JP2018173357A - Test method, test sample, test system, evaluation method, evaluation system, and evaluation program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To evaluate reliability of a semiconductor module in a short test period.SOLUTION: A test method includes: a determination step for determining a mechanical load that causes, on a semiconductor module, a change of stress, distortion, etc., according to deformation generated on the semiconductor module in which an insulating substrate is joined to a metal base by a joint material in a heat cycle test (H/C test); a testing step for performing a mechanical fatigue test in which the mechanical load is applied to a test sample containing a sample joint material corresponding to the joint material; and an evaluation step for evaluating reliability of the semiconductor module by estimating an H/C life curve for the semiconductor module in the H/C test from a result of the mechanical fatigue test. Thus, a test period can be shortened.SELECTED DRAWING: Figure 15

Description

  The present invention relates to a test method, a test sample, a test system, an evaluation method, an evaluation system, and an evaluation program.

  For example, power semiconductor modules (simply referred to as semiconductor modules) are incorporated in the main industrial fields such as renewable energy fields such as solar power generation and wind power generation, in-vehicle fields such as hybrid cars and electric cars, and railway fields such as vehicles. In addition, power conversion systems such as power conditioners (PCS), inverters, and smart grids are employed. Stable operation of these systems is particularly important in the main industrial field, and high reliability of semiconductor modules is required.

One of the reliability tests of a semiconductor module is, for example, a heat cycle test (H / C test). In this test, for example, a thermal load due to a temperature change (heat cycle) of −40 to 125 ° C. is given to the semiconductor module. For example, in the case of a semiconductor module configured by bonding an insulating substrate on which a semiconductor element is mounted to a metal base using a bonding material such as solder or resin, the insulating substrate and the metal base, each having a different linear expansion coefficient, have different temperatures. Bending stress (stress) is applied to the bonding material between them due to thermal expansion and contraction with changes, fatigue accumulates through temperature changes of several hundred to several thousand cycles, and from the end toward the inside. Cracks grow and the thermal conductivity from the semiconductor element to the metal base decreases, leading to failure. By evaluating the state of the bonding material, such as crack growth, with respect to the number of cycles of the heat cycle, the product life of the semiconductor module, that is, thermomechanical reliability (simply called reliability) is evaluated (for example, patents) Reference 1).
Patent Document 1 Japanese Patent Application Laid-Open No. 2003-21584

  However, in the H / C test, in order to hold the high temperature and the low temperature for about 40 minutes each in each heat cycle, the heat cycle is several thousand cycles or more in about 2 hours. By repeating, for example, a semiconductor module is tested over a long period of one year. If the module is evaluated as not satisfying the product standard by the test, the module is redesigned and tested again. This significantly delays product development.

  Conventionally, in accordance with the Coffin-Manson rule, a rod-shaped bulk material made of the same material as the bonding material used in the semiconductor module has been subjected to a tensile and compression fatigue test, and the result is used to perform a semiconductor module against heat cycle. May be evaluated for reliability. However, this method is not necessarily suitable for evaluating the reliability of the semiconductor module because the stress applied to the bulk material by tension and compression differs in behavior from the stress applied to the bonding material by heat cycle. In fact, reliability tends to be underestimated.

  Therefore, by modeling the module so that the reliability of the semiconductor module can be evaluated with only a few factors regardless of the details of its design, and reproducing the stress applied to the semiconductor module by heat cycle only with those few factors, A test method that can uniformly and quickly evaluate the reliability of a semiconductor module having a configuration is desired.

  In all fields other than the reliability test of semiconductor modules, reliability can be improved by performing a direct bending test on an object having a structure part in which a bonding material is bonded to a substrate, or by indirectly applying a stress by a heat cycle test or the like. May be tested. It is also desired to evaluate the reliability of such an arbitrary structure portion in a short time.

(Item 1)
The test method may include a determination step of determining a mechanical load that causes the evaluation object to change according to the deformation generated in the evaluation object in which the bonding material is bonded to the substrate in the reliability test.
The test method may comprise a test step of applying a load including a mechanical load to a test sample including a sample bonding material corresponding to the bonding material.

(Item 2)
The reliability test may be a heat cycle test or a power cycle test.

(Item 3)
The object to be evaluated may have a structure part in which the base body and the object to be joined are joined by a joining material.
The test sample may have a structure in which a sample substrate corresponding to the substrate and a sample bonded body corresponding to the bonded body are bonded by a sample bonding material.

(Item 4)
Each of the substrate and the sample substrate, the bonded object and the sample bonded object, and the bonding material and the sample bonding material may be formed of the same material.
The sample bonding material may have a thickness corresponding to the corresponding edge of the bonding material at one or more edges.

(Item 5)
The object to be evaluated may include a metal base as a base, an insulating substrate having a circuit pattern on the front surface and a metal plate bonded to the back surface, and a semiconductor element mounted on the circuit pattern.
You may have the structure part by which the metal plate and metal base as a to-be-joined body were joined by joining material.

(Item 6)
In the determination step, a mechanical load that mechanically provides a change amount according to deformation of one or more edges of the bonding material in the reliability test may be determined.

(Item 7)
In the determination stage, the amount of change according to the deformation of the evaluation object during the reliability test may be measured.
In the determination step, a mechanical load that applies a load corresponding to the measured change amount may be determined for the test sample.

(Item 8)
In the determination step, the amount of change corresponding to the deformation at one or more edges in the reliability test may be calculated by analysis.
In the determining step, a mechanical load that applies a load corresponding to the calculated amount of change to one or more edges of the bonding material may be determined.

(Item 9)
In determining the mechanical load, a mechanical load that mechanically applies a load corresponding to the amount of change to the test sample may be calculated by analysis.

(Item 10)
The test step may be performed by relatively pressing a region surrounded by two or more fulcrums in the plane of the sample substrate to impart a warp deformation to the test sample.
The sample bonded body may be a plate-like member having a corner portion where two sides approach each other in a direction in which the radius of curvature due to warping deformation is minimized within the surface of the sample substrate.

(Item 11)
The sample substrate and the sample bonded body may be square.
One side of the outer periphery of the sample bonded body may be oriented in a direction of a predetermined angle with respect to one side of the outer periphery of the sample substrate.

(Item 12)
The predetermined angle may be 45 degrees.

(Item 13)
In the test stage, the test sample may be deformed by repeating the same number of times as the number of cycles of the reliability test.

(Item 14)
In the test stage, the test sample may be repeatedly deformed under a predetermined temperature condition.

(Item 15)
The reliability test may be a heat cycle test.
In the test stage, the test sample may be repeatedly deformed under temperature conditions corresponding to the conditions in the heat cycle test.

(Item 16)
The reliability test may be a heat cycle test.
In the test stage, the test sample may be repeatedly deformed at a cycle equal to or shorter than the cycle time in the heat cycle test.

(Item 17)
The test method may further include an evaluation step of evaluating the reliability of the bonded structure portion of the evaluation object based on the state of the bonded structure portion in the repeatedly bent test sample.

(Item 18)
In the evaluation stage, the heat cycle life curve may be estimated based on the result of determining the mechanical load and repeatedly deforming it corresponding to each of the heat cycle tests under a plurality of temperature conditions.

(Item 19)
The test phase may perform a plurality of tests that apply different mechanical loads to each of the plurality of test samples prior to the decision phase.
The determination step may determine a test that gives the test sample a change corresponding to the deformation generated in the bonded structure portion of the evaluation object in the reliability test from among the plurality of tests.
The test method may further include an evaluation stage for evaluating the reliability of the structure part to which the evaluation object is joined based on the test result of the determined test.

(Item 20)
In the determination step, a test that gives a load corresponding to the amount of change according to the deformation of the evaluation object in the reliability test may be determined from among a plurality of tests.

(Item 21)
As the amount of change corresponding to the deformation, at least one of stress, stress gradient, strain, strain gradient, and temperature may be used.

(Item 22)
As a change amount according to deformation, a one-dimensional distribution, a two-dimensional distribution, or a three-dimensional distribution of stress or strain within a predetermined range of the evaluation object and the test sample may be used.

(Item 23)
The test sample may have a structure in which the sample base and the sample bonded body are bonded with the sample bonding material.
The test stage includes, for each of a plurality of test samples, a load applied to the sample substrate, a distance between a force point and a fulcrum that gives a determined load in the plane of the sample substrate, a thickness of the sample substrate, a sample bonded body A plurality of tests may be performed by applying a mechanical load using a test sample in which at least one of the thickness, the thickness of the structure bonded by the sample bonding material, and the size of the sample bonded body is different.

(Item 24)
The test sample may include a sample substrate corresponding to the substrate in the evaluation object having a structure in which the substrate and the object to be bonded are bonded with a bonding material.
The test sample may include a sample bonded body corresponding to the bonded body.
The test sample may comprise a sample bonding material that bonds the sample substrate and the sample workpiece.

(Item 25)
The test step in the test method may be performed by subjecting a test sample to warp deformation by relatively pressing a region surrounded by two or more fulcrums in the plane of the sample substrate.
The sample bonded body may be a plate-like member having a corner portion where two sides approach each other in a direction in which the radius of curvature due to warping deformation is minimized within the surface of the sample substrate.

(Item 26)
The sample substrate and the sample bonded body may be square.
One side of the outer periphery of the sample bonded body may be oriented in a direction of a predetermined angle with respect to one side of the outer periphery of the sample substrate.

(Item 27)
The predetermined angle may be 45 degrees.

(Item 28)
The test system may perform the test by the test method according to any one of items 1 to 23.

(Item 29)
The evaluation method may include an acquisition step of acquiring test results of a plurality of tests in which different mechanical loads are applied to each of the plurality of test samples including the sample bonding material.
The evaluation method is determined to determine a test in which a test sample is subjected to a change in accordance with the deformation generated in the bonded structure portion of the evaluation object in which the bonding material is bonded to the substrate in the reliability test from among a plurality of tests. There may be stages.
The evaluation method may include an evaluation stage for evaluating the reliability of the portion of the bonding material of the evaluation object based on the test result of the determined test.

(Item 30)
In the determination step, a test that mechanically applies a load corresponding to a change amount according to deformation of the evaluation object in the reliability test may be determined from among a plurality of tests.

(Item 31)
As the amount of change corresponding to the deformation, at least one of stress, stress gradient, strain, strain gradient, and temperature may be used.

