VAF Forums - AFP Fuel Injection (LONG)

VAF Forums (https://vansairforce.net/community/index.php)

- Fuel Injection Systems (https://vansairforce.net/community/forumdisplay.php?f=94)

- - AFP Fuel Injection (LONG) (https://vansairforce.net/community/showthread.php?t=14998)

AFP Fuel Injection (LONG)

On the Matronics rv-list, there was a thread the other day that spawned when somebody reported that a friend's RV-4 had a forced landing due to power loss. The cause ended up being in the builder's control...he had not safety wired a screw on the AFP purge valve which the installation manual clearly states should be lock wired. This is the stop screw on the purge valve that limits the travel of the valve. When this screw backed out, the purge valve was allowed to rotate further than it should have, and the engine was starved of fuel. This problem could have been prevented by simply safety wiring that screw as per the instructions.

However...the thread that ensued on rv-list started when somebody decided to "swear off" purge valves, citing that there was no good reason to have one, even on an AFP system.

I replied on the list that even if you don't use the purge valve for one of its intended functions -- purging vaporized fuel prior to a hot start -- the other purpose was to act as a true cutoff of fuel flowing to the engine. The Airflow Performance injection system is actually designed to still flow a bit of fuel at ICO (idle cut-off)...unlike other systems such as the RSA. On the rv-list I advocated reconsidering removing the purge valve from an AFP system for that reason. Some replies came, indicating that the "ICO leak" could be fixed.

Others, such as myself, knew that there had to be a good reason why the AFP system still flows a tiny bit of fuel at ICO, and so I asked Don Rivera of AFP about it. What follows is his response. I don't know too many manufacturers who take the time to address concerns & questions in such detail. I have his permission to post it where required, and I felt this forum would benefit from this in-depth explanation of the design's history. I know there are many people out there flying behind the AFP system (or considering it), and I think this type of info helps to understand it better.

Enjoy!

(post follows)

Here is Don's response...

Quote:

Originally Posted by Don at Airflow

Dan,

<snip>

You had made comment that you wanted to know more about the mixture control
valve design. To completely understand the reasoning behind this design you
have to know a little history of how this all got started. When we started
our company in 1984 I had already 10 years experience with aircraft fuel
injection systems at the Bendix Corporation. Being the under study of the
inventor of the RS and RSA fuel injection system and later being the project
engineer on that product line gave me insight into the manufacturing
problems and cost associated the RSA design. In Airflow's infancy, we knew
that we would have to design a system to satisfy a large range of horsepower
requirements with a minimum of part and tooling changes. Knowing that we
wanted to be able to run engines from around 80 HP to 1000 HP we designed
the present fuel regulator concept.

Studying the needs in the aviation field we constantly heard of the big draw
back to fuel injection was 1) initial cost, 2) hot starts, 3) high cost of
overhaul. In this design we determined that eliminating part count without
sacrificing performance would help with manufacturing costs, and overhaul
cost.

Studying various manufacturing techniques, we knew that plate valves were
expensive to make (high part count) were susceptible to scoring unless you
used some expensive materials and there's always the issue of making the
parts flat (specialized equipment). Rotary valves on the other hand were
easy to control in manufacture (OD grinding) and round bores were easy to
control with honing. This would allow parts that would not have to be hand
lapped or fitted. The round parts could be made with tight enough
tolerances that matched parts were not necessary. Having a through bore
that both idle and mixture valves ran in gave the bonus of getting cost out
of manufacturing as through bore honing would hold the bore straight and we
could easily hold + .0005" on the entire bore. Brass was chosen as the
material to run in an anodized honed bore. Designing the L/D of the valve
gave excellent bearing surface and I have to admit, we really haven't had
any problems with wear or scoring of these parts in 20+ years of service.
The only down side is continued actuation of the parts when dry can cause
galling of the valve. This is solved by oil flushing the parts after test,
and in service the parts are always in fuel. Of course with a rotary valve
there has to be clearance for the valve to rotate, therefore ICO cannot be 0
leak. We also only shut off the metered side of the circuit in the
regulator. This removed the additional parts required to mechanize an
additional valve to shut off this side of the circuit and since the decision
was made to use the purge valve as standard equipment, a zero leak mixture
valve was not required.

