There are several questions that got my mind pondering.

I know in the old systems there is a hose that goes from under the radiator cap to the overflow tube. That usually just dumps on the ground when the car gets hot and pukes out coolant. Then when it cools down, the coolant level drops leaving an air gap at the top of the radiator.

Does the radiator crate a vacuum and suck air in when cooling down, or does the coolant just shrink (not sure of the correct term, but it expands when hot) and the level settles where it needs to be?

If you have an overflow tank that has a hose entering from the bottom, when the system cools back down, it will draw coolant back into the system and no air. What about one that the tube enters from the top and goes to the bottom of the overflow tank? Wouldn't there be a small air gap where the tube enters the overflow tank and the top of the coolant in the overflow? Then when the vacuum is created the small air pocket would get sucked in as well as fluid leaving the radiator a tad lower than full, but more full than without a tank, correct? There should be either a vented cap or another hose going to the ground to keep the tank from collapsing on itself is what I am thinking.

If you were to blow into the overflow hose when the engine is cold (simulating vacuum draw), would you get any leaks or hissing sounds from the radiator cap area? What about if you blew in from the hose going to the ground from the overflow tank?

short answer is : the "open" system where it pukes on the ground, it just establishes the amount contained in the system when hot. the excess is purged out to the ground. this type of system has only one seal on the cap.
the "closed" system has a cap with two seals. when the coolant is hot, the excess is released into the overflow can/bottle by pushing up on the lower seal on the cap that seals on the bottom of the cap neck. when it cools, the same seal lifts off the bottom of the fill neck, and vacuum draws the coolant back into the system without any air. after a couple of cycles, the radiator and engine is full of coolant without air in the system, and will remain as such.
the overflow can come into the reservoir from either the bottom or top, but the tube must end at almost the bottom [about 1/2" or so] of the tank to remain effective not drawing in any air during the cool down process.
the reservoir is where it's vented, either through a small hole in the cap [usually found on the fancy metal ones] or the sloppy fit of the cap to the bottle. [factory plastic reservoir]
this prevents the container from collapsing while still allowing the coolant to return to the radiator.
hope this short answer helps.

Just to add, everything expands as it heats up.
water expands about 4 percent going from say 70 degrees to 212 degrees F.
so for 3 gallons of coolant, assuming all water, it adds about .12 gallons of volume of water to the system.
That increases pressure in an enclosed space, tipping past the say 12 or 15 pounds the cap is supposed to hold and bleeds out into the overflow tank.

When it cools that volume contracts the same amount, causing a vacuum to form in an enclosed space. So it then sucks air or coolant back through that tube.
You can look up the density of any material by temp to do the math on how much something expands or contracts due to temp of that item.

to take it back to cars, they didn't do the math right on the phenolthaic pistons they used in the calipers back in the l80's and when they heated up they jammed in the calipers.
had one seize up in an 83 m-body and had to power it into the garage smoking the whole way.

MoparX's explanation couldn't be better, seems to me.
Note that there's a minimum level for coolant in the overflow tank - even when rad is full - so the system won't suck air.
(Look at any OEM tank, you'll see "minimum" & "maximum" fluid levels).

Opposite of expand is condense, though in some cases that's not the most articulate, where shrink may better apply...

One other thing the overflow tank does is replace air in the system with coolant. Short of pulling a complete vacuum on the system and then filling it with coolant it is hard to get all the air out when doing an initial fill.

In use, the air in the system will make it's way to the radiator cap area and vent out the cap as the system heats. When the system cools and makes a vacuum the overflow tank supplies coolant to top off the radiator instead of allow the system to suck in air.

I believe the word you were looking for was contract as a verb.



Kevin

^^^ That (emphasis on 2nd syllable) is the more articulate & accurate word, depending on the application, thanks.
Must nore coffee, or fewer distractions

Here's the parts and photos of what I have installed. It's on a 1970 Monte Carlo and it is a popular choice according to Monte Facebook groups. It is a 1975-1979 Nova overflow tank and it mounts nicely in the passenger side area and looks more stock vs the universal ones from the parts stores. It mounts with one bolt and rests on the inner fender.

The overflow tube runs into the tank and it has a vent towards the back side to allow it to have any excess flow out. I had a vented cap on it, but upon hard acceleration, it would spill coolant out. I put a non vented cap on it and figure the vent on the top of the tank itself should prevent any pressure issues in the tank. I added a fitting to the tank vent to allow a hose to route anything that may come out to flow under the car and not in the inner fender area.

One thing that got my mind pondering is the tube molded into the tank. It comes from the top and the coolant in the tank and the molded tube are level. This tells me there is air in the molded tube part and may suck that air into the system upon recovery. That air will get pushed out next time and if there is still an air pocket, will get drawn back in. Yes, ideally it wouldn't draw any air in, but it's less than an open overflow tube.In an ideal situation, would the molded tube have coolant all the way up or will it level with the amount in the overflow tank?

I have a small leak in the radiator and that may cause the levels to be the same and/or not drawing back in like it's supposed to. Spring it will get addressed as far as the radiator and repairs.

When I blew into the overflow tank overflow tube, I could see coolant going up the molded tube and into the radiator with the radiator cap off. I put the cap on the radiator and did it again. I could see some bubbles coming out from under the radiator cap. I know that's not scientific, but should it be allowed to blow into the radiator if the system is cold and been sitting for a few days? I would think if the system is leak free, it would not do anything because the coolant can't push the air pocket out of the radiator because it has nowhere to go. Bubbles tell me there is something not right.

I may be overthinking it, but when I don't completely understand something, my mind wonders.

The cap that is on it:
https://www.oreillyauto.com/detail/...let-monte-carlo?q=radiator+cap&pos=0

The overflow tank:
https://www.classicindustries.com/product/all-years/chevrolet/nova/parts/551448.html


Non vented cap for tank:
https://www.classicindustries.com/product/all-years/chevrolet/nova/parts/371978.html

Vented cap for tank:
https://www.oreillyauto.com/detail/...coolant-cap/mtm0/82590?q=82590&pos=0

Not sure why the pics are sideways.