GB2514830A - Pulse tube cooler - Google Patents

Abstract

A cold head 2 for a pulse tube cooler has a regenerator 4, a pulse tube 14, and a heat exchanger 10 connected therebetween. The heat exchanger connects between a first end 12 of the pulse tube and a second end 8 of the regenerator, and a first end of the regenerator 6 is connectable to a compressor. A phase control device is connected to a second end 16 of the pulse tube, and in use controls flow dynamics (e.g. the relative phases of the pressure and mass flow variations) within the pulse tube to provide cooling at the heat exchanger, thereby maintaining a negative temperature gradient between the first and second ends of the regenerator. The pulse tube has a wall 15 which has a porous portion 28 to for allowing a working gas to enter or leave the pulse tube directly through the porous portion, as indicated by arrows 30. The porous portion is nearer to being parallel than perpendicular to the temperature gradient between the first and second ends of the regenerator.

Description

PULSE TUBE COOLER

The present invention relates to pulse tube coolers that have improved efficiency.

Pulse tube coolers have evolved as an alternative to Stirling cycle coolers. Their operation is very closely related; the main difference is that instead of using a "solid" piston or displacer to provide an expander component, this frmnction is provided by a column of gas that acts as a springy, deformable displacer.

The motion of the gas column in a pulse tube can be controlled mechanically by a warm end piston/displacer but more often the motion is controlled by fluid "phase control" components. These avoid the need for extra moving components and control the mass flow in the way that they respond to the applied pressure variation. Various mechanisms have been employed to provide this thnction. These include designs that use orifices or inertance tubes in combination with reservoir volumes. In addition a second inlet and orifice combination can be connected between the conipressor input and the warm end of the pulse tube to ftmrther modify the response.

Pulse tube coolers can be used in the same wide range of applications as Stirling cycle coolers.

Examples include those for cooling optical components both for space and terrestrial applications. Their main advantage over Stirling cycle coolers is the elimination or simplification of the moving expander,' displacer components. This can improve overall reliability and reduce vibration. Potentially it also offers lower cost as a pulse tube cold head does not require high precision.

A good general description of present pulse tube technology is given in: Proceedings of the Institute of Refrigeration (London) 1999-2000, "Development of the Pulse Tube Refrigerator as an Efficient and Reliable Ciyoeooler, Ray Radebaugh, NIST.

Figure 1 shows the cold head 103 for a conventional Stirling cycle cooler. The cold head 103 consists of a regenerator 104 and an expander cylinder 105 that are connected via a heat exchanger 110 (which also acts as a fluid connection between the regenerator and the expander cylinder) at the cold end (at temperature Te). An expander piston 113 is movably engaged within the expander cylinder 105. The working gas volume that includes the regenerator 104, cold end heat exchanger 110 and the expansion space 111 above the expander piston 113 is subjected to a pressure variation provided by an "AC" compressor (connected via connector 107). Typically, the compressor, or pressure wave generator, does not have valves and usually has an ambient heat exchanger (cooler) to reject heat from the gas before it enters the cold head 103. The compressor and cooler are not shown in Figure 1.

In operation, the movement of the expander piston 113 is arranged to be out of phase with the pressure variation so that net work is done by the gas on the piston 113. This loss of energy from the gas causes the gas temperature to decrease which in turn allows heat to be absorbed from the cold end heat exchanger 110. Over a cycle the net result is that heat can be absorbed from an external load via the heat exchanger 110 and this energy is transported to the warm end of the cold head (at temperature Th) as work done on the expander piston 113, which may in turn do work on gas in a compression space 115.

In the gamma configuration shown, the expander piston 113 becomes what is tenned a displacer. The displacer takes energy from the gas at the cold end as an expander and directly transfers it back into a gas volume at the ambient end as a comprcssor. The additional gas volume is usually connected to the input from the compressor so that the expansion work is "recycled" to improve efficiency. This energy recovery is not essential; for low temperature coolers the expansion power is relatively small and can be just dissipated through a damping arrangement.

In Figure 1 there is no indication of how the movement of the displacer 113 is controlled in order to ensure that the work is done by the gas in the expansion space. In practice a small diameter shaft is usually attached to the displacer and taken through a seal in the ambient end to a mechanism that imparts the required motion. Various means are available for driving the displacer and these are generally well established in the literature.

In discussing Stirling cycle coolers and particularly pulse tube coolers, it is useful to consider the operation in terms of the phase relationship between the applied pressure variation and the mass flow of the gas between the cold end heat exchanger and the expansion space/volume.

The work clone Wby the expander piston is given by W = J P.C/V over a cycle P is pressure, V is expander volume For simplicity, sinusoidal* variations are assumed for pressure and volume. This integral has a maximum when the pressure and first derivative of volume are in phase. This is equivalent to requiring that the pressure variation and mass flow are in phase. (Note: Actual pressure and mass flow variations will generally have additional harmonics but their magnitude will be small and they will not fundamentally change the operating mechanism).

