4 replies to this topic

[jimmason]

any ideas out there for how to make a pulse jet robustly not work?

we're having lots of problems with thermoacoustics now that we've put a tall stack on our gasifier swirl burner.  the swirl burner is 5" in diameter, 10" tall.  we've added a 4 foot or so stack of the same 5"
diameter.   it looks like this http://www.gekgasifi...hread.php?t=501  (though this one without the joint at 10")

when run, this howls rather loudly.  if art, it would be fabulous.  if a genset, not so good.

darrel tonight explained that we have made a rijke or sondhauss tube. or really, an unintentional pulse jet.  read about them here. http://en.wikipedia....wiki/Rijke_tube.   scroll down 3/4 to see
sondhauss close bottom tube, exactly as we have.

this apparently can happen with any pipe of one or two open ends, with a big temp difference between the ends.  a great enough temp diff between the ends alone will set up the standing wave.  you can make
the effect more pronounced with heat exchange meshes inside that augment the pulse in one direction, but not the other.

turns out the best place to site a mesh, or apply your heat source is about 1/4 length of the tube. (at least with the two end open type) hmmm, we put our always on silcone nitride igniter right about the 1/4
length point.  didn't think of that one.  no idea if it is enough heat to be making a difference in the resonance though.  we should try turning it off and seeing if the resonance changes.

an important thing that makes them work is the push and pull against atmosphere.  this suggests that one way to "detune" the tube would be to put holes in the sides around the top portion of the tube.  make the pulse hit atm at difference lengths, and thus scatter the wave, like one does in a car muffler.

or, maybe a flare outward would work?  but that seems only only help impedence matching, and make the trombone louder.

we put criss cross baffles down the tube tonight.  that helped a bit, but didn't solve it fully.

this pulsing is also likely why we have flame backing up into the mixing section/blower/etc in manners we never have before.  i think we did most our testing on the blower and flame arrester, integrated flare/exhaust with just the short stack.  the big resonance problem did not start until we put on the 4' extension.  we we need the long stack so we can keep the start up co away from the operator.

i wonder if we could insulate our way out of this problem.  insulate the stack from bottom to top, even out the temp differentials so it can generate less sharp thermal pulses.

other ideas for how to make it not work?

jim

[mikeotter]

Jim,
in my old house you seen ( not the dive I was living in) I had set up a radiant hydronic heating system with a boiler in the back yard shed.  I didn't us a damper control so the pipe was solid going out the stack. When I cracked the door it would oscillate violently and literally shake the dust off the ceiling in the boiler shed.. I tried different length tubing and got a different frequency....  I was about to give up and decided to try a draft corrector... This worked !!   You have to make a break in the front end of the pipe and deal with it there. They make Stirling engine refrigerators  with this same effect. Good luck !
OOE

[pgt7787]

1. get rid of swirl - just a tee will do
2. narrow the pipe to 50mm
3. put a little cone shaped cap above the end.

or

1. same
2. same
3. put a 45deg angle section for the last 300-400mm of the pipe

GT

[jimmason]

completing the circle on this question from a few months ago, here's
what we finally found solved the problem- insulation.  we found that
if we simply insulated the stack from bottom to top, eliminating the
hot/cold delta, the thermoacoustic effect went away.   more reading in
the rilke / sondhaus tube effect clarified that it needs this temp
differential, and if you eliminate the temp differential, the effect
stops (which it did).

this was one of the many good suggestions offered here.  we didn't try
all of them.  but some of note.

1.  cut a V in the top of the stack like done to quiet some pulse jets
2.  drill holes towards the top, or other shapes to soften the
transition at the end to reduce the pulse wave
3.  put a floating "deresonator" plate in the tube
4.  angle the tube, or otherwise put angle plates in it to break up
the standing wave.
5.  drill a hole in the side down low and put a rod across the flame path.
6.  experiment with expansion chambers like a two stroke exhaust,
except with the intent of "detuning".
7.  mike suggests a draft corrector.  what is that mike?

i don't know which of the suggested will work.  we ultimately decided
the easiest one was the insulation and tried that.  it worked, so we
stopped there for now.

as usual, thank you for the input.

and btw, what problem of physics do gasifiers not touch?

J