JP7024962B2 - Cooler - Google Patents

Description

The present invention relates to a cooler that cools a semiconductor element by a water flow.

In power semiconductor modules and the like used as switching devices for power conversion applications, coolers and the like for heat dissipation are used in order to suppress adverse effects due to heat generated from circuits.

For example, in the liquid-cooled cooler of Patent Document 1 below, the coolant flowing into the coolant passage cools the heat generating element via the heat radiation fins and the heat sink base member. A large number of heat radiation fins arranged vertically and horizontally are provided on the first surface portion of the heat sink base member, and the length direction of the heat radiation fins is a predetermined angle θ with respect to the direction orthogonal to the first surface portion. It is inclined toward the downstream side in the distribution direction (paragraphs 0019, 0020, FIG. 2).

Japanese Unexamined Patent Publication No. 2016-22555

The heat dissipation fin structure of Patent Document 1 has improved heat dissipation performance as compared with the conventional straight fin structure. However, depending on the structure and arrangement of the heat radiation fins, the pressure loss of the fluid (cooling liquid such as water) becomes large, and there is a problem that, for example, the motor output of the fan must be increased in order to make the fluid flow.

In view of such problems, an object of the present invention is to provide a cooler having a heat radiation fin structure having low pressure loss and high cooling efficiency.

In order to achieve the above object, in the cooler of the present invention, a plurality of plate-shaped fins are provided vertically on both the top plate and the bottom plate and in parallel with each other, and a semiconductor device is provided by a fluid flowing between the plurality of plate-shaped fins. The flat plate portion of the plate-shaped fin is inclined in a direction away from the flat plate portion from the upstream side to the downstream side in the flow direction of the fluid, and is arranged in a staggered manner on the flat plate portion. The cross-sectional portion having a protruding portion and formed in parallel with the flat plate portion of the protruding portion gradually widens from the upstream side to the downstream side, gradually narrows after reaching the maximum width A. It is characterized by having a shape that makes it.

In the cooler of the present invention, a plurality of plate-shaped fins are provided perpendicular to the bottom plate and the top plate of the cooler and parallel to each other, and a semiconductor device is provided by a fluid (for example, water) flowing between the plurality of plate-shaped fins. To cool.

The plate-shaped fins have protrusions arranged in a staggered manner on the flat plate portion, and the protrusions have a shape inclined from the upstream side to the downstream side in the flow direction of the fluid. Further, the cross-sectional portion of the protruding portion formed in parallel with the flat plate portion has a shape that gradually widens from the upstream side to the downstream side, reaches the maximum width A, and then gradually narrows. By forming the protrusion in such a shape, the fluid flows through the protrusion from the upstream side to the downstream side, and proceeds between the plate-shaped fins while keeping the pressure loss low. This makes it possible to obtain a cooler having a low fluid pressure loss and high cooling efficiency.

In the cooler of the present invention, the ratio C / A of the maximum width A of the cross section to the length C from the downstream end of the cross section to the straight line forming the maximum width A is 1. It is preferably a value of 5 or more and 10.0 or less.

Since the shape of the cross section of the plate-shaped fin affects the flow method and velocity of the fluid, it affects the temperature (heat dissipation) and the pressure loss of the fluid. In particular, the cooling efficiency of the cooler can be improved by setting the ratio C / A of the maximum width A and the length C, which is an element forming the cross-sectional portion, to a value of 1.5 or more and 10.0 or less. ..

Further, in the cooler of the present invention, the angle of the protruding portion with respect to the axis perpendicular to the flat plate portion is preferably 20.0 degrees or more and 85.0 degrees or less.

The angle of the protrusion of the plate fin also affects the fluid flow. When the angle is smaller than 20.0 degrees (close to perpendicular to the flat plate portion), a portion where the flow velocity is high and a portion where the flow velocity is slow occur, and the cooling effect is not enhanced. Further, when the angle is larger than 85.0 degrees (close to horizontal with respect to the flat plate portion), the flow velocity around the protruding portion rapidly decreases. Therefore, the cooling efficiency of the cooler can be improved by setting the angle to be 20.0 degrees or more and 85.0 degrees or less.

Further, in the cooler of the present invention, the ratio W / H of the height H of the protruding portion with respect to the flat plate portion and the thickness W of the flat plate portion is a numerical value of 0.6 or more and 1.2 or less. Is preferable.