It is noted that while this relationship will give the highest cooling effect it does not generally give the highest efficiency because losses in the cooler are also phase dependent. However it is generally the case that if the pressure and mass flow are 90 degrees out phase then there is no net work done on the gas and hence no cooling.

Figure 2 shows a conventional arrangement for a pulse tube cold head 102 in which the expansion work is transmitted via a column of gas contained in a tube 114 (which may be referred to as a pulse tube).

The gas column can be considered to consist of three different zones. One zone 121 (which may be referred to as a first tidal volume) behaves like the expansion space 111 of a Stirling cooler in that gas has a tidal movement into and out of the cold end heat exchanger 110. At the other end of the pulse tube 114 is a similar zone of gas 125 (which may be referred to as a second tidal volume) that has a tidal movement into and out of the ambient temperature phase control components. This tidal volume 125 behaves in a similar way to the compression space 115 of the Stirling cycle cooler. In between there is a third gas zone 123 that separates thc other two. Ideally this gas volume 123 simply acts as a deformabic displacer that transmits work but at the same time provides thcrmal isolation.

Cooling is generated by appropnate phasing of the volume 121 with respect to the pressure variation.

Achieving this is a more complex task than for a Stirling cycle machine and this aspect will be described more fUlly below.

For efficient operation of the pulse tube 114 it is desirable to minimise any turbulence so that the gas column remains stratified and convection by mixing is avoided. Flow straighteners 120 may be provided for this purpose, at either or both ends of the pulse tube 114 (as shown in the example of Figure 2). These provide low velocity, uniformly distributed axial flows.

In the above description of the Stirling cycle cooler it was shown that a key requirement for cooling is a combination of pressure and mass flow variations that are in phase. In a pulse tube the cooling process is more complicated for two reasons: * Firstly, the deformation of the "gas displacer" (labelled 123 in the example shown in Figure 2) with pressure requires additional mass flows into the pulse tube.

* Secondly, as one of the main attractions of a pulse tube is the elimination of moving components, it is very desirable to control the mass flow using a system of fluid phase control components rather than using pistons/displacers.

In the following, the term "phase control device" is used to refer to any device which allows the relative phases of the pressure and mass flow variations to be such as to allow cooling to take place. The phase control device may operate using pistons/displacers, a system of fluid phase control components that may not comprise any moving parts, or a combination of the two.

The mass flows required into the pulse tube can be divided into two components: those in phase with the pressure pulse and those out of phase. If it is assumed that the pressure variation has a sinusoidal wavefonri: P = P0Sin(oit) (Mean pressure level is ignored; pressure is assumed to be AC component only) then a general expression for the mass flow variation is: th = A.Sin(wt) + B.Cos(oit) The first con-iponent is in phase with the pressure pulse and this has already been established as the "cooling" component. The second component is mass flow required by the "deformation" of the "gas displacer". This component does not give rise to any work input or output; instead it acts as a spring component that stores energy during half the cycle and releases it in the other half This combination of two components of mass flow is illustrated in Figures 3A to 3C. The pulse tube in 3A has the total mass flow required for operation as a cooler. This can be likened to the combination of a displacer that is both * Reversibly expanding and coniracting -the spring component * Displacing the tidal gas volumes into and out of the pulse tube The operation of the pulse tube shown in Figure 3A can be imagined as the combination of the separate processes illustrated in Figures 3B and 3C.

Figure 3B represents the "spring" flows that occur when the mass flows are conipletely out of phase with the pressure variation and no net work is done on the gas. In 3B I the pressure is increasing and all the gas flows arc inward. In 3B2 tEe pressurc is decreasing and all tEe gas flows are outward. It is seen that the "gas displace?' changes shape in accordance with the distribution of the mass flows.

Figure 3C shows the gas displacer considered as a solid displacer that is causing tidal mass flows into and out of the pulse tube by displacenient without any pressure variation. In 3C1 the displacement is from the cold end at Tc (top) to the ambient end at Th (bottom). In 3C2 the displacement is the reverse of 3d. The flows into and out of the pulse tube are consistent with these displacements.

In Figure 3B the mass flows associated with the gas spring are shown equally divided between the two ends of the pulse tube i.e. the mass flow from the cold end is the same as from the ambient end. This does not need to be the case; the mass flow could be entirely from either end or divided in any proportion.

The key characteristic of the "spring" flows is that they are out of phase with the pressure variation. The direction from which the "spring" mass flows enter the pulse tube does not affect the aniount of refrigeration generated, however it does affect the losses and hence the overall efficiency of the cooler. It is generally preferable for at least some of the "spring" mass flow to enter from the ambient end. If it comes in from the cold end it has to do so via the regenerator adding more thermal load to this component and hence increasing its losses.

