i realize that there is interest in an exhaust gas exchanger that elminates the need for periodic
cleaning (yes Jens i feel you pain buddy)
the exchanger that i built last year about this time, has been working excellently since in testing, but
i have not been able to do a couple things that i feel is necessary before i release the plan set.
while the unit works very well at near full loading, i need to do more testing at low and part loads
to determine is efficiency and whether or not it can maintain itself (self clean)
and most importantly i need to work out a formula for sizing the unit for specific enigne's, hp output, rpm,
exhaust temps, and the output the engine is likely to be operating under.
because of its design "one size" likely will not suffice for all engine's, and likely not for all load levels.
in order to work out the formula at least get a working baseline to build from i need some input from you guys
with other enignes.
1. what engine would you want an exchanger to work on? (listeroid, changfa, other)
2. how many cylinders?
3. displacement?
4. rpm
5. exhaust port diameter (in inches if possible)
6. max hp rating
7. intended power output range ( for instance a 6/1 listeroid running at 2.0 kwatt average, etc)
8. exhaust temp as near the port as you can measure
9. orientation you would like to use the unit in, horizontal or vertical?
the unit i have working on my 195 changfa which is run at 90-100% loading 1800rpm, 7 plus kwatt output
has an input range of 640 degree's F. and an exhaust outlet of ~230 degree's F. it is approx 40" long
and approx 7 inches in diameter (with insulation installed)
i would expect that a listeroid 12/2 would be very close to the same physical size if run at near full load is a requirement
and a 6/1 might be half the physical size in length,, in any case the diamter will remain the same whatever we drive it with.
the last thing i want to do is release plans for something that might not perform as well as it should because of me not
accounting for the variables and sizing the internal parts inappropriate to what the engine can deliver.
btw for what its worth, my unit has been proven to remove ~75% of the waste heat from the exhaust stream, while this
is not the most efficient perhaps the fact that one does not have to clean it is the tradeoff.
also keeping the exhaust temp over 212F output keeps down condensation issues with the remainder of the exhaust system.
any input would be useful in trying to work out the final plans
thanks
bob g
Add one vote for standard 6/1 listeroid at nominal 2KW load. If loads drop off for any period of time (say 2 hours) but then pick up, do you think this would load up?
I would vote for one 6/1 in size and a second chanfa/12/2 in size.
One of each would be perfect for me!!!!
Thanks Bob!
Steve
bswartz:
i haven't tried that in testing yet, at least with any real observation
i will put that on the list of tests to run.
my thinking is maybe i can simulate what you are asking by dropping the load on my changfa
to around a kwatt for an hour or two and then ramp back up to full load to see if there is any indication
of degradation of performance or increase in exhaust back pressure.
i don't believe it would be a problem handling that, however
it probably would not be a good idea to run the engine for long periods of time at low or no load
because of buildup of tars and oils internally? i just don't know at this point.
for the unit to self clean it must be allowed to cycle from cold to full operating temperatures with some frequency.
especially if part load operation is the goal for the majority of the time.
interesting questions, have to give it more thought.
thanks
bob g
thanks for the input guys
Jens:
i am thinking the sizeing of my exchanger is very close to being appropriate for your listeroid 2 cylinder
your displacement is quite large, comparatively speaking to the changfa, but the flow works out to about
a wash between the two,
if you can, measure the heat leaving the Ypipe betwen the heads, i would like to know the peak temperature
at that point if possible.
also i am thinking the operating window is sufficiently large to obtain a relatively flat efficiency curve from 5-9kwatt
output, the target of 7 kwatts would be a good place to design to, and allow some dip on either side i would think
would be acceptable.
thanks
bob g
i think that would be fine, in my testing allowing the unit to cool to ~75 degree's seems to be good enough
to allow for the thing to do it cleanup function on restart, it might be that even a higher temperature at cold
startup would work as well.
the units need for a clean cycle seems to be more a function of running at very light or no loading for a protracted
period, i don't think it would need a cleanup cycle if it were to run at over 5kwatts and higher most of the time.
your application i am not as concerned with, i think it will work very well with it, however
most of my concern relates to the 6/1 lopeing along at a kwatt output with an oversized cooling system
where the exhaust temps are so low and exhause flows are low as well.
that is where sizing is very important
bob g
270C is about 500 degree F.
at what loading was the engine running at at this temp? 5kwatts?
bob g
ok that makes sense now, for some reason i was thinking you had a 12/2
but in reality you have a 20/2?
is there anyway to measure the temp right at about 1" out from the exhaust port on each head?
what i am thinking is you have air dillusion coming in from crossover on the opposing cylinder mixing at
the Y pipe and cooling the exhaust a bit?
