Daisy-chaining coolant containers

[philtipping]

New to forum so not sure where to post this... couldn't find a 'fluidics' section
The question relates to linking multiple coolant containers together so each one syphons into the next... will the levels in each container remain constant?

The photo shows the 3 x 5-litre containers, although only container 3 is connected due to the problem outlined below.

I started with containers 1 & 2 linked with flexible tubing pushed into holes at the top so their open ends are below the liquid surface. Same for containers 2 & 3. The tube from the spindle motor goes to container 1, and container 3 goes to the pump. All tubes are a tight-ish push fit but they're not airtight.

The assumption was that the levels in each container would remain equal when the pump was running... and this was in fact the case for the last few months of use.

The problem is I've just revamped the workshop so had to dismantle and re-connect everything, and now the levels are not staying equal when the pump is running. Container 1 rises higher than 2, and 2 rises higher than 3. I had to resort to just using container 3 to avoid container 1 overflowing.

I've confirmed all tubes are still below the liquid surface, so the questions are:-
1. Does it matter if the joints are airtight or not?
2. Does the amount of tubing below the surface matter?
3. Was it pure fluke it worked before?

Similar Threads:

• daisy chain small pieces
• Containers for chip/coolant recovery
• Need Help!- chaining a slot

[MO Metal]

You need to get all the air out of the tubes between 1-2 and 2-3 for it to siphon.

[philtipping]

Thank for replying Mo. All tubes are full, no air bubbles anywhere. Have tried more experiments... one possible change from previous layout was the pump was on the other table leg and was pumping down instead of up, so I've now remounted it horizontally. It's pumping *into* 3, so reverse the arrows in the photo.
I started the pump and the levels went slightly higher from 1 to 3, but they remained static. Ran for 10 minutes and no overflows, so looks like I'm back to the previous working state even though I don't fully understand how/why it's working.
Also noticed containers 2 & 3 were bulging slightly, and 1 was concave, so wonder if the system is more air-tight than I thought. Maybe the different air pressures inside the containers is affecting the flow, and every time I fiddle with the tubing, this affects the seals.
Slackening the tops of the containers 2 & 3 causes them to fill up rapidly and overflow, but slackening 1 has no effect!
The tube diameters are slightly different; the spindle tube coming from 1 is the narrowest, the link tubes are larger, and the pump tube larger still, but don't know if this is a factor.
If I can't rely on this 100% I may have to abandon it and buy a single large container. Daisy-chaining seemed the simplest solution at the time as I've got loads of 5L containers