Fuel Tanks & Mods

[flatbed]

Lenny, can you use just one ball in the middle of a long tee, hose out each side, but have the hose run to the oppisite side of the tank, Just like a tee, the ball is able to move freely to both sides and can not create a vacum,

[Kinarfi]

If I remember correctly, your proving some of my suspicions that caused me to follow Occam's razor and go simple, I kept seeing the possibility for valves not reopening and chambers not being able to refill or vapor lock.

I did do some testing today. My tank is out of the Trooper, so I put a quart bottle on the floor under the vent canister and put the fuel line hose and the return line into it and filled it with gas, started the engine and let everything stabilize, topped off the bottle, then move the return line to another quart bottle, 18 second later, the 2nd bottle was full. That's 1 gallon every 72 seconds flowing through the return line . I wonder how much of the gas flow is used to cool the fuel pump? and by making it a closed loop and not running back into the tank, will it get hot? We'll see.

Decided to move the fuel pump and mount it low to the panel behind the tank and also the fuel filter will be removed from the tank and mounted to the panel also.

Jeff

[flatbed]

I moved my filter and pump next to the tank quite awhile ago, have not had a problem since i did the gas tank thing. but------------ after i put my fuel tank back in and filled the tank full, with the pump next to the bottom of the tank, I could not get the pump to prime its self, I thought this was crazy and still do. That pump would not prime its self untill i lossend the hose at the inlet side and let the air out. I think if the pump is laying down it will not prime its self.

[Kinarfi]

I moved my filter and pump next to the tank quite awhile ago, have not had a problem since i did the gas tank thing. but------------ after i put my fuel tank back in and filled the tank full, with the pump next to the bottom of the tank, I could not get the pump to prime its self, I thought this was crazy and still do. That pump would not prime its self untill i lossend the hose at the inlet side and let the air out. I think if the pump is laying down it will not prime its self.

I'm actually following your lead you on this issue. Can you post a photo of your set up please? & thanks.

Oh yea, one more question, should you filter the before or after the pump, I would think that you want to protect the injectors from the wear off the pump (carbon off the brushes, copper off the commutator, etc) as well any dirt or debris from the tank, or do we need to protect the pump from the dirt and debris, or should we put one of those small clear plastic screen filter in the loop also? Questions, questions, question, an obsessed person could go nuts !!!!

Jeff

Good nite!

morning fuel flow - tank - plastic filter - fuel pump - fuel filter - pressure regulator - injectors ----- Just my idea

[flatbed]

filter goes before the pump. i will take some pictures tonight. the plastic filter will break so don't use one of those. to much fuel preasure.

[Guest Lenny]

Lenny, can you use just one ball in the middle of a long tee, hose out each side, but have the hose run to the oppisite side of the tank, Just like a tee, the ball is able to move freely to both sides and can not create a vacum,

flatbet, I thought of that too. The problem is that were putting the ball on the other side of the flow through the close off point, The suction will pull the ball open. I guess I'm back to the pipe idea.

kinarif, I don't think the parts of the motor portion of the pump would be exposed to the gas in any way. If it was, an air pocket going through combines with a spark could blow it. I'm sure the pumping chamber is totally seperate from the motor portion. As aresult, I don't think the fuel pump can add any contaminates.

The thought of the return line is interesting. Maybe run it into a canister that is low then run the stock tanks outlet to the canister also. When it needs gas from the tank it will have it. The return gas would end up in the tank but it first is filling the canister and pushing back through the line comming from the tank. Simple, no extra fittings on the tank and a canister of reserve gas always available.

Come on guys and girls, if we can put a robot on Mars, by putting all our heads together, we should be able to at least get gas to always come from the low side of the tank.

This is really fun. As we get older, heaven forbid, one wonders at what age do we have to find a replacement for sex. Could this be it. Thank goodness , so far iI don't think any of us are old enough for it to be an issue.

