Customer experiences -Plasma-Therm optimizes a new product using Kinematics/Dynamics Analysis -
a better automated manufacturing system still means accounting for the
inevitable need for human intervention, if only on occasion, to maintain
the system. Recently, Plasma-Therm, Inc. of St. Petersburg, Florida,
expanded its product line with a robotic batch loading system for a
plasma deposition and etching product called the Versalok. Plasma-Therm
machinery must perform its tasks in clean environments, using vacuum
chambers to bake silicon materials for flat panel displays, read-write
heads and other products.
Access to the automated robot loader and plasma chamber combination
was designed to be through a heavy, 25.5-inch-square lid that would
maintain the vacuum seal. However, the lid, which weighed 110 pounds,
had to be easy enough for one person to open while the robotic loader
needed servicing. Plasma-Therm mandated that gas shocks be used instead
of a servo motor lift mechanism to reduce the overall cost of the new
Versalock production system. The challenge was, what size shock would be
needed to make the lid easy to lift, yet be small enough to fit within
tghe restricted confines of the product. Because the Versalock's robotic
loading module had to interface with a vertical elevator holding the
silicon products to be treated, there were a lot of constraints on the
available locations for the shocks due to the cramped space.
Accesible to Engineers and Designers Alike.
Normally, a Plasma-Therm engineer with an extensive physics
background would have to do a lot of trial and error, solving equations,
to determine the necessary gas shock force to use and experiment with
prototypes to determine where to place the shocks for maximum
effectiveness. Instead, it took one designer with little physics
background just three days to use a kinematics/dynamics analysis program
to answer these questions and produce results. Jeff Barriault, a
mechanical designer with Plasma-Therm, chose WOrking Model from
MSC.Software to solve his design questions. Working Model is a motion
simulation program that applies the laws of physics to CAD models. Armed
with the software and some dimension constraints, Barriault proceeded to
attack the problem.
After sketching the basic chamber and lid configuration in AutoCAD,
Barriault used dimensions from a manufacturer's catalog to insert a gas
shock into his design. He then imported his preliminary AutoCAD layout
drawing into Working Model via a DXF file.
Barriault then added some constraints to his model so the
"virtual" shock would opperate just like a gas shock; e.g.,
the more the shock was compressed, the higher the pressure. "I
entered a formula supplied by the manufacturer in one of the Working
MOdel dialog boxes to create that effect, and then added other
constraints to limit the extension and compression of the shock."
Constraints are added neatly, with just a few mouse clicks, by using
Working Model's pulldown menus and dialog boxes that prompt the user to
enter either custom equations or select from default equations (based
upon the dimensions and other parameters assigned to the masses in the
Using a rope constraint and a separator constraint in the slot joint
of the gas shock, Barriault selected "Run" to watch how
forcefully the shock would open the lid when force was applied to the
handle. He then performed more iterations, placing the shock in
different locations, first locking down one end and then putting the
other end at new position on the lid to see the change in force in force
necessary to accomplish the same task of lifting the lid. (Barriault
planned to use two shocks, but he only used one in the simulations and
doubled the results.)
"At the pivot point, I put in a motor and moved it a constant
velocity," explained Barriault. "When the lid was at a certain
angle, I could see how much torque was required to keep it open because
I had Working Model display a meter while the simulation was running to
measure the force necessary." Barriault kept relocating the shock
and running the simulation again until he found the optimum spot, which
turned out to be a position where only seven pounds of force would be
required to open the lid, and 10 pounds of force needed to close it.
"That's not bad for a 110-pound lid," he added.
A fellow designer suggested that Barriault refine his test of the
shock by simulated shock wear, to see how much force the shock would
exert after six months of use. To make sure he had a ten percent margin
of safety, Barriault re-entered the force equation in Working Model with
a reduced force. Barriault ran the same simulation, using the optimum
position of the shocks, but with 10 percent less forceto make sure the
size of the shock and its location on the chamber would still be
sufficient to raise and hold the lid in an open position. It did.
Had he tried to simulate all these tests on paper, he simply coudn't
have, because of his own lack of training in physics. Barriault's
supervisor, himself an engineer, was impressed by Barriault's ability to
solve the design engineering task at hand using Working Model. The
supervisor insisted on doing his own equations to double check Working
Model's calculations and conclusions. They were the same.
"It would have taken a lot longer to get the Versalock design
correct," according to Barriault, "probably weeks."
Instead, Plasma-Therm's new Versalock is now in production after
Barriault worked on the vacuum chamber lid problem for just a few days.
Plasma-Therm plans to use Working Model to streamline, shorten, and
lower the costs associates with similar design engineering tasks in
future product launches.