Phenomena of hydrogen trapping and embrittlement in hydrogen changed iron are investigated throughout the theoretical models derived in this work and experiments like themal analysis technique and tensile test.
Theory of hydrogen retrapping which can predict and describe the behavior of supersaturated hydrogen in interstitial site of normal lattice is suggested, considering the variation of equilibrium state between the hydrogen trapped in trapping site and one in normal lattice site with the change of temperature, occurred during rapid cooling after hydrogen charging at high temperature.
Coexisting the grainboundary and microvoid in pure iron, the hydrogen trapped in grainboundary is detrapped throughout the accomplishment of dynamic local equilibrium with one remained in normal lattice site (reversible trap), but that in microvoid is released independently with lattice hydrogen (irreversible trap). Heating the hydrogen charged iron with 3K/min, peak of lattice hydrogen forming the dynamic local equilibrium with grainboundary and peak by microvoid appear at 385K, and their trap activation energies are 16.30KJ/mol, and 40.31KJ/mol, respectively.
In case of pure iron including the grainboundary and microvoid, lattice hydrogen content forming the dynamic local equilibrium with one trapped in grainboundary and hydrogen amount trapped in microvoid increase with hydrogen charging temperature, simultaneously. Amount of hydrogen dissolved into interstitial site of normal lattice forming the dynamic local equilibrium increases as the amount of microvoid decreases, and total amount of hydrogen existing in specimen is nearly constant. In 60% cold drawn Fe, that is, iron containing a large amount of dislocation, only one thermal analysis peak induced by dislocation is observed. These phenomena are in accord with results predicted by the theory of hydrogen retrapping.
Equation of apparent hydrogen diffusivity in polycrystalline iron, which is decided from the combination of apparent hydrogen diffusivity at 293K measured by evolution method with the activation energy of the hydrogen peak of normal lattice forming the dynamic local equilibrium with one in grainboundary, is as follow.
$D_A=3.919×10^{-4}$ exp (-16.30 (KJ/mol)/RT) (㎠/sec)
Hydrogen solubility equation in pure iron measured by thermal analysis technique is as follow.
$C_L$ = 25.88 exp (-25.84 (KJ/mol)/RT) (wt.ppm)
Holding the heat treated iron single crystal at 293K in vacuum atmosphere, which was charged with hydrogen at high temperature, the amount of hydrogen trapped in reversible trap decreases rapidly but that in irreversible trap, slowly. This is also consistent with the prediction by the theory of hydrogen retrapping. And in case which reversible trap is only existed, the apparent hydrogen diffusivity at 293K is $1×10^{-6}$㎠/sec. and that irreversible trap is only existed, $4×10^{-8}$㎠/sec. and for the existence of two types of traps, $2.5×10^{-7}$㎠/sec.
When the sintered iron with $Al_2O_3$ power is thermally analyzed with a uniform heating rate, 3K/min, evolution rate peak of hydrogen released from the interface of $Al_2O_3$ and closed micropore formed around a interface of $Al_2O_3$ appear at 853K and 653-708K, respectively. The trap binding energy and trap density that are determined by using the theories of hydrogen trapping and retrapping and the results of thermal analysis are 71.4KJ/mal, and $1×10^{19}$ site/gFe, respectively. Energy level of hydrogen around the interface of $Al_2O_3$ is suggested by using the trap binding energy and calculated trap activation energy, 78.96KJ/mol. Not-appearing the hydrogen peak for reversible trap in sintered iron with $Al_2O_3$ powder is agreed with the prediction in theory of hydrogen retrapping.
Rate equation for hydrogen evolution which can be applied to thermal analysis technique is derived by using the non-steady state evolution equation of hydrogen. The results computed from this equation are consistent with the experimental ones.
Evolution rate peak of hydrogen in lattice consistuting the dynamic local equilibrium with that in grainboundary moves toward lower temperature as the thickness of specimen becomes thinner. This is interpreted clearly throughout the computed results of rate equation of hydrogen evolution.
The types and amounts of lattice defects, acting as trapping sites of hydrogen in pure iron, generated during cathodic charging in the absence of applied stress are determined by thermal analysis technique, and the variation of apparent hydrogen diffusivity with the amount of lattice defect generated during cathodic charging is examined, using the mathematical model derived from the existing trap theory. From the result of thermal analysis, it is observed that internal microcracks or microvoids are generated more predominantly than dislocation and are the major trapping sites of hydrogen in pure iron, charging hydrogen cathodically. Internal microcracks and microvoids begin to be formed below the applied current density, 1 mA/㎠, during cathodic charging, and the blisters observed on the surface of iron specimen are found to be interconnected to the surface of specimen through microcracks remained in the vicinity of blisters. Apparent diffusivity of hydrogen at 458K decreases linearly with the reciprocal value of square root of applied current density during the cathodic charging. The amount of internal microcrack or microvoid induced by hydrogen during cathodic charging increases linearly with the square root of applied current density. This implies that the generation of internal microcrack or microvoid is directly related to the amount of hydrogen dissolved in interstitial site of normal lattice, considering the linear increase of lattice hydrogen solubility with the square root of applied current density.
A theoretical model by which a critical current density for the generation of defects during cathodic hydrogen chargingbe determined is presented by using the existing trap theory. From the curve fitting of the model and experimental results, it is found that critical current density for defect generation is in the range of 0.5-1.0 mA/㎠ in the cathodic charging of hydrogen into pure iron in 10% sulfuric acid solution adding a recombination poison of 10mg $As_2O_3/1$.
