Pipistrel on Fuel Vaporisation

[skippydiesel]

Below is an excerpt from a letter circulated by Pipistrel & published by RAA in its recent news letter (see full letter)  PVS-2304_2024-Circular_letter.pdf (raaus.com.au)

 

"Date: 23th April 2024 Subject: Vapor lock preventive actions, applicable to all Pipistrel aircraft equipped with Rotax 912, 912 iS and 914 engine types

 

NOTE: Steps specified below act as additional content to the existing procedures outlined in the applicable aircraft flight manual.

 

BEFORE FLIGHT:

• If temperatures rise in the area of run-up, it is advised to increase RPM above idle in order to force more air through the engine compartment, which will maximize the cooling.

• If aircraft is equipped with auxiliary/booster fuel pump, pay attention to its sound – any significant “non-standard” deviation in pitch of the sound could indicate presence of the trapped vapor.

• After takeoff, reduce climb angle to increase airflow through the engine compartment, which will maximize the cooling.

AFTER FLIGHT:

• If situation permits, park the aircraft with nose pointing into the wind.

• Before shutdown, let the engine cool down at idling.

• Keep the doors opened – this will allow heat to escape from the cabin, which is especially beneficial at configurations with fuselage tank installed."

 

Not having access to the Pipistrel "flight manual" I don't know what additional information it may containe on the subject/management of fuel vaporisation - I am concerned that the above letter does not go nearly far enough in that it has failed to refer to:

 

• the main/most likely time for this phenomena to occur, is on the ground, after flight, on a hot day 
• opening any inspection doors/hatches to allow the hot air from heat soaked engine to escape more readily.
• fuel pressure indicators (other than a vague reference to boost pump, if fitted, sound)
• hot starting problems
• the assistance a correctly fitted return fuel line will give in helping to minimise/alleviate this problem.
• the safety implications in not having a boost pump.
• the need  to run the boost pump until posative/acceptable fuel pressure is achieved - this may include during attempted start.
• extended run up time, to not only assess the engines condition but to further aid in clearing fuel vapour from the system.
• having a "plan" should the engine not deliver full power/die on take-off/climb out.

 

The advice to :

"..........let the engine cool down at idling." may not be good , in that most aircraft will taxi, inducing a flow of air through the cowling/cooling systems befor shut down. Further engine running while aircraft is stationary may only exacerbate the heating of fuel lines, leading to a greater chance of vaporisation. I would recomend stopping the engine as soon as aircraft has parked.

Keep the doors opened – this will allow heat to escape from the cabin, which is especially beneficial at configurations with fuselage tank installed." will do no harm , it sounds like BS to me. While it is desirable to keep all onboard fuel cool, the fuel stored in the tank(s)will not be influenced by the heat from the engine, which is the main generator of the problem, often referred to as "vapour lock"

[BurnieM]

You left out the bit about Pipistrel considering 26 degrees to be hot.

 

Agree, there is not much useful advise in this notice.

 

Did I not see similar warnings from other manufacturers 2-3 years ago ?

 

Edited May 3 by BurnieM

[skippydiesel]

I passed on my comments to Pipistrel - will be interesting to see of they reply/have anything to say.

 

I think RAA should be a tad more discriminating in the "stuff" they reprint & circulate to the membership.

Edited May 3 by skippydiesel

[Area-51]

The related issue is not likely causal of vapour lock. The cause is likely due to the design of the fuel delivery system between wing tanks and gasolator mounted on the firewall. These inherent flaws have been rectified within the 2016 Virus SW121 variant onwards by employment of an Andair twin coupled fuel/return selector valve and vapour accumulator tanks behind the rear cockpit bulkhead for each wing. All other earlier Sinus and Virus airframes employ a seperate fuel selector valve at each wing tank with the fuel return feeding back to the left wing only. This allows a selected right tank to be pumped dry within 15-20min flying if the left tank is full, causing premature unexpected fuel starvation. This fact is well documented upon other forums along with the official responses offered by the manufacturer, which, at the time was, "fly with the left fuel valve partially open"... It could be presented that the response is questionable, inadequately addresses a critical safety issue, has been recognised by the factory as a real issue due to the upgraded delivery system of the 2016 Virus SW121, and may leave the manufacturer exposed to serious legal action for not issuing an AD to rectify known existing threat to operator safety of pre SW121 produced airframes.

 

Images of fuel delivery systems attached for explanation. 

Edited May 3 by Area-51

[Thruster88]

7 hours ago, skippydiesel said:

I passed on my comments to Pipistrel - will be interesting to see of they reply/have anything to say.

 

I think RAA should be a tad more discriminating in the "stuff" they reprint & circulate to the membership.

They won't say anything because the less said the better for legal reasons.  

