Stirling Engine Design Considerations

[dr dodge]

I am doing some experimental work with stirling enging design, in order to add supliment the power I use at my house, and want to get some facts straight, and throw out some ideas for discussion (or destruction, as the case may be)

Basic assumptions:(so far)
The more heat differential would mean power potential power.
Thermal creep that allows the cold side to get hot and vise versa decreases efficiency.
Most "normal inefficiencies" that apply to IC recips, will also effect the stirling (drags, stroke/rod ratios, mass/inertia, etc)
"Dead volume" is inefficient

Now some questions.
Rather than a common path for the expand/contract, can separate paths be used, with check valves to control flow?
Does the displacer have to move at the same cycle as the power piston, only 90 degree phase angle?
Is there a formula of temp diff vs ratio of piston sizes?
Is there a formula, or rule of thumb for flywheel selection? (aka vs stroke, piston mass)

thanks,

dr

 

[CFDFEAGURU]

Hello,

I am not sure that I follow all of what you are asking but here is what I can say.

The more heat differential would mean power potential power.

From a very basic approach

q = m*Cp*dt

where q = power (kW)
Cp = specific heat of constant pressure (kJ/kg - C)
dt = the temperature difference (C)
m = mass flow rate (kg/hr)

From that equation you can see that more mass flow or a larger temperature differential will produce a higher power.

Thermal creep ...

How do you mean this? To me creep is phenomena that effects a material at high temperatures. It appears that you are defining it in a different manner.

Most "normal inefficiencies"

I am assuming that you are referring to mechanical efficiency. Is that correct?

What about the thermal efficiency of the Stirling process?

Thanks
Matt

 

[dr dodge]

what is m in the first equation?

thermal creep would be the amount of heat that over the process of operation migrates to the wrong side. In the case of my design, the cold side is Earth temp., the hot side very hot, so the more the cold side rises the more loss of efficiency takes place.

and yes, I was talking about mechanical efficiency. Example: lowering mass of recipricating parts would improve efficiency

I would welcome any discussion of the thermal efficiency, too

dr

 

[CFDFEAGURU]

Sorry about that.

m = mass flow rate (kg/hr) (I fixed that oversite in the post above.)

I will look into my thermal books regarding the Stirling cycle later today.

What is your source of heat? An industrial burner or some other application?

Thanks
Matt

 

[dr dodge]

eventually a solar panel, 35 ft x 35 ft square, using tubing (probably roughened copper) and aluminum reflectors. Hot oil will be the media, unsure of the temps that it would attain. flow would be maintained by a 12 volt pump, powered by a PV solar panel and battery system if needed.

dr

 

[CFDFEAGURU]

...unsure of the temps that it would attain

What type of oil? The main problem with using oil if the temperatures get high (you probably don't need to worry about this) is it can crack and cause the tubes to plug. You would want to set your upper temperature limit to be a safe distance away from the cracking temperature of your oil.

Thanks
Matt

 

[dr dodge]

the oil selection would be based upon the real world temps that I can attain. I would prefer (cleaned) used motor oil. (and I get eco points for that...lol)

dr

 

[Yeti08]

The more heat differential would mean power potential power.

Just a nitpick, but it is the increase in temperature difference that results in more power.

Rather than a common path for the expand/contract, can separate paths be used, with check valves to control flow?

You'd have to explain in more detail what you're thinking, but a Stirling engine is considered a valveless system and requires a regenerator (specific or integral) in order to operate, so I don't think that would be possible.

Does the displacer have to move at the same cycle as the power piston, only 90 degree phase angle?

In order for it to work correctly, yes, you need one expansion and one compression stroke for each cycle. The angle doesn't have to be 90 degrees either - also note the difference between piston and volume phase angles, especially relevant for the different alpha, beta, and gamma designs.

Is there a formula of temp diff vs ratio of piston sizes?

Not specifically piston size, but I would suggest looking at the Schmidt analysis (I've also seen it called the Schmidt approach) for Stirling engines. It relates, for an ideal cycle, the different parameters such as cycle work, temperature ratio, dead volumes, swept volumes, etc.

 

[chayced]

Here is the best example I found while researching the exact same idea:
http://www.suction.co.jp/stirling/ya2.html
Sadly, for home solar generation you either need a stirling engine to sit at the focal point of a lens/mirror or you need a very large stirling.
When I get the time, I plan on building an open brayton cycle for solar steam generation. No cooling medium, run straight air through the collectors, and off the shelf parts.

 

[Bob S]

From a very basic approach

q = m*Cp*dt

where q = power (kW)
Cp = specific heat of constant pressure (kJ/kg - C)
dt = the temperature difference (C)
m = mass flow rate (kg/hr)

From that equation you can see that more mass flow or a larger temperature differential will produce a higher power.

Isn't the Stirling engine a Carnot cycle engine?
See
http://en.wikipedia.org/wiki/Heat_engine
The theoretical maximum efficiency of any heat engine depends only on the temperatures it operates between. This efficiency is usually derived using an ideal imaginary heat engine such as the Carnot heat engine. The theoretical maximum efficiency is

Efficiency_max = 1 - T_cold/T_hot,

where T_cold and T_hot are in kelvin.

