Radiatiors: Series to Parrallel

[Bob Ramberg]

I am going to change my radiator set-up from two radiators in series to two in parrellel. I would like to downsize the water tubes from/to the engine. Because the flow to each rad will be cut in half this seams reasonable. What am I overlooking?

The existing tubes are 1.25" x .065" 6061T6 AL. I would like to go to something like 1"x.035" or even 7/8" x .035" 6061T6. Its easier to route, less expensive and saves a couple of pounds. Has anyone ever gone that small or thin on their water tubes?

[Steve Demeter]

Bob,
I don't know if downsizing the tubes from the engine would be a good idea, if they are the right size now. My logic is that the engine still needs the same flow in and out of it and going to smaller tubes would possibly reduce that flow.

But I am sure that there are people more knowledgeable than myself out there. Just my 2c.

[Bob Ramberg]

[size=2]The set up I have now is one 1.25" circuit with the two radiators in series with all the water flowing through both radiators. [/size]
[size=2][/size]
[size=2]In the new set up, the water would leave the engine in a 1.25" tube, then tee into two smaller tubes. There would be two 1" (or 7/8") circuits with the radiators in parrellel. Each radiator would only see half the flow. Total flow would be the same. The second radiator should also work better (higher dT across it) since it now has higher temperature water going into it.[/size]

[size=2][/size]
[size=2]The two 1" tubes have more flow area than in a single 1.25" tube. Two 7/8" are about the same area as one 1.25".[/size]

[RoadHazard]

I'm no fluid dynamics expert, so take this for what it's worth, but...

Why would the flow be divided in half? Just 'cause you split the two hoses with a tee doesn't mean the coolant will automagically split itself 50/50. More likely most of the coolant will flow through the "easier" radiator (the one that's closer, warmer, lower down, etc.) with very little coolant circulating through the other radiator. In fact, without two water pumps I can see one radiator getting no flow at all.

Apart from that, won't splitting the flow reduce your cooling effect? In other words, if the water is flowing through two radiators now, flowing through only one will cool it less, no?

[Bob Ramberg]

[size=2]If the resistance to flow of the two radiator circuits is the same, then the flow will be the same. This will require equal length and size tubing for the left and right side radiators. Not hard to do. [/size]
[size=2][/size]
[size=2]There will always be flow through both sides even in a poorly balanced system. As the flow through the "loose" side increases so does the resistance and pressure drop. Eventually the pressure drop throught the "loose" side equals that of the "tight" side and the flow rates stabilize although at uneaqual rates.[/size]

[size=2]As for cooling effectiveness it should actually increase. The pertinent parameters are the inlet/oulet temperature differences between the water and the air (refered to as the log mean temperature difference or LMTD). With two radiators in series, the second radiator sees water that has already been cooled by the first. The LMTD is lower and less heat is transferred to the air than by the first radiator.[/size]

[size=2]In a parellel set up both radiators see the same water inlet temperature and operate identically. They also see only half the flow so the residence time in the radiator is longer. They should cool better.[/size]

[Frank C]

Bob:
The LMTD's will be better, but the convection coefficient will drop with the reduced flow rate. If one assumes that the convection coefficient on the coolant side is changing with flow rate to the 0.8 power, at half the original flow rate, the coolant side convection coefficient will be about 0.57 the original value. The change may not be quite as big as you expect.
- Frank C

[Bob Ramberg]

[size=2]Thanks all for responding. The main reason I'm doing this is because I managed to pick up some nice llight AL radiators to replace the heavy brass units that came with the car and they are set-up for a parrellel flow system. The car currently has its rads in series. The space between the rear bulkhead and the engine is extreemly tight. (The bulkhead has a bump-out for the water pump pully!) Smaller tubing would be much easier to route through this tight area. IT is also less expensive and saves a few punds as a bonus. I wonder how thin is too thin on the tube wall for the smaller tubes?[/size]

[size=2]BTW, anyone know the flow rate vs. rpm curve for our pump?[/size]

[Dick R.]

Bob,

I have an 85 Van Diemen 1600 with side radiators plumbed in parallel. The rad side of the "Y" point splits uses 1 inch OD tubing for both inlet and outlet. The engine side of the splits is roughly 1 3/8 OD.

Dick

[Bob Ramberg]

Thanks Dick. That's good to know. I'd rather not break any new ground here. Do you know what wall thickness was used?

[Dick R.]

Thanks Dick. That's good to know. I'd rather not break any new ground here. Do you know what wall thickness was used?

Bob,

I don't know the wall thickness. Whatever the "factory" used. It is aluminum and certainly didn't seem "flimsy" the last time I had the cooling system apart. If it helps, the "Y" point is welded as is the fitting for the temp gauge sending unit.