(Item 32)
As a change amount according to deformation, a one-dimensional distribution, a two-dimensional distribution, or a three-dimensional distribution of stress or strain within a predetermined range of the evaluation object and the test sample may be used.

(Item 33)
The test sample may have a structure in which the sample base and the sample bonded body are bonded with the sample bonding material.
In the acquisition stage, for each of a plurality of test samples, the load applied to the sample substrate, the distance between the force point and the fulcrum that gives the determined load in the plane of the sample substrate, the thickness of the sample substrate, the sample bonded body The test results of a plurality of tests in which at least one of the thickness, the thickness of the structure bonded by the sample bonding material, and the size of the sample bonded body are subjected to different mechanical loads may be obtained.

(Item 34)
The evaluation system may execute the evaluation method according to any one of items 29 to 33.

(Item 35)
The evaluation program may cause the computer to execute the evaluation method according to any one of items 29 to 33.

  The summary of the invention does not enumerate all the features of the present invention. In addition, a sub-combination of these feature groups can also be an invention.

A configuration of a semiconductor module is shown as an example of an evaluation object. An example of the stress distribution added to the joining material of a semiconductor module by a heat cycle is shown. The cross-sectional structure of the test sample regarding the reference line L in FIG. 3B is shown. The lower surface structure of a test sample is shown. The strain distribution which generate | occur | produces in a joining material when a semiconductor module is set to low temperature -40 degreeC (upper stage) and high temperature 125 degreeC (lower stage) is shown. The state of the bonding material of the semiconductor module after the H / C test (upper stage) and the state of the sample joined body of the test sample after the mechanical fatigue test (middle stage and lower stage) are shown. The strain distribution (left) generated in the bonding material of the semiconductor module by the heat cycle in the H / C test is shown, and the strain distribution (middle and right) generated in the sample bonding material of the test sample by bending deformation in the mechanical fatigue test is shown. The crack (left) which generate | occur | produced in the joining material of the semiconductor module by the heat cycle in a H / C test, and the crack (middle and right) which generate | occur | produced in the sample joining material of the test sample by the bending deformation in a mechanical fatigue test are shown. An example of plastic strain generated in a sample bonding material with respect to a load applied to the sample bonding material in a three-point bending test of a test sample is shown. An example of the stress gradient and strain sensitivity applied to the sample bonding material with respect to the distance between two fulcrums supporting the test sample (distance between fulcrums) in a three-point bending test of the test sample is shown. An example of the stress gradient and strain sensitivity applied to the sample bonding material with respect to the thickness of the sample bonded body constituting the test sample in the three-point bending test of the test sample is shown. An example of the stress gradient applied to the sample bonding material and the strain sensitivity with respect to the thickness of the sample bonding material constituting the test sample in the three-point bending test of the test sample is shown. An example of a lifetime curve is shown. 1 shows a configuration of a test system according to the present embodiment. The structure of a mechanical fatigue test apparatus is shown in a side view. The structure of a mechanical fatigue test apparatus is shown in a top view. The flow of the 1st test method using the test system concerning this embodiment is shown. The flow of the 2nd test method using the test system concerning this embodiment is shown. The structure of the evaluation system which concerns on a modification is shown. The flow of the evaluation method using the evaluation system concerning a modification is shown. 2 shows an exemplary configuration of a computer according to the present embodiment.

  Hereinafter, the present invention will be described through embodiments of the invention, but the following embodiments do not limit the invention according to the claims. In addition, not all the combinations of features described in the embodiments are essential for the solving means of the invention.

  FIG. 1 shows an example of the configuration of a semiconductor module 1 as an example of an evaluation object whose reliability is evaluated by the test method and the evaluation method according to the present embodiment. The semiconductor module 1 includes a case 2, an insulating substrate 3, a bonding material 4, a semiconductor element 5, terminals 6 a, 6 c and 6 c, wires 7 b and 7 c, and a gel filler 8.

  The case 2 is a housing that accommodates the constituent parts of the semiconductor module 1, and includes a metal base 2a, a frame 2b, and a lid 2c. The metal base 2 a (an example of a base) is a plate-like member that forms the bottom of the case 2, and supports each component of the semiconductor module 1 thereon. For the metal base 2a, a copper (Cu) substrate, an aluminum silicon carbide composite (Al-SiC) substrate, or the like with high heat dissipation can be used. In this embodiment, a copper substrate is employed. The frame 2b is a frame-like member that forms the side portion of the case 2, and is erected on the periphery of the metal base 2a. The lid 2 c is a plate-like member that forms the upper part of the case 2, and is placed on the frame 2 b to enclose each part of the semiconductor module 1 in the case 2. A plurality of openings for projecting the upper ends of the terminals 6a, 6b, 6c out of the case 2 are formed in the lid 2c.

The insulating substrate 3 is a substrate on which the semiconductor element 5 is mounted. For example, a DCB (Direct Copper Bonding) substrate, an AMB (Active Metal Brazing) substrate, or the like can be used. The insulating substrate 3 includes an insulating plate 3a, a metal plate 3b, and a wiring board 3c. The insulating plate 3a is a plate-like member formed from an insulating ceramic such as aluminum nitride, silicon nitride, or aluminum oxide to a thickness of 0.2 to 1 mm, for example. The metal plate 3b (an example of an object to be bonded) is provided on the back surface of the insulating plate 3a with a film thickness of, for example, 0.1 to 1 mm using a conductive metal such as copper or aluminum. The wiring board 3c is provided on the front surface of the insulating plate 3a by using a conductive metal such as copper or aluminum in the same manner as the metal plate 3b. The wiring board 3c has three circuit patterns 3c ₁ , 3c ₂ , 3c ₃ . Circuit pattern 3c ₁ is arranged on the left side of the drawing on the insulating plate 3a. The circuit patterns 3c ₂ and 3c ₃ are arranged in parallel on the right side of the drawing (respectively and in front) on the insulating plate 3a.

  The bonding material 4 (an example of a bonding material) is a member for bonding the insulating substrate 3 to the metal base 2a, and solder (for example, Sn-5 wt% Sb) can be used. With the bonding material 4, the heat generated by the semiconductor element 5 provided on the insulating substrate 3 can be efficiently guided (that is, exhausted) to the metal base 2 a. Note that the bonding material 4 forms a fillet between the insulating substrate 3 and the metal base 2a that widens the bottom from the insulating substrate 3 side toward the metal base 2a side. A portion formed by bonding the metal plate 3b of the insulating substrate 3 as the bonded body onto the metal base 2a by the bonding material 4 is referred to as a structure portion 9.

The semiconductor element 5 is a switching element that operates by a control signal as an example, and includes an insulated gate bipolar transistor (IGBT), a power metal oxide semiconductor field effect transistor (MOSFET) having electrodes on the front surface and the back surface, respectively. Can be adopted. In the case of an IGBT (or MOSFET), the semiconductor element 5 has an emitter electrode (or source electrode) and a gate electrode on the front surface, and a collector electrode (or drain electrode) on the back surface. In this embodiment, IGBT is adopted. The semiconductor device 5, by joining the rear surface to the circuit pattern 3c ₁ wiring board 3c by joining material 5a, is mounted on the insulating substrate 3.

The terminals 6a, 6b, and 6c are terminals for inputting an external signal to the semiconductor element 5 or inputting / outputting a current from the semiconductor element 5 to the outside. The terminals 6a, 6b, and 6c are formed in a columnar shape or a flat plate shape using a conductive metal such as copper or aluminum. Terminal 6a is erected in the vicinity of the left end of the circuit pattern 3c ₁ wiring board 3c, is connected to the collector electrode of the semiconductor element 5 via a circuit pattern 3c _1. Terminal 6b is erected on the circuit pattern 3c ₂ of the wiring board 3c, is connected to the emitter electrode of the semiconductor element 5 via a circuit pattern 3c ₂ and the wire 7b. Terminal 6c is erected on the circuit pattern 3c ₃ of the wiring board 3c, is connected to the gate electrode of the semiconductor element 5 via a circuit pattern 3c ₃ and wire 7c. The terminals 6a, 6b, and 6c are erected on the circuit pattern by a bonding material (not shown) such as solder, and protrude from the case 2 through the opening of the lid 2c.

The wires 7b and 7c are wire-like members that join the surface electrode of the semiconductor element 5 to the circuit pattern. The wires 7b and 7c are formed using, for example, a conductive metal such as copper or aluminum or a conductive alloy such as iron aluminum alloy. Wire 7b connects the emitter electrode of the semiconductor element 5 to the circuit pattern 3c _2. Wire 7c connects the gate electrode of the semiconductor element 5 to the circuit pattern 3c _3.

  The gel filler 8 is a member for sealing each component of the semiconductor module 1, and silicone gel can be used as an example. The gel filler 8 is filled in the case 2 (on the metal base 2a) after the insulating substrate 3, the semiconductor element 5, the terminals 6a, 6b, 6c, and the wires 7b, 7c are disposed. . Furthermore, each component is enclosed in the case 2 by placing the lid 2c on the frame 2b.

  In the semiconductor module 1 having the above-described configuration, the metal base 2a, the bonding material 4, and the insulating substrate 3 have different linear expansion coefficients (for example, 16 ppm, 20 ppm, and 7 ppm, respectively), so that a heat load is applied. In such a case, the joined body deforms in a convex shape toward the metal base 2a with respect to the insulating substrate 3 at a high temperature, and protrudes toward the insulating substrate 3 with respect to the metal base 2a at a low temperature. Bending stress (stress) is applied to the bonding material 4 due to deformation generated in such a bonded body, fatigue accumulates through temperature changes of several hundred to several thousand cycles, and cracks grow from the end toward the inside, When the thermal conductivity from the semiconductor element 5 on the insulating substrate 3 to the metal base 2a is lowered, the semiconductor module 1 is broken. Since the product life of the semiconductor module 1 strongly depends on the reliability of the bonding material 4, even if a bonded body including the bonding material 4, the metal base 2 a that applies stress to the bonding material 4, and the insulating substrate 3 is substituted as an evaluation object. Good.

