Injected Diesel 56cc 2 Stroke, Will it ever work?"

[Lloyd-ss]

Joe, sure, I'd love to take a look.

 

[ajoeiam]

Joe, sure, I'd love to take a look.

Hmmmmm - - - - what is the best way to send such to you?
Do I use a private message (not sure about official rules and don't want sanctions!!!) or do I send them directly to you?
(You could send me your email addy in a private message - - - - maybe?)

 

[Lloyd-ss]

Joe. Conversation just started.

 

[Lloyd-ss]

The 6 lobe roots blower seems to be working ok. The inlet opening was enlarged for improved airflow, and it appears to have worked. I had to cobble up a crude locking collar for the "driven" gear so that the gears could be precisely timed with the rotors. The plastic gears and the teflon rotors, locked together as they are, do not play well with each other. The 4-40 cross bolt is secured into the shaft, but the gear is slotted to allow for approx +/- one half tooth of adjustment. Hand-fitting was tedious, but lessons-learned from the failures paid off.

I also ran a functionality test. Running at 1,510rpm, it filled a 125 liter plastic trash bag in approx 4 minutes, or 31.1 liters per minute. 31.1 liters divided by 1,510 revs = 20.6 cc/ rev. Calculations for the blower are 29.2 cc/rev (with no losses), so I guess 20.6cc/rev is reasonable for an actual output. Unfortunately, if the diesel cylinder volume is 56cc, that would require at least a 2.7to one overdrive to purge the cylinder every revolution, which might not be practical. Forward progress is slow, but the learning is healthy.

 

[Lloyd-ss]

The Latest Version of the Roots Blower is WORKING !

Finally! I have been around the world twice on this and ended up back with the original 3-lobed design, but totally revamped. I ended up with better bearings and gears, and tighter clearances. But that still wasn't enough. The ratio of the clearances to the effective swept volume was still too much to stop the "slippage" between the rotor lobes. See post #75 for more about this.
The last set of 3 lobed rotors are made out of Teflon so that the clearances can be kept to a minimum without galling, but they ALSO had to be made somewhat flexible such that under centrifugal force, they would expand in such a way that the mating rotors would conform to each other, and seal to each other, without binding up. This was accomplished by drilling larger thru-holes in the 3 lobes to make the walls of the lobes thinner in certain areas so that they had flexibility where needed, and rigidity where needed in other areas.

Here is a quickie video that I made of the roots blower powered by a little drill press at approx 2500 rpm. The tubing filled with blue tinted water serves as a pressure gauge. The Roots blower is a positive displacement blower that moves air, but doesn't build significant pressure (unless modified extensively). At most, the blower would build up maybe 20" of water. BUT, it pumped a lot of air.

I fastened a 125,000 cc plastic trash bag to the outlet of the blower and it filled in about 55 seconds, running at 2500 rpm. Previous versions had taken from 2 and half to over 4 minutes.

If we just say that the blower output 125,000 cc's in one minute, and divide that by the proposed 56 cc displacement, that is 2232 cylinder fills in one minute. Since this is for a 2 stroke, the blower rpm of 2500 might support an engine rpm of ..... maybe 1500?? Hard to say.

Here is the amateurish video, but I think it gets the point across. Honestly, I was a bit surprised at how well it filled the plastic bag.

 

[aka9950202]

Fantastic work. Amazing to see the bag fill and build pressure.

Cheers,


Andrew in Melbourne

 

[Lloyd-ss]

Fantastic work. Amazing to see the bag fill and build pressure.
Cheers,
Andrew in Melbourne

Andrew, thanks for the words, I appreciate that. And Steamchick offered encouragement several posts back with "perseverance." Yes, it paid off.

Here are a couple of detail pics. The end plates are pinned into place and each has 2 flanged, sealed, ball bearings positioned in flanged cups. Each cup is bored slightly eccentric to give some adjustability in where the center of rotation for each rotor and gear is located. Because one rotor is "driven", and one "follows", huge amounts of fiddling was required to get the clocking and alignment and clearances correct. But no matter what I tried, the air "slippage" made the output of the blower much lower than it needed to be. I was ready to go to another type of air pump, and was thinking about vane pumps and how they maintain continuous contact as they rotate.

For ref, the right hand rotor in the pic rotates CW and the left rotates CCW.

