Fuel run

[Jettison]

In my RV-7, the fuel runs into the engine driven fuel pump through a 90 degree bend at the pump. After the pump, it makes another 90 degree bend downward (where the pressure is measured), then another 90 degree bend to the horizontal, and then a 45 degree bend just before the fuel flow transducer. Seems to me, straighter lines would be best.

With just the engine driven pump running at cruise, my measured fuel pressure seems to bounce up and down some, and trends towards the low side (17-20 psi). I have the IO-360 angle valve engine.

Is straightening out this fuel run desirable? Might it be causing low/bouncing fuel pressure? What can be done about it? What fittings, etc?

 

[lr172]

In my RV-7, the fuel runs into the engine driven fuel pump through a 90 degree bend at the pump. After the pump, it makes another 90 degree bend downward (where the pressure is measured), then another 90 degree bend to the horizontal, and then a 45 degree bend just before the fuel flow transducer. Seems to me, straighter lines would be best.

With just the engine driven pump running at cruise, my measured fuel pressure seems to bounce up and down some, and trends towards the low side (17-20 psi). I have the IO-360 angle valve engine.

Is straightening out this fuel run desirable? Might it be causing low/bouncing fuel pressure? What can be done about it? What fittings, etc?

Bends can create turbulence and some here believe this can introduce air bubbles (I am not one of them). However, generally speaking bends in fluid pipelines do not reduce pressure. May be slight declines in velocity or flow rate, but the mech pump is so over sized (approx 45 GPH capacity) that you cannot experience enough flow reduction to end up seeing a pressure reduction. IMHO, your pressure issues lie elsewhere. Bends in fittings WILL NOT create oscillating pressures. Maybe one of our members is an engineer that studied fluid dynamics and can chime in. I did not.

 

[Carl Froehlich]

Bends can create turbulence and some here believe this can introduce air bubbles (I am not one of them). However, generally speaking bends in fluid pipelines do not reduce pressure. May be slight declines in velocity or flow rate, but the mech pump is so over sized that you cannot experience enough flow reduction to end up seeing a pressure reduction. IMHO, your pressure issues lie elsewhere. Bends in fittings WILL NOT create oscillating pressures.

90 degree bends, and other poor plumbing practices increase the risk of vapor lock (not air bubbles but fuel vapor bubbles). This is even more of a concern for those running mogas. The suction of the pump drops the pressure on the fuel to get it to flow. More flow restrictions cause more pressure drop. Once past the fuel pump the pressure is restored - so attention should be giving to the suction side of the pump.

90 degree fittings are also not compatible with most fuel flow sensors.

Carl

 

[rvbuilder2002]

I agree with Larry and Carl, so that leaves you with the likelihood that the performance of your mechanical pump has degraded.
I have seen the same thing in the past and replacing the mechanical pump makes pressures back to normal.

 

[TS Flightlines]

Agreed. Anytime you can have a relatively straight run is optimum. But with other constraints under the cowling, sometimes using an angled fitting is somewhat necessary. Best to use a sweeping angled hose end instead of a block style---IE--looks like a AN833 or AN837 fitting with a nut, but these types of fittings have been used by a MAJOR supplier for decades with no issues.

I would try against using multiple angled fittings where any leverage forces could loosen the fittings and create a leak. Try and keep the plumbing far away from exhaust heat if you can. There is a story behind this pic--Lets just say it could have been catastrophic.

Tom

 

[lr172]

More flow restrictions cause more pressure drop. Once past the fuel pump the pressure is restored - so attention should be giving to the suction side of the pump.

Carl

I agree with that statement, but a 90 fitting is NOT a restriction; Yes, flow direction changes and some velocity and flow rate is lost, but that is not a restriction (classic definition implies forcing the liquid through a smaller size) and won't reduce pressure until flow rate is low enough that the pump cannot keep up. Possibly you could cite the research that shows gasoline will move from a liquid state to gas state when going through an elbow. I have never seen data to support that. Plenty of data to say that the turbulence of the elbow will reduce flow rate, but nothing about state changes. WHile not a phycist, it doesn't seem to follow the general rules of physics. I have plenty of 90 elbows in my home water system and have never seen air bubbles produced.