  FIG. 2 shows an example of a stress distribution applied to the bonding material 4 of the semiconductor module 1 (here, a bonded body is substituted) by a heat cycle in the H / C test. For comparison, the stress distribution due to the pulling of the rod-shaped bulk material made of the same material as the bonding material 4 is also shown. The stress distribution of the bonding material 4 is given as a one-dimensional distribution with respect to a distance on a straight line from the end of the fillet (particularly, the end near the insulating substrate 3) to the inside. The stress distribution of the bulk material is given as a one-dimensional distribution with respect to a distance on a straight line from the rod surface toward the center. The stress applied to the bulk material by tension and compression is uniform over distance. On the other hand, the stress applied to the bonding material 4 by the heat cycle is a stress caused by the difference in thermal expansion and contraction between the insulating substrate 3 and the metal base 2a sandwiching it, and thus the spreading direction (and thickness of the bonding material 4). It can be seen that it has a finite distribution in the vertical direction and is particularly concentrated at the end (fillet surface). However, the stress is attenuated with respect to the distance from the end portion, and is distributed approximately equally and uniformly to the stress of the bulk material inside the bonding material 4. The stress distribution may depend on design details such as the linear expansion coefficient, size, shape, etc. of the bonding material 4, the insulating substrate 3, and the metal base 2a.

In order to accurately evaluate the reliability of the semiconductor module 1 as an evaluation object, stress applied to the bonding material 4 by the heat cycle in the H / C test or a strain generated in the bonding material 4 accompanying this (given to the bonding material 4) Therefore, the bonding material 4 in the semiconductor module 1, the metal base 2 a that applies stress to the bonding material 4, and the insulating substrate 3 (which may include the semiconductor element 5) must be reproduced. It is desirable that a test sample is constructed corresponding to the joined body including and used for the test. However, the insulating substrate 3 includes an insulating plate 3a, a metal plate 3b, and a wiring board 3c. The wiring board 3c has a plurality of circuit patterns 3c ₁ , 3c ₂ , 3c ₃ , and further testing is performed when the semiconductor element 5 is included. The composition of the sample becomes complicated and there are many factors for reliability evaluation. In order to evaluate the reliability of a semiconductor module with only a few factors, it is necessary to configure the test sample more simply.

  In order to reproduce the failure mode applied to the bonding material 4 by the heat cycle in the H / C test, a temperature field (that is, temperature or temperature distribution) and a stress field (that is, stress distribution or a strain generated in the bonding material 4 accompanying this). Distribution) may be given to the bonding material 4 independently. Furthermore, it is possible to easily evaluate the reliability of the semiconductor module 1 by parametrizing the stress field with a small factor regardless of the design details of the semiconductor module 1.

  Here, due to the difference in coefficient of linear expansion between the metal base 2a and the insulating substrate 3, the bonding material 4 is disposed between the metal base 2a and the insulating substrate 3 and reflowed to bond the insulating substrate 3 to the metal base 2a. In some cases, a joined body is formed in a deformed state at the time of thermal expansion, and stress accompanying the deformation may remain in the joining material 4. However, the stress remaining in the bonding material 4 (referred to as residual stress) only contributes to a negligible contribution to the reliability of the actual problem, H / C test by several hundred to several thousand heat cycles. Absent. Thus, the residual stress of the bonding material 4 is ignored, and the following test sample 11 is introduced as a test body for simulating the H / C test of the semiconductor module 1 that is the evaluation object.

  3A and 3B show the configuration of the test sample 11. Here, FIG. 3A shows a cross-sectional configuration related to the reference line L in FIG. 3B, and FIG. 3B shows a configuration on the lower surface. The test sample 11 includes the semiconductor module 1 in the structural portion 9 included in the test module 11 so that the reliability of the bonding material 4 in the semiconductor module 1 as an evaluation object can be evaluated with only a small factor regardless of the details of the module design. This is a correspondingly simplified test body, which makes it possible to uniformly evaluate the reliability of semiconductor modules having various configurations. The test sample 11 includes a sample base 12a, a sample bonded body 13b, and a sample bonding material 14.

  The sample base 12a is a plate-like member corresponding to the metal base 2a, and is made of the same material as the metal base 2a. The sample base body 12a is formed in a rectangular shape, in this case, as an example, a rectangular shape whose longitudinal direction is parallel to the reference line L.

  The sample bonded body 13b is a member corresponding to the metal plate 3b that is the bonded body, and is formed of the same material as the metal plate 3b. The sample bonded body 13b is formed into a square shape, in this case, a square shape as an example, and faces a corner where two sides approach each other in a direction parallel to the reference axis L on the lower surface of the sample base 12a. Accordingly, one side (for example, the upper left side) of the outer periphery of the sample bonded body 13b is predetermined with respect to one side (for example, the upper side parallel to the reference line L) of the outer periphery of the sample base 12a. The direction of the angle θ is, for example, 45 degrees. In a mechanical fatigue test to be described later, warp deformation is applied to the test sample 11 so that the direction in which the radius of curvature due to warp deformation is minimized in the plane of the sample base 12a is parallel to the reference line L.

  The sample bonding material 14 is a member corresponding to the bonding material 4 and is formed of the same material as the bonding material 4. The sample bonding material 14 has the same shape as the sample bonded body 13b and is arranged between the sample bonded body 13b and the sample base 12a to bond them together. Here, the sample bonding material 14 has a thickness corresponding to the corresponding edge portion of the bonding material 4 at one or more edge portions, and from the sample bonded body 13b side to the sample substrate 12a side, like the bonding material 4. To form a fillet that widens the hem.

  In order to use the test sample 11 to evaluate the reliability of the semiconductor module 1 with only a few factors, design conditions for the test sample 11 and test conditions for a mechanical fatigue test described later are determined.

  FIG. 4 shows a strain distribution generated in the bonding material 4 when the semiconductor module 1 is placed at a low temperature of −40 ° C. (upper stage) and a high temperature of 125 ° C. (lower stage) as an example. These strain distributions are obtained by thermal stress simulation and indicate the distribution on the surface of one corner of the bonding material 4. Under a low temperature, the strain concentrates on the lower side of the bonding material 4, that is, on the metal base 2a side with a small strain amount. On the other hand, under a high temperature, the strain concentrates on the upper side of the bonding material 4, that is, on the insulating substrate 3 side with a large strain amount. Therefore, in the heat cycle in the H / C test, it is expected that a large stress applied to the upper side of the bonding material 4 at a high temperature contributes predominantly to the reliability of the bonding material 4.

  FIG. 5 shows the state of the bonding material 4 of the semiconductor module 1 after the H / C test (upper stage) and the state of the sample bonding material 14 of the test sample 11 after the mechanical fatigue test described later (middle stage and lower stage). These were observed using an ultrasonic flaw detector. Here, the temperature conditions of the H / C test (that is, the heat cycle temperature) are a low temperature of −40 ° C. and a high temperature of 125 ° C., and the temperature conditions of the mechanical fatigue test are −40 ° C. (middle stage) and 125 ° C. (lower stage). It is. In any test, a crack grows from the end portion of the bonding material 4 (that is, the fillet surface) toward the inside. However, in the H / C test, cracks grow from the metal plate side (upper side of the fillet) of the bonding material 4, whereas in the mechanical fatigue test, when the temperature condition is 125 ° C., similarly, Cracks grow from the metal plate side (upper side of the fillet), and cracks grow from the metal base side (lower side of the fillet) of the bonding material 4 when the temperature condition is −40 ° C. From these observations, it can be seen that cracks occurred in the bonding material 4 at a high temperature in the H / C test. This is consistent with the knowledge obtained from the analysis of the strain distribution of FIG.

  From the above results, in the mechanical fatigue test, the high temperature in the heat cycle of the H / C test is used as the temperature condition, and the stress applied to the bonding material 4 or the strain generated in the bonding material 4 under the high temperature is reproduced. Thus, it is expected that the failure mode applied to the semiconductor module 1 by the heat cycle of the H / C test can be reproduced by the mechanical fatigue test.

  FIG. 6 shows a strain distribution (left) generated in the bonding material 4 of the semiconductor module 1 by the heat cycle in the H / C test, and is generated in the sample bonding material 14 of the test sample 11 by bending deformation in the mechanical fatigue test described later. Strain distribution (middle and right) is shown. These strain distributions are obtained by structural analysis such as three-dimensional finite element method (FEM) analysis. Here, the angle of the insulating substrate 3 with respect to the metal base 2a of the semiconductor module 1 used for the H / C test (corresponding to the angle θ described above) is 90 degrees, in the test sample 11 used for the mechanical fatigue test. The angles (equal to the aforementioned angle θ) of the sample bonded body 13b with respect to the sample base 12a are 90 degrees (medium) and 45 degrees (right). The temperature condition of the mechanical fatigue test is equal to the high temperature in the heat cycle of the H / C test. It can be seen that the bending stress applied to the bonding material in the heat cycle of the H / C test concentrates on the corners of the bonding material 4. On the other hand, the bending stress applied to the sample bonding material 14 in the mechanical fatigue test is 45 degrees in contrast to the side of the sample bonding material 14 being slightly biased toward the corner in the case of 90 degrees. In this case, it is concentrated at the corner of the sample bonding material 14. Therefore, it is expected that the bending stress applied to the bonding material 4 in the heat cycle of the H / C test can be reproduced by using the test sample 11.

  FIG. 7 shows a crack (left) generated in the bonding material 4 of the semiconductor module 1 due to the heat cycle in the H / C test, and a crack generated in the sample bonding material 14 of the test sample 11 due to bending deformation in the mechanical fatigue test described later. (Middle and right) are shown. These were observed using an ultrasonic flaw detector. Here, the angle of the insulating substrate 3 with respect to the metal base 2a of the semiconductor module 1 used for the H / C test (corresponding to the angle θ described above) is 90 degrees, in the test sample 11 used for the mechanical fatigue test. The angles (equal to the aforementioned angle θ) of the sample bonded body 13b with respect to the sample base 12a are 90 degrees (medium) and 45 degrees (right). The temperature condition of the mechanical fatigue test is equal to the high temperature in the heat cycle of the H / C test. It can be seen that cracks generated in the bonding material 4 in the heat cycle of the H / C test grow from the corner toward the inside. On the other hand, the crack generated in the sample bonding material 14 in the mechanical fatigue test grows inward from the side of the sample bonding material 14 in the case of 90 degrees, whereas in the case of 45 degrees, It can be seen that the sample bonding material 14 grows from the corner to the inside. Therefore, it is expected that the bending stress applied to the bonding material 4 in the heat cycle of the H / C test can be reproduced by using the test sample 11.

