Big Beta Stirling 1kw, 100degC to 400degC

[Bumpkin]

I could envision a brush-type seal to stop most of the flow, but don't know any way to totally seal off the dead space without a lot more resistance - so why not just let the annular exchanger/regenerator do its job? The path of least resistance would lead most of the air through the external exchanger anyway. I'm with Geoff as far as optimizing what you've got first though. By the way, a Beta with a sealed displacer and external exchangers could just as well be an Alpha, and perhaps better. Bumpkin

[zhivko]

That's huge hot heat exchanger with 3,5m^2 square area but what worries me is volume! It adds 5,56 l of dead volume. That would surely break proportions... Comments anybody?

Stirling2_1_with_exchanger.jpg (170.97 KiB) Viewed 8475 times

[Geoff V]

Zhivko

Tubular heat exchangers are very good at adding surface area, but by comparison to slotted heat exchangers (being inside and adjacent to the cylinder walls) they add a lot of dead volume, even worse they restrict the pass area.

The Philips 1002ca produces 500w at the crankshaft from 61cc working volume at 14bar on air and the heat exchangers consist of 160 slots 2.38mm deep x 0.4mm wide giving a pass area of 1.52sq cm. If we divide the working volume by this pass area we get 40:1 or each sq cm has to pass 40cc twice per revolution.

Your proposed tubular heat exchanger has 296 x 4mm bore tubes, so pi x 4 x 296 = 37.2sq cm, if we now divide 9000cc/37.2 we get 242:1 or six times more restrictive than slots, increasing the ID of the tubes won't help much as it will increase the dead volume and reduce the rate of heat transfer.

Large swept volume SE's are not the solution to getting useful power, large volume slow revving SE's with plain exchangers have a use as has been demonstrated by Peter Lynn but trying to (change horses in mid stream) will create more problems than it will cure.

Develop this engine by all means but if you now understand the importance of heat exchangers I would recommend starting with a 'clean sheet' of paper.

GeoffV

ps. Fitting a seal to the displacer does not change it from a beta to an alpha even hanging fairly lights around the displacer will not change it to an alpha!

[zhivko]

Hi GeofV,

I did some calculation and I saw that slotted heat exchanger indeed have better HeatArea to deadvolume ratio, but slotted actually heaving smaller heat transfer area meaning that this would result in lower temperature of heated air than in tubular heater design.

HeatArea to deadvolume ratio (calculated in attached excel sheet):
slotted: 5000
tubular: 1000

Rethinking both exchangers in manufacturing point of view - it seems slotted would be easier to manufacture.
In other hand tubular exchangers give more control when prototyping SE - one could reuse same SE parts (displacer, piston, cylinder) meaning keep them, and redesign heat exchangers and/or regenerator exclusively.

Regarding displacer sealing - I know that this measure would not change it to alpha type, but how come that Peter Lynn also tried to fasten displacer in LSM 11 version of his SE (explained in PeterLynn LSM_11)

quoting Peter Lynn

But, when using tube heat exchangers, in addition to the usual piston seal, it's necessary for the displacer to have a seal as well, creating extra friction. Fortunately, this extra frictional loss increases as the square of dimension, while swept volume (and power output, providing speed stays constant) increases with the cube of dimension, so larger engines of this design would not be effected as much by this as LSM11 is.
LSM11 probably hasn't achieved its target output because the cross sectional area of it's tubes (8.5 sq.cm total) restrict gas flow, limiting rpm (and therefore power). Using larger diameter tubes would be likely to cure this without other unacceptable consequences.
However, there is no compelling reason why swept volume can't be increased to achieve required output. Even if a 15kw engine/generator requires 2m x 2m of floor, this will not limit dairy shed applications.

Thanks in advance for clarifications...

[Bumpkin]

Bumpkin: "By the way, a Beta with a sealed displacer and external exchangers could just as well be an Alpha, and perhaps better."
GeoffV: "ps. Fitting a seal to the displacer does not change it from a beta to an alpha even hanging fairly lights around the displacer will not change it to an alpha!"

Misrepresenting other's points out of context and then exaggerating to silliness could almost be construed as bluster. But to clarify -
Though I prefer the Beta design, the obvious point here is that if you are going to seal the displacer anyway, there are numerous other ways to move gas through external exchangers while modulating displacement - some might conceivably be more appropriate to the task. Bumpkin

[Geoff V]

Bumpkin

I take it that's a NO to fairey lights then, pity.

GeoffV

[Geoff V]

Zhivko

From your calculations it seems you have selected slots just 2.38mm deep, why, they can be 10mm deep or 20mm deep?

You also seem to have used a value for H of 20 w/m2/k, probably about right for large bore tubes but 0.4mm slots will transfer nearer 160 w/m2/k.

Peter Lynn points out the limiting aspect of tubular heaters on his LSM11, namely the small cross section area and the effect of pumping losses on engine speed, interesting his later engines are plain heat exchangers.

I find it sobering to realise that all these engines produce just 0.1 watts/cc working volume unpressurised on air at 600rpm with the one exception to date, the Philips 1002CA which probably due to its slotted exchangers produces 0.17 watts/cc.

