## Abstract

Very high temperature reactors (VHTRs) are planned to be operated between 550 and $950∘C$ and demand a thermally efficient intermediate heat exchanger (IHX) in the heat transport system (HTS). The current technological development of compact heat exchangers (CHXs) for VHTRs is at the “proof of concept” level. A significant development in the CHX technologies is essential for the VHTRs to be efficient, cost-effective, and safe. CHXs have very high thermal efficiency and compactness, making them a prime candidate for IHXs in VHTRs. Photochemically etched plates with the desired channel pattern are stacked and diffusion bonded to fabricate CHXs. All plates are compressed at an elevated temperature over a specified period in the diffusion bonding process, promoting atomic diffusion and grain growth across bond surfaces resulting in a monolithic block. The diffusion bonding process changes the base metal properties, which are unknown for Alloy 800H, a candidate alloy for CHX construction. Hence, developing mechanical response data and understanding failure mechanisms of diffusion bonded Alloy 800H at elevated temperatures is a key step for advancing the technology of IHXs in VHTRs. The ultimate goal of this study is to develop ASME BPVC Section III, Division 5 design rules for CHXs in nuclear service. Toward this goal, mechanical performance and microstructures of diffusion bonded Alloy 800H are investigated through a series of tensile, fatigue, creep, and creep-fatigue tests at temperatures 550 to $760∘C$. The test results, failure mechanisms, and microstructures of diffusion bonded Alloy 800H are scrutinized and presented.

## 1 Introduction

Generation IV nuclear power plants are sustainable, economical, efficient, and produce minimal waste with enhanced safety [1]. These Gen IV reactors comprised various concepts with operating temperatures between 550 and $950∘C$, hence also known as very high-temperature reactors (VHTRs) [13]. High operation temperatures of VHTRs impose challenges in designing heat transport systems. Intermediate heat exchangers (IHXs) transfer heat from primary loop to secondary loop serving a key role in the heat transport system in VHTRs [4]. The traditional heat exchanger design concepts, such as tube-in-tube or shell-in-tube, result in bulky and inefficient heat exchangers, and thus, electricity generation becomes uneconomic [5,6]. The IHX technological development for VHTRs scored 3 out of 10, implying the significant need for technological development [2]. Hence, the design concepts of compact heat exchangers (CHXs) became a prime candidate for VHTRs.

First step in the CHX fabrication involves removing surface impurities, and lamination of polymer-based coating everywhere except the surfaces where channels will be etched [7]. These plates are immersed in etching solution to etch desired channel pattern. Alkaline solution dissolves the polymer-based coating from the plates. Next, a thin layer of Nickel is applied on plates to promote atomic diffusion [8,9]. A Gleeble system applies compressive stress to the stacked plates at elevated temperature for specified time [10,11]. These conditions promote the diffusion across the plate interface resulting in a solid bond between adjacent plates [12]. These fabrication steps provide a solid monolithic core of the CHX with thousands of channels of 0.5 to 2 mm and spacing/pitch of 0.5–2 mm [5,9,10]. This geometric configuration results in a highly compact and thermally efficient heat exchanger. Higher compactness and high thermal efficiency made CHX an attractive option for VHTRs.

Diffusion bonding results in the grain growth changing the base metal mechanical properties. Assessing the mechanical properties of diffusion bonded alloys is a key step for designing CHX for Gen IV nuclear power plants. A very few studies in the literature evaluated properties of diffusion bonded alloys through tensile and creep tests [9,13,14]. Li et al. [13] studied the bond performance of Alloy 617 through tensile tests at room temperature. They demonstrated ductile fracture and satisfied the minimum strength criteria at room temperature. A series of tension tests by Sah et al. [15] on diffusion bonded Alloy 617 at different temperatures showed bond delamination fracture mechanism. Mylavarapu et al. [9] conducted tensile and creep tests at room and elevated temperatures and demonstrated that the tensile strengths of diffusion bonded Alloy 617 at room temperature and $650∘C$ are weaker compared to the wrought Alloy 617. The diffusion bond tensile strength satisfied the minimum strength criteria at room temperature, but the % elongation was significantly smaller than the wrought alloy. The diffusion bonded Alloy 617 creep test at $900∘C$ failed prematurely with bond delamination fracture. Sah et al. [14] studied diffusion bonded Inconel 617 and Haynes 230 alloys through a series of tension tests at room temperature and $900∘C$ and showed that at elevated temperature, bond delamination is the primary rupture mode despite reasonable elongation of the specimens. They also demonstrated that the post diffusion bond heat treatment does not change this failure mode. Rozman et al. [16] demonstrated a similar rupture mode of diffusion bonded Haynes 230 through a series of creep tests at $800∘C$. Kim et al. showed that cold working on plates prior to bonding seems to enhance mechanical properties of Alloy 600; however, Ti precipitates hinder diffusion process [17]. The theoretical framework of diffusion bonded Alloy 800H studied by Mizia et al. [18] demonstrated microstructural evolution during the diffusion bonding processes of Alloy 800H. The above investigations on diffusion bonded alloys are based on a limited number of tensile and/or creep experiments at elevated temperatures. None of these studies accounted for the fatigue and creep-fatigue performances at different temperatures. VHTRs are planned to be operated between 500 and $1000∘C$ and anticipated to experience transient loading that may induce thermo-mechanical fatigue and creep-fatigue failure mechanisms. Hence, it is essential to investigate the fatigue and creep-fatigue failure mechanisms of diffusion bonded alloys, in addition to their tensile and creep failure mechanisms, in order to be able to effectively design and evaluate the performance of CHXs in VHTRs.

