Stirlingengineforum.com

Lesson 5 - 3 Experimental LTD Stirling engine with 100W output: part A

This is a Stirling engine course translated from Fujio Toda's YouTube channel:
https://www.youtube.com/@user-yw7eo3no6m/videos

I've covered Lesson 5-1 and 5-2 in this post:
viewtopic.php?f=1&t=5513

In previous lessons, the professor shared his experience on building 100W class Stirling engines and turning one of them into a more compact build.

In this lesson, a LTD Stirling engine will be discussed.
The temperature difference is about 100C and the target output is still 100W.
Simulation models and calculations started showing their value in this lesson.
The professor used them to decide how big this engine should be to reach the target output.
Since he and his team had the experience from previous models, this LTD Stirling engine successfully achieved their goals.

This part starts from 0:35 of the following video:
https://youtu.be/XymfL5zBo_c?t=34

[Translation starts here]:

(Picture starts from 0:35)
(Picture title: Very Lowe temperature differential Stirling engine) Here are some pictures of low temperature differential(LTD) Stirling engines.
That engine held by a person's hand was developed in 1990.
At the same period of time, Mr. James R. Senft from University of Wisconsin developed a similar engine.(Please look at the middle bottom of the screen.)
My research result at that time was the "Beaker engine" on the top right of the screen.
It requires slightly higher temperature difference to work.

And this is a can Stirling engine.
(picture starts from 1:28)
(picture title: Types of educational engine 2) It is often used in short-term workshops.

Besides this, models like waste-heat powered engines are also used.
(Picture starts from 1:51)
(Picture title: Very Lowe temperature differential Stirling engine---Waste heat powered type) These tiny models are powered by waste heat from other equipment.
On the left there's a model powered by heat from LED lamps.
(He started playing the video footage at 2:00)
LED can convert 80 percent of the electricity to light ant that 20 percent wasted heat is used by this tiny model.
Moreover, on the right there's a large pellet stove.
LTD Stirling engines can be powered by the heat from its surface.
Though that can engine was roughly built, it had no problem spinning with such low energy input.
That copper heat sink on its right is a Peltier device.
The Peltier module is underneath heat sink and connected to the pellet stove's surface.
Through thermoelectric effect, some electric power was generated and powered a motor.
You can see a blue propeller spinning on its right.

These LTD Stirling engines use styrofoam-made displacers and cannot tolerate higher temperature.
Their structures are usually fragile and unreliable.

Here's some data from educational Stirling engines.
(picture starts from 4:39)
(picture title: Test results of educational Stirling engines(Gamma type) ) (X axis represents rotation speed in rpm)
(Y axis represents output power in mW)
On top left is data from a can engine, delta T = 7K, Top power is about 8 mW
On top right is from a beaker engine, delta T = 50K, Top power is 11 mW
The bottom one is a test tube engine, delta T = 164K, Top power is 190~200 mW
The performance of educational models is quite low.

In order to get output power above one Watt, an experimental Stirling engine was built.
(Picture starts from 5:40)
(Picture title: 10W Class LTD Stirling engine) Hot side temperature: 130C
Buffer pressure: 3 atm
Shaft power output: 10Watts

This was our first try.
We made all of its parts besides cylinder liner and piston rings.
A square-shaped plate heater is used as heat source.

We wanted to go one step further to 100W and did some simulations to find out how big that LTD engine should be.
(Picture starts from 6:51)
(picture title: simulation results) A condition was set for this simulation:
Hot end temperature:100C
Cold end temperature:25C
Buffer pressure: 101.3kPa = 1atm
Bore stroke : 400 x 200
Phase difference: 90 degrees
Working gas is air.

First there's Isothermal model, which is used in Schmidt cycle.
And the curve of Adiabatic model is near it.
Below them is Quasisteady flow model.
These three models are considered.

There are two straight lines in this graph which represents ideal conditions.
We were unable to provide conditions for these two models to work.

With this simulation result, we decided that our target is to create a LTD model which can generate 100W or higher at 100~150 rpm.

Lesson 5 - 3 Experimental LTD Stirling engine with 100W output: part B


There's many detailed drawings of this huge LTD Stirling engine in part B.
The professor also reveled materials and other specs of this engine.

