This page contains information for the radiator and coolant expansion tank that C&R Racing will hopefully be able to build for me.

Here's some additional information:

  • 90 degree fittings will be used to route the condenser lines through the radiator. These could be holes or notched. Either way they must be properly spaced vertically and I must be able to adjust the evaporator left/right on the radiator mounts.
  • Key cooling system heights from road:
    • Bottom of radiator: 5”
    • Top of Radiator: 17-1/2” (note it’s angled)
    • LS7 head steam outlets: 22” (adapter and tube increase that to 23-1/2”)
    • Top of expansion tank: 31”
    • Bottom of expansion tank: 25”
    • Top of Heater Core: 20”
  • The way that Howard Jones' C&R radiator is mounted leaves no room for vibration isolation. The first three pictures are my car. Note that there is only room for bolt heads on the outside of the vertical aluminum fins. The fourth picture with the rubber isolatators is another builder's approach. I’d like to use McMaster versions in the notches of the side tanks.

Coolant Expansion Tank

I have replicated a friend’s coolant expansion tank in SolidWorks. I need to have one made and when I last checked there were several other SL-C builders that wanted one. The picture below shows a mock-up being held in place.


My car is missing lots of equipment and, as can be seen in my friend's car below, space is very tight when the car is built:


  • Standoffs are spaced to allow 10 mm aerogel insulation to be sandwiched between firewall and tank. The firewall will use nut plates so that the tank can be removed without removing the interior.
  • Open to suggestions on location of standoffs and how to prevent the bolts from transmitting heat to the firewall.
  • The coolant level indicator.
  • Should I have some type of restrictor on the radiator bleed line input?
  • Steam vents should connect to bottom of tank and have internal tubes welded so that steam exits near top of tank… there isn’t much room to run tubes on outside. Maybe I could join the vent and steam lines before they enter the tank.

The following section summarize research that I have done regarding coolant swirl tanks and plumbing. At the end of the research is a suggestion on how I think things should be plumbed.

Radium Engineering

Radium Engineering has an interesting approach that I'd like to replicate. According to their documentation:

"The coolant expansion tank features two internal chambers. These chambers are NOT isolated from each other. There is cross-over passages at the top and bottom. The Radium Engineering coolant header tank also has an internal swirl feature, as shown below. Hot pressurized coolant from the cylinder head is sent into the vortex chamber's inlet tangentially and accelerated to a high rate of rotation. This is to help aid in deaerating the system."

Since the tanks are coupled there is no lost capacity. It would seem fairly straight forward to weld a cylinder with several holes in it as well as an internal tube that goes from the bottom to top where it bends 90 degrees and introduces the air tangentially. 

I was able to learn the following from looking at an actual Radium tank. 

  • The swirl chamber ID is 2.13" and it's 4.8" tall
  • The inlet is at the top of the chamber and is, as expected, tangential to the ID.
  • There is a 5/16" in the top of the swirl chamber to allow steam to flow into the expansion chamber. The OD of the hole is about 0.1" from the top of the chamber. This was drilled (as opposed to cast) and is aligned with the edge of fillet, so this dimension might have been driven more by manufacturing than functional reasons. The hole was drilled 190 degrees from the tangential inlet. It would have likely been easier to drill it at 180 degree so there is likely a reason that it's at that angle (see picture below).
  • There is a similar hole in the bottom to allow fluid to flow into the expansion tank. It's 170 degrees from the the tangential inlet. I'm not sure what that's the case, but I don't think it's important because it should always be covered in fluid. 
  • The OD of the clear tube is 1/4"
  • The 1/8" NPT push-to-connect elbow fittings were all metal. They were stamped with "1/4 N 01-UK MADE". They don't have swivel joint so they take a little work to install per their installation directions.


    Carroll Smith

    The following scans are from two of  Carroll Smith's books. Key takeaways:

    • A de-aerating swirl pot between the pump and the radiator which is bled to the header tank is an absolute necessity on any race car. He provides a sketch of one on last page of the Prepare to Win scan
    • He like to run a SMALL line from the bottom of the header tank to the inlet side of the water pump to always provide a positive head of water at the inlet of the pump to prevent cavitation.
    • He strongly prefers to make the header tans non-circulating.
      • If non-circulating. He runs a -10 or -12 to facilitate filling (not sure if this is from the header tank or separate). I assume the former
      • If circulating, restrict to 1/4" ID
    • Header tank appears to have separate tangential swirl locations for the bleed from the top radiator bleed and the coolant swirl pot.


    Griffin has the best article that I've read on cooling systems located here (that said, I like C&R Racing's products more). Here are some extracts:

    • Bleed lines go to top of expansion tank in an air pocket.
    • Expansion tank location placed on the low-pressure side where coolant velocity is low to facilitate deariation. 
    • Bleed line from top of radiator on the low-pressure side.
    • Radiator should be shrouded
    • Radiator should be isolated from vibration
    • Highest coolant flow possible; it's a myth that coolant flowing too fast won't have time to absorb heat in the engine or dissipate it in the radiator;So, let's start with the tiny nugget of truth. If you had a sealed rad (no flow) full of hot coolant, and subjected that rad to airflow, yes, the longer you left the coolant in the rad, the more it would cool. However, if you were to plot that cooling over time, you would find that the RATE at which the cooling takes place is an exponential curve that decreases with the temperature difference between the hot coolant and the air. Put another way - when the temperature difference (delta-T) between the hot coolant and the airflow is large, heat transfer (cooling) initially takes place very, very quickly (almost instantaneously). But as that happens, and the coolant cools, the delta-T becomes less, and the RATE at which further cooling happens gets less and less until the point where the coolant and air are almost the same temperature and continued cooling takes a very long time. This is Newton's law of cooling.
    • Only use distilled water; it has the highest specific heat of all liquids commonly used for coolant. It also reduces electrolysis.
    • 2% air in the system results in 8% less heat transfer, but 4% air results in a whopping 38% less!!

