Long thin things fail by folding
Buckling in compression, and why it arrives without warning.
Long thin things fail by folding because pushing on something slender gives it a way to escape the load sideways, and once a long thin member starts to bow, that bow grows on its own until the member suddenly snaps sideways in a fold called buckling.
A bamboo cane pushed from the top
Pushing down on a long, thin bamboo garden cane standing upright on a table shows the failure plainly. For the first part of the push the cane stands there, seemingly rigid, giving no outward sign of how close it is to its limit. Then, suddenly and with no stretch of gradually increasing bend, it kicks sideways into a pronounced bow and folds, the whole failure taking well under a second once the critical buckling load is reached. A short, fat piece of the same bamboo under the same push simply crushes straight down, because there was never enough length for a sideways bow to develop before the material gave out. Cutting the long cane down bit by bit and repeating the push shows the change directly, from a sudden sideways fold at full length to a straightforward crush once it is short enough.
How a small bow runs away with itself
Any slender column carries some tiny imperfection, an almost invisible curve left from manufacture or a load slightly off its centre line, and that gives it a sideways nudge the instant it is loaded. The nudge bows the column slightly, and the bow makes matters worse, because the load now runs off the column's centre line and gains leverage on the bowed shape, pushing it to bow further. Below a certain load this feedback settles and the column stays gently bowed but stable. Above it the bow grows faster and faster until the column folds completely, at a load that can be a small fraction of what the same material would need to crush in a stockier shape.
That critical load falls with the square of the unsupported length. A column twice as long, with the same cross-section and the same end fixings, buckles at a quarter of the load, and one three times as long at a ninth.
Checking struts, and bracing them properly
Because buckling arrives suddenly, it gives almost none of the visible squashing that usually warns of a crushing failure. Compression members in a lightweight structure, such as a chassis strut or a suspension arm carrying a push, are therefore checked for buckling separately from the material's crushing strength, since a member sized only against crushing can look comfortably strong on paper and fold at a fraction of its calculated load. A designer used to tension members, where this danger does not exist, is most at risk of the mistake the first time a real compression member appears on a drawing.
Bracing a long compression member partway along its length is one of the most effective fixes, since the same square law works in the designer's favour: a brace at the midpoint halves the unsupported length and raises the buckling load about fourfold for very little added weight. The brace has to be stiff, though. One that can move sideways under load leaves the full length free to bow, and a brace bolted to a panel that itself flexes is a common way this goes wrong, the panel looking like a solid support on the drawing but providing little of the rigidity the calculation assumed.