nafems4/comments-nick.md

• strong background on in mathematics
• worth the time of an engineer need needs to be
• Though I like the prose I’m not sure I would say FE is romantic 😀
• Probably they would be be simple
• late XIX century — I would write this as “late 19th century”
• 5.2.3 — I think it would explain the fatigue and stress-life bit by noting that it is the relationship between stress (amplitude) and (cyclic) life -> stress-life.
• Young’s modulus
• 5.5.2 — the assumption is for linearity when computing seismic loading in the frequency domain rather than the time domain
• Figure 10 — frequency is misspelled
• 5.5.2.2 — “Each application should last two seconds.” I’m unclear what this refers to because the solution is in the frequency domain
• Should look to note that SRSS method means one loses the sign
• 5.5.3 — you use “weight” a lot when you mean to write weigh or weighs.
• The linear FEM problem leads of course of to a system of NG linear equations
• “It is called a sparse matrix because most of its elements are zero.” — is this not because there is only interaction locally to each element/node. Ie if node 1 is far away from node 10000 then node 1 will have zero effect on the displacement or stiffness at node 10000? (I’m looking to be educated here because I just don’t understand enough…)
• So the question is… how hard is it to solve a sparse linear problem?
• “Figure 16: Structure of the stiffness matrices for the same FEM problem with 10k nodes. Red (blue) are positive (negative) elements. a — 42k first-order elements, b — 15k second-order elements” — what’s a negative element in the context?
• Then it computes the stress tensor (atready already introduced in sec. 5.3.1)
• The link to my article on stress singularities is broken and is now here: https://nickjstevens.netlify.com/post/2019/stress-singularities/
• two seconds with one sign and the other two seconds with the opposite sing sign.
• Figure 24 — you present ax, ay and az, but also note SRSS in the caption. I was expecting to also see the SRSS fourth image in this figure - ie the combination of ax, ay and az
• “My second tip is: when faced to a new challenging problem, read, learn and ask for guidance to real people to see if you read and learned it right.” — this is so important we should all be screaming it from the rooftops - it’s so important to accept that continuous learning and growth is absolutely necessary for any engineer, so you need to be humble enough to ask for help and modest enough to challenge your own knowledge.
• 5.8.1 — folks might recognise the cumulative fatigue formula from college days as the “Miner’s rule” or Palmgren-Miner rule
• You use the word “automatise” where I would say “automate” instead.
• 5.8.1 — where you first talk about Git and scripting I completely agree with the message here and the value in that but think it would be useful to emphasise that this then becomes software that needs to be controlled, checked, approved, etc.
• 5.8.2 — CUF formula in second para has minor formatting where the “j” should be a subscript
• My guess is that they their underestimated they audience
• keep in mind that finite-elements are a meant to get an engineering solution, not and an end by themselves
• And finally, make sure that at the end of the journey from college theory to an actual engineering problem your conscience,is clear knowing that there exists a report with your signature on it.