500 degree's F is probably pretty close though for a 20/2 at 7.5kwatt loading
your at about 75% loading on a 20/2 at 7.5kwatt, so 500 / .75 = ~660degree F (which is about what i would expect
the engine to be doing at full 10kwatt loading.
a bit more number crunching and i think we can get the sizing formula worked out to get you right in the sweet spot
for your application.
glad i asked the question, now that i think about it. this has been something that has been nagging at me for many months
the design is interesting, but the reality is i got very lucky in sizeing the various important parts of the exchanger for my application
now the importance of sizing is more clearly defined in my mind at least, and i think i have enough info to work out the rest.
the only thing left that might shift the efficiency a bit at this point is the coolant temperature you are running at (engine coolant
not exhaust exchanger coolant), so long as you are running at or near 195-200 i don't think there will be effect.
bob g
I have a 12/2Metro with a 7.5 st head. It's not in operation yet,so no run data available,sorry. With no hardware in place yet,orientation ie,horizontal versus vertical is not an issue. I could plan either.
As you know Bob, I've been hankering after one of these for a while now :) So I'll definitely vote for a 6/1 sized unit.
At the moment, I can easily provide low load testing - because I don't actually have a load hooked up to my engine. I need to do a bit of work with temp sensors & the like, but now I've got my lathe & can do screw cutting, I'm well on the way there.
How about this for an idea? I'm assuming that, if this heat-x works out as well as you're anticipating, that you'd quite like to make a commerical system - either planset or finished items. If that's the case, maybe it would be an idea to get some "beta testers" together, who can build a unit up & provide a set of running data. The test results can then be correlated & a decent stab can be made at predicting some generalised performance figures. I think that to cover all of the variables involved, one - or even just 2 or 3 - testers will not be able to provide sufficient variability; but 10 or 15 may just be able to do it.
Anyway, gotta go, lunch is calling. More thoughts later...
Very ambitious project this, designing a waste heat recuperator... technically, a gas/liquid heat recuperator...
Some random thoughts...
I'm going to wager that the liquid side is clean enough that fouling won't be an issue. It's the exhaust gas side that's going to be problematic.
Shell and tube design for ease of cleaning? Since the pressures on the exchanger are low, durable/inexpensive/highly efficient materials can be used - like thin steel.
On-line water or solvent flush, lowering the need for shutdown and disassembly for cleaning? Or - off-line water or solvent "soak" (fill it, let it sit, drain it?) Combination of both?
Corrugated tube for turbulence? Tube side gas turbulence = higher heat transfer efficiency, but at the cost of more diffficult cleaning.
Single gas pass or double pass - or more? Single pass = long, narrow exchanger with low flow restriction, while multiple pass = shorter, larger diameter, higher restriction.
Shell side baffles? Increases the liquid side turbulence, for higher efficiency, but harder to make and increases flow restriction.
Sized properly, the outlet gas temperature should not fall below the condensation temperature of anything in the gas, nor pass the boiling point of anything on the liquid side. Tough to do if the load on either side of the exchanger varies - flow rates or temperatures. But possible.
There are some basic control schemes that can help with size versus load matching over a load curve....
One can put a recirculation bypass on the liquid side - joining the outlet back to the inlet - so if the total liquid flow (heat demand) is high enough to drop the outlet gas temperature below 230°F, some warm outlet liquid can be mixed back in with the inlet liquid to decrease the heat uptake...
And a low liquid heat load demand (flow) can be dealt with by bypassing some exhaust gas around the exchanger, so the liquid outlet temperature doesn't get too high...
There's an "el cheapo" formula for fouling, that we used to determine when an exchanger's heat transfer had become so low due to fouling that it needed cleaning...
you needed some baseline inlet and outlet temps (and differential pressures, if possible, but not mandatory) - then you monitored the fall of the outlet temp on the "cold" side (in your case, the water side) and rise of the outlet temp on the "hot" (exhaust) side...