The ultimate goal is to establish the ASME BPVC Sec. 3, Division 5 design rules for CHXs. Keating et al. [19] identified the current gaps in ASME Section III, Division 5 nuclear code for design and analysis of CHXs. Lack of experimental data on diffusion bonded alloys is one of the critical gaps in the current ASME design code relevant to CHX. During the inception of this study, performance evaluation of diffusion bonded Alloy 617 or 316H or 800H was considered. Preliminary data by authors and industry experiences point to diffusion bonding of Alloy 800H are most promising [20,21]. Hence, this study evaluates the mechanical performance evaluation of the diffusion bonded Alloy 800H through a systematic set of tensile, fatigue, creep, and creep-fatigue tests on standard ASTM specimens at temperatures 550–760$∘C$. In addition, the microstructural examinations of diffusion bonded specimens through optical, scanning electron, and backscatter electron microscope studies of specimens before and after mechanical tests are performed. The experimental results, rupture surfaces, and microstructure characterization are presented and discussed.

## 2 Experimental Plan and Analysis Procedures

### 2.1 Materials and Specimens.

This study started with testing of wrought base metal Alloy 800H (BM 800H) procured as hot-rolled bars and plates annealed at $1176∘C$ for 9.7 min. The average grain size of BM 800H plates before diffusion bonding was 75 μm, and the chemical composition is as shown in Table 1. Next, a set of base metal tensile, fatigue, and creep-fatigue specimens were machined from the procured bars and plates with ASTM specified dimensions as shown in Fig. 1. Finally, tests on BM 800H specimens were conducted at room and elevated temperatures to determine the baseline performance of the alloy.

Fig. 1
Fig. 1
Close modal
Table 1

Chemical composition of BM 800H plates used in fabricating a diffusion bonded block from which test specimens were machined

MaterialFe%Ni%Cr%C%Al%Ti%Mn%Si%Cu%S%
BM45.231.220.10.070.480.520.90.20.310.002
DB47.030.319.70.080.430.580.70.40.190.0003
MaterialFe%Ni%Cr%C%Al%Ti%Mn%Si%Cu%S%
BM45.231.220.10.070.480.520.90.20.310.002
DB47.030.319.70.080.430.580.70.40.190.0003

A total of six diffusion bonding trials with 1.6 mm thick Alloy 800H plates were conducted by a leading diffusion bonding company in the USA, Vacuum Process Engineering (VPE). Diffusion bonding was performed as per ASME Code Section VIII, Division 1, Code Case 2621-1 [22]. The specified techniques and parameters used in diffusion bonding are the company’s proprietary information and hence cannot be presented here. For each diffusion bonding trial, room temperature tension tests were performed to determine the yield strength (YS), ultimate tensile strength (UTS), and % elongation. The test data from the sixth trial specimens always satisfied the ASME minimum YS and UTS requirements, in addition, the % elongation exceeded the ASME requirement of 30%. Hence, the sixth trial diffusion bonding technique and parameters were selected for this study to fabricate a diffusion bonded solid block of 200 × 200 × 200 mm. Wire electrical discharge machining (wire EDM) was used to fabricate cylindrical blanks from the solid blocks. The cylindrical blanks were then machined to ASTM standard specimens shown in Fig. 1. The diffusion bonded planes of the specimens were always perpendicular to the loading axis, except for a few tensile specimens with diffusion bonding planes parallel to the loading axis.

### 2.2 Test Procedure and Matrix.

A thermomechanical servo-hydraulic test system was used to perform the uniaxial tensile, fatigue, and creep-fatigue tests. An induction heating technique with closed-loop temperature feedback from three thermocouples was used to heat, monitor, and control the specimen temperature. The thermal ramp rate of 1°C/s was used to heat all specimens for tensile, fatigue, and creep-fatigue tests. A high-temperature extensometer was used to monitor and control strain. A set of displacement controlled tension tests at room temperature and elevated temperatures from $450∘C$ to $700∘C$ with $50∘C$ temperature increment and at $760∘C$ are conducted. All fatigue and creep-fatigue tests are strain-controlled with a strain rate of 0.1%/s. A set of fatigue tests with strain range of 0.6% at 550, 650, and $760∘C$ are conducted. The creep-fatigue and creep test matrix is shown in Table 2. In creep-fatigue tests, maximum peak strain is held constant for desired strain dwell periods per test matrix. All tests are conducted in an open-air environment.

Table 2

Test matrix for diffusion bonded Alloy 800H

Creep-fatigueCreep
Temp. ($∘C$)Dwell time (s)# of testsσ (MPa)# of tests
5506003170, 180, 2002, 1, 1
6001381
650300, 6003, 1100, 116, 1601, 1, 1
700681
760120, 6003, 140, 46, 62, 682, 2, 1, 4
Creep-fatigueCreep
Temp. ($∘C$)Dwell time (s)# of testsσ (MPa)# of tests
5506003170, 180, 2002, 1, 1
6001381
650300, 6003, 1100, 116, 1601, 1, 1
700681
760120, 6003, 140, 46, 62, 682, 2, 1, 4

## 3 Test Results From Diffusion Bonded 800H Specimens

### 3.1 Tension Test Results.

The tensile stress-strain responses at different temperatures from the diffusion bonded 800H (DB 800H) specimens are compared to a response of wrought base metal 800H (BM 800H) specimen in Fig. 2. The DB 800H specimens were tested at room temperature with axial load perpendicular to diffusion bonds showing a reduction in YS and UTS compared to the BM 800H specimens. At room temperatures, the % elongation of the DB specimens is comparable to the BM specimen. The reduction in YS and UTS of DB 800H specimens is expected due to the grain growth by the diffusion bonding processes. Surface texture is observed in the gage length of tested DB specimens, which is a common occurrence in metals with large grain size at high strains [23]. The DB 800H tension test is performed with axial load parallel to diffusion bonds, and the response (DB 800H P in Fig. 2(a)) shows comparable YS, UTS, and fracture mechanism, but higher % elongation compared to the results from tests with axial load perpendicular to the diffusion bonds.