A research paper is based on this model.
You can get full text here:
https://www.nmri.go.jp/archives/eng/khi ... owtemp.pdf

This part starts from 9:28 to 18:06 of the following video:
https://youtu.be/XymfL5zBo_c?t=34


[Translation starts here]:
Here's one section view of the LTD Stirling engine.
You can see its specs in the table below.
(Picture starts from 9:28)
(Picture title: 100W class LTD Stirling engine) (It's a gamma type engine)

Since the displacer is huge, we had to use two shafts on it.
The flywheels are connected to the power shaft by couplings.
Scotch yoke is implemented in this engine because we want to shrink its size.
We wanted to test this engine and see if it matches the simulation model.

Compared to educational LTD engines, this one is quite huge.
Its width is 1227mm and height is 1616mm.

This table contains more specs of the engine.
(Picture starts from 11:34)
(Picture title: Specs of 100W class LTD Stirling engine ) The heater and cooler are identical, both made from copper finned tubes.
The regenerator was made from brass wire cloth(50 mesh).

Let's take a closer look at heat exchanger's section view drawings.
(Picture starts from 12:13)
(Picture title: Section view of Heat Exchanger) On the right side of the screen, a drawing shows detailed layout of the heat exchanger.
Eight hollow circles on the top are the heating wire and the eight on the bottom are the cooling wire.
Regenerator in the middle separates the heater and cooler.
The heating wire is twined and stacked around the displacer cylinder, so do the cooling wire.
Please note that air can still flow around the thin space between heater wire and wall of displacer cylinder.

This is Scotch yoke and power piston.
(Picture starts from 13:00)
(Picture title: power piston and Scotch yoke mechanism) Under the crank shaft is a crank pin.
This pin will rotate around the crank shaft.
Its movement will be regulated by two sets of slide bearings, one horizontal and one vertical.

The power piston underneath is linked to the yoke.
There are two rings on this piston.
One rider ring is on the top and below it is on-way piston ring.
(In this picture, only the on-way piston ring is marked by arrows)

Here's some more detailed drawing of sealing on this engine.
(picture starts from 14:12)
(Picture title: Sealing ) On the top left is sealing of power piston.
The rider ring's purpose is to keep the piston in place.
Under neath is the one-way piston ring we developed.

On the top right is displacer.
A soponge-type Teflon ring is used to keep this displacer in place.
This will force the working gas flow through that thin gap in heat exchanger.
(Please go back to 12:07 to see the drawing on the right)

One-way shaft seals are also used in this engine.
On the bottom left is the sealing for displacer's slide shaft.
On the bottom right is for power shaft.

These are the power piston and piston rings.
(Picture starts from 15:30)
(Picture title: Power piston and piston rings) The rider ring is removed from the piston and placed on the right side.
The one-way piston ring is not visible in this photo.

Here's the displacer and its seal.
(Picture starts from 16:04)
(Picture title: displacer and Seal) That yellowish green cylinder is the displacer.
Its diameter is 800mm.
Due to its size, it cannot use ordinary shaft position, otherwise it may wobble.
In this case we used two shafts.
That black and white belt is its seal.

You can also see the heat exchanger in this photo.
Its upper part is exposed and you can see a blue wire.

The displacer is made from hardened urethane foam.
This is a light-weight material, but the displacer still weights about 20Kg.
We were unable to make a hollow displacer.

The red machine in this photo is the engine which was about to be completed.
We were checking its condition.
(Picture starts from 17:17)
(Picture title: the completed engine) We made sure no parts can wobble and all of them got assembled correctly.
After that, we can finally run tests on it.

gitPharm01 wrote: ↑Mon Apr 03, 2023 2:21 am ... target output is still 100W.
Simulation models and calculations started showing their value in this lesson.

The professor used them to decide how big this engine should be to reach the target output.
...

Nothing personal,. I'm all for preserving history but 100 watts from something the size of a refrigerator? A 45 pound displacer?

This weed Wacker is over 2 horsepower (more than 1500 watts) and the engine is about the size of a baseball.


https://youtu.be/tMJjFAO4Uiw


Need I say more? I think this points up the need for innovation. I think we can do better.

Aside from the enormous amount of work going to waste hoisting a 45 pound SOLID displacer up and down, and one wonders why it was not possible to make a lighter weight hollow displacer, here is another glaring wonder of design genius worthy of a government/NASA grant/contract


Copper cooling tubes virtually right on top of copper heating tubes and a practically non-existent regenerator of Brass ?

I'm curious what is to prevent virtually all of the heat input from being transfered directly to these copper cooling tubes ?

In an electrical analogue that would be a direct short circuit, typical of these engines based on simulations based on centuries old theory and mathematics.