    Meziere and Saldana Products

    Cooling System Principles by Meziere and Saldana Products PDF.  The most relevant section is "The bottom of the [expansion] tank is plumbed to the low pressure (suction) side of the cooling system (after the radiator core and before the pump impeller). The smaller fitting on the upper portion of the tank is plumbed to the high points on the engine and radiator to remove trapped air and aerated water. This reservoir is located high and out of the main flow of water, and allows air to separate out of the water making your cooling system more efficient."

    Several Approaches

    It seems that everyone agrees that the surge tank should be on the low pressure side for two reasons: (1) bubbles are only going to migrate to an area of lower pressure and (2) a low-velocity, low-pressure air pocket will facilitate deaeration. With this in mind, the following routing seems to have the following problem. When the heater control valve is closed all of the flow goes directly into the expansion tank, thus placing it on the high-pressure side. I expect this to the case the vast majority of the time. 


    If you look at the Griffen article you will note that while the expansion tank in plumbed into the heater flow is done so after the heater core and is therefore on the low-pressure side of that circuit. If you look at the diagram below you will see that they are 100% clear that before the heater core is high pressure and after is low pressure.

    Therefore, it seems that the following plumbing is superior; (1) the expansion tank is on the low pressure side and (2) there is less plumbing. On the down side, a 5/8"to 3/4" adapter is needed between the HCV and the water pump. A bung would be welded on the 1.5" water pump supply line to take the 3/4" line from the expansion tank (easy to do if it planned up front).

    Pantera Cooling Post

    Air management in a motor’s cooling system is often taken for granted. Trapped air in the radiator or cylinder heads will inhibit the system’s ability to transfer heat and cool the engine. If the plumbing of the coolant system or the location of the system’s components tends to trap air the trapped air must be removed (i.e. vented). Air released in the cylinder heads as the result of coolant boiling must also be removed from the coolant because air in the coolant will reduce the heat transference properties of the cooling system.

    The primary element needed to vent trapped air in a cooling system is a tank connected to the lowest pressure zone in the cooling system, i.e. the suction of the coolant pump. This tank is commonly referred to as a header tank; however it is also referred to by various authors as an expansion tank, a surge tank, or a de-gas tank. Without a low pressure header tank the vents in a cooling system will have nowhere to vent because coolant or air will only flow to a zone of lower pressure. The header tank will be equipped with small connections for vent hoses near the top, and a larger connection at the bottom that can be “teed” into the coolant pump suction line. The header tank is operated with an air space above the coolant to allow for the thermal expansion and contraction of the coolant as it warms or cools, and to allow for surge as the engine speed rises and falls. The header tank is also topped with a radiator cap, making the tank both the fill point and the pressure relief point for the cooling system. The tank is mounted as high as possible within the engine compartment to maximize its efficiency as an air collection tank, a surge and expansion tank, and a coolant system fill point.

    Any air that collects within a cooling system after filling and “burping” is normally the result of coolant boiling in the cylinder heads. It is standard race car practice to use a swirl pot (aka a swirl tank) in the cooling system at the coolant outlet of the motor. If correctly designed, the swirl pot will spin the coolant around inside it, making any bubbles or air pockets collect in the middle, rise to the top, and exit via a small bleed tube in the top which is connected to the header tank. Thus the swirl pot keeps the cooling system free of air.

    The inlet and outlet nipples on the Pantera’s smaller cooling system tank are configured in such a way there is no mistake it was intended to be a swirl pot. Since a swirl pot is not functional without a header tank to vent to, it is easy to assume the taller cooling system tank was intended to be a header tank. The radiator cap, the large bottom connection and the volume of the taller tank reinforce the assumption it was conceived to be a header tank. These are the two standard tanks utilized in the cooling system of a race car. For an unknown reason the Pantera design team revised the use of the tanks. The header tank became a pressurized coolant recovery tank known as an “expansion tank”; the swirl pot became nothing more than a convenient location for the radiator cap which DeTomaso refers to as a “system tank”. With the header tank converted for use as a recovery tank the radiator no longer had a low pressure tank to vent into so the radiator vent was relocated to the swirl pot (aka the system tank), which was an inept decision since the system pressure in the swirl pot is higher than it is in the radiator.

    This change in the cooling system design was ill-advised. The Pantera’s cooling system has a problem with air being trapped in the radiator and it needs a functional vent system.


    Other research

    Coolant Level VIEW PORT

    I spent a lot of time looking for the right 1/4" NPT, 90-degree (i.e., right angle) push-to-connect fitting for the coolant level view port. Most are rated to only 250 degrees (temperature). The Legris 3609 was a good find:

    • Nickel-Plated Brass (reduces electrolysis vs. unplated brass)
    •  Swivel (makes it easy to install the tube!!!)