Radiator Heat Dissipation Testing

Comprehensive radiator tests with important tips for optimum results. — Bill Adams.

NOTE: This article has been corrected and updated 12/7/03.

SUMMARY:

This article will describe the results of comparative heat dissipation performance testing for several different types and sizes of radiators of interest to the overclocking community. Also briefly described are the equipment and the procedures used to generate the data.

The intention is to provide those designing their CPU watercooling system with some hard data by which more appropriate component selections may be made. It should be noted that all of the data was actually recorded by (reasonably calibrated) instruments, NOT manufacturer’s “ratings” which too often have little relation to the item’s real-world performance.

Radiators are one of the principle components (with the fan, pump and waterblock) of a CPU watercooling system, and may have a substantial effect on the system’s performance. May, rather than will, because watercooling systems are usually designed empirically as they consist of elements whose specific performance is often poorly defined (or understood, or applied, or all of the foregoing).

The goal of cooling “back” to ambient temperature necessitates substantial overcapacity and the “design process” is often a pragmatic serial selection of components to identify, and then replace, the limiting element.

Radiators do not function independently; rather their specific performance is dependant on each of the other watercooling system components’ capabilities and optimization. In particular, the selection and installation of fans is of primary importance, pumps are obviously crucial, and if the waterblock can’t extract the CPU’s heat then nothing else ever has a chance to perform.

But this article is not about fans or pumps (although some recommendations are made); their outputs are considered simply as variables in the determination of a radiator’s heat dissipation capability.

No amount of testing ever changed a product for the better, or made it any worse; but properly designed and performed tests can generate hard data to aid in the selection of a component suitable for the service conditions. Radiators (forced air-cooled crossflow heat exchangers) are tested under arbitrarily defined conditions to provide a consistent means of determining their absolute, and relative, performance capability. Good tests will utilize conditions closely approximating the actual conditions of use to minimize extrapolation and provide higher confidence levels for proposed applications.

Obviously, a radiator’s specific capacity will be determined by its frontal area, tube size, number and configuration, fin dimensions and spacing, and construction materials and methods, etc.; while its actual performance will be a function of how these “design” choices interact with the service conditions.

The principle variables affecting a radiator’s performance are liquid and air flow rates, and the temperature difference between them. A good description of the “Thermal Calculations for Heat Exchangers” can be found HERE.

These variables are not open-ended, as there are practical limits to the flow rates and temperatures; at high flow rates, the resistance (or pressure drop) becomes disproportionately high and the temperatures are normally dictated by the application and environment.

A decreasing rate of response to the change of a variable is indicated by an asymptotic curve (approaching a straight line / limiting value), the so-called “point of diminishing returns”. Performance data plotted as curves are useful for graphically indicating reasonable “limits” and some excellent examples for radiators can be seen HERE. A comprehensive write-up covering fan cooling design and specification can be found HERE.

So, given some well designed tests generating reproducible performance data, how does one determine what a radiator’s “rating” might, could, or should be?

An Engineer, Technician, Salesman, and User might all have different opinions, but the common ground should be that the radiator will “work” in the described conditions. Ah, the User too has an obligation: To know their (approximate) heat load, air flow rate and temperature, and the coolant flow rate and temperature in order to select an appropriate radiator; numbers are needed.

On the subject of numbers, it is worth remembering how the “work” that the radiator is performing is calculated:

• where Q = total heat transferred, Btu/sec
• Ww = coolant flow rate, lb/sec
• Cp = specific heat of coolant, Btu/lb°F
• T = exit temperature of coolant, °F
• Ti = initial temperature of coolant, °F

A more convenient restatement is:

Note that the above is NOT concerned with how close to ambient the “cooled” coolant temperature is. A radiator’s “rating” simply describes its heat dissipation capability (in Btus or Watts) at specific air and coolant flow rates. (“Btus” as used in this article are Btus/hr and can be converted to Watt*hours by multiplying by 0.2931)

The goal of CPU watercooling is to cool the baseplate of the waterblock as much as possible. To do this, a coolant temperature of only slightly above ambient is sought, the thought being that the lowest possible coolant temperature will maximize the temperature difference – and therefore the heat transfer potential. As the efficiency of the radiator is greatest with the largest possible temperature difference, this would suggest lower coolant flow rates to maximize the heat rejection by maximizing the contact time of the coolant and the radiator tube walls.