Example of/he Onf lee Pulse Tube One of the earliest designs used in pulse tube coolers is the orifice pulse tube, as shown in Figures 4A and 4B, in which gas is allowed to pass between the ambient end of the pulse tube and a reservoir 124 via an orifice 122. The orifice 122 is an example of a fluid phase control component. Figure 4A shows the pulse tube with the orifice 122 closed. There is no mass flow at the ambient end and the mass flow at the cold end consists of only the out of phase spring component; there is no net cooling: For P = 1)Sin(cot) thcpr = B.Cos(wt) W P.dV= 0 If the orifice 122 is opened (Fig 4B) then the mass flow at the ambient end will be determined by the flow characteristics of the orifice 122. An orifice generally behaves as a resistive flow component where the mass flow is proportional to the pressure drop across it. As the resen'oir volume 124 remains close to the mean DC level then the mass flow the through the orifice 122 and hence from the ambient end of the pulse tube will be given by = constP.Szn(cot) = A.Sin(mt) The mass flow into the pulse tube at the cold end is a combination of the spring mass flow and the orifice mass flow: co/dend = thspr + = A.Sin(wt) + B.Cos(wt) The cold end now has a mass flow component that is in phase with the pressure variation and there is a net cooling effect at the cold end.

(Note: As might be expected, the cooling is equal to the work lost in pumping gas backwards and forwards through the orifice 122) Other Types of Pulse Tube The orifice pulse tube is probably the simplest design that has succeeded in producing reasonable levels of cooling at low temperatures. However it will be seen from Figure 4A that the spring mass flow enters the pulse tube entirely from the cold end. This flow adds a considerable burden to the regenerator and limits the ultimate performance. Two configurations that have been developed to overcome this limitation will be briefly described.

Double In let Pulse Tube In the double inlet pulse tube configuration there is an additional connection between the ambient end of the pulse tube to the compressor via a second orifice. The effect of this is to cause a mass flow into the ambient end of the pulse tube that is in phase with the pressure drop across the regenerator and cold end heat exchanger. This can be used to allow some of the "spring" flow to enter from the ambient end. The reduction in flow through the regenerator reduces the regenerator loss and allows an overall improvement in performance.

Although good performance has been achieved with this arrangement it has been found that the second orifice is inclined to be asynunetrie in its operation. This attribute tends to generate net DC flows that circulate around the pulse tube and regenerator. These flows do not undergo thermal regeneration as intended and even at low levels can produce unacceptable losses.

Inerance Tube An alternative to the Double Inlet configuration that attracted interest is the inertance tube configuration. In this configuration the orifice of the Single orifice configuration descnbed above is replaced by a tube or assembly of tubes that connect the ambient end of the pulse tube to a reservoir volume. The tube or assembly of tubes is a fhrther example of a fluid phase control component(s).

The tube assembly is arranged to have both damping and significant inertia. It terms of an electrical analogy the inertance tube assembly is represented by a series combination of a resistor and an inductor whereas an orifice is represented by just a resistor. The inductive component has the effect of allowing some of the "spring" mass flow to enter the pulse tube fiom ambient end. This helps to reduce the thermal load on the regenerator without setting up the circulating flows that tend to occur with the Double Inlet design.

Coaxial Designs In Figures 1 to 4 the cold head gcometry is that of a U bend. Figure 5 shows a coaxial geometry applied to the general pulse tube arrangement shown in Figure 2. The opcration is exactly the same cxcept that the regenerator is now located in an annular volume surrounding the pulse tube. The wall of the pulse tube no longer has to withstand the full gas pressure instead it just acts as a partition. This arrangement is advantageous as it provides a more compact and convenient envelope for the cold hcad. Similar arrangements have also been widely used for Stirling cycle cold heads.

Pulse Tube Losses There are four main loss mechanisms directly associated with the operation of a pulse tube: * Heat transfer through the bulk of the gas: The minimum value is set by thermal conduction in stratified layers but this can be greatly incrcased if there is significant mixing/turbulence.

* Natural Convection: The temperature gradient produces a corresponding density gradient.

The latter tends to drive a circulation of gas within the pulse tube. This effect causes pulse tube performance to be very dependent on orientation -pulse tubes generally work better with their cold ends pointing down.

* There arc two recognised losses associated with the interaction of the axial oscillation of the gas and the pulse tube wall.

o One loss, referred to as a typc of shuttle loss, is due to the constantly reversing temperature gradient between the gas and the wall.

o A second loss that has been termed "Streaming convection" does not have any simple explanation but appears to derive from the velocity boundary conditions imposed at the interface between the gas and wall.