I agree that avoiding sharp turns in a fluid system is best practice, but struggle to see how violating that will create fuel vapor.

Larry

 

[Pilot135pd]

...I have plenty of 90 elbows in my home water system and have never seen air bubbles produced.

As a teacher I always like to give students real world examples that they can relate to so it's easier for them to grasp a concept. This one gave me a clear example of your view on this. Thanks.

 

[Freemasm]

I agree with that statement, but a 90 fitting is NOT a restriction; Yes, flow direction changes and some velocity and flow rate is lost, but that is not a restriction (classic definition implies forcing the liquid through a smaller size) and won't reduce pressure until flow rate is low enough that the pump cannot keep up. Possibly you could cite the research that shows gasoline will move from a liquid state to gas state when going through an elbow. I have never seen data to support that. Plenty of data to say that the turbulence of the elbow will reduce flow rate, but nothing about state changes. WHile not a phycist, it doesn't seem to follow the general rules of physics. I have plenty of 90 elbows in my home water system and have never seen air bubbles produced.

I agree that avoiding sharp turns in a fluid system is best practice, but struggle to see how violating that will create fuel vapor.

Larry

Comparing a pressurized water system and an induced flow gasoline system are really an apples/oranges comparison.

It's all about the local static pressure of the fluid. Water flow through your home piping is fed from a positive pressure source, usually associated with elevation (water tanks). Gasoline being drawn into a pump (many times up hill) is experiencing less than atmospheric pressure. Add losses associated with flow (restrictions through a non-functioning boost pump, check valve, filter, etc) further lowers that local static pressure. Get below the fluid's vapor pressure, liquid turns to gas phase. Our relevant fluid pumps can't pump vapor. Game over.

Someone mentioned real world examples. Want to see static pressure converted to dynamic (positive pressure becomes a vacuum?) Look at your garden hose-end sprayer. Total pressure is conserved (ignoring friction) so the dynamic pressure from velocity increases -`> static pressure decreases. 40 or so psig water is accelerated through a very small orifice in the device and the static pressure is now a vacuum drawing liquid from the attached canister.

In the shop tonight? The holes in the side of nozzle of your shop air gun are there to increase airflow. The 120 psig compressed air comes through a very small orifice. High velocity -> static P drops below 0 psig -> draws atmospheric air through the said holes.

Lots of threads on fuel system, suction-side layout. Not always enough to know that no vapor formed that test. What is the margin so when hotter days/operating environments, suction filters fowl, etc. will the powerplant keep running?

 

[lr172]

Comparing a pressurized water system and an induced flow gasoline system are really an apples/oranges comparison.

It's all about the local static pressure of the fluid. Water flow through your home piping is fed from a positive pressure source, usually associated with elevation (water tanks). Gasoline being drawn into a pump (many times up hill) is experiencing less than atmospheric pressure. Add losses associated with flow (restrictions through a non-functioning boost pump, check valve, filter, etc) further lowers that local static pressure. Get below the fluid's vapor pressure, liquid turns to gas phase. Our relevant fluid pumps can't pump vapor. Game over.

Someone mentioned real world examples. Want to see static pressure converted to dynamic (positive pressure becomes a vacuum?) Look at your garden hose-end sprayer. Total pressure is conserved (ignoring friction) so the dynamic pressure from velocity increases -`> static pressure decreases. 40 or so psig water is accelerated through a very small orifice in the device and the static pressure is now a vacuum drawing liquid from the attached canister.

In the shop tonight? The holes in the side of nozzle of your shop air gun are there to increase airflow. The 120 psig compressed air comes through a very small orifice. High velocity -> static P drops below 0 psig -> draws atmospheric air through the said holes.

