1971 MGB/GT with Buick 215cid V8 (owner: Curtis Jacobson)
"Measuring a Summer Day"                                                                     1971 MGB/GT with Buick 215cid V8 (owner: Curtis Jacobson)

Cooling System Design Considerations

as published in British V8 Newsletter, Volume XIV Issue 2, August 2006

by: Curtis Jacobson

The components we think of collectively as "the cooling system" are responsible for regulating engine temperature within a prescribed range (e.g. 185 to 205 degrees Fahrenheit). Besides simply cooling, this system must help the engine come up to temperature quickly and evenly. Especially in cool weather, it's important the cooling system not over-perform because cold engines wear rapidly, pollute more, achieve inferior fuel economy, and typically produce less power. Perhaps more people are familiar with the other performance extreme. Under-performing systems are prone to catastrophic failures such as burst hoses and warped metal under angry storm clouds of superheated steam. Engines that run cool cause uncomfortable passengers in cool weather, whereas engines that run hot cause discomfort on warm days. But for maximum performance (particularly on carbureted engines) achieving a very consistent operating temperature is extremely helpful for fine tuning overall engine performance.

Cooling system performance requirements are at least somewhat proportional to power output, so it seems intuitive that upgrading the engine in a British sports car would require increasing the capacity of the cooling system. Considering that many of the cooling systems on British sports cars are only marginally adequate with stock engines, it may seem like cooling system improvements would be mandatory. This article will attempt to present an overview of customary cooling system modifications, with some pragmatic and hopefully innovative suggestions to keep cost, complexity and effort within reason. With that said...

There are two primary factors to consider: air flow and heat transfer.

Air Flow:

Air moving through the engine compartment, but not through the radiator, provides very inefficient cooling. To ensure all incoming air goes through the radiator, recirculation shields around the radiator are called for. They don't have to be totally air-tight, but every bit of air that slips around the radiator is lost-potential.

By the same reasoning, it's an excellent idea to place a shroud and/or "fan ring" around cooling fans. An engine cooling fan's purpose is to move a current of air, not to simply "stir up" the air. Unlike paint being stirred in a can, the air moving past a cooling fan should come past the blades only exactly once.

Fan shrouds duct the air that passes through the radiator core down to the circular cross-section of the fan or fans. With a shroud fitted, the fans can draw air through the entire area covered by the shroud, and potentially through every square inch of radiator core. Fan shrouds are particularly helpful when the vehicle is stationary. At high road speed, in installations with a lot of natural airflow through the radiator, fan shrouds can sometimes restrict airflow significantly. Many car manufacturers install rubber flaps on their fan shrouds that act as valves. Air pressure forces the flaps open at road speed, so air bypasses the fan(s).

Enjoying this article? Our magazine is funded through the generous support of readers like you!
To contribute to our operating budget, please click here and follow the instructions.
(Suggested contribution is twenty bucks per year. Feel free to give more!)

Fans work most efficiently when fitted with fan rings. A fan ring is essentially a very short bit of ductwork (either separated from the tips by a small air gap and stationary, or less-commonly integrally attached to the propeller tips and rotating with them). Fan rings provide a straight and efficient current of air by restricting vortexes and turbulence at the blade tips.

Fan rings can be used in conjunction with fan shrouds, but in some installations it makes sense to use them on their own! Three factors that suggest use of fan rings without shrouds are: 1. the vehicle will mainly be traveling at high speed, 2. the fan's cross-sectional area is a relatively large fraction of radiator-core area, and/or 3. for whatever reason, the fans are mounted in front of the radiator where a shroud is usually too obstructive. When a fan ring is fitted without a shroud, it should extend right up close to the radiator fins.

Fan shrouds and rings have the additional safety benefit of helping to keep fingers intact.

What about the fan itself? There are many fan-related design decisions, including engine-driven versus electric, and diameter. The amount of room available is often a driving factor in these decisions. As some engine driven fans have been shown to provide insufficient airflow at idle, it has become more common to supplement them with one or more electric fans.