The "Streaming convection" loss is discussed in: JR. Olsen, G.W.Swift, Acoustic Streaming in Pulse Tube Refrigerator: Tapered Pulse Tubes, Cryogenics, Volume 37, Issue 12, December 1997.

It is an object of thc prcsent invention lo provide an improved pulse tube cooler which at least partially address one or more of the problems with the prior art discussed above, for example by reducing losses.

According to an aspect of the invention, there is provided a cold head for a pulse tube cooler, comprising: a regenerator having a first end connectablc to a compressor; a pulse I-nbc having a first end and a second end; a heat cxchangcr connected between a second end of the regenerator and the first end of the pulse tube; and a phase control device connected at the second end of the pulse tube for controlling the flow dynamics in the pulse tube to provide cooling at the heat exchanger, thereby maintaining a negative temperature gradient between the first and second ends of the regenerator, wherein: the pulse tube comprises a wall having a porous portion for allowing a working gas to enter or leave the pulse tube directly through the porous portion, the porous portion being nearer to being parallel than perpendicular to the temperature gradient between the first and second ends of the regenerator.

Thus, an arrangement is provided in which a porous portion allows gas to enter or leave the pulse tube substantially laterally (which may also be referred to as "radially"). This provides a number of

advantages relative to the prior art.

For example, there is a greater range of possibilities for the distribution of both the refrigeration process and the "spring" flows required: i) In prior art arrangements such as that shown in Figure 5 for example, cooling only occurs at the cold end of the pulse tube. In contrast, the porous wall of embodiments of the present invention allows cooling to be additionally or alternatively distributed along the length of the pulse tube. The rate of cooling depends on the magnitude of the flow through the pores. Therefore, the spatial distribution of cooling can be set as desired by appropriate selection of the spatial distribufion of porosity of the walls. Distributing some or all of the cooling along the porous wall can be advantageous as it allows the pulse tube to act as a multistage cooler. Cooling part way along the pulse tube can be used to remove heat at a higher temperature and hence with lower power requirement. This approach reduces the heat flow to the cold end and allows more of the low temperature refrigeration to be available for cooling of the load. Efficiency can therefore be i inproved.

ii) In thc prior art the gas "spring" flows can only enter from the cold end or wanu end (or both). However, the actual flow requirement is distributed along the length of the pulse tube. The provision of a porous wall allows the flow requirement to be fulfilled with more favourably distributed gas flows. The resulting pulse tube can be likened to a "multiple" inlet pulse tube.

A thrther advantage is rclated to the ability to reduce gas velocities and gas movement with respect to the thermal gradients: i) In a conventional pulse tube all the gas flow has to be distributed across the end faces of the pulse tube. The flow straighteners are designed to provide low velocity, evenly distributed flows but there is clearly a lower limit set by the cross sectional area of the pulse tube. The provision of porous lateral walls allows the gas flows to be distributed over much larger areas allowing gas velocities to be considerably reduced. This helps to maintain streamlined flow with minimal mixing and turbulence ii) In a conventional pulse tube the gas flows arc all axial and this requires the gas to have siguificant velocities in the direction of the thermal gradient. The porous walls of the present invention allow much of the flow to be radial where the gas velocity is perpendicular to the thennal gradient. This reduces the axial gas velocities and allows the establishment of a more favourable temperature distribution.

A further advantage is that, where there is radial flow, the gas flow will be closer to being perpendicular to the wall than parallel to it. The axial velocity component that interacts with the tube wall will be much smaller than for a conventional axial flow design. This will result in lower losses that are dependent on the gas flow/wall interface. For example the "shuttle" loss will be reduced as this loss is proportional to the square of the axial displacement. The other loss related to the interaction between the pulse tube wall and the axial gas velocity is the "Streaming Convection" and it is also believed that the large reduction in axial gas velocity adjacent to the tube wall will also result in lower values for this loss.

Furthermore, there is the effect the radial gas velocity components have on natural convection losses.

Natural convection occurs when buoyancy forces drive a circulation of gas. Typically this occurs when a hot surface is positioned below a cold surface -in a simple model the heated less dense fluid rises and the cooled denser fluid descends. The magnitude of the natural convection heat transfer is dependent on the Grashof Number -a measure of the balance between the buoyancy forces that drive the circulation and resistive processes that reduce it. In a conventional axial flow pulse tube, the tube's walls are the principal source of damping -the smaller the diameter the more the circulation is suppressed. The axial gas flows do not generally have much influence as they will add to the circulation on one side and subtract on the other. In the radial flow pulse tube it will be seen that there arc two effects that will tend to reduce natural convection: firstly, the radial flows close to the wall will disturb the axial flow and will tend to suppress it; and, secondly, the core of the gas displacer where circulation could become established has a significantly reduced diameter which will also tend to suppress circulation.