Bill Adams

From the excellent test data from Lytron (previously cited), the flow regimes to be investigated were thought to be fairly well understood and uniform data recording points were established based on practical conditions of use. The testing at (exaggeratedly) high air and water flow rates was necessary only to rough out the “limits” of practicality and have been truncated from the graphs shown.

The highest liquid temps tested (120 to 130°F) are the maximum (and may even exceed those) appropriate for the aquarium/utility pumps used by almost all computer water cooling systems; they were included to make the data somewhat more complete.

Plastics have an upper service temperature limit that is related to the applied stress, and this applies particularly to the vinyl tubing so prevalent. (The “cure” for vinyl tubing stress relaxation is called a hose clamp. The prevention of same is called silicone hose, again with hose clamps.)

In practice, both the air and coolant flow rates can be varied by the selection (or control) of the fan(s) and pump. These choices may affect the ambient (radiator cooling) air temperature, and will affect the coolant temperature. It should be appreciated that an attempt to return high temperature coolant to something close to the ambient air temperature is a “sizeable” undertaking and becomes progressively more difficult as the coolant temperatures rise – recall the diminishing returns comment above.

It is difficult, if not impossible, to assess radiators without becoming very involved with pumps and their characteristics. The first task is to measure their output and this is easily done by timing the filling of a gallon jug at various heights (Revelation #1); then one checks the flow through the test loop without a radiator (Revelation #2); finally one checks the flow through the complete system (Revelation #3).

These flow rates could be called low, very low, and vanishing; bring on the next pump. The plastic centrifugal pumps used by overclockers have relatively large impeller clearances and are intolerant of intake backpressure, and not much more capable on the discharge side. The flow rates at which data was collected are: 0.42, 0.65, 0.90, 1.23, and 2.06 gpm. These higher flow rates (through a complete system) will be beyond the capability of many (most?) OCing pumps.

The quantity of air moving through the radiator is one of the principle determinants of the unit’s performance; but while it is fairly simple in a laboratory environment to vary and quantify the air flow, for the user this is much more difficult. While it is the mass of the air that conveys the heat (which is loosely referred to as the volume flow rate, or CFM), it is the radiator’s backpressure resulting from this air flow that is the limiting factor (for a particular fan).

As these tests were designed to characterize the radiator’s performance and not the fan’s, the air flow was set based on the backpressure (measured as pressure drop in inches of water across the radiator) created by the unit being tested.

For these tests four “design basis” static pressures were selected; two representing reasonable low and high speed/thick bodied axial fan applications (0.015 and 0.05 in. H2O) resulting typically in 75% or so of “rated” output, and two much higher levels (0.15 and 0.25 in. H2O); typical perhaps of special, stacked, or centrifugal fan applications. Radiators supplied by the vendors with fan(s) were tested also that way, and the air flow recorded.

The equipment included:

• Haake A82 recirculating heater/chiller as the heat source (1000W heat capacity w/0.01°C RTD controller); with an EDC DC Voltage Standard (to 0.00001VDC) and HP 3940A dvm for remote control;
• Two DigiTec HT Series 5810 digital thermometers (0.01°C resolution) with YSI 703 linear dual thermister probes;
• Fluke 2190A digital thermometer (0.1°C resolution) and Y2001 Selector, with J and T thermocouples;
• Little Giant, Model #3E-12NYS pump, 500gph (mfgr’s rating);
• Lake Monitors digital flow sensor; 0.3 to 4.5 gpm, (with Fluke 8600A dvm and Acopian A12H1300 regulated power supply), physically calibrated (0.01gpm resolution, 0.1gpm accuracy);
• Omega HHP 701-20, 20psi digital pressure gauge (0.01psi resolution, 0.1psi accuracy);
• Fans; 2 Panaflo FBH-12G12L – 136 CFM, 4 Nidec TA600DC – 960 CFM;
• Energy Conservatory digital pressure gauge (direct and differential manometer, 0.1 Pa resolution, 1.0 Pa accuracy, where 1.0 Pa =0.000145 psi = 0.004015 in. H20);
• Kunz 441S (low velocity) thermal anemometer (mfgrs stated 5% accuracy assumed);
• a bucket of fittings, yards of hose, and way too much sheet metal ducting, pop rivets and 200mph tape.