Engine driven fans waste power and fuel by spinning when not needed, when the engine is below temperature and also at speed when ram-air makes them redundant. "Flex" fans are designed to address the later issue by moving proportionally less air per revolution at high RPM. Clutch devices for engine-driven fans are uncommon on sports cars.

Many installers elect to omit the engine-driven fan in lieu of one or more electrically driven fan. One advantage of this decision is that it's easier to achieve close fit between the propeller tips and fan shroud or fan ring. (No extra room is required to accommodate engine rocking on the motor mounts, which incidentally are rubber on street cars for a reason. Volvo trucks have a brilliant solution: their fan ring is mounted to the engine and connected to the fan shroud only by a rubber grommet.)

 Please support the sponsoring companies who make BritishV8 possible, including:

Pragmatic suggestion: Japanese carmakers typically fit wonderfully-engineered electric fans on their vehicles. They run quietly and move a lot of air for their size. Often, Japanese carmakers install a second, smaller fan on air-conditioning condensers. If you need small fans for your application, you should consider checking the Japanese section of your local junkyard. These fans are so reliable and long lasting, there's not much demand for them and junkyards sell them at very low prices. The only downside is that you'll probably have to fabricate your own bracketry. When selecting an electric fan, you should note that the fan blades are often directional by virtue of blade shape. If your application blows air in the opposite direction, the fan will be less productive. Irregular blade spacing is another interesting and typical feature; it's used to make the fan run quieter.

The CFM ratings on aftermarket fans should be read with great skepticism. If you can, buy and try several fans. Return the ones you don't like.

Electrical fans require a fair amount of electric current. Make sure wiring to the fans is of sufficient gauge (i.e. cross-sectional area) and consider switching the fans on and off with a relay because relays are designed to handle more current than typical toggle or thermostatic switches. Any voltage drop that occurs because of resistance in your wiring or switches will reduce fan motor speed. Finally, remember to observe polarity. The fan won't work so great if it's spinning backwards!

Air flow obstructions can also significantly reduce cooling system performance. If anything is blocking airflow into or out of the radiator core, look for ways to remove or relocate it. (Throwing away your car's cosmetic grille will give it that racecar look!) It is prudent, of course, to provide some sort of protection to keep debris out of the fans and radiator.

Air flowing into the engine compartment should have a clear path to flow out, or air pressure will build in the engine compartment and reduce airflow through the radiator.

When a vehicle is in motion there will always be areas of higher and lower air pressure around and within it. It makes good sense to use this fact to advantage. For example, the cowl area (outside and at the base of the windshield) is a high pressure area, which is why cowl fresh-air vents work so well. When it comes to engine cooling, it should be noted that (at road speed) the front wheel area is typically lower pressure than the engine compartment. Holes cut in the inner fender-wings typically draw hot air out of the engine compartment. (If one routes exhaust headers through the holes, they serve double-duty; they get exhaust heat out of the engine compartment by a very direct route. MGB-V8s fitted with "RV8-style" through-the-fender headers rarely have cooling problems.)

1971 MGB/GT with Buick 215cid V8 (owner: Curtis Jacobson)
1971 MGB/GT with Buick 215cid V8 (owner: Curtis Jacobson)

Figure 1:
          •     "Stock" MGB radiator (with oversized ports installed, altered mounting, and plugged cap.)
          •     Twin Honda Civic a/c condenser fans ("pushing") fitted with steel fan-rings
          •     Engine-driven fan omitted
          •     Simple mesh screen in lieu of original 1971 MGB grille
          •     Volkswagen Jetta pressure tank (aka "header tank")
          •     Plumbing as described in the text below
          •     Stock 1963 Buick water pump, with stock pulleys
          •     Oil cooler omitted (note: Buick V6 "metric" oil pump is used with directly-mounted spin-on filter.)
          •     Air damn fitted below bumper ("urethane" type with no air holes)
          •     Stock (un-louvered) MGB hood (removed for photograph)
          •     Un-vented inner-fender wings (i.e. headers exit the conventional route...)
          •     Heater omitted, and home-made "cowl-induction" system fitted in its place
          •     Note: In fifteen years of operation (in North Carolina and in Colorado) this cooling system has never over-heated.


