Prettiest FWE (?) - the 'Lady Anne Boleyn'

[milisavljevic]

My own feeling is that it is impossible to "unfold" or "unpack" classic FWE type engines (or Chinese, or Thermojet,
or concentric engines, or ...) in a way that makes a linear model that is truly equivalent.

Hello Larry --

We are in agreement on this, although I am confining my comments to UFlow1D.
Someone may yet write software that can describe this class of pulse reactor. ;-)

Best regards,
M.

[milisavljevic]

Wow! Go away for a couple of days and look what happens! I want to run your new layout through the model;
however, someone else is ahead of you in the queue. I hope to be able to model the revised layout Saturday.

I can hardly wait!

Hello Larry --

I have not forgotten you, but I am still chewing on a particularly tough cud* - a tasty pulse-switchgrass hybrid. =)

M.

*figuratively speaking ;-)

[larry cottrill]

All right, at last - here it is, for better or worse. M, I can hear you chuckling already ...

As already said, the initial gas temps are laid out pretty much intuitively, based on examples that have been used before. Error perpetuating itself. This is still the 23mm diameter intake pipe, incidentally - so the model is supposed to represent the Rev 01 plan as I actually posted it earlier, before some additional suggestions were tried.

Note in the last graphic (the "running out" picture): the sine-like figure of pressure at the front chamber plane; the nearly flat velocity line at that plane; the roughly sine-curved intake velocity; the near-perfect phasing of intake and exhaust planes in the velocity and mass flow curves.

To me, the timing looks about as good as it gets. Of course, it means nothing if the layout of temps turns ot to be way off. It seems to me that the intake temps are really critical in most engines, and that the exhaust megaphone temps would be pretty critical in this particular design. But, I'm just speaking from intuition, not knowledge.

L Cottrill

[milisavljevic]

All right, at last - here it is, for better or worse. M, I can hear you chuckling already ...

Hello Larry --

If I had wanted a chuckle at your expense, I would not have offered my two cents, whatever they are really worth.
Chuckles usually come from those folks sitting on their hands, away up in the bleachers. ;-)
Question: what temperature scale are you using here? (°C, °F, or °K)

Best regards,
M.

[larry cottrill]

Question: what temperature scale are you using here? (°C, °F, or °K)

Now, you caught me chuckling, to wit: If the numbers don't speak for themselves as to the intended units, they must be just a bit farther from reality than I imagined ;-) Why am I not surprised?

They are meant to be Kelvin temperatures (the only unit provided by the program), with standard outdoor air usually being represented as 380 or so (Edit: Mis-typed that; should have read 280 or so, though at the moment that seems to me a little cool - not much above freezing. Seems like 300 might be a better round figure for "standard" air that could at least be considered "comfortable").

I have the distinct feeling that major adjustments lie dead ahead ...

L Cottrill

[milisavljevic]

Question: what temperature scale are you using here? (°C, °F, or °K)

Hi Larry --

I missed one temperature option: °R! I had completely forgotten about the lack of choice in UFlow1D; °K it is.

I have the distinct feeling that major adjustments lie dead ahead ...

At first blush, your numbers do not look too unreasonable. More later.

M.

[larry cottrill]

Dear M -

We have seen recent anecdotal evidence that the intake gas temps are VERY cool averaged over time - particularly, it was reported that a fingertip could be placed in the visible intake flame very close to the flare without sensing uncomfortable temperature. I myself have tried the simple experiment of placing a match head in the intake and tail flames (this was with an FWE engine running at a low-to-moderate level - intake flame clearly visible but invisible tail flame under near-darkness lighting conditions). The match head did not ignite even when fully immersed in the intake flame, but fired instantly when placed behind the tailpipe exit (even though there was no visible flame there). Extremely interesting, I thought!

So, a few minutes ago, I did a quick test modifying the Rev 01 model with the last (rightmost) three Initial Gas Temps in the layout reduced to 400, 350 and 300 (from dome end to inlet). There was NO discernable change in the shape or timing of the curves, BUT there was one difference: The vertical scale in the Mach Number graph was altered to fit a slight INCREASE in the intake air speeds. Again, the SHAPE of the intake curve was unchanged - just a few percent increase in magnitude (maybe 3 or 4 percent)! Astonishingly, the intake Mass Flow curve was practically unaltered.

Interesting, don't you think?

L Cottrill

[larry cottrill]

From a few days back:

I know you have your reasons for moving the intake transition forward, but this would seem to reduce thrust.

M -

I remembered you said this, so I tried playing around with the intake pipe length and (only SLIGHTLY adjusted) location of Eric's 'Advanced FWE'. It is within a few mm of the very same layout. It runs out exactly the same as my longer, more forward one, and produces almost EXACTLY the same graphs, believe it or not. So, it is another layout that looks just as good, but no better (at least, as far as UFLOW1D can determine).