Hot starts were a common problem with low-pressure non-returning fuel
injection systems, and even some early mechanical automotive fuel injection
like the Bosch K Jetronic suffered from this problem. We determined that
the hot start problem was due to heat soak on the fuel system components on
the engine. Since fuel boiled at around 130 degrees F at sea level
pressure, after the engine shut down the fuel on the engine side of the fire
wall in the hoses, engine driven fuel pump, fuel control, flow divider, and
nozzle lines would be partially boiled away. Since the fuel metering system
was non-returning, there was no way to get rid of the hot fuel and vapor.
You had to start the engine flooded or when the engine started you had to
run it up excessively to pass the vapor through the metering system to keep
the engine running. Some people didn't have problems with this technique,
many did. The components that held the most volume of fuel were the
culprits. The #6 fuel hoses, the engine driven fuel pump and the fuel
control. Since our metering system metered fuel to the engine based on
engine airflow consumption there was a limit on how fast fuel would transfer
through the system when the engine was not running. On a typical 4 cylinder
Lycoming the normal calibration set up allowed about 1 cup of fuel to
transfer through the system in 45 seconds of purging with the throttle wide
open. This would pretty much exchange the fuel in the engine driven fuel
pump and the fuel control and hoses. At idle the fuel transfer would be
.038 cup of fuel in one minute. This is why idling the engine will never
get the air out of the system, well at least not for 26 minutes. This is
another reason we want to minimize the volume of fuel on the engine side of
the firewall.

The purge valve was designed on the premise that cleaning out the hot fuel
and vapor from the engine driven pump, fuel control and hoses would cure the
hot start problem. The first system was installed on an IGSO 480 in an
aerobatic airplane, which was pretty much unstartable when hot. The system
worked quite well with pretty much the same start routine hot or cold. Also
the benefit with the purge valve was that it would dump the fuel pressure
when the engine was shut off to keep fuel from bleeding into the engine
after shut down. This was a problem with engines using diaphragm fuel
pumps. We always had complaints of fuel dripping into the air box after
shut down on Bendix servos which basically dead head the fuel pump pressure
against a plate valve. When the plate valve scored a little leakage started
and the engine would not shut down clean. People whine and moan about this
now, but 30 years ago when I was working at Bendix we heard the same thing.
Thus, another reason for the design of the purge valve.

The purge valve design was not something we designed from scratch with a
fresh sheet of paper. The basic valve design was studied as to what design
in the field gave the most trouble free service. Looking at helicopter
service, we found that that seemed to get the most abuse. From both a
vibration and wear stand point this installation typically had fuel tanks
above the engine so the valve had to be near zero leak as possible, yet be
robust enough to withstand the harsh environment it was in. So the valve
bushing was used from a RSA-7 fuel regulator. This same design had been
used on all Hughes 300 and Beechcraft Baron 58P installations. With a few
million flight hours accumulated, there had been not one incident of
malfunction of the valve, let alone the screw backing out because it was not
lockwired. The idle valve bushing on these fuel servos had the same design,
that is, being held in by one screw. Thus the Airflow purge valve was
designed to mimic the Bendix design, with some minor changes in the venting
of the ports in the bushing, and of course a housing was designed to hold
the valve.

So there you have it. A history and reasoning behind the mixture control
and purge valve design. This design was done to satisfy requirements that
we determined customers wanted in the field. After all, if the status quo
was accepted, why build anything? It would not address any of the issues
that existed, and you would end up with a clone of the same 40-year-old
design. Kind of like a Silver Hawk. All of these parts were designed for a
reason and with lot of forethought. Are there other ways to do it? You
bet. Is there a better cost effective way to address the problems
associated with low-pressure non-returning fuel injection systems? Probably
not, with the market as it is today.

