Composite Material Tank Head Refit

A client was considering adding an agitator to the top of a tank by cutting a hole in the tank head, adding a neck, and then mounting the agitator to the neck flange. The tank was a composite material, as would be the neck and supporting composite layup. While the vessel appeared to be able to support the agitator's weight, it wasn't clear if the agitator's dynamic load would exceed the tankhead's strength. The vendor who would be contracted to perform the modification provided material information as well as specifications for the neck, reinforcing composite layup, and a suggested method for reinforcing the neck at the neck-head joint and distributing the load. This method was to add a vertical flange structure extending from below the neck flange to some radius away from the neck-head joint. While the vendor had guidelines for the various thicknesses, KEG was contracted to determine if the tank head modifications were feasible, and if so, determine actual reinforcing geometry.

Approach:  Using traditional engineering methods, the forces and moments induced by the motor were determined. While the agitator's dead weight and torque would be symmetrical with respect to the tank head, the precession loads would not be. Therefore, a 3-D FEA model was needed instead of a simple 2-D axisymetric model. There were two areas of interest: the head-neck joint and the tank head region immediately around the outside of the reinforcement. First, the vendor's guidelines were modeled. While the neck joint was sufficiently protected by the reinforcement, the stresses in the tankhead around the reinforcement were too high for the material. Another structure, the cone-shaped layup pictured below, would be more difficult to layup but would better distribute the stresses.  Several thicknesses and configurations were attempted. Again, the asymmetrical forces and moments due to precession exceeded the tank head's strength. The band of high stresses shown below, in conjunction with deformation and strain plots, indicated the agitator would tear through the tank head shortly after startup.

Solution:  The client and vendor were advised of the tank head's predicted performance. A steel structure external to the tank was used to support the agitator and its loads.

Go To Top

ASME Code Analysis for Reactor Coils

A client was required to submit ASME Section VIII, Div. 1 calculations for a series of reactor coils for a chemical process. Elevated temperatures, high pressure, and vacuum were varibable conditions. ASME code software can  automate Section VIII, Div. 1 vessel design with a great deal of accuracy and efficiency. However, these software packages do not account for the code requirements particular for coils. 

Approach:  Using the ASME Boiler and Pressure Vessel codes on CD-ROM and traditional engineering methods, a  MathCad file was developed to automate the process. The ASME code on CD-ROM, in which all text is fully searchable, allowed all sections to be searched, including comments, notation, and interpretations. As part of the search, it was determined while the stresses due to differential thermal growth from the specified operating conditions must be calculated, there was no code calculation method provided that would fit the situation. A simple FEA model of a section of the coil, spanning from support to support, was developed. The various conditions, including Maximum Allowable Working Pressure (MAWP) and temperatures, were applied. The results of the differential thermal growth as well as the induced stresses and strains were compared to manual calculations of the thermal growth of the support system and were determined to be within code specifications. 

Solution:  The client received a complete set of code calculations, including FEA plots and model descriptions, that satisfied all requirements by their ASME inspector.

Go To Top

Nonlinear Elastic-Plastic Analysis of Traffic Barrier

A client had developed a vehicle barrier gate to protect railroad crossings. When lowered, the gate was secured at both ends with three horizontal cables interlaced with smaller vertical cables, all encased in supporting metal tubing.  Instead of stopping a vehicle immediately, the barrier was designed to have the external tube structure break away and the cables stretch without breaking, cushioning the stop within specified nonfatal levels without allowing the vehicle to reach the railroad tracks. Specific performance criteria was based on Federal Highway Administration's NCHRP Report 350 TL-2 standards. Based on vendor-supplied cable test data, the horizontal cables nominal diameters were increased from the original design. The client wanted their design analyzed for reliability to ensure it fell within specified benchmarks prior to crash testing the prototype.  Forces on the cable anchors were also needed for vendor swedge specifications.

Approach:  The vehicle's kinetic energy is transferred to the cables' strain energy. Where most engineering is done in a material's elastic region, this design required the material to be loaded well into the plastic region. While material in the elastic region responds linearly to loads (doubling the load equals double the bending), the bulk of the energy in this case is absorbed in the nonlinear portion of cable's response. Using mathematical modeling and interpretation of the vendor's cable testing, which was a different diameter than the cables in the final design, a relationship was established between the cable's strain and the kinetic energy the cable could absorb as it stretched. This relationship yielded a projected travel distance and acceleration force (measured in terms of gravity, or "g-forces"). The travel distance was used in a nonlinear Finite Element Analysis of the cable barrier system to examine the resultant forces on the cable anchors as well as the the interaction between the horizontal main cables and the vertical lacing cables and the system response as a whole.  Careful consideration was given to analysis assumptions such as the vehicle dynamics, vehicle deformation, energy absorbed by the destroyed support tubing, and the anchor reliability.

Solution: The reliability analysis showed the largest test subject, a pickup truck, would travel 155 inches and experience 10.5 g's, assuming there would be no vehicle dynamics outside the crash test parameters. The FEA results provided cable anchor information needed for specifying swedges and indicated a 2.1 safety factor in terms of strain loading versus rupture. Crash tests conformed to predicted results within 5 percent, passing the required performance benchmarks with flying colors.

Follow Up:  During development, the barrier passed several physical crash tests using sedans and pickup trucks. The final design passed the pickup truck crash test. The client contracted KEG to show that pickup truck crash test of the final design, coupled with data from previous crash test results, demonstrated the barrier did not require additional testing. KEG conducted a detailed analysis of the design and the crash test results. KEG's analysis, coupled with the test reports from the testing facility, was accepted by the FHWA/NHTSA (National Highway Transportation Safety Administration) National Crash Analysis Center in place of additional testing of the final design, allowing the client to avoid additional testing expenses or delays in bringing their product to market.

Conceptual Design of Skid-Mounted Process

A client was in the process of developing a new skid-mounted system using steam to create product. The process engineering firm was able to provide plan and elevation drawings of the equipment, including various fittings and commercially-available components as well as the newly designed components; however, the firm could not provide any sort of illustration or rendering the client could use for presentations or marketing. The 2-D drawings were highly technical and required excessive explanation or required that the audience members were technically proficient in the specific field. 

Approach:  KEG used the 2-D drawing to develop a computer 3-D model to scale. Using our technical knowledge, we were able to make the majority of the illustration technically accurate while protecting proprietary process design and data. The primary areas of concern were the two vessels with concentric coils where the majority of the process takes place. These vessels were rendered transparent, showing these areas. Colors were selected to differentiate equipment items and identify specific process conditions, not for realism. The client worked with KEG to select the best angle, elevation and "camera lens" to suit their purposes. During the development of this illustration, a process change was introduced. Because of our expertise in this work, KEG was able to change the equipment layout to match the process revision quickly and easily without requiring a change order charge.