Disappointing turbo performance

[Banzaibob]

New plumbing.

[alexanderfoti]

Now thats a short run! Would be interested to hear what the differences are with the new plumbing.

[Banzaibob]

Hope to have all of the fittings and pipes finished by this evening. I do not think I will be able to do much of a test run however due to an approaching winter storm.

[BertTrack]

I applaud your fast work on that plumbing! Sweet!

[Banzaibob]

New images, plenum with reed valve installed. No performance results yet.

[Banzaibob]

Got the bike finished yesterday. Changed rear sprocket from a 38t to a 43t gearing it down about 10%. It appears to accelerate through the gears smoothly and with more authority. It appears to generate more boost at lower rpm's with modest throttle settings and modest speeds (3rd-4th gear) generating about 4 psi of boost. Full throttle high rpm gives 8-9 psi of boost. It pulls strongly through 4 gear and when I hit 5th it accelerates to 53mph and simply won't do anymore. I have backed the fuel stop all the way out and as it's struggling to maintain 50 or so mph it's leaving clouds of fuel smoke. It is almost as if it is running up against the governor.

I constructed the plenum chamber out of 22ga. sheet metal, brazed at the seams. Just before I was going to make my 3rd test run of the day I noticed that the reed cage was oscillating back and forth with the vibrations of the engine. Upon closer inspection this was not only because of the vibrations but also the turbo boost was deforming the metal ruining the seal I had between the reed valve and the chamber. I have to construct the plenum out of more robust material.

The bottom line is that everything appears to run well, it pulls a little stronger at lower rpm's than before, the turbo is generating boost, it appears to run much smoother at high rpm's. So why in the hell isn't this thing generating more power? Why is it STILL not as fast overall than it was before the turbo? If there was a glaring flat spot in its performance I could go to that but this thing is not giving me any clues.

[tappy]

You say you're using a IHI RBH31 turbocharger. A little searching of the internet threw up a compressor flow characteristic here http://www.scooterinvasion.net/forum/vi ... 429ceb057c.

A Yanmar L100 is 406cc, revving to 3600rpm. At ambient conditions (101.325KPa, 288.15K), that's swallowing 0.406 litres x 1800 intakes per minute x 0.001225Kg/litre = 0.015Kg/s airflow.
If you're getting about 9psi gauge of boost at about 60% isentropic efficiency then your pressure ratio is about 1.6 and the temperature leaving the compressor will be about 357K. So now the air density will be about 0.00158Kg/litre, so the engine can swallow about 0.019Kg/s.

If you look at the flow characteristic, and find where 1.6 (vertical axis) and 0.02 (horizontal axis) meet, you'll see the compressor is operating to the far left of it's flow map. The left most line on the map will probably be the stall line, so I'd expect that except when the engine's actively sucking air in at max revs, the compressor's probably operating very near to, or actually in surge, in which case it's not very happy. It's certainly not very efficient - it's sitting on the 60% efficiency contour.

I haven't been able to find a turbine chareristic for it, but it's safe to assume that it also won't be designed to run in this area either. Diesel engine's higher compression & expansion ratios mean that the exhaust gas hitting the turbo will be at lower pressure than that of a petrol engine. This means that if the turbo isn't turning very fast (because of the low efficiencies and lower exhaust pressure) then it might be choking - effectively causing a blockage in the exhaust. Hence your engine is smoking and not making enough power. This will be aggravated by the flow pulses as the engine will be trying to get all the flow out in a very short window.

For a bit of back-ground have a look here: http://www.turbomagazine.com/tech/0304_ ... essor_map/.

So basically, the statements that your turbo is too big for your engine, are probably about right. Without having a different turbine put into the turbo, other things you might do to get better results are:
1) Reduce pressure losses
Make well blended corners and changes in section - convergent sections aren't too bad, but divergent sections should be about 12-14degrees included angle. Think about how a 2-stroke exhaust looks. Your "dump" diffuser into your plenum probably ain't great, and I dare say there are steps in the flow section where the exhaust comes out of the exhaust port and then into the turbine NGV.

2) Cool the intake charge. This will reduce the compressor power, increase the charge swallowed by the engine and help the engine generally. You could try making a basic dry-ice or water-ice intercooler to see if it helps.