Here's a tip that may save you a lot of money. You probably don't actually need a custom-made radiator! For example, many people have found that the standard MGB radiator core is more than adequate for a 200+ horsepower V8 engine, provided the overall cooling system is well conceived and in good maintenance. It's relatively easy to modify old-fashioned OEM radiators to get the port sizes and locations you need. Old radiators do get clogged up, but your local radiator shop can unclog them, or you can buy a standard replacement radiator for most models of British cars (and typically they're cheaper than custom-made radiators.)

Your local radiator shop can also re-core your old radiator, and doing so is often economical. If space permits in you application, you might ask them about installing a thicker or longer core. Adding a couple extra inches between top and bottom tanks is generally better, for reasons explained below. An "MGB V8 radiator", as sold in some parts catalogs, is an MGB radiator with an extra long core.

It certainly doesn't hurt to have a radiator with extra capacity, provided you use it wisely. If you decide to go shopping for a radiator, whether from a different model car (e.g. 1965 Ford Falcon radiators are particularly popular on MGB conversions) or from a custom radiator manufacturer, here are some things to consider:

Copper is a slightly more efficient material for radiator cores than aluminum, although the difference is less than many people think. Many people also report that copper-and-brass radiators last longer before cracking/leaking. There are, however, many design parameters at play (e.g. tube size and wall-thickness, variation in solder, etc.) and we don't have access to any reliable numbers to quantify or prove either of these statements in practice.

Copper and brass radiators are certainly much easier to modify or repair than aluminum (because soldering is much easier than TIG welding.)

Aluminum is lighter weight than copper or brass. Due to location, extra pounds of radiator (and coolant) will affect the car's weight bias more than extra pounds of engine.

As a matter of common sense, thicker radiators can provide more heat transfer than thinner radiators, provided they aren't so thick that they restrict airflow. Re-coring three-row radiators with thicker four-row cores is commonly done, but the added row will not be as efficient as the first three. Air flowing through the last row will have been heated by the first three, so heat exchange (from coolant to air) will not be as great for these rows. For this reason, if room allows, a larger core area may be much more beneficial than a thicker core.

If you have occasion to visually compare two radiator cores, look at how the fins contact the tubes. Older-style tube-and-fin construction, where the tubes pass through holes in the fins, are generally less efficient than newer-style serpentine fins. On tube-and-fin radiators, there is a lot of variation in fin design and on how tightly the fins contact the tubes. A tight mechanical connection between tubes and fins is very desirable because it facilitates efficient heat transfer into the fins.

Radiator shops like to paint radiators black. They shouldn't. Paint reduces the efficiency of radiator cores. (Typically aluminum radiators are anodized, instead of painted, which is a good thing.) If you do paint your radiator, apply as thin a coat of paint as you can. Color choice won't noticeably affect performance. (If you have an optical pyrometer this is easy enough to prove to yourself.)

Interesting trivia: the total surface area of fins and tubes on a high performance radiator is typically well over 100 times the frontal area of the radiator.