[skippydiesel]

Area 51

 

Wow! the Pipistrel way of addressing, the fuel return system,  what I see as relatively simple concept, is very (unnecessarily?) complex. Perhaps there is just no space, closer to the engine, for a header tank.

[Area-51]

5 hours ago, skippydiesel said:

Area 51

 

Wow! the Pipistrel way of addressing, the fuel return system,  what I see as relatively simple concept, is very (unnecessarily?) complex. Perhaps there is just no space, closer to the engine, for a header tank.

The Pipistrel factory rectification adequately addresses the earlier causal issue of in flight fuel starvation; it provides switched return line routing to the selected source tank along with escape routing for vapour accumulation to the selected source tank breather at the filler cap.
 

An additional observed phenomena of the pre SW121 fuel supply system is asymmetric flow rates from each wing tank while operating in level cruise flight with both fuel taps selected; the right tank would still empty itself prematurely over a longer time frame. (static flow rates while not in flight measure equal). It should be noted there is no compensation porting employed between left and right wing tanks. Employment of engineered dynamic pressure enhancers at the filler cap breather aided flow rate symmetry however was not consistently reliable to mitigate asymmetric flow and premature exhausting of fuel reserve of the right wing tank. The only reliable performance procedure determinable to keep AC balanced was to fly on the left wing tank first until half reserve, then switch to right tank until half reserve then back to left, which will be almost full again after 10-15min, then back to right tank for 10-15min, it will be exhausted, back to left tank, which will be 3/4, for the remainder of flight if required... Alternatively, fly on left tank first till exhausted, then fly on right tank 15-20min till exhausted, then on left tank for remainder of flight. FINAL NOTE, flying with both taps selected can still result in fuel starvation due to suction of air into fuel delivery system when right tank becomes prematurely exhausted. Astute monitoring and management of fuel reserve levels is therefore critical. The post SW121 fuel delivery solution mitigates all these fore mentioned aspects.

 

It should be appreciated the above outcome of the pre SW121 fuel delivery system applies to any similar scenario where the fuel return supply has been routed back to a single tank only without provision of adequate compensation porting between additional selectable source tanks toward redistribution of reserve overflow.

 

Hopefully this knowledge can prevent pipistrel users ending up in the trees after take off or on approach.

[IBob]

On 04/05/2024 at 5:19 PM, Area-51 said:

The only reliable performance procedure determinable to keep AC balanced was to fly on the left wing tank first until half reserve, then switch to right tank until half reserve then back to left, which will be almost full again after 10-15min........

That seems to suggest the return line is delivering at a very high rate.
I believe Skippydiesel did a fair bit of work recently with his return line orifice, and may be able to give us some rates from the various sizes he tried.

As for uneven feed, I am told all aircraft suffer from this to some degree. I have been able to reduce this on my Savannah by doctoring the angle of the ends of the tank breather pipes. But if I was building again I would look at some means of cross-porting the upper tanks, with a single breather, as Cessna do. Having said that, I have seen some odd fluidics in industrial settings, where separate flows are siamesed together: sometimes the flow from one input would take over entirely, excluding flow from the other, though admittedly this was usually in much higher flow situations than an aircraft fuel system.

[FlyBoy1960]

Uneven feed is due to friction losses in the hose. If the hose from the right tank is longer than the hose from the left tank it is going to generate more friction losses which are very easy to calculate. Also not flying perfectly level will have an effect, if you are right wing down then you are probably going to use more from the left wing and vice versa because the higher tank will effectively have more head  in the fuel flow will find the easiest route.

 

Friction losses are significant in small tubes over any distance, an example most people can understand as the water pressure at the tap comes out screaming but at the end of the 10 m hose it is nowhere near the same pressure, this my friends is friction loss.

From the Google

1. Multiply the length of the pipe L with the volumetric flow rate Q raised to the power 1.852.

2. Divide this by the pipe diameter D raised to the power 4.87.

3. Divide this by pipe roughness coefficient C raised to the power 1.852.

4. If all dimensions are in metric units, multiply the result from Step 3 by 10.67 to get the frictional head loss. If you're using imperial units, multiply by 4.52 instead.

[skippydiesel]

"Friction losses are significant in small tubes over any distance, an example most people can understand as the water pressure at the tap comes out screaming but at the end of the 10 m hose it is nowhere near the same pressure, this my friends is friction loss."

 

Me thinks you are confusing flow & pressure - The pressure should remain constant, while the flow is reduced, due to such factors as friction.

 

In small (carburettor) aircraft we monitor the health of the fuel supply system in pressure (kph, psi, Bar) but actually use the fuel in volume (L/hr). All the pressure in the World may not deliver sufficient volume of fuel for your engine to run.

Edited May 6 by skippydiesel

[IBob]

FlyBoy certainly friction losses may be a significant element, though they are far more noticeable with higher flow rates.
 