  Since the test sample 11 having the above-described configuration is a bonded body including the sample base 12a, the sample bonded body 13b, and the sample bonding material 14, the bonding material 4 is obtained by heat cycle in the H / C test for the semiconductor module 1. It is possible to reproduce the bending stress applied to or the strain generated in the bonding material 4 accompanying this. However, since the sample base 12a corresponding to the metal base 2a and the sample bonded body 13b corresponding to the metal plate 3b are made of the same material, the residual stress when they are bonded by the sample bonding material 14 corresponding to the bonding material 4 Does not occur. Therefore, by performing a mechanical fatigue test, which will be described later, using the test sample 11, the strain can be uniquely controlled by the load applied to the test sample 11 and other small factors, so that it does not depend on the details of the design of the semiconductor module. It is possible to evaluate the reliability of semiconductor modules having various configurations with only a small number of factors.

  In the mechanical fatigue test, a three-point bending test is adopted so as to reproduce the stress applied to the bonding material 4 of the semiconductor module 1 by the heat cycle in the H / C test or the strain generated in the bonding material 4 thereby. Here, in order to evaluate the reliability of the semiconductor module 1 using the test sample 11 with only a small factor by the mechanical fatigue test, several test conditions are defined. Details of the mechanical fatigue test will be described later.

  FIG. 8 shows an example of plastic strain generated in the sample bonding material 14 with respect to a mechanical load (that is, load) applied to the sample bonding material 14 in the three-point bending test of the test sample 11. Temperature conditions are 25 ° C. and 125 ° C. Since the plastic strain of the sample bonding material 14 is proportional to the load, it can be seen that the plastic strain can be controlled by the load. Here, when the plastic strain generated with respect to the increase in the load applied to the sample bonding material 14 is strongly increased, the plastic strain is not uniquely determined due to a slight shift in the load, and there is a possibility that a large error is caused. On the other hand, when the increase in plastic strain is small relative to the increase in load, the load within the output standard of the test apparatus cannot generate a sufficiently large plastic strain in the sample bonding material 14 and the test result may not be obtained. is there. Therefore, the maximum plastic strain amount (referred to as the target strain amount) to be reproduced by the test is set as 10%, which is the strain amount at which the bonding material 4 can be broken, as an example, so that the plastic strain can be finely controlled. The condition is that a load unit (that is, an adjustable load unit), for example, 1 / 100th of the target strain amount can be obtained for 1N, and a target strain amount can be obtained for the maximum load, for example, 1000N. . From these, the strain sensitivity (plastic strain generated when a load of 1 N is applied) of 0.01 to 0.1% / N is determined as a test condition.

  FIG. 9 shows an example of the stress gradient and strain sensitivity applied to the sample bonding material 14 with respect to the distance between the two fulcrums that support the test sample 11 (distance between the fulcrums) in the three-point bending test of the test sample 11. Here, the stress gradient is the amount of change in stress within about 50 μm in the vicinity of the end of the sample bonding material 14 (end of the fillet), and represents the concentration of stress at the end of the sample bonding material 14. The stress gradient in the horizontal and vertical directions is almost constant with the distance between the fulcrums, whereas the strain sensitivity increases with the distance between the fulcrums. From now on, when reproducing the stress applied to the bonding material 4 by the heat cycle of the H / C test by the mechanical fatigue test by the three-point bending test, the distance between the fulcrums is set to 27 mm or less from the condition for the strain sensitivity previously determined. It turns out that is desirable.

  FIG. 10 shows an example of the stress gradient and strain sensitivity applied to the sample bonding material 14 with respect to the thickness of the sample bonded body 13 b constituting the test sample 11 in the three-point bending test of the test sample 11. The horizontal stress gradient is greatly reduced for thicknesses of about 5 mm or less and becomes substantially constant for thicknesses of about 5 mm or more. The vertical stress gradient increases slightly for thicknesses of about 5 mm or less and is substantially constant for thicknesses of about 5 mm or more. Strain sensitivity increases greatly for thicknesses of about 5 mm or less and becomes substantially constant for thicknesses of about 5 mm or more. From this, when reproducing the stress applied to the bonding material 4 by the heat cycle of the H / C test by the mechanical fatigue test by the three-point bending test, the thickness of the sample bonded body 13b is determined based on the conditions for the strain sensitivity previously determined. It turns out that it is desirable to set it as 5 mm or less.

  FIG. 11 shows an example of the stress gradient and strain sensitivity applied to the sample bonding material 14 with respect to the thickness of the sample bonding material 14 constituting the test sample 11 in the three-point bending test of the test sample 11. The horizontal stress gradient decreases with respect to the thickness of the sample bonding material 14. The stress gradient in the vertical direction increases with respect to the thickness, and gradually decreases with a maximum at a thickness of about 170 μm. Strain sensitivity decreases with thickness. From now on, when reproducing the stress applied to the bonding material 4 by the heat cycle of the H / C test by the mechanical fatigue test by the three-point bending test, the thickness of the sample bonding material 14 is 100 μm based on the previously determined conditions for strain sensitivity. It can be seen that the above is desirable. In addition, since the behavior of the stress gradient in the vertical direction changes with the thickness of about 170 μm or less and the stress distribution may change, the thickness of the sample bonding material 14 is set to 170 μm or less so that the stress distribution does not change. It turns out that is desirable.

  In addition, the size of the sample bonded body 13b may be arbitrarily determined, and the thickness of the sample base 12a is desirably 5 mm or less from the condition for strain sensitivity.

FIG. 12 shows an example of the life curve. In the mechanical fatigue test, a three-point bending test was performed using the test sample 11 in accordance with the test conditions determined as described above. Here, as an example, the material of the sample base 12a of the test sample 11 is oxygen-free copper (corresponding to H material), the size is 40 × 100 mm, the thickness is 3 mm, and the material of the sample bonded body 13b is oxygen-free. Copper (corresponding to O material), the size is 20 × 20 mm, the thickness is 3 mm, the material of the sample bonding material 14 is solder (Sn-5 wt% Sb), the thickness is 100 μm, and the fillet shape (tilt) is The distance between supporting points of the support jig was 20 mm, the mechanical load (that is, the load) was 1.0 kN, 1.5 kN, and 2.0 kN, and the temperature condition was 125 ° C. Three-point bending _in which equivalent plastic strain (also simply referred to as plastic strain) Δε _in was applied to the sample bonding material 14 was repeated, and the number of cycles in which a crack of 1 mm or more occurred in the sample bonding material 14 was defined as a life cycle. In the H / C test, a bonded body including the bonding material 4, the metal base 2a bonded thereto, and the insulating substrate 3 was used. The conditions for the metal base 2a, the insulating substrate 3, and the bonding material 4 were the same as the conditions for the sample base 12a, the sample bonded body 13b, and the sample bonding material 14 in the mechanical fatigue test. Moreover, heat cycle temperature was made into low temperature -40 degreeC and high temperature 105,125,150 degreeC. The heat cycle was repeated with respect to each heat cycle temperature (the plastic strain Δε _in generated in the bonding material 4), and the number of cycles in which a crack of 1 mm or more occurred in the bonding material 4 was defined as the life cycle. In the fatigue test, tension and compression were performed using a rod-shaped bulk material made of the same material as the bonding material 4. The temperature condition was 125 ° C. The number of cycles in which the stress applied to the bulk material was reduced by 25% with respect to the initial stress was defined as the life cycle.

  In contrast to the life curve obtained from the H / C test, the life curve obtained from the fatigue test using the bulk material has a small plastic strain and a large slope for the same number of life cycles. On the other hand, the life curve obtained from the mechanical fatigue test has a slightly small plastic strain for the same life cycle number, and the slopes are almost equal. Therefore, the mechanical fatigue test by three-point bending using the test sample 11 can reproduce the stress applied to the bonding material 4 of the semiconductor module 1 by the heat cycle in the H / C test or the strain generated in the bonding material 4 thereby. The reliability of the semiconductor module 1 can be evaluated instead of the H / C test.

Table 1 shows test conditions in a mechanical fatigue test by three-point bending using the test sample 11. From the viewpoint of simulating the H / C test of the semiconductor module 1, the material of the sample base 12a of the test sample 11, the material of the sample bonded body 13b, and the material and thickness of the sample bonding material 14 are the same as those of the metal base 2a of the semiconductor module 1. It is determined so as to match the material, the material of the metal plate 3b, and the material and thickness of the bonding material 4. The temperature is set to a high temperature of the heat cycle temperature of the H / C test. Other conditions are arbitrarily determined so as to reproduce the stress applied to the bonding material 4 of the semiconductor module 1 by the heat cycle in the H / C test or the strain generated in the bonding material 4 thereby.

  FIG. 13 shows the configuration of the test system 100 according to the present embodiment. The test system 100 reproduces the stress applied to the bonding material 4 in the semiconductor module 1 by the H / C test or the strain caused to the structural portion 9 by applying a mechanical load from the outside of the module. Is repeatedly added to the module in a short cycle time, and the test time is shortened. The H / C test apparatus 110, the mechanical fatigue test apparatus 120, and the control apparatus 130 are provided.

  The H / C test apparatus 110 is an apparatus that performs an H / C test on the semiconductor module 1, and includes a chamber (not shown), a heating / cooling unit 111, and a measurement unit 112. The chamber accommodates the semiconductor module 1 (may be a joined body) therein. The heating / cooling unit 111 heats the inside of the chamber to a high temperature of, for example, 100 to 250 degrees, and cools the chamber to a low temperature of, for example, -40 degrees. The heating / cooling unit 111 may measure the temperature in the chamber using a temperature sensor (not shown). The heating / cooling unit 111 is controlled by the control device 130. The measuring unit 112 measures the strain of the semiconductor module 1 using, for example, a strain gauge bonded to the semiconductor module 1 or calculates the strain by an image interphase method. The measurement result is transmitted to the control device 130.

  In the H / C test, the inside of the chamber containing the semiconductor module 1 is heated by the heating / cooling unit 111 to maintain the high temperature for the holding time, and then the inside of the chamber is cooled to maintain the low temperature for the holding time. Hold. Since the stress applied to the bonding material 4 may change with time, the holding time is, for example, about 40 minutes depending on the scale of time change (depending on the material of the bonding material 4, the structure of the semiconductor module 1, etc.). The temperature change of one cycle is repeated for several thousand cycles in about 2 hours, and a heat load is applied to the semiconductor module 1.