Lenny

[Kinarfi]

kinarif, I don't think the parts of the motor portion of the pump would be exposed to the gas in any way. If it was, an air pocket going through combines with a spark could blow it. I'm sure the pumping chamber is totally separate from the motor portion. As a result, I don't think the fuel pump can add any contaminates.

Lenny

One of my idiosyncrasies is to tear things apart, have you ever taken a fuel pump apart? They're interesting, the fuel enters and is taken up by a multivaned impeller and forced into this and channeled to that and fed into the motor stator and winding and past the commutator and brushes and out the output. The gasoline acts as coolant and lubricant for the pump motor. The air/fuel mix was correct and the brushes shouldn't be sparking (electrical background) anyway on this type of commutator and the burn would would be very small and (IMO) wouldn't cause any problem.

Jeff

[Guest Lenny]

One of my idiosyncrasies is to tear things apart, have you ever taken a fuel pump apart? They're interesting, the fuel enters and is taken up by a multivaned impeller and forced into this and channeled to that and fed into the motor stator and winding and past the commutator and brushes and out the output. The gasoline acts as coolant and lubricant for the pump motor. The air/fuel mix was correct and the brushes shouldn't be sparking (electrical background) anyway on this type of commutator and the burn would would be very small and (IMO) wouldn't cause any problem.

Jeff

I'm amazed, It goes against my common sense thinking. Thanks for filling me in.

Lenny

[Kinarfi]

I moved my filter and pump next to the tank quite awhile ago, have not had a problem since i did the gas tank thing. but------------ after i put my fuel tank back in and filled the tank full, with the pump next to the bottom of the tank, I could not get the pump to prime its self, I thought this was crazy and still do. That pump would not prime its self untill i lossend the hose at the inlet side and let the air out. I think if the pump is laying down it will not prime its self.

I got to thinking, oh oh, about your air lock problem and my fuel flow test. I said I put the fuel line hose and the return line into a quart bottle and filled it with gas, started the engine and let everything stabilize, what I saw was all the air getting flushed out of the system, as it left the return hose, it would get sucked up by the fuel line and recirculated until it all escaped, this was happening because of the close proximity and limited space of the bottle. To correct this problem, I need to put an air/fuel separation chamber in my system, this could be as simple as a barbed Tee or as complex as chamber with barb fittings.

IMO,Your's, Flatbed, should be OK, IMO.

filter goes before the pump. i will take some pictures tonight. the plastic filter will break so don't use one of those. to much fuel preasure.

I still like the Idea of a good filter after the pump. see drawing.

Jeff

[flatbed]

It takes very little to plug the screen in the fuel pump. I have not looked but we probably have a filter after the pump also. on my return line it goes into the bottom of the tank so the air floats away from the pick-up. The plastic filters are made to hold around 5 psi. Our fuel systems run at around 35-40 psi. I do believe when the pump is laying down it traps air and will not prime. My pump is maybe 18" from the tank and below it and it would not prime untill i let the air out of the inlet.

[flatbed]

lenny, with the tee long enought with a good size ball i do not think it would pull the ball away.

[Kinarfi]

If this can be made, I think it would work, no concern with the return, not sensitive to how high or how low it's mounted. Use a smooth bore tube with snug fitting steel ball that when it rolls to the lower end, it's weight seals up against the green seal, o-ring, and prevents air from being sucked into chamber. Two outlets so can't be stopped. As the hill gets steeper, the force of the ball on the seal get greater, sealing better.

Still working on this post, I'll add a drawing later.

Jeff

Picture this, 1/2" ID brass pipe internally threaded with 1/4 nptX5/16 barb on each end, each of then drilled just enough for an O-ring to into, two 1/4" holes drilled into the pipe about 3/8 c-c apart and angle away from each other, grind off a fitting to match the pipe and braze or solder it over the holes and put another barb fitting on it.

