Heat sink element for power module, power module and method for producing such heat sink element

Abstract

The invention relates to a heat sink element (1) for an electric power module (10), comprising: a base body having a first end face (S), which faces a surface to be cooled in the assembled state, and a second end face (S2), which lies opposite the first end face (S1); and a cooling channel (30, 30 ') which is embedded in the base body between the first end face (S1) and the second end face (S2), the cooling channel (30, 30') having a feed section (31), a deflection section (32) and a discharge section (33), the cooling channel (30, 30 ') being designed to form an overall flow course in order to convey a cooling medium in the feed section (31) towards the first end face (S1), and the cooling medium in the discharge section (33) towards the second end face (S2). The conveying section (31) is transferred in the deflection section (32) into the lead-out section (33) and is transported in the lead-out section (33) towards the second end face (S2), at least one transverse flow section (50) connecting the conveying section (31) to the lead-out section (33) being formed between the deflection section (32) and the second end face (S2).

Description

Technical Field

[0001] The present invention relates to a cooling element for a power module and a method for manufacturing such a cooling element. Background Technology

[0002] Power modules are widely known in the prior art and provide the feasibility of providing circuitry or electrical connections between electronic or electrical components via corresponding connection surfaces and printed wires. It has proven particularly advantageous to use ceramic as the insulating element, with the various metal segments of the device's metallized portion disposed on said insulating element. This ceramic has relatively high insulating strength and is particularly temperature resistant. With the increasing demands on such power modules and the resulting heat dissipation, a corresponding cooling device that ensures efficient heat removal is becoming increasingly challenging.

[0003] A classic method for providing cooling elements is to bond a cooling structure with cooling fins to the rear metallization of a cermet substrate. The cooling medium, particularly a coolant, is guided through the arrangement of the corresponding cooling fins to ensure heat dissipation. Bonding via the rear metallization is necessary because it is crucial for the bonding method from the metal layer to the ceramic element. Otherwise, especially when the metal layer is bonded to the ceramic element from one side, the cermet substrate will flex due to the different coefficients of thermal expansion and the resulting thermomechanical stresses at temperature changes, such as those during cooling after bonding.

[0004] As an improvement to this relatively simple fin structure, the teachings of DE 10 2013 109 246 B4 propose embedding multiple microchannels into the cooling body structure for planar cooling of the power module. The substantially annular extension of the cooling channels targets the cooling medium as close as possible to the surface to be cooled within the power module, while simultaneously redirecting the cooling medium away from the corresponding cooling surface. Preferably, the cooling medium is supplied to multiple such cooling channels via a common distribution structure.

[0005] Here, the annular cooling channels are specifically configured as microchannels, which are identically designed and arranged parallel to each other. This advantageously allows the cooling medium to be directed and targeted to the surface to be cooled, and in particular, ensures the most uniform possible distribution of cooling on the cold side of the cooling element.

[0006] Furthermore, it is common in the prior art to provide a metal-ceramic substrate, which serves as a power module, to a customer, who then bonds a cooling element to the metal-ceramic substrate, particularly to the rear metallized portion of the metal-ceramic substrate. It is generally not feasible to fit the cooling element to the metal-ceramic substrate as efficiently as possible by separating these two steps, since the corresponding manufacturing steps are separate from each other. Summary of the Invention

[0007] Therefore, the present invention, based on the prior art, aims to simplify the manufacturing method of a power module with cooling capability, and in particular to improve the cooling efficiency and compactness of the power module.

[0008] The present invention achieves the stated objective by means of a cooling element according to claim 1, a power module according to claim 9, and a method for manufacturing such a cooling element according to claim 10. Further description, drawings, and dependent claims encompass other non-limiting embodiments of the invention.

[0009] According to one aspect of the present invention, a cooling element for an electric power module is provided, the cooling element comprising:

[0010] - A base having a first end side and a second end side opposite to the first end side, the first end side facing the surface to be cooled in the assembled state, and

[0011] - A cooling channel, which is embedded in the substrate between a first end side and a second end side.

[0012] The cooling channel includes a conveying section, a deflection section, and an outlet section. The cooling channel is designed to form an overall flow direction, so as to convey the cooling medium toward a first end in the conveying section, transfer it to the outlet section in the deflection section, and convey it toward a second end in the outlet section. At least one transverse flow section is formed between the deflection section and the second end to connect the conveying section and the outlet section.

[0013] Unlike what is known from the prior art, it is proposed here that a transverse flow section be constructed between the conveying section and the drawing section. This transverse flow section provides the additional feasibility of guiding the cooling medium from the conveying section to the drawing section, particularly without having to guide a corresponding proportion of the cooling medium via a deflection section. Preferably, the transverse flow section is configured such that less than one-third, preferably less than one-eighth, and preferably less than one-fifteenth of the flow volume guided through the deflection section is guided through the transverse flow section, particularly through one or all of the transverse flow sections. It has been surprisingly proven that, through this transverse flow section, the thermal resistance of the entire cooling element can be advantageously adjusted. In particular, this also reduces the number of layers used, thereby reducing the second thickness of the cooling element, which has a favorable effect on adjusting the mass distribution on the front and rear sides of the ceramic element to compensate for flexural forces. Especially in manufacturing, the configuration of the transverse flow cross-section means that plastic deformation during cooling during the bonding process is favorable, and this plastic deformation has a favorable effect on minimizing flexural deformation. Preferably, the device includes multiple, particularly stacked, lateral flow sections. More preferably, multiple lateral flow sections are provided, which are stacked vertically along a direction perpendicular to the main extension plane and / or arranged side-by-side in a direction parallel to the main extension plane.

[0014] Specifically, it is proposed that the cooling element comprises multiple individual cooling channels, each having a corresponding deflection section, conveying section, and outlet section, and particularly a lateral flow section. Here, the individual cooling channels are preferably arranged side-by-side along two linearly independent directions to form a two-dimensional arrangement of the deflection region. Preferably, the cooling channels are arranged in a repeating pattern. In particular, the cooling channels are arranged parallel to each other along two row directions.

[0015] Preferably, the conveying section and / or the lead-out section are formed by multiple individual layers, namely a first layer and / or a second layer. Advantageously, the recesses arranged vertically within the layers are slightly offset from each other, thereby creating protrusions extending into the cooling channel. This can induce additional vortices within the cooling channel.

[0016] Furthermore, it is proposed that the deviation between the first flow cross section measured perpendicular to the flow direction in the conveying section and / or the second flow cross section in the exiting section and the second flow cross section in the deflection section is less than 15% of the second flow cross section in the deflection section, preferably less than 10%, and particularly preferably less than 5%. Here, the flow cross sections are typically measured in the respective sections in a plane perpendicular to the flow direction, preferably as an average value when the flow cross section varies along the flow direction. This advantageously achieves a cooling element with relatively large free space, thereby reducing the weight of the cooling element. Furthermore, it is proposed that the deviation between the flow cross section measured perpendicular to the flow direction in the transverse flow section and the second flow cross section in the deflection section is less than 15% of the second flow cross section in the deflection section, preferably less than 10%, and particularly preferably less than 5%.

[0017] Furthermore, it can be envisioned that the deflection segment extends substantially linearly in a direction parallel to the main extension plane.

[0018] Specifically, it is proposed that a first opening is formed in the conveying section on the second end side. Furthermore, it is proposed that the deviation between the flow cross-section perpendicular to the flow direction measured in the first opening and the first flow cross-section of the downstream section of the conveying section is less than 15% of the first flow cross-section in the downstream section of the conveying section, preferably less than 10%, and particularly preferably less than 5%. Here, the downstream section of the conveying section is preferably located between the deflection section and the first opening, particularly centrally located between the deflection section and the first opening. This ensures a particularly large volumetric flow that can be introduced into the conveying section and guided through the cooling channel.

[0019] Specifically, a second opening is formed in the outlet section on the second end side. Furthermore, the deviation between the flow cross-section perpendicular to the flow direction measured in the second opening and the third flow cross-section of the upstream section of the outlet section is less than 15% of the third flow cross-section in the outlet section, preferably less than 10%, and particularly preferably less than 5%. The upstream section of the outlet section is preferably located between the deflection section and the second opening, particularly centrally located between the deflection section and the second opening. This ensures that a particularly large volumetric flow can be guided through the cooling channel. Furthermore, nozzle effects are avoided. Here, the dimensions of the first opening and / or the second opening are preferably designed such that the flow cross-section constituting the aforementioned section...

