Where does the fuel go? - Page 3

AlexPeterson · #21 03-08-2015, 09:04 AM

I have suspected the erroneous readings were from reversing flow. Between the two pumps, there are a couple springs and accumulators in the low pressure pump systems. The rotor cannot distinguish between forward and reverse flow, so reported flow is erroneously high.

Just a theory...

bpattonsoa · #22 03-08-2015, 10:29 AM

I have the issue in my -10. It has the stock fuel system with a Flowscan meter left over from my racing days. With the pump on I get a significant increase of indicated flow.

The increase of flow is NOT real. The EGTs do not change when pump is switched off and the flow "decreases". The meter is calibrated with the pump off.

In the Reno racer I worked on, we had two flow meters because we used a fuel rail and flowed 90GPH in the rail at all times. Fuel flow was the delta. I put two Flowscan accumulators in the system and they made it much more accurate. Having straight inlet and outlet lines is also critical with the Flowscan. At best we had about a 5 to 10% error, up and down between readings. The JPI 930 engine instrument saved data at 2 second intervals, and the variation was between individual data points.

As I recall, having worked with a few Flowscan meters, they do not have a bypass.

Henryrifle · #23 03-08-2015, 06:25 PM

So glad this thread is active right now. I have this same issue when my boost pump is on in an O-320 equipped -6. When the boost pump is on fuel flow reads between 12 and 15 GPH depending on the phase of flight. With it off reading are normal between 8 and 10 GPH.

Glad to know there are some cures for this.

Hank

Weasel · #24 03-10-2015, 06:54 AM

I have not disassembled my spare fuel flow transducer so all of this information is based on inspection from the external. I was able to look inside of the In and Out ports and see what appeared to be a paddle style wheel protruding into the fuel stream. Made into the side of the paddle was what appeared to be an interrupt disk to trip a photo cell pick-up.

It is also possible that the part of the wheel that I am referring to as a Photo Cell Interrupt Disk may be a gear that is driving something lower in the housing that I cannot see

In the pictures the paddle appears to obstruct the flow but it looks like the cavity that holds the wheel is hollowed out and somewhat larger so that in the event of a locked wheel there can still be adequate flow.

I tried to draw a sketch to give an idea of what I see inside the port. The hole is fairly small and very hard to get a clear picture of the paddle.

DanH · #25 03-10-2015, 08:35 AM

Opened a conversation with EI yesterday, which I expect to continue today with emphasis on the "why".

So far...

(1) There is no side passage. Tech said flow is straight through (no swirl passage like a Flowscan), which is what Joel's photos seem to show.

(2) Must install with the wire facing up. Tech said the "why" is because facing down may trap air in the rotor cavity. Makes sense, looking at the photos, which seem to show the rotor on the bottom. I'm guessing that it is fine in any position that does not trap air.

(3) Rotor interrupts the beam from a light source, 68,000 per gallon, so the signal (as Matt said) works out to be a tad less than 19hz per gallon per hour (18.888). A 10 GPH flow would be 188hz. For my 390, min WOT fuel requirement per the manual is 17.5 GPH, or 329hz.

(4) The engine driven pump is driven by a half-speed idler gear, so it would oscillate flow in a start-stop pattern at one cycle = 1/2 RPM....for example, 22.5hz at 2700.

DanH · #26 03-12-2015, 06:38 AM

Not much progress with EI.

The official position is that the FT-60 (Red Cube) transducer is to be located downstream of all pumps (rather than between the boost pump and the engine driven pump, or between the tank and the pumps), because:

...the installation location that was referenced is a safety concern. When fuel is pushed thru the transducer there is little to no chance of vapor lock. When pulling fuel thru the transducer there is a decent chance of vapor lock if there is an air leak.

Regarding the "why" of increased flow indication with the electric pump ON:

...the indication that is being seen is only in the case of installing the transducer in the incorrect location. This is why we are not particularly interested in the source of the indication error.

Rupester · #27 03-12-2015, 07:57 AM

If the Red Cube is located between the FI servo and spider, there should no increase in flow upon increase in fuel pressure. The servo (also called the fuel regulator) has a diaphram-actuated ball and seat specifically to assure fuel flow is a function of only servo inlet air velocity .... and independent of fuel pressure.
My Red Cube is located between servo and spider and I see no change in flow with the booster pump on or off. I do see a ~2-3 psi increase in fuel pressure, however.

