Performance-centric Product Development:

Putting Engineering back into CAE

Uwe Schramm
Altair Engineering, Inc., Irvine, USA

1. Innovation and Enterprise

New and innovative products are necessary for the survival of businesses today. Profitable enterprises
need a constant stream of creative ideas moved into products to win new customers and satisfy
existing customers needs. What computer aided tools will be needed in the future to aid the creative
process of design and how will they be employed? Do current design tools actually inhibit the creative
process and how can this be rectified? How can computers aid the ideation phase of product
development? A new paradigm is that in the ideation phase, when the freedom to create and modify a
design is the largest, computer aided engineering can lead to quantification of design performance
based on the concept instead of unquantified decision making. Computational techniques become
tools for performance assessment and design creation rather than just tools for validating an idea.
Traditional CAD/CAE tools are used to document a verified design rather than driving the design which
is then verified later. This will require a fundamental shift in how organizations treat design and the
roles of the designers and analysts but in the end, this will lead to a more efficient design process
which will produce more innovative products. The requirements and technology for such a vision will
be discussed.
The survival of businesses in todays economy depends on new and innovative products. Profitable
enterprises need a constant stream of creative ideas moved into products to win new customers and
satisfy existing customers needs. For industries that market manufactured goods designing these
products is crucial to win in the market. The process that turns a physical principle that satisfies an
existing or perceived consumer need into a product is commonly called the Design Process or simply
Design.
Design is often used with different meanings. It stands synonymous for the conception, the aesthetic
styling (Industrial Design), the drafting of manufacturing drawings, or the entire process from ideation
to manufacturing readiness. In any case Design is always related to the creative portion of product
development.
There is no doubt that the enterprise that brings the most innovative products with the best
performance to market first will win (or which is perceived to do this). For some marketing is the means
to great business success, but not without a great product. The enterprise that best manages the
process of designing new products will end up with the right product first. Naturally, the needs for
product innovation, shorter time-to-market and best product performance are the biggest drivers for
improving the design process.
Certainly there are other factors making a business successful besides having great products. Factors
such as manufacturing cost, customer service, brand and more can be crucial for business success. In
the end anything that maximizes profit is a contributor to business success. However, there are many
examples where business success was based on product innovation. Examples are Apples iPod or
the recent success of Japanese car manufactures.

2. Improving the Design Process

If we just stick to the design process (our line of business), we can maximize our customers profit by
reducing cost (innovation, be more creative), time-to-market (shorter cycle), performance (get out the
last reserve) and/or driving up price (innovation, never seen before), time-to-market (be first),
performance (best). Naturally, the needs for product innovation, shorter time-to-market and best
product performance are the biggest drivers for improving the design process.
There are two ways of improving the design process. Firstly, it is an organizational matter to create a
company culture that allows for innovative ideas to turn into products fast. Secondly, it is a matter of
what design tools are used in the process.

An idealistic view of the product life has the following phases: Market analysis, Ideation, Creation,
Validation, Manufacturing, Rollout, Product life. Associate with that is of course product data that
needs to be managed. To manage all aspects of this is typically called Product Lifecycle Management
or PLM. Design is the portion of that comprised of Ideation (Conception), Creation, and Validation.
Validation includes performance and manufacturing validation.
In the old days pen, paper, clay and other raw materials where the main tools for ideation and creation,
while physical testing (make and brake) was used for validation. Design data was usually stored on
paper. Computers have changed this. Design creation and data storage became computer aided. This
actually reduced design freedom initially to what the software allows to do. If it can only create straight
lines the product has only straight edges. Over the years software has evolved that allows more
design freedom including styling and surface design. The design creation is called Computer Aided
Design or CAD, the data management Product Data Management or PDM.
Virtual testing using computational techniques made physical testing less important. These
computational techniques are frequently referred to as Computer Aided Engineering or CAE. This term
is somewhat overstating their role in the design process as virtual validation techniques.
Despite the computerization of design, the traditional process remained largely intact. The ideation still
remains a game of ideas based on experience and intuition. Naturally the number of design iterations
has not been reduced by the computerization. Costly iterations that involve changes to the concept
can not be avoided this way. The only thing that has changed is the tools the different design tasks are
performed with. Interestingly some call the computerization of the traditional process Virtual Product
Development, which leads to the idea of a highly parallelized creation and validation phase to reduce
cycle time. However, this might squeeze more out of each individual employee or make change
management easier, but it does not lead to an overall reduction in redesigns and design iterations.
In general the current design process consists of the phases Ideation, CAD, and CAE using PDM to
manage the data. CAE data remains largely unmanaged and therefore this vast source of knowledge
remains untapped.
The ideas of Virtual Product Development and Product Lifecycle Management are centered on the
geometric data of the product. More, the geometric data is a prerequisite for the design process to
proceed to the validation phase. Again, this does not change when virtual validation is parallelized with
the actual act of creating the geometry. The designer does the validation because he/she is close to
the geometry and (of all things!) knows how to drive the CAD software.
The idea that engineering managers pondered with is that early simulation, like early detection of
design flaws will lead to shorter cycle time. Hence, many have identified moving computational
validation early in the design process. This is a step in the right direction. However, existing concepts
of Virtual Product Development dont turn this into new ways of designing better they just introduce
another task to the concept phase.
Following the notion of bringing computational methods forward in the design process, why not base
the ideation also on computational methods? Why not computerize the ideation phase entirely? Why
not do all the validation on a concept of the design before recording and detailing the geometry in a
CAD system? This would move all the critical design decisions to the beginning of the design process
thus avoiding costly design iterations on the detailed geometry or late changes to a failed design
concept. Performance and manufacturability can be quantified before the time consuming CAD takes
place. Different concepts can be evaluated in a virtual environment. Plug and play allows addressing
flaws in the concept from the beginning. A virtual prototype could be subjected to early customer
feedback for unquantifiable performance measures.
The new paradigm is that in the ideation phase when the freedom to create and modify a design is the
largest computer aided engineering can lead to quantification of design performance based on the
concept instead of un-quantified decision making. Computational techniques become tools for
performance assessment and design creation rather than just tool for validating an idea.

