process candidates where Stirling has a higher order of merit due to an unconstrained volume ratio (Otto's adiabatic processes are thermal bound).
(D) isochoric input - isochoric heating, adiabatic expansion, isobaric cooling to ambient aka Lenoir cycle and not possible (just honorable mention) due to open cycle and isobaric cooling.
After nixing non-compression Lenoir cycle (D) we have a couple possible 4 process cycles and a few possible 3 process cycles (need a compression cycle even if only 2%). So, pick your poison...let's start with isobaric input A and B (above). The bugger with both is connecting the dots (completing cycle on PV plot) and each will have thermal ratios constrained by their volume ratios. In thermo slang, their 'run' (volume change on PV plot) is so short that there's little difference between both adiabatic and isochoric 'drop'. Next up is isochoric C which also has a fairly short run, but a greater 'rise' potential (input pressure change on PV plot). Despite short runs, all 3 are contenders since 2% expansion is a short run (almost a rounding error).
It's also possible that the cycle may vary depending upon input, maybe Stirling at one time and Otto at another. The only wild card cycle I can think of is adiabatic compression, isobaric heating, adiabatic expansion, isochoric cooling and this irregular cycle can connect a lot of dots (various PV plots).
Starting at BDC with an atmospheric engine at the zero point
1: Internal temperature lowered to Tmin at constant volume
2: Piston driven toward TDC from ATM pressure at constant temperature "compression" (more like returning to 14.7psi)
3: Internal temperature raised to Tmax at constant volume
4: Piston driven towards BDC from increased internal pressure at constant temperature expansion (again returning to 14.7psi)
At the zero point the engine is a quasi-compression engine. There would be compression if the gas temperature was not kept at Tmin during movement from 1 to 2. In the same way, it is a quasi-expansion engine while moving from 3 to 4.
From a thermodynamic standpoint, there is no Wneg, no back work.
This is the easy point to reach. The difficult part is going beyond the zero point in the right way.
The above mentioned Gama engine has gone above the zero point but in the wrong way, needing and open alcohol flame to reach a measly 80k delta T. If built to use the full 800k delta T available, the displacer volume would be 10 times smaller than it is now for the same power piston.
The previous wildcard cycle I referred to is the Diesel cycle
Diesel_PV.png (35.46 KiB) Viewed 6704 times
I chose this example due to simple plot and process callouts inline previous PV plots. Note that when Vr=constant, the 3-4 process can 'move around' (somewhat) within the same basic PV plot for both Diesel cycle and previous 3 Stirling PV plots. In this manner, there's sufficient wiggle room for various finite source and sink values.
I machined a steel hot cap so I could put more heat into the little thing. Of course steel conducts way better than glass so now it required active cooling with ice to prevent overheating.
It ran good but not much different than the glass with passive cooling. The measured expansion ratio never improved.
I then realized this is the path of the standard Beta stirling engine, NASA engine or even the Philips stuff. More heat more cooling more heat more cooling more charge pressure more heat sinks more dead space blah blah blah.
It's a constant battle between fire and ice and IMO a dead end without a major rethink in mechanical layout. My epoxy LTD is very effective but still can't handle high temperatures.
Something like the Rider Alpha, with distinctly separate hot and cold sides is a massive improvement from a thermodynamic standpoint but packaging is an issue.
Ive ordered this engine for testing after seeing how well it performs(4krpm with a torch) and realizing it's actually an Alpha and not a Gamma like I had thought.
Stirling kit.com actually rates it at 5 watts.
I think this with a piston port exhaust will be very power dense and the open cycle will drastically improve the actual thermal ratio without the need for exotic ceramics and the like.
20240328_125848.jpg (106.78 KiB) Viewed 6620 times
I drilled an exhaust port in the power piston cylinder of the high temperature Gamma engine configured for stock displacement.
It started running very fast and then a hole popped through the glass hot cap. A failure I've never seen anywhere before, indicating to me that internal pressure was higher than ever.
I'm still trying to see this as an alpha, especially due to 'cooling fins' below hot cap. Anyways, I've often wondered about such a simple experiment. Here's my thinking where volumes, temps, and phasing are whatever...
(1) if 1cc of engine gas exhausts to ambient and intakes 1cc of ambient, the cycle only remains constant if both 1cc have the same molar mass. It's simply all wrong to consider both as simply "1cc".
(2) since ANY 1cc exhaust will be dependent upon PP vs ambient pressure differential, there's only 2 comparative exhaust possibilities (A) low temp high molar mass (B) high temp low molar mass.
(3) if A occurs, then the gas mass will decrease and the engine stops, but if B occurs, then the gas mass will increase and the engine explodes. Both presuppose that system has no gas leaks and that heating supply per mass is constant.
I was recently pondering an Essex with ported hot PP and inlet 'snifter' valve to DP cold space, and only gaming backwork, when I saw there's a little more to consider (eyes roll). Nevertheless, any open cycle is frowned on by SE crowd, but I still consider it a major wildcard in cycle schemes. Most guys consider 'open cycle' akin ICE and limit concept to ambient vs 'open cycle' with LP reservoir.