[0020] Furthermore, it is proposed that multiple cooling channels, particularly multiple individual cooling channels, are connected to a common supply channel via conveying sections. Therefore, each deflection section has its own conveying section. The conveying section is thus defined in the area between the supply channel used by the multiple individual cooling channels and the deflection section. In particular, it is proposed that each cooling channel has exactly one lateral flow section.

[0021] Furthermore, it is preferable that the cooling element is closed at its first end side, or has no gaps or openings. That is, a continuous layer of material is provided at the first end side, which, in particular, is consistent in material composition with at least one layer forming the cooling element, namely the first, second, third, and / or fourth layers. This ensures a particularly durable connection, especially in critical areas subjected to intense temperature changes. Therefore, the deflection section, and consequently the cooling fluid, does not directly contact the object to be cooled. This advantageously prevents potential damage or harm in the interface area, thereby increasing service life.

[0022] In particular, it is proposed that the cooling element is constructed by means of tabs disposed between the conveying section and the outlet section to form a partition wall, and at least one transverse flow section, preferably multiple transverse flow sections, are disposed within the partition wall. This provides a particularly lightweight cooling element, which is also advantageous in terms of flexibility. Furthermore, the tabs preferably form partition walls between adjacent cooling channels. The tabs preferably define corresponding flow cross-sections, and in particular determine the shape of said flow cross-sections.

[0023] Preferably, the tab elements connect adjacent face segments. Adjacent face segments are preferably part of a column segment, which particularly preferably extends between a first end side and a second end side. Face segments advantageously ensure the necessary stability within the cooling element. Preferably, the face segments are distributed in a two-dimensional or three-dimensional arrangement within the cooling element. It is conceivable that the cooling element includes face segments that are connected in a plane to all adjacent face segments via tab elements, and / or face segments that are connected in a plane to less than 75%, preferably less than 55%, and particularly preferably less than 50% of all adjacent face segments.

[0024] The patch element should be understood in particular as a planar, interconnected sub-segment extending along a plane parallel to the main extension plane within a layer, the sub-segment having a preferred extension direction, the patch element having an extension along the preferred extension direction that is several times, preferably at least five times, preferably at least eight times, and particularly preferably ten times, the patch width measured perpendicular to the preferred extension direction. The face segment does not have a preferred extension direction, or may have an extension along the preferred extension direction, the extension not exceeding five times, preferably eight times, and particularly ten times the patch width.

[0025] Furthermore, it is preferably proposed that at least one transverse flow section has a flow cross-section having a value between 0.005 mm² and 0.07 mm², preferably between 0.01 mm² and 0.055 mm², and particularly preferably between 0.015 mm² and 0.04 mm². It has been proven that with such a flow cross-section, sufficient cooling medium can still enter the deflection section, and the opening constituting the transverse flow cross-section is also large enough to reduce the possibility of blockage by particles in the cooling medium. The flow cross-section is also defined here as a surface extending perpendicular to the transverse flow direction and preferably limited by the tab element.

[0026] Preferably, the cooling element has at least one first layer and a second layer, which are stacked one on top of the other to form at least one cooling channel. Preferably, the cooling element comprises a system having multiple first layers and multiple second layers, wherein the first and second layers may be geometrically different from each other. In particular, it is conceivable that the first and second layers are manufactured as shaped etched parts, stacked one on top of the other and subsequently joined together. It is also conceivable that the cooling element is constructed from only multiple first layers. The first and second layers preferably have a thickness between 0.1 mm and 0.8 mm, preferably between 0.1 mm and 0.5 mm, and particularly preferably between 0.1 mm and 0.3 mm.

[0027] Particularly preferably, the voids are embedded in the first and second layers, and these voids are strategically stacked vertically to form corresponding cooling channels, preferably forming cooling channel segments extending along the stacking direction. Here, for example, the tilt angle of the cooling channels and / or the spiral or stepped extension of the cooling channels can be set by the corresponding misalignment of the first and second layers or the first and second voids, which particularly facilitates vortices flowing within the cooling channels, thereby improving the efficiency of the cooling channels. It is conceivable that the voids are star-shaped and / or have protrusions extending into them to additionally influence the flow performance when flowing along the flow direction. Preferably, the first and second layers are offset from each other in a direction parallel to the main extension plane, preferably by a distance between 0.01 mm and 0.5 mm, more preferably between 0.01 mm and 0.25 mm, and particularly preferably between 0.01 mm and 0.15 mm. This allows for the desired offset, particularly for spiral extensions, which have proven to be especially advantageous for vortices in the cooling medium.

[0028] Furthermore, it is preferable that the misalignment between the first and second layers, or between the first and second voids, increases along the flow direction from the second end side towards the first end side. This correspondingly increases the vortex intensity in the cooling channel in a region located as close as possible to the first end side to be cooled.

[0029] Preferably, the first and / or second layers have face segments, which are connected to each other via tab elements. The face segments are arranged to overlap each other, particularly in the combined and stacked state of the first and second layers. Preferably, the face segments of the first and second layers are identically constructed, especially located in the same position. It is advantageous to implement, therefore, a columnar segment, preferably continuous from the first end to the second end, in the cooling element, which is particularly advantageous for the stability of the cooling element. Here, the configuration of the face segments, measured parallel to the main extension plane, can be designed rectangularly, polygonally, squarely, circularly, and / or ovally. Preferably, the face segments are arranged in a checkerboard pattern relative to each other, or in a defined two-dimensional pattern relative to each other. Tab elements connect adjacent face segments. Preferably, face segments are provided that are connected to adjacent face segments of the same layer in two linearly independent directions. Preferably, both the first and second layers also include such face segments, which have a reduced number of connectors with tab elements. For example, it is conceivable that the facial segment has fewer than three tab elements protruding from the facial segment along the main extension plane. Thus, by selectively omitting tab elements between the various facial segments, larger gaps can be selectively created in the first and second layers.

[0030] In particular, it is proposed that the face segments and tab elements are connected to each other in a grid-like manner, especially to form a first layer and / or a second layer. Preferably, the face segments and / or tab elements are designed to create star-shaped gaps. Thus, it is also advantageous to influence the deformation and / or stiffness of the cooling element.

[0031] Particularly preferred is that the tab elements connecting the face segments do not extend in a straight line along a plane parallel to the main extension plane. For example, the tab elements are designed to be angled and / or curved. Waveform designs are also conceivable. It has been advantageously proven that a certain degree of flexibility can be generated by non-linearly extended tab elements, which has proven particularly advantageous for mutually compensating forces acting on the ceramic element, especially forces acting on the ceramic element during cooling after bonding with the metal layer. Stiffness can be selectively adjusted, for example, through geometric configuration. This allows for advantageous means, for example, counteracting deflection or targeted reverse control.

[0032] Preferably, the first and second layers are arranged stacked vertically, such that the non-linear extensions of the stacked tab elements create openings to form lateral flow sections. In other words, corresponding openings are achieved through the stacked layers, particularly in sections with tab elements, allowing desired lateral flow. In a top view of the stacked first and second layers along the stacking direction, the stacked tab elements form a boundary section with a rhomboid cross-section. Alternatively, depending on the configuration of the tab elements, the cross-section may be at least elliptical or similar to an ellipse.

[0033] Furthermore, it is conceivable that the first opening in the first layer and the second opening in the second layer are laterally offset from each other along a direction parallel to the main extension plane, for example, without twisting, so as to achieve a lateral flow section. It is also conceivable that the first opening in the first layer and the second opening in the second layer are misaligned, without forming a lateral flow section, because in this manner, flexibility favorable to deformation also occurs in the cooling element. Preferably, it is also conceivable that the face sections are offset from each other. This, for example, can additionally ensure vortices.

[0034] Preferably, a plurality of cross-sectional flow sections are arranged in an overlapping manner, particularly at least one cross-sectional flow section for each pair of first and second layers. Advantageously, it is irrelevant whether the cross-sectional flow sections are blocked by implementing multiple cross-sectional flow sections. In this case, providing sufficient additional cross-sectional flow sections ensures lateral flow within a certain and desired range, thereby enabling a reduction in thermal resistance.