The height H of the protruding portion of the plate-shaped fin and the thickness W of the flat plate portion also affect the fluid flow. It is known that the height H and the thickness W are shape parameters having a large influence on the temperature (heat dissipation), and the ratio W / H of the height H and the thickness W is 0.6 or more. By setting the value to 1.2 or less, the cooling efficiency of the cooler can be improved.

Further, in the cooler of the present invention, the width E of the overlap of the cross section and the other cross sections adjacent to the cross section in the flow direction of the fluid, and the downstream side of the cross section. The ratio E / C with the length C from the end of the to the straight line forming the maximum width A is preferably 0 or more and 0.5 or less.

Since the length C and the overlapping width E are related to the distance between adjacent cross sections, they affect the flow of fluid. In particular, by setting the ratio E / C of the length C to the overlapping width E to be 0 or more and 0.5 or less, the cooling efficiency of the cooler can be improved.

Further, in the cooler of the present invention, the cross-sectional portion has a rectangle having the maximum width A as the long side and the long side on the downstream side of the rectangle as the base, from the bottom to the end on the downstream side. A first triangle having a length as a first height and a second having a long side on the upstream side of the rectangle as a base and a length from the base to the end on the upstream side as a second height. It is a hexagonal shape composed of the triangles of the above, and it is preferable that the second height is lower than the first height.

The cross section of the cooling fin has a rectangle having a maximum width A as a long side, a first triangle having a maximum width A as a base and a height of the first height, and a maximum width A as a base and a height. The cooling efficiency is good when the hexagonal shape (kite shape) including the second triangle having the second height is used. Further, the second height is lower than the first height, that is, the hexagonal shape is long toward the downstream side in the fluid flow direction, so that the cooling effect can be enhanced while suppressing the pressure loss.

It is sectional drawing of the semiconductor module which concerns on embodiment of this invention. Top view of the cooler constituting the semiconductor module from the top plate side (without top plate). Perspective view of the plate-shaped fins in which the cooler is arranged. (A) Surface view of plate-shaped fins. (B) The figure explaining the cross section part (kites type) of a plate-shaped fin. (C) The figure explaining the cross section part (quadrilateral) of a plate-shaped fin. (A) The figure which shows the simulation result of the temperature and pressure loss of a cooler. (B) The figure which shows the numerical ratio when the protrusion angle is 79.2 degrees. (C) A graph plotting simulation results. (A) A graph plotting the relationship between the ratio C / A and temperature. (B) A graph plotting the relationship between the ratio C / A and the pressure loss. (A) A graph plotting the relationship between the ratio W / H and temperature. (B) A graph plotting the relationship between the ratio W / H and the pressure loss. The figure which shows the insulation device for measuring the thermal resistance. The figure which shows the result of the thermal resistance between the conventional straight fin and the plate-shaped fin of this invention.

Hereinafter, embodiments of the semiconductor module of the present invention will be described with reference to the drawings.

FIG. 1 shows a cross-sectional view of a semiconductor module 100 according to an embodiment of the present invention. The semiconductor module 100 is mainly composed of two semiconductor elements 1a and 1b, a wiring board 3, a laminated board 5, a cooler 7, a case 9, and the like. As shown in the figure, the semiconductor elements 1a and 1b, the wiring board 3 and the laminated board 5 are housed in the case 9 and molded with the resin 8. Further, a cooler 7 for cooling the semiconductor elements 1a and 1b is arranged on the lower surface of the case 9.

The semiconductor elements 1a and 1b are, for example, an IGBT (Insulated Gate Bipolar Transistor) or a power MOSFET (Metal Oxide Semiconductor Field Effect Transistor). These transistors may be RB-IGBT (Reverse Blocking-IGBT) or RC-IGBT (Reverse Conducting-IGBT) formed in one semiconductor element in the vertical direction.

The wiring board 3 is arranged on the upper surface side of the semiconductor elements 1a and 1b. The wiring board 3 has a structure in which both sides of an insulating substrate are covered with metal foil, and the metal foil on the lower surface side is formed so as to face the semiconductor elements 1a and 1b. The insulating substrate is preferably a material having a low dielectric constant and a high thermal conductivity, and for example, a resin insulating material containing a resin such as an epoxy resin or a ceramic such as Si 3N ₄ , AlN _, Al ₂ O ₃ can be used. .. Further, the metal foil is preferably a material having low electrical resistance and high thermal conductivity, and for example, Cu can be used.

