Article written in June 1998 with S. Morino and F. Martinelli for 31st ISATA
Abstract
Today, market competition forces the optimization and reengineering of even traditional, consolidated mechanical components. The paper describes the experience we performed at the F.lli Ferrari Gru – a small company leader in the truck loader cranes – in the redesign of the cast steel-slewing group. We started with a traditional 2D drawing ready for production, generated by a simple 2D CAD system. Using a PC based CAD system, we generated a complete 3D parametric model of the crane’s rotating unit. We discussed such a model with the designers to optimize the component from structural, manufacturing, and cost perspectives. Using the new generation parametric and features-based CAD system, we have been able to quickly apply the changes to the three-dimensional model (i.e., to reorganize the design process and re-parametrize the basic components) and to verify and quantify the corresponding advantages obtained. This experience demonstrates that this approach, widely adopted by large mechanical companies, can be used successfully even in a small company.
Introduction
The recent evolution of the CAD market is proposing to the attention of technical managers and designers a set of new software tools: the mid-range CAD products. These new systems run on low-cost personal computers, based on the Microsoft Windows operating system, and powered by Intel or Digital Alpha processors. They offer functionality on par with the high-end area, at a price usually under 5,000 USD. As a result of high sales volumes and low selling prices, software quality is usually very high. In general, these systems are easy to install, to learn, and to use, user-friendly, well-integrated, and compliant with the rules and “traditions” of Microsoft Windows. Furthermore, they are based on robust, well-tested geometric engines, such as ACIS from Spatial Technologies or Parasolid from EDS. Obviously, they aren’t as powerful or as rich in functionality as the systems in the high-level area; despite these limitations, they provide a valuable solution for many application contexts. In particular, these systems can constitute a real opportunity for many small and medium-sized companies to adopt new and established CAD technologies, which were previously inaccessible due to cost and complexity.
On the other hand, large companies are only partially interested in these new systems; many have used the latest CAD technologies for a long time. Furthermore, large companies need a broader set of functionality and require a different type of service and interaction with the CAD supplier, characteristics that mid-range systems don’t provide.
Generally, small companies don’t have the financial resources required to acquire and manage a high-end CAD workstation and the corresponding software modules. In particular, companies from the mechanical field also lack the human resources for the long training typical of high-end systems. Moreover, these companies can’t afford the transition costs of integrating a complex CAD system into their design and production processes.
Starting from these observations, we decided to directly investigate the use of a mid-range CAD system in the design and production processes of a small company to verify the advantages and costs of such a choice. The target was not limited to testing the use of new geometric modeling capabilities in such a technical environment. Our aims were also to verify the capabilities of the CAD system in using and interacting with legacy data, in cooperating with traditional design tools and processes, in supporting a preliminary structural analysis, and in fully managing the parametric changes to the geometry as suggested by experimental and simulation tests.
The study case
To evaluate a mid-range CAD system for the design and production processes of a medium or small company, we first needed to identify and select a mechanical company in our geographic area. Then, a significant component among those designed and produced by the company should be chosen. We selected the F.lli Ferrari, a specialized small company, leader in the design and production of truck loader cranes. For some years, all design processes and all technical communications with subordinate suppliers have been based on technical drawings, on paper, produced by a well-known brand CAD system. Therefore, all parts libraries and the entire drawings archive are constituted by 2D CAD drawings. The company uses 2D representations even for more complex parts; this requires drawings with many cross-sections and views.
When we started the project, the technical management at the F.lli Ferrari was already investigating the mid-range products as a concrete alternative to their current non-parametric 2D CAD system. To evaluate the use of such systems to a significant extent, with designers and engineers at F.lli Ferrari, we analyze and select two components of the crane: the crane’s rotating unit and the corresponding support.
The support of a new crane’s slewing group is a component usually designed with reference to previous cast iron basements from F.lli Ferrari’s experience. For example, the starting point for designing the model-728 crane basement was the cast iron support of the model-722 crane. Because the shape of this component is very peculiar, for the designers it is rather difficult to employ some of the consolidated computational structural methods.
Typically, to define the basement component for a new crane model, the designer adjusts the dimensions of the old design by scaling them up or down as required to meet the new loads and performance requirements. The result is a preliminary prototype design that the design group is discussing, enabling the identification of the best solution from a structural, manufacturing, and economic point of view. This discussion and the calculations taken into account lead to the design drawings for the basement; usually, after a period of three to four months, it is possible to test the prototype unit that has been manufactured during this time.
However, the F.lli Ferrari’s prototype design procedure requires simulating the overall crane’s life. This, obviously, makes the designer sure that his assumptions and calculations are correct and ensures the safety of the basement. The cast iron support of the slewing group plays a strategic role in the birth of a new crane project, both for saving money and for achieving tighter fabrication schedules.