[0035] Preferably, the first and second layers, stacked one on top of the other, constitute a continuous column section. This column section is particularly useful for improving stability, especially along a direction parallel to the stacking direction. The column section preferably provides stiffness, which is essential, especially during the sintering process. Furthermore, the column section has proven advantageous for heat transfer. It is also conceivable that the stacked layers have face sections of different sizes, thus obtaining, for example, column sections that taper towards the second end. This also proves advantageous for the cooling efficiency of the cooling equipment. Here, the column section preferably tapers layer by layer in a stepped manner.

[0036] Preferably, the conveying section has a first flow cross section measured perpendicular to the flow direction, and the deflection section has a second flow cross section measured perpendicular to the flow direction, wherein the ratio of the second flow cross section to the first flow cross section is less than 0.5, preferably less than 0.4, and particularly preferably less than 0.3.

[0037] Compared to methods known in the prior art, the flow cross-section in the deflection region or deflection section is specifically modified, and particularly significantly reduced, compared to the flow cross-section in the conveying section and the outlet section. This deviates, in particular, from the approach intended to keep the flow velocity in the channel as constant as possible. By specifically influencing the flow velocity, the pressure drop experienced by the cooling medium as it passes through the cooling element can be advantageously reduced. This is crucial for pumps that pump the cooling medium through the cooling element. Where the pressure drop of the cooling medium between the inlet and outlet is low, it is advantageous to use a pump with less stringent performance requirements.

[0038] Preferably, the first and / or second and / or third flow cross-sections used to form the microchannels have values ​​between 0.1 mm² and 25 mm², preferably between 0.1 mm² and 12.5 mm², and particularly preferably between 0.1 mm² and 5 mm². Therefore, this relates to cooling channels configured as microstructures. In other words, it is preferably proposed that the channel structure composed of multiple channel elements is a microchannel structure, the microchannel structure being small enough that as many individual cooling channels as possible can be arranged side-by-side to ensure corresponding uniformity and uniform supply to the surface to be cooled. In other words, it is preferably proposed that the cooling body element includes a microchannel cooling structure. This is particularly different from finned structures, which are relatively large in size and provide a very large flow cross-section between the fins, especially in a plane extending perpendicular to the flow direction.

[0039] Furthermore, it is conceivable that the first flow cross-section and / or the second flow cross-section and / or the third flow cross-section have a polygonal shape. Preferably, the first flow cross-section and / or the third flow cross-section have a star-shaped cross-section. Preferably, the second flow cross-section also differs from the first flow cross-section and / or the third flow cross-section in geometry, and is designed, for example, as a rectangle. By changing the geometry, for example, from a star shape to a rectangle, it is advantageous to promote vortices during the transition from the conveying section to the deflection section, which can prove to be advantageous for heat transfer. Preferably, the first flow cross-section and / or the third flow cross-section have a geometry with more angles than the second flow cross-section. It has been proven that thermal resistance and pressure drop can be targeted and improved by means of different geometries.

[0040] Furthermore, it is proposed that the corresponding flow directions in the conveying section, deflection section, and exit section together constitute the overall flow direction. Here, the flow direction in the conveying section is particularly directly connected to the flow direction in the deflection section. The same applies to the flow directions in the deflection section and the exit section.

[0041] Furthermore, it is proposed that the deflection section preferably extends as parallel as possible to the surface to be cooled, particularly toward the first end side, in order to ensure corresponding point-like or surface-like cooling. The deflection section may also, for example, be inclined relative to the main extending plane and / or have a stepped and / or curved extension. The flow cross-section is particularly understood as the extension in the corresponding cooling channel in a plane extending perpendicular to the flow direction. Therefore, it relates to the surface defined herein and through which the cooling medium passes.

[0042] Preferably, a plurality of identically shaped cooling channels are provided, wherein the cooling channels are part of a plurality of identically shaped cooling channels. Preferably, more than 50 cooling channels are provided in the cooling body element, more preferably more than 100 cooling channels, and particularly preferably more than 200 cooling channels. Thus, it is advantageous to ensure cooling at the first end side as uniformly as possible. It is also conceivable that multiple cooling channels are concentrated in one sub-section. Corresponding clusters of cooling channels, for example, allow for targeted and more intense cooling of areas where higher heat dissipation is expected, especially areas where higher heat dissipation occurs during operation. This can thus ensure enhanced cooling in a targeted manner. Furthermore, it is advantageous that a smaller amount of cooling medium must be guided through the cooling body element. In principle, it is also conceivable that multiple sets of identically shaped cooling channels are constituted, wherein the corresponding sets differ in their type, i.e., have different configuration shapes for the conveying section, deflection section, and / or lead-out section.

[0043] Preferably, multiple cooling channels extend parallel to each other, preferably along the row direction, and particularly preferably arranged in a parallel and staggered manner. This advantageously allows for the common supply of transport sections to adjacent cooling channels and / or the containment of cooling fluid primarily flowing out from the outlet sections, by means of a suitable distribution structure, for example, the disclosure known and explicitly cited in DE 10 2016125 338. In particular, it is proposed that multiple rows of cooling channels are arranged side-by-side along the row direction in a predetermined direction through the main extension plane. This results in a regular pattern of cooling channels in the cooling body element.

[0044] Preferably, the outlet region has a third flow cross-section measured perpendicular to the flow direction, wherein the ratio of the third flow cross-section to the second flow cross-section is less than 0.5, preferably less than 0.4, and particularly preferably less than 0.3. More specifically, the outlet region has a third flow cross-section measured perpendicular to the flow direction, wherein the ratio of the third flow cross-section to the first flow cross-section is less than 0.5, preferably less than 0.4, and particularly preferably less than 0.3. Here, the third flow cross-section can be larger than the second flow cross-section, or vice versa. Particularly preferably, the first flow cross-section and the third flow cross-section are substantially the same. In other words, the flow cross-section decreases from the first opening to the deflection section and then increases again towards the second opening; the cooling medium enters the cooling element through the first opening, and the fluid exits the cooling element through the second opening.

[0045] Preferably, the conveying section begins in the first opening, particularly in the first opening embedded in the second end side, and / or the outlet section extends into the second opening, particularly into the second opening embedded in the second end side, wherein a first flow cross-section is defined in the region of the first opening and / or a third flow cross-section is defined in the region of the second opening. Alternatively, it is conceivable that the first and / or second openings are not located in the second end side, but are introduced from the side, i.e., from the side connecting the first and second end sides, and extend substantially along the stacking direction, for example. The configuration of the first and second openings at the second end side proves advantageous, particularly for the distribution structure correspondingly joined to the cooling body element, in order to ensure the delivery and clearing of the cooling medium. Preferably, to form a U-shaped orientation, the conveying section and the outlet section extend substantially parallel to each other, or, where an angle is formed, inclined relative to each other, wherein the angle has a value between 0° and 120°, preferably between 0° and 90°, and particularly preferably between 0° and 45°. In particular, the U-shaped configuration allows for the creation of cooling channels with the smallest possible lateral expansion. This enables as many cooling channels as possible to be arranged sequentially and side-by-side in rows extending parallel to each other. Specifically, the cooling channels are arranged in a two-dimensional configuration, preferably with a repeating pattern. The corresponding inclined orientation with the aforementioned angles can advantageously influence the flow rate, and further facilitates the compensation of manufacturing tolerances.

[0046] Furthermore, it is preferable that the deflection region has an additional contoured portion. For example, the additional contoured portion is a waveform design of the channel wall portion facing the first end side in the deflection region. This can be achieved, for example, by corresponding gaps or protrusions in the metallized portion or metal layer, which forms a covering layer of the cooling element at the first end side. Extending, tab-like additional structured portions are also conceivable, which, for example by protruding and / or spherical terminals, also contribute to improving heat transfer from the cooling element to the cooling liquid.

[0047] Furthermore, it is preferred that the conveying section and / or the lead-out section have additional structured parts that extend into the cooling channel, particularly for improving cooling efficiency, especially for improving or enhancing heat transfer from the cooling body element to the cooling liquid or cooling medium.

[0048] Preferably, the first flow cross-section decreases along the flow direction, particularly continuously, and / or the third flow cross-section increases along the flow direction, particularly continuously. Furthermore, it is preferred that the cooling element is closed at its first end. Alternatively, it is conceivable that the cooling element is open at its first end, and the rear side of the ceramic element forms part of a cooling channel in the engaged state.

[0049] Furthermore, it is preferably proposed that a partition wall be formed between the conveying section and the outlet section, the partition wall having a width measured parallel to the main extension plane, the width decreasing from the first end side toward the second end side. This allows for targeted modification of the corresponding lever action, while also taking into account that heat transfer from the cooling element to the liquid is most efficient, particularly in the first third or the region directly adjacent to the first end side.

