**3. The prediction of creep fracture life and the derivation of the converted stress to estimate the mechanical performance of creep strength of a double edge notched weld joint specimen for P91 steel based on the** *QL***\* concept**

#### **3.1 Material and specimen**

The material used for this study is P91 steel and matching weld metal of US-9Nb, which is a similar material as P91 steel. The chemical composition and mechanical properties are shown in **Tables 1** and **2**.


#### **Table 1.**

*Chemical compositions of P91 steel (wt%).*


#### **Table 2.**

*Mechanical properties of P91 steel.*

**Figure 7.** *Sampling site of specimen of base metal and weld joint. ① Base metal ② Weld joint.*

The specimen used is a double-edge notched specimen (DEN) as shown in **Figure 2**. Sampling sites of specimen of base metal and weld joint are shown in **Figure 7**.

#### **3.2 Experimental method**

The machine system was designed and developed to enable automatic real-time observational experiments using a CCD microscope [15]. Now, CCD microscope was replaced to digital microscope manufactured by KEYENCE corporation. Using this apparatus, in situ observation of creep damage progression was conducted and the images of the damage region were quantified using a PC. The tests were conducted under high temperature vacuum conditions of 10<sup>4</sup> Pa. The creep damage region around the notch tip was found to be a dark region, composed of voids and micro-cracks originating along grain boundaries, as shown in previous results for SUS304 stainless steel [10, 15]. The dark region is defined as a creep damage region owing to the following reasons.

The specimens were heated using infrared rays (IR) under vacuum conditions, as shown in **Figure 8**. Creep damage is caused by micro-cracking along a grain boundary, which is composed of voids at the grain boundary [10, 15] that are considered to be caused by vacancy diffusion [16, 17]. In the damage region, a specimen surface becomes irregular due to micro-cracking along a grain boundary. In this region, diffused reflection of light by the lamp of IR was caused and it shows as the dark region.

*The Quantitative Estimation of Mechanical Performance on the Creep Strength… DOI: http://dx.doi.org/10.5772/intechopen.106419*

**Figure 8.** *Schematic illustration of in-situ observational creep-fatigue testing machine [15].*

### **3.3 Experimental conditions and results**

Previously, experimental results were published in Japanese [18], however, more detailed analyses are needed by more accurate analyses. In this section, updated results are written.

#### *3.3.1 Similarity law of creep deformation*

Experimental conditions, their results of steady state RNOD rate and creep fracture life are shown in **Table 3**. These results show that the fracture life of weld joint takes 3.5 5% of that for the base metal.


*"Predicted" means that creep test was interrupted at the accelerated creep region and its final creep fracture was predicted from this point.*

*"\*" means that it almost covers total fracture life, however it is interrupt test.*

#### **Table 3.**

*Experimental conditions, results of steady state RNOD rate and creep fracture life of DEN specimen (Base metal and weld joint for P91 steel).*

#### **Figure 9.**

*Non-dimensional time sequential behavior of RNOD of the base metal and weld joint for P91 steel. t=tf : non-dimensional creep fracture life of each specimen.*

Non-dimensional time sequential characteristics of the RNOD curve (creep deformation) controlled by each fracture life are shown in **Figure 9**. For the base metal, the similarity law of RNOD curve caused, which is independent of applied stress. For the weld joint, the similarity law also caused for the case of *t=tf* <0*:*5.
