Wilcox Caloric, 1860, open-cycle, regenerator

[matt brown]

staska - very nice work !!!

list - staska has provided predictive results from computer simulations which consider various real values, but a tad hard to read and grasp if unfamiliar. Meanwhile, my crude graphics are rather limited and idealized via distinct events with a linear piston motion (when applicable). My method tends to exploit max ideals (despite unrealistic) and is usually devoid of any "reduction to practice" or optimization (competing issues).

Most computer simulations generate information overload where data can be inconclusive. When staska favors hot PP vs cold PP gamma (for various reasons), I wonder why not simply move to alpha?

alpha vs hot PP gamma.png (26.45 KiB) Viewed 2982 times

Here's the previous hot PP gamma vs typical alpha within similar values. During expansion process, the alpha has 2x MEP of gamma but only 1/2 the expansion "stroke", so crude 'indicated' work is equal. However, note Pmax for alpha indicates alpha will be a torque monster vs hot PP gamma when piston/slider-crank due to bellcrank effect. I think this one graphic proves that if you can tolerate hot PP engine, simply go all the way and go alpha.



spinningmagnets - did you download this Andy Ross book ?

http://www.stirlingengineforum.com/view ... 337#p23337

[staska]

Matt - alfa is always limited to high temperatures/ideal mechanism. Not 600/300, but 900/300 region.

Reason - compression piston consumes 70% of expansion piston made power. And all this is diluted by piston side pressures etc.

Where gamma or beta have only power extracted and returend to same piston. And beta of course is an excersize to minimize friction in complex mechanism. Now, and in next 1000h of running. Many tried it, fraction succeds :)

Matlab is easy. Much easier than try to machine some part.

[matt brown]

Matt - alfa is always limited to high temperatures/ideal mechanism. Not 600/300, but 900/300 region.

Reason - compression piston consumes 70% of expansion piston made power. And all this is diluted by piston side pressures etc.

Have you read the Andy Ross book "Making Stirling Engines" ? (link to free download was recently posted)

An alpha is more sensitive to dead volume than gamma and alpha tends towards Otto more than Stirling where adiabatic processes are restrained by low volume ratios. Here's a very interesting PV from Martini's "Stirling Engine Design manual" (I recently posted 2nd ed. link)

Martini pg 63.png (54.46 KiB) Viewed 2947 times

Note 2:1 volume ratio and 300-900k cycle. Assuming time constraint for heating, adiabatic processes will restrain a hot air engine relative it's volume ratio where the major PVT variable is monatomic vs diatomic gas.

I've studied this stuff for decades and divided major issues into order of merit schemes. Oddly, the most important issue passes 'under the radar' and is merely the volume ratio: no volume ratio = no output. Regardless of thermal ratio, typical piston SE are constrained ~2:1 volume ratio when continuous motion drive and this further constrains most other design issues. A discontinous "dwell" scheme may resolve some conventional SE issues, but this does not resolve inherent regen issues where regen 'load' (regen/input) dominates most Stirling cycle issues.

Where gamma or beta have only power extracted and returned to same piston. And beta of course is an excersize to minimize friction in complex mechanism. Now, and in next 1000h of running. Many tried it, fraction succeeds :)

Do you have a novel gamma ???

Matlab is easy. Much easier than try to machine some part.

My graph paper is easier than Matlab and gives me insight in small visual bits vs blinding avalanche...

[staska]

Have you read the Andy Ross book "Making Stirling Engines" ? (link to free download was recently posted)


Do you have a novel gamma ???

My first stirling which do run and produced some power. But hard to assemlby and harder to modify - too many places with silicone sealant. viewtopic.php?t=5223

But all was build using manual mill and lathe in a month or so.

Now, after some upgrades to vmc, doing another engine with more stroke and better balanced. And hopefully with any silicone sealant :)

On books and theory thing - yes i have read Martini books, and Andy Ross, But best was Alan J. Organ books - Stirling Cycle Engines: Inner Workings and Design. Costly, but it is a must have. Plus some private conversatio with Allan itself. And some good works Of Urieli, on stirling engine similarities / simulations.

https://content.e-bookshelf.de/media/re ... cd1521.pdf

https://cread.jd.com/read/startRead.act ... readType=1

So - any gamma or beta could be shown as analogue alpha and vice versa. Just true alpha will always have 0 added dead volume, while gamma have it always. Beta could negate this by having power piston in displacer volume. But on kinematic side alpha always loss - compression piston only take power, expansion only produce it at some portion of revolution. And all extra load goes to our bearings / piston sides etc. Same was shown in Andy works on making V twin engine in same casing - extra side load caused much less power output, compared to Ross linkaege which is close to zero side load of pistons.

[VincentG]

Beautiful shop.

[staska]

Beautiful shop.

Thank you.

[Jack]

Yeah that is a dream shop! Damn

[staska]

Do not be afraid of my workshop :) back to wilson engine. What i was able to show - extra volume do cost extra loses. Can we negate them other ways or ?
Can we add simple valve in single acting hot gamma? Or only twi hot gamma ?

[matt brown]

On another thread, I'm scheming a gamma (cold PP) proportional to Ian's legacy guidelines where DP/PP=1.66

It appears that Andrew Hall used Ian's proportions with PP=400cc and DP=600cc, so the following graphics compare an Ian/Hall Stirling with cold PP=400cc and DP=600cc vs Essex with hot PP=400cc but DP=400cc.

cycle A 300-450k.png (17.93 KiB) Viewed 2765 times

cycle B 300-600k.png (17.89 KiB) Viewed 2765 times

cycle C 300-900k.png (17.55 KiB) Viewed 2765 times

I used an esoteric m values for gas mass which is easy to follow thru processes BUT kept proportional to both DP volumes whereby all cycles have 1 bar charge when cold. All frames 1 and 4 are boring, all the action is comparing frames 2 and 3 between hot and cold PP vs thermal ratio.

The main takeaway is dT vs dP (thus dW) where cycle B has Thigh +33% Thigh cycle A, yet only increases cold gamma dP at S-3 by 10% vs cycle B increases hot gamma dP at E-3 by 33% (similar an alpha). This difference increases with thermal ratio whereby cycle C has Thigh +100% cycle A, yet cold gamma dP at S-3 only increases by 33% vs hot gamma dP at E-3 by 100%.

So, at least within these proportions AND these thermal ratios, the hot PP gamma Essex model follows the thermal ratio directly while the cold gamma Stirling model falls behind.

[matt brown]

Does anyone see why the cold gamma is "falling behind" as the thermal ratio increases ?

staska - computer simulations have their place, but this is the type of stuff that computers miss

BTW awesome shop !!!

[matt brown]

On another thread, I'm scheming a gamma (cold PP) proportional to Ian's legacy guidelines where DP/PP=1.66

correction...on another thread I'm using DP/PP=1.5 where the compression ratio is 1.66

IOW the DP/PP volume ratio is NOT the compression ratio despite many consider it so...