[0050] Preferably, the distance between the deflection segment and the ceramic element is less than 300µm, more preferably less than 200µm, and particularly preferably less than 100µm. This allows for a particularly close arrangement of the deflection segment to the ceramic element. This is only feasible when bonding via the rear metallization is abandoned, especially when the first thickness of the device metallization is greater than the distance between the deflection segment and the ceramic element. The direct bonding of the cooling element provides the advantage of avoiding heat diffusion and shortening the thermal path to optimize cooling performance. It is also conceivable that the deflection segment is directly adjacent to the ceramic element. In this case, the cooling element directly bonded to the ceramic element is open at its first end. Heat can then be directly transferred to the cooling medium via the ceramic element.

[0051] Furthermore, it is proposed that at least one transverse flow section connecting the conveying section and the lead-out section is formed between the deflection section and the second end side.

[0052] According to another aspect, a power module having a cooling element, particularly a cooling element according to the present invention, is proposed, the power module comprising:

[0053] - Ceramic components,

[0054] - Device metallization, and

[0055] - Cooling element components,

[0056] The ceramic element, the device metallization section, and the cooling element extend substantially along a plane parallel to the main extension plane and are stacked vertically in a stacking direction perpendicular to the main extension plane. The ceramic element is disposed between the device metallization section and the cooling element.

[0057] The device metallization portion has a first thickness measured in the stacking direction, and the cooling element has a second thickness measured in the stacking direction, wherein the ratio between the first thickness and the second thickness is less than 0.4, preferably less than 0.3, and particularly preferably less than 0.2.

[0058] The cooling element, particularly through the targeted arrangement of hollow regions within the cooling element and / or through targeted mass distribution within the cooling element, is designed such that the deflection of the power module is less than that of the reference power module, being one-fifth, preferably one-eighth, and particularly preferably one-tenth. The reference power module has dimensions corresponding to those of the power module, and the reference power module has a solidly formed cooling element. Specifically, the second thickness of the reference power module corresponds to the second thickness of the power module. The difference between the reference power module and the power module is essentially only that the reference power module is constructed as a solid, cavity-free metal layer with the second thickness of the power module.

[0059] Contrary to methods known from the prior art, this proposal utilizes the configuration of the cooling element, particularly its internal configuration, to ensure compensation for thermomechanical stresses acting on the rear and front sides of the ceramic element throughout the power module. This differs from common methods in the prior art, where the cooling element is designed independently of the power module and simply bonded to the rear metallization after the cermet substrate is manufactured. This allows for simultaneous bonding of the cooling element to the ceramic element, preferably during a common bonding or bonding process. This avoids the additional step of post-bonding the cooling element. However, this is only meaningfully achieved if the cooling element is correspondingly designed such that deflection is reduced compared to a reference power module, i.e., a power module with the same geometry but a solid cooling element. Those skilled in the art will understand "solid" particularly as a cooling element without hollow areas, i.e., a substantially unstructured metal layer or block.

[0060] Therefore, it has been proven that, despite significant differences in thickness—specifically, significant differences in the second thickness of the cooling element and the first thickness of the device metallization portion with a given ratio—it is entirely feasible to ensure that the deflection of the power module does not impair, and in particular, limit, the functionality of the power module, or make it usable. Deflection is particularly understood as a deviation relative to a perfectly flat plane. Deflection is typically of an arch type, such that it can preferably be considered as an angle formed between a plane and a point that extends tangentially through the apex of the arch, and that the point deviates or is spaced to the maximum extent from or spaced from the extension of the flat plane extending through the apex. This angle can, for example, serve as a criterion for determining the corresponding deflection. Here, the angle should be determined relative to the same distance from the apex for both the reference power module and the power module.

[0061] For example, the first thickness has a value between 0.1 mm and 4 mm, preferably between 0.1 mm and 2 mm, and particularly preferably between 0.2 mm and 0.8 mm, and / or the second thickness has a value between 0.5 mm and 6 mm, preferably between 1.0 mm and 5 mm, and particularly preferably between 1.5 mm and 4 mm or even between 2 mm and 3.5 mm.

[0062] Preferably, the cooling element can be formed from one or more metal layers. However, it is also conceivable to construct a composite structure in which metal and / or ceramic alternately define the configuration of the cooling element. Cooling elements composed of ceramic layers are also conceivable. Each layer can be made of metallic and / or ceramic materials. Materials for the metal layers include copper, aluminum, molybdenum, tungsten, nickel, and / or their alloys, such as CuZr, AlSi, or AlMgSi, as well as laminates such as CuW, CuMo, CuAl, and / or AlCu, or MMC (metal matrix composite), such as CuW, CuM, or AlSiC. Preferably, the ceramic layer or ceramic element uses Al2O3, Si3N4, AlN, HPSX ceramic (i.e., ceramic with an Al2O3 matrix comprising x percentage shares of ZrO2, such as Al2O3=HPS9 with 9% ZrO2 or Al2O3=HPS25 with 25% ZrO2), SiC, BeO, MgO, high-density MgO (>90% theoretical density), and TSZ (tetragonal stabilized zirconia) as materials for the ceramic. It is also conceivable that the ceramic layer is configured as a composite or hybrid ceramic, wherein multiple ceramic layers, differing in their material composition, are stacked one on top of the other and assembled into an insulating element to combine different desired properties.

[0063] It is particularly essential to ensure a certain porosity within the cooling element through corresponding cavities, thereby providing a similar mass distribution on both the upper and rear sides of the ceramic element to meaningfully compensate for deflection. This compensation is especially feasible when the cooling element is correspondingly finely and precisely designed. Preferably, the ceramic element, particularly when it is part of the cooling element, has a thermal conductivity greater than 80 W / mK, preferably greater than 100 W / mK, and particularly preferably greater than 120 W / mK. This allows for the achievement of advantageous thermal resistance, which is essential for the effective removal of heat.

[0064] To reduce deflection, the cooling element preferably has a plurality of three-dimensionally arranged tab elements that together form cooling channels. The tab elements are assembled such that they form corresponding cavities that exert a corresponding effect on deflection. The tab elements preferably have a tab width of less than 250 µm, more preferably less than 200 µm, and particularly preferably less than 150 µm. It has been advantageously proven that plastic deformation can be achieved during cooling in the bonding process precisely with such a thin tab width. This helps to reduce mechanical stress and positively influences the degree of deflection. In particular, the tab elements are designed to form wall structures or partitions, especially thin wall structures or partitions, which, for example, separate the delivery section from the lead-out section. These wall structures or partitions can be plastically deformed relatively easily during cooling and can contribute to compensation within the desired range in the aforementioned manner.

[0065] Preferably, the patch element has a cross-section along a plane measured perpendicular to the preferred direction, the cross-section being rectangular and / or square and / or circular, and / or occupying an area of ​​less than 0.08 mm², preferably less than 0.05 mm², and particularly preferably less than 0.038 mm².

[0066] Preferably, the cooling element, as a body, defines a first volume by its outer circumference, and a hollow region within the cooling element occupies a second volume (e.g., which can be defined as completely filling all cooling channels with liquid), wherein the ratio of the second volume to the first volume has a value greater than 0.5, preferably greater than 0.6, and particularly preferably greater than 0.7. It has been shown that increasing the proportion of the hollow region within the cooling element can improve the compensation for thermomechanical stresses within the cooling element.

[0067] Preferably, the power module is configured as a metal-ceramic substrate, wherein at least one metal layer bonded to the ceramic element in the manufacturing state is structured. For example, it is proposed that structuring also be performed after the bonding step, for example by laser, etching and / or machining, by means of which printed wires and / or connectors for electrical or electronic devices are realized.

[0068] In particular, it is proposed that the cooling element is directly bonded to the ceramic element. This direct bonding can be seen, for example, in the fabricated cermet substrate, resulting in a uniform grain size distribution, especially from the first end to the second end, during the extension of the cooling element. In other words, it is not apparent that a different grain size is achieved in the sub-segment of the cooling element directly adjacent to the ceramic element compared to the rest of the cooling element. This is evident, especially in the fabricated cermet substrate, where the cooling element itself is bonded directly and adjacent to the ceramic element, particularly to the rear side of the ceramic element. This direct bonding proves particularly advantageous because it eliminates the costly intermediate step in which the rear metallization portion must first be sequentially applied, and then the cooling element bonded to the rear metallization portion.

