DIY tunnel drive

Sorry, CDK, a quick correction on the pump theme. In a naturally aspirated diesel the induced massflow of air is varying practically linearly with rpm,s as long as the atmospheric pressure is constant. But with the turbocharged engine, the air massflow is varying with rpms AND load (=fuel flow).

Imagine rolling at constant speed on the highway, level and flat. Engine power is constant, so is fuel flow and turbo-rpms (=inlet pressure). Enter an uphill. To keep speed constant you increase fuel flow.

The air flow in the NA engine is the same as before, due to rpm dependance, but the fuel/air ratio is increased.

In the turbo-engine, the increased fuel flow generates increased exhaust flow, the turbine is speeding up and the air flow is increasing until a balance is reached. Fuel/air ratio is kept more or less constant all the way!

If there were no sensor for inlet pressure, a defect turbounit would send massive smoke as soon as a load was applied. This really makes me wonder what engines you are stuck with!

Also, the internal feed pump is there to maintain a controlled housing pressure. It is not designed to feed the engine in a boat.

Please describe your fuel system a bit more in detail. Are there any common parts (filters, rubber hoses, whatever) or are the systems completely separated all the way? Have you checked the tank outlet for dirt/bacterial gel or other debris?

 

Anyway, the pressure compensator (also called "smoke limiter") is not there, and that´s it. It is not as crucial in a marine engine as in a road vehicle, since the propeller load has about the same power characteristics as the charger unit, so we can get the same power when everything else is ok.

One problem with the system without a separate feed pump is air leaks. Since it is ten times as difficult to make pipe connections tight against air leaking into a sub-atmospheric system pressure, than stopping fuel to leak out at overpressure, you are asking for trouble in the tiny internal priming pump.

So, next exercise for you: Disconnect the fuel delivery tube at the pump. Install 20 cm transparent, fuel resistent tubing in between, with a clean, white paper behind. Run (both) engines under full load and watch carefully for any trace of gas bubbles in the incoming fuel.

And I just have to ask the obvious: is your throttle control really pulling the control lever to full flow position, and if there is a mechanical stop; does it really go all the way back in the "run" position? If there is just a solenoid; does it really get its full 12 Volts to open fully?

 

At or below 50 hp @ liter, the amount of fuel injected is substantially lower than the available airmass can handle, so there is no need for an aneroid. The VE pump is a marvelous piece of mechanical engineering that - with just a handful of springs, rollers and a plunger - does the right thing at the right moment. The same engine in a VW Golf-3 or Passat is labeled AAZ. It generates more power, has a stronger spring in the waste gate actuator for more pressure at higher rpm; there they installed an aneroid to prevent a fuel overdose.
And yes, a faulty turbo charger manifests itself clearly in the rear mirror.

Internal pressure is of little concern in the VE pump. The submerged plunger that meters the amount to be injected takes always the same volume, both at idle and at full power. A sleeve around it has a waste port and can slide back and forth whenever the governor requires that.
In the VW transporter, the pump draws fuel from the tank near the rear wheels and under all driving conditions keeps the pump body filled.

In my setup the only common part is the 300 ltr fuel tank, located directly behind the bulkhead. There are two fuel lines, filters and water separators screwed against the bulkhead and very short rubber hoses connect them to the engines. The fuel returns are also separate.

The tank is clean, it was used for gasoline in the past.
The fuel lines are transparent, so show any air that passes. None does!
The actuators that control the pump setting are (of course) not simple solenoids but electronically controlled motors with spindles, limit switches and potmeter feedback. They release all tension at idle and can reach more than 95% of "full throttle". I left a few mm of travel unused to prevent the motors from drawing current continuously.

 

Well, now is the time to make a summing up of round one. So far we seem to have produced a decent amusement factor for our friends (which is not bad per se) but no solution to your problem. The tacho calibration issue may be temporarily settled. The high idle of 4000 rpm (~8 % over max power rpm’s) is what convinced me that your readings are in the right region, if not with the best of accuracies.

So, we have a propeller working at 1050 rpm. With available info on the operating point and propeller data, the absorbed power might be up to abt 19 hp (make it 20 with transmission losses) in the worst of cases. At this speed, your engines should produce at least 36 hp according to available information, which leads to conclusion no one: both engines are at least 45 % low on power.

If it had been only on one side, it could have been the occational sh-t, but it is on both, which makes it more likely that it is caused by ”the last hand on the bolt”. Then, what makes a diesel run; enough fuel to produce the reqired power, enough air to burn that fuel and big enough holes to get rid of combustion gases. Everything else is a question of life span. First round survey of the installation revealed ”too small holes for the residues”, but no complaints on air inlet conditions (”mice are absent…”). If air inlet flow had been low, you would most certainly have reported black smoke, but you haven’t! Now, with the reduced power, the outlet should not be too critical and the chargers are not yet really alive, which leaves us with our noses pointing to the injection pumps and fuel system (”not enough fuel to produce………”).

