Converting an LTD stirling to a stove fan - hopefully

[Stroller]

Hello everyone, I'm new here. My name is Rog, I have a mech eng background as a precision machinist, and tinker with classic bikes etc.

Owning a stirling engine with a useful practical purpose is an itch I've wanted to scratch for many a year. My thermoelectric stove fan is ok, but a bit weedy for pushing air airound my 14 x 13 foot living room. Scouring ebay, I found a 'spares or repairs' stirling engine of good dimensions, with nice bronze details and decided to bid. I could see there was no connecting rod between crank and displacer, and hoped there were no other major issues.

It arrived yesterday, and all seemed to turn freely, so I jury rigged some single core copper wire as a temporary displacer rod, retimed the piston rod connection to the crank collar which had come loose, and put a gentle heat under it on my gas cooker hob.

https://www.youtube.com/shorts/_Dt1QqYmvRs

She runs! After a fashion, but power output is very low.

Some dimensions:
Overall diameter 9" Total weight ~5lb
Flywheel 8" x 1/8" aluminium with bronze balance weights.
Power cylinder bore x stroke 1.25" x 0.5"
Displacer cylinder depth 1"
Displacer stroke 0.5"
Displacer counterweight 2oz


I don't know how old this engine is, but the brown cylinder wall is turned hardwood. The shrouded bearings are certainly long postwar, but could be from the 70s onwards. Screwthreads are imperial BA.

So... I think this engine is going to need modernising if it's going to yield enough power to spin a fan blade with enough vigour to outperform my thermoelectric fan. I've yet to go inside the belly of the beast to see what the displacer is made from. Judging by the lift amplitude, it may be a pretty slim disc.

Maybe I can improve power output by making the cylinder deeper, and making a better insulated displacer which incorporates some wire wool around its perimeter for some regeneration?

Maybe I could replace the aluminium power cylinder with glass to reduce friction? The piston weight is currently enough to pull the flywheel round and sink from BDC to TDC in the bore. Would a lighter piston reducing reciprocating mass make much difference to performance?

I'd be grateful for any suggestions you could offer for improving this nice old stirling engine.

[Stroller]

Well I opened up the displacer cylinder, and found the remains of the expanded polystyrene ceiling tile displacer fused to the hot side in a melted mess. I'm guessing it had already been pre-shrunk by a previous owner before I finished the job off. That would be why the performance wasn't great then!



Does anyone have any info on how to calculate the optimal swept working fluid volume for a stirling engine with a power piston swept volume of around 10cc ?

Ideas for a high(ish) temperature capable, lightweight displacer are welcome. Wire cage with bonded beercan alloy core core and and steel wool periphery?

[VincentG]

That is a very nicely built engine with an elegant displacer drive mechanism. With the wood frame, I can't see it being meant for an open flame or even to sit on a wood stove.

These engines are pretty forgiving of a wide range of volume ratios. My own modified LTD ran exceptional on a tea candle with a displacer volume 9 times the power piston. If it were mine, I would make the displacer as large as possible while avoiding contact with the top and bottom plates.

Estimating the displacer is .5" tall and a chamber diameter at 6"(I assume it's probably larger still) with a stroke of .5", you have a swept displacer volume of 230cc, bringing your true volume ratio to 1:23. That is a perfect ratio for both low and medium temperature operation.

Note that the porosity of the displacer will change the true volume ratio. Tom has had great success with grill stone, but your steel wool idea is well tested and should work fine. For(under 250F) low temperature operation, nothing is better than xps home insulation foam.

[Stroller]

Many thanks Vincent. I'm just glueing up a temporary displacer with two crossed plys of corrugated cardboard and a facing of beer can aluminium while I gather materials for something more permanent. I do want to run this on my stovetop, so I'll manufacture an aluminium sidewall with high temperature gaskets too.

[Tom Booth]

There are various materials you could make a high temperature displacer out of. Since the displacer is already counterbalanced, weight is not really an issue.