So I thought I'd take a risk (pass-fail modification) and using a utility knife, cut all the way down thru the trailing-side of each rotor lobe. My thought was that the centrifugal force would sling the weakened rotor out enough to maintain contact throughout rotation. It didn't exactly work, so I thinned the internal wall of the leading-side of each lobe to weaken them to add some flexibility. That worked. A lot of hours in this madness.



This pic shows the PVC adapter installed onto the outlet of the blower with 2-faced tape. The small tube coming out of the outlet adapter is a pressure tap that goes to the water-column pressure sensor.

 

[Lloyd-ss]

Will the Blower Pump Enough Air?

Here are some thoughts and questions about this Roots blower.
First off, we know this application works in real life in the 100's of thousands of 2-stroke Detroit Diesels that were produced. I remember here in the US, that the Greyhound buses would always go flying by on the interstate, with that familiar blown 2-stroke sound.
I know this blower pumps air. The 125,000 cc plastic bag was stretched tight after just over a minute of pumping at 2,500 RPM.




Here is the inside of the blower and each lobe space calculates out to 11.3 cc's, and one complete revolution should pump six lobe spaces, or (at 100% efficiency) 67.8 cc. At 2500 rpm, that is 169,500 cc's. Does that mean the blower is 125,000/169.500= 74% efficient. That is extremely hard to believe, but maybe since the pressure only builds up to 15 or so inches of water, maybe that is true.



So here is the real question(s). How many revs of the blower would be needed to purge a 56 cc cylinder? And how about a method to test to see if it will actually purge 56cc while at operating speed? This is a head-scratcher.
Lloyd

 

[a41capt]

The Latest Version of the Roots Blower is WORKING !

Finally! I have been around the world twice on this and ended up back with the original 3-lobed design, but totally revamped. I ended up with better bearings and gears, and tighter clearances. But that still wasn't enough. The ratio of the clearances to the effective swept volume was still too much to stop the "slippage" between the rotor lobes. See post #75 for more about this.
The last set of 3 lobed rotors are made out of Teflon so that the clearances can be kept to a minimum without galling, but they ALSO had to be made somewhat flexible such that under centrifugal force, they would expand in such a way that the mating rotors would conform to each other, and seal to each other, without binding up. This was accomplished by drilling larger thru-holes in the 3 lobes to make the walls of the lobes thinner in certain areas so that they had flexibility where needed, and rigidity where needed in other areas.

Here is a quickie video that I made of the roots blower powered by a little drill press at approx 2500 rpm. The tubing filled with blue tinted water serves as a pressure gauge. The Roots blower is a positive displacement blower that moves air, but doesn't build significant pressure (unless modified extensively). At most, the blower would build up maybe 20" of water. BUT, it pumped a lot of air.

I fastened a 125,000 cc plastic trash bag to the outlet of the blower and it filled in about 55 seconds, running at 2500 rpm. Previous versions had taken from 2 and half to over 4 minutes.

If we just say that the blower output 125,000 cc's in one minute, and divide that by the proposed 56 cc displacement, that is 2232 cylinder fills in one minute. Since this is for a 2 stroke, the blower rpm of 2500 might support an engine rpm of ..... maybe 1500?? Hard to say.

Here is the amateurish video, but I think it gets the point across. Honestly, I was a bit surprised at how well it filled the plastic bag.

View attachment 135788

Now THAT is cool!

Congratulations!!!

John W

 

[ajoeiam]

Will the Blower Pump Enough Air?

Here are some thoughts and questions about this Roots blower.
First off, we know this application works in real life in the 100's of thousands of 2-stroke Detroit Diesels that were produced. I remember here in the US, that the Greyhound buses would always go flying by on the interstate, with that familiar blown 2-stroke sound.
I know this blower pumps air. The 125,000 cc plastic bag was stretched tight after just over a minute of pumping at 2,500 RPM.
View attachment 135828



Here is the inside of the blower and each lobe space calculates out to 11.3 cc's, and one complete revolution should pump six lobe spaces, or (at 100% efficiency) 67.8 cc. At 2500 rpm, that is 169,500 cc's. Does that mean the blower is 125,000/169.500= 74% efficient. That is extremely hard to believe, but maybe since the pressure only builds up to 15 or so inches of water, maybe that is true.
View attachment 135830


So here is the real question(s). How many revs of the blower would be needed to purge a 56 cc cylinder? And how about a method to test to see if it will actually purge 56cc while at operating speed? This is a head-scratcher.
Lloyd

Good go there sir!!!!!!!!!!!

