ROTAX 912 Fuel Pump

[Blueadventures]

14 minutes ago, skippydiesel said:

Please explain why you feel there may be  an issue with fuel delivery (pump to carb)

Just about how the hose run will go, I expect not over top of engine.  Some data that may be worth gaining is the fuel temp from the following locations 1 at through firewall penetration, 2 at ‘T’ to each carb and 3 at banjo on Carb.  Data could be a simple fridge temp sensor unit. In my aircraft there is brass fittings at each place so could fix the temp sensor there.  This way you will see how much temp you are wanting to reduce.  It could be almost zero or could be something. If very little temp increase then just fit and route the Rotax way.  Look forward to any data you gain, if you do such.

[IBob]

34 minutes ago, skippydiesel said:

Mitigation usually takes the form of;

 

My list was short two important mitigation tools ;

• A metered (small orifice) fuel return line, may encourage vaporised fuel to exit the up stream side of the fuel delivery.
• The use of a fuel "boost" pump may assist the expulsion of gases fuel from the system and assist in the suppression of further vaporisation.

 

Skippydiesel, I have experienced vaporisation of fuel in the engine bay lines twice that I know of, both during an engine restart some minutes after shutdown. So the engine bay would have been heat soaked.

 

Since part of managing this depends on the configuration of the aircraft, here is my (engine bay) layout:

The fuel line runs forward under the aircraft belly, and is piped into the bottom of the aux fuel pump, which is located low on the firewall and fitted with a shroud which is open top and bottom.
From the top of the aux pump, the fuel is piped up and forward over the top of the engine, to the Rotax fuel pump.
From the Rotax pump, the fuel is piped back over the top of the engine to a splitter at the firewall.
The splitter has four ports: one for each of the carbs/ a single port with small orifice for 'fuel return': this is piped back to a point high on the RH wing tank/ a single port with a small orifice, piped to the fuel pressure gauge.

 

Here is part of my standard startup routine, and how the vaporised fuel affected it:

 

Check tank levels/tank select valves/emergency shutoff valve.
Power on.

Power on aux fuel pump and observe fuel pressure rise to > 2PSI. Normally the aux pump runs at a slow steady rate, and the pressure rises promptly. However, with vaporised fuel, the pump 'chatters' at a high rate for 5 to 10 sec before settling down to a steady rate with the fuel pressure then rising. What is happening here is that the vaporised fuel is being driven through the orifice of the return line, and back to tank.

Power off aux pump and observe fuel pressure drop to zero. (This verifies that the fuel return orifice is not blocked: there is a check valve in the aux pump, if the return line is blocked, the fuel pressure will not drop.)
Start engine as normal. Cool air is now moving over the upper engine.
Do not take off immediately. Allow time for cool air and flow of fresh fuel to cool fuel delivery pipework etc.

NB: It should be noted that the fuel return line, with orifice, is piped back to a main fuel tank. There are several accounts online (including of an EFATO with inversion) where the operator has piped his return line to the gascolator in the engine bay. All this does is recirculate vapor and hot fuel in the engine bay, rather than flushing them back to the tank.

 

[skippydiesel]

IBob;

 

It sounds like you have just the two fuel lines (in/out of pump) over the top of the engine with the fuel manifold to carbs & return line on the firewall - If my reading is correct, this is a significant improvement over most set ups, that I have seen, where the fuel manifold and carb lines are also over the engine.

 

Would it not be a further improvement to have the fuel pump in/out lines under the engine ?

 

To do this the pump spigots would have to be pointing down ie rotate the pump cap.

 

As for your start & pre takeoff regime - sounds much like what I do, however on two occasion, when I was on an extended away trip, I experienced FVL on my take off run.  From memory this was my third departure, starting in the cool of first light, and the last before final destination. I had refuelled on my last stop but this one was for "comfort" Temperatures had risen to around 40 C . Engine start was normal, however she soon coughed/spluttered & "died". I was fairly certain the problem was FVL. Restart took some effort despite prolonged boost pump, I think I even tried choke. Once the engine started there was no more hesitation and I did an extra  long pre-flight check,  with engine rpm to 4000 in an attempt to clear any residual vapour. All good, lined up and commenced my run - engine lost power close to rotation speed. I aborted TO. Engine still running, I departed the runway at taxi speed, returning to the parking area. I was pretty sure what the problem was but removed my upper cowling for a visual check with special attention to fuel  filter condition - all good. Start was normal as was taxi & extended run ups.  Lined up power on, instant cough splutter, recovery & this time we continued with normal climb out  Rest of trip okay.