The design group cannot afford any safety risk with this component; consequently, the design’s strength is often achieved at the expense of weight, which means higher costs.
The shape of the basement is so complicated that the weight penalty may be significant, and cannot be easily reduced due to problems in using standard formulas in the science of construction. It may happen that a lot of cast iron material is used in areas that aren’t very stressed, and these situations can’t be easily identified.
The modeling and analysis activities
First, we model the crane slewing group geometry using a modeling system. Based on our previous experience and simple availability criteria, we chose to use SolidWorks, from SolidWorks Corp., as the reference CAD system. This system is generally considered one of the best representatives of the mid-range CAD system category.
Using the parametric and feature-based approach, typical of the mid-range category, we were able to quickly define a fully parametrized model of the rotating unit and the support. Due to the complex shape of the support and some limitations of the CAD system, we need to explore alternative construction procedures before we can complete the geometric model. In this phase, we get the main part of the data from some preliminary drawings and directly from the F.lli Ferrari design team.
Once the CAD model was available, we discussed it with the design team, collected and applied improvement suggestions, and consolidated it. Using an FEA (Finite Element Analysis) module integrated into SolidWorks, we obtained a map of stress values.
The redesign activities
After the modeling activity was completed, we evaluated the possibility of performing a deep redesign of the selected parts. The main target was to redesign the support, using as input data the actual constraints and loads (both already verified on the crane prototype unit loaded to its maximum lifting capacity).
The first aim was to assess the thickness of the basement body. In fact, if it were possible to reduce the wall thickness, even by one millimeter, a 5% reduction in weight and production costs could be achieved. The analysis of the 3D model led us to conclude that reducing the wall thickness was possible, but not across the entire body uniformly. We obtained some advantages, especially in the upper and lower parts of the cast iron basement, so that we could save some eight Kilos of material.
Then we have paid attention to the stiffening ribs. The internal one, connecting the lower bearing support to the forward pin of the basement, was over-dimensioned, so we could reduce its thickness, resulting in a less encumbering shape. This solution not only saves material but also enables a simpler (and thus safer) casting procedure by minimizing the length of the stiffening rib joining the forward fairing wall.
These two steps were a significant upgrade of the existing cast-iron support.
The following, and really innovative step, was to modify the shape of the fairing wall, generating a structural frame that connects the bearing supports (the upper and lower ones) to the forward pin (obviously, again, to save both weight and money).
This was obtained from the analysis of inertial data from different sections of the fairing wall. Starting from a continuous wall, we took away material all around the neutral fibers, optimizing shapes and profiles. We could generate a kind of fairing structural frame, easy to manufacture using casting technology, 10 Kilos lighter, and with a good, innovative look.
The performance of this new solution has been calculated by applying maximum loads across twelve different angular configurations of the crane. For each of them, the 3D parametric model showed the stress distribution in the basement. This work required many iterative changes and calculations to obtain the maximum economic savings for the production runs. In this phase, the integration between parametric modeling and finite element analysis modules allows us to quickly implement our optimization hypothesis as changes to the geometry and directly obtain updated results from the analysis model.
Of course, safety considerations have always had the maximum priority. The computer-aided modeling and finite elements stress analysis allowed us to reach these promising results in just a few days. This new solution, obtained by modifying the original design, demonstrates how the 3D-stress analysis approach, if well integrated in a powerful CAD system, makes it possible to reengineer even traditional and consolidated mechanical components.
Conclusions
The new mid-range CAD system, with its rich set of integrated engineering tools, can really enhance the design and production processes of small and medium companies. They constitute a real opportunity for these companies to adopt the latest CAD technologies and begin the long, expensive integration and reengineering of their internal processes.
The described experience demonstrates that, in medium- or small-sized companies, these new systems can successfully and gradually replace more traditional 2D CAD systems and provide valuable support for all design and engineering activities.
Acknowledgements
We wish to thank the designers, the production engineers, and the managers at F.lli Ferrari involved in this project for their collaboration and support.
References
- U. Cugini, G. Bocchi, F. Folini e M. Galloni. Rapid prototyping and parametric CAD systems. In: Proceedings of 27th ISATA International Symposium on Automotive Technology and Automation, Dedicated Conference on Rapid Prototyping for the Automotive Industries and Laser Applications for the Transportation Industries, (Aachen Germany, October 31–November 4 1994), pp. 51–57, ISATA, Automotive Automation Limited, Croydon UK, 1994.
- R. Wysack. Designing parts with SolidWorks. CAD/CAM Publishing, San Diego, CA, USA, 1997.[3] I.Bozza and F.Folini. Proposal for a meta-model of product-development processes.To appear on: 2nd International Conference on Planned Maintenance, Reliability and Quality, (Oxford, England, 2nd-3rd April 1998).
