Multimedia

IMAGES

Close up annotated shot of the device showing the swinging impactor system.

Annotated picture of the device as manufactured, assembled, and tested. The dolphin carrier is in the background!

Modal analysis of OptiStruct Interpretation; modes and stiffness (not shown) were in line with expectations.

Comparison of bottom of plate as machined (left) and as designed (right). Yay!

SolidWorks interpretation of OptiStruct output; some load cases were not considered in the optimization study, so engineering intuition had to be used to "fill in" material where real samples might be mounted.

OptiStruct result for ~0.7 GN/m stiffness (left) and 0.9 GN/m stiffness (right). Note the 0.9 GN/m figure is approaching the theoretical maximum of a solid plate (hence the distinct lack of pocketing).

Image showing a stamped tube. This is tube 15-2 (-2 indicates the right side of the frame). The arrow points to the rear of the frame.

SolidWorks render of the foam/slider; some of the delrin contactors can be seen through the transparent wood track.

Results for linear buckling of frame at with impactor at 45 degrees during swing. FoS > 18 for all modes. This is the lowest factor of safety for all tested load cases.

Results for linear buckling of frame at bottom of impactor swing. FoS > 30 for all modes.

Results for linear buckling of frame during lifting of impactor; FoS > 50 for all modes.

Close up of frame mesh, highlighting boundary conditions and element size; shear panel, blue, removed from final model due to manufacturing time, cost, and lack of need.

HyperWorks 1D beam mesh of the impactor frame; some changes from a 4 foot model (not depicted) are highlighted; further modifications (removal of tubes and squaring of frame dimensions) took place after analysis results.

Final assembly overview (some modifications were made later).

Standardized tab preview (4 of 12 tabs shown).

Closeup of impactor/excitation surface interface.

Impactor/excitation interface, shows where impactors (not shown) are attached.

Waxed wood/Acetal Delrin® slider.

Underside of wood platform; shows Acetal Delrin® contactors and foam bonded to platform.

Interpretation of optimized model, including deflection (stiffness) constraints. Approximately 60kg (versus unoptimized 74kg).

Altair Hyperworks OptiStruct results: 4x cyclic patterning with internal symmetry, stiffness constraint of 8E8 N/m, and 2.5kHz and 2.0kHz minimum natural frequency, respectively.

Demonstration of an non-ideal system: fixture plate has modes too low and foam has modes too high.

Demonstration of ideal system: fixture plate has high modes (noisy, but easily filtered), foam has low modes (large displacement over a time scale much bigger than what is observed).

CAM file preview for an early optimization.

Early optimization iteration, with no stiffness constraint; natural frequency of approximately 2.7kHz, mass of approximately 48kg.

Altair HyperWorks OptiStruct result: 4x, internally symmetric (left one is about center of the sides, the right one is about the diagonals) cyclic patterning for the 7.62cm (3in.) thick plate. Natural frequencies of about 2.7kHz (aiming for 3kHz).

Altair HyperWorks OptiStruct result: 2 plane symmetry (left, 7.62cm or 3in. thick plate) and 4x internally symmetric cyclic patterning for the thicker (10.16cm or 4in.) plate; Aiming for 3kHz and 4kHz natural frequencies, respectively.

Meshing setup: Checking for stiffness.

Meshing setup: checking for modes with test article mass attached.

Altair HyperWorks OptiStruct setup: Design and non-Design spaces indicated and optimized

Mesh setup for modal analysis of foam block with plate mass on top. Plate modeled as point mass and constrained by rigid spiders to foam surface.

Possible fixture plate construction method: aluminum plates sandwiching aluminum bars, spot welded at small intervals; an alternative to a large plate with pockets machined out from the bottom.

Old method for doing modal analysis with plate weight; artificially increasing foam density in region where plate sits. Modes appear too low this way.

Examples of the first 2 modes for the foam as compressed and uncompressed; old analysis NOT including plate mass on top; additionally, modes shown will likely be damped (friction against fixture plate would prevent "shear" movement between plate and foam).

Detail view of compressed foam modes. Again, no plate mass included.

Nonlinear static analysis of plate mass on foam showing overall compression of foam due to weight.

Mesh setup for nonlinear and linear static foam deflection.

ATA specified chart showing pulse shapes of current table and 3 design targets highlighted in red borders. Yellow regions show where more prototypes similar to our project could extend the range of shock testing.