On your pic’s the details of the pump and its connections are not quite visible, but I suggest we concentrate our efforts there. I take it that any immobilization paraphernalia (if originally present) has been removed without destruction of anything essential. As previously noted, the housing pressure is used for injection timing control (please don’t argue on that!). The return connection (center bolt in banjo conn.) has means for controlling this pressure (varying from calibrated nozzle to spring loaded non-return valve). Are all the original connection parts in place?

Also, I have a vague idea of having seen some VW diesels with Lucas/CAV rotary injection pump, could this be the case here?? If so, I think Lucas sometimes use additional electrical inputs for cold starting and timing control. There may also be a fuel flow limiter, reducing injection quantity until engine has reached normal operating temp. Questions: are there any unused (as for now) electrical connections on the pumps, looking like the fuel sol. connections? Can you verify the pump supplier and part number? Is there anything else that might indicate something sensing engine coolant temp? Any extra, unused temp sensors in the engine water jacket that might be there to ”report” to the injection pump?

A couple of years ago we had a similar problem with a brand new VM diesel pulling a waterjet. With fairly exact data on the jet, we knew the engine was some 35 % low on power. It turned out that these engines, although ok for the purpose, are in fact ”overcooled” for our climate (big, mediterranian duty oil coolers). The pump was electronically controlled rotary Bosch, and the temp sensor reported some two degrees C too low engine temp, which caused control circuits to limit fuel flow. Any associations of thoughts?

 

Don't you think that is a bit condescending? If you don't want me to argue, I'll leave it for what it is, but please try to remember I studied mechanical engineering and they even gave me a very official document to confirm that.

There was an immobilizer circuit on the pump which communicated with a coil around the key lock in a very complicated fashion. All it did was keep the fuel valve from opening if the proper key wasn't present. I dissected the black box and re-installed just the solenoid that operates a needle valve.

If there was a Lucas pump installed, don't you think I would have told you?
Both pumps have their original banjo bolts, the one with the restrictor on the return side. The fuel lines are short and of adequate dimensions and so are the Volvo Penta filters/ water separators.

At the beginning of this discussion, I mentioned that the turbo charger behaviour may have changed because I use water jacketed exhausts.
Although I wasn't able to retrieve any factory data, I remember that forums like the dieselschrauber-community mention full turbine pressure at 1600 or 1800 rpm, which means that the waste gate should start to open at that point.
My theory is, that with 90 C. water circulating a light alloy manifold, the exhaust temperature will be considerably lower than with a cast iron one under a heat shield. At higher gas velocities the difference will decrease, but around 2000 rpm there certainly is less gas volume to turn the turbine and there will be less intake air. Not to the point of black smoke forming, but still noticeable as a reduced performance. The manometer will provide the answer.

 

Thinking about my own last post I must admit it is not as strong a case as I thought when writing it....

 

Ouch, I tried to avoid rubbing your nose with the housing pressure issue, but you really want it don't you? Thought it was waste of my time, because the info is available netwise, but I can scan the information from the "Bosch VE information page" if you really insist......

And, in fact I cannot find that you have mentioned the pump type before at all; it was my own automatic first assumption that there were only Bosch pumps in VW engines. Anyway, the VE pumps also have a variety of temperature compensating bits and pieces. For instance there are pumps with a small watercirquit for engine coolant, that acts on injection settings, depending on engine temperature. So my questions about surplus temp sensors et c are still relevant.

Of course you have a point on the reduced xh temperature, but the turbine power is a product of mass flow times enthalpy difference, and the massflow is still there. The temperature of the bulk of gas is not changed dramatically during its short residence time in the manifold. Your turbine housing (the "spiral" housing) is NOT watercooled on your engines and you might add some heat insulation around it to keep gas temperature up, but the effect is marginal. Instead, I still believe that the xh back pressure is too high; it remains an open question until measured. But again: it should not be critical at the low rpm end.

Having full turbopressure at 1800 rpms is b------t! Possible in a dualturbo installation but not here.

 

The pump type was a matter of prior knowledge. I wrote all about it in another thread:http://www.boatdesign.net/forums/diy-marinizing/marinizing-vw-turbodiesel-engines-19118.html , but of course you couldn't know that.

Do not spend time searching for the housing pressure issue. I've read lots of texts about these pumps, the most detailed ones are in German. I also looked inside such a pump and decided to leave mine untouched as long as possible because at re-assembly too much can go wrong. And in the gray past I briefly worked at Robert Bosch KG in Amsterdam investigating diesel fuel system malfunctions. But back then they only made inline pumps.