One end of the pin 4 is joined to the upper surface side of the semiconductor elements 1a and 1b by the metal joining member 2a, and the other end is used for connection with the wiring board 3. For the pin 4, a metal having a low electric resistance and a high thermal conductivity, for example, Cu can be used. The metal joining member 2a may be a member having metal fine particles such as solder and silver.

As shown in the figure, it is preferable that a plurality of pins 4 are arranged for each of the semiconductor elements 1a and 1b. By doing so, it is possible to reduce the electrical resistance and improve the heat conduction performance.

The laminated substrate 5 is composed of an insulating substrate 52, a first conductive plate 51 formed on the upper surface side thereof, and a second conductive plate 53 formed on the lower surface side thereof. As the insulating substrate 52, a material having excellent electrical insulation and thermal conductivity can be used. Examples of the material of the insulating substrate 52 include _{Al2O3 ,} AlN, SiN and the like. In particular, for high withstand voltage applications, materials having both electrical insulation and thermal conductivity are preferable, and AlN and SiN can be used, but the material is not limited thereto. For the first conductive plate 51 and the second conductive plate 53, metal materials such as Cu and Al, which are excellent in conductivity and processability, can be used. In the present specification, the second conductive plate 53 made of Cu may be referred to as a back surface copper foil. For the purpose of rust prevention or the like, Cu or Al that has been subjected to a treatment such as Ni plating may be used. Examples of the method of arranging the conductive plates 51 and 53 on the surface of the insulating substrate 52 include a direct bonding method (Direct Copper Bonding method) and a brazing material bonding method (Active Metal Brazing method).

Further, the laminated substrate 5 is arranged on the lower surface side of the semiconductor elements 1a and 1b. The laminated substrate 5 has a structure in which both sides of the insulating substrate 52 are covered with a metal foil such as Cu, and is electrically separated from the metal foil due to the insulating property of the insulating substrate 52. The peripheral edge of the insulating substrate 52 preferably protrudes outward from the peripheral edges of the conductive plates 51 and 53.

The lower surface side of the semiconductor elements 1a and 1b and the first conductive plate 51 on the upper surface side of the laminated substrate 5 are electrically and thermally bonded by a metal bonding member 2b. The metal foils on the upper surface side and the lower surface side of the laminated substrate 5 are electrically separated, but the heat conduction between them is good. Further, the second conductive plate 53 on the lower surface side of the laminated substrate 5 and the outer wall (top plate) of the cooler 7 are joined by a metal joining member 2c. The metal joining members 2b and 2c may be members having metal fine particles such as solder and silver.

Next, the details of the cooler according to the embodiment of the present invention will be described with reference to FIGS. 2 to 4.

FIG. 2 is a top view (without top plate) of the cooler 7 as viewed from the top plate 7a (see FIG. 1) side. The cooler 7 is composed of a top plate 7a, a bottom plate 7b, a side frame 7c, and a plate-shaped fin 7d, and the material is a metal having high thermal conductivity such as Al and Cu. The shape of the side frame 7c is not limited to an octagon, but it has an inlet 7e and an outlet 7f for the fluid, and the direction in which the fluid (cooling liquid such as water) flows in and the longitudinal direction of the plurality of plate-shaped fins 7d are different. They are arranged so as to be parallel. The cooler may have a U-shape with the fluid inlet and outlet on the same side.

Next, FIG. 3 is a perspective view showing how some of the plate-shaped fins 7d1 to 7d3 constituting the cooler 7 are arranged, and the plate-shaped fins 7d of a part (three pieces) of FIG. 1 are shown. It is in the taken out state.

As can be seen from the plate-shaped fins 7d3, each plate-shaped fin 7d is composed of a flat plate portion 10 and a protruding portion 11. The flat plate portion 10 is a so-called straight fin, which alone plays the same role as a conventional heat sink. The flat plate portion 10 has two parallel surfaces, and a protruding portion 11 is formed so as to project from one surface (surface) toward the opposite surface of the adjacent flat plate portion 10. The protruding portion 11 may be formed so as to connect the adjacent flat plate portions 10, and in that case, the protruding portion 11 may be arranged so as to penetrate a plurality of parallel flat plate portions 10. Further, the cooler 7 of the present invention employs a plate-shaped fin 7d in which a large number of protrusions 11 are arranged in a staggered manner on the flat plate portion 10, and a large number of the plate-shaped fins 7d are arranged so that a fluid is introduced between the plate-shaped fins 7d. I try to shed it. The fluid flows in from the front side in the figure of the plate-shaped fins 7d1 to 7d3 and flows out to the back side (direction of the arrow). The direction of the arrow through which the fluid flows (the inflow direction of the fluid) is the X-axis direction.

