Yamagishi Complexity Torus LabYamagishi Lab

[Paper 617] If a Valve Closes Faster Than a Round Trip Along the Pipe, the Water-Hammer Pressure Does Not Depend on How It Closes

Abstract

When water flowing in a pipe is suddenly stopped by a valve, the stopped water is compressed, and a pressure wave forms and travels back and forth along the pipe (water hammer). The size of the pressure rise, and how it changes with the time taken to close the valve, are obtained by two routes, formula and numerical computation. No new theorem or law is claimed.

First, the pressure rise is the product of density, wave speed and the velocity stopped——stopping $1$ m/s instantly raises the pressure by $\mathbf{1.341653}$ MPa in a steel pipe (bore $100$ mm, wall $5$ mm), $0.374397$ MPa in a PVC pipe and $0.198028$ MPa in a polyethylene pipe. Steel and PVC differ by $\mathbf{3.583508}$ times (Section 2).

Second, the wave speed is not the speed of sound in water——water with a bulk modulus of $2.2$ GPa has a sound speed of $1484.576418$ m/s, but the wave is slowed as much as the pipe wall expands under pressure: $1344.072336$ m/s in steel, $375.071665$ m/s in PVC and $198.384869$ m/s in polyethylene (Section 2).

Third, and this is the core. If the valve closes in less than the round-trip time of the pipe, the pressure does not depend on how it closes——in a $500$-m steel pipe the wave takes $0.744008$ s to go from the valve to the reservoir and back. Computing the closure by the method of characteristics, closing instantly, in half that time ($0.372004$ s) or in exactly the round-trip time all gave the same head rise, $\mathbf{137.057235}$ m (Section 3).

Fourth, closing more slowly than the round trip lowers the pressure in inverse proportion to the closing time——closing in twice the round-trip time gave $68.528618$ m, in $5$ s $20.394324$ m and in $10$ s $10.197162$ m, each matching $2Lv/(gT)$ within $10^{-6}$. The pressure-lowering wave returning from the reservoir reaches the valve before it has finished closing (Section 3).

Fifth, the separator is whether the closing time or the round-trip time is longer——if shorter, the valve cannot be told from an instant closure; if longer, the way it closes sets the pressure directly. The longer the pipe and the slower the wave, the longer the round trip (Section 4).

Keywords: water hammer, Joukowsky equation, pressure wave, pipeline, valve closure time.

Scope (what is not claimed)

No new theorem or law is claimed——the formula for the water-hammer pressure (Joukowsky, $1900$), the formula for the slowing of the wave by the elasticity of the pipe wall (Korteweg, $1878$), and computation by the method of characteristics (in the textbook by Wylie and Streeter and elsewhere) are all known. Friction and cavitation are not treated——pipe friction, the formation of bubbles when the pressure drops, and the detailed relation between valve opening and flow are left out; only a closure that reduces the velocity in proportion to time is considered. The properties of water are rounded——the bulk modulus is set at $2.2$ GPa and the density at $998.2$ kg/m$^3$. No design is recommended——neither the choice of pipes nor guidelines for closing times are given. Relation to earlier papers: Paper 474 showed that the speed of sound changes with air temperature, so distant sounds carry on warm-sky nights——here the speed of the pressure wave in water changes not with the water but with the softness of the pipe wall. What is added is setting the wave speed and pressure rise side by side for four pipes, matching a method-of-characteristics computation with the formulas and showing in numbers that the behaviour switches at the round-trip time, and placing the separator on the closing time versus the round-trip time.

1. Notation and the Question

Setting: water flows at velocity $v$ from an upstream reservoir through a pipe of length $L$ to a valve at the downstream end. Let $\rho$ be the density of water and $c$ the speed at which a pressure wave travels in the pipe. Stopping the flow at the valve raises the pressure of the stopped water by

$$\Delta p=\rho\,c\,\Delta v$$

(a head of $c \Delta v/g$). The wave speed follows from the bulk modulus $K$ of water and the bore $D$, wall thickness $e$ and Young’s modulus $E$ of the pipe:

$$c=\frac{\sqrt{K/\rho}}{\sqrt{1+KD/(Ee)}} .$$

Question: how much does the pressure rise, and how does it change if the valve closes slowly?

2. Wave Speed and Pressure

Computation 2.1: with a bulk modulus of $2.2$ GPa and a density of $998.2$ kg/m$^3$, the speed of sound in water is $1484.576418$ m/s. For pipes of bore $100$ mm and wall $5$ mm, the wave speed and the pressure rise from stopping $1$ m/s instantly were found.

pipeYoung’s moduluswave speedpressure risehead
steel$200$ GPa$1344.072336$ m/s$\mathbf{1.341653}$ MPa$137.057235$ m
ductile iron$170$ GPa$1323.183668$ m/s$1.320802$ MPa$134.927184$ m
PVC$3$ GPa$375.071665$ m/s$0.374397$ MPa$38.246666$ m
polyethylene$0.8$ GPa$198.384869$ m/s$0.198028$ MPa$20.229627$ m

Observation 2.2 the wave speed is set not by the water but by the pipe wall. In steel it is about nine tenths of the sound speed in water, in PVC about a quarter and in polyethylene about a seventh. As the pressure rises the wall expands, and instead of compressing the water the wave pushes the wall outwards——since the pressure rise is proportional to the wave speed, steel and PVC differ by $\mathbf{3.583508}$ times.

