Automotive air-conditioning industryhas been experiencing significant changes in recent years due to environmental concerns. Concerns about the damaging impact ofthe chloroflurocarbon refrigerant (CFC) to the ozone layer have led to its complete replacement by the hydroflurocarbon refrigerant (HFC) R134a. Although HFC refrigerant is known to have limited ozone depleting effects, it still has a signific sully high global warming potential, about 1, 300 times higher than carbon dioxide gas of the same amount. This is increasingly becoming environmentally unacceptable. Such environmental pressure has led to international treaties, protocols and proposed legislation in some countries to either ban completely the use of HFC refrigerants over a short period of time, or to penalise its usage by levying a hefty environmental tax. Therefore, increased efforts and investment have been made by the automotive industry to seek new alternative refrigerants, which can eliminate or alleviate the environmental impacts of mobile HVAC system.

Carbon dioxide has been shown to be one of the most promising candidates for an environmentally friendly refrigerant. CO2 has been used before the adoption of CFC refrigerant and it is still used as a refrigerant in deep freezing industry today. Due to the particular thermophysical properties of carbon dioxide (its low critical temperature of about 31 °C), however, when used for mobile air conditioning, it has to be operated in a transcritical cycle in most mobile air conditioning usage, when the ambient air temperature is likely to be high. Thus in a mobile CO2 air conditioning system, the heat exchanger that dissipates heat to the ambient air (also known as gas cooler) has to operate in a high pressure supercritical state, usually over 100 bar. Its operation in supercritical state also means that in part of the gas cooler the temperature tends to be very high and that the refrigerant ternperature varies significantly (known as temperature slide) through the gas cooler. It is known that in a transcritical air-conditioning cycle, the cooling capacity and energy efficiency, known as coefficient of performance, which is defined as the ratio of cooling capacity to the power consumed, depends strongly on high gas cooler effectiveness and low gas cooler pressure loss. This can only be realized through careful refrigerant tube design and optimal combination of tube with the airside heat transfer surfaces such as louvered airways.

Flat tubes with multiple circular section geometry channels have been designed for both the conventional and CO2 air conditioning systems. While the circular section geometry chancel offers good strength characteristics, it is not most effective in maximizing heat transfer conductance for the same tube cross-section.

An improved arrangement has now been devised.

According to a first aspect, the present invention provides a substantially flat tube for an automotive heat exchanger, the tube comprising a plurality of flow channels extending in the longitudinal direction of the tube, the channels having substantially non-circular cros s- sectional geometry.

The present invention provides flat tube designs with non-circular cross-section channel geometry, that are narrow to reduce airside pressure drop, strong to withstand high pressure, of light weight and that also offer higher heat conductance and lower pressure drop compared to flat tube with circular cross-section channel with the same tube cross-section size.

It is preferred that the tube is extruded, the channels preferably being formed in the extrusion process. The tube is preferably used in an HVAC gas cooler having a working fluid operating in a substantially supercritical state. Preferably the tube material is aluminium. Preferably the refrigerant is CO2.

For optimum performance, it has been found that the ratio of the maximum channel dimension in section in the tube minor axis direction to the tube minor axis dimension is less than a factor of the ratio between the bursting pressure and the tube material yield stress: H/Ltminor <A oyeild/ (oy-eild + Pburst); where A is a safety factor and A < 1.

For optimum performance, it has also been found that the ratio of the sum of the length of all the channel widths to the tube major axis dimension is less than a factor of the ratio between the bursting pressure and the tube material yield stress : Swi/Ltmajor < B yeild/ (6yeild + Pburst) ; where B is a safety factor and B < 1.

The invention will now be further described, in specific embodiments, by way of example orily and with reference to the accompanying drawings, in which : Figure 1 is a schematic sectional view of a tube in accordance with a first embodiment of tie invention (triangular cross-section channels) ; Figure 2 is a graphic representation of effects on channel width on performance compared with a comparable circular section channel tube; Figure 3 is a schematic sectional view of a tube in accordance with a second embodiment of the invention (rectangular cross-section channels) ; Figure 4 is a graphic representation of effect on channel width for the second embodiment of tube (rectangular section channels) compared with a comparable circular section channel tube ; Figure 5 is a schematic sectional view of a tube in accordance with a further embodiment of the invention (waisted rectangular cross-section charnel) ; Figure 6 is a graphic representation of effect on channel width for the further embodiment of the invention compared to a comparable circular section channel tube; and Figure 7 is a graphic comparison between refrigerant temperature in a triangular cross-section channel tube and a comparable circular cross-section channel tube.

