strange combustion in 510rpm 4stroke marine diesel engine 2

[CapeNor]

Very strange,

We have experimental v12 engines, 9000hp each, but they smoke like hell. People think we're burning cokes or something.

If we look at the P-a diagram of the combustion (pressure vs crank angle)We see that the ignition delay is rather long. The initial ignition is very "agressive". It rises very fast, more an explosion than a combustion. And most strange of all, the rest of the cycle, the controlled combustion, has always a tripple peak curve. Somehow the pressure in the cylinder goes up and down in peaks, instead of just going up or down. If I look at it, I think there are 3 explosions, a big one followed by 2 small ones instead of one smooth combustion. Does anybody has ever seen anything like this before?? Does anybody knows a web adress where combustion problems are explained by P-V or P-a diagrams.

I will try to upload a graph later.

I must admit that the engines are running at 70% max, because of crankpinbearing problems. Maybe caused by the same strange combustion????

And also I will admit, because the turbochargers are surging a lot, we reduced the scavenge air temperature to a very low 30°C only. We use 180 cSt HFO, but even on MDO, she smokes like hell.

The TDC pressure is 60 bars, pressure peaks during combustion go up to 95 bars.

I hope anyone has any idee???

 

[21121956]

Hello everybody:

Hi Norcape. Running medium speed engines on residual fuel oils, under high operating temperatures conditions, piston crown and exhaust valves deposits build-up and corrosion are a significant problem. There are also problems with deposits and corrosion on turbochargers (nozzles and blades) and in the exhaust gas boiler (when a waste heat boiler is installed).

But, in addition to the stated above,you must have to take into account that deposits (like that on the piston crown of the pictures) can be a result of one or a combination of the following facts:

- Problems with the combustion: rich or lean mixture.
- Injectors tips damaged or leaking.
- Injection pumps working out of order.
- Earlier or late injection timing.
- Delayed (more than normal) ignition: gives as result unburnt fuel.
- Inadequate fuel viscosity.
- Turbochargers malfunctioning.
- High water concentration in the scavenge air (if the water separator and cooler drainage system are not
operating correctly).

Regarding with your piston, the exterior area of the flower shape shadow on the piston crown is an indication of unburnt fuel. The piston deposits can be, as well, residue from the thermal degradation of cylinder lubricant.

It would be advisable to check out the lube oil quality, this is TBN (Total Base Number), Calcium content, etc., and of course, the sulphur content of your fuel oil. Sometimes, depending on the values of these components and the conditions of operation, they become together in really "bad boys".

As a general information on this subject, you can visit this site:

http://energy.bmstu.ru/eo2/diesel/d11eng.htm#examples

I hope this can help you.

 

[21121956]

Hello everybody:

I beg your pardon, the correct address is as follows:

http://energy.bmstu.ru/e02/diesel/d11eng.htm#examples.

Once again, excuse me.

 

[CapeNor]

Hello 21121956,

This is truly a educating site. I've seen and learned a lot. I will need some time to process everything

Tomorrow I go back to the ship for another 14 days. I will search and find as much details about the current injection nozzles as possible. There must be some details somewhere. Injection angle, velocity, ... even if it was just a production number...

I will also search for details from the newbuilding. What tips did they use with what kind of injection pattern. There must be something somewhere.

And finally I will try to convince the boss to ask the nozzle facturer if these are the correct nozzles for the engine.

I hope it is succesfull

Greatings

Norcape

 

[JimCasey]

I like what jbthiel said. There is no turbulence in this chamber to mix the air with the fuel. Look at the squish band in the cat chamber in his second post.( Black, flat area on the right side) When the piston gets to TDC, the air trapped between the head and the squish band is expressed into the combustion bowl at high velocity, creating a highly turbulent 3-dimensional swirl shaped like a doughnut.

Also, the pattern on top of your piston indicates direct droplet impingement on the relatively cool piston top. This probably delays combustion significantly and results in low efficiency. Instead of tiny droplets bursting into flame, you have a puddle of fuel on top of the piston sizzling like a drop of water in a hot skillet. So a lot of your diesel fuel just gets scorched, and goes out the stack as soot instead of going into your prop shaft as mechanical work.

Flatter injector pattern, smaller droplet size, are things you can do easily. New pistons with Different piston crown and a Ricardo chamber would probably be the best thing, but then you really WOULD have a prototype engine. Fooling around with injector timing migh pay off: either inject it at the instant of greatest turbulence, or advanced enough so that it does not squirt directly onto the piston crown.

 

[MikeHalloran]

ISTR that the fuel used is often heated just to make it flow, so this idea might be problematic:

I'm wondering if refrigerating the fuel going into the injectors would raise the injection pressure and improve atomization?



Mike Halloran
Pembroke Pines, FL, USA

 

[WGJ]

"There is no turbulence in this chamber to mix the air with the fuel."

