I have had several instances of occurrences which I believe probably relate to fuel vaporization (vapor lock). Most have been minor- ranging from fuel pressure variations at cruise after long climbs- to engine stumbles and corresponding fuel pressure drops when shutting off the second fuel pump after reaching altitude. One occurrence resulted in a complete engine shut down after takeoff after the fuel pressure went to 4 psi. All of these examples were associated with high ambient surface temperatures of at least 90 degrees F as well as the use of auto fuel.
On the last occasion, I experienced an engine stumble and fuel pressure going below 40 psi upon shutting down the second fuel pump after reaching altitude. The surface temperature was 75 degrees F and I assume that the auto fuel was probably still winter blend with a higher Reid Vapor Pressure.
My aircraft is a high wing with gravity fuel feed. Hoses of 7mm diameter deliver fuel to a selector valve that has a RIGHT, LEFT, and BOTH selection. The return fuel hose is of 7mm diameter and is routed from the engine back to the fuel selector which returns the fuel to the tank that is currently feeding the valve. The fuel selector is on the cockpit side of the firewall. The engine receives fuel from the fuel selector by a 5/16" diameter hose through the firewall to an AN 6 fitting. On the firewall side the fuel system is aluminum with AN 6 fittings. Once on the engine side of the firewall the fuel is routed through a course fuel filter/sump combination and then to the Rotax dual fuel pumps. Once leaving the fuel pumps the fuel line "T's". The top of the "T" is attached to the restricted orifice to bleed any vapor/air and is routed back to the fuel selector with the return fuel from the engine.
The pressurized fuel is routed to the standard Rotax fine fuel filter and ultimately the engine. Parallel with the Rotax fine fuel filter is the check valve. All of this is in compliance with the Rotax recommended design.
I have tried to control, or at least monitor engine compartment heat by wrapping the engine exhaust and installing heat shielding between the engine and the fuel system as well as fire sleeves over the fuel lines and silicone heat shielding around the course fuel filter/sump. I added a temperature sensor on the firewall to monitor under cowling air temperatures as well as a temperature sensor on the course fuel filter/sump that is plumbed to directly monitor fuel temperature.
Upon analyzing the temperature and fuel pressure data after the last occurrence it appeared that the recorded temperatures for both under the cowling and the fuel were not high enough to support fuel vaporization (at sea level pressure). This leads me to conclude that if the temperatures were not high enough then the fuel pressure must be lower. Using rough calculations the fuel pressure must have been around -8 to -9 PSI to allow the fuel to vaporize at the recorded temperature. Since the occurrence happened at altitude 8,500' the ambient pressure would have been approximately 4 PSI below sea level pressure. I am assuming that the remaining -4 or -5 PSI must be the result of the vacuum from the inlet side of the fuel pumps.
I have flow tested my system and recorded 20 gallons per hour of flow through the open sump. This is obviously sufficient to feed the engine. However the fuel pumps are capable of much higher flows and would therefore create a low pressure on the inlet side in order to pump more than the 20 gallons the system is capable of providing with gravity flow.
I have switched to using Avgas (100LL) at least until the summer blend of auto fuel is widely available.
I have tried to think of ways to transfer the suction from the "hot" side of the firewall to the "cool" side. That would require installing an additional fuel pump capable of supplying at least 30 gallons per hour at a pressure of approximately 4 PSI.
I would install the fuel pump inside the cabin and located between the fuel selector valve and the firewall. My concern is that I may be just pushing the problem further and causing some other problems by introducing more friction and turbulence into the fuel system.
I thought about using ram air to the fuel vents and thereby pressurizing the fuel tanks but at 100 mph the ram effect would only result in approximately .22 PSI (according to my web source).
Anyone with experience with adding additional fuel pumps would be appreciated.
Also, I am not an engineer and don't have any education or training in the fields related to these issues. All of the numbers and figures I mention are rough calculations from various web sources and are not representative of expert opinions.
On the last occasion, I experienced an engine stumble and fuel pressure going below 40 psi upon shutting down the second fuel pump after reaching altitude. The surface temperature was 75 degrees F and I assume that the auto fuel was probably still winter blend with a higher Reid Vapor Pressure.
My aircraft is a high wing with gravity fuel feed. Hoses of 7mm diameter deliver fuel to a selector valve that has a RIGHT, LEFT, and BOTH selection. The return fuel hose is of 7mm diameter and is routed from the engine back to the fuel selector which returns the fuel to the tank that is currently feeding the valve. The fuel selector is on the cockpit side of the firewall. The engine receives fuel from the fuel selector by a 5/16" diameter hose through the firewall to an AN 6 fitting. On the firewall side the fuel system is aluminum with AN 6 fittings. Once on the engine side of the firewall the fuel is routed through a course fuel filter/sump combination and then to the Rotax dual fuel pumps. Once leaving the fuel pumps the fuel line "T's". The top of the "T" is attached to the restricted orifice to bleed any vapor/air and is routed back to the fuel selector with the return fuel from the engine.
The pressurized fuel is routed to the standard Rotax fine fuel filter and ultimately the engine. Parallel with the Rotax fine fuel filter is the check valve. All of this is in compliance with the Rotax recommended design.
I have tried to control, or at least monitor engine compartment heat by wrapping the engine exhaust and installing heat shielding between the engine and the fuel system as well as fire sleeves over the fuel lines and silicone heat shielding around the course fuel filter/sump. I added a temperature sensor on the firewall to monitor under cowling air temperatures as well as a temperature sensor on the course fuel filter/sump that is plumbed to directly monitor fuel temperature.
Upon analyzing the temperature and fuel pressure data after the last occurrence it appeared that the recorded temperatures for both under the cowling and the fuel were not high enough to support fuel vaporization (at sea level pressure). This leads me to conclude that if the temperatures were not high enough then the fuel pressure must be lower. Using rough calculations the fuel pressure must have been around -8 to -9 PSI to allow the fuel to vaporize at the recorded temperature. Since the occurrence happened at altitude 8,500' the ambient pressure would have been approximately 4 PSI below sea level pressure. I am assuming that the remaining -4 or -5 PSI must be the result of the vacuum from the inlet side of the fuel pumps.
I have flow tested my system and recorded 20 gallons per hour of flow through the open sump. This is obviously sufficient to feed the engine. However the fuel pumps are capable of much higher flows and would therefore create a low pressure on the inlet side in order to pump more than the 20 gallons the system is capable of providing with gravity flow.
I have switched to using Avgas (100LL) at least until the summer blend of auto fuel is widely available.
I have tried to think of ways to transfer the suction from the "hot" side of the firewall to the "cool" side. That would require installing an additional fuel pump capable of supplying at least 30 gallons per hour at a pressure of approximately 4 PSI.
I would install the fuel pump inside the cabin and located between the fuel selector valve and the firewall. My concern is that I may be just pushing the problem further and causing some other problems by introducing more friction and turbulence into the fuel system.
I thought about using ram air to the fuel vents and thereby pressurizing the fuel tanks but at 100 mph the ram effect would only result in approximately .22 PSI (according to my web source).
Anyone with experience with adding additional fuel pumps would be appreciated.
Also, I am not an engineer and don't have any education or training in the fields related to these issues. All of the numbers and figures I mention are rough calculations from various web sources and are not representative of expert opinions.