Large sheets of pressed mica bonded with silicone are available as "cut to fit" wave guide cover replacement for microwave ovens, or you might be able to pull some out of an old toaster.

https://youtu.be/SQCxZ8XEMns?si=IS5mKrlWXTEnt2qf

Mica might not be a good insulator though.

It might be a good idea to back it up with some ceramic fiber

https://youtu.be/GY4-pgSI2VA?si=-jlKSr8ux7N7ssUx

The ceramic fiber board is an excellent insulator and comes in various thickness but is probably too fragile. The mica is strong and very rigid, though somewhat brittle.

A sandwich of the two together should make an excellent thin displacer. Though somewhat expensive.

I would NOT bond them together with supper glue as super glue does not hold under high heat.

[Tom Booth]

BTW, what are these protrusions that look like stacks of coins:

ultra_LTD_stirling.jpg (207.49 KiB) Viewed 229 times

[Tom Booth]

Many thanks Vincent. I'm just glueing up a temporary displacer with two crossed plys of corrugated cardboard and a facing of beer can aluminium while I gather materials for something more permanent. I do want to run this on my stovetop, so I'll manufacture an aluminium sidewall with high temperature gaskets too.

I'd avoid aluminum, especially for the sidewalls.

I'm sure the wood was used for the very low heat conductivity. Aluminum would conduct a LOT of heat between the hot and cold plates.

Stainless steel would be much better in the metal category, but something like ceramic would be better or glass is OK if you want to see what's going on inside. (Or just stick with the wood, which can be surprisingly heat resistant).

Aluminum would conduct heat across or up to the cold plate as well as out of the engine altogether to the surrounding air which all adds up to lost power.

That looks like a P19 "ultra LTD" or modeled after it, originally designed to run on extremely low heat source, like the warmth from a hand, or even a (cold) wet sponge to provide a temperature difference.

[Stroller]

BTW, what are these protrusions that look like stacks of coins:

ultra_LTD_stirling.jpg

They're solid brass finned heat exchangers, presumably to cool the top plate. Once my aluminium fan blade set arrives from China to replace the flywheel with, a breeze across them should improve the differential.

There are five more on the hot side. Not sure what they're for, but they certainly prevent the engine from running off someone's palm. Not that you'd hold a 5lb engine up on your palm for long anyway!

I get what you're saying about the wood being a good insulator. I'm thinking it might be an idea to silicone some beer can aluminium to the inside, to seal any air leaks from splits in the wood (there are a couple), and provide a smooth surface for the regenerator wire wool to run against. I'll leave a gap top and bottom to minimise heat conduction between hot and cold side.

I couldn't find a P19 "ultra LTD" on the net. Do you have any old brochures with information about them?

Thanks for the advice.

[Tom Booth]

...
I couldn't find a P19 "ultra LTD" on the net. Do you have any old brochures with information about them?
...

It was really just one prototype engine, I think. But had a counter balance similar to yours.

Some information and links in this post:

viewtopic.php?p=16558#p16558

And more about "dwell". The P19 had a dwell, or simple "lost motion" mechanism.

Having "cooling fins" on top AND bottom seems rather unusual. Maybe to help absorb heat? Makes me curious what the application for that might have been.

Maybe it ran backwards with its "feet" in a birdbath.

[matt brown]

Does anyone have any info on how to calculate the optimal swept working fluid volume for a stirling engine with a power piston swept volume of around 10cc ?

My response to similar post on another recent thread...
___________________________________________________

Here's the basic equation that is best considered parity among various PVT values:

(Thigh-Tlow)/Thigh = PPvol/DPvol

Going with your DP/PP=1.5 let's say DP=30cc and PP=20cc, so what's Thigh and Tlow ?

Remembering that T must be Kelvin, let's use conv'l Tlow=300k whereby we find that Thigh=900k as in:

(900-300)/900 = 20/30

If you play with this above equation thru a wide range of values you'll see why common LTD where DP/PP= 40 (or more) can run on such a small temperature spread (source vs sink). I'm not saying you must follow these proportions, I'm just saying that these are the relative variations (kinda an order of merit = 1 thing).