This may be time for the equivalency testing situation.

I also have no idea of the calculations involved - - - but if you compare the volume created by a blower for a Detroit diesel of its mated size you would have at least some kind of idea of how close you are in your endeavors.

At the very least you are one huge heap closer to getting something that works - - - your iteration does something - - - quite impressively I would add.

How much volume are you purging - - - how much are you providing (from your blower)?
If your volume supplied (maybe per minute for each - - - dunno - - - maybe per second) is say 115% of required - - - - well - - - I'd say - - - try it!!

(Sometimes trial and error beats a lot of calculations and might even be faster - - - grin!)

 

[Lloyd-ss]

Now THAT is cool!
Congratulations!!!
John W

Good go there sir!!!!!!!!!!!
This may be time for the equivalency testing situation.
.............
(Sometimes trial and error beats a lot of calculations and might even be faster - - - grin!)

Thanks gents. The encouragement is appreciated and sometimes that is what keeps me going on this.

I crunched some numbers for the inlet (scavenge) port timing to see how long the inlet port would have to purge the cylinder. Even with a fixed port size, there is a way to vary the port timing.

This engine might be too large for a model, but I don't know. The plan was 56cc, 33mm bore x 65.5mm stroke.
I found many examples of inlet port and exhaust valve timing in a 2 stroke diesel One in particular was given for the DD. Exh valve opens 83 deg BBDC, inlet port opens 49 deg BBDC, inlet port closes 49 deg ABDC, Exh valve closes 62 deg ABDC. That means that the power stroke only lasts 97 deg of crank rotation, and the compression takes place in only 118 deg of crank rotation. Doing some calcs, at 1000 rpm, the exh valve is open 24.4 msec, and the inlet port is open 16.3 msec. At 2000 rpm the inlet port is open for only 8.2 msec.

So, at 2000 rpm, the air has to flow thru the cylinder at a velocity of 4 meters per second to theoretically totally purge the cylinder. Thinking in the imperial system, that converts to 787 feet per minute, or 9 miles per hour. I don't know if my numbers are correct, but a velocity of 9 mph for air seems achievable.

One of the variables I mentioned has to do with conn rod length. The longer the conn rod, the longer the piston will dwell at the top and bottom of the stroke (do I have that backwards??) [ darn, I am pretty sure I stated that backwards. the shorter rod has more piston dwell at the top and bottom of the stroke.], and this affects the actual length of the inlet ports. With a minimally short conrod of 68.6mm (65.5mm stroke), the inlet port if timed for 49 deg BBDC, will be 6.6 mm long. If the conn rod is made very long at 114mm, then the port length is increased by 2mm to 8.6mm long. That is 30% more area in the inlet ports. It seems like a longer conn rod and longer inlet port will help the engine breath better. Maybe.
Lloyd

 

[Steamchick]

Fag-packet calculation:
Engine displacement and rpm:
56cc x 2500rpm: One cylinder fill per Rev.
=> 140litres per minute:
Your pump makes 125litres per minute... but at 1.04barG. = 130l/min.
So maybe the engine will run? but won't run really well. So can you make a blower half as long again, (or whatever?) to exceed the displacement (at atmospheric pressure) and even give you some boost?

Probably of no use whatsoever...but, as an aside,
I think that modern motorcycle scooters a la Vespa and Lambretta, but now made by Piaggio, and lots of others (not the electric toy scooters like single wheeled skate-boards with a steering handle), have 125cc engines with small blowers to boost their inlets. I have also seen an electric blower to boost and engine...
Universal 3" Electric Turbocharger Air intake for Car/Motorcycler/Truck/ATV/RV – Grandado.com ...
Mini Turbocharger For Small Engines Motorcycle RHB31 For ALTO | eBay
It may be possible to get enough information from these to understand what actual volume and pressure they produce, and for what size/power of engine, so you can get some better clues as to sizing your blower?

Thanks for previous compliment, but I don't really deserve any, as I have more "stupid" ideas issuing from my brain than "sensible".
Enjoy what you are doing, as we enjoy your posts!
Thanks,
K2

 

[awake]

Impressive - both the results and the perseverance!

 

[Lloyd-ss]

Fag-packet calculation:
Engine displacement and rpm:
56cc x 2500rpm: One cylinder fill per Rev.
=> 140litres per minute:
Your pump makes 125litres per minute... but at 1.04barG. = 130l/min.
So maybe the engine will run? but won't run really well. So can you make a blower half as long again, (or whatever?) to exceed the displacement (at atmospheric pressure) and even give you some boost?