The preceding equipment list was not easily arrived at, and is the result of several complete upgrades of almost all components. Mechanical gauges were exchanged for electronic, which were in turn were replaced with others of greater resolution and accuracy. Very small pressure, flow, and temperature increments necessitate rather sophisticated equipment.

The testing procedure was fairly straightforward:

• Liquid utilized was distilled water with 5% Water Wetter added.
• Set coolant flow rate (+, – 0.05 gpm) and air side static backpressure (+, – 0.2 Pa).
• Desired coolant temperature set at bath, then adjusted after 1/4 hr to achieve desired temperature at the radiator inlet under steady-state conditions.
• Coolant radiator inlet temperature continuously adjusted to maintain 10°C differential (+ or – 0.01°C max) with ambient air as test series progressed. (Ambient air temperature fluctuation was + or -1°C, or less.)
• All temperatures, coolant and air flow rates, and pressures recorded at each test point.

The testing procedures evolved as the equipment’s accuracy increased. When distinction is being made between 0.01°C increments, the assurance of steady-state conditions is important and quite dependant on comprehensive procedures consistently followed – in particular anything affecting the air circulation within the test room.

Bill Adams

The following are descriptions of the radiators tested. Note that these descriptions are substantially incomplete with respect to the description of water flow cross-sectional area and heat dissipation surface area (both internal and external); all essentially radiator design parameters. The internal volume is listed for each radiator and is usefully compared with the waterblock’s capacity to determine their relative flow velocity.

• Opening (Area): 5 x 5 in. = (25 sq.in.)
• Thickness (Volume): 3/4 in. = (18.8 cu.in.)
• Internal capacity (liquid volume): 72 ml
• Connections: 3/8 in. barbed hose connectors
• Tube: 4 parallel 3/8 in. copper tubes in series, on 1 1/4 in. centers
• Fins: crimped on corrugated aluminum fins, 16 per in.
• Outside Dimensions: 5 x 6 1/2 x 1 3/8 in. (w/connections)

• Opening (Area): 3 1/8 x 3 5/16 in. = (10.4 sq.in.)
• Thickness (Volume): 1 in. = (10.4 cu.in.)
• Internal capacity (liquid volume): 35 ml
• Connections: 1/8 in. CTS straight tube
• Tube: 8 offset 1/8 in. parallel copper tubes in series, on 5/8 in. centers
• Fins: crimped on flat copper fins, 15 per in.
• Outside Dimensions: 5 1/2 (w/connections) x 3 9/16 x 1 1/4 in.
• Factory fan: 11 cfm (measured)

• Opening (Area): 4 1/4 x 3 7/8 in. = (16.5 sq.in.)
• Thickness (Volume): 4 1/2 in = (74.1 cu.in.)
• Internal capacity (liquid volume): 172 ml
• Connections: 1/4 in. CTS straight tube
• Tube: 15 offset parallel 1/4 in. copper tubes in series, on 1 in. centers
• Fins: crimped on flat aluminum fins, 8 per in.
• Outside Dimensions: 6 3/4 x 4 3/4 x 5 in. (w/connections)
• Factory fan mounting: Panaflo FBH-12G12L (67 cfm “rating”), to be added shortly

