Podcast #128: We chat with Kent C Dodds about why he loves React and discuss what life was like in the dark days before Git. Listen now.
37

Beware of overfitting. A more accurate model of gathered data from a system may not be a better predictor of future behavior of a system. The above image shows two models of some data. The linear line is somewhat accurate on the training data (the points on the graph), and (one would expect) it will be somewhat accurate on the testing data (where the ...


26

The most obvious downside is cost, all engineering projects have a finite budget and spending more money than you need to is clearly a bad thing not to mention wasting time. There can also be more subtle issues. Things like FE analysis are always approximations and sometimes adding unnecessary detail can introduce artefacts and make it more difficult to ...


23

I have written mostly about CFD in this answer, however same points should also work for FEA or other simulation techniques. CFD is mostly used for design optimization and parametric study of the design. Following are a few examples showing how engineers use simulations Selection of a design: Read: A conceptual study of airfoil performance enhancement ...


13

There are a few reasons. From a purely pragmatic perspective, it's due to time constraints. The requisite time to solve a model increases far, far faster than the level of precision, and whichever level is adopted is subjective, anyway. This is also affected by the fact that excessive accuracy is mostly useless. After all, your model might be 99.999% ...


12

To summarize the other answers: An engineer needs to know qualitatively how the simulation will go, but he still needs to run the simulation to get the quantitative answer. Also, simulation allows the engineer to vary the parameters slightly (Monte Carlo simulation) in order to evaluate the stability or margin of error of the solution. This is frequently ...


11

The turbulence model can make a big difference in your simulation. There are many turbulence models around. It becomes a tough job to select one out of them. There is no perfect turbulence model. It all depends on several parameters like Reynold's number, whether the flow is separated, pressure gradients, boundary layer thikness and so on. In this answer, ...


8

Before I start, let me disclose that I don't know a lot about the details of infrastructure projects or your file format, but I think the general considerations are fairly universal. There are a number of considerations when providing your digital model to a contractor, and in my work I have decided to offer them my model in some situations, but not in ...


8

Engineers should have a general idea of the expected result (Balpark values, expected behaviour)when using a complicated computer model. Most of the time these conclusions are based on a (much) simpler model, that preferably can be checked by hand. Biggest reason for this is to eliminate the possibility of human error in constructing the model itself. Using ...


8

There are a number of techniques for meshing complex domains for Finite Element Analysis. They generally fall into two categories: Structured vs. Unstructured. For structured meshes, basically the entire mesh can be mapped directly to a 3D array of XYZ coordinates, whereas unstructured grids cannot. There is a good description of the classifications with ...


7

Yes, getting a second opinion can be useful. This is done routinely in weather forecasting where exact solutions are unknown, and there is some judgement about how to apply various factors. There will be less wiggle room in something like a finite element mesh stress analysis because the iterative equations for solving it will be basically the same no ...


7

As my Fluids lecturer said many years ago, “if the mathematics doesn't agree with reality the mathematics is wrong”. You can easily substitute the words model, theory or simulation for the word mathematics.. Engineers who use simulations should have a very good idea of what to expect for a solution, not necessarily know what the answer will be for a ...


6

I write this from the perspective of an engineer who develops simulation software. I think the practice described is bad, and I recommend you do not use two different softwares to "confirm" the results. In general, two different modeling softwares can not be used to confirm much anything other than their similarity. Two softwares could easily both get two ...


6

Splitting a uniform load into separate pieces that are still continuous will have no effect. This is frequently done. As far as your question about bending in HE and EB, there shouldn't be any bending because all of the forces are balanced. A sum of the moments at H or E will show that the moments from the beams on either side are opposite and equal. That ...


6

The most common way to model this structure is as follows (ignore the fact that the proportions are a bit off): So, all of the diagonals are pinned-pinned. You'll notice in the schematic, however, that the chords (including the diagonals from the supports to the top chord) are not segmented at every point of intersection with the diagonals. Indeed, the ...


6

An extremely accurate model may require a prohibitive amount of input data. It might be possible to generate an excellent model of weather systems, for example, by taking as input the position and velocity of every gas molecule in the atmosphere. In practice, such a model would not be useful, since there's no realistic way to generate the proper input. A ...


5

In my particular field (buried culvert design), we run finite element analyses constantly. We almost never change a design based on the results; we know going in (from a variety of factors, mostly prior experience and conservative assumptions) whether or not the design is good. We run the analyses to demonstrate to others that our design is good. We may ...