 

So the moral of the story is, with every mitigating tool (short of AvGas & delaying my departure to the cooler afternoon) available to me I still had a couple of FVL heart stopping moments.

 

Seems to me fuel lines under the engine might be worth a try BUT only if either Rotax make such a pump or it is safe fore me to rotate the cap.

[IBob]

Skippy, I believe I can see what you are getting at:
If the lines to and from the Rotax pump were under the engine, they would be less likely to heat and gas (in a static heat-soak situation) in the first place?

That seems reasonable to me too.
However, with my installation, since cold air comes in to the top of the engine and passes down over it, this would place the lines in a warmer place once the engine was running.

So there appear to be pros and cons with each of the two arrangements.

 

In my circumstances, I am comfortable with the arrangement I have.

But your circumstances are clearly different, and given your account, I can see where you are seeking to reduce the chances of a repeat.

 

Can I ask what the layout of your fuel system is in the engine bay, the various components and the location of them?
It would also be interesting to know how your fuel is piped from tanks to engine bay.

 

Attached are pics of my layout:

The aux pump is behind the square white shroud at lower left of the firewall.

The splitter is just above the shroud and to the left (against the firewall secured to the engine mount with a cable tie). You can see the line to the LH carb, with filter, coming forward from it.

 

[skippydiesel]

Veeery nice IBob !

 

I don't seem to tire from looking at Rotax installations and the subtle variation's initiated by either builder or factory.

 

The above story was when I had an 2000, ATEC Zephyr/Rotax 912ULS however the fuel reticulation lay out was/is similar to my new Sonex/Rotax 912ULS (yet to fly) with one significant exception. The exception is the inclusion of a fixed aperture fuel return line from the fuel manifold. The Zephyr had an adjustable pressure relief valve set up, quite large  & heavy, mounted on the firewall (this was before the simple return system had been developed/widely used)- it did work and had a fuel return to the tank. It is possible that the additional plumbing that this older system required  somehow retained fuel vapour, that resulted in the delayed engine problems that I experienced. 

 

 

[IBob]

Skippydiesel, that is the standard installation for the Savannah kit of the time, but with a great deal of fine engine detail taken from pics supplied by Reg Brost, the Australian agent I bought the kit from.
I hadn't realised how valuable the pics would be until I got well into the build: I worked latterly with an old laptop on the bench so that I could refer to them, also fire off questions from time to time to Reg and other builders.
I have no way of knowing how that setup would perform in the conditions you mention, but generally speaking it seems to me the Sav fuel system is very well thought out.

 

If your Zephyr had a pressure relief valve it perhaps would not have vented gas buildup as effectively as a fixed aperture.

And it's worth noting that, since the dynamic viscosity of gas is much smaller than the dynamic viscosity of liquid.... while the fixed aperture returns only a small volume of fuel (I've not measured it, but figures like 2L/hr are quoted) it will return or vent gas far more rapidly: hopefully you will find it works very well indeed.

 

The other factor that occurs to me is that much of the fuel system, including the hoses, has fairly low thermal conductivity. What that means is that it does not heat through quickly.........but once heat soaked, it will not cool quickly either. Furthermore, the hoses are shrouded in a protective sleeve, so not well able to dissipate heat. And what that means is that, with a hot start, it is probably the fresh fuel passing through the hoses that cools them. Initially, that may make more gas. But provided that can be vented, the system should flood and stabilise.

 

Keep us posted!

 

 

[Area-51]

Fuel vaporisation occurs due to the fuel sample reaching its boiling point (it will return back to liquid state as soon as it cools relative to line pressure); this can be determined by using a barometric tester such as a Doc Hodges Fuel Tester... Avgas has very tightly controlled specifications; so the vaporisation temperature/pressure can be predetermined; Mogas to a definitive level cannot... so two variables need to be addressed being 'Line Pressure' and 'Line Temperature'. However the vapor lock issue does not end there.