[flatbed]

I do not see where you need 4 ports, It should work with only 3. Center is the suction line, front port goes to the back of the tank, and the back goes to the front. Make the center port so there is no way the ball seals on it. Make the tube around 8" long.

[Guest Lenny]

Well here is what I did. I pretty much followed the ideas being thrown out and decided to make one up. The pics show a peice of 3/4 electrical conduit. Used this because the ID is just a little over 3/4" and thus I can use a 3/4" steel bearing. The rest is prettu much self explanitory. I used a slot instead of 2 outlets so the ball cannot get sucked up to close off the outlet. The ball I used is about .680" and thus doesn't hit the o-ring centered, It needs the 3/4" ball.

I went with a fairly large ball so it would have some weight and I used o-rings I had.

Here is what I found out. When I turn the assembly vertical so the ball finds and centers itself on the o-ring and I plug the other intake and suck on the outlet, I was susprised at how very little pressure it took to dislodge the ball and suck air. I can blow on the outlet and the seal holds. It dislodges easily enough so that I'm not comfortable with it as is. There is the option which will help and that is to use a smaller o-ring. This will reduce the area that atmospheric pressure can push on it from the backside. going to a larger ball will also help. Obviously if I go to a bowling ball, I'm not going to suck it off.

Lets get some more input.

Lenny

[flatbed]

maybe you should try a baseball, i was going to say football but that might be hard to seal

[flatbed]

Lenny, what do you think of making slide with a rod instead of a ball, Have the rod have a needle seat on the end. with the steep hills it is going to go one way or the other

[Guest Lenny]

Lenny, what do you think of making slide with a rod instead of a ball, Have the rod have a needle seat on the end. with the steep hills it is going to go one way or the other

That is typically the way most hydraulic and air 2, 3 and 4 way vlves work. I've been think the same thing. On a typical valve there is a pilot signal which pushes the slide ( I beleive it's caled a shuttle) from side to side. In our case, we have to have gravity do it for us. I'm thinking about a fairly heavy shuttle that is set is polyethylene runners or something like that so it slides freely. Like you say, have a needle seat on each end. With a bar type shuttle, a needle seat is possible and far superior. I also think it would be better to use a rectangular cross sectioned seal so we seal against a square corner for more reliable sealing with less pressure. I'll think about different ways to make the shuttle move freely. If you have any ideas, post them. We want it to slide without having to be on a big slope. The troopers viabraition may be a big factor in this. If when its out of its seat, it has some wiggle to it, It should move nearly as easily as a ball. I'm convinced that the three port consept is correct with center port modifications. Also, I'm convinced that if we can get the shuttle to slide freely, were done. It's fairly easy nfor me to prototype something.

Lenny

[Kinarfi]

That's pretty much what I had in mind.

Add these to the ends and tighten up the clearance between the tube and ball so the gas works to slow the ball down as it hits the valve. These are about $30 / pair with a .150 hole holley 6-519-2 Needle & Seat Kits http://www.holley.com/data/Catalogs/Holley/91.pdf

Jeff

[Guest Lenny]

I like your needle and seats real well. Are these available in a larger hole size? The reason I'm concerned is that I put a clunker in my tank a while back just to try it. I didn't want to add new fittings to the tank yet so I used a peice of soft rubber hose that I threaded through the tank outlet. When I was done I still had about a 1/8" dia. hole going through. The trooper would miss every once and a while and seamed a little sluggish. I took the clunker and tube out and everything ran normal. It didn't seem to like the smaller line. You would think the 1/8" line should be sufficient -- 1 gallon every 72 seconds / 50 gallon per hour flowing through the return line. I'm also wondering why you wanted to slow the ball down. Are you concerned with the needle seat not taking the hit of the ball? I think I have an answer for a free moving shuttle. We could use 3 or 4 balls to get the weight without the drag. Your needle seat may work with one ball.