Fig. 2
Fig. 2
Close modal

Stress–strain responses at $500∘C$, $600∘C$, and $700∘C$ are presented in Fig. 2(b), and those at $550∘C$, $650∘C$, and $760∘C$ are presented in Figs. 2(d)–(2f), respectively. In Fig. 2(b), gradual strength and % elongation reduction of DB 800H specimens with temperature are observed. Tension tests of DB 800H specimens at $450∘C$ demonstrated the reduced strength but increased % elongation of DB specimen when compared with BM specimen (Fig. 2(c)). Note in Fig. 2(c) that the DB specimen elongation at failure is more than 50%. The $450∘C$ tested specimens showed signs of surface texture and failed by necking with a cup and cone ductile fracture mode (not shown). The gradual strength and % elongation reduction of DB 800H specimens with temperature are also observed in Figs. 2(d)–(2f). However, note in Fig. 2(f), the performance of the DB 800H with axial load parallel to diffusion bonds (DB 800H P). The YS and UTS are comparable to the other DB 800H specimens, but the % elongation is larger than the BM 800H specimen. These results indicate that the diffusion bonding processes primarily cause the strength reduction of diffusion bonded specimens, but diffusion bond delamination governed failure mechanism (see below) is the primary reason for the reduction of % elongation of the diffusion specimens tested at $550∘C$ and above.

All three DB specimen tests at room temperature demonstrated ductile fracture mechanism, where specimen necks followed by the cup-cone fracture with shear lips. The fractured surfaces of room temperature tensile specimens are shown in Fig. 3(a). The necking and fracture were distributed over a wider region encompassing multiple diffusion bonded planes. At $500∘C$ and below, all DB 800H specimens failed with a ductile fracture mode similarly as shown in Fig. 3(a). However, a change in fracture mode is observed with an increase in temperature above $500∘C$. Tests at $500∘C$ and $550∘C$ demonstrated a small initial necking, but the final fracture occurred by bond delamination. Above $550∘C$, fracture occurs primarily by bond delamination with a relatively flat fracture surface as shown in Fig. 3(b), where drastic difference in failure modes of BM and DB specimens is observed. It is important to note in Fig. 2(f) the significant difference in % elongation of the DB 800H 1, 2, 3 specimens from the BM 800H and DB 800H P (axial load parallel to diffusion bonds) specimens. Note in Fig. 3(b) the occurrence of multiple bond delamination before final fracture. It is important to point out here that such a fracture mode cannot be categorized as brittle fracture because the elongation at rupture is 15%. As demonstrated later, this failure mode is related to the voids and microstructure of the diffusion bonds obtained by the current techniques.

Fig. 3
Fig. 3
Close modal

The Young’s modulus (E), yield stress (σy), ultimate tensile stress (σu), % elongation and local ductility determined from the tensile tests at room and elevated temperatures, and failure modes observed are listed in Table 3. The local ductility is defined as a ratio of the most deformed plate thickness in a ruptured specimen to the undeformed plate thickness. An optical microscope measured the distance between two adjacent bonding planes for local ductility calculation. Multiple bond delamination near the fractured surface, as demonstrated in Fig. 3(b), made it easier to locate the plate boundaries. The local ductility value indicates the accumulated plastic strain after fracture, and hence, the fracture mode observed for DB 800H is not a brittle mode. The local ductility values are calculated for specimens tested at 500–$760∘C$. Similarly, the tested specimens’ average specimen ductility or % elongation values are calculated for the specimen shank length. These % elongation values are shown in Table 3, where consistently higher local ductility values are observed at all temperatures at $500∘C$ and above. This observation indicates the localization of strains in the plates near the bond rupture region, which accelerates the flat bond rupture of diffusion bonded specimens once bond delamination is initiated.

Table 3

Properties of diffusion bonded Alloy 800H determined from the tensile tests

Temp. ($∘C$)MaterialE (GPa)σy (MPa)σu (MPa)% Elong.Local ductilityRupture surface
Tests perpendicular to diffusion bonds
20DB119718752851.2Ductile
DB219316050855.6Ductile
DB319517852751.5Ductile
DB419918553155.1Ductile
450DB116812345357.9Ductile
DB217912545155.1Ductile
500DB117312541137.561.4Mixed
550DB116811632635.859.8Mixed
DB215411633035.9
600DB115812131722.741.3Delamination
650DB115911129520.434.1Delamination
DB215911328920.6
700DB114811122713.529.3Delamination
760DB114412013816.333.3Delamination
DB213410313815.035.7
DB31349713815.730.1
DB41329513914.3
Tests parallel to diffusion bonds
20DB19818052555Ductile
760DB1619915840Ductile
Tests on wrought 800H
20BM20426158052.8Ductile
450BM17121754749.9Ductile
550BM15920850341.2Ductile
650BM15519437631.5Ductile
760BM13916620138.4Ductile
Temp. ($∘C$)MaterialE (GPa)σy (MPa)σu (MPa)% Elong.Local ductilityRupture surface
Tests perpendicular to diffusion bonds
20DB119718752851.2Ductile
DB219316050855.6Ductile
DB319517852751.5Ductile
DB419918553155.1Ductile
450DB116812345357.9Ductile
DB217912545155.1Ductile
500DB117312541137.561.4Mixed
550DB116811632635.859.8Mixed
DB215411633035.9
600DB115812131722.741.3Delamination
650DB115911129520.434.1Delamination
DB215911328920.6
700DB114811122713.529.3Delamination
760DB114412013816.333.3Delamination
DB213410313815.035.7
DB31349713815.730.1
DB41329513914.3
Tests parallel to diffusion bonds
20DB19818052555Ductile
760DB1619915840Ductile
Tests on wrought 800H
20BM20426158052.8Ductile
450BM17121754749.9Ductile
550BM15920850341.2Ductile
650BM15519437631.5Ductile
760BM13916620138.4Ductile