 

High carburettor body or float bowl temperatures will result in vapor lock instantly as higher pressurised fuel enters the float bowl, boiling then spirting out through the main and idle jet into the venturi; rough running coughs and spits until temperatures cool down. Hence the use of rubber isolating mounts for the carburettor and at times additional heat shielding plate between manifold and carburettor. V12 jaguars had the aircon system routed around the fuel lines for this reason; it worked.

 

As stated earlier in this topic the orientation of the upper pump body on a diaphragm pump has no affect on pump operation. Fixings may generally be torqued to a standard specification relevant to a specific thread sizing. However as the modifier you would have to be prepared to accept all responsibility for any subsequent loss or injury resulting from an operational failure.

 

That being said the threat of vapor lock occurring at any point in the fuel supply system is always present; especially at the float bowl, and the top of the fuel pump body. You can only do so much to prevent it, and it will always be caused by a rise in temperature beyond a certain point relative to pressure (somewhere for Mogas in the vicinity of 75C at 0'AMSL density altitude).
 

The affect of the return line and restrictor AD on the rotax 912 is to allow fresh fuel to continue flowing through the supply lines at line pressure, at all times, rather than lay somewhat static and suffer heat soak leading to vapor lock. The greater the volume of fuel that can be continually cycled through the lines at required operating pressure the better. Other than proximity to external heat sources and mechanical moving components, how the lines are routed, over or under, is irrelevant.

[facthunter]

If the fuel in the Carb. bowl is being vapourised the density falls there and the float sinks causing flooding. This is often not recognised when fault finding.. Carburetters on aero engines are risky for many reasons . There's good reasons to eliminate them from the equation. . Nev

[Thruster88]

Having heavy floats, we had a pair from a customer 912 weighing 12grams just last week, would exacerbate carburetor problems in hot weather.

 

One reason not to point the inlet and outlet down on the mechanical fuel pump is it would make it more difficult and maybe impossible to get rid of vapour,  vapour likes to stay at high points in a system.

 

Placing the electric boost pump in a cool location close to the tank would be ideal, not in the engine bay for heat soak reasons.  

[Area-51]

Was just editing and addressing this fact; true...

 

Hot restarts after being parked up are probably best undertaken by letting the auxilary fuel pump run for at least 2-3 minutes with the throttle set wide open prior to engine pre-start checks... and of course your engine pre-start checks require you to cycle and close the throttle back to idle...

 

if your aircraft has a tendency toward vapor lock it will still splutter a little on initial startup but will probably settle down within 5-10 seconds.

 

its important to be mindful that vapor lock is not a permanent condition; once the fuel vapor cools down it returns back to its previous liquid state and the "air bubble" disappears.

Edited September 24, 2022 by Area-51

[facthunter]

What does having the throttle set wide open do?  Nev

[Thruster88]

10 minutes ago, facthunter said:

What does having the throttle set wide open do?  Nev

In a fuel injected lycoming it dumps fuel under the aircraft, lots of fuel. Running the boost pump with the engine not running can be problematic for the 912 as well if carb float valve does not shut off. Engine fires while trying to start 912 powered aircraft are not in heard of.

Edited September 24, 2022 by Thruster88

[facthunter]

I know what it does on a fuel Injected Lycoming.   The 912 Carb is what I'm interested in . Nev

[Area-51]

Setting the throttle to wide open (* this will flood engines with constant pressure needle slide carburettors and certain mechanical injection) on CV carburettors allows any fuel bubbling up out of heat soaked carburettors to enter the inlet manifold and remain in situ in vaporised form ready for when cranking commences. The floats will become operational again as soon as fresh fuel reaches the float bowls preventing flooding (this has nothing to do with the subject or affect of "heavy floats")

 

If the carburettors are not suffering heat soak then no fuel will enter the inlet manifold however fresh cooler temperature fuel will still cycle through the pump and supply lines flushing out any existing vapor prior to cranking.

 

Pulling choke to mitigate vapor lock will have no affect on Bing64 but will have marginal affect on SU and Stromberg CV carburettors.