Probably of no use whatsoever...but, as an aside,
I think that modern motorcycle scooters a la Vespa and Lambretta, but now made by Piaggio, and lots of others (not the electric toy scooters like single wheeled skate-boards with a steering handle), have 125cc engines with small blowers to boost their inlets. I have also seen an electric blower to boost and engine...
Universal 3" Electric Turbocharger Air intake for Car/Motorcycler/Truck/ATV/RV – Grandado.com ...
Mini Turbocharger For Small Engines Motorcycle RHB31 For ALTO | eBay
It may be possible to get enough information from these to understand what actual volume and pressure they produce, and for what size/power of engine, so you can get some better clues as to sizing your blower?

Thanks for previous compliment, but I don't really deserve any, as I have more "stupid" ideas issuing from my brain than "sensible".
Enjoy what you are doing, as we enjoy your posts!
Thanks,
K2

Steam,
I agree that at a 1 to 1 drive ratio, that the blower will be a bit anemic for a 56cc 2 stroke. But with a 2 to 1 overdrive, it might work ok...... as long as it doesn't blow up. I have not figured out why Detroit Diesel decided on a Roots blower back in 1935. Was that the best option? Seems like a difficult item to manufacture for just blowing air.
Real calcs and specs for diesel blowers are hard to dig up..... but I probably just haven't dug in the right place yet. I need to cobble up some kind of flow test to get some better data. Those tiny blowers are kind of sketchy on specs. The AMR500 min roots blower claims 500cc of air per rev and they recommend it for a max engine size of 1.6 L. My blower might produce 67 cc/ rev, but I am pretty sure that the AMR500 can be driven at major overdrive ratios. So they are easily putting out 500cc x overdive ratio per rev.

Time to go outside and work in the garden and ponder the blower situation while performing mindless tasks. That's about as good as my multitasking gets, LOL.

P.S. Wrote this reply yesterday and then forgot to post it. So much for multitasking, LOL.

 

[mu38&Bg#]

AMR500 will produce 1.8 pressure ratio or ~12psi of boost. High performance engines will have peak primary pressure ratio of ~1.5 just before transfers open. There must be pressure to push the charge through the ports against cylinder pressure which takes time to drop. Slower engines give a bit more time for blowdown and transfer. Obviously, engines run at part throttle when crankcase pressures are much lower, but the effect on power is clear.

The key will be making the blower components as precise as possible.

Here's some motivation. superchargedsaitopics

https://www.rcuniverse.com/forum/glow-engines-114/1580599-roots-supercharged-saito-engine.html gets interesting at post 18.

 

[Steamchick]

Just a curiosity: Nissan made a March/Micra with a Super-turbo" set-up - for rallying/performance. A guy in work had one and raced it with he=is local rally club. The Supercharger worked up to 3000rpm (or something?) then a magnetic clutch disengaged and the turbo took-over boosting the engine. I am not sure if the waste-gate closed automatically, (open during supercharging) or by some electronic actuator. But the guy said he could never feel the changeover as engine performance, but he could hear a slight change of turbo note - probably as it came on/off load? Not that he was interested, as he was fully focused on racing...
I'm sure the supercharger took a bit of a pounding when the clutch engaged as the engine revs dropped through the transition point?
I think that basically, the turbo and supercharger were in parallel.
Nissan Micra K10 March Super Turbo Technical Specs, Dimensions (ultimatespecs.com)
Enjoy!

 

[Lloyd-ss]

AMR500 will produce 1.8 pressure ratio or ~12psi of boost. High performance engines will have peak primary pressure ratio of ~1.5 just before transfers open. There must be pressure to push the charge through the ports against cylinder pressure which takes time to drop. Slower engines give a bit more time for blowdown and transfer. Obviously, engines run at part throttle when crankcase pressures are much lower, but the effect on power is clear.

The key will be making the blower components as precise as possible.

Here's some motivation. superchargedsaitopics

https://www.rcuniverse.com/forum/glow-engines-114/1580599-roots-supercharged-saito-engine.html gets interesting at post 18.