Again, I think an interesting result. The shorter pipe must have a higher natural frequency, but perhaps its lower Q plus the FWE's "ability" to suppress the upper harmonics takes away any advantage from what otherwise might be a "better" location.

L Cottrill

[milisavljevic]

Interesting, don't you think?

Indeed! I will prepare an alternate temperature profile for you, but please be patient...

Thanks!
M.

[larry cottrill]

For everyone's further amusement, here is an alternate method of modeling the unfolded front-end, borrowed from what I did to model an earlier version (straight-tailpipe) FWE (probably the Smooth Lady). The 'transition' is modeled as a reflected segment of the front of the chamber, and a 'flat plate' is placed to hold the intake pipe at the appropriate distance.

At a glance, this looks like it would be a poor representation of the modelled engine because of wave reflection from the "shoved forward" front plate; in reality, the curves it generates are almost identical to what has been shown above, except that the intake velocity is several percent greater in magnitude. I think this comes from the effective increase in chamber volume, which would provide a little more power for the same chamber pressure. So, it is probably less realistic than Graham's smoother "transition cone" method, though it does produce reasonable results (good running out, velocity node still in the right place, etc.).

The reason I first tried this is that I thought it would model the intake more realistically (since in the real engine I was originally modeling, the pipe broke sharply into the chamber; i.e. there was no physical 'transition' segment). In the present case, however, it turns out to offer no real advantage. It is interesting that it produces very similar results in the shape of the graphs; this was not expected.

I think I would still advocate this method for any FWE that has 'Thermojet' style "transitionless" intakes penetrating forward into the chamber. Elsewhere, Eric (perhaps among others) has explained his dislike for such intake geometries, which I used in the 'Smooth Lady' and 'Sveldt Lady' FWEs as well as in my Fo Mi Chin II (FMC re-design).

L Cottrill

[larry cottrill]

I decided to re-model the engine with my original "curved pipe" intake, using the dimensions based on M's idea of what part is an "intake" and what part is a "transition". This is certainly a better model (because it's based on a better mental picture!) of the intake than what I had done originally. The effect is that the transition cone shown earlier is lengthened and the straight intake pipe is shortened (but not enough to completely compensate - the whole engine model is just a few mm longer).

The result was exactly like the version above (the one where all the graphs are shown) except that the intake rim Mach Numbers are just slightly higher (maybe 2 or 3 percent). M is undoubtedly right, though, that a 'welded pipe' intake should in reality be better than a 'smoothly curved pipe' intake, since the tuning of the straight section will be a little 'sharper'. This subtle "smoothness" difference cannot be brought out in the simple UFLOW model being used; I was only interested in what the length and location changes might show us.

L Cottrill

[hagent]

Dear M -

We have seen recent anecdotal evidence that the intake gas temps are VERY cool averaged over time - particularly, it was reported that a fingertip could be placed in the visible intake flame very close to the flare without sensing uncomfortable temperature. I myself have tried the simple experiment of placing a match head in the intake and tail flames (this was with an FWE engine running at a low-to-moderate level - intake flame clearly visible but invisible tail flame under near-darkness lighting conditions). The match head did not ignite even when fully immersed in the intake flame, but fired instantly when placed behind the tailpipe exit (even though there was no visible flame there). Extremely interesting,

Hi Larry,

I just had an idea about why the intake blue flames don't appear to be hot.
I'm just pulling this out of my *&! so I'm probably wrong.

All the different wave lengths of light that are created in the burning process by electrons dropping into specific orbits. Blue light which has a higher frequency will drop more energy levels than an IR wave lenghth.

So who is to say that all the electrons are dropping into their orbits in a random fashion? Is it possible that the IR light is given off first and then some time later the remaining electrons take a bigger jump and emit the blue photons in a different location within the PJ? Hense we see some come out of the intake, but we don't feel the IR photons because they already released their energy and were absorbed by the gas or CC walls.

Does this make any sense at all?

Fun to think about.

Cheers

[larry cottrill]

Does this make any sense at all?

I don't know. Back in my days at Drake University, when I was preparing to become a world-famous astronomer, I got to study things like that. But, Physics, even at that level, was simpler in those days. Chemistry, too - we had about three kinds of valence holding molecules together. Now, my 23-year-old son can explain about seven or eight variations of valence to you, at considerable length. Even the complex paths of electrons around the atoms in a molecule have been modeled with Physics animation tools. Fascinating stuff, which I have no practical way of keeping up with.

Fun to think about.

Yes, it is.

L Cottrill

[larry cottrill]

I requested Durawear Corp. to give me an idea of feasibility and ballpark estimate of setup costs for molding Lady Annes out of this hot cast material:
http://www.durawear.net/ceracast.html

I got an answer back yesterday that they felt it was too small and the dimensions too critical for production. Note that this would have to be cast as a mighty thin shell to keep the weight down. Would have been interesting, though - look at the properties of this stuff!

L Cottrill

[Graham C. Williams]

The numbers are a percentage of corrected length. This was not one of the better models but interesting.

Graham.