When a client located a crack on a critical piece of manufacturing equipment midway through a production run, they implemented an immediate repair and shut the machine down. Before production could resume, the client needed to know if the equipment would run in a safe mode for the required amount of time to complete their current order. We set about breaking the single task down into a series of discrete stages. Each stage would feed into the next and would build an understanding of the underlying problem and therefore define the potential solution. Ultimately our analysis showed that the part needed to be replaced, but we managed to provide a temporary secondary solution that would allow them to fulfil their order to their customer.
When a client located a crack on a critical piece of manufacturing equipment midway through a production run, they implemented an immediate repair in an attempt to keep the equipment going so they could complete the current order. The immediate repair consisted of drilling holes and fixing two stiffening plates at either end of the crack, which on face value makes sense, but the validity of the repair had not been underpinned with any supporting calculations or detailed analysis per se. Before production could resume, the client needed to know if the equipment would run in a safe mode for the required amount of time to complete the current order. Safety was of paramount importance and time was clearly of the essence. This is where we came in.
We set about breaking the single task down into a series of discrete stages. Each stage would feed into the next and each stage would build an understanding of the underlying problem and therefore a definition of the potential solution.
Four stages were defined, namely;
- evaluate the stress field in the component before and after the crack developed,
- evaluate the effectiveness of the stiffening plates,
- estimate the likelihood of further crack propagation and
- make recommendations on the machine operating window to extend the remaining life.
We would look at three models:
- the intact component with no crack present,
- the cracked component with no stiffening plates added
- the cracked component with the 2 stiffening plates added.
We would then do four assessments:
- an ASME VIII Div. 2 linear elastic strength assessment for the intact component,
- an S-N fatigue analysis on the intact component,
- a strength and Linear Elastic Fracture Mechanics (LEFM) assessment on the cracked component without any stiffening plates added and
- a strength and Linear Elastic Fracture Mechanics (LEFM) assessment on the cracked component with the 2 stiffening plates added.
The linear elastic stress assessment on the intact component confirmed the membrane and membrane + bending stress Utilisation Ratios (UR) for all key Stress Classification Lines (SCL) were below 1 implying that the intact component was acceptable with respect to the global criterion assessment for plastic collapse. The fatigue analysis on the intact component determined;
- the alternating stresses in the assembly components when subjected to the applied loads,
- identified the critical “hot-spot” locations and
- estimated the intact component fatigue life using the S-N fatigue design methodology and the appropriate fatigue design curves.
The strength assessment of the cracked component with no stiffening plates added determined the developing stress field around the drill holes at each end of the crack. For the LEFM assessment, we applied a shallow semi-elliptical crack of 1mm length at the surface of each drill hole (i.a.w. the BS 7910 guidelines) at the location of the peak stress and assessed the stress intensity factor (SIF) at the defect as the component was subjected to the operational loads. The developing SIF was also compared to the material fracture toughness in order to verify the component against the risk of brittle fracture. Finally, we derived the crack propagation rate and the estimated the remaining fatigue life of the component using the appropriate crack propagation curve; this involved a number of models where the physical geometry of the crack was accurately modelled so that representative plots of ‘SIF versus crack length’ and ‘SIF versus crack depth’ could be generated. We then repeated the above strength and LEFM assessment on the model of the cracked component with the strengthening plates fitted to determine the effect of the strengthening plates on the crack propagation rate and the remaining fatigue life.
The analysis showed that one of the plates was very effective in arresting the crack propagation rate at one end of the crack, but the second plate located at the other end of the crack had little effect on the crack propagation rate. We concluded that if the current operating conditions of the equipment remained unchanged, the cracked component with the stiffening plates added would eventually fail in a brittle and potentially unsafe manner. The primary recommendation was to cease production and wait until the cracked component had been replaced. However, a secondary recommendation, which was subject to further analyses and a comprehensive non-destructive testing regime, was also made to operate the equipment at a reduced duty cycle; the cracked component still needed to be replaced, but the reduced duty cycle would enable the existing production order to be fulfilled, albeit over a longer time frame.
- Provided technical assurance on the current status of the compromised equipment.
- Maintained short term production capability to complete an existing order by defining a reduced duty cycle supported by inspection intervals of sufficient (or ‘prescribed’) frequency to ensure safe on-going, but short term, operation.
- Underpinned the technical need to make a financial investment in replacement parts in the short term.