3) Reduce exhaust heat losses before the turbo. That heat tape is probably helping, consider more of it, and if you change the exhaust tubing then make sure it's still lagged.

4) Use an exhaust plenum

5) Improve the basic engine's breathing - porting, valve timing & lift etc.

[larkout]

But the numbers do plot within the design range. I think we should say that this turbo is not ideal ( this whole project is marginal in so many ways and I think that is a lot of its appeal) but that its really ok and certainly the best we got right now.
My short ramble (longer ramble in the 'end of the road' thread) is that its getting boost, which means its getting more air in the cylinder than n/a. Its getting fuel. We presume that since theres soot coming out the back that its running rich, too much fuel stoichiometrically speaking, but what if its really just not burning the fuel quickly enough? If thats the case how do we get it to burn faster/ more thoroughly? Like a stove with a fresh log, but this little stove is getting 1800 fresh logs every minute...

[tappy]

I see your point, and on the face of it I'm inclined to agree - it IS in the range. BUT, it's running very innefficiently so just because it's building pressure, doesn't mean it's flowing more air.
If the pressure loss as the exhaust flows through the turbine is dominating the pressure rise through the compressor then all you're really doing is increasing the temperature of the charge going in - you're not increasing the mass flow. Hence the smoke.
If the engine's running at higher pressure then the combustion should be happening faster anyway.
As I said in my previous post - there are things you could do - make sure you're not creating pressure losses, and get the incoming charge nice and cold for a start...

[larkout]

My simple logic is that 7 pounds boost = 50% more air into the engine = 50% more power. (I know the math is off due to temperature increase of the intake charge) Its easy to see that working the turbo would take up some of that power but I have a hard time imagining it taking up ~4-5 hp.
If there is enough backpressure to make it so that measured boost in the intake is not indicative of a proportional amount of usable air going into the cylinder, this would be because scavenging has become so poor that the cylinder is still full of old exhaust when the intake opens again and we're just trying to press fresh air on top of that?
Are there effective ways to cool the intake while avoiding the extra plumbing (and backpressure?) of an intercooler? Is there any practical benefit to a tuned exhaust length with a turbocharger?

[toyotaracer9]

I dont think the engine has enough heat energy to spin the turbine wheel fast enough causing the exhaust to be restrictive, like Ive said since the beggining of the post if you are not spinning the turbine wheel fast enough, each blade is acting like a wall holding back exhaust flow. Look how the turbine wheel is made.

[tappy]

[quote="larkout"]My simple logic is that 7 pounds boost = 50% more air into the engine = 50% more power.

OK, I'm going to be blunt.

No!

50% more pressure only equals 50% more flow if the turbo compressor isn't having to push against 50% more flow restriction. If the turbine is adding an extra 50% flow restriction on the exhaust side (not impossible with such pulsing flow) then all that's happening is that you're squeezing the air harder from the inlet and exhaust side. It might all be at higher pressure, but there's no extra flow. And with no extra flow, there's no extra power.

There is no direct correlation between boost and flow if your turbo is badly set up. If the turbo set up was much better with much less flow restriction but only 25% more pressure on the inlet side, then you'd get more flow and more power.

Does that make sense? People need to stop correlating pressure with flow and therefore power.

[tappy]

The more I think about it, the more I'm convinced that the pulsing nature of the exhaust flow out of the engine is a major part of the problem.

It's a reasonable assumption to assume that both the compressor and turbine have been matched to deal with the same mass flow going into and out of the engine. Yes the pressure, temperature and density will be different, but fundamentally the mass of gas is practically the same (tiny bit of fuel mass in there too).

A plenum has been used to smooth out the flow on the inlet side. At the average flow rate I calculated that the turbo is only just flowing as much as it would like to. If there was no plenum then for 540degree of engine rotation there would be zero flow from the turbo, and for 180 degrees (intake stroke) there would be a sudden flow. The plenum will help smooth this out, but it's possible that for say - 360degrees of engine rotation there is no flow, and then for 360 degrees there's about the flow the turbo was designed for. So on the inlet side at the moment, things might be OK.