  During the repetition of the heat cycle, the semiconductor module 1 is taken out from the chamber at regular intervals, and the state of fatigue of the bonding material 4, particularly the degree of crack growth, is observed using an ultrasonic flaw detector. The number of cycles in which the crack has reached a length of 1 mm is defined as a life cycle (also simply referred to as a life).

  Prior to the H / C test, a strain gauge is bonded to the joint portion of the semiconductor module 1, and the plastic strain applied to the joint portion of the semiconductor module 1 by the temperature change of one cycle using the strain gauge by the measurement portion 112 ( Simply referred to as strain). Alternatively, the strain may be calculated by an image interphase method. This plastic strain is proportional to the thermal load applied to the semiconductor module 1 per cycle (that is, the temperature difference between the high temperature and the low temperature). By measuring the plastic strain and the number of life cycles for several temperature differences, the relationship of the number of life cycles to the plastic strain, that is, a life curve is obtained.

  14A and 14B show a configuration of the mechanical fatigue test apparatus 120. FIG. Here, FIG. 14A and FIG. 14B show configurations in a side view and a top view, respectively. The mechanical fatigue test apparatus 120 is an apparatus for performing a mechanical fatigue test simulating an H / C test on the semiconductor module 1 by applying a mechanical load to the test sample 11. A chamber (not shown) and a support jig 121 are used. , A pushing jig 122, a heating / cooling unit 123 (see FIG. 13), and a measuring unit 124 (see FIG. 13).

  The chamber is a member that accommodates the test sample 11 and includes a support jig 121 and a pushing jig 122 inside.

  The support jig 121 is a member that supports the test sample 11 in the chamber, and includes a base 121a and two support members 121b. The base 121a is a rectangular parallelepiped block that receives a load applied to the test sample 11 via the support member 121b. The two support members 121b are semi-cylindrical members that support the test sample 11 on the base 121a. Each of the two support members 121b is perpendicular to the reference line LL on one side and the other side in the direction parallel to the reference line LL. In this direction, the side surface is in contact with the upper surface of the base 121a and is disposed on the base 121a. The distance between the fulcrums of the two support members 121b (that is, the distance between the fulcrums) can be arbitrarily changed. The test sample 11 is supported on the support jig 121 by supporting both ends of the test sample 11 from below by the two support members 121b. Both the base 121a and the two support members 121b can be formed using, for example, stainless steel.

  The support jig 121 is not limited to supporting the test sample 11 with the two support members 121b, and may support the test sample 11 with three or more support members 121b. Further, the shape of the support member 121b is not limited to a semi-cylindrical shape, and the test sample 11 may be supported by three or more fulcrums by three or more support members. Further, one end of the test sample 11 may be clamped and cantilevered.

  The pushing jig 122 is a member that pushes the test sample 11 supported on the supporting jig 121 downward from above. The pushing jig 122 has a round-top bar shape, is supported with the tip directed downward on the center of the two supporting members 121b of the supporting jig 121, and is driven by a driving device (not shown). It is paid out downward. A three-point bending deformation is given to the test sample 11 by clamping between the pushing jig 122 and the two support members 121b.

  Note that two or more pushing jigs 122 may be provided to give the test sample 11 four or more multi-point bending deformations. Further, the pushing jig 122 is not limited to a rod-shaped member, and may be a plate-shaped member having a semi-cylindrical end.

  The heating / cooling unit 123 heats or cools the inside of the chamber within a temperature range of −40 to 250 ° C., for example. The heating / cooling unit 123 may measure the temperature in the chamber using a temperature sensor (not shown). The heating / cooling unit 123 is controlled by the control device 130.

  The measurement unit 124 measures the strain of the test sample 11 using, for example, a strain gauge bonded to the test sample 11 or calculates the strain by an image interphase method. The measurement result is transmitted to the control device 130.

  In the mechanical fatigue test, first, the test sample 11 is placed on the support jig 121 in the chamber. Here, the length of the test sample 11 is oriented in a direction parallel to the reference line LL, the center of the sample bonded body 13b is positioned directly below the pushing jig 122, and both ends of the test sample 11 are held by the two support members 121b. To support.

  Next, the inside of the chamber containing the test sample 11 is heated or cooled by the heating / cooling unit 123 and held under a predetermined temperature condition. Here, the predetermined temperature condition is a temperature condition according to the condition in the H / C test, and is a temperature between a high temperature and a low temperature in one cycle.

  Next, the driving device is driven to feed the pushing jig 122 downward, abut against the center of the upper surface of the test sample 11 and press downward, and are surrounded by the fulcrums of the two support members 121b that support the test sample 11. The region in the upper surface of the test sample 11 is pushed relatively, thereby giving the test sample 11 a downward warping deformation. This is repeated as many times as the number of cycles in a cycle of 1 cycle time or less in the H / C test, and the test sample is repeatedly deformed to apply a mechanical load.

  During repeated deformation, the test sample 11 is taken out from the chamber at every fixed cycle, and the state of fatigue of the sample bonding material 14, particularly the degree of crack growth, is observed using an ultrasonic flaw detector. The number of cycles in which the crack has reached a length of 1 mm is defined as a life cycle (also simply referred to as a life).

  Prior to the mechanical fatigue test, a strain gauge is adhered to the sample base 12a or the sample bonded body 13b of the test sample 11, and the sample base 12a or the sample is subjected to deformation due to a single mechanical load applied by the measuring unit 124. The plastic strain generated in the sample bonding material 14 is calculated by measuring the plastic strain generated in the bonded body 13b and performing structural analysis such as FEM analysis so as to reproduce the measurement result. Alternatively, plastic strain generated in the sample base 12a or the sample bonded body 13b may be calculated by an image correlation method. This plastic strain is proportional to the mechanical load applied to the test sample 11. By measuring the plastic strain and the life cycle number for several mechanical loads, a relationship between the life cycle number and the plastic strain, ie, a life curve, can be obtained.

  In order to evaluate the life of the test sample 11 from the fatigue state of the sample bonding material 14, the number of cycles in which the crack generated in the sample bonding material 14 reaches a length of 1 mm was evaluated. You may decide. Further, the fatigue state is not limited to the cracks generated in the sample bonding material 14, and the life of the test sample 11 may be evaluated based on, for example, a change in the structure of the sample bonding material 14. Further, the life of the test sample 11 may be evaluated not only from the fatigue state of the sample bonding material 14 but also from the fatigue states of the constituent parts of the test sample 11 other than the sample bonding material 14 or the entire test sample 11.

  The control device 130 is a device that controls devices constituting the test system 100 and evaluates the reliability of the semiconductor module 1 based on the test results. The control device 130 includes a determination unit 132, a test unit 134, and an evaluation unit 136. The control device 130 expresses each functional unit by causing an information processing device including a computer, a microcontroller, etc. to execute a control program stored in a storage device such as a nonvolatile memory or a recording medium such as a CD-ROM. And function as a control device.

  The determination unit 132 determines a mechanical load that causes the semiconductor module 1 to change according to the deformation that occurs in the semiconductor module 1 that is the evaluation target in the H / C test. Details of the function of the determination unit 132 will be described later.

  The test unit 134 performs a mechanical fatigue test on the test sample 11 including the sample bonding material 14 corresponding to the bonding material 4 of the semiconductor module 1 using the mechanical fatigue test apparatus 120. Details of the function of the test unit 134 will be described later.

  The evaluation unit 136 evaluates the reliability of the structure unit 9 of the semiconductor module 1 based on the state of the sample bonding material 14 (or the entire sample) in the test sample 11 that has been repeatedly subjected to bending deformation by the test unit 134. Details of the function of the evaluation unit 136 will be described later.

  FIG. 15 shows a flow of a first test method using the test system 100 according to the present embodiment.

  In step S <b> 102, the determination unit 132 determines a mechanical load that causes the semiconductor module 1 to change according to the deformation that occurs in the semiconductor module 1 that is the evaluation target in the H / C test. Here, as a change corresponding to the deformation, a stress (including a stress distribution) applied to the bonding material 4 of the semiconductor module 1 or a strain (including a strain distribution) generated in the bonding material 4 due thereto is employed. By adopting these state quantities, the mechanical load applied to the test sample 11 in the mechanical fatigue test can easily correspond to the thermal load applied to the semiconductor module 1 in the H / C test, and the test sample 11 was used. The H / C test of the semiconductor module 1 can be reproduced by the mechanical fatigue test. However, the present invention is not limited to these, and a fine structure change, hardness (including hardness distribution), or the like may be adopted as a change according to deformation.

  The determining unit 132 determines the amount of change according to the deformation of one or more edges of the bonding material 4 in the H / C test for the semiconductor module 1. Here, the determination unit 132 uses the H / C test apparatus 110 to heat cycle the semiconductor module 1 (or a bonded body including the bonding material 4 and the metal base 2a and the insulating substrate 3 bonded thereby). In addition, the amount of change corresponding to the deformation of the bonding material 4, that is, the strain distribution generated at the edge due to the deformation is measured. The strain measurement is as described above. Alternatively, the determination unit 132 may calculate a stress distribution or a strain distribution applied to the edge of the bonding material 4 by a heat cycle by performing a thermal stress simulation.

  Then, the determination unit 132 determines a mechanical load that applies a measured strain distribution or a load corresponding to the calculated stress distribution or strain distribution to the edge of the sample bonding material 14 in the test sample 11. Here, the determination unit 132 calculates the mechanical load by structural analysis such as FEM analysis on the test sample 11, that is, the stress distribution or strain distribution applied to the edge of the sample bonding material 14 by the three-point bending test of the test sample 11. Of the three-point bending test in which the calculated stress distribution or strain distribution sufficiently matches the stress distribution or strain distribution applied to the edge of the bonding material 4 by the heat cycle previously calculated by the thermal stress simulation. Determine the conditions (see Table 1). In particular, the driving force (corresponding to a mechanical load or a load) for feeding the pushing jig 122 or the feeding amount is determined.

  In addition, since the crack of the bonding material 4 is generated from the end portion where the stress is concentrated, the coincidence of the stress distribution or the strain distribution is, for example, the stress and the stress gradient at the end portion of the bonding material 4 or the sample bonding material 14 corresponding thereto. Or you may judge by the agreement of the distortion | strain and distortion gradient corresponding to these.

