By Michael Rigby
Chief Consultant, Owner
AnJen Solutions
Westford, Mass.

A design optimized for some property such as minimum weight demands use of advanced CFD to predict airflow, temperature, and heat transfer, well in advance of physical prototypes. A good example comes from a recent thermal simulation for an elevator with linear synchronous motors (LSM). There are no cables, pulleys, or counterweights. The elevator, from MagneMotion, travels 148 fpm and has a verticallift capacity of 24 tons.

Conventional horizontal material transporters work in benign environments and have payload weights that are relatively low. They just need 2D FEA to predict system temperatures. Because such designs handle light loads, an approximation of the heat transfer to the surrounding air is good enough.

An LSM elevator, on the other hand, carries more weight and typically works in a severe environment. The elevator consists of four rails mounted vertically in the elevator shaft. The rails contain copper coils and attach to a heat sink. Stators (stationary magnets) mount to the elevator platform. Electric current passing through the copper coils generates a magnetic field which propels the elevator platform. A control system synchronizes the traveling magnetic field and elevator platform.

Overall weight of the new LSM was critical and 2D FEA wasn’t accurate enough to assess the heat sink surfaceto- air conditions. But Flotherm CFD software from Flomerics provided the details to understand the heat transfer from rail to heat sink and heat sink to surroundings.

LSM geometry came in the form of a STEP file. The model consisted of seven individual aluminum heat sinks conductively coupled with aluminum blocks at the heat-sink bases. The software meshed the structure using cuboid elements.

The total power entering the heat sink was 600 W, uniformly distributed across the back surface (the interface with the LSM rail). The distance from the heat-sink back surface to the wall of the elevator shaft was 130.25 mm. The flow cross section was 174.92 431.6 mm. The only flow condition considered was natural convection induced by the 600-W heat load. Material properties included the thermal conductivity of aluminum and air, as well as air’s viscosity, density, specific heat, and expansivity.

Flotherm solved the complete thermal problem. It determined the heat conduction from the motor through the mechanical structure and the heat sink, and the heat convection from the mechanical structure and heat sink to the air. The software also solved the underlying Navier-Stokes equations to determine the airflow caused by the heat loading.

The initial heat sink weighed 68 lb, with a 94.65-mm fin height, 11.47-mm fin thickness, and 10.2-mm base thickness. The temperature at the interface between the motor and the structure was 114.7°F — safely below the maximum interface temperature of 150°F.

The goal was to reduce the weight of the heat sink without unduly raising the interface temperature. We evaluated 11 different design scenarios, varying the heatsink fin count, spacing, and thickness. When all was said and done, a fin count of 15 and a fin thickness of 3 mm minimized weight with an acceptable temperature. The optimized heat sink weighed only 39 lb.

Additional tests helped evaluate design alternatives. For example, shortening the fins raised the interface temperature. Boosting fin count to 30 and cutting fin thickness to 1 mm did not help. Bolting the heat sink directly to the wall of the elevator just made things hotter. Using one large vent hole instead of smaller holes slightly cooled the interface but was more expensive because of the costly machining process. And eliminating the vent holes had a negative impact.