Engineers use a wide range of tools and techniques to ensure that the designs they create are safe. However, accidents sometimes happen and when they do, companies need to know if a product failed because the design was inadequate or if some other cause, such as user error, was to blame. Whether a manufacturer incurs the cost of damages, recalls, replacements, and potential legal liability for injuries often depends on the cause of the accident.
Forensic analysis specialists at Herrera, Stafford and Associates, LLC (HS&A) of El Paso, TX, are often called upon to determine why products fail. Recently, they used finite element analysis (FEA) software to evaluate the reasons for the failure of a compressor head cover at a Texas oil well.
HS&A was hired to research the accident's cause and determine whether the cover was flawed or user error had played a part. HS&A principal Juan Herrera, Ph.D., gathered specific information indicating that the compressor had been used for many years prior to the accident. However, the cover failed after the compressor was restarted following repairs and maintenance operations, releasing natural gas, which found an ignition source and resulted in an explosion and fire that seriously injured several workers and caused millions of dollars of property damage.
It was determined that a pneumatic impact wrench had been used to tighten the bolts on the head cover instead of the recommended torque wrench. Tests conducted in the HS&A laboratory using a similar impact wrench revealed that the recommended torque (80 ft-lbs.) was obtained with only 20 psi in the air-line supply. After his research and study of fragments from the failed cover, Herrera concluded that several errors had been committed in assembling the compressor head cover: the bolt/nut assembly had not been lubricated; the torque values were not equal in all eight bolts; the bolts were not equally screwed in, as each had different screw depths; the recommended torque value was exceeded; and the recommended torquing sequence had not been followed. As a result, the aluminum seal on the bottom of the cover deformed unevenly around the base.
With this information in hand, Anselmo Najera, HS&A design and analysis engineer, turned to the task of conducting a series of FEA analyses to study how different geometric features and loading conditions affected the stress results. He wanted to study design variations with and without the groove that is present on the top surface of the cover to discover whether that design feature led to high stress concentrations. In addition, there were concerns that manufacturing variations might be causing high stresses.
When a cast iron part such as the cover is produced, the mold often twists a degree or two. This results in a slight asymmetry in the thickness of the cover and location of the dome's apex. To study whether the asymmetry would affect stress results, Najera planned to create several variations of the model to represent the dome as designed, as manufactured with a shift in the dome apex, and as manufactured with both a shift in the dome apex and asymmetry of the thickness.
Najera created structured-mesh models using ALGOR's finite element model drawing tool, Superdraw. The structured mesh was chosen to specify the exact location of nodes for the loads and constraints. The asymmetry of the dome was created by incrementally modifying the angle while extruding the mesh. After one-quarter of the model was created, it was mirrored to complete the geometry.
To simulate friction between the cover and its aluminum seal, Najera modeled a thin layer of elements and gave it weak material properties so it would easily deform. For all models, the aluminum seal was completely restrained and material properties were applied based on information from the manufacturer. There were two sets of loading conditions that needed to be considered: the force of the bolts holding the cover in place and the pressure within the compressor.
A variety of bolt force loads were considered, ranging from 1,024 to 15,000 pounds. Najera also considered a scenario in which the bolt forces were not the same for all of the bolts. The maximum pressure of the compressor — 600 psi — was applied on the inside surface of the dome for most of the models and omitted for several analyses to isolate the effect of the bolt loads. In all, 18 different variations of the model were analyzed.
For each linear static stress analysis, Najera looked at the maximum principal stress and compared it to the yield stress of the material. He also looked for areas of stress concentrations. It was found that neither the groove nor the asymmetry of the dome was significant enough to cause a failure. The most significant factor was the force used to tighten the bolts. In the laboratory, Najera tested each of the 18 FEA cases to verify the simulation results, which showed good correlation between the FEA and laboratory results.