In an embodiment, a phase control device is provided that comprises a piston configured to move within a cylinde-n In comparison with prior art Stirling cycle coolers, the piston can be confined to a greater extent to the warm end of the tube than can the displacer of the Stirling cycle cooler. This allows the arrangement to be made lighter and/or easier to manufacture, while also tending to lower vibrafion and/or manufacturing cost. In comparison with prior art pulse tube coolers, this approach allows for a more compact arrangement because there is no need for a reservoir volume to be provided to implement the phase control.

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which corresponding reference symbols indicate corresponding parts, and in which: Figure 1 depicts a cold head for a prior art Stirling cycle cooler; Figure 2 depicts a pulse tube cold head according to the prior art; Figures 3A-C depict operation of the pulse tube shown in Figure 2; Figure 3A illustrates a combination of spring action and displacement action, Figure 3B illustrates the spring action in isolation, Figure 3C illustrates the displacement action iii isolation; Figures 4A and 4B depict an orifice pulse tube according to the prior art; Figure 4A illustrates the pulse tube wilh the orifice closed, Figure 4B illustrates the pulse lube with Ihe orifice open; Figure 5 depicts a coaxial geometry version of the pulse tube cold heat! of Figure 2; Figure 6 depicts a pulse tube cold head according to an embodiment of the invention, in which a coaxial geometry is adopted; Figures 7A-C show how the flow regime of the pulse tube cold head of Figure 6 can be seen as a combination of spring and displacement actions in an analogous manner to the flow regime depicted in Figures 3A-C; Figure 7Adepicts the combination of spring action and displacement action, Figure 7B depicts the spring action in isolation, Figure 7C depicts the displacement action in isolation; Figure 8 depicts a variation of the embodiment shown in Figure 6 in which the radial flows have becn extended to the cold end heat exchanger and the connection to the phase control device; Figure 9 depicts an embodiment in which a warm end displacer piston is used to control the mass flows in place of fluid components; Figure 1 OA and I OB depict a heat exchanger configured to pass gas into the pulse tube in a radial direction: Figure IDA is a side sectional view, Figure 1DB is an end view; Figure 11 depicts an embodiment in which the pulse tube wall contains an electrofonned screen that is solid at the warm end and then becomes more permeable towards the cold end.

In an embodiment, a cold head for a pulse tube cooler is provided in which the flow of at least a portion of the working gas between the pulse tube and other components is diverted through a porous portion of a wall that is nearer to being parallel with the tcmperature gradient of the regenerator than being perpendicular to it. Preferably, the porous portion is within 5 degrees of being parallel, preferably within 1 degree of being parallel. The result of this arrangement is that the working gas enters or leaves the pulse tube in a predominantly lateral direction (perpendicular to the temperature gradient). As discussed above, this allows operational losses to bc reduced. An example of such an arrangement is shown in Figure 6.

In this particular embodiment, the cold head 2 comprises a regenerator 4 which has a first end 6 and a second end 8. In use a temperature gradient will be maintained along the regenerator 4 such that the second end 8 will be colder (at Tc) than the first end 6 (at Th). The second end 8 may therefore be referred to as the cold end and the first end 6 may be referred to as the warm end. In an embodiment, the warm end 6 is al ambient temperature in which case it may be referred to as the ambient end.

The regenerator 4 is connectable via passageway 7 to a compressor (not shown in Figure 6) at the warm end 6. As dcseribed above with refcrenee to Figures 1 and 2 the compressor may be configured for example to provide an AC pressure wave input, for example in the form of a sine wave (or a form in which the dominant component is a sine wave). The cold end 8 is connected to a heat exchanger 10 which in turn is connected to a first end 12 of a pulse tube 14, optionally via flow straightener 20.

The second end 16 of the pulse tube 14 is connected to a phase control device via connection 26, optionally via flow straightener 20. As described above with reference to Figure 2, a phase control device in this context refers to any device which allows the flow dynamics (e.g. the relafive phases of the pressure and mass flow variations) to be such as to allow cooling to take place. The phase control device may operate using a piston/displacer, a system of one or more fluid phase control components without solid moving parts, or a combination of the two.

In the embodiment shown, the pulse tube 14 has a cylindrical form and comprises a wall 15 defining a pulse tube volume. In other embodiments, the pulse tube nrny take other forms. In the case where the pulse tube is cylindrical it will be understood that references to the axial direction refer to the axis of the cylinder. In embodiments where the pulse tube is not cylindrical it will be understood that references to the axial direction refer to an axis of elongation or "long axis" of the pulse tube. Typically, the axis of the pulse tube 14 will be substantially parallel to the direction of the temperature gradient in the regenerator (i.e. the direction of steepest temperature gradient -the vertical direction in the orientation of the figures).