### 3.2 Creep Test Results.

A set of standard creep tests on DB 800H specimens under steady force at 550, 650, and $760∘C$ are conducted. The Larson–Miller parameter (LMP) using Eq. (1) is calculated to study the LMP and creep stress relationship for the diffusion-bonded Alloy 800H. In LMP equation (Eq. (1)), T is in the absolute temperature in Kelvin, tr is rupture life in hr, and “C” is a constant. Note that a wide spectrum of creep tests are needed to accurately characterize the” C” value, which is out of scope of this study. Hence, the “C” value of 15.21 determined by Swindeman [24] for BM 800H is used for DB 800H.
$LMP=T[log(tr)+C]$
(1)
The LMPs of DB 800H and BM 800H and their linear fits are compered in Fig. 4, where the creep rupture data of BM 800H are collected from Refs. [24,25]. Based on the DB 800H LMP linear fit in Fig. 4, few additional creep tests at 550, 600, and $700∘C$ are conducted to validate the life prediction LMP linear fit. All these additional tests are discontinued after the predicted rupture life based on the DB 800H LMP linear fit is exceeded. These tests are denoted by DB 800H (D) in Fig. 4.
Fig. 4
Fig. 4
Close modal

The creep rupture in all DB specimens is governed by bond delamination as shown in Fig. 5. The creep strain time histories of the tests performed at 550, 650, and $760∘C$ are shown in Fig. 6. At $550∘C$ and $650∘C$ responses, primary creep regimes are clearly observed, followed by secondary creep regimes. However, at $760∘C$, the primary creep regimes are not observed in Fig. 6. Typically, a tertiary creep regime of wrought alloys consists of void accumulation and growth over long periods, eventually resulting in a micro-crack initiation followed by necking and rupture. However, DB specimens showed negligible to no tertiary creep at all temperatures. Consequently, DB specimens failed prematurely due to bond delamination without necking, resulting in negligible or no tertiary creep regime. Also, note the initiation of multiple bond delamination before final fracture in Fig. 5(b). This observation eliminates the possibility of inadequate bonding at one particular bond plane. A microstructural evaluation is performed to investigate the diffusion bond creep rupture mechanisms through pre-test micro-tomography, scanning electron microscope/energy dispersive X-ray spectroscopy (SEM/EDS) secondary electron (SE), and backscatter secondary electron (BSE) images.

Fig. 5
Fig. 5
Close modal
Fig. 6
Fig. 6
Close modal

The variable pressure SEM is used to scan the rupture surfaces of creep specimens. The EDS images of the creep ruptured surface at $550∘C$ is shown in Fig. 7, where the SE images show the topographic features on the ruptured surface. Voids observed in the 100X magnification SE image match with the gray-colored regions of BSE image rich in elements with low atomic numbers. These EDS images show small pockets of Al and Nb distributed along the ruptured surface. The presence of Nb was not expected as ASTM specifications do not indicate any % Nb content. The Nb concentration could be the result of the alloy recycling as indicated by Mizia et al. [18]. The Nb leads to the formation of Ni3Nb, which is a γ” precipitate which could be detrimental to creep performance [26,27]. The Ti deposition predominantly in the form of γ” precipitates are observed in the voids seen in the SE image (Fig. 7). It can be argued that the void formation occurred during the creep deformations. Void formation usually results in an accelerated strain accumulation indicated by tertiary creep regime [28,29]. None of the DB 800H creep tests performed show a clear tertiary creep regime. Hence, it can be inferred that the Al- and Ti-rich sites prevent diffusion bonding uniformly, resulting in initial voids distributed along the bonding plane. The Ti precipitates could have been formed during the diffusion bonding process, as the bonding temperatures exceeded 1100°C [30]. EBSD scans by Mahajan et al. [21] showed the grain growth across the diffusion bonded interface. However, at some locations, new grain boundary was demonstrated to be formed along the bonded plane. If the plate surfaces are not cleaned before diffusion bonding, Al and Ti precipitate on plate surfaces may result in void formations along the bonded plane. This study makes a similar observation of precipitate accumulation and resulting void formation along diffusion bonded planes.

Fig. 7
Fig. 7
Close modal

Figure 8 shows the SE images of ruptured surfaces after creep tests at 550, 650, and $760∘C$. The 500X magnification SE image (Fig. 8(a)) at 550°C shows voids rich in precipitates. The fractured surface show mixture of relatively flat regions and signs of grain debonding. Similar observations are made from the 650 and 760°C test images in Figs. 8(b) and 8(c). It is hypothesized that the Ni interlayer used to facilitate the diffusion bonding process resulted in local diffusion, resulting in grain debonded rupture. None of the ruptured surfaces resemble a commonly observed void accumulation and growth-assisted creep failure mechanism. Hence, it can be concluded that voids at the bonded planes were created during the diffusion bonding process due to the presence of surface precipitates.