• Opening (Area): 6 x 4 in. = (24 sq.in.)
• Thickness (Volume): 3 1/2 in. = (84 cu.in.)
• Internal capacity (liquid volume): 221 ml
• Connections: 1/4 in. CTS straight tube (modified to 3/8 in. OD)
• Tube: 16 offset 1/4 in. parallel copper tubes in series, on 1 in. centers
• Fins: crimped on flat aluminum fins, 10 per in.
• Outside Dimensions: 8 1/4 (w/connections) x 4 1/8 x 4 1/8 in.
• Factory fans: to be added shortly

• Opening (Area): 5 1/2 x 5 1/2 in. = (30.3 sq.in.)
• Thickness (Volume): 1 1/8 in. thick = (34.1 cu.In.)
• Internal capacity (liquid volume): 365 ml
• Connections: 1/4 in. female NPT (modified by drilling out the 0.25 in. orifices to 0.4 in.)
• Tube: 30 flat copper tubes in 15 parallel pairs with 5/16 in. gap (2 rows, in line)
• Fins: soldered folded copper fins, 16 per in.
• Outside Dimensions: 7 3/4 x 6 3/8 x 2 3/8 in. (w/connections)

• Opening (Area): 6 x 6 in. = (36 sq.in.)
• Thickness (Volume): 2 in. thick = (72 cu.in.)
• Internal capacity (liquid volume): 220 ml
• Connections: barbed 3/8 in. hose connections
• Tube: 13 full thickness corrugated brass “plates”
• Fins: soldered slit folded copper fins, 12 per in.
• Outside Dimensions: 7 3/8 (w/connections) x 6 x 3 in. (w/connections)

• Opening (Area): 9 5/8 x 4 3/4 in. = (45.7 sq.in.)
• Thickness (Volume): 1 7/8 in. thick = (85.7 cu.In.)
• Internal capacity (liquid volume): 334 ml
• Connections: 1/2 in. BS tubing compression fittings (but 1/2 in. NPT works ok)
• Tube: 16 parallel single flat 1 7/8 in. wide aluminum tubes with 3/16 in. gap
• Fins: brazed folded aluminum fins, 19 per in.
• Outside Dimensions: 14 1/2 (w/brackets) x 6 (w/connections) x 2 in.

• Opening (Area): 4 7/8 x 4 3/4 in. = (23.2 sq.in.)
• Thickness (Volume): 0.625 in. thick = (14.5 cu.in.)
• Internal capacity (liquid volume): 87 ml
• Connections: straight 1/4 in. CTS, 1/2 in, long (for 3/8 in. ID hose)
• Tube: 3 parallel flat 1/ 2 in. wide copper tubes, 4 series passes
• Fins: folded soldered copper fins, 20 per in.
• Outside Dimensions: 5 1/4 x 6 x 1 1/2 in. (w/connections)
• Factory fan mounting: Panaflo FBH-12G12L (68 cfm “rating”), 39 cfm (measured)

• Opening (Area): 4 x 4 in. = (16 sq.in.)
• Thickness (Volume): 3/4 in. thick = (12 cu.in.)
• Internal capacity (liquid volume): 53 ml
• Connections: barbed 3/8 in. hose connections
• Tube: 19 diagonal channel stacked aluminum plates, 2 1/2 in. long
• Fins: none, connected to adjacent plates at diagonal flow channels
• Outside Dimensions: 6 3/8 (w/connections) x 4 x 3/4 in.

• Opening (Area): 8 x 4 in. = (32 sq.in.)
• Thickness (Volume): 3/4 in. thick = (24 cu.in.)
• Internal capacity (liquid volume): 97 ml
• Connections: barbed 3/8 in. hose connections
• Tube: 38 diagonal channel stacked aluminum plates, 2 1/2 in. long
• Fins: none, connected to adjacent plates at diagonal flow channels
• Outside Dimensions: 10 3/8 (w/connections) x 4 x 3/4 in.

