Dave Baker Principal Engineer

Agile product engineering

Dave Baker, Principal Engineer at KD, discusses how to resolve conflicting demands of innovative design, rapid development and robust engineering.

As industry demands faster and cheaper routes to bring products of increasing complexity to market, the methods of the engineering design teams must adapt and evolve. Here, we consider some recent developments in the available tools and their place in the product development process.

Product development is typically approached in a staged system; establishing strategy and ideas before screening, detailed development, testing and finally commercialisation. The fundamentals of this staged process remain as relevant as ever, allowing the designer to become more knowledgeable at each stage and to make a better assessment of project risk. However, advanced tools for the visualisation of new ideas, the virtual proving of a design and rapid prototype verification allow engineers to drive down risk and accelerate progress at each stage. For example, can a functional 3D print prototype be used as a means of verifying a basic idea the same day as its conception, as a sketch on a piece of paper? By proving aspects of an idea at such an early stage, improved screening decisions can be made and development risk reduced.

By identifying and solving problems in the early stages of product development – so called ‘front-loading’ – it becomes possible to reduce development time and to become more ambitious in the degree of technical complexity within the product at an acceptable risk. The methods identified to achieve this are typically categorised as:

Project-to-project knowledge transfer: The influence of team experience and post-project learnings.

Rapid problem solving: The application of technology to reduce the cost and time of solving a problem.

This article considers tools for rapid problem solving, exploring some benefits but also examining their constraints. However, it is important to note that project post-mortem reviews, that carry learnings from one project to the next, can be just as effective as the most advanced virtual prototypes in early stage problem solving.

An apt definition of virtual prototyping for the design engineer is: a software based visualisation of a mechanical system under real-world conditions. Advances in computational power and the availability of previously specialist simulation software to the hands of the design engineer allows early stage feasibility assessment of a design. The use of virtual prototyping, such as thermal simulation and structural analysis, allows the design to be progressed without resorting to iterative physical prototype and testing.

One of the great attractions of virtual prototyping is the ability to perform detailed sensitivity analysis during development at reasonable cost. Repeated iterative design changes can be made and their impact on the system performance assessed, opening the possibility for a design to progress from ‘good-enough’ to truly optimal. The most striking example of this high-speed iterative process is the computational optimisation of structures. The example illustrated here, from Dassault Systèmes, illustrates a part created by a computer algorithm that removes any superfluous material to create a super-lightweight structure. Just over a decade ago, this process was the preserve of university research programmes; now it comes packaged for routine use, by non-specialist design engineers.

Computer Algorithm creating a lightweight part
Figure shows a computer algorithm creating a lightweight part (right) by removing all superfluous material from a user supplied starting point (left).

This poses the question of whether handing control to less specialist users makes the results of this process sufficiently robust to have a meaningful role in the product development process. Take the example of the optimised bracket with its lightweight slender elements. Does the designer, in this case a computer algorithm, consider the new possibility of buckling failures? The virtual prototype may predict a high strength, stiff structure; the reality could be one that folds and collapses in operation.

One response to the pitfalls of virtual prototyping is the manufacture of experimental prototypes using metal and plastic 3D printing methods. Continued advances in materials and processes mean the performance of printed prototypes can approach that of real parts, making them useful for proof-of-principal testing. However, their use as a means for final design verification, particularly where combinations of high strength and toughness are needed, remains an ambition rather than reality.

In summary, new tools available to the product development engineer can provide substantial benefits: accelerating time to market, reducing risk and opening the door to greater product capability. However, the fundamental process of robust product development remains, particularly in light of the present inability to fully replicate, either virtually or in low cost prototypes, the performance of many production assemblies in their intended environment. As technology continues to progress, it will be increasingly important to continue questioning these norms as greater possibilities emerge.

To view a digital copy of FOCUS and read more articles like this one, click here

Find out more?

To hear more about how agile product engineering can help resolve your business demands contact Ben Arlett our Head of Engineering