Since there needs to be a pressure-relief valve somewhere in the cooling system, a remote "pressure tank" (aka: header tank) is commonly fitted. Although metal pressure tanks are popular, they aren't recommended. Provided it's mounted as the highest point in the cooling system, a strong, transparent, relatively high-volume, plastic pressure tank of the kind found on many newer European cars makes it possible to verify coolant level at a glance and to see if contaminants are in the system. Junk yards give these tanks away for free, or you can buy them economically at the VW parts counter. You'll know you're buying a pressure tank because a relief valve will be mounted in the cap, and the cap will be labeled with its pressure-rating. These tanks normally have two ports: a smaller one at the top and a larger one at the bottom. A hose (or preferably rigid tubing) should travel uphill all the way from the top tank of your radiator to the pressure tank. A second, larger hose (or preferably rigid pipe) is highly recommended from the inlet port on your water pump uphill to the pressure tank. (See photo above for illustration of these recommendations. Please note that second pipe mentioned is part of a strategy to reduce cavitation at the pump, which is discussed below.)

Air trapped within the cooling system severely compromises its performance. For filling coolant and purging air, there needs to be a vent at the highest point in the system. It's common to see a bung (with threaded plug) soldered to the radiator top tank for this purpose, but if the pressure tank is mounted high enough and the radiator top tank is vented to the pressure tank (as previously suggested), a bleed plug on the radiator isn't necessary.

In practice, the pressure relief valve itself really should be the "vent" at the highest point in the cooling system. If the pressure relief valve is lower in the system, there's little hope for it releasing air without releasing coolant too. Ideally coolant flow should be relatively still in close proximity to the pressure relief valve so air bubbles will accumulate (and not be swept away.) When the pressure relief valve is in contact with flowing water, it is far more likely to open at a lower-than-intended system pressure (i.e. open due to a "local pressure spike").

Big problems are likely if the highest point in the cooling system is within the engine itself. This problem may sometimes be avoided by fitting the remote pressure tank high on the firewall. Don't forget that the heater is part of the cooling system too, and that it may also need to be purged of air.

Make sure all water hoses throughout the cooling system are routed so that they won't become crushed or kinked. Formed metal piping, joined with short lengths of Gates green-stripe hose, is really the hot ticket. The coolant ports on your intake manifold and water pump define what hose diameters you should be using. Your local radiator shop can quickly and easily replace the input and outlet ports on your radiator with larger diameter ports (salvaged from an old radiator.)

Make sure your engine thermostat opens fully. (You can test it by putting it in a sauce pan of water on your kitchen stove.) If you have a candy thermometer you can verify that it opens at the rated temperature.

Use high quality coolant, mixed 50/50 with water. High quality coolant already contains a generous cocktail of "water wetters" and corrosion inhibitors, so additional additives aren't recommended. Propylene glycol is better for the environment than ethylene glycol, and you should use it. (Besides being something you don't want in your local groundwater, ethylene glycol apparently tastes good to animals. It causes kidney failure and painful death. Propylene glycol is relatively inert. It's an ingredient in many foods, including Cool Whip.) Neither kind should go down your drain; take used coolant to a recycling center. Don't mix the two kinds because they have two different specific gravities, so you won't be able to reliably verify coolant concentration by measuring specific gravity. As for the water itself, distilled water is generally recommended over tap water to further prevent ionization that could result in corrosion and pitting of aluminum components.

It has come to our attention that some aftermarket parts suppliers categorically recommend over-speeding factory water pumps by 30 to 35 percent by fitting aftermarket pulleys with non-OEM ratios. In our opinion, that may be very bad advice for British V8 readers. It's certainly bad advice for many high performance (e.g. road or endurance racing) applications. Our advice is to maintain the OEM pulley ratio. If you feel compelled to tinker with pulley ratio, we advise that production-car engines which are routinely operated at high RPM should generally slow down coolant flow. For one thing, water pumps are prone to cavitate when operated at too high speed.

When a pump cavitates, it spins without pushing water, which creates vapor bubbles. A more generic explanation is that cavitation tends to occur when a pressure-drop and a temperature-rise occur simultaneously. Cavitation most frequently occurs right at the water pump. Vapor bubbles that are pressurized as they pass through the pump, collapse or implode with destructive force. The superheated steam bubbles created by cavitation can cause severe damage throughout the cooling system, including especially aluminum heads and radiators. (Incidentally, modern water pump impeller profiles are engineered to help reduce cavitation. Those of us that choose to use forty year old water pumps are more vulnerable.) If in doubt about your application, leave your pulleys at the factory-designed ratio.