Fig. 8
Fig. 8
Close modal

A 10 mm long cylindrical region from the gauge length of a test specimen was scanned using an X-ray micro-tomography system. The X-ray scanning of the cylindrical specimen demonstrates the micro-void concentrations along bonding planes as shown in Fig. 9(a). The void size versus a corresponding number of voids is plotted in Fig. 9(c), where a high concentration of 25 to 30 μm sized voids are observed along the bond planes. Literature has shown that Al and Ti presence resists diffusion, resulting in voids along the bonding plane [9], as also observed in tested specimens in Fig. 8. In the creep regime, the precipitates induced voids and grain boundary formation along bond planes first initiate multiple bond delamination near specimen surface. Under tensile stress, the delamination gradually propagates to ultimately rupture the weakest bond of the DB 800H specimens.

Fig. 9
Fig. 9
Close modal

### 3.3 Fatigue Test Results.

One strain-controlled cyclic test at each of 550, 650, and 760°C and with a 0.6% strain range is first performed on wrought base metal (BM) 800H specimen to develop the baseline for performance evaluation of DB 800H test responses. Following, strain-controlled cyclic tests at each of these temperatures and the same strain amplitude are performed on DB 800H specimens. The stabilized stress-strain hysteresis loops from the tests are plotted in Fig. 10, where it is observed that the stabilized peak stress of the DB 800H specimen at each temperature is less than that of the BM 800H specimen. The strength reduction is consistent with the tension and creep test data. Stress amplitudes against the cycle number from the DB 800H tests are presented in Fig. 11. In this figure, it is observed that at 550 and 650°C, the material reaches the stabilized peak stress after few hundred cycles. On the other hand, at 760°C, the DB 800H cyclically hardens quickly over the first few cycles, followed by very slow cyclic hardening until reaching the peak stress just before failure of the specimen. The fatigue failure is defined as the peak force drop of 20%, which occurs little before the rupture of the specimens. Also, it is noted in Fig. 11 that fatigue lives from repeat experiments vary significantly. Also, the fatigue lives of DB 800H are smaller than the BM 800H (not shown). At each of 650 and 760°C tests, abrupt stress drop is observed, whereas at 550°C, a more gradual stress drop is observed. In all cases, a flat fracture surface along a bonding plane is the failure mode. In addition, all fatigue specimens showed delamination at multiple bonding planes before fracture, which is consistent with creep specimens.

Fig. 10
Fig. 10
Close modal
Fig. 11
Fig. 11
Close modal

The secondary electron (SE) and back scatter electron (BSE) images of DB 800H fatigue tested rupture surfaces are shown in Fig. 12. BM 800H fatigue fracture surfaces show striations, which is typical of fatigue fracture [31,32]. DB 800H specimens fracture surfaces are mostly flat with intermittent mini-striations. For example, the fatigue specimen at 650°C (Fig. 12(c)) shows small striations with small voids. As discussed earlier, these voids were precipitate rich, which may have been formed during the diffusion bonding process, accelerating the fatigue fracture process. The fractured surface at $760∘C$ also shows signs of grain debonding fracture.

Fig. 12
Fig. 12
Close modal

Similar to creep specimens, EDS maps of fatigue specimens indicate precipitation of Nb and Ti on the bonding plane. As stated earlier, these precipitates have likely to be formed during the diffusion bonding process. The EBSD scan in a previous work [21] indicated some grain growth across the bonding plane. However, at some locations, grain boundary formation at the bonding plane is observed. The micro-voids at these grain boundaries during the diffusion bonding promote the crack growth along the bonding plane. This is the reason that bond delamination fractures are observed in all fatigue specimens.

### 3.4 Creep-Fatigue Test Results.

Strain-controlled tests with tensile peak strain dwell are performed at 550, 650, and $760∘C$ to evaluate the creep-fatigue performance of DB 800H. Different dwell times at different temperatures are prescribed to ensure measurable stress relaxation during strain dwell. BM 800H specimen creep-fatigue tests at each temperature are conducted to develop the baseline performance for DB 800H fatigue-creep performance evaluation. The 100th cycle hysteresis loops and stress relaxation histories during strain dwell from the BM 800H and DB 800H creep-fatigue tests are compared in Fig. 13. In Figs. 13(a), 13(c), and 13(e), it is observed that the stress peaks of DB 800H are consistently lower than those of BM 800H, which is expected from the fatigue test hysteresis loop responses in Fig. 10. A dwell time of 10 mins is set for the $550∘C$ creep-fatigue test load history. At this temperature, creep rates are low, resulting in very small stress relaxation during strain dwell for both the DB 800H and BM 800H (Figs. 13((a) and 13(d)). A dwell time of 5 mins is set for the $650∘C$ creep-fatigue test. The 100th cycle hysteresis loops of BM 800H and DB 800H for this temperature are compared in Fig. 13(b). Although the stress relaxation in BM 800H and DB 800H specimens have different dwell stresses, similar stress relaxation rates are observed in Fig. 13(e). A dwell time of 2 mins is set for the $760∘C$ creep-fatigue test. At this temperature, the alloy creep rates are higher, resulting in higher stress relaxation as shown in Fig. 13(f), where it is observed that the stress relaxation rates of the BM 800H and DB 800H are comparable. The stress peaks in the creep-fatigue tests of DB 800H specimens at three temperatures are plotted against the number of cycles in Fig. 14, where it is observed that the stress peaks of DB 800H do not stabilize for any of the temperatures under creep-fatigue loading. This phenomenon is unlike the fatigue loading responses at 550 and $650∘C$ when the stress peaks of DB 800H stabilized within few hundred cycles (Fig. 11). The failure mode of all creep-fatigue tests is bond delamination with sudden stress drop over a few cycles.