Tumper · #28 03-12-2015, 08:27 AM

I started a conversation with one of the manufactures of flow sensors a few days ago as a result of this post and I thought it was worth sharing. I am going to leave out names for the obvious reason.

< Me> Dear ?Company Representative?,

I have just built a Richard VanGrunsven designed RV-9 airplane and I have a question about the fuel flow. My plane has a Lycoming Fuel injected engine with an electrical fuel pump ahead of a mechanical fuel pump. I have also installed a ?Company? sensor after the electric fuel pump.

My fuel flow goes up noticeable when I turn on the electric fuel pump and it has been suggested that if I install a pulsation dampers it might correct the problem. Do you have any suggestions or thoughts?

Thanks for your time.

Dean Eiland

<Company Representative> Hello Dean,

This sounds like an interesting problem. The most common issue when a reading goes high is that bubbles are getting in the fuel line and going past the sensor. If it is this type of problem then adding a pulsation damper will not help.

A pulsation damper would help if the issue is caused by pulsation coming from the fuel injection pump downstream. However from your email it sounds like the issue is more closely related to the powering of the electric lift pump.

My supervisor has left for the day but I will ask him tomorrow on his thoughts about this. He has a longer history with the product and may have some additional thoughts. In the meantime can you give me a rough idea of how far (in fuel hose) is the sender from the electric lift pump and the sender from the engine. Also is it possible to move the sender so that it is upstream of the electric lift pump but downstream of any fuel filters. This could fix the issue.

Talk to you soon,

?Company Representative?

<Me> Dear ?Company Representative?,

Thanks for the reply. I have attached a picture of my electric fuel pump assembly and flow meter, while it was still on the table. The mechanical fuel pump is about 2? down line of the flow meter.

I don?t think there would be enough straight pipe in front of the electric fuel pump to install the flow meter between the fuel selector and the electric fuel pump.

Thanks again,
Dean

NEXT DAY
<Company representative> Hi Dean,

?Company? builds and sells aircraft sensors only to authorized avionics companies and aircraft manufacturers. When an order is placed aircraft sensors are manufactured to the customer?s requirements. ?Company? does not sell aircraft sensors to end users, (pilots, owners, mechanics etc.).

All Customer Service and Technical Support is provided to the customer by the avionics company which sold the sensor. These companies are the ones that have aviation engineers and other aviation experts on staff.

Best Regards,

?Company Representative?

<Me> Attorney got to ya?

Thanks anyway,

Dean

<Company Representative> Yes. I almost was fired.

<Me> Sorry. I should have told you up front I have a contagious incurable disease.

bpattonsoa · #29 03-12-2015, 09:00 AM

Thumper

Your mistake was not telling them it was installed in an airboat.

I suspect the issue is a high frequency pulse (vibration?) in the fluid produced by the electrical fuel injection pump. It would introduce oscillations in the spinning very low mass rotor. The sensor would then double count some, not all, of the tabs on the rotor.

Unfortunately measuring this type of pressure variations requires very sophisticated instrumentation. The pump manufactures most likely would have this information, if they ever measured it.

When the flow sensor is after the engine pump, which operates at a much lower frequency, the high frequency pulsations are damper by the length of fuel line and by the diaphragm in the pump.

I plan on changing to a red cube and moving it into the engine compartment.

rvbuilder2002 · #30 03-12-2015, 10:55 AM

This phenomenon has been seen on RV's since the early 90's when more economical fuel flow computers hit the market and quickly became popular with RV builders.
The primary two I can think of is Electronics International (E.I.) FP-5 and Matronics Flow Scan.
At that time I think all manufacturers (even Shaden, Vision Micro Systems, etc.) used the flowscan brand transducer.
It has been only in recent years that I.E. began producing their own because the price of the flowscan (which had always been way over priced) began to get totally ridiculous.

So, this is nothing new. It occurs with the E.I. sensor or the flowscan.
I have a feeling that the problem also exists to some degree in a manufacturer recommended installation (sensor after the engine driven pump), but since that pump is never turned off, it is not noticed.... During calibration, K factor adjustments get made to compensate, and since the level of error is constant (engine drive pump is always on), there is never any noticeable error or variation.
I think it might have been Matt (Matronics) that originally introduced the use of an anti hammer device to resolve errors on installation of his units. It is no surprise that contemporary sellers do not care what the cause is, and instead just say "follow the recommended installation procedures".