The toroidal motion you mention was certainly something that was aimed for in European truck diesel engines a few years back.

The re-entrant bowl with centre peak in the crown of the piston was there to get the squish motion to provide this fast-moving air 'toroid' into which the fuel was injected.

Norcape mentions the swirl plates fitted to the ports of these big marine engines - the 6 litre truck diesels I had a minor involvement with had a fairly complex twist cast into their inlet ports to get the inlet air to rotate in the cylinder during the inlet stroke and the majority of the compression.
In doing this, the 'toroid' was not only rotating in on itself but rotating about the vertical axis of the cylinder as well, with the object of maximising mixing.

JimCasey, when you mention the 'Ricardo' chamber, aren't you referring to the Ricardo Comet chamber used in indirect-injection engines?
This ship's engine piston, etc., looks just like it's smaller direct-injection truck brothers to me.

http://mmc-assess.tuwien.ac.at/data/sfrm/alsaffil/kolben1.jpg

For what it's worth, IMHO (hate that, but it seems right here):
White/grey smoke:
Engine too cold
Injection too late
Light levels of (dark) smoke:
Under rated load, performing nicely
Lots of smoke (per the pictures)
Injector holes too big
Injection pressure too low
Injection too late / continues too long
Injector not suited to the type of fuel in use
(e.g. fuel too viscous/cold?)


Bill

 

[jbthiel]

I fully agree about the intake port shape, but I didn't want to get into that because of Norcape's experience with excessive contamination and the unlikelyhood that this radical of a modification being done. I have seen the "dog leg" intake ports on everything from a 1.9L VW TDI to a med-large (~15L?) CAT engine.
On many of the new 4-valve diesels the valves are twisted compared to the traditional gasoline design. The short runner goes straight in, but the long one has a "dog leg" to induce the swirl.

As for the piston, it doesn't have the vertical lip like the latest pistons. (mexican hat vs. toriodal) I think that will limit the amount of mixing that is going to happen. If you had the swirl plates installed right before the pictures were taken it doesn't look like they were doing anything. (except increasing your workload )

I think the best thing to concentrate on is the injection system (pressure, angle, # of holes, etc.) You might also ponder if there could be improvements made to the swirl plates to induce more swirl without adding a ton of restriction.

WGJ - agree with everything, but I would add that adding fuel too quickly can also result in incomplete combustion. (though that doesn't appear to be an issue here) The inner part of the stream doesn't get a chance to properly atomize.
See my thread 71-26352 from July02.

 

[FJTOM]

Changing the injector nozzle angle would definetly help the impingment problem. I would also be sure that you are using the correct gasket for the injector. You may want to experiment with thicker gaskets to reduce impingment. Are there any inpingement marks on the head?
Measuring the peak injection pressure would be useful. If it is indeed low, then reduceing the nozzle area would raise the pressure and lengthen the injection period. On high speed diesels the injection period usually occurs before combustion is initated. Medium and slow speed diesels can control burn rates based on injection rates.

More holes of smaller diameter, keeping the same wall thickness in the injector, will also increase the length of the injection spray. Hard to tell from picture if there is any impingement on the cylinder walls?
Looking at the photos, infact perhaps more penetration is required. Making the L/R ration bigger would help in this situation. Leaving the nozzle wall thickness (L) the same, and reducing the hole size results in more penetration. The resulting spray pattern would be longer and narrower
(reducing impingement on the piston). More holes of smaller size, so that the total nozzle area is the same, would not change injection pressure.

I would assume that changing the injection cams is not feasible.

Try changing the injection timing from pump port closing of 14.5 degrees BTDC, to about 12 degrees. Later injection should result in less fuel being injected before combustion initiation. Less ignition delay, smother pressure rise rates. Shiming pumps to achieve late injection may be possible.

 

[cengrtecs]

Norcape,

Good day. There is nothing could be changed on your engine now unless your management or your owner are willing and it would be costly. Please check first and try to analyse shop test of this engine. If you can find this shop test result, you can see different parameters at different loads. And you will see power curves. All ships engines are tested before and during sea trial, this will not pass classification society where vessel belongs if during those tests, engine emitted too much smoke.

Brgds,
cengrtecs

 

[TJNBKA]

Hi all!

cengrtecs is absolutely right, the original testbed report is the best starting point in order to analyse properly the problem. Best thing to do is to compare, at similar rev's and loads, the actual engine parameters to those at the factory's testbed.