Fig. 13
Fig. 13
Close modal
Fig. 14
Fig. 14
Close modal

The secondary electron (SE) images of creep-fatigue test rupture surfaces are shown in Fig. 15. In this figure, the SE image at 550°C show traces of slip planes indicating localized ductile failure. Mini-striations are observed intermittently at 550°C and $650∘C$ fractured surfaces. These striations are localized far apart from each other, unlike those from fatigue tests on BM 800H, indicating accelerated crack propagation during strain dwell [33]. The creep-fatigue rupture occurs with a sudden stress drop, unlike the traditional void-assisted gradual failure. The fractured surface showed several voids rich with Ti and Nb as indicated by the EDS map of a fracture surface tested at $650∘C$. The Ti precipitates located in the voids prevent diffusion bonding processes, resulting in micro-voids along bonding planes and the quick rupture. Similar to the creep specimen, the creep-fatigue specimen shows Al and Nb’s local pockets along the fracture surface.

Fig. 15
Fig. 15
Close modal

To investigate the influence of dwell time on creep-fatigue responses, two more tests are performed at 650 and $760∘C$ with a 10-min peak strain dwell period, which is same as the $550∘C$ creep-fatigue test. Recorded data from these three tests are compared in Fig. 16. At 650°C, the specimen experiences higher stress relaxation during the strain dwell than at 550°C as shown in Fig. 16(b). Stress relaxation during the load cycle is also different at different temperatures and influences the hysteresis loop shape. At 760°C, the hysteresis loop shape is much different from those observed at 550 and 650°C because of the higher stress relaxation as shown in Fig. 16(b). Stress relaxation during strain dwell increases dislocation density, which influences the initial stiffness as observed in Fig. 16(a). The stress peaks of these three tests are compared in Fig. 16(c), where significant cyclic hardening with comparable hardening rates are observed at 550 and $650∘C$, but lower cyclic hardening rates at $760∘C$. For 550 and $650∘C$, the specimen failed before reaching the cyclic stabilization, whereas the $760∘C$ test reaches the stable stress state after 30 cycles.

Fig. 16
Fig. 16
Close modal

Two sites are selected on the rupture surface creep-fatigue specimen tested at $650∘C$ for detailed element distribution analysis through 7000X magnification images as shown in Fig. 17. In this figure, Site 1 shows a void with precipitates, and Site 2 shows a metal matrix. The energy dispersive X-ray map of these two sites are presented in Fig. 17. Site 1 has a higher concentration of Ti, C, and Nb compared to Site 2. This Nb and Ti reach site indicates the presence of precipitates yield void formation. This precipitate distribution type is observed in all specimens. As most of these precipitates usually form after a long duration of intermediate temperature exposure, these are unlikely to be formed during creep-fatigue loading. As discussed earlier, these precipitates might have formed during the diffusion bonding process and induce micro-void along bonding planes and initiated the bond delamination rupture observed in all specimens.

Fig. 17
Fig. 17
Close modal

## 4 Discussion

This study explores the mechanical and microstructural performances of diffusion bonded Alloy 800H (DB 800H) under tensile, creep, fatigue, and creep-fatigue loads at elevated temperatures. All tension test responses demonstrated a reduction in yield and ultimate tensile strengths of DB 800H compared to BM 800H. Up to $450∘C$, all tension tests with axial loading perpendicular to diffusion bonds resulted in ductile fracture. From $500−550∘C$, a mixed-mode fracture with the combination of ductile and bond delamination failure modes is observed. All tensile specimens above $550∘C$ showed only the bond delamination rupture mechanism. However, the plastic strain accumulation in diffusion bonded plates after fracture of specimens indicates local ductility which suggests that the fracture mode of DB 800H specimens was not brittle. The tension tests with axial load parallel to diffusion bond planes shown a ductile fracture at both the room temperature and $760∘C$. The %elongations of DB 800H specimens tested with axial load perpendicular to bond planes are always lower compared to those of BM 800H for temperatures above $550∘C$. On the other hand, the %elongations of the DB specimens tested with axial load parallel to bond planes are a little higher than the BM 800H specimens. These results indicate that the strengths of DB 800H specimens are reduced compared to the BM 800H primarily because of the diffusion bonding processes, whereas the %elongations of DB 800H specimens at elevated temperatures are reduced because of the bond delamination governs the failures.

Creep tests with the target rupture time ranging from 1000 to 3000 hours are conducted for temperatures 550 to $760∘C$. DB 800H minimum creep rates are comparable to BM 800H; however, DB 800H specimens show very little to no tertiary creep and always fail by bond delamination rupture. The X-ray tomography scans indicate a large number of micro-void concentrations along bond planes. The highest micro-void concentration was for the void sizes 25–30 μm. The LMP demonstrate significant creep life reduction of DB 800H compared to BM 800H. The SEM images show precipitate formation at grain boundaries along bonding planes.

The fatigue tests on DB 800H at elevated temperatures also show life reduction compared to the BM 800H tests. Small striations are observed on the rupture surface, but most of the surface show bond delamination type fracture assisted with micro-voids formed during the diffusion bonding process. The strain dwell tests with different dwell periods are conducted to study DB 800H creep-fatigue performance at elevated temperatures. The dwell period directly influenced the reduction of the creep-fatigue life. Again, small striation separation is observed, and the fracture surface consistently shows bond delamination features. The SEM fractography of specimen tested at 550°C showed slipping traces indicating localized ductile deformation. However, at the macro-scale, the fracture is bond delamination.