Excellent post Dan and good information for us to know.
For those with the purge valve, make sure those set screws on the stops are safety wired. If they back out, its bad news.
Best,

Photo of safety wired stop screw.
[IMG]

[/IMG]

Thanks for doing the legwork on this, Dan. It's so nice to get info straight from the horse's mouth on such things.

Thanks for posting Dan and thanks also to Don at AFP for taking the time to provide an authoritative answer on fuel injection systems. Were it not for install issues I had with a Ryton sump/nosegear install with linkages and exhaust system interferences, I would have stayed with an AFP system. As it turned out, after my Ryton sump broke and I installed an Aero cast aluminum sump, effectively turning my installation into an M1B, I could have stayed with the AFP system. I am sold on the purge installation for all the reasons Don mentions and elected to retain it in my installation, along with the AFPin my Bendix system. Don was great in helping me find a rebuilt Bendix system since RSA wouldn't exchange or allow their servo to be installed in conjunction with a purge system and AFP distribution block and AFP injectors. Throughout my tribulations, Don was one of the best suppliers I have dealt with.

Regarding my experiences using the purge setup, I am sold on it completely. For my own piece of mind, I installed a return spring in case anything came unhooked under the plenum. The purge system shuts the engine down much better than simply using mixture and I am happy to know I'm not backing pressure up into the servo and engine pump.

Back when the sun was shining and it was 100F out I watched my friend with his V tail gas guzzling Bonanza cranking, and cranking and cranking. He then yelled outside the window to the rest of us patiently waiting in line for $4.00 100LL..."Hot start"...It was then I had my opportunity...Crank, crank brmmmm...Purr....:)

Incidently Vince if your reading this, order the rest of that rv8 kit would ya?..

Frank
N484H 7a, 30 hours of IFR training and I finally figured out holding petterns..>:)

I elected not to install the purge valve. (incidently, my neighbor had an engine out due to not safetying the stop nut a number of years back on his purge valve. Safely made a local field. Saftied the nut, no further issues.)
It is my belief, and hope, that the hot start procedures I have witnessed will actually work for me as well as they worked for those I have seen demonstrate it.
Full rich, full throttle, boost pump on 5 seconds, boost pump off, full lean, start, engine catch, throttle down, mixture in. ****, "do I have that backwards?"
Wish me luck.....

I remain a fan of the purge valve. Seems the troubles here have been from leaving the stop screw unsaftied; is that AFP's fault?

I was sanguine about the valve, honestly, until I started flight test. In my Sportsman, the fuel selector was exposed, as were the firewall-aft lines. One day at Chino, typically hot day, I was purging after buying fuel. Let it purge for about 60 seconds, and then, out of curiousity, reached down to feel the temp of the fuel through the lines. I could NOT hold my hand on the alumminum lines, the fuel was so hot. I let it purge a bit more, gave it a bit of prime, and it started right up.

This is in stark contrast to a Bellanca Viking I flew for awhile (IO-540-K) that was an absolute PITA to hot start. I never got just the right technique down with that airplane.

The only downside I've seen to the valve itself is in a high-wing airplane like mine. If you somehow manage to nudge the purge control forward even a little and don't notice, you'll come out next time to a puddle of fuel under the engine. I now make sure to shut off the fuel selector if I'm doing work in the cockpit, no big deal.

As for the gas-guzzing Bonanza mentioned elsewhere, that's a pilot problem. Big-bore Continentals are EASY to hot start if you do it right because the mixture controller allows the same kind of purging--just not as much of the system--as the AFP purge valve. Don't blame the airplane or the engine for the pilot's inability to start the Bonanza.

--Marc

Quote:

Originally Posted by JonJay

Well for a normal bendix system (Piper Arrow for example), on a hot start, there is NO prime (Engine is in a flooded condition already), I go throttle 1/2, mixture ICO, Crank until she fires, ease the mixture up to 1/2, start sliding the throttle back to 1200 RPM, and bring the mixture the rest of the way up at the same time.

Works for me every time.