|
Comments added 6/23/03.... I have since reconsidered this plan, however I will leave it here for informational purposes. I believe I may end up sticking with the original design, since it has flown very successfully in most planes, and is very similar to other certified aircraft installations. I will maybe adding wing sump low points on the outboard side on the wing, as Jim Bede himself recommended in later newsletters, if I can figure out a way to retro-fit with only a single pickup in each wing.
*********************************************************
BD-4'er's, (written 2/28/03, updated 3/5/03)
Problem Review
Over the past few months, I have been researching all the published information I could find related to fuel delivery on the BD4. I have reviewed FAA accident reports, many historic newsletters, information on the BD-4.org website, and many discussion threads with the e-mail chat group and its archives. Overall, I'm not convinced the stock system is the best solution in terms of getting fuel from the wing tanks to the engine consistently and reliably. Many BD4 pilots with significant expirence feel the originally designed system will work fine, IF and ONLY IF the pilot does not exceed any of the system's inherent limitations. Through analysis only, I agree with them. Unfortunately, it seems the fuel system has limitations at lower margins than other performance aspects of the the plane's design, making it possible for the pilot to fly the plane in such a way as to cause fuel starvation. I believe a system should be designed to operate as independant of pilot error as possible, so that even if the pilot makes a mistake such as departing with a high angle of attack or flying ball un-centered, the system will still deliver fuel without a problem.
While many have had no problem flying the stock system, the accident reports show that one of the most common themes seems to involve fuel delivery issues. The specific reasons vary, some possibly due to pilot error, others possibly to contaminated fuel or sloppy construction, and some to to high angles of attack while departing on less than 1/2 tanks, etc. It seems if one follows good piloting techniques, properly clears the lines of air after refueling, maintains reasonable AOA or full tanks so as not to unport both pickups - then the stock system works satisfactorily. However mistakes are made, we are all human.
Header tanks provide a "buffer" for some shortcommings, weather human or design related. However, header tanks are really just a band-aid type fix that results in a more complicated system still subject to a different set of problems, and dangerously bring fuel into the cockpit. I believe a solution that is simple to operate, and more forgiving of pilot error, and functions properly through the plane's normal operating envelop is still needed.
I have tried to address these primary challenges:
 Fuel flow stopping (breaking siphon) at high angles of attack
 Airlock when filling the tanks
 Un-porting both pickups in the wing selected to deliver fuel such as in a slip or uncoordinated turns
 Vapor Lock in gasolator due to engine driven mechanical fuel pump suction
 Engine driven mechanical fuel pump failure
 Water or debris in the system
Proposed Solution Description
 Route fuel lines for original dual pickups on each side down rear door post and tee the lines together at floor along rear door post. This resolves fuel flow stopping beyond these tee's in high "accellerated" angles of attack such as can occur when tees are positioned behind panel as in original design. Refer to Side View Drawing.
 Place a Left / Right / Both / Off fuel valve just forward of the tee at the rear door post along the left side of the pilot's seat. Keep it as far back as possible but easy for pilot to reach. Bring a line from the tee on the right (passenger's side) over along the floor (maybe under the seats) to the fuel valve. Check valves and clear fuel filters can be optionally added to each line just before it enters the fuel valve. Check valves would prevent cross-flow from one wing tank to the other with fuel valve in a 'Both' position. Fuel filters prior to the valve would allow fuel to continue to flow from one wing tank even if the other's filter or line became plugged with debris. Additionally if the fuel filters are clear, it provides a quick visual point to confirm fuel flow and absence of airlock. The both position does not require pilot interaction, and will continue to feed from one tank even when both pickups in the other are unported such as in slips or uncoordinated turns. Refer to Top View Drawings.
 Place a drain, or gasolator just forward of the fuel valve as a rear system low point to remove water that might enter the lines. Keep as close to fuel valve as possible. Additionally, will break airlock in lines if existant after filling tanks (similar to Piper's). Refer to Top View Drawings. I also plan to use Stephan Pelgar's flush drains in the wing tanks or convert the doors to fold up like the new TVAP BD-4's which will allow me to add sumps and drains to the bottom of the wings.
 Place a Facet fuel pump just forward of the gasolator / drain but as close as possible. Fuel will be delivered by gravity at all times to the inlet of the pump, limiting vapor lock before it. This pump should always be running if the engine is running (connect power to ignition switch). It asks as a backup to the engine driven pump in case of a failure, and vice-versa. Also since it now is pressurizing the line before it, it will limit the possibility of vapor lock in the gasolator on the firewall. Since it is self priming, it will help clear any airlock that might have been overlooked such as if checking drain was forgotten during preflight as well as help reestablish fuel flow rapidly in situations where both pickups become unported such as might happen with very little fuel remaining in the tank. Refer to Top View Drawings.
 Route the fuel line forward, along the floor under the door frame on the left side of the pilot's seat. Since the facet pump is pushing fuel forward, the line may be optionally run up to the lower panel through a simple section of clear line built into the panel to act as a visual fuel flow gage. Refer to Top View Drawings.
 Cross vent between tanks (adjust any air vents as neccessary to obtain similar fuel consumption rates from both tanks in flight in the 'Both' position'.
 Optional fuel level sights may be added just inside the fuselage in the wing root area using tee's rather than elbows for where the fuel and vent lines run through the fuselage wall. Connect the extra upward facing tee leg of the fuel line to the extra downward facing leg of the cross-vent tee using a clear fuel line. See pictures on BD-4.org of Carlos Serodio's installation. Note Carlos has two sight gages since he has a 2nd independant auxillary tank outboard of the main tank.