[0069] Preferably, the metallized portion of the device has a first mass, and the cooling element has a second mass, wherein the deviation between the second mass and the first mass is less than 160%, preferably less than 120%, and particularly preferably less than 80%. This corresponding mass distribution proves particularly advantageous in resisting deflection. It has been shown that, despite significant differences in thickness, the difference between the first and second masses is relatively small. Furthermore, the entire power module is particularly lightweight. This is preferably coupled with the corresponding orientation of the various wall structures within the cooling element to, for example, generate a corresponding levering effect. This, for example, cannot be guaranteed in conventional methods.

[0070] Furthermore, it is conceivable to influence the thermomechanical stress at the front and rear sides of the ceramic element through particle size and / or similarly consider the thermomechanical stress. Particularly preferably, the device metallization has an average first particle size, and the cooling element has an average second particle size, wherein the ratio of the average first particle size to the average second particle size is less than 0.7, preferably less than 0.5, and particularly preferably less than 0.25. In other words, the device metallization comprises fine particles, while the cooling element has coarse particles. At least, the corresponding particle size distribution is considered in the configuration of the cooling element.

[0071] It is also conceivable that the mass distribution in the cooling element could be influenced by the patch width modulation of layers that are stacked one on top of the other and connected to each other before they are connected to each other, along a plane extending parallel to the main extension plane, so as to selectively increase density in specific areas and induce a corresponding levering effect. Furthermore, it is conceivable that flexural deformation could be influenced by reinforcing the layers that are connected to each other to form the cooling element.

[0072] Particularly preferably, the ceramic element has a third thickness measured in the stacking direction, and this third thickness has a value up to 500 µm, preferably up to 300 µm, and particularly preferably up to 250 µm, or even up to 200 µm. Thus, corresponding limiting values ​​are included, namely 500 µm, 300 µm, 250 µm, and 200 µm. It has been proven, and even feasible, to construct the ceramic element thinly through a suitable configuration of the cooling body element. For example, this would be impractical if flexing as expected for the reference module were to occur, because in this case, the relatively thin ceramic element might break, and the functionality of the power module would not be achievable.

[0073] In particular, considering the influence of thermomechanical compensation stress, the corresponding lever action can also induce thermomechanical stress compensation through a corresponding mass distribution on the side of the cooling element. Here, those skilled in the art utilize, in particular, their empirical values ​​regarding the bonding process and / or, for example, corresponding simulations considering the corresponding material and geometric properties of the cooling element. It is particularly proposed that the cooling element be designed to be porous, especially with multiple cooling channels. It is advantageously feasible by constructing multiple cooling channels to ensure a corresponding mass distribution by having multiple corresponding cavities in the cooling element, which reduce the mass without affecting the thickness of the cooling element. It is also feasible to provide a correspondingly smaller second mass without reducing the thickness of the second cooling element. Preferably, the cooling element has more than 30, more preferably more than 90, and particularly preferably more than 150 cooling channels, especially active cooling channels.

[0074] To this end, it is specifically proposed that cooling channels be embedded in the cooling element to influence the thermomechanical stress at the rear side of the ceramic component. Correspondingly, it is conceivable, for example, to resist corresponding deflection by strategically placing the cooling channels. In particular, the deflection of the power module is reduced by positioning the cooling channels.

[0075] Furthermore, it is preferably proposed that the device metallization portion has a first extension extending parallel to the main extension plane, and the cooling element has a second extension extending parallel to the main extension plane, wherein the ratio of the first extension to the second extension has a value between 0.9 and 1, preferably between 0.95 and 1, and particularly preferably between 0.98 and 1. In other words, it is preferably proposed that the device metallization portion and the cooling element are only slightly offset in the direction perpendicular to the stacking direction, especially to avoid levering as much as possible. Here, viewed in the stacking direction, the first extension and the second extension are not flush with each other, so that a pullback portion can be provided, particularly on the side of the ceramic element facing the device metallization portion. It is particularly preferably proposed that the cooling element is flush with the ceramic element on the outer circumference of its planar boundary extending parallel to the main extension plane. In other words, the ceramic element does not protrude relative to the cooling element in the direction perpendicular to the stacking direction.

[0076] Preferably, the cooling element has a hollow region, which is not necessarily used for guiding the cooling medium. The hollow region not intended for guiding the cooling medium is preferably a continuation of the cooling channel and / or preferably located in the edge region of the cooling element.

[0077] Specifically, it is proposed that the cooling element has hollow regions that are not designed for guiding the cooling medium. In other words, in addition to the cooling channels that guide the cooling medium during operation, other hollow regions are provided, which are preferably used only for reducing the secondary mass and achieving a favorable mass distribution. For example, the hollow regions have the shape of cooling channels. In other words, in the cooling element, the structure or arrangement of cooling channels is continued, without providing specific cooling channels for actually guiding the cooling medium. The cooling channels form so-called dummy structures, which advantageously contribute to mass distribution. Preferably, such unused cavities or dummy structures are embedded in the edge regions of the cooling element.

[0078] Preferably, the cooling element has cooling channels on its second extension parallel to the main extension plane. Cooling channels are provided even in areas not intended for cooling. Therefore, the cooling channels are preferably evenly distributed across the entire second extension of the cooling element. This particularly relates to the configuration of the main unit. Thus, the connecting structure that would connect the various cooling elements intended for corresponding individual substrates within the main unit is preferably omitted.

[0079] Another subject of the invention is a method for manufacturing a cooling element according to the invention. The advantages and characteristics described for the cooling element are similarly applicable to the method, and vice versa. Attached Figure Description

[0080] Other advantages and features will become apparent from the following description of preferred embodiments of the subject matter according to the invention, with reference to the accompanying drawings. Here, within the scope of the invention, various features of the various embodiments can be combined with each other.

[0081] The attached diagram shows:

[0082] Figure 1 A schematic diagram of a power module having a cooling element according to a first exemplary embodiment of the present invention is shown.

[0083] Figure 2 A schematic diagram of a power module according to a second exemplary embodiment of the present invention and a cross-sectional view of a cooling element are shown.

[0084] Figure 3 A schematic diagram of a power module according to a third exemplary embodiment of the present invention and a cross-sectional view of a cooling element are shown.

[0085] Figure 4 A partial schematic diagram of a cooling element for a power module according to a fourth exemplary embodiment of the present invention is shown, and

[0086] Figures 5a and 5c show partial schematic diagrams of a cooling element for a power module according to a fifth exemplary embodiment of the present invention, and

[0087] Figure 6 A partial schematic diagram of a cooling element for a power module according to a sixth exemplary embodiment of the present invention is shown. Detailed Implementation

[0088] exist Figure 1 The diagram illustrates a power module 10 according to a preferred embodiment of the invention. This power module 10 includes a device metallization portion 20, a ceramic element 25, and a cooling element 1. The device metallization portion 20 is preferably a structured metal layer in which printed wires and / or electronic connectors are formed. This allows the device metallization portion 20 to be used in circuits or circuit systems. For this purpose, electrical devices or bonding wires are connected, for example, to the connection surfaces and / or printed wires of the device metallization portion 20. Here, the structured portion (not shown) ensures electrical insulation between the various metal segments in the device metallization portion 20. For this purpose, the structured portion extends to the ceramic element 25, which acts as an insulating element to ensure insulation between the metal segments, which are then bonded to the ceramic element 25.

[0089] The primary purpose of the cooling element 1 is to ensure that the heat generated at the device metallization 25 during the operation of the power module 10 is effectively dissipated, so as to prevent corresponding damage to the power module 1 and, in particular, not to jeopardize or degrade the performance of the power module 1. In the power module 1 described herein, especially in the current material selection for the insulating element, it has proven to be a common practice to bond the metal layer bonded to the ceramic element 25 to form the device metallization 20 together with the rear metallization at the ceramic element 25 opposite to the device metallization 20.

[0090] This serves to resist thermomechanical stresses on the front side of the ceramic element 25, which are caused by the different coefficients of thermal expansion of ceramic and metal and would cause deflection in the absence of a rear metallization. Without corresponding compensation, the different coefficients of thermal expansion, especially during temperature variations during the bonding process, would cause deflection of the cermet substrate. Therefore, it is ideal to achieve corresponding compensation when the metal layer bonded to the device side is substantially symmetrical to the rear metallization, i.e., similar or identical in size, i.e., in thickness and extension. The substantially mutually compensated thermomechanical stresses act on the ceramic element 25 on opposite sides in the manner described above.