In the particular embodiment shown the regenerator 4 coaxially surrounds the pulse tube 14 (as in the embodiment of Figure 5). However, this is not essential. In other embodiments, the regenerator may surround the pulse tube but not have a conimon axis, may only partly surround the pulse, coaxially or not, or simply be positioned adjacent to the pulse tube.

In an embodiment, the pulse tubc shares a wall with one or both of the regenerator 4 and the heat exchanger 10 and/or, where provided, a flow distributer 11 between the second end of the pulse tube 14 and the phase control device (see Figure 8 for example). Any combination of these shared walls may comprise porous portions for allowing lateral passage of working gas into or out of the pulse tube 14 directly through the porous portion. In the embodiment shown in Figure 6, the pulse tube 14 shares a wall 16 with the regenerator 4 only. In the example shown the shared wall is a radially inner wall of the regenerator 4.

Figure 8 shows an example embodiment in which the pulse tube shares a wall with the regenerator 4, the heat exchanger 10 and a flow distributer 11.

In an embodiment, at least a portion (the uppermost region in the marked range 28 in the example shown) of the shared wall 15 is porous to a working gas such that a portion of the working gas can enter or leave the pulse tube 14 (indicated by arrows 30) through the porous portion of the wall 15. Allowing some of the gas to enter into the pulse tube 14 directly from the regenerator in this way does not change the distinction between "spring" type mass flows that are out of phase with the pressure variation and "displaeemenf' flows that are in phase, it just alters the way in which the gas displacer deforms and the distribution of the cooling effect. In the example shown, the additional radial inward flows in the range 28 effectively change the shape of the first tidal volume (the equivalent of the volume 121 of Figure 5 that does work on the gas displacer 123) so that it comprises both a component 121A in the form of a solid cylinder (as in Figures) and a component 121 B having an annular fonn located adjacent to the wall 15.

Figures 7A to 7C show how the new flow regime can be considered as a combination of "spring" flows and "displacement" flows analogous to the purely axial flow arrangement illustrated in Figure 3.

Figure 7B represents the "spring" flows that occur when the mass flows are completely out of phase with the pressure variation and no net work is done on the gas. In 7B1 the pressure is increasing and all the gas flows are inward. In 7B2 the pressure is decreasing and all the gas flows are outward. It is seen that the "gas displace?' 23 changes shape in accordance with the distribution of the mass flows.

Figure 7C represents the mass tiows when the gas displacer 23 is just displacing gas without any pressure variafion. In 7C1 the displacement is from the cold end (at Ic) and the radial inlets (the pores in the wall 15), which are at intermediate temperatures, to the ambient end at Th. In 7C2 the displacement is the reverse of 7CL The flows into and out of the pulse tube 14 are consistent with these displacements.

Although the gas displacer 23 does change shape it is important to note that it does not change volume.

Figure 7A is intended to represent the combination of 7B and 7C when the pulse tube 14 is operating as a cooler with both in phase and out of phase flows occurring with respect to the pressure variation.

In an embodiment, the flows through the porous portion of the wall are such as to cause minimal mixing or turbulence. The ideal is that the gas flows follow streamlines so that over a cycle they tend to return to the point where they entered the pulse tube 14. This (or close to it) is achieved by having low gas velocities and well distributed apertures/pores.

In an embodiment, the porous portion of the wall 15 extends from the second end 8 (the cold end) of the regenerator 4 along a portion, but not all, of the regenerator 4 towards the first end 6 (the warm or ambient end) of the regenerator 4. This is the ease for example in the arrangement of Figures 6 and 7A-C. In this and other embodiments the porosity of the porous portion of the wall 15 may be configured to decrease as a function of increasing separation from the second end of the regenerator, or vary in some other manner.

In the arrangement of Figures 6 and 7A-C the working gas flows into the pulse tube 14 through a combination of radial flows (through the porous wall) from the regenerator and axial flows from the heat exchanger 10. However, other combinations of flow, optionally of radial and axial flows or of flows directed in other ways, are possible. Figure 8 depicts an example in which the heat exchanger 10 is arranged to coaxially surround the pulse tube 14 at the cold end. A wall having a porous portion is provided (e.g. a mesh -see Figures 1 OA and 1 0B discussed below) to guide the flow from the heat exchanger 10 into the pulse tube in a radial direction.

In the example of Figure 8, a flow distributer 11 is provided for distributing flow from the second end of the pulse tube 14 to the phase control device. The flow distributer 11 may have a heat transfer thnction in addition to a flow distributer function. The flow distributer 11 may therefore optionally have the same or a similar configuration to the heat exchanger 10. The arrangement described below with reference to Figures 1 OA and 1 OB maybe applied, for example, to the flow distributer 11.