So what do we have thus far. Kinarif's needle seats look perfect except for maybe size. flatbed's three port concept also looks like the way to go. Using multiple balls gives us the weight needed with freedom of movement.

What aren't we sure of. Needle seat size and can the needle seat take a hit from the ball or balls.

Were close. When we settle on something, I'll make one up for testing. I do feel that it's very preferrable if we can have a design that most anyone can do for themselves. In other words, they don't need more than common household tools.

Lenny

[Kinarfi]

Lenny said, I took the clunker and tube out and everything ran normal. It didn't seem to like the smaller line. You would think the 1/8" line should be sufficient -- 1 gallon every 72 seconds / 50 gallon per hour flowing through the return line. IMO 1/8 is on the shy side.

A friend saw the holes in my old tank and told me that if you use isopropyl alcohol as cutting fluid, it cuts real clean, I was using a hole saw.

Keep the idea of have the return line with a fuel / air separator being tied directly back to fuel pump line in mind, that makes the smaller orfices more acceptable. that may have made Lenny's clunk work OK.

Hex or square tubing with multiple balls - no chance of the ball plugging the outlet, but now you the possibility of a ball stuck at each inlet again.

Interesting page: http://www.pneumadyne.com/cirValves.cfm?cirControlID=8

The blue line is the valve attached to a weighted arm that hangs straight down.

Just Rambling, still tinkering,

Jeff

[Guest Lenny]

Your hex/square tube would work fine but it is complex to build. The interesting site is a typical 2 way valve. It still requires a signal at port "A" or "B". Take a look at http://www.pneumadyne.com/cirValves.cfm?cirControlID=2

In this valve, which as it stands wouldn't work for us, The pilot signal, in our case is the steel ball(s) which would hit the shuttle. The difference here is that we are using the ball(s) to open the good fuel side instead of closing off the bad side. This is just the opposit of what we have been looking at. This particular valve does not have an outlet port but small holes around the outside of the fitting. There isn't any good way to connect to the output. You can almost picture one of these at each end of a tube with a ball(s) inbetween. So far I feel that the needle valves, 3 port assembly with ball(s) has the the best chance of working. I'm going to set that aside for now and think about how opening the good gas side might be able to work.

Check this out. pretty close to where we are at. http://www.pneuhydro.com/Shuttle_Valves.htm

Lenny

[Kinarfi]

Took some photos of my intended klunk and moved fuel pump. This gets you to the first of 13, some have comments typed on the photo.

additional: go to http://www.mcmaster.com/ in find product box type Push-Button Valves. put 2 of the normally open, button to button with a lever in between and the lever can be as heavy or long as needs be with stops to limit the button travel.

This could mounted anywhere, level not critical

[flatbed]

I think the holly needles will be way to small, remember this is on a suckion line for a fuel injected engine, you need to have alot of fuel flow, Lenny, i think it would be a good idea for the fuel to be able to flow around the body that holds the needle, that way it wont be importand for the needle to be at one end or the other when cruzzing.

[Kinarfi]

I think the holly needles will be way to small, remember this is on a suckion line for a fuel injected engine, you need to have alot of fuel flow, Lenny, i think it would be a good idea for the fuel to be able to flow around the body that holds the needle, that way it wont be importand for the needle to be at one end or the other when cruzzing.

The hole for the needle valve is .150 which IMO MAY work and WILL work if it does not have to handle the return fuel volume. I don't support the use of the needle valve because of complexity of the flow pattern.

How would you use it without making a special chamber for it to be screwed into. The way I see it, we have 3 individual fuel flows to be considered.

1 engine fuel

2 return fuel

3 pump fuel = 1+ 2

4 tank fuel = 1 or 1+ 2, depending on what is done with the return fuel.

I'm still going with Occam's razor is often paraphrased as "All other things being equal, the simplest solution is the best." a central, low as possible junction with an air / fuel separator in the return line.