It is noted here that the tests conducted in this study explore the performance of DB 800H at a basic level. Therefore, these tests are limited and act as guidance for future test program development. The test data acquired for DB 800H will be used for developing allowable stresses and isochronous curves for the ASME Section III, Division 5 design codes.

## Acknowledgment

This research is being performed using funding received from the DOE Office of Nuclear Energy’s Nuclear Energy University Program (Award No. DE-NE0008576), and Integrated Research Program (Award No. DE-NE0008714). Authors would like to acknowledge Mr. Aaron Wildberger from Vacuum Process Engineering for his help with the diffusion bonded Alloy 800H block. The microstructural examinations were performed at the Analytical Instrumentation Facility (AIF) at North Carolina State University, which is supported by the State of North Carolina and the National Science Foundation (Award No. ECCS-2025064).

## Conflict of Interest

There are no conflicts of interest.

## Data Availability Statement

The authors attest that all data for this study are included in the paper. Data provided by a third party listed in Acknowledgment.

## References

1.
US DOE,
2002
, US DOE Nuclear Energy Research Advisory Committee and Gen IV International Forum, A Technology Roadmap for Generation IV Nuclear Energy Systems.
2.
Collins
,
J.
,
2009
,
Next Generation Nuclear Plant Project Technology Development Roadmaps: The Technical Path Forward
.
Technical Report, Idaho National Laboratory (INL)
.
3.
Sabharwall
,
P.
,
Kim
,
E. S.
,
Siahpush
,
A.
, and
Patterson
,
M.
,
2012
, “
Preliminary Design for Conventional and Compact Secondary Heat Exchanger in A Molten Salt Reactor
,”
ASME 2012 Heat Transfer Summer Conference Collocated With the ASME 2012 Fluids Engineering Division Summer Meeting and the ASME 2012 10th International Conference on Nanochannels, Microchannels, and Minichannels
,
Rio Grande, Puerto Rico
.
4.
Lillo
,
T.
,
Williamson
,
R.
,
Reed
,
T.
,
Davis
,
C.
, and
Ginosar
,
D.
,
2005
,
Engineering Analysis of Intermediate Loop and Process Heat Exchanger Requirements to Include Configuration Analysis and Materials Needs
,
Technical Report, Idaho National Laboratory (INL)
.
5.
Ravindran
,
P.
,
Sabharwall
,
P.
, and
Anderson
,
N. A.
,
2010
, “
Modeling a Printed Circuit Heat Exchanger with relap5-3d For The Next Generation Nuclear Plant
.”
Office of Scientific and Technical Information Technical Reports
.
6.
Oh
,
C. H.
,
Kim
,
E. S.
, and
Patterson
,
M.
,
2010
, “
Design Option of Heat Exchanger for the Next Generation Nuclear Plant
,”
ASME J. Eng. Gas Turbines Power
,
123
(
3
), p.
032903
.
7.
Eesa
,
M.
, and
Marriott
,
G.
,
2018
, Photochemical Etching of Heat Exchanger Plates.
8.
Li
,
X.
,
Smith
,
T.
,
Kininmont
,
D.
, and
Dewson
,
S. J.
,
2009
, “
Materials for Nuclear Diffusion-Bonded Compact Heat Exchangers
,”
Proceedings of ICAPP’09
,
Tokyo, Japan
, p.
9
.
9.
Mylavarapu
,
S. K.
,
Sun
,
X.
,
Christensen
,
R. N.
,
Unocic
,
R. R.
,
Glosup
,
R. E.
, and
Patterson
,
M. W.
,
2012
, “
Fabrication and Design Aspects of High-Temperature Compact Diffusion Bonded Heat Exchangers
,”
Nucl. Eng. Des.
,
249
, pp.
49
56
.
10.
Sabharwall
,
P.
,
Clark
,
D. E.
,
Mizia
,
R. E.
,
Glazoff
,
M. V.
, and
McKellar
,
M. G.
,
2013
, “
Diffusion-Welded Microchannel Heat Exchanger for Industrial Processes
,”
ASME J. Therm. Sci. Eng. Appl.
,
5
(
1
), p.
011009
.
11.
Jahn
,
S.
,
Sändig
,
S.
,
Dahms
,
S.
, and
Gemse
,
F.
,
2014
, “
Diffusion Bonding Systems
,”
Materialwissenschaft und Werkstofftechnik
,
45
(
9
), pp.
807
814
.
12.
Mizia
,
R.
,
2010
, “
Scoping Investigation of Diffusion Bonding for NGNP Process Application Heat Exchangers
.” PLN-3565, July, 9.
13.
Li
,
X.
,
Kininmont
,
D.
,
Le Pierres
,
R.
, and
Dewson
,
S. J.
,
2008
, “
Alloy 617 for the High Temperature Diffusion-Bonded Compact Heat Exchangers
,”
Proceedings of ICAPP
,
Anaheim, CA
, Vol. 8, pp.
282
288
.
14.
Sah
,
I.
,
Kim
,
D.
,
Lee
,
H. J.
, and
Jang