  Note that the determination unit 132 may determine the mechanical load (load) using the strain sensitivity of FIG. The determination unit 132 measures or calculates the plastic strain generated at the edge portion of the bonding material 4 in the H / C test for the semiconductor module 1 as described above, and applies the mechanical strain to the strain sensitivity to generate a mechanical load that causes the plastic strain. decide.

  The determination unit 132 determines the mechanical load corresponding to each of a plurality of temperature conditions (that is, the heat cycle temperature) in the H / C test. The plurality of determined mechanical loads are transmitted to the test unit 134.

  In step S <b> 104, the mechanical fatigue test is performed on the test sample 11 including the sample bonding material 14 corresponding to the bonding material 4 of the semiconductor module 1 by the test unit 134 using the mechanical fatigue testing apparatus 120. Here, as described above, the test sample 11 includes the sample base 12a and the sample bonded body 13b formed from the same material as the metal base 2a, the metal plate 3b, and the bonding material 4 of the semiconductor module 1 as the evaluation object. , And a sample bonding material 14.

  The test unit 134 repeatedly warps the test sample 11 by a three-point bending test under a predetermined temperature condition with the same number of times as the number of heat cycles in the H / C test and a cycle equal to or less than the cycle time. give. Here, the temperature condition is any temperature between a high temperature and a low temperature in one cycle of the heat cycle depending on the conditions in the H / C test, and here, the high temperature in the heat cycle determined from the analysis according to FIG. To do. The mechanical load applied to the test sample 11 by the three-point bending test is a plurality of loads determined in step S102, and a mechanical fatigue test is performed on the test sample 11 for each load. The details of the three-point bending test of the test sample 11 by the mechanical fatigue test apparatus 120 are as described above, and the result (that is, the life cycle) is transmitted to the evaluation unit 136.

  In step S106, the evaluation unit 136 evaluates the reliability of the structure unit 9 of the semiconductor module 1 based on the state of the sample bonding material 14 (or the entire sample) in the test sample 11 repeatedly subjected to bending deformation in step S104. To do. Based on the result of the mechanical fatigue test obtained in step S104 for each of the plurality of mechanical loads determined in step 102, the evaluation unit 136 determines the sample bonding material 14 of the test sample 11 based on each of the plurality of mechanical loads. By obtaining a life cycle relationship with respect to the plastic strain generated in the semiconductor module 1, that is, a life curve, a relationship between the life cycle with respect to the plastic strain generated in the bonding material 4 of the semiconductor module 1 due to the heat cycle of a plurality of temperature conditions in the H / C test, that is, H / C Estimate life curve.

  In the first test method described above, the heat cycle in the H / C test is controlled by independently controlling the temperature field (that is, temperature distribution) and the stress field (that is, stress distribution) applied to the bonding material 4 of the semiconductor module 1. Thus, the stress applied to the semiconductor module 1 was reproduced, and the H / C life curve was estimated. Furthermore, by using the test sample 11 in which the semiconductor module 1 is simplified, it is possible to evaluate the reliability of the semiconductor module 1 with only a small factor regardless of the details of the design. Here, since there are few factors for reliability evaluation, a temperature curve and a stress field are given to the small factors in advance using the test sample 11 to construct a life curve database, and the semiconductor to be evaluated from the database. By selecting a life curve corresponding to the module 1, the reliability can be evaluated without performing a reliability test.

  FIG. 16 shows a flow of a second test method using the test system 100 according to the present embodiment.

  In step S202, a mechanical fatigue test is performed by the test unit 134 to give different mechanical loads to each of the plurality of test samples 11 using the mechanical fatigue test apparatus 120, and a life curve database is created.

  As described above, the test sample 11 has a structure portion corresponding to the structure portion 9 of the semiconductor module 1 in which the sample base 12a and the sample bonded body 13b are joined by the sample joining material 14. The test sample 11 is used for the material of the sample base 12a (for example, copper, aluminum silicon carbide composite material etc. used for the metal base 2a) and the material for the sample bonded body 13b (for example, for the metal plate 3b). A plurality of combinations of the material of the sample bonding material 14 (for example, solder such as Sn-5 wt% Sb employed for the bonding material 4). In addition, in order to control the stress distribution (see Table 1), for example, the size and thickness of the sample base 12a, the size and thickness of the sample bonded body 13b within the range determined from FIGS. 8 to 11, and The test sample 11 may be prepared for the design conditions of the test sample 11 such as the shape of the fillet of the sample bonding material 14.

  Using the mechanical fatigue test apparatus 120, the test unit 134 repeatedly applies warp deformation by a three-point bending test to each of the plurality of test samples 11. Here, as test conditions (see Table 1), some of the temperature conditions within the temperature range determined in the H / C test, the load applied to the test sample 11 by the three-point bending test, and the distance between the fulcrums of the supporting jig 121 A combination is defined and a three point bend test is performed for each condition. With these conditions and further test conditions including the design conditions of the test sample 11 described above, the amount of change according to the deformation applied to the test sample by the three-point bending test, such as stress, stress gradient, strain, strain gradient, and temperature, Be controlled. The details of the three-point bending test of the test sample 11 by the mechanical fatigue test apparatus 120 are as described above, and the result (that is, the life cycle) is transmitted to the evaluation unit 136.

  In addition, according to the test conditions described above, a one-dimensional distribution, two-dimensional distribution, or three-dimensional distribution of stress or strain in a predetermined range of the test sample 11, for example, an end portion (that is, a fillet surface) of the sample bonding material 14 is obtained. It is good also as controlling. Here, the one-dimensional distribution is, for example, a distribution on a straight line from the upper end portion (a point on the fillet surface) of the sample bonding material 14 where stress is concentrated at a high temperature to the inside. The two-dimensional distribution is a distribution on a cross section from the end face (fillet surface) of the sample bonding material 14 toward the inside. The three-dimensional distribution is a distribution in a region including an end portion (fillet) of the sample bonding material 14. When an impact test is employed as the reliability test, a strain rate may be employed as the amount of change corresponding to the deformation.

  The evaluation unit 136 calculates a change amount according to the deformation applied to the sample bonding material 14 of the test sample 11 by the three-point bending test with respect to the above-described test conditions by structural analysis such as FEM analysis. For example, the evaluation unit 136 calculates a two-dimensional distribution of stress (or strain) at the end of the sample bonding material 14, selects a point at which stress (strain) is most concentrated on the fillet surface, and starts internal processing from that point. Stress and stress gradient (strain and strain gradient) may be calculated from a one-dimensional distribution on a straight line toward By editing the result of the mechanical fatigue test by the test unit 134 using the calculation result, the amount of change (for example, the end portion of the sample bonding material 14) according to the material, temperature condition, and deformation of each component of the test sample 11 The relationship of the life cycle to the plastic strain generated in the sample bonding material 14 (ie, the life curve) is obtained. The plastic strain generated in the sample bonding material 14 is determined from the load applied to the test sample 11. The evaluation unit 136 stores the obtained life curve in a storage device (not shown) and creates a database.

Table 2 shows test conditions for selecting a life curve in the database created from the test results of the mechanical fatigue test using the test sample 11. With respect to the material of the sample base 12a, the material of the sample bonded body 13b, the material and thickness of the sample bonding material 14, the material of the metal base 2a of the semiconductor module 1 to be evaluated, the material of the metal plate 3b, and the bonding material 4 Select the one that matches the material and thickness. For stress and stress gradient (or strain or strain gradient), those corresponding to the stress distribution (or strain distribution generated in the bonding material 4) applied to the bonding material 4 of the semiconductor module 1 by the heat cycle in the H / C test. select. For the temperature, any temperature between a high temperature and a low temperature of the heat cycle temperature is selected. Thereby, a life curve that reproduces the result of the H / C test of the semiconductor module 1 is extracted from the database.

  If the life curve can be selected according to the above test conditions, a database is created from the test results of the H / C test of the semiconductor module 1 as well as the test results of the mechanical fatigue test by the three-point bending test of the test sample 11. Alternatively, a database may be created by combining the results of both tests. If a life curve database is created for the test conditions shown in Table 2, a life curve obtained by an arbitrary test method can be taken into the database and used for designing the life of the semiconductor module 1 having an arbitrary configuration. can do.

  In step S204, the determination unit 132 changes a test sample according to the deformation generated in the bonded structural unit 9 of the semiconductor module 1 in the H / C test from the results of the plurality of mechanical fatigue tests performed in step S202. The test given in 11 is determined. The determining unit 132 determines the material of the metal base 2a of the semiconductor module 1 to be evaluated and the material of the metal plate 3b with respect to the material of the sample base 12a, the material of the sample bonded body 13b, and the material and thickness of the sample bonding material 14. A material matching the material and thickness of the bonding material 4 is selected. The determination unit 132 performs, for example, a thermal stress simulation, a change amount according to the deformation given to the bonding material 4 of the semiconductor module 1 to be evaluated by the heat cycle of the H / C test, for example, stress and stress distribution (or strain or strain). (Gradient) is calculated and the one that matches or is equal to the calculated one is selected. The coincidence of the change amounts is not limited to a corner portion in the sample bonding material 14, and may be determined for a plurality of locations. The determination unit 132 selects a high heat cycle temperature with respect to the temperature. Thereby, a test giving a mechanical load (load) corresponding to the calculated change amount is determined from a plurality of mechanical fatigue tests. Note that the mechanical load corresponding to the calculated change amount is not necessarily the same load, and may be the most equal load. The determined test is transmitted to the evaluation unit 136.

  In step S206, the evaluation unit 136 evaluates the reliability of the bonded structural unit 9 of the semiconductor module 1 based on the test result of the determined test. The evaluation unit 136 extracts the test for the test determined in step S204 from the life curves in the database created in step S202, and outputs them to a display device (not shown). Thereby, the user can evaluate the reliability of the semiconductor module 1 by the H / C test.

  In the second test method described above, a life curve database was created by performing a mechanical fatigue test on the test sample 11 using the mechanical fatigue test apparatus 120 prior to evaluating the reliability of the semiconductor module 1. May be obtained, and the reliability of the semiconductor module 1 may be evaluated using the database.

  FIG. 17 shows a configuration of an evaluation system 200 according to a modification. The evaluation system 200 is a system for evaluating the reliability of the semiconductor module 1 using a life curve database, and includes a storage device 240 and an evaluation device 230.

