Withstanding the elements with FEA

I am the structural engineer at Anchor Industries Inc., a leading manufacturer of tents and other lightweight, portable buildings and structures for more than a century. With expertise in numerous markets, we wanted to get back into the military market for Vehicle Maintenance Shelters (VMS).

The fully-assembled VMS (shown above) stores and shelters military vehicles, protecting personnel working on the vehicles from extreme weather conditions. It may also be used for a portable military command post and for short-term housing of troops.

Our first VMS that we designed years ago and since discontinued would have been overpriced in today's market. We needed a more competitive, efficient and optimized shelter design. From past experience, I knew I could quickly and inexpensively model and test VMS concepts using Finite Element Analysis (FEA) software, so I decided to use ALGOR FEA to design a new VMS for the U.S. Military.

In this article, I will walk you through the three steps I took to model and test the structure.

Step 1: Building the Model

Using ALGOR software, I created an FEA model based on the use of existing extrusions. The frame was composed of two parts: arches (legs and rafters) and purlins (horizontal beams that connect the arches and support the fabric). I defined the arches and purlins as beam elements, using elements from 1 to 2 ft in length. I modeled the fabric for the walls and roof using membrane elements. I used a rectangular mesh on the lower ends of the cover and a triangular mesh for the apexes of the ends and specified the entire cover as 0.02-inch-thick, vinyl-coated fabric. Since the splices are critical stress points, I used shorter elements to correctly model both the leg and purlin splices. The complete model incorporated about 1,200 membrane elements for the cover and 600 beam elements for the frame.

Step 2: Defining and Simulating the Loads

With the initial FEA model constructed, I began to define loads. The dead loads, representing the hanging accessories inside the structure, were constant forces of 100 lbs applied at the center of each rafter and at the peak, for a total of 900 lbs. I then applied wind loads as high as nearly 7 psf and a snow load of 10 psf loaded vertically on the roof. I applied five load configurations to the model, including three separate wind loadings, one snow load and one combination of wind and dead loads (see load table). The ground was fixed using boundary conditions.

I then used ALGOR's Mechanical Event Simulation (MES) software to accurately simulate the deflection of the fabric and the nonlinear effects of the X-cables included in the concepts (see Model Iterations). Although the forces were steady, each loading comprised a timed event lasting 20 seconds — consisting of one second of rest to permit the initial tension in the X-cables to distribute itself, then 17 seconds of increasing load, followed by two seconds of rest. The rest period allowed transients in the model to settle.

Step 3: Verifying the Model

After several design iterations, I built a prototype of the frame. To verify the ALGOR analysis results, I applied a 240-lb load at the purlin splices, the most vulnerable point in the design. I previously hand-calculated this load and applied it to a FEA model to verify that it produced stresses marginally exceeding those revealed in any of the five load combinations tested in ALGOR. In the worst case, the snow load, roof purlin splices were very near the yield strength. I was anxious to verify that the splices performed as the FEA software predicted and found that the software was correct.

The beams that the ALGOR analysis determined we could use were much lighter than expected, representing a 40 percent overall weight reduction. This provided us with an extremely competitive product.