Yesterday, after moving the filter and pump, I tested every thing using a bottle as the gas tank and since I'm using 5/16 clear super thane fuel line, I noticed the fuel drained from the line after shut down.

Another question, when you start your engine, do you wait for the system to do it's checks before you hit the starter or just turn it on and hit the starter?

[Guest Lenny]

Well, I'm ready to make a, hopefully, a correctly working version. Here is what i'm doing. I'm using a peice of 1" pvc pipe just long enough for the balls to only have 3/4" of travel from one end to the other. This will will keep it from hammering itself to death. there will be four 1" dia. steel balls. Using multiple balls has the benifit of a heavy elongated shuttle bar with the free movement of balls. I calculated the pressure the balls would hold back based on the slope of the tube. I'm hoping to see it begin to hold somewheres between 10 and 15 degrees. The lower the slope, the closer to empty we can get. Each end of the tube will have a polyethylene plug which screws in. The outside end of the plug will be tapped 1/4" NPT to receive a barbed outlet fitting. These plugs will be bored out on the inside end to receive a needle guide which will also screws in. The needle guide will also be polyethylene and have a center sleeve whch guides the needle then a relief 1/8" wide going around the needle sleeve with a depth 1/8" less then the length of the guide and an outter wall threaded to screw into the plug. The end of the needle guide will have eight 1/8" dia. equally spaced holes through the connecting space between the inner sleeve and the outer wall. The needle itself will be steel with a 1/4" dia. stem about 7/8" long and a 3/8" dia. head 3/16" long. The needle head will have a 1/8" x 1/8" camfer on the end which will be the seating portion. This needle will slip into the needle guide, then guide with needle will screw into the plug with the large end of the needle going in first. Thus the needle will be captured while the 1/4" stem end will protude 1/4" past the valve guide towards the balls. The needle will seat against a reamed 1/4" dia. hole in the plug. This is reamed so I get a very clean crisp corner on the inside of the hole facing the needle. Once the needle assembly is screwed in place in the plug, I will lightly tap the needle to create an aprox 1/32" seat on the crisp cornered hole. This should provide a pretty good seal. The needle itself will be able to withdraw 3/16" away from it's seat. This will provide a 1/4" dia. port that the needle seats against, a large space between seat and needle (when open) and eight 1/8" holes for gas to flow through to get into the pipe. As flatbet sujested, I designed it so the gas would go around the needle valve and its guide rather then flowing inbetween the needle and its guide providing good volumn flow. The PVC tube wil have 2 outlets rather then 1. This is so the gas doesn't have to squeeze around the balls even though there is sufficient clearance to do this. These outlets will be just past the inside end of the valve guide and will not be able to be blocked by the balls.

This assembly still needs to be mounted low. When looking at kinarfi's swinging weight, the concept is pretty good and would work ok except that it, as shown, cannot be mounted low enough. If the valves are above the tanks outlets, when one of the valves is turned off, the gas in its respective line would tend to drain back into the tank. If you were to drive up a steep hill and quickly crest over and head downhill. The air in this line could get sucked up by the pump. I'm not a car guy so maybe the pump can take a slug of air without causing problems. Howeverthe weight on kinarfi's drawing could be over the valves which would allow them to mount low. Also, everybody can build his fairly easily.

Sorry this is so long but I figured that by the time anyboody deciffers it I'll have it done. So, i'll be glade to have your comments good or bad. In the mean time, this is the way it will be made. I'll come back to the forum when I'm done. Speak to the Gods an throw down sacrafices so this thing works. Gee, wouldn't a simple sketch have been better than a thousand words. Go ahead, I deserve to have some darts thrown at me. When I started, I didn't think it would be this long.

Lenny

[Kinarfi]

It will take even longer to build because you MUST take photos of each step completed. My testing shows that the pump CAN handle a big slug of air, like I said, were moving 50 gallon of gas per hour, a quart every 18 seconds.