The protruding portion 11 is an oblique prism inclined in a direction away from the flat plate portion 10 from the upstream side to the downstream side of the water flow with respect to the flat plate portion 10. In other words, the protruding portion 11 is inclined toward the downstream side at a predetermined angle φ (protruding angle) with respect to the axis Z perpendicular to the flat plate portion 10. The protrusion angle is preferably 20 to 85 degrees, more preferably 40 to 80 degrees. Although the details will be described later, since the temperature of the fluid is low in this range, the cooling capacity is excellent and the pressure loss can be suppressed to a low level.

Further, the tip portion of the protruding portion 11 is formed in parallel with the flat plate portion 10, and this tip portion is referred to as a cross-sectional portion 11a. The cross-sectional shape cut on the plane parallel to the flat plate portion 10 is also the same as the cross-sectional portion 11a. The cross-sectional portion 11a has a shape that gradually widens from the upstream side to the downstream side in the flow direction of the fluid, reaches the maximum width portion (maximum width A in FIG. 4B), and then gradually narrows. There is.

The minimum length of the upstream end and the maximum width is preferably shorter than the minimum length of the downstream end and the maximum width. The minimum length of the upstream end and the maximum width is the length from the upstream end to the maximum width side in the case of the maximum width side closest to the upstream end. In other words, it is the length from the upstream end to the straight line forming the maximum width A, and corresponds to the width B in FIG. 4 (b).

The same applies to the downstream side, and in the case of the maximum width side closest to the downstream end, the length from the downstream end to the maximum width side is the downstream end and the maximum width. The minimum length of the part. In other words, it is the length from the downstream end to the straight line forming the maximum width A, and corresponds to the width C in FIG. 4 (b). That is, the cross-sectional portion 11a has a shape extending to the downstream side. The cross-sectional portion 11a preferably has an X-axis symmetric shape, and more specifically, it has a hexagonal shape and a quadrilateral shape as described later.

The cross-sectional portion 11a is preferably in contact with the flat plate portion (back surface) of the adjacent plate-shaped fins. The back surface of the flat plate portion is a surface on the side having no protruding portion. The intersection of the vertical axis and the horizontal axis of the cross-sectional portion 11a is the axis (see FIGS. 4 (b) and 4 (c)), and the angle φ is the origin and the axis in the plane including the axis X and the axis Z. It is an angle formed by a straight line connecting the hearts. Further, the straight line connecting the tip end (upstream end) and the rear end (downstream end) of the cross-sectional portion 11a with respect to the fluid inflow direction (X-axis direction) should be parallel to the X-axis. preferable. Further, the inflow direction of the fluid is preferably parallel to the plane including the axis X and the axis Z.

The thickness W of the flat plate portion 10 of the plate-shaped fin 7d is preferably 0.5 (mm) or more and 1.5 (mm) or less. When the thickness W is 0.3 (mm) or less, heat conduction to the fluid is insufficient, and when the thickness is 1.75 (mm) or more, the contact area with the fluid becomes small, and again. Insufficient heat conduction. The optimum thickness W of the flat plate portion 10 is 0.75 (mm) or more and 1.25 (mm) or less.

Further, the height H of the protruding portion 11 with respect to the flat plate portion 10 of the plate-shaped fin 7d is preferably 0.9 (mm) or more, 1.5 (mm) or less, and more preferably 1.05 (mm) or more, 1 It is .35 (mm) or less. The height H is the same as the distance between the plate-shaped fins 7d1 and 7d3 in FIG. Details will be described with reference to FIG. 7, but as a condition for improving the cooling efficiency, the ratio of thickness W / height H is preferably a numerical value of 0.6 or more and 1.2 or less.

The cross-sectional portion 11a of the central plate-shaped fin 7d2 is fixed by coming into contact with the flat plate portion 10 of the adjacent plate-shaped fin 7d3. Further, the protruding portion 11 is arranged so as to penetrate the adjacent plate-shaped fins d7.

Next, FIG. 4 is a surface view of one plate-shaped fin 7d. As shown in FIG. 4A, the protruding portions 11 are formed in a staggered manner on the upper surface of the flat plate portion 10. In addition, the staggered shape means an arrangement in which two rows are arranged alternately (zigza).

