|
At Textron's AAI Division in Hunt Valley, MD, unmanned boats for military and commercial
applications are a specialty. Recently its electronics designers developed a common marine power
panel for installation in a boat engine room. The six-sided panel could house any of several
circuit configurations, some of which produced high heat loads. When the configurations were
submitted for thermal analysis, the electronics group dictated that the circuits were fixed, and
nothing, not even the thermal pads could be changed.
The need to accommodate multiple component layouts left the thermal engineer with only two
possible vent locations. There was no room for a heat pipe, or a bigger heat sink. "The
only option," he explains, "was to manage air flow as efficiently as possible in order to remove
the maximum heat."
First, he verified that airflow direction would have no impact on cooling. Then he split the
ducting into 2 paths, biasing flow to the area with the highest heat. Using Coolit, he
identified the most effective wall location to deliver adequate airflow to the hot side, with
enough leftover to handle the secondary path.
As the thermal analysis proceeded, the electronic circuits were still in a state of flux.
There seemed to be new versions constantly coming in for analysis. "Many times I had to
rebuild the whole model," explains the thermal engineer. "Fortunately, that proved pretty
easy to do. With Coolit, I could start a new model from scratch, have it up and running, and
deliver results in the same day."
"Sometimes Coolit surprises you," he observes. "You expect one thing to happen and another
thing does. One surprise was that small changes in object orientation made a significant impact
on cooling."
At some point during the process, the electronics designers learned that the thermal engineer
assigned to their project was a first time Coolit user.
"They expressed their concerns to me," notes the thermal engineering manager, " because they
couldn't understand how our thermal engineer could be getting valid answers so quickly, when
they couldn't. Another group within their general organization had been using a different
thermal package and spent weeks running thermal analyses. Often the problems ran overnight, and
even then, they couldn't get conclusive answers."
"When thermal engineer and I showed them how quickly we could obtain a Coolit solution,
they were amazed. Not only was Coolit at least five times faster than what they were using, our
answers were conclusive and theirs weren't."
"I used Coolit ribbons," explains the thermal engineer, "so they could visualize the air
flow, and velocity and temperature changes. That proved a definite eye opener."
Though component changes were considered a no-no, he went on to prove that a slight change in
the thermal pads would greatly decrease the amount of cooling required. The presented evidence
was so strong that the designers relented and changed the pads. "They wouldn't have changed
their minds if I hadn't been able to prove it," he notes.
When the design was completed, the worst case scenario proved the predictions to be
accurate; there was more than a 5 degree buffer between maximum operating temperature and
design limits.
|