There is a slight overpressure in the pump housing to prevent vacuum forming around the fast moving and rotating plunger. Early models suffered from decreased output volume near max rpm, so they resolved that with an insert in the banjo bolt or a washer if there were axial connectors. If I remember correctly there should be a minimum of 100 mm H2O, but that may not apply to all pumps.
You are correct, there are many different types with or without appendages and even the optically identical ones have different springs.

But the pumps on ABL diesels are very basic. There was this immobilizer with two wires, one for 12V the other for serial data from the key lock sensor. No further electrical connections, no vacuum, just the "throttle lever" and a bowden cable for the starting device that advances the injection and increases the minimal fuel amount so idle is slightly over 1000 rpm.
I ignored that last one because I thought it would be required for low temps only, but that was a mistake: starting without the cable pulled results in poor idling with lots of smoke from unburnt fuel leaving the exhaust. That stops as soon as the engines are loaded, but when idling you can make a blue smoke curtain with the sharp smell of unburnt fuel, so there probably is still injection after the exhaust valves open. I solved that with a linear motor and a switch on the instrument panel.

In the above thread I also gave some details about how I inject cooling water in the turbine housing and waited with bated breath for the cracks to appear, but they didn't.

"Having full turbopressure at 1800 rpms is b------t! Possible in a dualturbo installation but not here."

Take a turbine designed for 0,9 bar at 3000 rpm and replace the waste gate spring with one that opens at 0,5 bar. Remember that the engine output difference between SAP and turbocharged is only 5 kW (1996/97) or 10 kW (1998/99).
The manometer hasn't arrived yet from Germany.

 

For your info on the VE pumps, I dug down into documentation sediments since I worked (long ago) as a r&d engineer with the green diesels, arranging fuel mapping and timing optimizing among other tasks.

In the attached document you find the function of the injection timing device and how it works, using housing pressure as a rpm signal. You also find the recommended supply pressure range.

I also compiled a diagram, showing typical air inlet pressures (expressed as a percentage of the respective max pressure) for two diesels, operating in the same rpm range as the VW. The Volvo TMD31 is a 4 cyl. without wastegate and intercooler, with total volume of 2.4 l. Power is 100 hp (ie 42 hp/l.). The VM is a 6 cyl 4.2 l. engine with intercooler and wastegate (60 hp/l). It has a 4 cyl brother with exactly the same mean brake effective pressure. There is no such thing as full turbo-pressure at less than half rated rpm's in a standard application! Since I am not allowed to disclose the VM figures in absolute terms, I made the comparison nondimensional regarding pressure levels.

Your VW will have a turbocharacteristics similar to, but slightly lower than the VM. It would probably have a relative pressure of 0.4 to 0.45 of pmax (ie ~0.35 to 0.4 b.) at 2100 rpm. The "waste-gate engines" show a very marked pressure kick when their "undersized" turbo's start working. When the w-g opens, the pressure increase flattens to a linear rise up to max. pressure. Within a very limited rpm range, the power increase is very steep, and these engines are often notorious for bad behaviour under propeller load conditions. The Volvo TAMD 63L is a typical example, and your VW belongs to this cathegory as well.

It also has the consequense, that even a relatively small deviation from ideal conditions will result in a dramatic change of performance. So after checking your installation, there are two areas that deviate from established marine standards; the exhaust diameters are small and creating a high back pressure to the turbine, and the supply pressure to the injection pumps are insufficient.

It is of no use to discuss propellers and tunnels until the engine installation is corrected, but from your photos there seems to be some details that have to be changed as well, but let's take this when your engines are up and running!

PS If you use paper cartridges for your fuel filters: Take care that they are stowed dry, with undamaged moisture-tigh wrapping until installed. They tend to absorb H2O moisture when lying around unprotected. Some types swell from this, blocking fuel flow after installation!

 

Baeckmo, the 1999 VE pumps Bosch brochure has slightly different pictures of the automatic timing device and detailed description of the operation.
The vane pump pressure is used to determine rpm and advance timing up to 24 crankshaft degrees, starting at 300 rpm. In their paper there is a spring-loaded valve at the exit side of the chamber, so there is an evenly distribution of pressure throughout the pump and they use the pressure from the chamber the high pressure plunger is in.

My knowledge is a bit dated I'm afraid. There have been - probably still are - pumps where the vane pump connects to the chamber with a small bore and the timing advance pressure came from the vane pump directly. That timing plunger bore also has a channel to the main chamber, partly closed by the plunger until the spring was compressed. It resembles the drawing in your document: the vane pump pressure determined the advance, the chamber pressure didn't matter.

So I stand corrected, even more so since I saw the new drawing.