Lots of threads on fuel system, suction-side layout. Not always enough to know that no vapor formed that test. What is the margin so when hotter days/operating environments, suction filters fowl, etc. will the powerplant keep running?

don't disagree with anything you said, but a 90* bend is not going to drop the pressure much, as it is not a restrictor per se, just a flow reducer and the -6 lines most often used have a flow capacity greatly exceeding the engines requirements. A 10 mmHg reduction in pressure creates a .5*F reduction in gasoline's initial boiling point. According to the feds, avgas has an initial boiling point of 122-158*F. It is going to take a significant restriction with a major vacuum (i.e. pressure drop) to drop the boiling down into the 80-100* typical cockpit temps. So, while I agree with your science I do not agree that a few 90* elbows are going to cause vapor in the vast majority of cases. If you are talking about a 90 elbow under the cowl after a good heat soak, then it MAY have a very small impact on whether or not vapor is produced. However, given the typical 180* under cowl temps after a heat soak, vapor is all but guaranteed regardless of elbows.

Not advocating here that folks don't follow best practices, only trying to counter this urban myth that an elbow WILL cause vapor lock.

 

[Pilot135pd]

Someone mentioned real world examples. Want to see static pressure converted to dynamic (positive pressure becomes a vacuum?) Look at your garden hose-end sprayer. Total pressure is conserved (ignoring friction) so the dynamic pressure from velocity increases -`> static pressure decreases. 40 or so psig water is accelerated through a very small orifice in the device and the static pressure is now a vacuum drawing liquid from the attached canister.

Never seen vapor or steam come out of my garden hose, even in our 107 degree temps this year. It's still liquid with no bubbles when it comes out.

 

[Freemasm]

Never seen vapor or steam come out of my garden hose, even in our 107 degree temps this year. It's still liquid with no bubbles when it comes out.

Hose end sprayer like for applying lawn chemicals. Water pressure drops below atmospheric pressure (due to very high velocity) to draw your lawn care carcinogens out of the attachment.

 

[Desert Rat]

To the Op's original questions;

If I'm understanding your description correctly, it sounds like you've got a 90 on the inlet side of the mechanical fuel pump, then a 90 on the outlet with an extra port (it's in vans online store as part number KB-090-T) to take the fuel pressure reading, which is a pretty standard setup.

However, if you've got another 90 attached to that one to aim the output somewhere else, thats just a bad practice due to the potential for the engine rocking about the crankshaft to cause them to work against each other and start to loosen up.

If you've got to have that setup, making the best of a less than ideal situation would be to position the fittings so that the legs of the fittings are perpendicular to the crankshaft to minimize the potential for them to loosen up.

Why the second 90? Is it just to get your fuel line pointed back outboard? If so, there are other better solutions.

 

[lr172]

Hose end sprayer like for applying lawn chemicals. Water pressure drops below atmospheric pressure (due to very high velocity) to draw your lawn care carcinogens out of the attachment.

pretty sure that the water pressure DOES NOT drop below atmospheric pressure. In fact it is WAYYY higher. That high velocity water goes through a venturi that is has another tap in the center of it. The high velocity water creates a low pressure state in that tap and it is that which creates the suction on the tube going down into the carcinogens. You are confusing the venturi principles used with unpressurized air, where the pressure does drop below atmo in the neck of the venturi (how a carb works). The pressure into the device you showed starts around 50 PSI . Yes, it drops some in the venturi, but NOWHERE close to atmospheric.

 

[Freemasm]

pretty sure that the water pressure DOES NOT drop below atmospheric pressure. In fact it is WAYYY higher. That high velocity water goes through a venturi that is has another tap in the center of it. The high velocity water creates a low pressure state in that tap and it is that which creates the suction on the tube going down into the carcinogens. You are confusing the venturi principles used with unpressurized air, where the pressure does drop below atmo in the neck of the venturi (how a carb works). The pressure into the device you showed starts around 50 PSI . Yes, it drops some in the venturi, but NOWHERE close to atmospheric.