Here, the cross-sectional portion 11a may have a hexagonal shape (kite shape) as shown in FIG. 4 (b). Since the direction of the water flow is as shown in FIG. 3, the width A of the cross-sectional portion 11a is the actual vertical direction, and the vertical width A is 0.5 (mm) or more and 2.5 (mm) or less. .. The vertical width A corresponds to the "maximum width" of the present invention.

Here, the distance between the center O of the portion having the maximum width A of the cross-sectional portion 11a and the end portion on the upstream side is the width B, and the distance between the center O and the end portion on the downstream side is the width C. As shown in FIG. 4B, the center O connects the intermediate points of the portions having the maximum width A on both sides when the portion having the maximum width A continues for a predetermined length in the X-axis direction. It shall be the intersection of the line and the line connecting the upstream end and the downstream end. Further, as shown in FIG. 4 (c) described later, when the portion having the maximum width A has vertices on both sides, the line connecting the vertices and the upstream end and the downstream side It is the intersection of the lines connecting the ends. In this case, it is preferable that the lateral rear width C is longer than the lateral front width B.

Further, the hexagonal shape has a rectangle having a vertical width A as a long side and a long side on the downstream side (upper side of the figure) of the rectangle as a base, and from the base to the end on the downstream side (one of the lateral rear width C). A first triangle having a first height (part) and a long side on the upstream side (lower side of the figure) of the rectangle as a base, and from the base to the end on the upstream side (a part of the lateral front width B). ) Can be said to consist of a second triangle having a second height. That is, the rectangular portion is a portion where the vertical width A is maintained. The hexagonal shape and rectangular corners described above include rounded shapes.

Further, it is preferable that the second height is lower than the first height. This corresponds to the relationship of lateral rear width C> lateral front width B in FIG. 4 (b). Further, the length of the short side of the rectangle having the vertical width A as the long side is preferably shorter than the horizontal front width B. Specifically, the length of the short side is preferably 0.1 mm or more and 0.3 mm or less. By providing the rectangular portion in the cross-sectional portion 11a in this way, the fluid can flow smoothly.

The cross-sectional portion 11a may be a quadrilateral as shown in FIG. 4 (c). In this case, the short diagonal of the quadrilateral is the vertical width A. This vertical width A corresponds to the "maximum width". Further, this quadrilateral has a first triangle having a vertical width A as a base and a first height from the base to the end on the downstream side (horizontal rear width C), and an end on the upstream side from the base. It consists of a second triangle whose second height is up to (horizontal front width B). As in the case of the hexagonal shape described above, it is preferable that the second height is lower than the first height. That is, also in FIG. 4C, there is a relationship of lateral rear width C> lateral front width B. It is preferable that the apex of the quadrangle in the Y-axis direction is rounded. The radius of curvature R of the apex is preferably A / 6 mm or more and A mm or less. The rounded vertices in that range allow the fluid to flow smoothly.

The lateral front width B is 0.5 mm or more and 2.5 mm or less, and the lateral rear width C is 2.0 mm or more and 6.0 mm or less. Details will be described with reference to FIG. 6, but as a condition for improving the cooling efficiency, the ratio of the lateral rear width C / vertical width A is preferably a numerical value between 1.5 and more and 10.0 or less.

Further, as shown in FIG. 4A, one cross-section portion 11a2 and the adjacent cross-section portion 11a3 (the same applies to the cross-section portion 11a1) overlap each other by the width E in the inflow direction of the fluid. More specifically, the width E is the overlap width in the X-axis direction between the upstream end portion of the cross-sectional portion 11a2 and the downstream end portion of the adjacent cross-sectional portion 11a3. The overlap width E is preferably 0 (mm) or more and 1.0 (mm) or less.

Further, the distance F (flow path distance) in the inclined plane direction between one cross-sectional portion 11a1 and the adjacent cross-sectional portion 11a2 is a numerical value depending on the lateral rear width C and the overlapping width E, but is 0.9 (mm). ) Or more, preferably 1.5 (mm) or less. It is necessary to secure at least 0.9 (mm) or more for the flow path spacing F so that clogging by particles does not occur, and more preferably 1.05 (mm) or more, 1.35 from the result of cooling efficiency. It is (mm) or less.

As the material of the plate-shaped fin 7d of the present invention, Al, Al alloys, Cu, Cu alloys and the like having excellent thermal conductivity and excellent workability are used. As described above, the plate-shaped fin 7d has a protrusion 11 formed on the surface of the flat plate portion 10. The shape of FIG. 3 is also formed by, for example, providing a predetermined hole in the flat plate portion by a method such as pressing and inserting a rod-shaped protrusion into the hole. The same shape can also be formed by a 3D printer.

