 |
Research in this
area will include the development of analytical tools for body shop
processing, fixture process and design, material handling, and simulation of
body assembly systems. Special emphasis will be on the development of
scientific methodologies for managing and controlling variation in body
design and manufacturing, e.g., systematic methodologies for predicting and
simulating variation in multi-leveled assembly systems, incorporating
dissimilar materials and their inherent process capability, robust design,
and process monitoring, diagnosis, and feedback. Two
projects are being carried out in this thrust area.
Current Project 1
DIMENSIONAL
VARIATION ANALYSIS FOR LIGHT WEIGHT VEHICLE STRUCTURES
As the demand for highly fuel-efficient and lightweight automobiles
increases, aluminum alloys are considered to be the alternative material
competing with steels. Even though aluminum has been used in the manufacture
of automotive assemblies for many years, its use has been limited to
low-volume vehicles and closure panels such as doors, hood and deck lids.
Nevertheless aluminum has a number of advantages for auto body application
compared to steel. Since the density of aluminum is just one-third that of
steel, reductions in the weight of body-in-whites can be achieved up to 50%.
With
reduced overall vehicle weight, fuel consumption can be significantly
reduced. However, the intensive use of aluminum on vehicles structure
creates new challenges for manufacturing engineers due to its different
formability and weldability. Unlike conventional steel bodies, which have
been mostly sheet based and resistance spot-welded, aluminum bodies can use
a combination of fabrication and joining techniques. For example, aluminum
can be cast, extruded, or stamped, and these components can be joined
together using Gas Metal Arc Welding (GMAW), Laser welding (LW), Resistance
Spot Welding (RSW), or mechanical joining, such as self-piercing rivet (SPR).
As in steel body development, it
is important to predict the dimensional variation of the bodies during
product design and manufacturing system development. But there is added
complexity in variation analysis for aluminum intensive vehicles due to the
various component fabrication and joining techniques mentioned above. Some
of the components can be very stiff, and therefore, its variation may play a
more significant role in the overall dimensions of the assembly. In
addition, the natural process variation and required tolerances for these
component fabrication processes are not known to auto manufacturers since
these processes are relatively new. This project seeks to develop new
improved variation simulation models applicable for aluminum intensive
bodies and derive design guidelines for these component fabrication and body
assembly processes.
The
University of Michigan has been developing methodologies for variation
simulation of compliant assembly since 1995. Specifically, under the
support of the GM Collaborative Research Laboratory at the University of
Michigan from 1998-2002, the variation simulation approach has been expanded
to system level by combining linear mechanics, statistics, and a state space
representation. Three sources of variation were analyzed, part variation,
fixture variation and weld gun variation. However, the methodology was
based on sheet structures and the joint models were resistance welding.
This project will focus on assembly variation analysis for aluminum
intensive vehicles including structural elements different than sheet metal
parts.
Current
Project 2
Adaptive
Control of Assembly Quality using Programmable Tooling
The dimensional
quality of automobile body is important to product functionality, appearance
and down-stream assembly processes. In the past, dimensional control for
auto body assembly has been performed using a combination of two
approaches: (1) robust design, e.g., through the use of slip planes. (2)
Diagnosis and variation reduction using in-process measurement and
multivariate statistics.
A programmable body assembly tooling system
has been developed at GM to provide flexibility in assembly. It allows
adjustment of the locators and clamps on a part-to-part basis, and is being
implemented in production operations at a number of GM assembly plants. It
provides an opportunity to integrate a multivariate process model with
feedback adjustment. The objective of this project is to explore the use of
this and other programmable tooling concepts in controlling the dimensional
quality of auto body assembly processes.
|
Assembly