  The storage device 240 is a device that stores a database of life curves, and is connected so as to be communicable with the evaluation device 230 (the acquisition unit 238 included therein). The storage device 240 may be a reading device that can read a recording medium that stores a database of life curves. For example, it is assumed that the database is created by the same procedure as in step S202 described above.

  The evaluation device 230 includes an acquisition unit 238, a determination unit 232, and an evaluation unit 236.

  The acquisition unit 238 accesses the storage device 240 and acquires a test result (that is, a life curve) of the mechanical fatigue test.

  The determination unit 232 and the evaluation unit 236 are configured similarly to the above-described determination unit 132 and evaluation unit 136, respectively.

  FIG. 18 shows a flow of an evaluation method using the evaluation system 200 according to the modification.

  In step 302, each of the plurality of test samples 11 having a bonded portion in which the sample bonded material 13b is bonded to the sample base 12a by the sample bonding material 14 from the database stored in the storage device 240 by the acquisition unit 238. Test results (that is, life curves) of a plurality of mechanical fatigue tests with different mechanical loads are obtained. Here, the acquisition unit 238 has a temperature condition within a temperature range determined in the H / C test for each of the plurality of test samples, a load applied to the test sample 11 by a three-point bending test, and a fulcrum of the support jig 121. Further, the design conditions of the test sample 11 (the size and thickness of the sample base 12a, the size and thickness of the sample bonded body 13b, and the thickness of the sample bonding material 14 (the total thickness of the test sample 11) Or at least one of the shape of the fillet, etc.) to obtain test results of a plurality of mechanical fatigue tests in which different mechanical loads are applied.

  Note that the amount of change according to the deformation applied to the test sample by the three-point bending test, such as stress, stress gradient, strain, strain gradient, and temperature, is controlled by the above conditions. Furthermore, a predetermined range of the test sample 11, for example, a one-dimensional distribution, a two-dimensional distribution, or a three-dimensional distribution of stress or strain at the end of the sample bonding material 14 (ie, the fillet surface) may be controlled. Details of these distributions are as described above.

  In step 304, the determination unit 232 determines a test that gives the test sample 11 a change corresponding to the deformation generated in the structure unit 9 to which the semiconductor module 1 is bonded in the H / C test from among a plurality of mechanical fatigue tests. To do. The determination unit 232 calculates the amount of change according to the deformation given to the bonding material 4 of the semiconductor module 1 to be evaluated by the heat cycle of the H / C test by structural analysis such as FEM analysis, and the calculated amount of change A test with a mechanical load (load) corresponding to is determined from a plurality of mechanical fatigue tests. Note that the mechanical load corresponding to the calculated change amount is not necessarily the same load, and may be the most equal load. The determined test is transmitted to the evaluation unit 236.

  In step S306, the evaluation unit 236 evaluates the reliability of the portion of the bonding material 4 of the semiconductor module 1 based on the determined test result of the mechanical fatigue test. The evaluation unit 236 extracts one for the test determined in step S304 from the test results (that is, life curves) of the plurality of mechanical fatigue tests acquired in step S302, and outputs them on a display device (not shown). To do. Thereby, the user can evaluate the reliability of the semiconductor module 1 by the H / C test.

  In the test method and the evaluation method according to the present embodiment, solder is used as the bonding material 4 for bonding the insulating substrate 3 to the metal base 2a in the semiconductor module 1 which is an example of the evaluation object. The metal may be used. Alternatively, the insulating substrate 3 may be bonded to the metal base 2a using a resin material. The evaluation object is a semiconductor module including a joined body of two metals, that is, the metal plate 3b and the metal base 2a included in the insulating substrate 3. However, the present invention is not limited to this, and two resin materials are bonded. A semiconductor module including a joined body may be used. The evaluation object may be a structure having an arbitrary configuration in which a plurality of members are bonded by an arbitrary bonding material.

  In the test method and the evaluation method according to the present embodiment, the heat cycle is applied as the reliability test of the semiconductor module. However, the present invention is not limited to this, and an arbitrary reliability test such as a power cycle test may be applied.

  Various embodiments of the invention may be described with reference to flowcharts and block diagrams, where a block is either (1) a stage in a process in which the operation is performed or (2) an apparatus responsible for performing the operation. May represent a section of Certain stages and sections are implemented by dedicated circuitry, programmable circuitry supplied with computer readable instructions stored on a computer readable medium, and / or processor supplied with computer readable instructions stored on a computer readable medium. It's okay. Dedicated circuitry may include digital and / or analog hardware circuitry and may include integrated circuits (ICs) and / or discrete circuits. Programmable circuits include memory elements such as logical AND, logical OR, logical XOR, logical NAND, logical NOR, and other logical operations, flip-flops, registers, field programmable gate arrays (FPGA), programmable logic arrays (PLA), etc. Reconfigurable hardware circuitry, including and the like.

  Computer readable media may include any tangible device capable of storing instructions to be executed by a suitable device, such that a computer readable medium having instructions stored thereon is specified in a flowchart or block diagram. A product including instructions that can be executed to create a means for performing the operation. Examples of computer readable media may include electronic storage media, magnetic storage media, optical storage media, electromagnetic storage media, semiconductor storage media, and the like. More specific examples of computer readable media include floppy disks, diskettes, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM or flash memory), Electrically erasable programmable read only memory (EEPROM), static random access memory (SRAM), compact disc read only memory (CD-ROM), digital versatile disc (DVD), Blu-ray (RTM) disc, memory stick, integrated A circuit card or the like may be included.

  Computer readable instructions can be assembler instructions, instruction set architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state setting data, or object oriented programming such as Smalltalk, JAVA, C ++, etc. Including any source code or object code written in any combination of one or more programming languages, including languages and conventional procedural programming languages such as "C" programming language or similar programming languages Good.

  Computer readable instructions may be directed to a general purpose computer, special purpose computer, or other programmable data processing device processor or programmable circuit locally or in a wide area network (WAN) such as a local area network (LAN), the Internet, etc. The computer-readable instructions may be executed to create a means for performing the operations provided via and specified in the flowchart or block diagram. Examples of processors include computer processors, processing units, microprocessors, digital signal processors, controllers, microcontrollers, and the like.

  FIG. 19 illustrates an example of a computer 2200 in which aspects of the present invention may be embodied in whole or in part. The program installed in the computer 2200 can cause the computer 2200 to function as an operation associated with the apparatus according to the embodiment of the present invention or one or more sections of the apparatus, or to perform the operation or the one or more sections. The section can be executed and / or the computer 2200 can execute a process according to an embodiment of the present invention or a stage of the process. Such a program may be executed by CPU 2212 to cause computer 2200 to perform certain operations associated with some or all of the blocks in the flowcharts and block diagrams described herein.

  A computer 2200 according to this embodiment includes a CPU 2212, a RAM 2214, a graphic controller 2216, and a display device 2218, which are connected to each other by a host controller 2210. Computer 2200 also includes input / output units such as communication interface 2222, hard disk drive 2224, DVD-ROM drive 2226, and IC card drive, which are connected to host controller 2210 via input / output controller 2220. Yes. The computer also includes legacy input / output units, such as ROM 2230 and keyboard 2242, which are connected to input / output controller 2220 via input / output chip 2240.

  The CPU 2212 operates according to programs stored in the ROM 2230 and the RAM 2214, thereby controlling each unit. The graphic controller 2216 obtains the image data generated by the CPU 2212 in a frame buffer or the like provided in the RAM 2214 or itself so that the image data is displayed on the display device 2218.

  The communication interface 2222 communicates with other electronic devices via a network. The hard disk drive 2224 stores programs and data used by the CPU 2212 in the computer 2200. The DVD-ROM drive 2226 reads a program or data from the DVD-ROM 2201 and provides the program or data to the hard disk drive 2224 via the RAM 2214. The IC card drive reads programs and data from the IC card and / or writes programs and data to the IC card.

  The ROM 2230 stores therein a boot program executed by the computer 2200 at the time of activation and / or a program depending on the hardware of the computer 2200. The input / output chip 2240 may also connect various input / output units to the input / output controller 2220 via parallel ports, serial ports, keyboard ports, mouse ports, and the like.

  The program is provided by a computer-readable medium such as a DVD-ROM 2201 or an IC card. The program is read from a computer-readable medium, installed in the hard disk drive 2224, the RAM 2214, or the ROM 2230, which are also examples of the computer-readable medium, and executed by the CPU 2212. Information processing described in these programs is read by the computer 2200 to bring about cooperation between the programs and the various types of hardware resources. An apparatus or method may be configured by implementing information manipulation or processing in accordance with the use of computer 2200.

  For example, when communication is performed between the computer 2200 and an external device, the CPU 2212 executes a communication program loaded in the RAM 2214 and performs communication processing on the communication interface 2222 based on processing described in the communication program. You may order. The communication interface 2222 reads the transmission data stored in the transmission buffer processing area provided in the recording medium such as the RAM 2214, the hard disk drive 2224, the DVD-ROM 2201, or the IC card under the control of the CPU 2212, and the read transmission. Data is transmitted to the network, or received data received from the network is written in a reception buffer processing area provided on the recording medium.

  Further, the CPU 2212 allows the RAM 2214 to read all or a necessary part of a file or database stored in an external recording medium such as a hard disk drive 2224, a DVD-ROM drive 2226 (DVD-ROM 2201), an IC card, etc. Various types of processing may be performed on the data on the RAM 2214. Next, the CPU 2212 writes back the processed data to the external recording medium.

  Various types of information, such as various types of programs, data, tables, and databases, may be stored on a recording medium and subjected to information processing. The CPU 2212 describes various types of operations, information processing, conditional judgment, conditional branching, unconditional branching, information retrieval, which are described in various places in the present disclosure and specified by the instruction sequence of the program with respect to the data read from the RAM 2214. Various types of processing may be performed, including / replacement etc., and the result is written back to the RAM 2214. Further, the CPU 2212 may search for information in files, databases, etc. in the recording medium. For example, when a plurality of entries each having an attribute value of the first attribute associated with the attribute value of the second attribute are stored in the recording medium, the CPU 2212 specifies the attribute value of the first attribute. The entry that matches the condition is searched from the plurality of entries, the attribute value of the second attribute stored in the entry is read, and thereby the first attribute that satisfies the predetermined condition is associated. The attribute value of the obtained second attribute may be acquired.