Food for thought.

GeoffV

p.s. If you get fed up with beta's, don't waste time with alpha's, go straight to double acting gamma.

[zhivko]

Ok, I changed dimension of slots to (drawing attached):

slot wide 0,4 mm
slot deep 15 mm

CylinderInox.jpg (81.84 KiB) Viewed 8447 times

For 260 slots I got area of 2m^2. With 160W/m^2K this exchanger is quite capable. For 160W/m^2K and 500K (if I assume that air is to be heated from 100degC to 600degC) that comes: 2*160*500=160000W and that's quite a lot.
But mass of 590mm long slotted cylinder came to be 68kg - whaauu.. Who could lift such a heavy thing...
Maybe I could reduce some weight on that account? Am I right?

I am thinking maybe cooling bottom part of cylinder with water doesn't require bottom half of cylinder to be in slots (so I could maybe reduce weight) ... Am I right? What do you think?

I think to achieve those 160W/m^2K I should force all air into those small slots - so displacer should be quite tight fit ?

[Geoff V]

Zhivko

This thread started with a desire to build a 1Kw stirling engine, the latest drawing is of a slotted cylinder with a heat transfer capability of nearer 160Kw, even at just 10% internal thermal efficiency this would equate to a power output sixteen times more than your original plans. Not that it would work, because the pass area of the slots is far to restrictive for the working volume as I explained earlier in the thread, and the engine speed would therefore be far too slow to realise much power output at all.

I answer to your questions, it is advisable to fit a sleeve inside the slots to ensure the gas passes through the slots.
The cooler section can be slightly shorter than the heater as there is less rejected heat and heat transfer solid to liquid is easier.
But where is the regenerator casing???????????

GeoffV

[Ian S C]

Tio those who have access to James G. Rizzo's book "The Stirling Engine Manual" (1), there is an interesting example in chapter 1, headed "How to construct 'Prova II, a competition type co-axial Stirling Engine", this one is not pressurised, but has a concentric regenerator sleeve around the cylinder. Ian S C

[zhivko]

GeofV and rest of SE entusiasts,

Yes, missed the point - since I thought that long cylinder can act as regenerator by itself, but in this version (try my best today to draw it in inventor), I included regenerator chamber.
It has 260 slots, 0,4mm wide, 15mm deep. Heater is 265mm long, cooler is 180mm long.

Since I added water cooling I reduced lengths also - in this version thermal shortage should not occur. Still - weight of all parts together is 50 kg... :)

Maybe I could reduce length a bit but I don't see any way to reduce weight more...

Newest cylinder:

CylinderInox.jpg (203.71 KiB) Viewed 8425 times

Still with this cylinder - I must seal displacer somehow to force air flow through slots - Any ideas ??? Image or sketch of sealing would be much appreciated... Thanks!

It seems I would better buy wire erosion (EDM) machine - I found one localy for 1200€ :) Hour of work on this machine locally is 50€ per hour.


Ps: I like beta since it has single piston sealing on cold side only

[zhivko]

Bumpkin: "By the way, a Beta with a sealed displacer and external exchangers could just as well be an Alpha, and perhaps better."
GeoffV: "ps. Fitting a seal to the displacer does not change it from a beta to an alpha even hanging fairly lights around the displacer will not change it to an alpha!"

Misrepresenting other's points out of context and then exaggerating to silliness could almost be construed as bluster. But to clarify -
Though I prefer the Beta design, the obvious point here is that if you are going to seal the displacer anyway, there are numerous other ways to move gas through external exchangers while modulating displacement - some might conceivably be more appropriate to the task. Bumpkin

Bumpkin do you know for an elegant sealing ring that would not press to much to the cylinder wall and be adaptive enough to compensate displacer unprecise manufacturing?

Thanks,
Klemen

[Bumpkin]

zyvko, sorry, didn't see this before answering your personal message. For the record here - nope. Bumpkin

[vamoose]

Hey guys,

I've been playing around with silicone bulbs and have been using them in a double acting alpha configuration (so far with limited success. But am getting some positive momentum (and hoping for some more)).



The right kind of silicone compound can tolerate some decent temperatures up to and above 300 deg C. Also they have very low friction with certain textured materials and/or smooth surfaces, and seem very flexible.

For example- the compound mix they use for oven trays.
Not sure how much more they will tolerate above 300C, but some may, and seem to remain stable??



There are a lot of different type and size oven (silicone) trays available. Maybe it might be worth a try to scrapyard and cut some of them up, to experiment around with, for a loose fitting seal??
Just a school-yard suggestion, don’t take it too seriously.

vamoose

p.s. i don't recommend the Pink ones specifically, i just happened across them, ......'honest'..

[zhivko]

I am currently thinking about idea to increase sloth thickness to 1mm, and make sure that displacer gap to cylinder wall would be around 0.5mm - In this configuration - it seems air would be more likely to go through slots than flow through displacer/cylinder gap.

Is my thinking correct?