Customer experiences -Electrical OEM puts motion analysis software to the test -According to Sergei Fedorjaczenko, a design engineer at Carlingswitch, circuit breaker analysis is
similar to conducting an autopsy; much of the data is destroyed during the
triggering of the switch. Working Model allowed Fedorjaczenko to simulate the
motion of a circuit breaker and collect all the data he needed -- streamlining
new product testing.
Carlingswitch
of Plainville, Connecticut is an original equipment manufacturer of magnetic
circuit breakers and electrical switches for the appliance, transportation, and
HVAC industries. One of the more tedious, time-consuming tasks of switch and
breaker design is making prototypes for testing the responses of new designs.
This "build and try" method can take months just to make and assemble
the prototype parts before a test is run, a test that is only milliseconds in
duration. In that short span of time, the prototype is often deliberately
destroyed through bursts of high current, and then the painstaking work of
analyzing the failure begins. At this stage, the designer would be in for many
more weeks of refining and testing, until a new method of testing was
discovered.
Carlingswitch experimented with this
new testing method, often called "virtual prototyping," which utilizes
motion analysis software from MSC.Working Knowledge of San Mateo, California.
Working Model® allows engineers to test a mechanism on a desktop computer and
know with a high degree of accuracy whether it is feasible to build and whether
the device will perform as intended. Carlingswitch runs Working Model on the
Macintosh, but an equally powerful Microsoft® Windows version is also
available.
Testing Speed & Power
Carlingswitch circuit breakers are analyzed for two types of response:
speed and power, measuring how fast contacts open to prevent a surge of damaging
current. The circuit breaker's response time has to be measured either by
onboard instrumentation or by high-speed photography.
A power test involves measuring current
to determine if a breaker will withstand interruptions. If not, the sparks and
the parts fly. Only through these testing methods were Carlingswitch personnel
able to see a prototype circuit breaker's effectiveness.
The problem with these previous testing
methods is that arcing can hamper the photographic method of analysis by causing
a white-out effect that makes photographic results difficult to interpret, while
fried prototype components make instrumenting just as difficult to tell the
sequence of a destructive event (determining which springs and contacts failed
when).
What makes the analysis even more
difficult is that Carlingswitch circuit breakers are not big: the smallest ones
are approximately 1-1/2" square and 1/2" thick; the largest are
3" x 5" x 1" thick. Instrumenting and photographing such small
prototypes is not an easy task.
Given these testing and analysis
problems, design engineer Sergei Fedorjaczenko decided to put Working Model to
the test to determine whether virtual prototyping could save Carlingswitch money
and improve the company's design-to-market time.
Fedorjaczenko
said, "I've been building models of some of our existing products,
running simulations and comparing the results with data we have on the actual
performance of these products. This allows me to develop a confidence factor
in the Working Model simulation engine and to fine tune the simulation so that
it represents what actually happened with a particular mechanism. With that
confidence factor, I can work with a new design, begin changing geometry and
properties like weights of parts and so forth, try different spring forces,
and then use the simulator to predict how performance will be affected."
"Excellent GUI."
Developing a simulation in Working Model was a logical, straightforward process.
Fedorjaczenko imported one circuit breaker design from his Ashlar Vellum CAD
system using DXF file formats. He then assembled the circuit breaker's various
components in their proper relationship to each other. The Working Model menu
let Fedorjaczenko insert joints, stops, and spring forces in the proper
locations.
"I've used other simulators
before," explained Fedorjaczenko, "but the thing I like most about
Working Model is that it has an excellent graphical user interface [GUI]. The
geometry of a simulation is easy to create within Working Model, but it's just
as easy to import my own geometry. And you don't have to prepare any kind of
input file. I generally try and simplify geometry wherever possible so I don't
bring in a great deal of unnecessary data. I don't have to show intricate
details on the parts, as long as the part functions as it should and is
represented with the proper mass and inertia.
"What
Working Model does is break down the simulation into finite steps, performing
a series of calculations at each time-step or interval. That is one reason you
don't want to overburden your mechanism with too much detail. The more things
the simulator has to check at each time-step, the longer the simulation will
take."
Once the model is built, it's easy to
update and change variables, such as the mass of an object, or the force applied
to a rod or actuator. Working Model will automatically assign values of mass and
friction to objects in the simulator, or modify those variables easily because
the GUI is simple to use. For example, Fedorjaczenko used Working Model to
develop an equation of his own that let him simulate the magnetic force or pull
exerted by a solenoid on a circuit breaker armature.
"I took data from a previous test
we had done, because I knew how fast this armature responded to the pull of the
solenoid. The equation was then adjusted until the simulated armature responded
as it did in real life. Once this equation was developed, I was able to run a
simulation and determine what the responses would be when the current was
altered."
Working Model's powerful simulation
engine allowed Fedorjaczenko to confirm his suspicions about the responses of
this particular circuit breaker: "It verified that we were getting abrupt,
sharp impacts with the contact -- something we thought might be happening but
really had no way to instrument or verify precisely using any other
method."
Checking the Data.
Fedorjaczenko likens examining a destroyed circuit breaker prototype to
an autopsy, noting, "In the event of a failure, we try to see what failed
first and why. These events usually involve an interruption of current, which
means an event lasting no longer than 10 to 20 milliseconds. Very often, the
conclusions of this autopsy are speculative and not very precise. You get a feel
for it, but you're still not quite sure."
Fedorjaczenko used Working Model's
on-screen graphic simulation to check that the simulation was running correctly.
The tables and other data output provided by the simulation itself let him
evaluate what needed to be changed in order to improve the results. Working
Model can write to external files, including data files or to Microsoft Excel.
"A lot of the analysis
portion of prototype testing takes place after the simulation, when you sit
and look at a velocity curve or some other piece of data and decide what's
required next, and Working Model easily generates that kind of output,"
said Fedorjaczenko. "It only takes about five or ten minutes to change a
variable and run a particular simulation again, so you're able to do a lot of
'what ifs'. I went through literally dozens of these 'what ifs' in a matter of
hours."
Now that he's tested Working Model 2.0,
Fedorjaczenko believes in the product and its capabilities as a way to
streamline the testing phase of new products. By reducing the number of
destructive tests necessary to study a new circuit breaker design, Carlingswitch
will save countless worker hours and modeling materials, a significant savings
when factored over the course of just a few months. The motion analysis provided
by Working Model is highly accurate, so Fedorjaczenko will be able to rely on
the results of future simulations with confidence. Fedorjaczenko confirms,
"I think Working Model is a great tool. In just three months I've already
verified for myself and my management that this is a tool we should all be using
on a regular basis in the design of our new products." |