From the radiators’ dimensions, the cooled frontal area and volume were calculated. Additionally, the fluid capacity was determined (by weighing, full and empty). End tanks or un-finned tubing were not included as part of the cooled area, but were included in the volumetric determination. This data is graphed below for comparison.

It is worthwhile to appreciate the significant differences between these radiators, not just in terms of their physical size but also with regard their different designs, materials, and fabrication methods.

A note on copper tubing sizes (CTS): The table below describes the dimensions, but it seems clear that the 3/8 in. “tubing” size often attributed to watercooling radiators is, in fact, 1/4 in. CTS – whose OD nicely (if loosely) accepts 3/8 in. ID vinyl tubing.

Bill Adams

The assessment of a radiator’s capability, and its selection as a part of a cooling system, cannot be done without an understanding of its hydraulic characteristics. Since flow resistance is related to velocity (among other things), it is a simple task to obtain the data to plot the pressure drop, or head loss (in psi) vs. the flow rate (in gpm). Likewise the air flow can be varied to develop the data for a plot of the static air back-pressure (in in. of H2O) vs. air flow (in cfm).

The liquid side flow resistance graph below is a plot of the pressure drop in psi vs. the coolant flow rate in gpm for all of the radiators tested where the data is plotted for direct comparison of the different units.

In the above graph, the lower resistance/higher flow radiators are to the lower right (Rads G and F), while those with higher resistance/lower flow are to the upper left (Rads B and H). The high performers are parallel flat tube and fin units, designed for water. The lowest performers are the small 1/4 in. dia. unit and the newly designed “for computers” segmented flow tube and fin radiator.

The graph below shows the increasing static backpressure caused by the radiators’ resistance to air flow.

As with the previous graph, the radiators with the lesser flow resistance are to the right (Rads G and E). Note the steep slope of the flow curves for the low flow units: Bigger fans aren’t going to make much difference, whereas the larger higher flowing units can pass more air – but at even higher back-pressures. It is interesting that the air flow resistances of the BE Cooling 5×5 and the Big Momma heater core are virtually indistinguishable from each other, despite being of completely different construction.

For each radiator, a set of curves was generated of the heat dissipated in relation to varying air flow (using pre-determined static backpressure levels) and varying coolant flow rates (again at pre-determined levels). The order is from low to high capacity and, as with most things, it can be concluded that bigger is generally better.

The OCWC 4×4 was strongly hyped in some quarters, but has the lowest capacity of any of the rads tested.

Another very small radiator but with excellent materials (all copper) delivers surprisingly good performance.

Twice as big as the 4×4 with twice the performance, which isn’t saying much.

At low flow rates, the BE Cooling 5×5 delivers good performance despite its simple construction; the 5×10

at low flow rates should be twice as good, with substantial capacity.

The Danger Den Cube and the Senfu (below) are unique among the radiators tested in that they exhibit a specific “point” of higher performance. The gain for the Cube is substantial, +30%, if the coolant and air flow can be set just right. These round-tube multiple-pass radiators have relatively low wetted-wall areas and many changes in direction, so it seems plausible that a certain flow regime will be the most effective in promoting coolant-to-wall contact.

Interestingly, the Senfu’s optimum is at a higher coolant flow rate than the Cube’s, despite (or perhaps because of ?) having smaller diameter tubing. Both radiator’s “sweet spots” necessitate greater than conventional air flow rates.

Bill Adams

Fairly good cooling in a compact package, but the unit’s high flow resistance (pressure drop) must be addressed during the pump selection. The new version being twice as large should be about equal to a typical heater core, but with much much higher coolant flow resistance.

This unit is a miniature copy of an automotive type radiator with good performance, particularly at low flow rates.

From the onset of CPU watercooling it has been recognized that heater cores are inexpensive, come in a variety of sizes, and work well; the Big Momma is no exception.

A somewhat different design, this radiator needs a higher flow to function efficiently – as it’s very low flow resistance suggests; the highest performing unit, but also the most costly were it to be purchased new.