[0091] As is common in the prior art, the cooler is then subsequently attached to the corresponding rear metallization section to ensure the corresponding cooling performance for the power module 10 in the manner described. This is typically done on the customer side and is separate from the manufacturing process of the cermet substrate. This not only introduces undesirable additional working steps but also affects the maximum possible efficiency and cooling performance of the coolant element 1, thereby impacting the overall performance of the power module 10, especially without optimal coordination.

[0092] It is currently preferred that the cooling element 1 be directly bonded to the ceramic element 25, or that the cooling element be directly bonded to the ceramic element. It has been particularly proven feasible to bond the cooling element 1 to the ceramic element 25 without causing flexural deformation that renders the power module 1 unusable or reduces its performance efficiency. Specifically, it is proposed that the device metallization portion 20, the ceramic element 25, and the cooling element 25 extend along a plane parallel to the main extension plane HSE, and are arranged stacked vertically in a stacking direction S perpendicular to the main extension plane HSE. Here, the ceramic element 25 is disposed between the device metallization portion 20 and the cooling element 1. The device metallization portion 20 has a first thickness D1 measured in the stacking direction S, and the cooling element 1 has a second thickness D2 measured in the stacking direction S. The ceramic element 25 has a third thickness D3 measured in the stacking direction S. Figure 1In the embodiment shown, the cooling element 1 has a second thickness D2 greater than the first thickness D1 of the device metallization 20. This size of the cooling element 1 and the device metallization 20 typically causes significant deflection due to a considerably greater accumulation of material on the rear side of the ceramic element 25, especially when the cooling element 1 with the second thickness D2 is directly and adjacent to the rear side of the ceramic element 25. However, it has now surprisingly proven feasible that the cooling element 1 is designed such that the deflection of the power module 10 is less than that of a reference power module, which is one-fifth or less, preferably one-eighth or less, and particularly preferably one-tenth or less, corresponding to the size of the power module 10, wherein the reference power module is designed with a solid cooling element 1. To reduce the deflection of the power module 10 compared to a reference power module of the same size, it has been proposed, for example, to ensure a reduction in the corresponding deflection of the power module 10 by embedding multiple cooling channels 30, 30, particularly microchannel cooling channels, into the cooling element 1.

[0093] In particular, it has been proven that deflection can be reduced in this manner, and it is also feasible to attach the cooling element 1 adjacently and directly to the rear side of the ceramic element 25. For this purpose, it is particularly preferred, for example, that the device metallization 20 has a first mass and the cooling element 1 has a second mass, wherein the deviation of the second mass from the first mass is less than 160%, preferably less than 120%, and particularly preferably less than 80%. In other words, it is proposed that, despite a significant difference between the first thickness D1 and the second thickness D2, the cooling element 1 is designed such that the mass distribution at the front and rear sides of the ceramic element 25 is similar in thermomechanical terms, for example, by embedding a corresponding number of cooling channels in the cooling element 1. Thus, it is also advantageously feasible, through the additional geometric distribution of the corresponding mass, to induce a corresponding levering effect, which is also advantageous in order to compensate for the corresponding deflection. Those skilled in the art here utilize corresponding empirical values ​​and / or simulations, which induce a corresponding mass distribution in the cooling element 1, which correspondingly acts on the deflection of the power module 10 in a reduced manner. Here, an exemplary configuration of the cooling element 1 can be partially seen in Figures 5b or 5c. It is also conceivable that other material distributions should be detected here as well. Therefore, as a difference from common methods, it is proposed that the cooling element 1 is bonded, for example, adjacent to and directly to the ceramic element 25, preferably in a bonding process common to the metal layer intended for the device metallization 20, in such a way that the mass distribution on the rear side of the ceramic element 25 is correspondingly designed and constructed for the cooling element 1 to reduce deflection. Preferably, it is also proposed that the first extension A1 of the device metallization 20, measured substantially parallel to the main extension plane HSE, is proportional to the second extension A2 of the cooling element 1, extending parallel to the main extension plane HSE. Preferably, the ratio of the first extension A1 to the second extension A2 has a value between 0.9 and 1, preferably between 0.95 and 1, and particularly preferably between 0.98 and 1. Correspondingly, it has proven advantageous that the second extension A2 of the cooling element 1 is larger than the first extension A1 of the device metallization portion 20, so as to minimize the lever effect on the rear side relative to the front side of the ceramic element 25. In particular, the manufactured metal-ceramic substrate or power module 10 is first a large card, and then individual metal-ceramic substrates or power modules 10 are produced from the large card by segmentation, preferably along the desired fracture region generated, for example by means of an ultrashort pulse laser. Here, the large card is particularly characterized in that the large card already has the joined cooling element.

[0094] exist Figure 2 The diagram illustrates a power module 10 according to a second exemplary embodiment. In particular, in... Figure 2The extension of feasible cooling channels 30, 30' is illustrated exemplarily. In particular, Figure 2 Two different feasible types of cooling channels 30, 30' are illustrated exemplarily, the cooling channels having two different overall flow directions. It is preferred in principle that cooling channels 30, 30' of the same type, and especially only a single type, are embedded in the power module 10. To illustrate the different types of cooling channels 30, in Figure 2 The cooling channels are simplified within the common cooling element 1. Preferably, the cooling element 1 has a first end side S1 and a second end side S2. The first end side S1 and the second end side S2 are positioned opposite each other along the stacking direction S in the power module 10, with the first end side S1 facing the ceramic element 25, and thus towards the surface to be cooled. Cooling channels 30, 30' are embedded or integrated into the substrate of the cooling element 1 and are disposed between the first end side S1 and the second end side S2. Figure 2 In the embodiment shown, the cooling element 1 has a first opening 41 at the second end side S2 and a second opening 42, which is also preferably provided in the second end side S2. Cooling medium, especially a fluid cooling medium, flows into the cooling channels 30, 30' via the first opening 41, while the medium used for cooling exits the cooling element 1 via the second opening 42. Preferably, the cooling channel 1 has a conveying section 31, a deflection section 32, and an outlet section 33. The conveying section 31 guides the cooling medium along the flow direction SR to the deflection section 32. Here, the conveying section 31 guides the cooling medium particularly toward the first end side S1 of the cooling element 1. The deflection section 32 then guides the cooling medium along the flow direction SR to the outlet section 33, and the outlet section 33 preferably guides the cooling medium to the second opening 42, i.e., particularly toward the second end side S2.

[0095] Furthermore, it is preferably proposed that the deflection section 32 extends substantially parallel to the first end side S1 of the cooling element 1 to ensure as planar cooling as possible. In other words, the actual cooling effect, or a major contribution to the cooling effect, begins in the deflection region 32, which is preferably located as close as possible to the ceramic element 25. Preferably, the distance A3 between the deflection region 33 and the ceramic element 25, measured in the stacking direction S, is less than 250µm, preferably less than 150µm, and particularly preferably less than 100µm. The successive flow directions SR in the conveying section 31, the deflection section 32, and the outlet section 33 preferably determine the overall flow direction in the cooling channels 30, 30'.

[0096] In the embodiment shown on the left, the overall flow direction is substantially U-shaped, while in the embodiment of cooling channels 30, 30' shown on the right, a substantially V-shaped flow is observed. Here, the flow direction SR in the conveying section 31 is inclined or angled relative to the flow direction SR in the outlet section 33, wherein the angle between the flow directions SR in the conveying section 31 and the outlet section 33 preferably has a value between 0° and 90°, more preferably between 0° and 60°, and particularly preferably between 0° and 45°.

[0097] In particular, it has proven advantageous to control, and especially to increase, the flow velocity in the deflection region 32. For this purpose, it is particularly preferred to adjust the flow cross-section accordingly during flow along the flow direction. It has been particularly preferred that the first flow cross-section Q1, measured perpendicular to the flow direction SR in the conveying section 31, is larger than the second flow cross-section Q2, measured perpendicular to the flow direction SR in the deflection section 32. In the cooling channel 30... Figure 2 In the embodiment shown on the left, the measured flow cross sections are oriented perpendicular to each other due to the overall flow direction. Preferably, the first flow cross section Q1 is determined in the region of the first opening 41. Furthermore, it is preferred that a third flow cross section Q3, with an outlet section 33 and measured perpendicular to the flow direction SR, is formed, particularly in the region of the second opening 42. It has proven particularly advantageous, especially for pressure drop in the flow direction, that the ratio of the second flow cross section Q2 to the first flow cross section Q1 is less than 0.5, preferably less than 0.4, and particularly preferably less than 0.3. Here, the flow cross sections are measured in particular as follows, the surface being defined by the corresponding channel walls belonging to the cooling channels 30, 30'. It is also conceivable that the first flow cross section Q1 decreases toward the deflection section 32, particularly in a stepwise and / or continuous manner, and / or the third flow cross section Q3 increases toward the second opening 42, particularly in a continuous and / or stepwise manner.