Some other feelings:
injection timing:
14° seems rather late for me! for example, we operate 600rpm stationary units on IF380 HFO, those units start injection at 25° (initial timing to be checked in testbed report)

turbocharger surging:
this occures mainly in two situations: actual air volume delivered by the TC is too high for the engine > generates surging!
The TC can deliver too much air to the engine when the exhaust nozzle ring is seriously clogged, what often results in exessive TC speed and charge air pressure...

injector opening pressure:
400 Bars is not so bad, our above mentionned engines have the injectors adjusted at 350 Bars, after 2000hrs, we find some injectors that dropped to 260 - 280 Bars and service them

smoke:
understood your engines run relatively low loads upto 60-70%. In that load area, smoke is quite normal (but not as much as shown on your pictures) Our 600rpm engines (8500 HP) are allmost smoke free from 80-85% upto full and overload.
Boost pressure is around 2,2 Bars at full load.

Could you put the initial tesbed reports in this thread together with your actual measurements & readings?

 

[CapeNor]

Hello everybody,

Thanx for your repies:

I'm back from the ship and i've got the following items:

We did tests on the injector test bench and found the average injection angle to be 124 degrees. This matches the center part of the mark on the piston. Together we searched all the books we have but cannot find an official injection angle.
There exist a modified injector. They made the bore of the hole longer. Is this to reduce the velocity of the fuel?? The term low velocity nozzle has been used. This modified type exist since 1989, and it was especially for smoke reduction.
Right now we used the old type nozzles, but the chief engineer said this was only because the previous time there where no modified nozzles ready to use??

We did a unit last week, it was already 3 months overdue! I took some pictures, and this is what we found:

Obviously one of the injector holes was plugged by the carbon.

If we have a look at the side of the piston, we can see plenty carbon build up. It has been "grinding" the liner with a damaged liner (worn surface) as a result. The top piston ring was broken on one spot. Probably it was stuck by the carbon. (increased blow by?)

Once the piston was out, I took some detailed pictures of the marks. The wite ruler is placed directly on top of the mark. It shows the 3D form of the piston on the place of the mark.

Overhere a cut out detail. The ruler above shows it dimensions in inches.

So what do you guys think of these pictures??


We mentioned the swirl plates at the inlet valves before. There seems to be a little misunderstanding, so I took a picture as well. The pipe in the center is the support of the valve seat grinding tool. But it shows the position of the inlet valve stem. The white dots everywhere is dust from this grinding process.
Originally there where 3 vanes, but the one in the back is grinded away already, because it was obstructing the air flow to much.

And now the "good news". The ship has been spotted in the port of Rotterdam for smoking. Now there is a bit of official pressure on the company to do something.
A specialist has been invited to do some measuring of the engine. It will take a couple off months before everything will be organised.

Some of you mentioned the original test results. I can make a copy, but...

Original test were made on diesel oil, not on heavy fuel.
Since the original test, there has been some modifications (nozzle rings on the turbo, increased cooling water temperature, modified fuel pumps (the original ones could not handle the HFO as there were supposed to de) ...

But, what we see today is: The max combustion pressure is less high then original. The exhaust temperatures are higher then originally. The boost air pressure is also higher then originally.

I will try to take a copy with me next time.

Greatings and happy xmas to all of you

 

[globi5]

Here's a link to a company (DUAP) which is specialized in Diesel injection systems on large Diesel engines. They might also be able to help you:

http://www.duap.ch/

(The company is better than what their website might indicate.)

(I haven't read the entire thread, but regardless of the size of the engine, fuel should always be kept away from the 'cold' cylinder walls. The piston crown looks rather flat to prevent this from happening).

 

[stenis]

Nice investigation work! IMHO the green spray should be much better than the red one. Fuel that is laying on the piston will not burn. You want air and fuel mixing in the combustion chamber. One problem you may run into is spraying fuel onto the cylinder walls but looking at your recent pictures I cannot imagine that cylinder wear will get worse than it is now.

If the injector opening pressure really drops from 400 to 260 bars the injection event must be totally different. But I assume that the engines are still smoking with 400 bars opening pressure and new nozzles so this is not the root cause.

The swirl plates look very interesting but I do not think they are more than plasters and they cannot fix the original problem. If they really need cleaning every month there must be quite a lot of backflow through the intake valves. Either the valve timing is really wrong or then the turbines or exhaust pipes are clogged. Do you measure the exhaust back pressure?

It is almost like you are breaking some law of physics in your engine now. Higher charge air pressure should give higher cylinder pressures and lower exhaust temperatures. If the root cause is a delayed combustion then there will be more energy to the turbine and the boost will go up. The exhaust temperature may go up too but cylinder pressure will go down. Clogging of the turbine may also give a similar effect. The fuel pumps could be worn or wrongly made or then the TDC mark on the flywheel can be wrong.

I think there may several design issues that are wrong with these engines and then adding some wear and clogging then you end up where you are now.

 

[MikeHalloran]

The swirl plates seem to have been optimized for ease of manufacture. It appears that the halves of each plate will fight each other, e.g. one half is bent to induce cw swirl and the other half is bent to induce ccw swirl.