Several items are apparent: Higher air flow increases the heat dissipation (no surprise here), BUT higher coolant flow decreased the heat rejection rate. Not by much, but this was so – with some perturbations – for all radiators tested. The decrease due to higher coolant flow was between 10 and 20% for the various radiators.

The two lower dissipation curves, at static backpressures of 0.015 and 0.05 in.H2O, are those attainable with conventional single axial fans and represent real-world capabilities. Before considering the use of the upper curves, investigate what it takes to deliver air at a fan backpressure of 0.25 in.H2O.

And here we are at the issue of “ratings”: Is it appropriate for a vendor to specify air and coolant flow rates that are far beyond those generally attainable with the components in general use by CPU watercoolers?

The comparative graph below shows the heat dissipation curves for all the tested radiators at the attainable static backpressure of 0.05 in.H2O. (The required fan air delivery can be read off the “Static Pressure vs. Air Flow” graph for the radiator of interest.)

The above graph has some intriguing anomalies; several radiators show increased heat rejection at “moderate” coolant flow rates. This was apparent during testing and the values were checked for accuracy. Of the four, two are round tube and two are flat tube – but there are other round and flat tube units that had no rise.

The data suggests that some radiators have an optimum flow rate, which must be experimentally determined (presumably by the manufacturer). Is it worth the effort? The peak is 10 to 30% of total capability.

Two of the units tested (the OCWC 4×4 and 4×8) were exactly the same except for one being twice as large as the other, and the results were consistent with the size difference. The cooling performance of the Black Ice II can probably be extrapolated from the Black Ice data, but the unit’s very high pressure drop will go up as well – bit of a challenge at the vendor’s spec’d 1 gpm.

Two of the radiators were provided with fans from the manufacturer, the AquaCool and Senfu, and the air flow was measured to get an idea of their capabilities. The AquaCool with 11 cfm of air flow will have a capacity of about 200 Btus (58 Watts) if a rather low coolant flowrate is assumed. The Senfu data will be added at a later date.

Another two units had provisions for the mounting of 120 mm fans, so a Panaflo FBH-12G12L was mounted to observe the deviation from the fan manufacturer’s “rating” – how much air actually passed through the radiator. These aluminum frame fans had a nominal rating of 68 cfm (actually 67.14 @ 3.6 mmH2O) and on the Black Ice flowed 39 cfm at a static backpressure of 0.05 in. H2O, indicating a capacity of 615 Btus (180 Watts) with a coolant flow rate of 0.65 gpm, or 550 Btus (160 Watts) at the vendor’s recommended 1 gpm. The Danger Den Cube data will be added later.

The test results graphed above are directly applicable only to those radiators actually tested, but should serve as a useful guide to estimating the performance of others of similar construction and materials. The results cannot be applied to a different type of radiator than that tested, but can be readily extrapolated to characterize similar units of different sizes by calculating the volume or area to estimate a unit’s cooling capability.

This article is flirting with a thermal characterization known as Specific Dissipation (SD), which is defined as the heat transfer rate of a heat exchanger divided by the maximum temperature difference across the heat exchanger.

As I have not the ability to display equations, nor most readers the inclination to view them, those interested in the “theory” are referred to an SAE paper, 2000-01-0579, on “The Effect of Changes in Ambient and Coolant Radiator Inlet Temperatures and Coolant Flowrate on Specific Dissipation” which can be found HERE.

Bill Adams

Everything has a cost and the work required to benefit from a radiator is the effort to pump coolant through it and air across it. It is clear that different radiators have different flow resistances, and pump and fan selection should be made accordingly. These differences between units illustrate the difficulty in selecting a component based on a “rocking” consensus – if ALL the other system components and conditions are not the same as those for which the unit “rocks”, neither will be the results.

Both pump and fan selection should be based on:

1. A determination of the desired flow rate,
2. An understanding of the consequential pressure drop/backpressure based on all of the actual in-system components, and

3. Pump and fan selections made on the basis of their ability to deliver the desired flow AT the predicted backpressure.

All fan and pump manufacturers have P-Q curves, use them (or suffer mucking about with different units and B.S. numbers from “expert posters”).