And here's my latest idea and I need your thinking too.

Imagine your heading home with low fuel and you drop over a hill going down at 29 degrees, that what my drawing ended up being, the red line indicates the gas level. Now, take a look at the drawing and tell me whats going to happen, Please. I will take your post, copy it, and move it into this post and delete yours.

Kinarfi says: first the fuel will drain from the rear down to the red line pushing gas through the junction back into the tank via the front outlet, then as gas is drawn of by the engine, the junction will stay full until the gas level at the front drops to where air enter the line, then the engine dies.

Lenny says:

Flatbed says:

........... says:

[flatbed]

first, the needle seat needs to atleast draw 50 gallons and hour, all the gas that is used, engine and return fuel has to come throught the needle seat, also are you reverseing the fuel lines, front to back and back to front, I have heard that the stock fuel pumps do not like air, the vanes go out easy, mine has had air in it and has done fine. Would a spring between the balls help in the pounding of the needle seat, the seat might be better to be made of a plastic instead of metal so it will take the abuse. with your stup the air is goin to get out because it will want to rise no matter what angle the car is. I would also be trying to make something but i am working on my hot rod model a. your doing good lenny.

[Guest Lenny]

first, the needle seat needs to atleast draw 50 gallons and hour, all the gas that is used, engine and return fuel has to come throught the needle seat, also are you reverseing the fuel lines, front to back and back to front, I have heard that the stock fuel pumps do not like air, the vanes go out easy, mine has had air in it and has done fine. Would a spring between the balls help in the pounding of the needle seat, the seat might be better to be made of a plastic instead of metal so it will take the abuse. with your stup the air is goin to get out because it will want to rise no matter what angle the car is. I would also be trying to make something but i am working on my hot rod model a. your doing good lenny.

My flow orfices are all larger then what is on the tank stock. I plan on reversing the fuel lines, front to valve rear and rear to valve front. The seat itself is Polyethylene. The needle is steel. The steel against the poly should easily give it enough cushon. The balls will be running through a gas filled chamber and thus be somewhat dampened. Keep in mind that the balls typically wont hit like if you shook the assembly but instead probably roll slowly from neutral to an end as you begin transition into tip. It won't be as if you went from level and dropped the front or rear off a dropoff. I don't anticipate air to be any problems to what I'm doing.

Good luck on your model A, Sounds like your sort of a glutten for punishment too.

Lenny

[flatbed]

Kinarfi, Im sorry, i am laughing, but how many times do i have to read the picture. I don't know if that would work as it might sifon air, but it mite. That is sort of what i have except mine is in a sump in the middle of the tank with the return line returning in the middle of the sump. I do like your method of comunicating

[Guest Lenny]

Gosh, I'm getting dizzy. When I responded before to the drawing, I couldn't see the yellow lines good enough to see what you were doing. Anyway scratch that answer. I better get this right because you probably tested it and know. It's a bit of a brain teaser. First off, you will get some flow from upper tank outlet to the lower. I'm going to take this a step at a time and think it outloud. That flow will go down to at least the tank gas level. Your going to have a little inertia from the this flowing gas which will bring its stopping level a little lower then the tank level. If this is actually where it stops, it should go back up to tank level. But, your going to have suction from the pump which will encougage it to continue on torwards the junction. When the pump is running, it will draw from the side with least resistance. Initally,the gas providing the least resistance will be from the line comming from the rear tank outlet. If that gas has enough help so that the pump pumps that line empty, once the pump caught air I would think it would continue to draw from the easier supply which is now the air. Air being at the junction would allow the lower side to siphon back.

I can tell you what it will do if your parked on a downhill but there are dynamics working here with the pump sucking midway in the circuit that I'm not sure what it would do.

You been drinking or something. What kind of question is this. Now it will drive me nuts until you tell us how your test of it did.

I have to admit, you are doing some really good thinking. Very interesting concept.

Lenny