Next, with reference to FIG. 5, a simulation result in which the temperature and pressure loss are measured by changing the shape parameters constituting the plate-shaped fin 7d of the present invention by analysis software will be described.

As shape parameters that influence the temperature T and pressure loss PL of the plate-shaped fin 7d, the vertical width A, the horizontal front width B and the horizontal rear width _C of the cross-sectional portion 11a, the thickness W of the flat plate portion 10, and the protrusion with respect to the flat plate portion 10. There are a height H of the portion 11, an overlap width E of the adjacent cross-sectional portions 11a in the direction of the water flow, a flow path spacing F in the inclined surface direction of the adjacent cross-sectional portions 11a, a protrusion angle φ, and the like.

As a result of repeated simulations, the inventors have found that the vertical width A, the horizontal rear width C, the thickness W, and the height H of the cross-sectional portion 11a and the flat plate portion 10 of the protruding portion 11 affect the temperature T (cooling efficiency). I learned that is a large shape parameter. Further, it was found that the vertical width A and the horizontal rear width _C are shape parameters having a large influence on the pressure loss PL. The cross-sectional portion 11a has a shape symmetrical with respect to the X-axis.

More specifically, in the protruding portion 11, the ratio (C / A) of the lateral rear width _C to the vertical width A of the cross-sectional portion 11a and the protruding angle φ have a large influence on the temperature T and the pressure loss PL. It was a parameter. Further, the flow path width of the distance between the flat plate portions 10 is the height H of the protruding portion 11 with respect to the flat plate portion 10 of the plate-shaped fin 7d, and the thickness W of the flat plate of the flat plate portion 10 and the height H of the protruding portion 11. The ratio ( _W / H) was also an effective shape parameter for controlling the temperature T and the pressure drop PL.

This time, the vertical width A of the shape parameters is 1.48 (mm), the horizontal front width B is 2.42 (mm), the horizontal rear width C is 2.90 (mm), and the thickness W is 0.85 (mm). The height H was 0.92 (mm), the overlapping width E was 0.24 (mm), and the flow path spacing F was 1.31 (mm) (see FIG. 5A). The cross-sectional portion 11a has a hexagonal shape, and the short side of the rectangle having the vertical width A as the long side is 0.1 mm. Then, the temperature T and the pressure loss _PL were measured by changing only the angle φ of the protrusion 11.

Here, as the temperature T, the temperature near the center of the cooler 7 was measured using a heat insulating device 12 including a heater 14 described later (see FIG. 8). The output of the heater 14 was set to 1,110 (W), pure water was used as the coolant, and the inflow amount was set to 4.0 (L / min). The pressure loss PL was evaluated as the difference (pressure difference) between the pressure at the inlet _7e of the cooler 7 and the pressure at the outlet 7f under the same inflow condition.

When the protrusion angle φ was 0 degrees, that is, when the protrusion 11 was upright with respect to the flat plate portion 10, the temperature T was 69.44 (° C.) and the pressure loss PL was _2679.21 (kPa). The lower the temperature T, the higher the heat dissipation and the higher the cooling efficiency. Further, the smaller the pressure loss, the more the fluid flows without loss.

Next, when the protrusion angle φ is 10.0 degrees, the temperature T is 69.28 (° C.), the pressure loss PL is _2637.70 (kPa), and the protrusion angle φ is 20.0 degrees. The temperature T was 69.11 (° C.) and the pressure loss PL was _2551.20 (kPa). As described above, it was found that as the protrusion angle φ was increased, the temperature T decreased and the pressure loss _PL also decreased.

Further, when the protrusion angle φ is 60.0 degrees, the temperature T is 68.40 (° C.), the pressure loss PL is _2080.00 (kPa), and when the protrusion angle φ is 85.0 degrees, the temperature. T was _69.18 (° C.) and pressure loss PL was 910.00 (kPa). From this result, it was found that the minimum value of the temperature T is that the protrusion angle φ is between 60.0 degrees and _85.0 degrees, and the pressure loss PL becomes smaller as the protrusion angle φ becomes larger.

As a result of investigating the protrusion angle φ at which the temperature T is the minimum value, the temperature T is 68.30 (° C.) under the best condition when it is 79.20 degrees, and the pressure loss PL at that time is _1429.91 . It was (kPa). Therefore, from a series of simulation results, 79.20 degrees was obtained as the optimum value of the protrusion angle φ, and it was confirmed that the cooling efficiency was the highest at this time (see FIG. 5 (c)). Based on the above, from the results of the temperature _T and the pressure loss PL, the protrusion angle φ of the protrusion 11 is preferably an angle of 20 to 85 degrees, more preferably an angle of 40 to 80 degrees.