Mike Halloran
Pembroke Pines, FL, USA

 

[patprimmer]

Mike is right. The swirl plates won't work. I would try it without them, or fabricate ones that will work.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[rmw]

In even a well designed compression ignition engine you can make it smoke by over fueling it. All that black stuff going up the stack is unburnt fuel which is not contributing anything to crankshaft horsepower.

I am not a boat person, but I have owned trucks and I had drivers that thought that power was everything and they would have injectors from higher HP engines put into my engines by their buddies and they would smoke like all get out. Whether they actually got more power by doing that or just the placebo effect because they thought they had souped the thing up is debatable. (They did manage to burn up some expensive engines.)

When I would put the right injectors back in, it would run as designed without excess smoking in all but the most abusive driver error situations (mashing the gas when there weren't enough rpm's to suppport the fuel being added.)

I would suggest having the injectors modified to put in slightly less fuel so that the fuel that is put in can combust and produce the power you want. Has anyone done any combustion calculations regarding fuel air ratios?

rmw

 

[CapeNor]

Hello, everybody

There seems to be a little misunderstanding about the swirl plates.

Originally, they were not installed. It was a original Mitsui modification, because they thought the fuel was not mixing with the air. Or the scavenging cycle left some parts of the combustion space with exhaust gasses.
So they made this swirl plates to let the air "swirl in the cylinder". Not around the valve.

So I put this extra picture of the cylinder head, You can see both air inlets with the swirl plates. Everybody can imagine the purpose. Question is will it work if the valves are in front of it???

The swirl plates are getting very dirty. I know when the engine is idling, we have some back flow through the inlet ports. And this is a ferry. We are in port every day and every day, we have to pass a lock. During the time in the lock, the engines are idling for half an hour, sometimes even longer.

RMW: For the question about overfueling it. I don't think were overfueling cause the engine runs only at 70%. This is because we had serious problems with the crankpin bearings.

But even when arriving in port, with a hot engine and reduced power, she still smokes like hell!

One extra question rose about the fuel spray on the piston. The engine is stopped after idling. And maintenance is always done after we passed the lock. Is it possible that what we see on the piston burns away when the engine is working full power?? What we see is only the result of the manouvring through the lock??

Greetings

Norcape

 

[globi5]

Diesel needs to be injected onto the piston cavity in order to get it to disperse properly. Before they were capable to manufacture injection nozzles with several orifices, they used to spray one single jet straight down directly onto the piston cavity, where it was dispersed and redirected to the center of the combustion chamber. Injection angle always depends on the shape of the cavity on the piston crown. (Fuel speed is too high such that it would stop and sit on the piston crown.)

Also, if the diesel is injected such that it hits the cylinder walls (too flat) there's a high probably that fuel will not be thrown back to the center of the combustion chamber, rather it will be dispersed onto the cylinder walls (cooler and larger area) without getting the hot air in the center needed to burn properly.
Diesel engines always run lean. Ideally all the fuel is in the center of the combustion chamber and there is a fuel-free insulating air cushion sitting on the cylinder walls in order to increase combustion temperature and ensure complete combustion.

In the picture from jbtiel the cavity in the piston crown prevents fuel from reaching the cylinder walls eventhough the injection angle is relatively flat. Also, dispersing and evaporating fuel on piston crown will assist cooling it.

 

[patprimmer]

Now I see.

The right hand port will swirl the charge into mostly free space, but the left hand port will swirl the charge into the head of the right hand valve.

The right hand valve is swirling against a cylinder wall which may carry fuel to the wall. I wonder if aiming it in slightly might help. Probably not as to get effective swirl, the air must run around and so will be contained by the cylinder wall.

Does swirl really help or does it just act as a centrifuge to throw more fuel against the cylinder wall.

My thoughts for what there worth bearing in mind I am not a diesel guy and especially not a big diesel guy.

A fuel shot, through high pressure multiple fine aperture injectors at a steeper angle more to the centre of the piston.

Inject fuel at higher pressure for shorter duration.

Ensure injection timing is optimal.

Ensure valves are sealing.

Ensure cam timing is correct and suitable.

A deeper smaller diameter dish in the piston.

Maybe higher coolant temperature.

Maybe preheat the fuel.

Maybe higher engine oil temperature (to let pistons get hotter.

Remove one or even both swirl plates.

Increase compression ratio or turbo boost.

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[MikeHalloran]

Ah. Now I see what the engine guys were trying to do.

The swirl plates as shown would induce some swirl in the chamber, with the valves full open. With the valves less than completely open, they just get in the way, and in the presence of swirl, might induce the backflow you detected.

I wonder if removing one set of swirl plates from each cylinder would help? Have you tried that?





Mike Halloran
Pembroke Pines, FL, USA