Saving money with FEA

FEA can be used to do more than just help designers, writes Tom Shelley Finite element analysis has come dramatically to the rescue of one manufacturer who was able to prove that failures of its heavy equipment were not down to design flaws, but allegedly resulted from customer mishandling. Moreover, the manufacturer could demonstrate exactly how this had happened. Not only did its actions save the company a small fortune by being able to make repairs under warranty, but it also afforded the opportunity to sell the customer a stronger (and more expensive) machine that would not break down – while enabling the customer to avoid damaging its machinery in the future. The tale of the mysteriously cracking coal haulers was revealed at the SolidWorks World event by Richard Wand, vice president of consulting engineer MJ Engineering. Coal haulers are large, very low profile trucks, designed to work in underground coal mines. The one in question was rated at 25 tonnes, with a chassis designed to support 16 tonnes of load, but capable of carrying a lot more, if heaped up. The problem was that the frames were cracking after only 90 to 180 days’ service. The cracks were not the welds, said Wand, so production processes were not at fault, but went through steel plates ½ to 1½ inches (13 to 38mm) thick. The customer claimed this was due to a design flaw and the manufacturer was concerned this might be the case. “They had previously changed some key components in the frame, and added some cutouts without reinforcements in areas they felt were over-designed,” confirmed Wand. The solution was to mesh the model and use Cosmos to undertake repeated FEA analyses at different loads, and to employ Cosmos Motion to simulate wheel loads under various conditions: all four wheels on the ground, three on the ground, and then two. According to Wand, they had a great deal of trouble with the 3D model of the vehicle frame as supplied, and first had to remove all mates from the model tree and make it consistent before they could mesh it. They were then obliged to run the model as a mixed surface and volume mesh, otherwise the complexity of the shape would have made it impossible to analyse. The frame was made out of A572 Grade 50 high strength steel, with a yield strength of 50,000psi (345MPa) – which was “a little on the brittle side”, conceded Wand, but quite adequate for purpose. The manufacturer first undertook static FEA analyses, starting with 20 tonnes on all four wheels, increasing this to 25 tonnes, 30 tonnes and 38 tonnes. Meshing took 60-90 minutes and FEA run times were two to three hours. Even 38 tonnes did not produce any stresses high enough to cause cracks. They then considered the ground – there was a change in gradient outside the mine entrance – and used Cosmos Motion to model loads on wheels, as a result of driving the machine with a 45 tonne load at 8mph diagonally over the bump. The model showed that, when the coal hauler went over the change in gradient this way, it threw all the weight on to two wheels, one at the front and the opposite one at the back. It was this situation that was producing the loads that caused the cracks. “They were able to explain to the customer that future repairs would be done at the customer’s expense,” said Wand, “They strongly suggested to the customer that they upgrade to bigger units, capable of carrying bigger loads.” Whether the customer purchased new, stronger machines, or whether they told their operators to be more careful or just flattened out the change in gradient outside the mine entrance was not evident. However, what was clear was that the actions taken saved the manufacturer a substantial amount of money, as well as establishing there was nothing wrong with the design of the coal hauler.