Further, as shown in FIG. 5B, when the ratio of each numerical value under the above-mentioned optimum conditions is calculated, the ratio B / A of the horizontal front width B to the vertical width A is about 1.635 and the vertical width A. The ratio C / A of the lateral rear width C is about 1.959, the ratio W / H of the thickness W to the height H is about 0.923, and the ratio E / C of the overlapping width E to the lateral rear width C is about 0. It was 082. By bringing each ratio close to the value of the optimum condition, the temperature T and the pressure loss _PL can be reduced. Moreover, this tendency did not change even if the shape parameters other than the protrusion angle φ were changed.

Here, FIG. 6 describes a simulation result in which the temperature T and the pressure loss PL are measured by changing the ratio _C / A of the vertical width A to the horizontal rear width C. In the measurement, the shape parameters other than the vertical width A and the horizontal rear width C were fixed.

FIG. 6A shows the relationship between the ratio C / A and the temperature T, and the result was obtained that the temperature T decreased as the value of the ratio C / A was decreased. In the example of FIG. 5A, the temperature T is as low as about 70 (° C.) when the ratio C / A is set to a value of about 2.0, as the ratio C / A was about 1.959. It became a value. Further, when the ratio C / A was made larger than 10.0, the temperature T increased to 77 (° C.) or higher, so 10.0 was set as the maximum value of the ratio C / A.

Further, FIG. 6B shows the relationship between the ratio _C / A and the pressure loss PL. Here, it was obtained that the pressure loss PL decreased as the value of the ratio _C / A was increased. When the ratio C / A was set to a value of 1.5 or less, the pressure loss PL exceeded 2,000 (kPa), so 1.5 was set as the minimum value of the ratio _C / A. From the above, the ratio _C / A is preferably 1.5 or more and 10.0 or less, and more preferably 4.0 or more and 8.0 or less in terms of the temperature T and the pressure loss PL.

Next, FIG. 7 will explain the simulation results of measuring the temperature T and the pressure loss PL by changing the ratio _W / H of the thickness W to the height H. In the measurement, the shape parameters other than the height H and the thickness W were fixed.

FIG. 7A shows the relationship between the ratio W / H and the temperature T, and the result was obtained that the temperature T decreased as the value of the ratio W / H was increased. When the ratio W / H was set to a value of 0.6 or less, the temperature T increased to 74 (° C.) or more, so 0.6 was set as the minimum value of the ratio W / H.

Further, FIG. 7B shows the relationship between the ratio _W / H and the pressure loss PL. Here, the result was obtained that the pressure loss PL decreased as the value of the ratio _W / H was lowered. When the ratio _W / H was set to a value of 1.2 or more, the pressure loss PL suddenly increased and exceeded 1,000 (kPa). Therefore, 1.2 was set as the maximum value of the ratio W / H. .. Based on the above, the ratio _W / H is preferably 0.6 or more and 1.2 or less, and more preferably 0.8 or more and 1.0 or less in terms of the temperature T and the pressure loss PL. Further, the ratio E / C of the overlapping width E to the lateral rear width C is preferably a numerical value of 0 or more and 0.5 or less. In this range, the temperature T and the pressure loss _PL do not deteriorate. When the overlap width E is larger than 0.5, the flow path becomes narrow, which is not preferable.

Finally, with reference to FIGS. 8 and 9, the results of comparing the thermal resistance of the plate-shaped fin of the present invention and the conventional straight fin are shown.

In order to measure the two types of thermal resistance, the heat insulating device 12 shown in FIG. 8 was used. FIG. 8 shows a state in which the cooler 7'composed of straight fins 7d'is housed inside the frame 13 of the insulator, and the heater 14 is placed on the upper surface of the cooler 7'. The heater 14 is adhered to the cooler 7'with a thermal grease 15 having a small thermal resistance. The cooler 7'is a replacement of only the plate-shaped fins 7d of the above-mentioned cooler 7, and the top plate 7a, the bottom plate 7b, the side frame 7c, the inflow port 7e, and the outflow port 7f have different shapes and materials. It is the same.