,
C.
,
2013
, “
The Recovery of Tensile Ductility in Diffusion-Bonded Ni-Base Alloys by Post-Bond Heat Treatments
,”
Mater. Des.
,
47
, pp.
581
589
.
15.
Sah
,
I.
,
Hwang
,
J.-B.
,
Kim
,
W.-G.
,
Kim
,
E.-S.
, and
Kim
,
M.-H.
,
2020
, “
High-Temperature Mechanical Behaviors of Diffusion-Welded Alloy 617
,”
Nucl. Eng. Des.
,
364
, p.
110617
.
16.
Rozman
,
K.
,
Carl
,
M.
,
Kapoor
,
M.
,
Doğan
,
Ö.
, and
Hawk
,
J.
,
2019
, “
Creep Performance of Transient Liquid Phase Bonded Haynes 230 Alloy
,”
Mater. Sci. Eng. A.
,
768
, p.
138477
.
17.
Kim
,
S. H.
,
Kim
,
C.
, and
Jang
,
C.
,
2018
,
“Diffusion Bonding of a Cold-Worked Ni-Base Superalloy,“ Pressure Technology, vol. 40764, p. V001T02A004. American Society of Mechanical Engineers
.
18.
Mizia
,
R. E.
,
Clark
,
D. E.
,
Glazoff
,
M. V.
,
Lister
,
T. E.
, and
Trowbridge
,
T. L.
,
2013
, “
Optimizing the Diffusion Welding Process for Alloy 800H: Thermodynamic, Diffusion Modeling, and Experimental Work
,”
Metall. Mater. Trans. A.
,
44
(
1
), pp.
154
161
.
19.
Keating
,
R.
,
Nestell
,
J.
,
McKillop
,
S.
,
Allen
,
T.
, and
Anderson
,
M.
,
2019
, “
Potential ASME Code Case for Construction of Compact Heat Exchangers in High Temperature Reactors
,”
The Proceedings of the International Conference on Nuclear Engineering (ICONE) 2019.27
,
Tsukuba, Japan
.
20.
Mahajan
,
H. P.
,
Devi
,
U.
, and
Hassan
,
T.
,
2019
, “
Material Property Evaluation and Response Simulation of the Diffusion Bonded Materials of Compact Heat Exchanger
,”
Proceedings of the International Conference on Nuclear Engineering (ICONE) 2019.27
,
Tsukuba, Japan
.
21.
Mahajan
,
H. P.
,
Elbakhshwan
,
M.
,
Beihoff
,
B. C.
, and
Hassan
,
T.
,
2020
, “
Mechanical and Microstructural Characterization of Diffusion Bonded 800H
,”
Pressure Vessels and Piping Conference
,
Virtual, Online conference
.
22.
2017 ASME Boiler & Pressure Vessel Code, Section VIII Division 1
”.
American Society of Mechanical Engineers
.
23.
Frommert
,
M.
, and
Gottstein
,
G.
,
2009
, “
Mechanical Behavior and Microstructure Evolution During Steady-State Dynamic Recrystallization in the Austenitic Steel 800H
,”
Mater. Sci. Eng. A.
,
506
(
1–2
), pp.
101
110
.
24.
Swindeman
,
R. W.
,
Zamrik
,
S. Y.
, and
Maziasz
,
P. J.
,
2007
, “
Effects of Long-Term Service on the Microstructure and Tensile Properties of Alloy 800H
,”
ASME 2007 Pressure Vessels and Piping Conference
,
San Antonio, TX
, pp.
483
492
.
25.
Booker
,
M.
,
1978
,
Analytical Representation of the Creep and Creep-Rupture Behavior of Alloy 800H
.
Technical Report
,
Oak Ridge National Laboratory
,
TN
.
26.
Lai
,
G.
, and
Kimball
,
O.
,
1978
,
Aging Behavior of Alloy 800H and Associated Mechanical Property Changes.[538/sup 0/to 816/sup 0/c]
.
Technical Report
,
General Atomic Co
.,
San Diego, CA
.
27.
Donachie
,
M. J.
, and
Donachie
,
S. J.
,
2002
,
Superalloys: A Technical Guide
,
ASM International, Materials Park, OH
.
28.
Chow
,
J.
,
Soo
,
P.
, and
Epel
,
L.
,
1978
,
Creep and Fatigue Properties of Incoloy 800H in a High-Temperature Gas-Cooled Reactor (HTGR) Helium Environment, Brookhaven National Lab., Upton, NY
.
29.
Tavassoli
,
A.
, and
Colombe
,
G.
,
1978
, “
Mechanical and Microstructural Properties of Alloy 800
,”
Metallurgical Trans. A
,
9
(
9
), p.
1203
.
30.
Jones
,
W.
, and
Allen
,
R.
,
1982
, “
Mechanical Behavior of Alloy 800 At 838 K
,”
Metallurgical Trans. A
,
13
(
4
), pp.
637
648
.
31.
Villagrana
,
R.
,
Kaae
,
J.
, and
Ellis
,
J.
,
1981
, “
The Effect of Aging and Cold Working on the High-Temperature Low-Cycle Fatigue Behavior of Alloy 800H: Part II: Continuous Cyclic Loading
,”
Metallurgical Trans. A
,
12
(
11
), pp.
1849
1857
.
32.
Soo
,
P.
, and
Sabatini
,
R. L.
,
1984
, “
High-Cycle Fatigue Behavior of Incoloy Alloy 800H in a Simulated Htgr Helium Environment Containing High Moisture Levels
,”
Nucl. Technol.
,
66
(
2
), pp.
324
346
.
33.
Hour
,
K.
, and
Stubbins
,
J.
,
1989
, “
The Effects of Hold Time and Frequency on Crack Growth in Alloy 800H At 650°C
,”
Metallurgical Trans. A
,
20
(
9
), p.
1727
.