Sender: <marv@lancaironline.net>
To: lml@lancaironline.net
Date: Sat, 16 Jun 2007 11:30:15 -0400
Message-ID: <redirect-2106418@logan.com>
Received-SPF: pass
 receiver=logan.com; client-ip=68.168.78.211; envelope-from=glcasey@adelphia.net
Mime-Version: 1.0 (Apple Message framework v752.2)
In-Reply-To: <list-2104005@logan.com>
References: <list-2104005@logan.com>
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Content-Transfer-Encoding: 7bit
From: Gary Casey <glcasey@adelphia.net>
Subject: Re: Boost switch

I've been reading this thread with interest and finally thought I  
could add something - if only just my opinion:
>
> No. My point is that the TCM fuel system design is faulty. It has
> an extraordinary sensitivity to its input fuel pressure. That
> problem can be made worse by bad setup, but even with "good" setup
> it is still a bad design.
>
I don't think I would go that far, since safety is relative, as all  
designs have faults and the only thing that can be done is to choose  
the best overall balance.  The TCM fuel system is what we used to  
call a "speed-theta" system in that fuel flow is primarily  
proportional to engine speed and throttle angle.  Add in various  
orifices to tweak the system and that's it.  Very simple, few moving  
parts and presumably reliable and durable.  An integral part of the  
system is the engine-driven fuel pump, which is a positive- 
displacement (vane) type sized to the engine flow requirements.  The  
throttle-controlled metering valve is shaped so that the fuel flow  
increases roughly in proportion to the throttle area.  There is  
nothing about the system that is related to air flow, and I think  
that's one reason why there is all the attention given to "set-up".   
Since flow is a function of all the orifices in the system the  
injectors themselves are part of the flow control.  Automotive users  
of this method include the classic Hilborn used in old Indy and other  
race cars (that's why they had "carburetor day" at Indy, because the  
fuel systems had to be set up just before the race) and Magnetti  
Marrelli (FIAT and Harley Davidson).  I think that today only H-D and  
Continental use this methodology.

Another approach is the "speed-density" system that makes the fuel  
flow proportional to rpm times manifold pressure (density).  This  
takes away the finicky throttle-controlled valve and makes the fuel  
flow more likely to be proportional to air flow, especially at part  
throttle conditions.  However, there is still no connection between  
fuel flow and air flow.  This system is used by a number of  
automotive manufacturers, notably Chrysler and Honda.  The addition  
of the "aneroid" to the Continental system adds partial manifold  
pressure compensation to the speed-theta system.

The Precision Airmotive (PAC, a.k.a. "Bendix", a.k.a. "Lycoming",  
although Lycoming has no connection with PAC) system is completely  
different and could be termed a "Mass Air Flow" (MAF) or just "Air  
Flow" system.  Air flow is measured by creating a pressure  
differential in a venturi and that pressure differential is used to  
meter flow through a fixed metering valve.  This is pretty much like  
in a carburetor except that the fuel remains under pressure, allowing  
it to be distributed to individual injectors.  Fuel flow is only  
dependent on air flow and is totally independent of the engine.  One  
nice byproduct of this is that the fuel system can be bench tested  
and just bolted on the engine - no "set-up" required other than an  
idle mixture adjustment.  In fact, there is no mixture adjustment  
possible(with the mixture control full rich, of course) in the field  
- the metering orifice has to be changed.  Of course, this doesn't  
mean it's perfect, but it doesn't suffer from some of the problems  
being discussed in this thread.  Since the FLOW of the fuel is  
controlled in the metering system, it is completely (well, almost)  
insensitive to upstream pressure.  It is so insensitive that the  
typical pump used is a diaphragm-type just as was used in cars for  
many years.  The diaphragm pump can be oversized so much that one  
part number is used for most 4 and 6-cylinder engines from 150 to  
400hp.  Since the system measures air flow, there are only two models  
(the -5 an the -10) that manage all these engines.  Most cars  
incorporate MAF systems, but non use venturis to measure air flow.   
The air flow measurement is sensitive to air flow pulsations, which  
is why the calibration of the systems is somewhat engine-specific.

There are some differences in the failure modes:  With the PAC system  
the total flow is controlled by the metering orifice, not the  
injectors, so if one injector plugs the flow in the others goes up,  
making those cylinders rich.    At full throttle the Continental  
system can suffer almost any type of air leak - like an intake tube  
that falls off -without serious problem as even that cylinder will  
likely keep running.  Even a hole in a piston or a cylinder head  
blowing off won't effect fuel flow unless the injector departs with  
the cylinder.  The same failures with an air flow system will  
dramatically reduce the air flow through the servo, creating a lean  
or even no flow condition.  The big difference has been thoroughly  
discussed here - the sensitivity to inlet pressure.  With an air flow  
system you can turn on the boost pump any time, including at idle,  
and there will be no effect.  The pump is a single speed and there is  
no worries about whether it is on or off.  Without the aneroid (I  
think all IO-550's have them??) the fuel flow in a Continental system  
is completely independent of altitude, making manual mixture control  
very important.  The PAC system compensates by the square root of air  
density, so it "half-compensates" for altitude.
>
> In previous messages, a few remedies have been offered.
> The two most practical are:
> 1) regulate the pressure of the fuel just upstream of the injector  
> servo,
> 2) use a Lycosaurus IO-540 fuel injection system (or the whole  
> engine).

I chose number 2, although I thought a lot about putting the PAC  
system on a Continental engine.  To imply that one system is "safe"  
and the other is "unsafe" is unrealistic as safety is relative.  In  
my opinion one is "more safe" than the other mainly in that it relies  
less on the least reliable component of the aircraft (that would be  
me) to keep everything working.

Another related note - there has been discussion about power  
interruptions when switching tanks with the continental system.   
Since its fuel pump relies on a steady supply of fuel if that supply  
is interrupted for even a very short time (like a few milliseconds)  
it is possible that the pump will create vapor at the inlet and that  
vapor might cause the pump to "vapor lock", or quit pump for a few  
seconds.  The fuel selectors shut off the fuel from the old tank  
before they turn on the fuel from the new tank.  Is it possible that  
some people don't have trouble because they have a habit of moving  
the selector very rapidly, reducing the "off" time?  Here again, the  
PAC system is fairly insensitive to this since there is an  
accumulator built into the pump, allowing it to continue delivering  
fuel even though the inlet is momentarily shut off.  That's why I  
made my system so that the new tank can be turned on before the old  
tank is shut off.

Gary Casey
ES 157, N224SG