  The programs or software modules described above may be stored on a computer readable medium on or near computer 2200. In addition, a recording medium such as a hard disk or a RAM provided in a server system connected to a dedicated communication network or the Internet can be used as a computer-readable medium, thereby providing a program to the computer 2200 via the network. To do.

  As mentioned above, although this invention was demonstrated using embodiment, the technical scope of this invention is not limited to the range as described in the said embodiment. It will be apparent to those skilled in the art that various modifications or improvements can be added to the above-described embodiment. It is apparent from the scope of the claims that the embodiments added with such changes or improvements can be included in the technical scope of the present invention.

  The order of execution of each process such as operations, procedures, steps, and stages in the apparatus, system, program, and method shown in the claims, the description, and the drawings is particularly “before” or “prior to”. It should be noted that the output can be realized in any order unless the output of the previous process is used in the subsequent process. Regarding the operation flow in the claims, the description, and the drawings, even if it is described using “first”, “next”, etc. for convenience, it means that it is essential to carry out in this order. It is not a thing.

1 ... semiconductor module, 2 ... case, 2a ... metal base, 2b ... frame, 2c ... lid, 3 ... insulating substrate, 3a ... insulating plate, 3b ... metal plate, 3c ... wiring _board, 3c 1, _3c 2, 3c ₃ ... circuit pattern, 4 ... bonding material, 5 ... semiconductor element, 5a ... bonding material, 6a, 6b, 6c ... terminal, 7b, 7c ... wire, 8 ... gel filler, 9 ... structure part, 11 ... test sample , 12a ... sample substrate, 13b ... sample bonded material, 14 ... sample bonding material, 100 ... test system, 110 ... H / C test device, 111 ... heating / cooling unit, 112 ... measurement unit, 120 ... mechanical fatigue test device, DESCRIPTION OF SYMBOLS 121 ... Support jig, 121a ... Base, 121b ... Support member, 122 ... Pushing jig, 123 ... Heating / cooling part, 124 ... Measuring part, 130 ... Control apparatus, 132 ... Determination part, 134 ... Test part, 136 ... Evaluation Part DESCRIPTION OF SYMBOLS 200 ... Evaluation system, 230 ... Evaluation apparatus, 232 ... Determination part, 236 ... Evaluation part, 238 ... Acquisition part, 240 ... Memory | storage device, 2200 ... Computer, 2201 ... ROM, 2210 ... Host controller, 2212 ... CPU, 2214 ... RAM 2216 ... Graphic controller, 2218 ... Display device, 2220 ... Output controller, 2222 ... Communication interface, 2224 ... Hard disk drive, 2226 ... ROM drive, 2230 ... ROM, 2240 ... Output chip, 2242 ... Keyboard.

Claims (35)

A determination step of determining a mechanical load that causes the evaluation object to change in accordance with deformation generated in the evaluation object in which the bonding material is bonded to the substrate in the reliability test;
Applying a load including the mechanical load to a test sample including a sample bonding material corresponding to the bonding material;
A test method comprising:   The test method according to claim 1, wherein the reliability test is a heat cycle test or a power cycle test. The evaluation object has a structure part in which the base body and the object to be joined are joined by the joining material,
The test method according to claim 1, wherein the test sample has a structure portion in which a sample substrate corresponding to the substrate and a sample bonded body corresponding to the bonded body are bonded by the sample bonding material. Each of the base body and the sample base body, the bonded body and the sample bonded body, and the bonding material and the sample bonding material are formed of the same material,
The test method according to claim 3, wherein the sample bonding material has a thickness corresponding to a corresponding edge of the bonding material at one or more edges. The evaluation object is
A metal base as the substrate;
An insulating substrate having a circuit pattern on the front surface and a metal plate bonded to the back surface;
A semiconductor element mounted on the circuit pattern,
5. The test method according to claim 3, wherein the metal plate as the object to be joined and the metal base have a structure part joined by the joining material.   6. The mechanical load according to claim 3, wherein in the determination step, the mechanical load that mechanically provides a change amount according to deformation of one or more edges of the bonding material in the reliability test is determined. Test method. In the determining step,
Measure the amount of change according to the deformation of the evaluation object during the reliability test,
The test method according to any one of claims 3 to 6, wherein the mechanical load that gives a load corresponding to the measured variation is determined for the test sample. In the determining step,
In the reliability test, the amount of change corresponding to deformation at one or more edges is calculated by analysis,
The test method according to any one of claims 3 to 6, wherein the mechanical load that applies a load corresponding to the calculated change amount to the one or more edges of the bonding material is determined.   The test method according to claim 7 or 8, wherein in the determination of the mechanical load, the mechanical load that mechanically applies a load corresponding to the change amount to the test sample is calculated by analysis. The test step is performed by applying a warp deformation to the test sample by relatively pressing a region surrounded by two or more fulcrums in the plane of the sample substrate,
10. The sample-bonded body is a plate-like member having a corner portion where two sides approach each other toward a direction in which a radius of curvature due to the warp deformation is minimized within the surface of the sample base. The test method as described in any one. The sample substrate and the sample bonded body are square,
The test method according to claim 10, wherein one side of the outer periphery of the sample bonded body is directed in a direction of a predetermined angle with respect to one side of the outer periphery of the sample substrate. The test method according to claim 11, wherein the predetermined angle is 45 degrees. The test method according to any one of claims 3 to 12, wherein in the test stage, the test sample is deformed repeatedly by the same number of times as the number of cycles of the reliability test.   The test method according to claim 3, wherein in the test stage, the test sample is repeatedly deformed under a predetermined temperature condition. The reliability test is a heat cycle test,
The test method according to claim 14, wherein in the test stage, the test sample is repeatedly deformed under a temperature condition corresponding to a condition in the heat cycle test. The reliability test is a heat cycle test,
The test method according to any one of claims 3 to 15, wherein, in the test stage, the test sample is repeatedly deformed at a cycle equal to or less than a cycle time in the heat cycle test.   17. The method according to claim 13, further comprising an evaluation step of evaluating reliability of the joined structure portion of the evaluation object based on a state of the joined structure portion in the repeatedly bent test sample. Test method according to item. In the evaluation stage, a heat cycle life curve is estimated based on a result of determining the mechanical load and performing the repeated deformation corresponding to each of heat cycle tests of a plurality of temperature conditions. The test method according to 17. The test stage performs a plurality of tests that give different mechanical loads to each of the plurality of test samples, prior to the determination stage,
The determining step determines a test that gives the test sample a change corresponding to a deformation generated in the joined structure portion of the evaluation object in the reliability test from the plurality of tests,
The test method according to any one of claims 1 to 18, further comprising an evaluation step of evaluating the reliability of the joined structure portion of the evaluation object based on the determined test result of the test.   The test method according to claim 19, wherein the determining step determines a test given a load corresponding to a change amount according to deformation of the evaluation object in the reliability test from the plurality of tests.   21. The test method according to claim 7, wherein at least one of stress, stress gradient, strain, strain gradient, and temperature is used as the amount of change corresponding to the deformation.   The amount of change according to the deformation is a one-dimensional distribution of stress or strain, a two-dimensional distribution, or a three-dimensional distribution within a predetermined range of the evaluation object and the test sample. The test method according to any one of 20 and 21. The test sample has a structure in which a sample base and a sample bonded body are bonded by the sample bonding material,
The test step includes, for each of a plurality of the test samples, a load applied to the sample substrate, a distance between a force point and a fulcrum that applies a load determined in a plane of the sample substrate, and a thickness of the sample substrate. A plurality of tests applying a mechanical load using the test sample in which at least one of the thickness of the sample bonded body, the thickness of the structure bonded by the sample bonding material, and the size of the sample bonded body is different The test method according to any one of claims 19 to 22. A test sample for testing by the test method according to any one of claims 1 to 23,
A sample substrate corresponding to the substrate in the evaluation object having a structure in which the substrate and the object to be bonded are bonded by the bonding material;
A sample bonded body corresponding to the bonded body;
A test sample comprising: a sample bonding material for bonding the sample substrate and the sample bonded body. The test step in the test method is performed by relatively pushing a region surrounded by two or more fulcrums in the surface of the sample substrate to give warpage deformation to the test sample,
25. The sample bonded body is a plate-like member having a corner portion where two sides approach each other in a direction in which the radius of curvature due to the warp deformation is minimized within the surface of the sample base. Test sample. The sample substrate and the sample bonded body are square,
The test sample according to claim 25, wherein one side of the outer periphery of the sample bonded body is directed in a direction of a predetermined angle with respect to one side of the outer periphery of the sample base. 27. A test sample according to claim 26, wherein the predetermined angle is 45 degrees.   A test system for performing a test by the test method according to any one of claims 1 to 23. Obtaining a test result of a plurality of tests with different mechanical loads applied to each of the plurality of test samples including the sample bonding material;
A determination step of determining a test in which a change in accordance with a deformation generated in the bonded structure portion of the evaluation object in which the bonding material is bonded to the base is given to the test sample in the reliability test among the plurality of tests; ,
An evaluation method comprising: an evaluation step of evaluating the reliability of the portion of the bonding material of the evaluation object based on the determined test result of the test.   30. The evaluation according to claim 29, wherein the determining step determines a test that mechanically applies a load corresponding to a change amount according to deformation of the evaluation object in the reliability test from the plurality of tests. Method.   The evaluation method according to claim 30, wherein at least one of stress, stress gradient, strain, strain gradient, and temperature is used as the amount of change according to the deformation.   The amount of change according to the deformation is a one-dimensional distribution, two-dimensional distribution, or three-dimensional distribution of stress or strain within a predetermined range of the evaluation object and the test sample. The evaluation method described. The test sample has a structure in which a sample base and a sample bonded body are bonded by the sample bonding material,
In the obtaining step, for each of the plurality of test samples, a load applied to the sample substrate, a distance between a force point and a fulcrum that gives a load determined in the plane of the sample substrate, a thickness of the sample substrate A test result of the plurality of tests in which at least one of the thickness of the sample bonded body, the thickness of the structure bonded by the sample bonding material, and the size of the sample bonded body is subjected to different mechanical loads. The evaluation method according to any one of claims 29 to 32.   An evaluation system for executing the evaluation method according to any one of claims 29 to 33.   An evaluation program for causing a computer to execute the evaluation method according to any one of claims 29 to 33.