In the embodiment shown, the flow distnbuter 11 comprises a wall having a porous portion that allows for a radial flow between the pulse tube 14 and the phase control device (via the flow distributer 11).

In the arrangement of Figure 8, therefore, flows into and out of the pulse lube 14 are predominantly (or entirely) radial and the expansion and compression spaces 121,125 (first and second tidal volumes) are radially outside of the gas displacer 123.

All the pulse tube arrangements described above have assumed the use of a phase control device comprising a fluid phase control component, which avoids additional moving components. For example, a combination of incrtance tube and gas reservoir may be used.

Figure 9 shows an example of a complete pulse tube cooler, including a compressor 45 and a phase control device 47, in which the phase control device comprises a piston 32 (which may be referred to as a warm end displacer) configured to move within a cylinder 34. A warm end tidal volume region is no longer required. The pressure variation within the cold head causes a net force to act on the warm end displacer 32 that is equal to the product of the pressure and piston face area. The dynamics of the piston assembly are arranged so that there is both inertia and damping. The resulting motion gives the desired mass flows at the boundary between the cold end heat exchanger 10 and the pulse tube 14. The damping is provided in this example by a damper piston 40, mounted via a shaft located by suspension spring 46, and a flow restrictor 42. The damping can be adjusted by varying the flow restrictor 42.

This use of a warm end displacer 32 can be regarded as a gamma configuration Stirling cycle cooler in which the thermal insulation thnetion of a conventional displacer is transferred to a dcformable "gas displacer" -the "gas displacer" effectively becomes an insulating piston crown. One advantage of this arrangement over a conventional Stirling cycle cooler is that the displacer 32 is confined to the warm end.

This allows it to be both lighter and easier to make with lower vibration and lower manufacturing costs.

At thc cold end the operation of the pulse tube 14 shown in Figure 9 is the same as the embodiment illustrated in Figure 8. Sonic of the gas flow enters the pulse tube 14 via a radial connection with the cold end heat exchanger 10. The remaining gas flow is distributed along a section of the pulse tube wall 16 and enters via connecting apertures between the regenerator 4 and the pulse tube 14.

The warm end displacer arrangement will not be the first choice for all pulse tube applications because of the additional moving components. However it possibly more compact as there is no longer a requirement for a reservoir gas volume which can occupy a significant volume.

In the example shown, the compressor 45 (which may also be referred to as a linear pressure wave generator) is configured to impart a modulating pressure on the cold head 2 via connector 7 (in this case a length of connecting tube). The compressor 45 comprises a piston cylinder assembly driven by a linear motor 50. The moving components arc mounted on flexurcs (suspension springs) 46 so as to allow a close but non-touching fit between the compression piston 52 and cylinder 53. This arrangement is generally referred to as a "clearance seal" -there is leakage but it is small enough to be acceptable. The compressor 45 does not have valves and the modulating pressure can be regarded as analogous to a voltage that has both AC and DC components. The basic operating frequency may typically be in the range 50 to 100 LIz.

The configuration of the phase control device of the example of Figure 9 is independent of compressor configuration and could be replaced by one or more fluid phase control components without any modification to the compressor.

Implementation of Radial Flows For the radial flows into the pulse tube 14 there are number of approaches that can be used. Figures 1 OA and 1 OB show one possible arrangement for the case where the flows enter from the heat exchanger 10 (arrows 61). In this embodiment, the axial gas flow 63 from the regenerator 4 enters slots 60 in the heat exchanger 10 that direct the flow radially into the pulse tube. A layer of mesh 64 between the slots 60 and the pulse tube 14 is used to even out and/or straighten the flow distribution. As mentioned above, a similar arrangement can be used for flows between a pulse tube and a flow distributer (e.g. flow distributer 11 of Figure).

For radial flows between the regenerator 4 and the pulse tube 14 the porosity of the pulse tube walls 16 needs to be controlled so as to give the required flow distribution. As the pressure drop across the pulse tube wall 15 varies along the regenerator 4 it is expected that porosity will also need to vary significantly.

The gas flows within the regenerator 4 and porous wall 16 are expected to be laminar and this allows them to be modelled using simplified methods -e.g. software intended for mathematically similar processes.

A finite element thermal conduction model was used to estimate the range of permeability that would be needed to achieve the required gas flows. For a pulse tube 14 where the radial flow was comparable with the axial flow as per the arrangement shown in Figure 6, the range of local permeability varied by a factor of 100.

If the volume flow dV/dt through the regenerator 4 is given by: dt j.t dx where A1is flow area, icr is permeability of regenerator 4, i is viscosity of gas and dP/dx is pressure gradient, the permeability of the pulse tube wall 15 would need to vary in the range O.Olic, to ic. The regenerator mesh used in pulse tube coolers is usually ver fine. A typical mesh specification is: Mesh Number (No. Of Wires per inch): #400 Wire diameter: 0.030mm Aperture dims: -0.034 x 0.034 Porosity: 0.59.