Tom Booth wrote: ↑Tue Apr 04, 2023 9:25 pm
Nothing personal,. I'm all for preserving history but 100 watts from something the size of a refrigerator? A 45 pound displacer?

This weed Wacker is over 2 horsepower (more than 1500 watts) and the engine is about the size of a baseball.

Need I say more? I think this points up the need for innovation. I think we can do better.

I have to point out that this huge engine is powered by merely 80C Temperature difference.
It showed possibility to utilize low density waste heat.
And of course the cost efficiency must be evaluated before any attempt to do it for commercial purpose.
A simulation model predicting how huge that engine is needed.
And that engine had proven that this model worked.
So if there is a potential renewable heat source out there, we can first evaluate it with this model and see if it is worthy of harvesting by Stirling engines.

This is an experimental model, not something ready to hit the market.
The purpose of this course is to give students some basic knowledge about this engine design, not showcasing revolutionary breakthrough.

I admire your persistent effort in finding ways to improve and innovate Stirling engine.
Many scientists around the world are also trying to fix its problems.
And I consider a complete and solid knowledge base is beneficial for members of this forum to make the decisive breakthrough in the future, not repeating the failures that had been made.

gitPharm01 wrote: ↑Tue Apr 04, 2023 10:27 pm
So if there is a potential renewable heat source out there, we can first evaluate it with this model and see if it is worthy of harvesting by Stirling engines.

That's my point exactly, and according to the "modeling" the answer would inevitably be "no". It would not be worth investing the time and resources for such a meger return on the investment.

And I consider a complete and solid knowledge base is beneficial for members of this forum to make the decisive breakthrough in the future,...

I agree completely with that. We may not, though, be in agreement as far as what constitutes "complete and solid knowledge".

I've been told a million times that this efficiency equation, which is at the heart, I believe, of many modeling algorithms constitutes "established science" or an actual thermodynamic physical LAW of the universe.

η = (Th - Tc) / Th * 100%

Read all about it:

https://physicscalc.com/physics/carnot- ... alculator/

If this bit of the supposedly "complete and solid knowledge base" is wrong, then all the modeling based on it is also wrong and will give a false as well as a very bleak and discouraging picture.

Theoretically I could greatly improve the efficiency of my Ford pickup by dumping all the gasoline on the road. I can even write some equations on the efficiency gains to be had by eliminating all that excess weight.

This is similar to the result of much of these modeling efforts ultimately based on obsolete theory and mathematics.

How much heat can we dispose of as quickly as possible seems to be the grounding philosophy. As if heat in a heat engine were a bad thing.

Cooler: 14866.3 cm3
Regen: 2424.6. Cm3

The cooler, disposing of "waste heat" is seven times the regenerator intended for retaining heat.

...not repeating the failures that had been made

Yes, but what is being presented as a "failure" and what is being presented as "complete and solid knowledge" ?

The knowledge includes approaches that had been tried and tested.
In this course, the professor presented these things and did not hide his failed attempts(in Lesson 5-1).

He did not overemphasize the power of simulations and if you read my translation closely you will find out he did NOT use those two ideal models (Ideal isothermal and Ideal adiabatic). Physical laws and building materials' physical limits were well respected during calculation.

That extremely low temperature difference condition was kept and he managed to create an engine design that may achieve the designated goal: deliver power output more than 100W.
This engine was then built in 1996 and he and his team did their best to build up a functioning engine with technologies that are available at that time.
Moreover, they recorded materials and mechanisms they used.
With these in one's knowledge base, he will less likely to reinvent the wheel.

And if he, with new technologies at hand, still insist on rebuilding this engine with original materials and designs,
that would be a failure.

gitPharm01 wrote: ↑Thu Apr 06, 2023 7:19 am ...

He did not overemphasize the power of simulations ...

My rant was not so much against the use of simulations or mathematics generally, but rather in defense of the trial and error approach.

I think there has been a decidedly heavy bias towards a dependence on simulation and mathematical models and against making "random changes". So my feeling is a counterpoint view is in order.

Just for example:

This development process includes two things: analysis and simulation.
Analysis means getting measurements and data from the models.
Simulation means building mathematical models and using methods like Simpson's rule.
(The professor also mentioned many other methods but I don't know these things due to my lack of knowledge, sorry! )
Analysis and simulation are both necessary, they can't work without each other.
Without data from analysis, a simulation model cannot be made.
Without any simulation, you will blindly make random changes on the prototype engine and get no meaningful data.