3. Toolset for Performance-based Design Process

The current toolset of CAE includes software for meshing (turning geometric data into a grid of nodes
for computational evaluation), modeling (application of product attributes as materials, loads), solving
(the actual computation), visualization, reporting, and collaboration. Most tools are equipped with
programmable interfaces (API) to adapt to the user environment and for automation. Interfaces exist to
the CAD world to retrieve design and product data from PDM systems.
Are these tools really dependent on CAD data to be useful?
Computational models store geometric information is stored in form of a mesh or facets. It can be dealt
with like with geometry, it can be morphed and manipulated. Parts can easily be exchanged or
modified to study the effect of design changes. It only depends on the sophistication of the morphing
and other manipulation tools. Penetration checking, simple and sophisticated performance and
manufacturing feasibility evaluation, cost estimation, rendering can be performed just like in a
computer game of sorts.
If such models are to be used to create designs, how should that work?
Firstly, of course, there exists data from past products, typically a large contributor to the development
of new designs. This data and the knowledge associated with it can be utilized to create and validate
new ideas.
Secondly, computational tools are available to create geometry out of packaging and loading
information. The optimal material layout of designs can be found this way, in essence creating a
design from functional requirements and design space information. At this point also the methods of
manufacturing can be considered to arrive at concepts that are manufacturable and perform to
specification. This is a completely new approach where the functional requirement drives the ideation.
Thirdly, human interaction and creativity is involved in this process. Because of automation of standard
CAE tasks and software with superior user interfaces the human interacts with the virtual design as if it
is real, immediately being able to validate performance and manufacturability.
In essence the entire process of designing a product can be done without actually detailing the
geometry in CAD. All design decisions can be made based on a simplified model suitable for
computational evaluation. Certainly, the multi-disciplinarity of the operation environment, the role of
uncertainties in production and usage and manufacturability must be included in the evaluation of the
design. Efficient communication and collaboration interfaces are necessary to allow participation of
field experts and decision makers.
A framework of engineering tools is needed to change the way design is done following the vision
detailed above. Elements of an engineering frame work for product design are:
Automated modeling and performance assessment
Engineering data management (engineering data is the keeper of design knowledge)
Knowledge capture and data mining
Multi-disciplinary reporting, trade-off and collaboration
Computational ideation, Plug and play
Optimal material layout technology with manufacturing considerations
Robust and reliability-based design
Mathematical modeling and computational simulation for multi-physics attributes
Virtual manufacturing validation

4. Conclusion
In a phase of product development where the freedom to make changes to a design (or design
requirements) is the largest ideation, creation and validation are performed on a single object. The
resulting concept is validated for performance and manufacturability. To be able to manufacture the
design on the shop floor detailed manufacturing drawings can then be made in CAD. Final validation
will only confirm the validity of the initial concept and therefore not require much design iteration or
even redesign.

However, as mentioned above the tools alone will not make a company create better designs. The
vision laid out is disruptive in a sense that the traditional process of designing is challenged and
change is required. This has proven difficult in many organizations. It changes the role of CAE in an
organization. In the past the person putting the design in CAD is called Designer. In the new process
the designer assumes a new role as that of the creator and that of the engineer formerly known as
analyst.
CAE has finally arrived to do the Engineering instead of just the design validation.