The permeability of porous material is generally: * Proportional to flow area and hence porosity * Inversely proportional to hydraulic diameter/aperture size.

As it is required to reduce the permeability by a factor of 100 from material that already has only 34 micron pore size it will be seen that it is necessary to produce a tube wall 16 with apertures of 1 micron or larger. A variation of I to 34 microns does not give the required permeability range so it is also necessary to alter the effective porosity.

Although there arc a number of technologies that can be used to produce apertures in sheet material, many are not suited to the small dimensions required. One approach that is suited is the technology of electrofonning. For example an electroformed screen can be made to give a varying permeability by controlling both the size and density of the apertures (the density effectively determines the flow area/porosity). The thickness of the screen needs to be between 5 and 10 microns to allow apertures of-1 micron to be defined. The screen is sandwiched between two layers of fine mesh which can then be formed into a tube and installed between the regenerator and pulse tube volume. The mesh is used for two reasons: * To produce a more robust assembly that can be handled * To give a symmetric flow characteristic so as to avoid any tendency to produce net circulations between the pulse tube and the regenerator as these will tend to generate losses.

The range of permeability that can be defined with a single electroformed screen is large but if necessary more than one screen can be used in conjunction with additional layers of mesh to thrther reduce the permeability.

Figure 11 shows an example where the pulse tube wall 15 comprises an electroformed screen that is solid at the warm end (portion 66) and then becomes more permeable towards the cold end (portion 64) so as to provide a controlled radial flow into pulse tube 14.

Claims (16)

1. CLAIMS1. A cold head for a pulse tube cooler, comprising: a regenerator having a first end connectable to a compressor; a pulse tube having a first end and a second end; a heat exchanger connected between a second end of the regenerator and the first end of the pulse tube; and a phase control device connected at the second end of the pulse tube for controlling the flow dynamics in the pulse tube to provide cooling at the heat exchanger, thereby maintaining a negative temperature gradient between the first and second ends of the regenerator, wherein: the pulse tube comprises a wall having a porous portion for allowing a working gas to enter or leave the pulse tube directly through the porous portion, the porous portion being nearer to being parallel than perpendicular to the temperature gradient between the first and second ends of the regenerator.
2. 2. A cold head according to claim 1, wherein the porous portion is within 5 degrees of being parallel to the temperature gradient between the first and second ends of the regenerator.
3. 3. A cold head according to claim 1 or 2, wherein the pulse tube is cylindrical and the porous portion is within 5 degrees of being parallel to the axial direction of the pulse tube.
4. 4. A cold head according to any of the preceding claims, wherein the porous portion is a part of a shared wall that is shared between the pulse tube and one or more of the following: the regenerator, the heat exchanger, a flow distributer between the pulse tube and the phase control device.
5. 5. A cold head according to claim 4, wherein the regenerator coaxially surrounds the pulse tube and the porous portion is part of a shared wall that is a radially inner wall of the regenerator.
6. 6. A cold head according to any of the preceding claims, wherein the pulse tube is substantially cylindrical and the regenerator has a substantially annular cross-section.
7. 7. A cold head according to ally of the preceding claims, wherein the heat exchanger coaxially surrounds the pulse tube.
8. 8. A cold head according to any of the preceding claims, wherein the porous portion extends from the second end of the regenerator along a portion of the regenerator towards the first end of the regenerator, without reaching the first end of the regenerator.
9. 9. A cold head according to any of the preceding claims, wherein the porosity of thc porous portion of the wall decreases as a function of increasing separation from the second end of the regenerator.
10. 10. A cold head according to any of the preceding claims, wherein the phase control device is configured to control the flow dynamics using one or more fluid phase control components to control the flow of gas into and out of the second end of the pulse tube, the tluid phase control components being configured to operate without using any solid moving parts.
11. 11. A cold head according to any of the preceding claims, wherein the phase control device provides either or both of damping and inertia.
12. 12. A cold head according to any of the preceding claims, wherein the phase control device comprises a piston and cylinder in fluid communication with the second end of the pulse tube.
13. 13. A cold head according to any of the preceding claims, wherein the porous portion of the wall is formed from an electroformed sheet.
14. 14. A cold head according to claim 13, wherein the electrofonned sheet is sandwiched between layers of mesh that act to even out and/or straighten the flow of gas passing through the porous portion of the wall.
15. 15. A pulse tube cooler comprising a cold head according to any of the preceding claims and a compressor connected to the first end of the regenerator.
16. 16. A cold head or pulse tube cooler constructed and arranged to operate substantially as hcreinbcfore described with reference to and/or as illustrated in the accompanying drawings.