[0098] Furthermore, it is preferably proposed that the third flow cross section Q3 is larger than the second flow cross section Q2, and / or substantially coincides with the first flow cross section Q1. Those skilled in the art will understand, in particular, that the deviation between the first flow cross section Q1 and the third flow cross section Q3 is no greater than 10% of the average value of the first flow cross section Q1 and the third flow cross section Q3, preferably no greater than 5%, and particularly preferably no greater than 2.5%. Those skilled in the art will also understand, in particular, that the flow cross section is such that... Figure 2 The length of the cross section shown is only a portion of the flow cross section that should be used.

[0099] exist Figure 3The diagram illustrates a power module 10 according to a third exemplary embodiment of the invention. In particular, the power module 10 shown here is characterized by a cooling element 1, wherein cooling channels 30, 30' are designed to form a transverse flow section 50. Here, the transverse flow section 50 is disposed between the second end side S2 and the deflection section 32, and connects the conveying section 31 and the outlet section 33, especially upstream of the deflection section 32, i.e., before a portion of the cooling medium reaches the deflection section 32. Thus, before reaching the deflection section 32, a portion of the cooling medium has already been deflected and outleted into the outlet section 33. Surprisingly, it has been shown that this can improve the overall cooling efficiency of the power module 10. In particular, it has been shown that this can improve the thermal resistance of the cooling element 1. Furthermore, Figure 3 The power module is characterized in that the cooling element 1 is, preferably completely, flush with the ceramic element 25 at the outer circumference. This is also the preferred embodiment for other variations of the figures.

[0100] exist Figure 4 The diagram illustrates the arrangement of a first layer 60 and a second layer 70, which are stacked vertically along the stacking direction S to form part of a cooling channel structure in the cooling element 1. In particular, the first layer 60 and / or the second layer 70 have tab elements 61 and 71. These tab elements 61 and 71 preferably connect adjacent face segments 62 and 72, and particularly preferably, the face segments 62 and 72 of the stacked first layer 60 and the second layer 70 are stacked overlapping each other, particularly to form a pillar segment 67 in the cooling element 1. Therefore, this pillar segment 67 proves particularly advantageous because it additionally serves to stabilize the cooling element 1, especially along a direction extending parallel to the stacking direction S. Figure 4 In the embodiment shown, the face segments 62, 72 are substantially square. It is also conceivable to choose circular or elliptical geometries and / or rectangular and / or polygonal cross-sectional shapes. Furthermore, it is preferably proposed that the tab elements 61, 71 connecting the face segments 62 extend non-linearly. Figure 4 In the embodiment shown, the tab elements 61 and 71 are configured at an angle, particularly between 80° and 140°, preferably between 90° and 120°, and especially preferably between 90° and 100°. In other words, the tab elements 61 and 71 comprise two sub-segments extending at an angle to each other. Alternatively, it is conceivable that the tab elements 61 and 71 have a curved orientation as a non-linear direction.

[0101] exist Figure 4The diagram shows a top view of a first layer 60 and a second layer 70 arranged vertically stacked along the stacking direction S. It is possible, through the corresponding dimensions or configurations of the first layer 60 and the second layer 70, or through their corresponding orientations, that the tab elements 71 are oriented in opposite directions to each other. In particular, the tab elements are designed such that the tab elements 61, 71 form openings in the top view, the openings being upper bounded by the tab elements 61 of the first layer 60 and the tab elements 61, 71 of the second layer 70 in a viewing direction parallel to the stacking direction S. These openings form a lateral flow section 50. Figure 4 In the embodiment shown, the opening is particularly designed in a diamond shape. It is also conceivable that other geometries, particularly polygonal, elliptical, or circular geometries, can be designed here.

[0102] The tab elements 61 and 71 should be understood in particular as planar, interconnected sub-segments extending parallel to the main extension plane within the layer, said sub-segments having a preferred extension direction, along which the tab elements 61 and 71 have an extension several times, preferably at least five times, the tab width B measured perpendicular to the preferred extension direction. The face segments 62 and 72 do not have a preferred extension direction, or in any case have an extension along the preferred extension direction, said extension not exceeding five times the tab width B.

[0103] Especially in Figure 4 In one embodiment, face segments 62 and 72 are arranged in a checkerboard pattern facing each other, and the respective face segments 62 and 72 are connected to adjacent face segments 62 and 72 via tab elements 61 and 71 facing all sides, i.e., all four possible sides. Therefore, the first layer 60 and / or the second layer 70 are preferably designed in a grid pattern.

[0104] Figures 5a and 5c illustrate another design feasibility of the cooling element 1 according to an exemplary embodiment. Specifically, Figure 5a shows a top view of a first layer 60 and a second layer 70 arranged vertically stacked along the stacking direction S. In principle, it is conceivable that the first layer 60 and the second layer 70 are arranged alternately and repeatedly stacked vertically, which is suitable for… Figure 4The embodiments shown in Figures 5a to 5c are described. Alternatively, it is conceivable that a first number of first layers 60 are stacked vertically, and a second number of second layers 70 are stacked vertically, with the first and second layers stacked vertically in turn. In other words, the first number of first layers 60 is disposed above the second number of second layers 70. This allows for the determination of the corresponding positions of the cross-sectional flow sections 50 or the number of transverse flow sections 50. Here, the first number of first layers 60 need not coincide with the second number of second layers 70. It is particularly preferred that the cooling element 1 is constructed of at least a plurality, preferably at least five, particularly preferably at least eight, and particularly preferably at least ten individual layers, wherein the plurality of individual layers particularly preferably have only first layers 60 and second layers 70 to constitute the conveying section 31 and the lead-out section 33. The deflection section is preferably constructed by a third layer 80 and / or a fourth layer (not shown) or a plurality of third layers 80 and / or fourth layers. The first layer 60, the second layer 70, the third layer 80, and / or the fourth layer differ, for example, in the orientation of the tab elements 61, 71 and / or the number and / or size of the tab elements 61, 72 and / or in the arrangement, size, and / or number of the face segments 62, 72. Thus, the alternating arrangement creates an uneven inner side of the cooling channel, which facilitates eddies in the cooling medium.

[0105] Figure 5b shows a first perspective view of the stack consisting of a first layer 60 and a second layer 70. Here, the column section 67, which preferably extends continuously from the first end side S1 to the second end side S2, is particularly noticeable. Furthermore, it can be seen that multiple lateral flow sections 50 are formed by the stacked tab elements 61 and 71. Figure 5c also shows another perspective view of the cooling element 1. In particular, the embodiments of Figures 5a to 5c are similar to... Figure 4The difference in the embodiment is that face segments 62 and 72 are provided, which are not connected relative to adjacent face segments 62 and 72 in at least one direction. Preferably, face segments 62 and 72 are connected to adjacent face segments 62 and 72 along a first direction via tab elements 61 and 71, and are not connected to adjacent face segments 62 and 72 along a second direction, wherein the first and second directions lie in a plane parallel to the main extension plane HSE and are arranged at an angle to each other, particularly extending at right angles relative to each other. Thus, it is advantageous to determine, in particular, to increase the flow cross-section Q1 and / or the third flow cross-section Q3, so that a larger first flow cross-section Q1 or third flow cross-section Q3 is also correspondingly produced compared to the second flow cross-section Q2 in the deflection segment 32. It is particularly preferred that the conveying segment 31 and the lead-out segment 33 are implemented by corresponding stacking of single layers, wherein the deflection segment 32 is designed such that, for example, in the case of two adjacent cooling channels 30 and 30', the lead-out segment 33 is arranged side by side. In other words, the two outlet sections 33 of two adjacent cooling channels 30, 30' are located between the two conveying sections 31 of the adjacent cooling channels. This particularly simplifies the extraction of the cooling medium via the corresponding distribution structures 30, 30', since the two outlet sections 33 can share a common outlet in the distribution structure. For example, it is also conceivable that the first layer 60 and the second layer 70 have identical structures and are twisted relative to each other, preferably rotated 180°, stacked and joined. It is also conceivable that the first layer 60 and the second layer 70 are offset relative to each other in a direction parallel to the main extension plane HSE, preferably offset by a distance having a value between 0.01 mm and 0.5 mm, preferably between 0.01 mm and 0.25 mm, and particularly preferably between 0.01 mm and 0.15 mm.