In the embodiment referred to, a heat exchange tube (1), typically of extruded aluminium material is provided with a series of substantially parallel working fluid (reirigerant) channels 2. Typically the tubes extend between headers and are stacked in a row having air-gaps between adjacent tubes. An airway or fin matrix may be provided in thermal contact with adjacent spaced tubes in order to maximise heat transfer. Typically, where aluminium is used, the heat exchanger is brazed together.

One of the preferred realisations of the invention is a flat tube 1 with multiple channels 2 of triangular cross-section. The second realisation is a flat tube of multiple channels of rectangular cross-section and yet a third realisation is a flat tube of multiple channels with a shape modified from rectangular cross-section.

The performance of the three realisations are compared to the benchmark round channel flat tubes under the same overall tube cross-section size. The maximum channel dimension in the tube minor axis direction is fixed as the same as the round channel diameter. The variation of the total heat conductance for each tube, incorporating a typical airside surface design and airside flow condition, and of the pressure drop through one tube, for a fixed tube length, are calculated with the variation of the channel width. The merits of each design are judged by these two performance parameters compared to the benchmark circular cro ss-section channel flat tube.

To achieve light weight, under a fixed overall tube cross-section, it is necessary to maximize total channel cross-section area. this maxirnisation of total channel area also helps in reducing the refrigerant pressure drop. Compared to the benchmark circular cross-section channel tube, all three realisations result in increased total channel cross-section area and therefore provide a lighter tube design.

The maximisation of total channel cross-section area needs also to result in a tube design having sufficient structural robustness. Therefore, the maximum dimension of the channel in the tube minor axis direction should be no more than a fixed fraction of the tube minor axis dimension. Similarly, over any intersection line between the tube end surface and any plane perpendicular to the tube end surface, the ratio of the total length falling into the channels to the total line length should not be greater than a fixed fraction. The exact value of these fraction numbers should ideally be a function of the desired burst pressure and the yield stress of the tube material.

The performance of this realisation, compared to the round channel case, is shown in Figure 2. It shows that for a fixed tube width illustrated here, a channel width (triangle base dimension) greater than 0.5 mm offers improved heat transfer conductance at the same or very slight increase in pressure drop. Considering, however, the structural requirements, the ratio of channel width to the triangle height should be limited by a maximum of about 3. Thus the optimal range of channel dimension should give 30° < oc < 65°.

To verify the preferred triangular channel design, some three dimension CFD simulations have been done to compare with the benchmark round channel flat tube. Figure 7 shows the result of the refrigerant temperatures in a flat tube of preferred triangular channel and in a benchmark circular cross-section channel flat tube. It can be seen that preferred triangular channel tube is more effective and cools the refrigerant to the same extent as a circular cross-section channel tube of a third longer length. For the same length, the new triangular channel tube will give a heat transfer rate about 7°/O higher.

Figure 3 shows the second realisation with rectangular channel. In this realisation the total channel cross-section area is increased by 27% compared to the benchmark round channel case and correspondingly the weight of the tube is reduced significantly. To maintain the structural robustness, the limitation on the two ratios as discussed in first realisation still holds.

As in Figure 2, the performance of the flat tube of the second realisation versus the channel width, compared again to the benchmark round channel tube is shown in Figure 4. The optimal performance range, offering better or equivalent under the same or much lower pressure drop, is 0.5 < w/H---2. 2.

A further and third realisation is schematically shown in Figure 5. The rectangular channel in the second realisation is modified to improve structural robustness of the rectangular channel. This design still offers weight reduction compared to the benchmark round channel flat tube. To maintain the structural robustness, again the limitation on the two ratios a-s discussed in first realisation still holds.

The performance of the flat tube of the third realisation versus the channel width is shown in Figure 6. Again a optimal ratio of channel width to the maximum channel height is identified as between 0.6 and 1.6.

In general, for optimum performance, it has been found that the ratio of the maximum channel dimension in section in the tube minor axis direction to the tube minor axis dimension is less than a factor of the ratio between the bursting pressure and the tube material yield stress: H/Ltminor <A ayeild/ (6yeild + Pburst) ; where A is a safety factor and A < 1.

For optimum performance, it has also been found that the ratio of the sum of the length of alL the channel widths to the tube major axis dimension is less than a factor of the ratio between the bursting pressure and the tube material yield stress : Swi/Ltmajor < B ayeild/ (6yeild + Pburst) ; where B is a safety factor and B < 1.