Diesel,
That was a nice read from ptechllp about his build of the supercharger for the Saito engine. I especially appreciated his candor about the entire process.
Several take-aways from that discussion. One is that it was a major learning experience. I can attest to that. Proper (tight) clearances are hard to achieve. I agree. And your statement about the need for precision is definitely true. That is one of the major reasons for the 4 different versions that I have built. And with a CNC, I don't know if I can do much better on my manual machines.

But there are some major differences for the diesel that I am "trying" to build compared to the Saito engine.
One main factor is that the air from my blower into the diesel must be clean and dry, with no lube in it. I don't want the diesel to pre-ignite or run away. To help with that I have used sealed ball bearings and Teflon rotors. The Teflon allows for the "rubs" that the rotors have with certain parts of the housing, and with each other, to happen without the parts galling together. The sealed bearings should operate without lube. The Saito supercharger has lube in the air stream passing thru it to keep it happy.

Another very important difference is the RPM of the 2 systems. The Saito operates up to 9,000 rpm. My diesel might get up to 2,500(?) or maybe a little more. The reason that this is so important is because of the handicap it puts on the low RPM engine. The static pressure output of the blower increases with the square of the increase in RPM. Triple the RPM, and the pressure output goes up by a factor of 9. That is major.
Also, the flow rate increases proportionally with the RPM of the blower. Triple the RPM and the flow volume goes up by a factor of 3. And that RPM vs air flow volume is pretty much a "must have" for the application.

And yes, I totally agree that in order for the scavenging to work at all, the pressure from the blower has to be noticeably higher than the residual pressure in the cylinder. Having sufficient pressure and flow is my major concern right now. I want to verify that, somehow, before i get much deeper into the project. I am not at all sure what that residual pressure be, but I imagine it will vary with RPM, just like the blower output pressure.
Lloyd

 

[Lloyd-ss]

Here are a few pics and a short video from testing the Roots blower with a rotating shutter to simulate one of the standard Detroit Diesel inlet port timings of : Open 49 degrees BBDC, close 49 degrees ABDC. 98 degrees total open time. For reference, the exh valve(s) opens 83 degree BBDC and closes 62 degrees ABDC. Therefore, the exh valve opens 34 degrees before the inlet ports are exposed so that the blower isn't hit with massive backpressure from the combustion process.

The blower is being driven at 2580 RPM and the shutter is at approx 700 RPM. No particular reason for the shutter speed, but that is what that drill operates at. The shutter window looks narrower than 98 degrees because the diameter of the hole that it is exposing has to be added to the opening of the shutter.
Static pressures started at 3" of water at 1248 RPM, 6"at 1812 RPM, 10" at 2580 RPM, and 27" at 3600 RPM. The pressure is supposed to go up with the square of the RPM,but there was quite a bit of leakage around the shutter that affected the pressure readings. Still, the pressure seems pretty low for the task at hand. More research is needed.

Here is the rotating plastic shutter for the tests. The thin tubing is for pressure tap.


Here is the how the drill drives the shutter to simulate the opening and closing of the inlet (scavenging) ports.



And here it is in operation, simulating the opening and closing of the inlet port as the piston moves down and up.

 

[Lloyd-ss]

Here is a composite video about the development of the Roots Blower for the 2 stroke diesel project. I think it will be adequate for the task, but if needed, I could improve on it down the road.
The next challenge will be seeing how difficult the fuel injection will be.

 

[Lloyd-ss]

Experimenting with Injectors now
Trying out a Yanmar Injector

I bought a knock-off Yanmar replacement injector, 186FA, which is compatible with the 189cc to 406cc single cylinder 4 stroke air cooled engines. Even though the bulk of the injector is too large for my 56 cc engine, otherwise it might be ok. I might be able to machine some excess metal off the injector to bring the size more in proportion to the project. Or at the very least, I can learn a lot from it.

I had a pressure test fixture from a previous project and wanted to see how the injector functioned. This part is just an injector and needs an outside pump to control the pressure and the stroke volume. From the testing, it looks like the injector fires cleanly at 2800psi, as specified.

The firing happens near the end of the video at about 23:00. This looks promising.

It's just a static set-up and only fires at pressure. Just for testing, nothing more. The volume isn't controlled, yet. But I do also have the matching Yanmar injector pump, which is cam driven thru the big tappet cup at the top, which controls both the pressure and the injection volume. The little ball knob on the right side of the pump rotates around to vary the injection volume. The brains are pretty much all in the pump. The injector just delivers the dose dispensed by the pump.

Here is the Injector and its matching cam-driven pump. Again, both are too large for the project, but something to study.