I'll see if I can find them, if anybody wants them...
Hi Bob
I'll cast votes for one each of a 6/1 and a design that would accomodate the Redstone engine, which it sounds like you have. It so happens that I have a Goodway heat exchange tube cleaning tool ( http://www.goodway.com/client_ucs/products/133-psm-500.aspx ). I am hoping to find a heat exchanger design that can be maintained with it. Sizing the tubing to their brushes would be my issue there.
Regards,
Will C
the design has no "tubes" to have to clean, and the prototype is fully welded and sealed so that disassembly
is near impossible without destroying the thing.
i am not at all sure there would be any benefit to building it so that is could be taken apart except for maybe making
the alterations internally to allow for operation under other parameters.
basically it doesn't operate anything like conventional heat exchangers do.
bob g
I have both a Changfa 195 and a Listeroid 6/1.
I vote for both. !
Veggie
So... even more ambitious than I first thought... a non-conventional exchanger design! I'll be very interested in seeing the proposed unit.
Every gas/liquid design I've seen where the gas is the "dirty" exchange medium is a shell and tube exchanger of some sort, even if it might be 5 storeys tall and look - and act - like a boiler. Cleaning takes many forms - on-line steam cleaners that rotate, water injectors, burners in the gas stream fired intermittently to burn off accumulated crud, rattles that spall off hard stuff - there are many solutions to this problem. But they all end up requiring disassembly and power-washing, at some point. At least on the tube side...
Jens:
thanks for the added info, if my back gives me a break tonight i plan on crunching some numbers and get a start
on a working formula that i can use to do some projections, then maybe this weekend i can test the projections against
actual measured results to see if the formula works.
i think the bandwidth is sufficiently large so that the exchanger can operate over a fairly wide range and still be within
the flat part of the efficiency curve.
much easier to build to a broad operating curve than one that is "peaky" in my opinion, especially since folks will want to operate
over a fairly broad output range from their engine.
this was not something i originally did much testing on, because i don't plan on running my unit at anything other than 90-100%
capacity, save for warmup after startup and cool down before shutdown.
bob g
Quote from: Jens on December 16, 2009, 05:22:02 PM
BTWE Bob, I took some more temperature measurements and it looks like we are closer to 300C but it's really tough to know exactly what is going on because I get different readings all over the place. The latest readings have me thinking that maybe cylinder #1 is not carrying it's fair share of the load. That is especially strange since #1 has always run slightly warmer than #2 and no modifications/changes were done to the fuel rack. The only thing different between #1 and #2 is that #1 has the high pressure fuel line heated.
Jens
Jens,
Seeing as you're taking heat readings....
Can I talk you into taking a reading 4 ft. from the exhaust port. How hot is the pipe ?
Thanks in advance,
Veggie
According to the "American journal of thermodynamics and food groups" Apples and Corn Flakes can in fact be compared. ;D
I look forward to your readings. (on the metal.....not the insulation ;) )
Veggie
Jens:
on your twin, does the oem mufflers screw into the head? or are there flanges that bolt on to the head
that the pepper pots screw into?
in any case, can you measure the id of one of the pepper pots on the threaded end?
from that i can determine the design port velocity from the head, this much i must know to size the exchanger
properly.
then if anyone else knows what the id of a pepper pot at the theaded end for a 6/1 that also would be useful
i don't need the id measurement down to .001" but within a 1/16" or about a mm, would be close enough.
but if i can get more accurate measurement i would sure use it.
got some work done tonight on a sizing formula, and a chart made up to input some test data to see if i can get
a relative correlation between test results and predictions from the formula.
i think the relationship between the internal spec's are related to port velocity and cylinder displacement factored against rpm
and the fuel consumption rate per hour.
what it appears as now is there is a compromise wherein one might give up some full load efficiency to operate at max efficiency at low loading, basically to get maximum efficiency at 1kwatt for a 6/1 i would expect to take a hit in effiency at full 3kwatt output.
the question is how much of a hit in efficiency? that i don't know, however
your 20/2 running between 5 and 9 kwatts, with an average of 7.5 appears to be a much easier to calculate and build a unit that is centered at ~7.5 kwatt, so that the efficiency might be off a bit at both 5kwatt and a bit less at 9 kwatt. i feel very confident in its ability
to operate well in that application. i think you would have good heat operating from ~60-90% loading capacity, rather
than the 6/1 running from ~33-90+%.