Now turn to the exhaust. If the flow out the exhaust port were constant, then again we'd expect that the turbine was seeing just about enough flow to spin the compressor. But in practise we're seeing no flow for 540 degrees, and then 4 times the "average" flow for 180 degrees. Given that the exhaust valve doesn't open & close instantaneously, and the piston speed is sinusoidal, we might actually be seeing a peak of 6 times the average flow. This is probably too much for the turbine, so the turbine chokes and doesn't let the exhaust out.

So back pressure builds up in the engine, stopping fresh air in.
So it smokes like a bastard, and makes no more power despite reading 7psi manifold pressure.

SOOO, my previous advice stands - make an exhaust plenum too. And for both inlet and exhaust plenum reduce their pressure losses (I can probably advise), and consider a spot of intercooling.

[larkout]

Tappy- This is good stuff. Theres a lot of new systems here for me to wrap my head around, and yeah, the lights not always shining brightly in all the corners of the room. You've helped a few things click into place.

[gearhead1951]

Tappy , this is exactly why I recommended an engine driven supercharger a good few posts back !!

[toyotaracer9]

belt driven s/c have a parasitic loss due to a belt drive a 10hp engine losses 2hp to gain 3 = 1hp gain?

[Banzaibob]

The first incarnation of this turbo installation had a large volume pipe between the exhaust manifold and the turbo inlet of about 500cc or so. We made it larger on purpose for the very reason that you explain. Are you saying that perhaps 500cc's was not enough? By the way, I have already removed the turbo but it is a hop, skip and a jump to reinstall. I, as well as toyotaracer9 still have serious doubts about whether this turbo overall will work on this engine under any circumstances.

[tappy]

Yeah - I'd be thinking you'd need about 2000cc for a 600cc engine. That's a bit of a stab in the dark, but essentially - the bigger the better. You can always make it smaller again. It would cause exhaust pressure losses, but that should be offset by the engine breathing better and the turbo getting a more consistent flow into it.
I can understand your feeling that it might never work, but if you do feel like trying again some time then that's where I'd start.

[DieselFly]

Just going to throw a crazy idea out there. Has anyone measured the crankcase pressure? I know the v twins have an outrageous amount of blow by when turboed. There is no way the little stock piss hole vent can handle all of the back pressure. This could be part of the problem that the singles are having. When you do open up the vents be prepared to make an air/oil separator.

Sean

[klondikekid]

I don't claim to know alot about turbo set up etc. but my 3cyl. Kubota sure didn't like it when I had a gt15 turbo on it, didn't seem to want to spool up, it was too big, then went to a smaller one gt12, made a big difference, just sitting still and revving it up you could hear the exhaust whistle. Maybe like having to big a carb on a gas engine..bigger isn't always better. I'm not familiar/know what size turbo your using, has other people used that combination with good results? if so,disregard my rambling. Good luck.

[UAofE]

When you do open up the vents be prepared to make an air/oil separator.

1500 miles and my clear CCV hose is still totally clean inside. I can feel gas venting from the end so I know its working. Or maybe Kubota just has a really good CCV system. I know it uses a reed valve instead of a check ball.


Seems to me that a "wet" nitrous system using propane as the fuel is the solution for those that want more power for acceleration and passing speed.

Benefits:
You won't have to overfuel the engine, so tank fuel economy and visible emissions are not affected.
No added heat or mechanical stress on the engine during non-maximum load conditions.
Much cheaper than a turbo/supercharger.
Can be easily tailored to exactly how much extra power you need.
Weight neutral compared to a turbo/supercharger system.
Tanks can be located anywhere for better weight distribution.

Risks:
Ungoverned fuel source can overspeed the engine if used without maximum load.
Consumable "air" source costs about US$5 per pound to refill.
Nitrous/propane could run out just when you need it most.
Nitrous is illegal to use on public streets in some jurisdictions.
Nitrous tank can be very dangerous if pressure gets too high or the tank gets damaged.

To prevent thread hijack, I will take that topic here: https://www.dieselbikeforum.com/view ... =32&t=3137

[coachgeo]

Last two post are informative and thanks for posting Butttt... mostly do not relate to this thread at all. This whole thread is about issues with doing turbo on a "one cylinder" engine. This info fits better elsewhere in here if you've not done so already

[klondikekid]

The principles should be the same you would think?

[UAofE]

This whole thread is about issues with doing turbo on a "one cylinder" engine.

Which is why I proposed my alternative.