Dick

[Mark H]

My SF 84 Reynard has 1" id hose for all the plumbing to the rads. Never cracks 190 deg. at speed.

[Bob Ramberg]

[size=2]My new rads are from a FC so I probably have more surface than I need. I think I could go as small as 7/8" and be ok on flow and pressure drop. Only problem is the reducers (ie. silicone hose reducers from Pegasus) would cost me a small fortune. I'm going to try to sorce some OEM type automotive radiator hoses that have reducers built into them and cut them apart.[/size]

[Mark H]

Gates green stripe hose is as good as the blue stuff.
I know this seems like a small rig up but 1" hose can be clamped down to 7/8", if in doubt use 2 hose clamps. This will last at least 1 season and counting.... I know.

[Bob Ramberg]

Mark,

Is your car using 7/8" water pipes?

[Brian]

I use 16 ga. aluminium tube, 6061 T6, and anneal it if I have to bend it - I sand bend it if it's a tight bend. you can anneal if you want to bead the ends as well, makes it a little easier. 16 Ga aluminium is .050" I think.

I use silicone hose exclusively if I have any choice at all. It is far superior to regular hose in longevity and reliablity, in my opinion. I've used it as long as five years so far, with no degradation.

I have had zero success getting silicone hose to "tighten down" on the next smaller size of tube nor getting it to stretch over the next size larger hose. Tried really hard once, and ended up just making aluminium adaptors from tube when I failed...the adaptors worked just fine!

Most recently I got my hose bits from Pegasus, like the web site a lot so I don't really bother with other shopping...

Brian

[DanW]

I have used both 1" 16 gauge (0.065") wall thickness and 18 gauge (0.049") tubing for cooling lines. If you anneal to bend, take care not to anneal the ends on the thinner wall tubing where you attach the hose. If you apply a bead, the welding does anneal the end until it age hardens in a few weeks.

If the tubing is too thin, the tube may collapse from the clamping force of the hose clamps, particularly in tubing that has been annealed. There is not a big weight savings between .028"/.035" tube and .049/.065" aluminum. Most of the weight is the coolant inside the tube.

[Mark H]

True Brian, the regular black rubber hose is not too good, but the Gates Green Stripe is dubble thick and has a lot less rubber and more reinforcement.
Bob, no I have not used 7\8, my rads are 1 1\4" and my tubes are 1", so I just use 2 clamps on the tube.

[Bob Ramberg]

Brian - how do you go about sand bending tubing? I've heard of it but have never tried it or have seen it done.

[tstrong]

I have used 6061-T6 .035 tube successfully for years for straight water pipes. Have never tried to bend. Have always annealed the ends to form beads and have never had a problem with crushing the ends. The hose clamps load the tubing pretty uniformily. Todd

[Bob Ramberg]

How did you bead the end of the tube? All of the beading tools are pretty expesive, except perhaps the Earl's tool.

[Brian]

What I did: with 1.25" OD .050" wall (plain 16 Ga. tubing, AL gauge is thinner than steel gauge for some reason) 6061-T6 tube...

I filled the tube with dry sand, vibrated it to settle, used wooden plugs in both ends, have an air hole so you don't get a bomb when the air heats up (don't ask how I know this is bad). You want the tube as full as possible.

I anneal the bend location. I've used a camp stove and oxy-acetelene torch. You can put soot on the tube using a rich flame from the torch and heat till the soot disappears, or you can heat till you can char a pine stick. Yu can air cool or quench - I quenched because I am too impatient to wait for it to air cool. The result is T-0 condition, or dead soft, and four weeks later it age hardens back to T-4 condition. Don't overheat or your nice water tube in waiting will become a blob on the floor...messy.

I used a form made from plywood to set the radius that I wanted. I just worked the bend in gradually, I probably used a 7" radius or so. Most of my bends were 20 degrees or so, to follow the line of the chassis, but each tube had a 90 degree bend at the end to point at the engine.

I didn't bead those tubes. I found that with the silicone hose the grip was so good that beading was irrelevant. we have subequently beaded tubes with a Parker tool and with an Earls tool, I liked the Earls bead better, but I liked the Parker tool better.

I recently sand bent an ARB from 3/4" .125" wall steel tube. I filled it with sand as above, and then flame bent it at dull red heat exactly as if it was a solid bar. Worked awesome, very smooth even bends, no collapse of the tube at all. Don't even hesitate to try this.

Next I want to try to sand-bend 1.5" dia .040" wall mild steel for exhaust headers. I think sand bent headers are totally classic in looks, very beautiful.



Brian

[Charles Warner]

The easiest and cheapest way to bead a tube is to take it to a radiator shop. Costs about $10 per end and the bead will be perfect.

[Bob Ramberg]

Thanks all. Great info as usual from a great site.