 

Position of return line restrictor can also affect ability to arrest vapor lock; too far away downline from the carburettors will prevent vapor quickly exiting back to lower pressure environment; so placing it back at the fuel tank is not a good idea.

 

Return line exit position is also critical; it needs to be far away from any heat source to allow sufficient time for fuel to cool down. It's also critical to ensure that where ever it is returning to cannot overfill and cause continued pumping of fuel out through fuel vents as the 912's pump will happily empty a 40L tank dry in about 15 minutes.

[IBob]

How about opening the oil inspection hatch on shutdown, to avoid the heat soak in the first place?

[Area-51]

Anything that gets air moving over heated components helps; keeping the fuel cool and moving through fuel lines helps the most. If vapor lock continues to be experienced during taxi/flight op's then you either have a fuel quality issue, line pressure issue, or a heat soak/source issue. The situation doesn't get any more simple than that.

 

Departing frequency area...

[skippydiesel]

6 hours ago, IBob said:

How about opening the oil inspection hatch on shutdown, to avoid the heat soak in the first place?

Always done but it doesn't " avoid" heat soak just helps to reduce it's effect along with all the other strategies.. 

[onetrack]

Has anyone considered coating their exhaust system with Cerakote heat reducing ceramic coating? I have no experience of the product, but I know people who have utilised it on vehicle engines, and it has made a substantial difference to engine bay temperatures.

[facthunter]

It does reduce radiated heat but UP's the pipe temperature which may cause other issues.   Nev

[skippydiesel]

6 hours ago, onetrack said:

Has anyone considered coating their exhaust system with Cerakote heat reducing ceramic coating? I have no experience of the product, but I know people who have utilised it on vehicle engines, and it has made a substantial difference to engine bay temperatures.

Not sure but have heard that Rotax advise against exhaust wraps/coatings (check out Nev's comment above)

[skippydiesel]

Soooo!

 

Has no one else;

• Considered rotating the fuel pump "cap" ???
• Got an opinion on the pro's/cons of doing so ???

[BrendAn]

its funny after reading this thread i watched a youtube video last night about a bloke travelling across the usa in a grand torino. when he was in the mountains it was vaporising the fuel in the thin air,  he purchased a fuel filter with a tank return nipple on it and ran a return hose and it fixed the problem straight away.

[Blueadventures]

This information is from a very experienced and knowledgeable Rotax operator.  It may be worth considering if anyone experiences a hard to restart situation. (I'm adding the info into my trouble shooting information.)

 

"Sometimes it’s not vapor lock - it’s the carbs getting really hot from radiated heat from the exhaust headers after shut down when there is no cooling flow.

A way of checking this is if it starts fine straight away and from 5-20mins it can be very difficult to start.  So, either avoid this time period for restarting or try it with around 1/3 throttle opening.  If that doesn’t work leave it to cool.

The timing retard for soft start on the later engines seems to help a lot. Some people have made heat shields under the carbs - which helps a little.

It’s pretty common on all cowled 912 installations. Even un-cowled trikes can suffer it on occasion."

Edited September 27, 2022 by Blueadventures

[facthunter]

All this amounts to "Houston, we have a problem". When every decent motorcycle is injected why use 2 crude carburettors on this motor?  The location is wrong from the  flow pulsing and heat. Inject and it's fixed.  Nev

[skippydiesel]

1 hour ago, facthunter said:

All this amounts to "Houston, we have a problem". When every decent motorcycle is injected why use 2 crude carburettors on this motor?  The location is wrong from the  flow pulsing and heat. Inject and it's fixed.  Nev

All true Nev.

 

Unfortunately, for the extra dollars demanded,  the "fix" is the Rotax 912S/iSc Sport & the 915/916 iSc at many many thousands of dollars more than the carb versions.

 

"You gets what you pay for" and for most sport pilots (at least in Australia) the carburettor variants deliver the best bank for the buck.

 

If you really want the extra fuel econamy & freedom from the potential for carb ice, one of the aftermarket fuel injection systems may be the way to go but then this is only applicable to "experimental" aircraft and probably wont achieve any significant improvement in power - so hard to justify for those doing less than a few hundred flying hours /year.