In an article in April 2004's Motor Age magazine, Kevin McCartney provides a further warning: "High-performance water pumps often increase cavitation. The damage may show up in the intake manifold or elsewhere instead of the pump itself. Standard water pumps provide all the flow required for proper engine cooling. A wide-open thermostat actually restricts coolant flow below the capacity of a standard water pump. So a pump designed to increase capacity will usually increase pressure changes and cavitation, while doing very little to actually increase flow unless the thermostat is removed. Of course, this is not recommended, as removing the thermostat will cause other problems."

Race car engineers have implemented various solutions to prevent or minimize the effects of cavitation. Venting the inlet port of the water pump to the pressure tank, as described above seems to help. If you feel you may have a bigger problem, research coolant "swirl pots". (We'll leave the rest to you.)

What about secondary heat exchangers? Oil coolers are a traditional option. If you're fitting a BOP/R engine in an MGB, consider omitting the oil cooler because it's not needed and it introduces problems of its own. These engines are happy to have all the oil flow they can get, and the oil cooler can significantly reduce flow. (It's desirable that cold engines come up to temperature quickly, so the oil cooler should have a thermostatic valve, but thermostats and valves add their own respective failure modes.) Hoses are expensive, and they're a pain to route. Modern motor oil formulations are much more resistant to breakdown from heat than the motor oils of the 1960's, so "sludge" isn't the issue it used to be. If you feel you must mount an oil cooler, put it someplace where it doesn't obstruct airflow to the radiator, where it gets enough airflow to work, and where it's protected from curbs and road debris.

Although engine tuning is outside the scope of this article, we'd be remiss not to mention that engine tuning problems very frequently cause cooling problems! If you have a cooling problem, consider checking that your ignition's vacuum advance is functioning properly. The manifold vacuum advance system is an important aid to idle and low speed cooling. Secondly, verify your air/fuel mixture. Lean carburetor jetting will cause an engine to run hot. Troubleshooting this can be complex because carburetors have different jetting at idle and cruise speeds. A jetting error should especially be suspected in any car that runs cool at low engine speeds, but runs hot at cruise speed. All these factors are interrelated - that's part of the challenge of engineering your own car!

Disclaimer: This page was researched and written by Curtis Jacobson, but it was inspired by a few paragraphs formerly included in Dan Master's essay on design considerations for those contemplating an engine swap. Views expressed are those of the author, and are provided without warrantee or guarantee. Apply at your own risk.

Photos by Curtis Jacobson. All rights reserved.

Note: If you like this article, you'll probably also like "Tuning for Temperature Control" by Jim Blackwood.

Editor's note: Looking for tricks and tips? Check out the British V8 Photo Gallery to see how pretty a Ford Falcon radiator looks in Bruce Mills' 74.5 MGB, how Martyn Harvey created a simple, effective de facto duct between the slam panel and radiator on his 1980 MGB, how Mikel Moor notched and reinforced his frame to make room for an extra-wide cross-flow radiator, how Dale Spooner fabricated an elegant aluminum fan shroud, how Robert Milner slipped a nifty and powerful "side-winder" belt-driven electric fan into a very tight space, how Robert Milks bleeds air from his top hose, how Jim Stuart used a Volvo pressure tank on his 1973 MGB-GT, how Bill Jacobson (no relation) mounted his pressure cap very high and also fitted trick chromed-steel coolant pipes, etc., etc. There are over fifteen hundred high-quality color photos in our gallery, so you'll find many more ideas!
British V8 Home:        Read the Magazine        Photo Gallery        Web Forum        Annual Meets        Contact Us        Site Map