The effect of hydrogen induced damages produced during the cathodic charging of hydrogen into pure iron to fracture strain and UTS (ultimate tensile strength) is investigated, removing the hydrogen dissolved in specimen throughout the heat treatment in vacuum. Fracture strain and UTS of pure iron decrease with increment of input current density. Lattice defect affecting dominently to the fracture strain and UTS is the surface blister, following is the internal microcrack or microvoid.
3-6
Chapter 1의 수소의 retrapping 이론을 통해 예측했던 thermal analysis의 이론적 결과는 여러 종류의 격자결함을 포함하고 있는 순철시편의 thermal analysis 실험결과와 일치하였다.
첫째, 정상격자내의 수소는 grainboundary 에 trap된 수소와 dynamic local equilibrium을 이루면서 확산하였고,
둘째, microvoid내의 수소는 정상격자에 있는 수소와 무관하게 확산하며, (irreversible trap) microvoid의 양이 감소할수록 grainboundary와 dynamic local equilibrium을 이루는 수소양은 증가하였으며, 시편내의 전체 수소양은 일정하였다. 이것은 식(44-b)가 성립한다는 것을 암시한다.
세째, 수소주입온도가 증가할수록 각 trap에 trap되는 양은 증가하나, grainboundary와 같은 reversible trap만이 주로 존재하는 경우, 대부분의 수소는 reversible trap과 dynamic local equilibrium을 이룬다. (Fig.19, Fig.26).
네째, 어떤 한 종류의 trap density가 매우 큰 경우, reversible trap에 의한 수소 peak은 나타나지 않았다 (Fig.22).
Thermal analysis 방법으로 수소용해도 및 수소확산계수를 측정하였으며, 이때의 확산계수 및 chapter 2의 thermal analysis 의 rate equation을 이용하여 계산된 thermal analysis의 peak온도는 실험치와 거의 일치하였다.
4-4
대부분의 실험결과가 chapter 3에서의 것과 일치하였으며, retrapping 이론과 일치하였다. 두 종류의 trapping site가 존재할 때, 각 trapping site가 수소확산에 미치는 영향을 살펴보았다. reversible trapping site 만이 존재하는 경우, 293K에서의 겉보기 수소확산계수는 $1×10^{-6}$ ㎠/sec 이었고, microvoid peak의 경우 $4×10^{-6}$ ㎠/sec 이었다. 그리고 두 종류의 trapping site가 공존하는 경우 는 $2.5×10^{-7}$ ㎠/sec 이었다.
5-4
$Al_2O_3$가 첨가된 소결된 철 시편의 경우, $Al_2O_3$ 계면 및 그 주위에 형성된 closed micro-pore가 수소의 주된 trapping site인 것으로 생각된다.
Trapping 및 retrapping 이론과 thermal analysis 결과로 부터 결정된 $Al_2O_3$ 계면에서의 수소 trap binding energy는 71.4KJ/mol이었다. Trap activation energy, 78.96KJ/mol 을 이용하여 $Al_2O_3$ 계면에서의 수소가 갖는 energy level을 제시하였다.
Retrapping 이론과 trap binding energy를 이용하여 $Al_2O_3$가 첨가된 소결된철 시편에서 reversible peak이 나타나지 않는 현상을 설명할 수 있었다.
6-4
Thermal analysis technique을 통하여 cathodic charging시 순철 내에 생기는 격자결함의 종류 및 양을 결정하였고, 이 결과로 부터 유도된 수학적 model을 이용하여, 이들 격자결함이 수소확산계수에 미치는 영향을 살펴보았다. 그 결과 internal microcrack 혹은 microvoid가 dislocation보다 더 생성되었으며, 또한 수소의 주된 trapping site로 작용하였다. Internal microcracks과 microvoid는 cathodic charging시 1mA/㎠ 근처에서 생성되기 시작했으며, 시편 표면에 생긴 blister는 blister주위의 microcrack들을 통해 시편의 표면과 연결되어 있는 것으로 관찰되었다. 그리고 458K에서의 수소의 겉보기 확산계수는 cathodic charging시 전류밀도의 역수의 제곱근이 작아짐에 따라 감소하였다. Internal microcrack 혹은 microvoid의 양은 전류밀도의 제곱근에 따라 직선적으로 증가하였으며, 이것은 격자수소용해도와 전류밀도의 제곱근과의 직선적인 관계를 고려할 때, internal microcrack 혹은 microvoid의 생성은 정상격자내에 용해된 수소양과 직접적인 연관관계가 있을 것으로 믿어진다. 그리고 firstT.A.시 수소주입시간 및 주입전류밀도가 증가할수록 낮은 온도쪽으로 microvoid에 의한 peak이 이동한 것은 microvoid내의 수소 압력이 증가하였기 때문으로 믿어진다.
7-4
기존 trap 이론을 이용하여, cathodic charging 동안 시편내에 주입된 수소에의해 격자결함이 생성하기 시작하는 임계전류밀도를 결정할 수 있는 일반적인 이론 model을 유도하였다. Thermal analysis 방법으로 얻어진 실험결과와 model의 curve fitting으로부터, 10mg $As_2O_3/ℓ$의 recombination poison이 첨가된 10% 황산 용액에서 순철내에 수소를 cathodic charging할 때, 격자결함이 생기기 시작하는 임계전류밀도는 약 0.5~1.0mA/㎠이었다.
8-3
순철내에 수소를 전해주입한 후 진공 열처리하여 시편내의 수소를 제거하여, 전해주입시 시편내에 생성된 격자결함만이 UTS와 fracture strain에 미치는 영향을 알아본 결과, 주입전류밀도가 증가할수록 UTS와 fracture strain을 감소하였다.
Fracture strain과 UTS에 가장 큰 영향을 미치는 격자결함은 surface blister이었고, 다음이 internal microcrack 혹은 microvoid 이었다.