I've never seen this equation before, but suspect Senft has it in one of his books (he's a math guru). I found it last year during my deep dive on gammas (I'm mainly an alpha guy).
____________________________________________________

With your PP=10cc constant, then DP=100cc has parity Thigh=330 as in

(330-300)/330 = 10/110

vs some higher input values

(340-300)/340 = 10/85
(350-300)/350 = 10/70
(360-300)/360 = 10/60

Again, just parity values via PV=mRT

[Stroller]

Thanks Matt.
My PP/DP is 10/200. I guess this means my stirling engine is the wrong shape for the application I want to put it to.
Oh well, it' a fun experiment, and hasn't cost me much so far. I just lashed out another £20 on a non-contact tachometer and some old Meccano brass gear wheels. I'm thinking a 2:1 reduction for the final drive to the 13" 5 blade fan might be about right. I have enough gears to make it 4:1 if necessary.

[Tom Booth]

Thanks Matt.
My PP/DP is 10/200. I guess this means my stirling engine is the wrong shape for the application I want to put it to.
Oh well, it' a fun experiment, ...

I wouldn't abandon the idea before trying it out.

I have a fairly recent thread on here about converting an LTD to high temperature.


https://youtu.be/RSZV11MNxjI?si=YkQw4jD9UpLgpXIJ


That's mostly going on a tea candle, with intermittent help from a propane torch.

I'm pretty sure your much larger engine has the potential to do more than light a string of LED Christmas tree lights.

[Stroller]

Thanks for the encouragement Tom. Thats impressive output from your small LTD engine. Just awaiting some stainless steel wire wool to finish the edge of the temporary displacer now, but I'm away abroad from Tuesday for a fortnight, so I'll be returning to the project towards the end of the month. I also found a roll of .002" titanium foil in my tinkering box, so that'll make a neater job of lining the wooden cylinder wall to give the wire wool a smooth surface to contact without conducting much heat.

[matt brown]

Tom's right, it's still worth trying to improve. But here's the nickel tour you may be missing...

forced work PV plot_1.png (31.14 KiB) Viewed 178 times

Consider this a PV plot for common LTD where ambient pressure bisects work area. Let's forget about the forced work aspect and focus on this plot itself. Tom and others are suckered into LTD illusions since ambient pressure supplies compression force as if by magic. If you have ME background then you'll know that buffer pressure doesn't change work area, only where Wpos vs Wneg occurs. The funny thing about these PV plots is that they're totally bogus...just try to sketch a PV plot of ANY gamma like you would any other engine. Nope, gamma PV plots are a composite of both work volumes and here's the clue that there's some major fudging going on.

Let's put some real PV stuff under the microscope...

flag PV.png (147.1 KiB) Viewed 178 times

Consider some common notation where Tr = thermal ratio and Vr = volume ratio. This "flag" like plot is from the PV work app on Geogebra site (free online app) and shows a common alpha 300-600k cycle where Tr = Vr = 2. Note that pts 2 and 4 are equal such that we'll use this to represent ambient pressure when LTD. Now, if we extend the volume ratio whereby Vr = 4 but retain Tr = 2 we have this "banner" like plot

banner PV.png (150.32 KiB) Viewed 178 times

Yep, "Houston, we have a problem..."

Due to the PV relationship of isotherms, there's no way to have pts 2 and 4 on the same isobar except when Tr = Vr !!! And if we move away from isotherms to adiabats (Tom's current love affair) pts 2 and 4 spread further/faster.

If you really have some time to kill, simply spend some time messing around with this Geogebra app and your calculator vs that 'parity' DP/PP equation and you'll be scratching your head as to how these buggers 'run' at all. It's probably due to their out-of-phase dynamics that coincides with some voodoo thermo cycle.

[matt brown]

BTW my DP/PP parity equation is based off PV=mRT but common LTD does appear to be designed around Tr = Vr. If I ever chat with Senft again (been nearly 20 yrs) I'll ask him if that was his basis. Note how...

DP/PP = 20 on par 300-315k cycle
DP/PP = 30 on par 300-310k cycle
DP/PP = 40 on par 300-307.5k cycle