I believe the difference of pressure in separately vented tanks to be far more relevant:
Eg: The fall from my fuel tanks to my fuel selector manifold is approx 250mm (the selector is on the RH baggage wall, where I can see it). Add approx 200mm for full tanks = 450mm which gives a fuel pressure at that point of just 0.49PSI
So very small differences in upper tank pressure due to slightly different venting, prop swirl etc can result in proportionally large differences in fuel pressure from L and R. And that will only get worse as tank levels fall.
Aircraft with floor mounted fuel selectors are marginally better off. In mine that would double the pressure at the selector, but certainly not enough to minimise the effects of unbalanced venting.

[facthunter]

Pressure doesn't remain constant unless there is no flow.. That's when you have a confined fluid. Pressure measuring at various points would show reducing pressure . Nev

[skippydiesel]

I agree with you Nev however many get confused by the relationship between flow & pressure. 

 

In a 912 ULS (fitted to a Pipistrel) no fuel pressure is required to run the engine. However to get the fuel from wherever it is being stored (tank), pressure (gravity & or pump) will be required to deliver the fuel.

Depending on the height above/below delivery point (carb float bowl) that the supply is, plus the effects of tube/hose ID and length (frictional losses) will determine how much pressure is required to deliver the appropriate volume/flow, to the engine at full power (max demand).

 

Note: Other effects will be pipe bend radius/valve types & number/ filters types / return line jetting.

 

Hope I have covered all/most of the issues.

[FlyBoy1960]

But there is only gravity flow, therefore very low pressure, and significant friction losses.  The fuelis only a dribble until it gets to the Rotax fuel pump.

[facthunter]

The pump can't create volume. It can produce negative pressure on the suck side and positive on the other. Gravity produces pressure also from energy of position. Potential energy. like height in an aeroplane.   Nev

[skippydiesel]

11 minutes ago, FlyBoy1960 said:

But there is only gravity flow, therefore very low pressure, and significant friction losses.  The fuelis only a dribble until it gets to the Rotax fuel pump.

No gravity flow in most low wings - pump all the way.

 

In a 912ULS installation, I would question the sanity of anyone relying on gravity feed - particularly when it comes to fuel vaporization mitigation.

[facthunter]

Who does but it's fine on a C-150 with a Cont. 0-200.  Fluid won't flow uphill without assistance.   Nev

[Area-51]

regarding the return line orifice. Rotax have there recommendations clearly stated in SB's. The Pipistrel uses the factory recommended size. It takes approximately 20min to pump the 50L right tank dry at a 12L/hr cruise burn. So that works out to around 135L/hr flow through the little hole at 5psi; about 2L/min circulated back to the left wing tank.

[IBob]

Area51 yes I know Rotax have a recommended return line orifice.

But 2L a minute doesn't pass the reality test: that's about a cupful every 6 seconds, go fill a cup in 6 secs and you will see there is no way that volume could come out of such a tiny hole...

Come to that, why would anyone design an engine that takes 17LPH in cruise, and circulate fuel at 135LPH???

Edited May 6 by IBob

[IBob]

Skippy has taken some actual measurements on orifice size vs flow.........where is he when we need him?

[Area-51]

5 minutes ago, IBob said:

Area51 yes I know Rotax have a recommended return line orifice.

But 2L a minute doesn't pass the reality test: that's about a cupful every 6 seconds, go fill a cup in 6 secs and you will see there is no way that volume could come out of such a tiny hole...

Math don't lie... 🤷🏼‍♂️ 45L of fuel to zero fuel in 20min, its only going out through the little hole.

[facthunter]

What actual figures are being encountered?  You'd have to have a pretty good idea of the reality. Using fuel bleed to cool the fuel seems to add risk..   Nev

[IBob]

Area51 I think you're missing my point: I'm not disputing whatever may have been quoted above, I'm wondering what size orifice they are using for that sort of flow.
My Savannah has individual tank valving, but returns only to the RH inner. And I can guarantee that the return rate is not a fraction of the figure you mention.

[IBob]

Scroll to last entry here:
https://www.rotax-owner.com/en/912-914-technical-questions/8008-fuel-return-line-for-a-912ul?start=10

[Area-51]

My own AC came with 0.50mm orifice in return line and maintains 3.5psi on the engine pump and 5.0psi on auxiliary with each flowing at 60 and 105 L/hr. The return feeds back into the reserve tank so what ever is left in main always directs back to reserve before overflowing to main; so fuel has plenty opportunity to return to liquid state.

 

Anyhow thread has now drifted some what. Throw the formulas out the window.

 

Get a bucket, disconnect return line, start engine and measure 1 minute of fuel quantity and multiply by 60. Reconnect return line, and check for leaks and hose clamp security before further flight op's.