I have come to believe turboing a single cylinder diesel and gaining power is not possible because there is just not enough time for the exhaust to operate the turbo cycle.
Its been done successfully on small gas engines, but they have far greater operating RPMs to reduce the time between cycles.

Supercharging works, but the stress added from making an additional 3hp to put only 1hp to the wheels would detract from the long life of the engine and significantly increase fuel consumption even during low load situations.

Both depend on the injection system being set to overfuel to gain power from the additional air, which severely increases emissions during throttle transitions causing negative public opinion of diesels.

[coachgeo]

The principles should be the same you would think?

one would think but after reading this thread and others of similar discussion it appears singles have other factors not present with multi-cylinder engines. Points are good with supercharger... a catch 22 as is LPG since it works only so so without some form of compressor.

[BoxerOtto]

wait a minute. who said turbos on a single cannot ever work? VW and their large budget made it work for the 1 litre car. putting together random bits from the internet that were never engineered to work together is what doesn't work well. make sure your system actually works,performs, and has the longevity, you say it does before you condemn other systems. it's obvious you havn't ridden a well set up supercharged single, there's not many out there, on mine i'm gathering information and passing it along once it has proven itself worthwhile. it takes a lot of time to make it right and opinions are only opinions until proven as fact.

[UAofE]

VW and their large budget made it work for the 1 litre car.

opinions are only opinions until proven as fact.

The first version was an SDI (non-turbo). The two later versions used an 800cc inline-2 TDI.

As has been proven fact several times with R2V840 twin, turboing an 800cc twin is fairly easy with existing turbo models.
Turboing a 400cc single properly would require a turbo smaller than even a Garrett GT0632. Especially if any altitude above sea level is involved.

[UAofE]

Another thing to consider is that overfueling hurts power.
On my engine it actually gained top end power by REDUCING the fuel quantity limit to a slight gray haze. Whereas before, it would noticeably bog down if I tried to increase throttle beyond the RPM equilibrium point and make it smoke.

[mark_in_manchester]

I may be re-opening a can of worms here, but anyway...

This thread seemed to be getting around to saying that boost pressure in the intake plenum of 7-9psi was all very well, but a restricted exhaust would mean that maybe that potential to deliver extra volume flow would not get used since cylinder would not scavenge. Well, OK, but doesn't the fact of 7-9psi in the intake plenum imply that a fairly serious exhaust flow is taking place in order to spin the exhaust impeller that fast, so as to generate that intake pressure?

While my bike is apart, I am almost minded to add tapped holes for pressure measurements across the exhaust impeller - across the intake impeller is already available. I used to work in acoustics and this is really interesting. But I have no access to instrumentation any more (or rather, only scrap stuff I saved from the skip) - so it might amount to going back to the 80s and recording pressure signals on a stereo cassette (!) and playing back through an oscilloscope in the 'lab'. Saying that, I would need a d.c. coupled measurement chain...bum. There are clever people on here - do d.c. coupled sound-cards exist for engineering purposes?

[pietenpol2002]

For the history of this board we've cited Mr. Macinne's work below as the basis for recommending plenum/capacitor sizing to be 2-3 times displacement.

http://www.duccutters.com/pdf/PlenumSizing.pdf

However, the results of the research below suggests that for a single, a plenum 4-5 times displacement is far more effective. And that plenums/capacitor sizes that fall below or above that range are far less effective. I can imagine that sizing above the recommended displacement would result in an incremental decline in efficiency. Whereas, sizing too small could result in the turbo ultimately being a restriction. You'll note that Banzaibob's plenum in the pics above is almost exactly 3 times displacement.

http://dspace.mit.edu/openaccess-dissem ... 21.1/98220 (opens as a PDF)

The Yanmar L100 single in the link below uses a variable-geometry turbocharger. They swapped out the original fuel injector for a Fiat Grande Punto MJTD 1.3 L bar injector that's controlled by a Bosch MS15.1 Diesel ECU through a Bosch fuel rail (model #261-B1-135-201). The pics in the article are from the pre-turbo era. I only wish they had published the results from the modifications made. I've emailed the authors in the hope they can supply us with more results and will post if forthcoming.

https://www.researchgate.net/profile/Ch ... ion_detail (once it opens, click on "download full-text" to the right)