Bob S

 

[dr dodge]

Sadly, for home solar generation you either need a stirling engine ... need a very large stirling. .

thats the plan, very large, slow

Last night i discovered the one of first hot air engines (buckett) used valves.
link to google books

http://books.google.com/books?id=XQ...CAcQ6AEwAA#v=onepage&q=cayley buckett&f=false

dr

 

[CFDFEAGURU]

Bob S,

I was not giving a calculation for efficiency, I was giving one for a simple quick calculation of the q value.

Thanks
Matt

 

[dr dodge]

I friend sent me to here, very interesting
it compares and improves on the schmidt model

http://sesusa.org/DrIz/isothermal/isothermal.html

I have gotten some progress done on my experimental model, using an old clutch slave from my 59 dodge firetruck (2.25 inch stroke at 1 inch dia) and have gotten my 6" aluminum flywheel just about done (with roller bearings) the goal is to make the power piston/flywheel (and sterno can) assy the "non-variable", then to change the hot side/displacer in various ways (bore/stroke relation and displacement). planned on using a simple strap type drag brake, weighed with nickels as the dyno to determine power output values.
(this is what happens when you don't watch tv (other than an hour of star trek...lol))

dr

 

[CFDFEAGURU]

Thanks for posting the link. Good webpage.

Thanks
Matt

 

[eduardomhay]

i am trying to come up with a setup that can produce a mechanical (shaft)output around 1KW.
i have an automotive turbine (the turbine side of a regular turbocharger), and this should drive both: it's compressor and the abovementioned shaft output.
i removed the compressor part, because it would never give me sufficient compression ratio that could make sense in a Brayton cycle configuration (the original compressor is meant to give very little boost, since otherwise pre-ignition would occur, diessel TCs don't go very much higher), from various literature i see that i will need something over 8 as compression ratio. i am considering designing, and building a centrifugal compressor, with forward looking vanes(to get higher pressure output).
i calculate i need to handle some 50 #/minute of air.
since i must use a dynamic compressor (no positive displacement options), i ask...
am i in a right track?, has someone looked into something similar?, is there something already available?

tks
eduardo

 

[Vespa71]

Last night i discovered the one of first hot air engines (buckett) used valves.
dr

Another invention-suggestion: The Stoddard-engine http://en.wikipedia.org/wiki/Stoddard_engine

I, personally, refer to The Stoddard Engine at the original hot-air-engine, maybe I'm wrong. It is in principle a valved two-lane "Stirling", or an "all-vapor Rankine", if you will.

For your use, I suggest setup with parallel-array-garden-hoses cooled underground or rather in running water if available (low pressure-section). Then the cooled air enters high pressure/low capasitity-piston-compressor. Air now holding elevated Temp. from compression, receives your T-in from sun (must of course be higher than compression-derived T of medium). Expantion then runs a larger displacement piston cylinder( or same size at higher rev.) towards low pressure, and cooling as described above.

Compressor may be valved with one-way-valves, like a normal air compressor, while engine must be valved like a air/steam engine, in principle. Compressors and air-motors are shelf-products, yet the air-motor must handle temperatures higher than normal in garage.

In short: Pressureelevation determined by compressor-capability or T-in, combined with good cooling, determines theoretic max-thermal eff. (DeltaT/Absolute T1) It improves efficieny further to heat expansion-cylinder with T1 , and cool compression cylinder with T2, like in a Stirling. To approach Carnot-cycle (and avoid flaws of Stirling) you would also need adiabatic compression and expansion on final stage of each, suggesting at least 2 cylinders for each action whereof the first stage isobaric, and final stages temperature-isolated.

This machine is in my opinion closer to Ideal Carnot Cycle than the Stirling, because I find that the Stirling has serious flaws to the Carnot-cycle during gas transmission. Recuperator would have no effect, if the machine was perfect. Recuperators only works because pressure and temperature are at faulted values between actions. The religious item recuperator (which really is a half-effective crossflow heat-exchanger), reverses some of the uncalledfor effect (which is heating at low pressure, cooling at high pressure). In this aspect the Stoddard is better, as it performs not the same counteractive actions (providing adiabatic stages present). But it has valves (causes friction and entropy-increase) and therefore relies on good compressor/motor-mechanical efficiency (low friction)to work. That was difficult in 1933, but perhaps possible now.

I suggest to check (research, test) mechanical efficiencies on hardware you chose, to know what total mechanic loss you will encounter. Since Stoddard is loopish, no strangling of lines are necessary to avoid dead-water, as in Stirling, losses elsewhere can remain low.

Steam: It is conceivable to replace air with saturated steam (continuosly vaporized), complicating the system and reducing possible T1, yet solving lubrication and sealing-issues. present with air at high pressure/temperature. (Train-steam-engines used no oil in motors). Condensate formed can be picked up from low point at condenser and sucked into compressor with injector.

Oil for lubrication is also possible, but limits T1 to oilburn-temp. as mentioned by CFDFEAGURU. Might help to separate oil (cyclone?) after compressor and introduce it to engine after sunheat-collector, thereby avoiding partially very high temp-exposure between sun and oil.

Material-problems reducing possible T1-p1 was, and is, the main problem to reaching high Carnot-eff. in heat-engines

Best of luck with important work.