I missed the word "local" but was clarifying the hose end sprayer from a hose nozzle for someone. That said, I respectfully disagree Sir. First off, suction doesn't do the actual work but that's the way most people conceptualize it. Like current, fluid flow goes from high pressure (potential) to lower. The canister has a vent to atmosphere. That atmospheric pressure is what motivates the chemicals into the very low, local pressure in the venturi; No more, no less.

Dynamic pressure increase by the square of the fluid velocity so the effect can be profound. Static pressure decreases by that same amount; in this case < atmospheric in the venturi. Can't work any other way.

 

[Jettison]

To the Op's original questions;

If I'm understanding your description correctly, it sounds like you've got a 90 on the inlet side of the mechanical fuel pump, then a 90 on the outlet with an extra port (it's in vans online store as part number KB-090-T) to take the fuel pressure reading, which is a pretty standard setup.

However, if you've got another 90 attached to that one to aim the output somewhere else, thats just a bad practice due to the potential for the engine rocking about the crankshaft to cause them to work against each other and start to loosen up.

If you've got to have that setup, making the best of a less than ideal situation would be to position the fittings so that the legs of the fittings are perpendicular to the crankshaft to minimize the potential for them to loosen up.

Why the second 90? Is it just to get your fuel line pointed back outboard? If so, there are other better solutions.

Here is a picture of the outlet side. Yes, after exiting the pump, the fuel makes a 90 and goes straight down, then another 90 to go horizontal again. Why not come straight out of the pump and have a t fitting for the fuel pressure measurement?

 

[Jettison]

Why are my pictures always posted 90 degrees off?

 

[avrojockey]

Here is a picture of the outlet side. Yes, after exiting the pump, the fuel makes a 90 and goes straight down, then another 90 to go horizontal again. Why not come straight out of the pump and have a t fitting for the fuel pressure measurement?

Most installations have a straight line coming off the 90 fitting so it doesn't make two immediate turns. However, if I remember correctly, the 90 fitting and its clocking is intentional to prevent loosening of the fitting due to rotational movement of engine

 

[Biggiesauls]

Comparing water to gasoline is crazy talk. Key issue here is the vapor point of the two liquids. Gasoline has a much lower vapor point and WILL develop bubbles (the start of vapor lock) at a low temperature point. Throw in a couple 90 degree METAL fittings to soak up more heat and pressure pulses will happen occasionally from the induces gas bubbles.

I did not stay at a Holiday Inn last night…. But battled through fuel/heat/stumble issues on my newly minted RV-9A.

 

[Desert Rat]

Here is a picture of the outlet side. Yes, after exiting the pump, the fuel makes a 90 and goes straight down, then another 90 to go horizontal again. Why not come straight out of the pump and have a t fitting for the fuel pressure measurement?

Well, That isn't actually two 90's it's one 90 and the hose has a 90* end. A subtle difference maybe, but still not two 90's working against each other.

I'd imaging it was done that way initially because that fitting coming off the fuel pump in readily available from Vans. If you want o clean that up with a straight fitting and/or a straight hose end, TS flightlines can hook you up. Tom might also suggest something else entirely.

I will say that I don't see anything particularly awful about the way it is now, since the second 90 you describes is part of the hose, not a separate fitting which was what I was envisioning. About the only issue it could cause is the previously mentioned loosening due to engine rotation, but it appears to be clocked the best way to minimize that possibility. Also I don't really see how it could be causing fuel pressure fluctuations.

Pictures are sideways because the imbedded orientation data you get taking them on a phone doesn't play well with VAF. Open them up in preview and rotate them through 360 degrees before you attach them to your post and they won't do that.

 

[bobnoffs]

so how does the word ''cavitate'' fit into this discussion?

 

[Freemasm]

Oh it would. Don’t forget “flashing”, “phase” plus some others. No need to go there. I’ll settle on “Happy Thanksgiving”.

 

[lr172]

so how does the word ''cavitate'' fit into this discussion?

cavitation is mostly a problem with impellers. I don't think you are going to get cavitation from a 90* elbow at the flow rates we see.