3. The Closing Time

Computation 3.1: for a $500$-m steel pipe (wave speed $1344.072336$ m/s), a velocity of $1$ m/s and no friction, reducing the velocity at the valve to $0$ in proportion to time was computed by the method of characteristics (dividing the pipe into $50$ sections), and the maximum head rise at the valve was found. The time for a wave to go from the valve to the reservoir and back is $2L/c=0.744008$ s.

closing timehead rise (computed)formula
$0$ (instant)$\mathbf{137.057235}$ m$c v/g=137.057235$
$0.372004$ s$137.057235$ m$c v/g$
$0.744008$ s$137.057235$ m$c v/g$
$1.488015$ s$68.528618$ m$2Lv/(gT)=68.528618$
$5$ s$20.394324$ m$2Lv/(gT)=20.394324$
$10$ s$10.197162$ m$2Lv/(gT)=10.197162$

Instant closure, half the round trip and exactly the round trip matched Joukowsky’s value, and the three longer closures matched $2Lv/(gT)$, all within $10^{-6}$.

Observation 3.2 this is the core: closing faster than the round trip, the pressure does not depend on how the valve closes. From the start of closing until a round trip has passed, no word from the reservoir reaches the valve. If closure ends within that time, the valve sees it as instantaneous——so $0.372004$ s and $0.744008$ s give the same $137.057235$ m as an instant closure.

Observation 3.3 closing more slowly than the round trip, the pressure is inversely proportional to the closing time. The wave reflected at the reservoir returns in the pressure-lowering sense. If it arrives before closure ends, it starts cancelling what the continuing closure adds——the maximum rise is $2Lv/(gT)$; a $5$-s closure holds it to $20.394324$ m, about $0.148802$ times the instant value.

4. The Separator

Observation 4.1 the separator is whether the closing time or the round-trip time is longer. If shorter, any closure gives the same pressure as an instant one. If longer, the closing time directly sets the pressure——the round-trip time is twice the pipe length divided by the wave speed, so it is longer for longer and softer pipes. Softer pipes also have a smaller pressure rise to begin with.

Observation 4.2 beside Paper 474. Paper 474 showed that because the speed of sound changes with air temperature, distant sounds bend back to the ground on nights when the air above is warm——the speed of the pressure wave here changes not only with the properties of the water but with the softness of the pipe wall around it. Both what carries the wave and what surrounds it set the speed.

Observation 4.3 what is not said. Pipe friction, bubbles formed when pressure drops, and the relation between valve opening and flow are not included. The bulk modulus of water is rounded to $2.2$ GPa. No choice of piping is recommended——what is shown is that the pressure formulas match a numerical computation, and that the behaviour switches at the round-trip time.

5. Roll Call and Closing

  1. stopping $1$ m/s in steel gives $\mathbf{1.341653}$ MPa (Computation 2.1): density times wave speed times velocity.
  2. $0.374397$ MPa in PVC (Computation 2.1): a difference of $\mathbf{3.583508}$ times.
  3. sound in water, $1484.576418$ m/s (Computation 2.1): the wave in a pipe is slower.
  4. $1344$ m/s in steel, $198$ m/s in polyethylene (Computation 2.1): the wall decides.
  5. round trip in a $500$-m steel pipe, $0.744008$ s (Computation 3.1): the boundary.
  6. closing faster than the round trip keeps $\mathbf{137.057235}$ m (Computation 3.1): the same as instant.
  7. independent of how the valve closes (Observation 3.2): the core.
  8. closing over $5$ s gives $20.394324$ m (Computation 3.1): $2Lv/(gT)$.
  9. computation and formulas within $10^{-6}$ (Computation 3.1): two routes.
  10. the separator is closing time versus round trip (Observation 4.1): longer for long and soft pipes.

Closing: Stopping flowing water with a valve sends a pressure wave back and forth along the pipe. The pressure rise is the product of density, wave speed and the velocity stopped——stopping $1$ m/s gives $1.341653$ MPa in steel and $0.374397$ MPa in PVC, because the wave speed is set not by the sound speed in water but by the softness of the pipe wall. If the valve closes faster than the round trip along the pipe, the pressure does not depend on how it closes——closing faster than the $0.744008$-s round trip of a $500$-m steel pipe, the method-of-characteristics computation gave the same head as an instant closure, $137.057235$ m; closing more slowly, it fell to $2Lv/(gT)$, $20.394324$ m for $5$ s. The separator is whether the closing time or the round-trip time is longer——during the round trip, no word from the reservoir reaches the valve. Placed among the earlier papers——where in Paper 474 the speed of sound changed with the air, here the pipe wall around the wave changes its speed. To be honest——neither friction, bubbles nor valve characteristics are included, and the properties of water are rounded.

References

  1. Joukowsky, N. (1900). Über den hydraulischen Stoss in Wasserleitungsröhren. Mémoires de l’Académie Impériale des Sciences de St.-Pétersbourg.
  2. Korteweg, D. J. (1878). Über die Fortpflanzungsgeschwindigkeit des Schalles in elastischen Röhren. Annalen der Physik und Chemie.
  3. Wylie, E. B., & Streeter, V. L. (1993). Fluid Transients in Systems. Prentice Hall.
  4. Yamagishi, Y. (2026). On Nights When Distant Sounds Carry, the Sky Is Warm [Paper 474]. Zenodo. 10.5281/zenodo.22733583

Reference

Yamagishi, Y. (2026). If a Valve Closes Faster Than a Round Trip Along the Pipe, the Water-Hammer Pressure Does Not Depend on How It Closes [Paper 617]. YOI: https://www.yamagishiyuuki.com/yoi/617
DOI: 10.5281/zenodo.23014494
ORCID: 0009-0006-8587-3450

PDF (English) · PDF (Japanese) · Date proof of the English PDF (OpenTimestamps .ots) · SHA256SUMS

The record page for this paper (in Japanese, with the PDFs in each language and the date proof): YOI 617. This paper is a preprint and has not been peer reviewed.