In addition, the inventors prototyped a metal (aluminum) plate-shaped fin 7d using a 3D printer for measurement. At this time, the various shape parameters of the plate-shaped fin 7d were the conditions described in FIG. When measuring the thermal resistance of the plate-shaped fin 7d of the present invention, the cooler 7 is housed inside the heat insulating device 12.

The thermal resistance Rth of the plate-shaped fin 7d or the straight fin _7d'is given by the following equation 1 when the temperature of the heater 14 is _Th , the water temperature at the outlet 7f is _Tw , and the electric power of the heater 14 is P. Be done.
R _th = ( _Th -T _w ) / P ... (Equation 1)
The fluid (water) flowed in from the inflow port 7e at a flow rate of 1.5 (L / min).

This time, the thermal resistance of the straight fin 7d'of the cooler 7'was measured using a flat plate portion having a thickness W'of 0.9 (mm). As a result, as shown in FIG. 9, the thermal resistance was 0.239 (° C./W). On the other hand, the thermal resistance of the cooler 7 was 0.197 (° C./W), and the thermal resistance was successfully reduced by about 18%.

Although the refrigerator made of the plate-shaped fins 7d has been described above, the present invention is not limited to the embodiments described so far. For example, the number of plate-shaped fins 7d inside the cooler 7 is arbitrary, but is about the same as the number of conventional straight fins. The material of the plate-shaped fin 7d is not limited to Al and Cu, and a material having high thermal conductivity can be applied.

Further, the protruding portion 11 of the plate-shaped fin 7d is an oblique prism whose rectangular parallelepiped is tilted in the diagonal direction (see FIG. 4A), but may have a shape that gradually becomes thicker in the protruding direction. .. The shape of the cross-sectional portion 11a of the protruding portion 11 is preferably a kite-shaped hexagon or a quadrilateral long in the direction in which the fluid flows (see FIGS. 4 (b) and 4 (c)), but even if it is a rhombus or a square. good.

1a, 1b Semiconductor element 2a-2c Metal bonding member 3 Wiring board 4 pin 5 Laminated board 7,7'cooler 7a Top plate 7b Bottom plate 7c Side frame 7d, 7d1-7d3 Plate-shaped fin 7d'Straight fin 7e Inflow port 7f Flow Outlet 8 Resin 9 Case 10 Flat plate 11 Protruding 11a, 11a1 to 11a3 Cross section 12 Insulation device 13 Frame 14 Heater 15 Thermal grease 51 1st conductive board 52 Insulation board 53 2nd conductive board 100 Semiconductor module

Claims (6)

In a cooler in which a plurality of plate-shaped fins are provided vertically on both the top plate and the bottom plate and parallel to each other, and the semiconductor device is cooled by a fluid flowing between the plurality of plate-shaped fins.
The plate-shaped fin has a flat plate portion and a protruding portion, and the protruding portion is inclined in a direction away from the flat plate portion from the upstream side to the downstream side in the flow direction of the fluid, and is formed on the back surface of the adjacent flat plate portion. It extends in contact with each other and is arranged in a staggered pattern on the flat plate portion.
The cross-sectional portion of the protruding portion formed in parallel with the flat plate portion has a shape that gradually widens from the upstream side to the downstream side, reaches the maximum width A, and then gradually narrows. A cooler that features that. The ratio C / A of the maximum width A of the cross section to the length C from the downstream end of the cross section to the straight line forming the maximum width A is 1.5 or more and 10.0 or less. The cooler according to claim 1, wherein the value is. The cooler according to claim 1 or 2, wherein the angle of the protruding portion with respect to the axis perpendicular to the flat plate portion is an angle of 20.0 degrees or more and 85.0 degrees or less. From claim 1, the ratio W / H of the height H of the protruding portion with respect to the flat plate portion and the thickness W of the flat plate portion is a numerical value of 0.6 or more and 1.2 or less. The cooler according to any one of 3. The width E of the overlap of the cross section and the other cross sections adjacent to the cross section in the flow direction of the fluid, and the maximum width A from the downstream end portion of the cross section. The cooler according to any one of claims 1 to 4, wherein the ratio E / C to the length C to form a straight line is a numerical value of 0 or more and 0.5 or less. The cross section has a rectangle having the maximum width A as the long side and a long side on the downstream side of the rectangle as the base, and the length from the base to the end on the downstream side as the first height. It is a hexagonal shape consisting of a first triangle and a second triangle having the long side on the upstream side of the rectangle as the base and the length from the base to the end on the upstream side as the second height. ,
The cooler according to any one of claims 1 to 5, wherein the second height is lower than the first height.

Images