Of course I have not yet flight tested these ideas, Consider it at your own risk! That said I currently I plan to install my system in this manner if noone can give me reasons why it is inferior or will not work! Of course I'm very open to improvements! I'm interested in critism from those with more experience! E-mail me your thoughts!
Drawings and Detail Discussion Rationale
Side View Drawing
One must remember of course that height is all relative and is really just a function of acceleration forces. The acceleration force we think of in terms of one liquid mass position being "higher" than another position normally is gravity. However in the case of an airplane, as others have mentioned in newsletters, and elsewhere, there are also acceleration forces acting on the airframe in various vectors whether taking off, diving, or slipping must be taken into account.
Where the fuel lines route is really in-significant once the lines are full of fuel, but significant when first full of air and then fuel is added above resulting in airlock. If one draws a side view of the BD-4, and then draws straight lines to various points on the airframe where tee's have been mounted or could be mounted, we can see and determine that if both lines ran down the rear door frame and were tee'd together as low as possible in the corner where the floor meets the rear door frame, this position would allow the plane to be at almost a 90 degree climb angle or a 60 degree dive before fuel would quit feeding beyond the tee. Acceleration from a powerful forward takeoff roll would only helps push fuel there rather than when the tee is mounted at the bottom of the forward door post.
Top View Drawings
 
So from 4 pickups and lines we are now down to 2 fee lines on each side of the cabin. Then where to bring these two lines from each wing together? This of course is done using the selector valve. If the valve is mounted on the panel, we have just defeated what we previously gained. Any nose high attitude that would bring the valve higher than the pickups in a wing tank relative to acceleration vector forces that was empty (or in a strong slip) will cause the fuel to quit flowing in a both position, or even if one switches from one tank to the other and there's an air pocket in one of the lines. The solution is to bring the lines from the tees at the bottom corner of the door frame directly to meet in the center of the cabin on the floor at about the same station as the rear door post. Or to the pilot's side of the cabin alternatively (see drawings) due to the push-pull tube and rudder cables in the center. Then a facet pump is placed on the outlet (forward side) going to the gasolator and engine to push fuel to the engine driven pump. A drain or gasolator should also be added at this low point between the fuel selector valve and facet pump of course. This should allow for fuel to feed uninterrupted to the engine even in uncoordinated steep bank angles, etc. provided a BOTH position is used and there is fuel in both tanks. To prevent cross flow, one could optionally add check valves to both lines before they ran into the fuel selector valve.
In this way, a both position can be used, and in any angle of attack and accelleration vector a BD-4 flys through normally, the fuel will always reach the facet pump even if only 1 pickup of the four is covered with fuel. During tank filling, the Gasolator between the fuel selector valve and fuel pump will bleed any air locked in the lines out similar to the Piper systems (see writeup on Cessna and Piper systems on BD-4.org) Further, the self priming facet pump should help 'prime' the system. In a test I did with a facet pump, I plugged it into a totally clear line, dropped anothe line down a few feet into a small 1 gallon tank. I also have a glass fuel filter just before the carb on the engine. It took the facet pump 4-5 seconds to draw fuel up the line, push the air out the carb and charge the line completely with fuel.
To further eliminate the airlock issue when filling the tanks, an OPTIONALY vent line (smaller than the main fuel line and clear for visual verification - 1/4" maybe) could be tapped into the top of the fuel valve or gasolator (ideal if possible) and vented up to both wing tanks. Of course with a both position the tanks need to be cross-vented. Both my tanks are currently vented individually, so by adding the cross vent, there is a possibility that the tanks could feed at different rates in the both position. If this occurred, either the vent on one wing tank can be adjusted in some way or completely blocked and both tanks vented through the cross vent and a single vent from one tank (see Tony Bingelis, Firewall Foreward).
I would go one step further for safety with a visual fuel sight gage created by using tees for the line connections through the fuselage wall in the wing root area. One line goes out to the tank, another clear line goes from the front pickup up to the vent line tee. The third connection for the vent tee would be connected to the other side for the cross-vent, and the third branch of the fuel line tee would be coonected to the line to deliver the fuel Carlos S. did this and has pictures to show it on BD-4.org although it would not be required or effect function, just a nice to have to quicly an reliably determine fuel available.
Since fuel is delivered essentially by gravity to the facet pump, and from there its pushed forward through the gasolator to the engine driven pump, it should be run all the time, and in this way should virtually elminate vapor lock. Additionally, as it would not effect the fuel delivery anymore, from the single line running forward along the side (or center) of the cabin, I would run it up along the panel for about 4 to 6" through a clear piece of line mounted through the panel to act as a simple, reliable fuel flow gage.
Front View Drawing
The resulting maximum angles through which fuel will continue to be delivered since a tee will not 'break flow' will be:
 Max Angle of Attack = approximately 80 degrees
 Max Dive Angle = approximately 65 degrees
 Resulting System Limits Drawing
Has anyone run the fuel lines in this manner before? Why not? Now everyone can analize and point out if and why I'm wrong or if have overlooked some important fact! Please send your comments and critism to me at mark@felling.us, all will be welcome!
References
 Roger D. Mellema & Jim Huber - Newsletter #17 pg. 26
 Monroe McDonald - R.M. Newsletter #18 pg. 4
 John Brecher - R.M. Newsletter #? pg. ?
 Tony Bingelis - Firewall Foreward, 1988, pgs. 158 to 172
 www.BD-4.org archieves
 Participant comments on bd4@northwestareo.com e-mail discussion group
|