III. Mathematical modeling versus physical prototyping
https://www.youtube.com/watch?v=UPVm3v5ChR4
Assuming that model preparation is complete, metrics are then identified for measuring a product's performance. We must now further the development process by developing a model to test our concepts. At this stage, an issue when considering a product model is descriptiveness, also known as detail or fidelity. How much work should we as engineers put into constructing a model that reflects the product design problem? The product development team does not have infinite resources.
This decision comes down to the comparison of model descriptiveness versus model construction difficulty:
• Should we use a lumped mass model or a finite element model?
• Should we model an idealized steady-state condition, or use a transient model?
• Should we make detailed numerical analyses of all the product options?
• Should we construct a physical model or a virtual (mathematical) model?
Determining the level of a model is a trade-off decision between model descriptiveness versus model construction and solution determination time.
https://www.youtube.com/watch?v=Ee4CKIPkIik
Example
For a fingernail clipper product, consider a
scenario where we may ultimately derive a model that represents
finger force.
https://www.youtube.com/watch?v=edYLW5kqXBU
We must choose between an analytical or physical model (Table 13.2). Combinations of analytical and physical prototypes are also used. An example is to embed sensors in physical prototypes to provide feedback to computer simulations to enhance the accuracy of the simulations. The other continuum determines whether the prototype has some, most, or all the attributes of the final product (Table 13.3).
The other continuum determines whether the prototype has some, most, or all of the attributes of the final product.
O
ne
way to consider the decision from these continua is depicted in
Figure 13.3. As we initiate a project involving the construction of a
formal model, typically we will have a project resource depletion
constraint. We do not have infinite time or money to build a
"perfect" description of our product, infinite in detail.
Thus, we have a project management decision to make, namely, "How
detailed should we model?"
To answer this question, we have data from past history or design process prediction studies. We must use our engineering judgment to decide how complex the performance metric is in terms of the design variables and how much safety margin we desire using our resources.
In general, it is smart to make detailed engineering models when proto typing is expensive and when we have reasonable expectations in obtaining an accurate model. This idea is depicted in Table 13.4, adapted from Ulrich and Eppinger.
IV. Advanced topic: what is a product model?
As introduced above, we never analyze a real system but only an abstraction of it. The abstraction is a model it abstracts the real world, giving approximations of complex physical phenomena as part of physical systems.
http://vimeo.com/53848170