5

While the other guys explained the theoretical framework behind meshing, the practice is markedly different and it is not at all automatic in industries where quality of mesh is of utmost importance given that finite element analysis results cover a great deal of the product development process. Let's first understand how meshing is done: Meshing for ...


5

This may be only a partial answer since I don't have any idea what to do for the Young's modulus of the fluid. Consider the paper Dynamic Pressures on Accelerated Fluid Containers by G. W. Housner. This is referenced by ASCE 4-98 Seismic Analysis of Safety-Related Nuclear Structures, Commentary Section C3.1, for analysis of hydrodynamic loads on tanks. In ...


5

Kinematics and dynamics Those are the steps to solve problems of this nature. Analize the kinematics of the system. $\hspace{5.em}$ $_{o}\vec{r}_{OP}$ = $_{o}\vec{r}_{OR}$ + $_{o}\vec{r}_{RP}$ $\hspace{5.em}$ $_{o}\vec{r}_{OP}$ = $_{o}\vec{r}_{OR}$ + $R(\varphi) _{B}\vec{r}_{RP}$ $\hspace{5.em}$ $_{o}\vec{r}_{OP}$ = $\big(x_{k}î + 0j + 0k \big)$ + $\big(...


5

You can't calculate friction coefficients this way. At the atomic-scale level, the friction coefficient depends on the interactions of the two materials across the interface. If you measure the behaviour of the interface between materials A and B, and A and C, you can't say anything about the interface between B and C. In any case, friction coefficients are ...


5

"Too accurate" is not monotonic. It can actually create an illusion of fidelity which makes you think it's worth pumping more money into the simulation. This becomes very important when you're presenting data from mixed-fidelity models, where some parts are very detailed and other parts are very coarse. A real life example I had involved sampling ...


4

I think this is a good practice overall. By using two different softwares, you may be able two avoid two kind of errors: 1) errors that come from an inaccurate software (which should not be overlooked), 2) errors that come from the lack of habit of the user with the software (hidden options, default settings...). If the softwares are different enough, the ...


4

My guess is that you probably need another differential equation for the angular movement, that will involve the inertia, such as: $$m_G l^2 \ddot{\varphi} = m_G g l \sin(\varphi)$$ which yields: $$ \ddot{\varphi} = \frac{g}{l} \sin(\varphi)$$ You can then maybe use the small angles approximation: $$ \sin(\varphi) \simeq \varphi$$ Check out the ...


4

You can write two ODEs describing the relationship, and that would be in the time-domain. Alternatively you could describe things in the frequency domain and get the relationship as a transfer function. I will show the latter. Using Kirchhoff's current law at the two nodes K and A, we get: $$ \frac{U_E-U_K}{R}=\frac{U_K-U_A}{L s}+C_1 s U_K$$ $$\frac{U_K-...


4

Real-world engineering example To my knowledge there are no simple analytical methods of determining whether an object such as a sword can cut cleanly through any arbitrary object. In manufacturing engineering we would rather cut many small pieces of material than take one big cut (i.e. we generally use saws to cut material instead of giant guillotines). ...


3

Adding just a slight notion: Your experience would always expect the beam HE to buckle, because it is a Eulerian buckling beam (don't know if it's also called like this in English). The numeric only sees an evenly distributed load, so there will be no buckling... Try to alter your forces to the left and right of H a bit and you will get a moment which will ...


3

As modal analysis is performed in the frequency domain, your modes and amplitudes are also in the frequency domain.The modal amplitudes ($q_i$) are scaled by the same amount as the applied load. If you want to know the displacement at each mode, then just multiply $q_i$ by whatever the magnitude of the load (for the mode) is scaled by. The maximum amplitude ...


3

I would expect the modeling as a single load to be accurate. Force per linear area is the same expressed either way. You could look at a linear load on a single beam and just add more points of integration analytically and try it in ANSYS to see it. The HE and BE segments will undergo buckling as its deformation mechanism after modest compression. The ...


3

I design electric motors and I use electromagnetic FEA as part of that design process. Motor designers have a lot of good analytic techniques that get us very close to the actual performance of the motors for certain key parameters (torque, current draw, speed, etc.). However, this requires that we make certain assumptions that may or may not be valid. ...


3

Section 2.1.14 of the Abaqus Example Problems Manual is 'Water sloshing in a baffled tank', which might be useful (even if you don't use Abaqus at your own institution, the manuals are hosted by many institutions that allow public access). The Abaqus water model uses an equation of state (EOS). If you download the input file from the manual, you can cut ...


Only top voted, non community-wiki answers of a minimum length are eligible