The first thing I will do is unscrew the banjo bolts to see what is inside/underneath. Then I will read through some papers from the "Dieselschrauber Community" to see if it is safe to remove the advance mechanism with the pump in situ. I expect to find sediment in at least one engine blocking the plunger's travel.

 

The channel from vane pump to the spring side of the plunger (lower left in pic) is the suction side. The advancing pressure enters from pump housing to the right side of the advancing plunger through the small nozzle seen just to the right of the pin in the plunger center.

I think the banjo with the ball (non-return) valve primarily was used in applications where the return fuel was led to the filter body, to prevent pump draining during standstill. I don't think it is worth your time to bother with that.

The spring side of the advance mechanism is available under the low cover at the side of the pump. It can be opened with the pump in situ, but beware the "jumping spring". You are correct that in older engines, maybe having been unused for some time, "slime" is collected here, as it is the lowest part of the wet inners. But, if my memory is correct, you should not try to withdraw the plunger; it is bl--dy hell to get it back in position!

 

CDK, came to think of your return lines. You omitted the original fuel filters with their thermostatically controlled shunt, having quite narrow bores. Could it be that your "new" return has too low flow resistance, compared to the original combination: banjo ball valve>thermo shunt>filter, so that the filling of the injection pump cylinder is not complete? That would certainly slow you down!

After a marinizing project of an Iveco ohc engine long ago (similar power and VE-pump as your VW's), I saved the following data for the pump housing pressure:

Engine rpm 1000, pressure 2.4 to 3.0 b.
1500, 3.5 to 4.1
2000, 4.6 to 5.2

I expect your engines to run with the same levels of housing pressure. Check the old filter heads for possible nozzles or equal! Unfortunately, there is no pressure connection in to the housing. I was allowed to borrow a longer center bolt and a manometer adaptor from a local diesel pump shop, but I guess that is not an option for you.

 

I give up CDK do you want to check out these engines to be sure you have no stupid problem or not ....yesterday baeckmo did not know his waste gate from his power valve now he is a diesel god ..before you do any of this dismantling you are taking about you need to check fuel and air ..its a diesel for gods sake !!!
simple basic things first .....

I have seen no post of boost pressure and of course it will have maxed at 1800 rpm
Nor have I seen any details of checks on the power valve settings ,,,

you need a pressue gauge 0-15 psi or 1 bar usually a fuel pressure gauge is the cheapest or a vac gauge with a reverse scale ....whats all this manometer crap .... bye till you call

 

CDK, Just spoke to a friend who is servicing a fleet of MAN vehicles in Norway. He confirms that there were two systems for the fuel return in light commercial vehicles. On the one you have there is an additional constriction in the original fuel filter head, together with this thermovalve. This version has never reached the marine environment (except through DIY marinizers).

Easiest way to correct your system might be either to reinstall the old filter head and connect both return paths to tank, or to check the orifice size (would be something in the 0.5 mm dia range) and make two plugs to fit into your tank connection.

As said before, low housing pressure will upset injection pump performance totally, so imho this is the first thing to adress.

 

Baeckmo, I think the riddle is solved.
I have a Brooklands Workshop manual for the VW transporter models 1996-1999. It says that the fuel line connectors are marked and must not be mixed up because the return one has a restrictor. It also has a description and several pictures of the control valve on top of the fuel filter, that should open one way at 12 C. and the other way at 31 C.
That is all crap. Maybe they employed a bible-translator who understood no technical German, some text got lost or they are just ignorant fools, but that doesn't alter the fact that it is crap.

When the sun went down yesterday evening and you were probably composing your post I lifted the floor boards on my boat and unscrewed the banjo bolts on one engine and guess what, they are identical. In
the rapidly falling light I think I saw some markings like E and A, but certainly nothing that remotely looked like an orifice to reduce fuel flow.

The Bosch VE brochure says that the mechanism advances the injection by 24 degrees, starting already above 300 rpm. As you may remember when asked for black smoke I responded there is only blue smoke and the smell of unburnt fuel when the engines are started without the "Starthilfe" lever being pulled.
That figures! If the commence of injection at idle rpm occurs that late in the power stroke, there is still some fuel injected after the exhaust valves start opening, hence the smell and smoke. And if there still is 24 degrees missing at higher rpm, more that a quarter of the power stroke passes before the injection starts.

Unfortunately I do not have the VW filters and valves because the engines came from a recycler where they used bolt cutters to free the engine from its environment. But I have a lathe and a variety of materials, so I can make something with an adjustment screw, a narrow passage or even a spring loaded valve if I must. Some guidance would be helpful because I have no example at hand.

I take the liberty to add that leaving engine timing to a remotely mounted plastic device is very poor engineering!