Designing Welded Joints by the Numbers

About this book

Every welded structure is only as good as its joints, and every joint was sized by someone, or not. On many drawings, the weld size is a habit: a 6 mm fillet all round, because that is what the shop usually welds. Sometimes the habit is generous, wasting weld metal, time and distortion; sometimes it is far too small for the load, and the joint cracks. The arithmetic that sizes a weld is not difficult, and knowing it lets a designer put weld metal where it does the work and nowhere else.

This book is for designers, drafters, engineers and fabricators who size and specify welds on machine frames, brackets, bases and light structures. It assumes basic statics and stress analysis in SI units. It explains how load passes through a welded joint, how butt and fillet welds are sized from their throat, how weld groups carry eccentric loads, twisting and bending, how repeated loads lead to fatigue, how weld size drives cost and distortion, and how to tell the welder clearly what to make.

Early tests as a starting point

In 1910, Herbert L. Whittemore of the University of Illinois tested oxy-acetylene welds in steel strip and found their strength, as a fraction of the plate’s, to be both lower and more variable than the enthusiasts of the new process claimed. Whittemore’s bulletin is one of the first careful measurements of weld joint efficiency, and Testing Welded Joints by the Numbers discusses it at length. The welding manuals of the following decade, including George W. Cravens’s treatise of 1921 and the Davis-Bournonville course of 1919, taught welders how to prepare joints so that the weld could reach the full thickness. Those early lessons, that a weld is only as strong as its fused area and that the shape of the joint decides that area, are where this book begins.

The running case

This is an illustration. A small engineering company designs the steel base frame, motor bracket and gearbox mounting for a conveyor drive, and the steel supports for a walkway beside it. Its designer wants to size every weld from the loads rather than habit, cut the weld metal where it is not needed, and make sure the joints that see vibration will not crack.

The company, its design, loads and results are invented for teaching. They are not KEVOS® work. Real structures and machines are designed by competent engineers to the applicable standards, such as AS 4100 for steel structures, with welding to AS/NZS 1554.1 or the standard the design calls up.

How to use the book

Chapter 1 traces the load through a joint. Chapter 2 sizes butt welds and Chapter 3 fillet welds, and Chapter 4 compares longitudinal and transverse fillets and deals with weld size limits. Chapters 5 and 6 cover weld groups under eccentric load and bending, and Chapter 7 combined stresses and the choice of filler strength. Chapter 8 covers fatigue, Chapter 9 weld volume, cost and distortion, and Chapter 10 the drawing and the design record. Every chapter ends with an exercise and its worked answer.

Limits and professional responsibility

This book is educational. The design strengths, factors and allowances in it are illustrations chosen to make the arithmetic clear, not values for design. Structures, lifting equipment, pressure equipment and machines whose failure could injure people must be designed and checked by competent engineers to the applicable standards and regulations, using the capacities, factors, fatigue categories and minimum sizes those documents give.

An original treatment

READOS introduces welded joints for fabricators in a collection chapter, covers weld testing in Testing Welded Joints by the Numbers, and weld metal volume in Arc Welding by the Numbers; this book links to them rather than repeating them. The explanations, the conveyor drive case, the calculations, figures and exercises are original KEVOS® writing. The historical sources are public-domain works in the United States; their findings are summarised in new words, and none of their text, tables or figures is reproduced.