SWBPIPE

Pipe stress analysis you can see all the way through.

SWBPIPE · free and open source

A program for piping flexibility and stress analysis, with an agent built in that works in the model beside you. Learn the method, try the alternative, and understand the answer before you hand it on.

macOS download coming soon Read the open manual

MIT licence · macOS · signs in with ChatGPT

Chapter I

Ask it why.

The nozzle load at 130 looks high. You say so. The agent reads the model and the last run, and finds what a senior engineer would: the rigid rod at 80 is holding the header down against the riser’s thermal rise.

It proposes a variable spring in its place, shows its reasoning and what it still does not know, and waits. You accept the change a row at a time and run again.

The agent is there to widen what you can see, not to see for you.

Isometric sketch of the Loop 4 header, from pump P-401 to vessel V-402 with a branch to the header tie-in. A proposed variable spring at node 80 replaces a rigid rod, and the nozzle load at node 130 falls from 7.4 to 3.9 kilonewtons. 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 PUMP P-401 DISCHARGE VESSEL V-402 NOZZLE HEADER TIE-IN SPRING H1 REST, GAP 3 GUIDE PROPOSED SPRING AT 80 REPLACES RIGID ROD OPE1 LOAD: 7.4 kN → 3.9 kN LOOP 4 HEADER NOZZLE LOAD AT 130 IS HIGH IN OPE1: (1) THE RIGID ROD AT 80 HOLDS THE HEADER DOWN AGAINST THE RISER'S THERMAL RISE, ABOUT 6 MM AT T1. (2) A SPRING SET AT THE W LOAD CARRIES THE WEIGHT AND LETS THE HEADER RISE. (3) TO BE SETTLED: TRAVEL AT 80 NEEDS A RUN WITH THE SPRING IN PLACE.
Fig. 1.—Loop 4 header. The agent’s reasoning for its proposal, set as the note on the figure. An illustrative model; the values are not engine output.
The SWBPIPE window. A table of hangers on the left; on the right the agent's proposal to replace the rigid support at node 80 with a variable spring, with its rationale, the constraints it considered, and two open questions.
Plate I. The proposal as it arrives: each changed row, the rationale, the questions still open. Interface in development.

Chapter II

Nothing is hidden.

Every part of the answer can be opened and read, by you or by the agent on your behalf.

2.1The model is a table.

One row per node, the way stress engineers have always written a system down. The drawing is made from the rows, so what you see is what you entered.

2.2The mechanics are open.

A 3D frame solver with bends, nonlinear supports and load-case algebra. Read any line of it. Sixty-four benchmark cases are published with the hand calculations they are checked against.

2.3The agent can do what you can.

Build a system from a line list, place supports, run a case, trace a result back to the element and the equation behind it. Nothing it does enters the model until you accept it.

The SWBPIPE window in its Both view: the layout table with one row per node on the left, the drawing made from it in the middle, and the inspector for node 20 on the right.
Plate II. The layout table and the drawing made from it. Interface in development.

Chapter III

Your code, your data.

SWBPIPE ships with no code tables.

Allowables, stress intensification factors, load combinations and component data come from the standards and catalogues you already have the right to use, held in a rule pack you build. A guide and a template show you how.

The Chirality workflows built into SWBPIPE let the agent help you turn your own spreadsheets and data sheets into rule packs and libraries, and keep them current as your sources change. What you build stays yours, as files on your machine.

Chapter IV

Design on every desk. Validate on one.

What SWBPIPE is not

SWBPIPE does not judge compliance with any code, and it does not vouch for the accuracy of its results. It is for learning, designing and exploring alternatives. Do not use it as the basis for professional work.

When a design needs validation, export it and validate it in the solver your industry accepts. In the first release that means a .mbf model batch file for CAEPIPE. More formats will follow.

In SWBPIPE

Route, support, iterate and learn, as many people as need it, at no cost.

In your accepted solver

One seat validates the design your team settled on, under your engineer’s judgement.

The name

Named for Shipman, Watts and Burrows.

In 1941 the M. W. Kellogg Company published a general method for the stresses, reactions and deflections of a piping system in space.

Its acknowledgment credits W. H. Shipman, who first solved the single-plane problem systematically, and G. W. Watts and W. R. Burrows of Standard Oil (Indiana), who extended his method to three planes.

“The most significant step towards the general solution.” Design of Piping Systems, 1941, on Watts and Burrows

For decades engineers learned that method by hand, and the ones who knew it best went on teaching it long after they retired. The standards the industry relies on today were built by people like them, one careful detail at a time. SWBPIPE is made to keep that knowledge in the open: the method, the mechanics and every detail within reach of anyone willing to learn it.

Read the open manual →