that spread looks like it might be a problem for the 6/1 application, does anyone know how hot the exhaust is at the about an inch out from the port running at 1 kwatt loading? if it is down much below 300 degree's F, i don't think any exchanger should be used in that range, you get to where you overcool and loading up the exchanger is the likely result.
not saying it couldn't be done, but would be much easier to work with between about 2 kwatt and 3.x kwatt output on a 6/1
basically i would have to be very careful in sizing for a 6/1 running at a consistant 1 kwatt load, and in that case it would probably be
a good idea to build another prototype for that specific application and put it into beta testing with someone running long hours.
i think i know of a board member that runs lots of hours with a 6/1, perhaps i should build one for that application and get it to him for testing. it would be something i would feel better about doing before a plan release for the 6/1,
just about all other applications i feel much better about and don't think it necessary to got to as much trouble to size properly
bob g
Bob;
I would need a range - for a 6/1, a 10/1 and a S195.....
John.
Quote from: mobile_bob on December 17, 2009, 02:59:08 AM
then if anyone else knows what the id of a pepper pot at the theaded end for a 6/1 that also would be useful
i don't need the id measurement down to .001" but within a 1/16" or about a mm, would be close enough.
but if i can get more accurate measurement i would sure use it.
Bob - the ID of a pepperpot for a 6/1 is a gnats cock over 2" (2.050" to be exact - but that extra 0.050" may be wear in mine).
so... a gnats cock is .050"? ;)
On average ;D
I believe the standard Lister and Listeroid exhaust flange is threaded 1.5 inch BSP.
Regards,
Ian
so is BSP approx 2.050 od, and about 1.5" id?
that would make sense to me.
bob g
"Although a few species are up to 1/2-inch long, adults commonly are about 1/16 to 1/8-inch long"
So .050 would be almost as large as it's whole body, that's a studly gnat!
Quote from: mobile_bob on December 17, 2009, 11:25:34 AM
so is BSP approx 2.050 od, and about 1.5" id?
that would make sense to me.
bob g
Not on my setup it ain't....
I measured 2.050 ID on a pepperpot exhaust; the threads are on the outside of that. I've locked the workshop up now & it's too cold & late to go back in there & measure the OD. I can do that tomorrow.
Do you need the ID of the flange bit as well? Or instead of? I'd say that was less than 2"; probably the 1.5" that Ian quoted.
i suppose what would be of most use would be the most restrictive area in the system,
1.5 inches seems more appropriate for a 6hp engine than a 2" id
even 1.5" is really too large for a 6/1 in my opinion, but that ok, if thats the way it is.
thanks guys
bob g
Quote from: mobile_bob on December 17, 2009, 11:25:34 AM
so is BSP approx 2.050 od, and about 1.5" id?
that would make sense to me.
bob g
Bob, FYI ... I screwed a 1-1/2" NPT nipple directly into the threads of my exhaust flange. (Model 6/1)
Veggie
Jens,
Wow! 480 degrees f ....Much higher than I thought it would be.
Veggie
Bob
On my 6/1:
Exhaust port. 1.47" dia
Flange ID. 1.74" dia
Peppercan inlet pipe ID. 1.68" dia
Peppercan inlet pipe OD. 1.88" dia
But by far the most restrictive part of this exhaust system are the twenty .275" outlet holes on the peppercan. These 20 holes have a total area of 1.19 SQ/IN. As a comparison, the 1.47" diameter exhaust port has an area of 1.69 SQ/IN. 1.19 SQ/IN equates to a 1.23" ID pipe...
Jens,
I agree with all your points. I wanted to see just how hot the gases can get that far from the head.
It suggests that your insulation is doing a very good job.
Something we all need to be aware of as we progress with Bob's exchanger ideas.
Thanks for doing that,
Veggie
for the exchanger to operate at its optimum, it needs to be mounted as close to the head as possible, the pipe feeding it
ideally will be insulated, so that the maximum heat can be delivered to the exchanger at the highest velocity consistant with
that of the exhaust port.