Many people hear, I believe, very much enjoy and derive a great deal of benefit and insight trying out new designes or new methods, approaches, theories etc. I know I for one am prone to experiment with whatever random idea pops into my head.

With that approach, it seems likely to me that a vast swath of previous, time honored theory and methods are likely best discarded opening up a world of new possibility and potential.

Nearly everything I ever learned and studied about how Stirling engines work turned out to be completely wrong, begining with a Stirling engine cannot run without a flywheel, which was a universally accepted assumption. Even considered an absolute physical limitation., that could be mathematically demonstrated. For me that was only the first domino to fall.

Thank you for sharing. This is very promising output considering the extreme mass of the displacer and the low dT.

VincentG wrote: ↑Sun Apr 23, 2023 6:18 am Thank you for sharing. This is very promising output considering the extreme mass of the displacer and the low dT.

I find the design of this engine a little strange, as well as apparently horribly inefficient, if I'm interpreting the diagram right. Maybe not.

But take #6 the "expansion space":
Normally, in an engine with an enormous bottom (LTD / pancake) this huge surface area would be utilized for heat input. It appears however, that in this engine the heat input is confined to the lower portion of the side walls.

So, as the displacer moves up, the cold air moves down through the cooling coils, down through the regenerator, briefly through heater pipes and then after passing through the regenerator and heat pipes, passes into this enormous unheated(?) "expansion space".???

A huge unheated cold side, another enormous unheated "expansion space" in addition, active cooling coils/pipes and a virtually non existent regenerator which, due to the arrangement appears to be mostly working backwards.

Heat picked up from the heating pipes as the gas moves upward passes through the regenerator and then the cooling pipes????

Then on the way back down the cold air is cooled before passing into the hot regenerator, heated, then passed into another cold "expansion space".???

This, is the model of excellence?

Add in the dead weight of this gargantuan solid displacer, which cannot be made hollow/lighter why???

No explanation.

The engine appears to be so crippled by poor design choices, it's a wonder it could ever run and produce any power output at all.

As some kind of lesson on what NOT to do, that I could understand, but when someone suggests that whatever methodology was used to produce such an engine is the pinnacle of achievement and the only way to go, I can't help but wonder.

Compare this with a SunPulse type design in the .5 to 1.5 kw range.


https://youtu.be/c5sc_v9RkPY


I'd venture to say that this may not be an ideal design either, but

Tom I think the takeaway here is that the engine made this power despite its design shortcomings. I'm sure he made use of the materials and techniques available to him at the time.

VincentG wrote: ↑Tue May 02, 2023 4:47 pm Tom I think the takeaway here is that the engine made this power despite its design shortcomings. I'm sure he made use of the materials and techniques available to him at the time.

That's fine, just trying to put things in perspective. The engine is basically being required to lift a stack of about 10 bricks. (A brick averages a little over 2 kilograms)

From the begining of the presentation, the general "takeaway" seems to be, all manner of trials were conducted in an attempt to produce a running engine, all endings in complete failure. Finally with the help of some wonderful computer modeling this is the grand achievement and really, nobody should think they could do any better. If I recall, a number of big names were mentioned, authorities in Stirling engines.

Kind of reminds me of the Tour de Sol I went to. Kids built solar powered cars and drove them across the country to attend a competition in Great Barrington Massachusetts. There was a DOE "energy bus" at the site. It had about 10 big solar panels on the roof.

Inside there were displays detailing information about different energy sources, Oil, Gas, Nuclear then at the end there was a tiny little fan, the kind that could run on a little AAA battery with a placard that stated that the fan was powered by the solar array on the food of the bus.

Taped to the solar display was a piece of paper that read "OUT OF ORDER"

"roof of the bus" that should be.

I found this LTD experiment quite interesting, especially due to the extreme values...

25 L piston swept volume
40 L displacer swept volume
2.5 L regenerator volume

300K - 375K thermal cycle, thus 1.25 dT ratio
1.6 compression ratio if we isolate and idealize events where (40 L + 25 L)/40 L = ~1.6
.8m/sec piston speed @ 120rpm

The 1.6 compression ratio is above the 1.25 thermal ratio, so ideal regen 'load' will still exceed ideal input (IOW the heat being recycled inside regenerator is greater than input) and this will be hard to capture with low dT of cycle (low gradient across regenerator) despite the relatively massive volume of each blow.

If this bugger was pressurized and given a little more heat, it should come to life nicely with (1) more power per stroke (2) more speed (3) more efficiency, thus (4) a lot more power (maybe 2HP/3Kw).