rjcthree
Well Known Member
This is long, there’s a lot to digest.
I’ve got to go back down the erratic fuel pressure rabbit hole. Something changed on my airplane, and I believe it needs to be explained. I understand and have read a lot of history on this topic, but I don’t see much root cause confirmation in those threads, and those that are there seem to point to early onset of vapor lock and do not close out with resolutions – other than fly it and see what happens, because the carb only needs 0.5 psi of pressure and modestly high pressure (<10 psi) doesn’t seem to matter – the engine stays happy. I get it, but I can see something changed, so ‘it’s just is like that’ is hard for me to accept.
Aircraft details: RV-9A, 250TT, O-320-H2AD, dual electronic ignition, 4SPA carb, fuel selector, boost pump and collator in the typical locations. The engine driven fuel pump is unique on the H2 being left top of the engine. Fuel pressure is monitored by a Kavlico three-wire sensor to a Dynon D180. It burns 100LL exclusively with a smidge of MMO. The collator has a shroud and blast tube. All fuel lines are fire sleeved and everything that can see radiant heat input of significance is also reflective sleeved over the top of fire sleeve.
30 hours ago the fuel distribution was changed from and Ellison -4 to the carb. At that time, the erratic fuel pressure was the first indication of trouble. I sought guidance from this forum then. During that investigation, the Ellison was found to have faults requiring repair, and the lack of support drove the change to a carb. At that time of those issues, the fuel pressure variation was confirmed - kinda. The mechanical parallel gauge temporarily installed was fluid-damped, but it did see sufficient variation during ground running to highlight something going on. In an attempt to clear the board, the Kavlico sensor was replaced, this did not change the pressure behavior. I have not been able to induce any variation by messing with wires, confirming grounds, checking power, etc. The engine driven fuel pump was changed at that time, for the same reason. For the next 30 hours flying with the freshly OH’d 4SPA, all was well, fuel pressure exhibited nominal, expected and stable readings, until 8/5.
In the chart there are two flights, to the left of the yellow vertical line is considered ‘normal’ and starts mid-flight, including a landing and takeoff in the middle. To the right of the yellow line is the suspect flight. The furball that is fuel pressure (scale is in psi on the left axis) on the 8/5 flight is quite obviously different than the 7/30 flight. RPM is provided as an indicator of fuel demand, scale in rpm is on the right axis. OAT of both flights ranged from 66F to 73F. Cruise speed was >120kt IAS. The data is right off the Dynon, three second intervals. The boost pump is on almost the entire time (I run that way). I did turn the boost pump off during the suspect flight on 8/5 to see what would happen; in the 15 or so seconds I watched, my impression was the variation got visibly greater with boost pump off, but I can’t pick that out of the data. I changed tanks about halfway into the flight, the behavior did not change. As part of first troubleshooting, I completed a ground runup on 8/6, engine ‘cold’, OAT 75F. Prior to the runup, the cowling was removed, and the fuel system was purged through the primer lines (~16oz run through), no air was observed in the purge. During the runup, the oil temp reached a high of 96F. The runup replicated fuel pressure variation shown with medium fuel demand, varying between 5.4 psi and 2.6 psi. I’m using this finding to generally rule out pure heat-induced vapor lock as the singular root cause, as the engine was not hot and the entire engine was exposed to turbulent cooling air.
I have not checked the vent screens, but I recently cap level filled the tanks (oops) and got puddles on the floor, so I believe them to be flowing tank to vent.
My plan is to install a new 0-10 psi undamped mechanical gauge in such a way to make it flight worthy. I’m not sure I will fly with it, but I want to know I can if required. I assume it will confirm the variation. Once that happens? I really don’t know. I’m considering replumbing, as I pick up fuel pressure at the mech pump outlet fitting, maybe moving that to teeing off the carb elbow. I don’t have any proof plumbing is part of the problem, but I know I have 90 deg fittings in the fuel path (qty 4 between the collator and the carb). I don’t think that makes my setup unique.
A theory I’m considering is that there may be a loose enough fitting somewhere pulling in some air, probably downstream of the fuel selector but before the boost pump itself. This should be the only place that fuel is not under pressure (I run boost pump 99% of the time). I don’t recall hearing the boost pump barking like it does when it gets air though, but that is not conclusive. Air introduced there would cause some chaos in the mech fuel pump as it moves downstream. I just can’t wrap my head around how that changed over time. I’ll do the clean towel fuel stain fitting search anyway.
A complication that may be aligned to the theory above: I have an Ellison-based habit where I shut fuel off before shutting the engine off by a few seconds (5-10 seconds) at idle. I usually see the fuel pressure drop to below 2 psi when I do it. With the Ellison, it resulted in less post-running fuel drain. Maybe this is pulling air in on the inlet side of the pump? If so, it could be pulled in anywhere between the selector and engine driven pump. I know I did this process at the end of the 7/30 flight, but I do it on most flights so it’s not new per se.
I also see some blue staining on the seam of the carb bowl to body, no evidence of dripping, but it seems that’s a common thing on OH’d carbs that needs to be snugged. I don’t think this is related, but I will take care of it.