The trend lately in pump selection seems to be towards over-sizing, with the rationale that the output can then be throttled back to find “the sweet spot” where the maximum cooling (minimum temperature) is obtained. It has been shown that some (few) radiators do indeed seem to have a point of higher performance, but this is not so for most.

To understand the watercooling system “sweet spot” concept, another bit of information is needed – the relationship between the coolant’s flow and heat transfer in the waterblock. The graph below illustrates this for a popular waterblock, where the difference is shown between the waterblock baseplate and coolant (inlet) temperatures vs. the flow rate.

The temperature was recorded with a type T thermocouple embedded in the baseplate over the die area (per AMD’s recommendations) with a 77 Watt heat load applied through a 100 mm² heat die. For this test, the inlet temperature was held constant, so the reduced temperature difference is in fact the reduction in baseplate temperature attributable to the higher coolant flow rate.

Since the flow through the system will be the pump capacity at the total head resulting from the sum of the individual flow resistances of all the components, the effect of higher velocity should be determined for all components. HIGHLY RECOMMENDED to those interested in such is the Bible for pipers: “Flow of Fluids through Valves, Fittings, and Pipe”, Technical Paper No. 410, by the CRANE Co (also available as interactive software).

Enlarging the connections of the waterblock tested above made a very substantial reduction in its pressure drop.

Reducing the overall system flow resistance is beneficial, if only to permit using a smaller pump. But the key to greater watercooling effectiveness lies in providing higher flow through the waterblock

It is useful to consider the different tube and plate configurations tested: Round, Flat, and Plate:

• Round Tubes are used for higher-pressure applications or for manufacturing economy, obviously the latter for CPU watercooling. The fins for round tubes are mechanically compressed around the tubes (or the tube is upset).
• Flat Tubes are used to reduce the internal volume, thereby increasing the wetted wall surface to coolant ratio, and are the norm for higher performance water cooling radiators. Fins for flat tubes are soldered or brazed. Both the round and flat tube radiators are affected by shrouding when one tube lays behind another, hence the offset tube arrangements seen in newer designs.

• Plates are a type of flat tube but run the full length and width of the radiator, which is made up of a stacking of said plates. Stacked plate radiators may have brazed fins for increased performance. The older style heater cores are made up of corrugated brass plates soldered together to form the coolant and air passages.

Each of these types represents a particular combination of performance, materials, and manufacturing costs, and each was designed with a specific end use and market in mind. It is to be expected that “radiators” for water, oil, or freon will be different and, when all put to the same task, perform differently. But, with various compromises, they will all “work”.

Consider oil coolers handling the high-pressure output of lower volume positive displacement pumps, with high fluid viscosities, and much higher fluid/air temperature differentials; yet, if properly sized, they can still provide adequate performance.

The performance difference between copper (actually brass with copper fins) and aluminum radiators is not clear. On a strength-to-weight basis, aluminum has the advantage but this is not a factor for CPU watercooling applications. Aluminum proponents (radiator manufacturers!) claim a 15% benefit, but no comparable data seems to be available (at least to the writer).

The highest performing unit in this group is a good example of a high dollar water cooling radiator: Made by Serck in the UK, it has ½ in. connections, extremely low flow resistance (an aluminum stacked plate design with brazed fins and very low air backpressure) and great cooling performance. It is from the water cooling system of a piece of medical equipment and cost $15 off eBay with a pair of ½ in. liquid-tight quick disconnects. High performance does not have to cost big bucks.

Bill Adams

A cost effective system is one that could be said to be efficiently performing its intended function; it’s got to work, and its elements’ capabilities should be reasonably balanced.

With radiators of similar cooling capacities, the lowest flow resistance unit is to be preferred to maximize the (potential) flow; this applies to both air and coolant.

