I've been lurking to see expert answers to this.

The situation here is that you need two flowmeters to get reasonable results.  One goes in the fuel line, and I'd recommend putting it between the filter/separator and the injection pump so that anything that might clog it up or corrode it gets taken out.  The other one goes in the return line from the injectors to the fuel tank.

Both the instantaneous and averaged rates are the difference between these flows, and my opinion (subject to correction!) is that you can integrate the signals off these at the same rate to get the result in either case.  There is a short lag between the flow-rate measured in the fuel line and the flow that is seen in the return, but I think this lag is so short (and the dead volume involved so small) that it doesn't materially affect accuracy.

I had a 1985 diesel Lincoln (!) which used the 2.5 liter BMW turbo engine (I forget their internal engine code for it)with the Lincoln digital dashboard.  This vehicle contained all the flowmeters, logic, connectors, etc. for the application (including nifty vacuum-fluorescent displays).  There were quite a few Mark VIIs built with that engine during the mid-1980s which I think also have all the pieces.  While I wouldn't expect the flowmeters from something that small to work on, say, a 3208, it's quite possible that everything else would, and it's the difference signal that determines the range, not the intake or return flow magnitudes.  So finding flowmeters with the appropriate style of encoder on them would be the only research, aside from a scrapyard trip for the Lincoln, that you might need even for a fairly big truck...

Thanks a lot Wizlish.

I've been reading about this for a while now, and I think I'm getting closer to it.

Thanks for the help.

you could use a differential pressure transmitter across a restrictor - which can be an orifice.  you wouldn't want to take a lot of pressure drop, but enough to give you decent results.  omega makes a FLR1000 turbine flowmeter for less than $200 which operates on 12 VDC.  The output is a 0-5 VDC signal.  you need to review the wetted parts list or talk with omega to make sure that this will be compatible with diesel.  i wouldn't mount this unit under the hood because of the heat.  if you need anymore help, let me know.

hduncan - How would that cope with the flow in the return line?

If (and I realise it is a big if) you have electronic fuel control then why not just pick up the injector pulse signal and average it?

Cheers

Greg Locock

I feel rather stupid asking this: but why would you want to measure fuel flow anyway? It seems to me that you a stuck with the problem of measuring the small difference in flow between the main fuel line and the return line. A typical fuel injected vehicle (petrol or diesel) will cycle the contents of the tank many times in an hour. We know that our flow meters will have some degree of error, usually in the range of 2-5%. The problem gets exaggerated when we have multiple flow meters at work. Seems to me you'd be better off picking up one of those cheap MPG gauges and be done with it.

Good Luck
Bryan Carter

Seems to me that the 'cheap MPG gauges' are vacuum gauges.  You may recall that we're talking about diesels.  Compression-ignition engines do not have manifold vacuum correlated with fuel consumption... in fact, shouldn't have manifold vacuum at all ;-}

Also remember that the compression for ignition is provided by fuel combustion, which means that fuel burn out of any proportion to intake-tract airflow or 'gauge' manifold pressure is required to turn the engine at any advanced speed.

You're reduced to measuring fuel rate (in gph, in the USA) and then deriving mpg via speed data averaged over time to give distance.  There is, quite simply, no valid way to get fuel rate on a diesel without gauging the actual fuel delivery through the injectors.   If all the fuel delivered to the injection pump actually  went through the injectors and into the engine, a single flowmeter going to the injection pump (or common rail) would suffice.  But that doesn't happen, for reasons I won't go into here, but which are easily accessible in the literature.  It's better to return any fuel not injected at high pressure directly to the tank, at low pressure, than to try to pipe it back to the injection pump ahead of the fuel-line flowmeter.  We put a second flowmeter in the return line to measure the return flow and subtract it from delivery flow.  The 'missing' amount is what has been injected...

Seems clunky -- but I assure you that extensive experience has shown it to be the best approach!

RME

Thank you all for your kind replies.

As I said, I've been reading for a while know, and I'm getting closer to it.

Can anyone please suggest suitable sensors for this app ? I googled a lot (flowmeters, flowsensors, fuel consumtion, etc) and found almost nothing.

Thanks again