Ron:
thanks for the specific's, that info provides all i need to know to work out the sizing requirements.
those twenty holes equating to less than the area of the exhaust port likely is still larger than the area of the port
at the valve/seat at full valve opening.
back to crunching numbers here.
thanks guys
bob g
Bob,
I can understand your need for temperature but diameter seems a bit superfluous? The important things appear to be gas volume per unit time and available temperature drop. Your exchanger can modify the exhaust gas speed by simply having a reducer or increaser(?) before the actual exchanger?
Regards, RAB
RAB:
i have a different design philosophy, which mirrors conventional "pulse" maniflold design
in that you do everything you can to keep the exhaust gas velocities as high as possible
without adding (appreciably) to back pressure.
by increasing diameters one increases surface area, which in turn allows for moreheat to escape
than would be the case with smaller amounts of surface area. you also set up standing waves that reverse
back up the pipe with each step in diameter in the exhaust runner, if i don't have steps and keep a consistant
runner diameter with that of the port it is one variable or set of variables i don't have to contend with in my
opinion.
i want the highest available heat, at the highest veloctiy to enter into my exchanger, for reasons that
will become apparent once you see the design.
my thinking is that there is 6hp in the exhaust of a 6/1 (running at peak load) and i don't want any degradation
of that power if i can help it, i want all 6hp presented directly into the exchanger. the closer i can get to this ideal
condition the more efficient the exchanger is at getting the hp converted to btu input to the water, and the better
able the unit is in ability to keep itself clean.
there is also cylinder scavenging, which is generally always desirable, and particularly so where we have something downstream
that might ordinarily present back pressure. while the effect might very well be slight, it appears to be enough so to offset any
pumping losses associated with the use of the exchanger.
that was one of my concerns in early testing, it is all good to harvest heat so long as there was not a penalty in fuel economy
that made it more costly than need be. as it turns out there is actually a slight improvement in fuel economy using my exchanger
on my changfa 195 over that of the oem muffler, and the unit turns out to be a better muffler than the oem to boot. this might be
explained by the obvious standing wave and reversion provided by the oem muffler making scavenging not as complete (which is
theory i realize, but until someone can find another explanation it seems as plausible as any for now).
my changfa port size is right at 1-5/8" so i chose schedule 40 black pipe that was 1-1/2 because it measures out to be right at
1-5/8". i port matched my flange and the pipe has one 45 degree elbow that is a nice smooth bend, the end result is a system that
works very well, so i am hesitant to continue to work the design by introducing other variables that might cause unforseen issues
that might be hard to sort out later. if that makes sense?
i just want to continue the project building on what works, produce the plans for distribution, and later should someone that has a set of plans decides he knows better and makes alterations for whatever reason, well... he can either be the beneficiary of the improvment or suffer with something may not work as well as it should. my responsibility is to try to make it so that if someone buys as set of plans and assembles an exchanger by following the directions he will be able to get the same results i have.
ymmv
bob g
I'm all excited to see the results! Do you have an ETA for release of plans, and a approximate MSRP? :)
Quote from: mobile_bob on December 17, 2009, 09:06:49 PM
for the exchanger to operate at its optimum, it needs to be mounted as close to the head as possible, the pipe feeding it
ideally will be insulated, so that the maximum heat can be delivered to the exchanger at the highest velocity consistant with
that of the exhaust port.
Ron:
thanks for the specific's, that info provides all i need to know to work out the sizing requirements.
those twenty holes equating to less than the area of the exhaust port likely is still larger than the area of the port
at the valve/seat at full valve opening.
back to crunching numbers here.
thanks guys
bob g
Yep, those were the numbers, along with bore and stroke, that I was also interested in when I was crunching these same numbers.
Yes, you are probably right about the peppercan hole area being larger than the actual exhaust valve exit area.
IMO, probably the most important area in design of an exhaust exchanger, is to have adequate volume in the first chamber to completely absorb all of the cylinder volume, sent there over the course of 1/4 of a cycle(exhaust stroke). After that you have the remaining 3/4's of the cycle to allow the exhaust gas to pass into the remainder of the heat exchanger. In my planning, I was aiming to have at least the same passage area, if not more than the existing exhaust port and peppercan exit area.
Ron:
the prototype worked out to well in excess of displacement, looking back i could have done with approx 3x displacement, maybe.
but yes that was one of several factors that went into the design.
bob g