If larger radiator connections (to ½ in) are an option, go for it – can’t hurt, may help; but note that larger connections do NOT increase cooling, they reduce the flow resistance thereby increasing the coolant flow rate – which may not be what the system needs for its optimization.
The widely used heater cores are certainly the most cost effective solution, and often the best performing.

Radiators cannot tell if the air is being pushed or pulled; such depends only on the fan, presumably selected for those characteristics.
Two pushing and pulling (series) fans together do NOT provide twice the air flow, but will increase it.

Fans should be set at least 1 in. away from the radiator to minimize shrouding by the motor and the noise from fan blade turbulence.
Since the static backpressure is a consequence of the radiator’s design (and the air flow velocity), the air flow rate can only be changed through fan selection or speed.

Design the fan’s installation with appropriate standoff and ducting; don’t just cable tie it to the radiator.

The above are not limits – one can select a fan or pump to push more or less – but its output may decrease the efficiency of some other part of the system.

It is not advised that a radiator’s Btu (Watt) “capacity” or “rating” be directly equated with the heat generated by the CPU. The overall efficiency of a CPU watercooling system will depend on many factors; the individual components’ capacity, their compatibility, the installation “details”, the heat input by the pump, the environment, etc., etc.

For low heat CPUs, a medium sized radiator will suffice. However, OCed T’Birds and all TECs should go for the largest radiator that can be tolerated, as both pump and fan sizing will be facilitated. And the greater the radiator’s overcapacity, the closer to ambient the coolant can be cooled.

Bill Adams

In reviewing smaller radiators designed for water cooling (but not CPU “watercooling”), some trends become apparent:

• Air flow recommendations are limited to a static pressure (head) of 0.25 in.H2O or less.
  BUT: Note that the axial fans used by overclockers have NO OUTPUT at this pressure. A more useful (but inaccurate) guide is to limit the drop to 0.015 in. H2O (for about 75% of “rated” flow) for low speed fans, and 0.05 in. H2O (again for about 75% of “rated” flow) for high speed (and thick body) fans.

  A better procedure is to check the actual PQ curve for that fan, they are NOT all the same – and all manufacturers have them for all models. An interesting example is
  HERE.

• Liquid flow pressure recommendations are limited to a drop of 7 psi or less.
  BUT: Note that many centrifugal pumps used by overclockers have NO OUTPUT at this pressure (7 psi equals 16 feet of head), and this pressure drop is for only the radiator. Select the radiator with the lowest possible pressure drop that will provide the required cooling, otherwise the pump selection becomes most difficult.

  Two small radiators in series will double the hydraulic flow resistance and have a marked effect on (reduction of) pump output. If two must be used, run them in parallel.

• Liquid flow rates range between 1/2 and 2 gpm.
  I suspect that CPU watercooling systems with a 2 gpm flow rate are VERY few; 1/2 gpm (or less) is probably typical, and 1+ gpm could be said to be “high performance” – all quite relative, of course.

  Time the filling of a gallon jug from the reservoir input (or pump return, but through the entire system) and you will have a real number. Larger lines, sweep elbows and fittings with chamfered edges may be cheaper than upsizing the whole system again later.

Some quote oil or transmission cooler specs as if they were applicable to watercooling applications, ignoring the differences in fluid properties, flow rates, temperature differentials, etc. Be skeptical of all data generated for non-CPU watercooling applications – OCing conditions are rather different than the industry norm.

It must again be observed that the manufacturers’ “ratings” for pumps and fans in no way reflect their installed performance.

This is not really deceit as much as a reflection of the very large differences between the manufacturers’ test conditions and the users’ service conditions. The principal use for “ratings” is to compare relative performance between units and to estimate anticipated performance, rather than to believe that the “rated” flow is that actually being achieved.

CPU watercooling radiator “ratings” by the vendors (where such exist) are about worthless, and often quite misleading.

This is due to a total absence of standardized testing conditions (unlike the pump and fan industries). It is up to the consumers to insist on product descriptions based on real data from actual testing.

Comments and questions are welcome, but please post in the
Overclockers.com Cooling Forum so that others may view.

Bill Adams