Originally Posted by
Bush65
Low carbon steel (AKA mild steel) has a definite yield point, high strength alloy steels do not.
I don't have pics with me, but will attempt an explanation of what happens in a tensile test, starting with low carbon steel.
As tensile load increases, the elongation (stretch) increases proportionaly until the proportional limit is reached - the slope of this straight line corresponds with the elastic modulus (E = stress/strain where stress = load/area and strain = elongation/original length so E = (L x l)/(A x e) ) curve. Near limit of proportionality the elastic limit is reached - where original length will be achieved if load is released. Just after the elongation of the test piece will start to increase with no increase in load - this is yielding and the curve goes horizontal. As further elongation occurs the load starts to drop (because cross sectional area has become smaller - waisting) and the curve dips down. The material begins to work harden and the load starts to rise again to increase elongation. Load starts to fall again because waisting becomes severe. During the last 2 stages the curve resembles an arch. Finally the test piece breaks.
With high strength alloy steel, the curve starts the same and the slope until the limit of proportionality is reached is the same because with steels E is much the same (small differences do occur but are usually neglected except for stainless) regardless of alloy or heat treatment - the value of load at proportional or elastic limit is higher but slope is same. Then as waisting occurs the curve has a similar arch shape until the break occurs. There is no yield point where elongation increases for no increase in load.
Because yield strength changes (material property) we commonly base design strength = 0.nn x yield strength. We avoid yield because it will result in permanent deformation (also formation of a plastic hinge which can turn a ridgid structure into a mechanism (resulting in collapse).
As the high strength steels have no yield point, we create an approximate value on the load/elongation curve by drawing a line parallel to the linear slope, but offset by 0.2% elongation. Where the 0.2% offset line cuts the curve from the tensile test, we have a value to use with our much loved equation (design strength = 0.nn x 0.2% offset strength).
For bolted joints in steel structures, it has become normal and more economical to use high strength structural bolts these are different to common high tensile bolts which are called hexagon precision bolts in the relavent Australian Standard (AS1110 from memory). HSS bolts are designed to be tensioned to the prescribed proof load (near as damb the yield or more correctly 0.2% offset) and the bolt proportions (particularly the head) were developed after much research by the international committee - the Aus Standard is identical to the ISO standard.
The Steel Structures Code (AS4100) and it's commentry have a section on tightening bolts. For bolted joints designed as fully tensioned (tension to the proof load) one of the acceptable tightening methods is part turn (snug tighten then further angle as given in table for bolt length). The other acceptable methods involve tension measurement and torque control (e.g. tension wrench) is generally not permitted.
It is well documented that the part turn method often results in tension exceeding the proof load, but it is not an issue. The standard allows bolts that have been fully tensioned to be re-used once, but only if they are used in the same bolt hole as they were removed from - because permanent deformation of threads, etc. can prevent proper tension being achieved if used in a different location.
With mechanical equipment, where parts are required to be pulled apart and re-assembled more than once, the bolted joints are usually designed for bolts tightened to approx 65% proof load and tables give tightening torque for achieving this tension.
When bolts are tightened to the proof load, the friction forces become so large that torque control results in errors of resulting bolt tension something like 25% - this error has been found in many tests making it unacceptable.
To avoid fatigue failure of bolts subjected to cyclic loads, the best practice is to use bolts (quantity and diameter) so that when tightened their pre-tension is at least 2 times (up to 5 times) the external applied load (in the bolt). Together with this tension, the joint must be designed so that it is considerably stiffer than the bolt. Then during pre-tensioning the bolt elongation is much more than the elongation of the joint. During the load cycles the variation in the tension in the bolt will be small and the variation in the compression of the joint will be large (in proportion to the relative stiffeness). The sum of change in tension of bolt and change in compression of joint equals the external applied load. So fluctuation in the bolt tension (the most important factor for fatigue strength) is much less than the fluctuation in applied load.
Note: it is not a requirement to tighten bolts to yield to achieve ar pre-tension 2 to 5 times the external applied load (we normally use 65% proof load, but where it is not possible to use more bolts or larger dia bolts then we use greater pre-tension, but achieving the pre-tension becomes more difficult.
