Reality dawns

The development process for the BMW3 series involved five and half years, 2.6 million man-hours, 130 hand-made system level hardware prototypes and some 2,400 new components. Yet the overall development cost for new vehicles like these are projected to pl

The answer is virtual prototyping which is as relevant to all manufacturing industries as it is to the automotive industry.

Simulation-based design practices, all of which can be networked electronically, allow design engineers to quickly assess form, fit, function, and manufacturability of new products from concept to completion. The traditional stages of prototype building, instrumentation, testing, and modification become redundant. Instead, designers can construct accurate virtual prototypes in less than a week, and exercise the models through hundreds of tests with thousands of varying parameters.

Traditional CAD/CAM/CAE tools and processes have been embraced and implemented throughout major industries. For the most part they lived up to their promise of dramatically improving part design. In the automotive industry, for example, suppliers reported a 40% reduction in part defects over a recent five year period. This was accompanied by a corresponding drop in development and manufacturing costs. But during the same five-year period that the suppliers enjoyed, the vehicle manufacturers (OEMs) who were using these parts to assemble vehicles experienced only a 20% reduction in warranty costs.

This was a surprise to OEMs who expected to inherit the same savings. But optimal part design doesn’t necessarily lead to optimal system design. It is the interaction of form, fit, function and assembly of all parts that contributes to overall product quality. And so we may be reaching levels of diminishing return in applying CAD/CAM/CAE technologies to part design. The big opportunities now lie at design and streamlining of whole manufacturing systems.

Some software providers extended their software to address system level designs, but even simple extensions rendered the performance of many systems too slow.

Present, workable methodologies for system level design are broken down to include digital mock up tools to investigate product form and fit, functional virtual prototyping (VP) to assess function and performance, and virtual factory simulation to assess manufacturability of the product.

This combination of methods provides a means for realising an effective transition from hardware prototyping practices to software prototyping.

The functional virtual prototyping (VP) stage makes use of 3D component solids models and modal representations of component finite element models to accurately predict the performance of the product in virtual lab tests and virtual field tests. Within this VP stage its deployment typically involves five stages: build, test, validate, refine, and automate.

During the build phase, virtual prototypes are created of both the new product concept and any target products which may already exist in the market. Active models of the product concept are kept simple and are most often driven by desired response curves rather than specific product types. These should be driven by customer demand studies that identify desired performance. For example, in the initial design of a suspension system, the VP data curves embody the response of the chassis under varying conditions of speed, weight, turn radius and road conditions.

Successful virtual prototyping dictates that we need to also construct virtual test rigs that reproduce the test procedures and boundary conditions of the real fixture and machine. Here the initial response curves are put through their paces.

Test results are then validated. The companies with the best records in making effective use of VP have invested time and resources in building a validation library to see how simulation assumptions match test results, both physical and virtual. Once a comprehensive library is built, physical testing can be phased out and a greater reliance on virtual prototyping can be made.

Now that an understanding of the product’s robustness is available, the design itself can be refined. This understanding can help drive component selection and overall design parameters. This will be further refined as issues of noise, comfort, vibration and durability are addressed. A template-based design system that allows for quick and easy exchange of various subsystems is of paramount importance for effective design selection and refinement.

Using statistics-based Design of Experiment (DOE) methods, the entire universe of part selection combinations can be estimated to give a statistically relevant prediction of operating performance.

The first few virtual prototypes will be expensive but once engineering analysts have worked through some prototyping cycles (from build, through test, validate and refine), the virtual prototyping environment can be automated within the company through the use of templates, validation libraries and company standards.

One OEM that has made substantial progress in using virtual prototyping, to save time, money and enhance quality, is VW. VW relied on the use of a robust virtual prototyping process throughout the chassis and powertrain development groups for the Beetle. They made extensive use of system simulation software to evaluate thousands of design variations for vehicle ride and handling, durability, safety systems, as well as engine, clutch, and transmission performance.

Robert Ryan, president and chief operating officer of simulation specialist Mechanical Dynamics, writes for Design Engineering about the opportunities available to cut the costs of physical prototyping

Once a comprehensive library is built, physical testing can be phased out and a greater reliance on virtual prototyping can be made. A template-based design system that allows for quick and easy exchange of various subsystems is of paramount importance for effective design selection and refinement. This of course may place added pressure on suppliers to create parts and systems to pseudo standard sizes and performance, depending on the OEMs demands.

A typical parameter study with virtual prototype uses data curves to embody the response of the chassis under varying conditions of speed, weight, turn radius, and road conditions

Mechanical Dynamics Tel: +44 (0) 1926 420 230