Pump mystery

When you say a "stinger", I am thinking you are meaning that basically no tank pressure is pushing the fluid through the pipe at all and the pump is doing the "sucking" to get the flow going to begin with? Are both tanks at the same height on the ground? Are both tanks under or above the pump itself?

Are you certain that both tanks are vented properly ... no obstruction or restrictions? Is there a separate venting port, or do you rely on the annular space around the stinger pipe?

Dave

Yes, the pump draws the fluid up through the stinger pipe, through an ell fitting and hose barb, into the hose, though another hose barb into the T, into 25mm pipe, thence into the pump. Paul

Yes. Both tanks are vented to the atmosphere very well, and both tanks are mounted at the same elevation on the same level platform. Paul

It looks like you have a funny thing happening then alright, Let me try to make up some "physics" excuse for it.. :)

The longer length pipe will have more "static" energy. (higher mass total = higher energy potential) so as the pump starts it will want to flow more than the shorter length would want to.

Sounds good here.. Anyone else think it sounds silly enough to be true? :)

I forgot to address your other questions: Yes, both tanks at same height. These tanks are about 1m high. The pump inlet is at about 80cm, so, sometimes the pump is below the liquid level(s) and sometimes above it, i.e., when both tanks are 80% full, their liquid levels are at the same height as the pump inlet. However, what always happens is that Tank B gets nearly emptied before Tank A even begins to subside. Paul

Paul Mathews wrote: [snip]

One thing to think about is that the "up, over, and down" flowpath describes a siphon. If the siphon is initially full of air, then the pump must develop enough suction to pull the fuel up over the top and start the siphon. Once the siphon is "started", and fuel begins to flow down the other side, the amount of suction required to keep the fuel flowing is reduced.

In your case, you have two siphons in parallel, connected to the same source of suction. Suppose that one siphon starts before the other. Once flow is established in one siphon, the other siphon may not start.

This explanation has a couple of flaws:

1. It says that all the fuel flow would be from one tank, and none from the other. This is not what you observe, apparently.

1. It doesn't explain why it happens the same way each time.

Something to think about, anyhow. If you could find a transparent hose that is compatible with the fuel, it might be interesting to take a look and see if there are any air pockets or bubbles in the suction lines.

Olin Perry Norton

Just curious. How is the T fitting configured?

-jim

Well, once the flow is higher from one tank, the other tank can't really force it's flow to enter the faster flow.

Anyways, it sounds like a great physics problem for schools :)

Yes this may be the answer. In general if the hoses are in good condition and you don't run the tanks dry then the siphon effect won't play any role. Both hoses will stay full of fuel at all times. There will never be any air in the lines. But what if the line to tank A has a tiny leak at the highest point? No fuel will leak out but when it sits between fueling it will slowly develop an air pocket as the fuel runs back into the tank.

-jim

How do the two lines feed into the pump, is it with a "Y" or a "T" connection?

I would think without a flow regulator on the two lines that, as someone else explained, you could have greater suction on one side than the other.

Try reversing your connections where the two meet.

"Paul Mathews" wrote in news: snipped-for-privacy@p10g2000cwp.googlegroups.com:

Another probable explination may be the flow within the T itself, as there will be a turbulence issue with the convergence of the flows from the tanks inside the T. One flow path may become dominant within the T, due to slight differences in the intersections of the 3 flow paths. What does the inside of the T look like? If there is a sharp ridge on once side, and a smoother edge on the other, the difference in resistance of flow from one inlet to the other may explain it. This application would probably be better with a Y instead of a T.

My bet, you have an air bubble on line A in the U. a pressure will be required to shift the bubble far enough for flow to commence. If I am right, if both tanks start level, tank B will empty with no flow from tank A until a level difference is reached, at which point both tanks will start to empty maintaining the level difference.

Detail port configuration of the tee could do it. It doesn't take much.

Brian Whatcott Altus OK

Years ago, there was a short-lived trend to fluidic logic. Something like a tee could serve as an 'either/or" arrangement.

Brian Whatcott Altus OK

The hoses are transparent, and the system is free of air, both tanks being refilled well before the stinger openings are exposed. Paul

I wondered about this myself, and I may yet change the T to a Y, but I reasoned (to myself) that the T might make a lot more difference if it weren't for the fact that the pump is sucking the fuel up from the T. It seems to me that this should guarantee that the flow takes a kind of Y shape, regardless of the fact that the fitting itself is a T. However, the T is iron pipe, and it's a classic T shape. Paul

Rats! I thought we were on to something.

I guess the thing to do now is to focus on the T. Either switch the connections or replace it with a Y.

I don't have a new solution for you. My only guess is that there is some non-obvious restriction in one of the lines or stingers...

But here's a question for you. If I understand this correctly, there are two lines, one from each tank, which are joined together in a T before the pump. Are there any check valves in the system? If not, I'd expect the two tank levels to equalize every time the pump is shut off. (I presume the pump doesn't run continually.) Essentially, when the pump is shut off, if there are no check valves, there should be a siphon between the two tanks (you said there's no air in the hoses), which should run by itself until the levels in the two tanks are the same. (By same, I mean relative to some independent reference plane.) Does this happen?

-Paul