I’d appreciate input and thoughts.
I’ve got to go back down the erratic fuel pressure rabbit hole. Something changed on my airplane, and I believe it needs to be explained. I understand and have read a lot of history on this topic, but I don’t see much root cause confirmation in those threads, and those that are there seem to point to early onset of vapor lock and do not close out with resolutions – other than fly it and see what happens, because the carb only needs 0.5 psi of pressure and modestly high pressure (<10 psi) doesn’t seem to matter – the engine stays happy. I get it, but I can see something changed, so ‘it’s just is like that’ is hard for me to accept.
Aircraft details: RV-9A, 250TT, O-320-H2AD, dual electronic ignition, 4SPA carb, fuel selector, boost pump and collator in the typical locations. The engine driven fuel pump is unique on the H2 being left top of the engine. Fuel pressure is monitored by a Kavlico three-wire sensor to a Dynon D180. It burns 100LL exclusively with a smidge of MMO. The collator has a shroud and blast tube. All fuel lines are fire sleeved and everything that can see radiant heat input of significance is also reflective sleeved over the top of fire sleeve.
30 hours ago the fuel distribution was changed from and Ellison -4 to the carb. At that time, the erratic fuel pressure was the first indication of trouble. I sought guidance from this forum then. During that investigation, the Ellison was found to have faults requiring repair, and the lack of support drove the change to a carb. At that time of those issues, the fuel pressure variation was confirmed - kinda. The mechanical parallel gauge temporarily installed was fluid-damped, but it did see sufficient variation during ground running to highlight something going on. In an attempt to clear the board, the Kavlico sensor was replaced, this did not change the pressure behavior. I have not been able to induce any variation by messing with wires, confirming grounds, checking power, etc. The engine driven fuel pump was changed at that time, for the same reason. For the next 30 hours flying with the freshly OH’d 4SPA, all was well, fuel pressure exhibited nominal, expected and stable readings, until 8/5.
In the chart there are two flights, to the left of the yellow vertical line is considered ‘normal’ and starts mid-flight, including a landing and takeoff in the middle. To the right of the yellow line is the suspect flight. The furball that is fuel pressure (scale is in psi on the left axis) on the 8/5 flight is quite obviously different than the 7/30 flight. RPM is provided as an indicator of fuel demand, scale in rpm is on the right axis. OAT of both flights ranged from 66F to 73F. Cruise speed was >120kt IAS. The data is right off the Dynon, three second intervals. The boost pump is on almost the entire time (I run that way). I did turn the boost pump off during the suspect flight on 8/5 to see what would happen; in the 15 or so seconds I watched, my impression was the variation got visibly greater with boost pump off, but I can’t pick that out of the data. I changed tanks about halfway into the flight, the behavior did not change. As part of first troubleshooting, I completed a ground runup on 8/6, engine ‘cold’, OAT 75F. Prior to the runup, the cowling was removed, and the fuel system was purged through the primer lines (~16oz run through), no air was observed in the purge. During the runup, the oil temp reached a high of 96F. The runup replicated fuel pressure variation shown with medium fuel demand, varying between 5.4 psi and 2.6 psi. I’m using this finding to generally rule out pure heat-induced vapor lock as the singular root cause, as the engine was not hot and the entire engine was exposed to turbulent cooling air.
I have not checked the vent screens, but I recently cap level filled the tanks (oops) and got puddles on the floor, so I believe them to be flowing tank to vent.
My plan is to install a new 0-10 psi undamped mechanical gauge in such a way to make it flight worthy. I’m not sure I will fly with it, but I want to know I can if required. I assume it will confirm the variation. Once that happens? I really don’t know. I’m considering replumbing, as I pick up fuel pressure at the mech pump outlet fitting, maybe moving that to teeing off the carb elbow. I don’t have any proof plumbing is part of the problem, but I know I have 90 deg fittings in the fuel path (qty 4 between the collator and the carb). I don’t think that makes my setup unique.
A theory I’m considering is that there may be a loose enough fitting somewhere pulling in some air, probably downstream of the fuel selector but before the boost pump itself. This should be the only place that fuel is not under pressure (I run boost pump 99% of the time). I don’t recall hearing the boost pump barking like it does when it gets air though, but that is not conclusive. Air introduced there would cause some chaos in the mech fuel pump as it moves downstream. I just can’t wrap my head around how that changed over time. I’ll do the clean towel fuel stain fitting search anyway.
A complication that may be aligned to the theory above: I have an Ellison-based habit where I shut fuel off before shutting the engine off by a few seconds (5-10 seconds) at idle. I usually see the fuel pressure drop to below 2 psi when I do it. With the Ellison, it resulted in less post-running fuel drain. Maybe this is pulling air in on the inlet side of the pump? If so, it could be pulled in anywhere between the selector and engine driven pump. I know I did this process at the end of the 7/30 flight, but I do it on most flights so it’s not new per se.
I also see some blue staining on the seam of the carb bowl to body, no evidence of dripping, but it seems that’s a common thing on OH’d carbs that needs to be snugged. I don’t think this is related, but I will take care of it.
I’d appreciate input and thoughts.