[0106] exist Figure 6 The diagram illustrates another design feasibility of the cooling element 1 according to an exemplary embodiment. Here, specifically only the first layer 60 is stacked vertically to form the conveying section 31 and the lead-out section 33. The tab elements 61 are linearly configured and connect to the checkerboard-shaped face sections 62. This results in rectangular gaps that define the first or third flow cross-section in the conveying section 31 and the lead-out section 33. This design has proven particularly advantageous because it can be implemented relatively easily.

[0107] exist Figure 6In the embodiment shown, a plurality of third layers 80 are arranged vertically stacked to define the deflection section 32 by means of the gaps created by their vertical stacking arrangement. Here, the third layers 80, arranged vertically stacked with substantially the same structure, are positioned relative to each other such that the gaps overlap. This creates a substantially smooth inner side in the deflection region 32. Advantageously, the third layers 80 are staggered relative to each other and / or a fourth layer (not shown) is provided to create a contoured section on the inner side of the deflection section 32, or in the transition from the conveying section 31 to the deflection section 32, and / or in the transition from the deflection section 32 to the lead-out section 33. This specifically induces eddies in areas particularly important for heat transfer. Preferably, the third layer 80 differs from the first layer 60 and / or the second layer 70 in that, compared with the first layer 60 and / or the second layer 70, at least one tab element 61 connecting adjacent face segments 62 to each other is omitted, in particular in order to achieve a larger gap constituting the deflection segment 32.

[0108] List of reference numerals

[0109] 1 Cooling element

[0110] 10 power modules

[0111] 20 Device Metallization Section

[0112] 25 ceramic components

[0113] 30, 30' cooling aisle

[0114] 31 Conveying Section

[0115] 32 deflection section

[0116] 33 Introduction Section

[0117] 41 First Opening

[0118] 42 Second opening

[0119] 50 Lateral Flow Section

[0120] 60 First Floor

[0121] 61, 71 splice elements

[0122] 62, 72 facial segments

[0123] 67 column section

[0124] 70 Second Floor

[0125] D1 First Thickness

[0126] D2 Second Thickness

[0127] D3 third thickness

[0128] A1 First Extension

[0129] A2 Second Extension

[0130] A3 spacing

[0131] B-type splice width

[0132] S-stack direction

[0133] S1 First Side

[0134] S2 Second Side

[0135] Q1 First flow cross section

[0136] Q2 Second Flow Cross Section

[0137] Q3 Third Flow Cross Section

[0138] SR flow direction

[0139] HSE Main Extension Plane

Claims

1. A cooling element (1) for an electric power module (10), the cooling element comprising: - A base having a first end side (S) and a second end side (S2) opposite to the first end side (S1), the first end side facing the surface to be cooled in the assembled state, and - Cooling channel (30, 30'), the cooling channel being embedded in the substrate between the first end side (S1) and the second end side (S2). The cooling channel (30, 30') has a conveying section (31), a deflection section (32), and an outlet section (33), wherein the cooling channel (30, 30') is designed to form an overall flow direction for conveying the cooling medium toward the first end side (S1) in the conveying section (31), transferring it to the outlet section (33) in the deflection section (32), and conveying it toward the second end side (S2) in the outlet section (33), wherein at least one transverse flow section (50) connecting the conveying section (31) and the outlet section (33) is formed between the deflection section (32) and the second end side (S2).

2. The cooling element (1) according to claim 1. The cooling channel (30, 30') is provided in a plurality of identically formed cooling channels (30, 30'), wherein the cooling channel (30, 30') is part of a plurality of identically formed cooling channels (30, 30'), and wherein the plurality of cooling channels (30, 30') extend in particular parallel to each other and are preferably staggered from each other along the row direction.

3. The cooling element (1) according to claim 2. The cooling channels (30, 30') are formed separately from each other in a plurality of identically shaped cooling channels (30, 30').

4. The cooling element (1) according to any one of the preceding claims. The second end (S2) has a first opening (41) in the conveying section (31), wherein the deviation between the flow cross section in the first opening (41) measured perpendicular to the flow direction (SR) and the first flow cross section (Q1) of the downstream section of the conveying section (31) is less than 15% of the first flow cross section (Q1) in the downstream section of the conveying section (31), preferably less than 10%, and particularly preferably less than 5%.

5. The cooling element (1) according to any one of the preceding claims. Each cooling channel (30, 30') has exactly one transverse flow section (50).

6. The cooling element (1) according to any one of the preceding claims. The cooling element (1) is directly bonded to the ceramic element (25), or the distance between the deflection section (32) and the ceramic element (25) is less than 300µm, preferably less than 200µm, and particularly preferably less than 100µm.

7. The cooling element (1) according to any one of the preceding claims. The cooling element (1) is closed at the first end side (S1).

8. The cooling element (1) according to any one of the preceding claims. At least one of the transverse flow sections (50) has a flow cross section having a value between 0.005 mm² and 0.07 mm², preferably between 0.01 mm² and 0.055 mm², and particularly preferably between 0.015 mm² and 0.04 mm².

9. The cooling element (1) according to any one of the preceding claims. The device includes multiple transverse flow sections (5), which are arranged vertically stacked along a direction perpendicular to the main extension plane (HSE) and / or arranged side by side with each other in a direction parallel to the main extension plane (HSE).

10. The cooling element (1) according to any one of the preceding claims. The cooling element (1) is constructed by means of tabs (61, 71) which are configured to form a partition wall between the conveying section (31) and the lead-out section (32), wherein at least one transverse flow section (50), preferably multiple transverse flow sections (50), are formed in the partition wall.

11. The cooling element (1) according to claim 5. The patch elements (61, 71) connect adjacent face segments (62, 72) and / or the patch elements (6) have a non-linear extension, especially an angled or curved extension.

12. The cooling element (1) according to any one of the preceding claims. The cooling element (1) has at least one first layer (60) and a second layer (70), which are stacked one on top of the other to form the cooling channel (30, 30'). The first layer (60) and / or the second layer (70) have tab elements (61, 71) that extend substantially along a plane parallel to the main extension plane (HSE) and are particularly preferably non-linear, especially angled or curved. The first layer (60) and the second layer (70) are preferably stacked one on top of the other such that the non-linear extensions of the stacked tab elements (61, 71) form openings to form the transverse flow section (50).

13. The cooling element (1) according to any one of the preceding claims. The conveying section (31) has a first flow cross section (Q1) measured perpendicular to the flow direction (S), and the deflection section (32) has a second flow cross section (Q2) measured perpendicular to the flow direction (S), wherein the ratio of the second flow cross section (Q2) to the first flow cross section (Q1) is less than 0.5, preferably less than 0.4, and particularly preferably less than 0.

3.

14. A power module (10) having a cooling element (1) according to any one of the preceding claims, the power module further comprising: - Ceramic components (25). - Device metallization section (20), and - The cooling element (1). The ceramic element (25), the device metallization portion (20), and the cooling element (1) extend substantially along a plane parallel to the main extension plane (HSE) and are stacked vertically in a stacking direction (S) perpendicular to the main extension plane (HSE). The ceramic element (25) is disposed between the device metallization portion (20) and the cooling element (1). The device metallization portion (20) has a first thickness (D1) measured in the stacking direction (S), and the cooling element (1) has a second thickness (D2) measured in the stacking direction (S), wherein the ratio between the first thickness (D1) and the second thickness (D2) is less than 0.4, preferably less than 0.3, and particularly preferably less than 0.2, wherein the cooling element (1) is designed, in particular, by means of a hollow region in the cooling element (1) and / or a targeted mass distribution within the cooling element (1), such that the deflection of the power module (10) is one-fifth, preferably one-eighth, and particularly preferably one-tenth of the deflection of the reference power module, wherein the reference power module has dimensions corresponding to the power module (10), wherein the reference power module has a solidly formed cooling element (1).

15. A method for manufacturing a cooling element (25) according to any one of the preceding claims.

Images